Climate Variability, Modeling Tools and Agricultural Decision-Making [1 ed.] 9781608767915, 9781606927038

Citation preview

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Climate Change and its Causes, Effects and Prediction

CLIMATE VARIABILITY, MODELING TOOLS AND AGRICULTURAL DECISION-MAKING

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

CLIMATE CHANGE AND ITS CAUSES, EFFECTS AND PREDICTION Global Climate Change Revisited Harace B. Karling (Editor) 2007. 1-59454-039-X Climate Change Research Progress Lawrence N. Peretz (Editor) 2008. 1-60021-998-5 Climate Change: Financial Risks United States Government Accountability Office 2008. 978-1-60456-488-4 Post-Kyoto: Designing the Next International Climate Change Protocol Matthew Clarke 2008. 978-1-60456-840-0

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Economics of Policy Options to Address Climate Change Gregory N. Bartos 2009. 978-1-60692-116-6

Climate Change and its Causes, Effects and Prediction

CLIMATE VARIABILITY, MODELING TOOLS AND AGRICULTURAL DECISION-MAKING

ANGEL UTSET

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

EDITOR

Nova Science Publishers, Inc. New York

Copyright © 2009 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Suasteguí, Angel Utset. Climate variability, modeling tools and agricultural decision-making / Angel Utset. p. cm. Includes index. ISBN 978-1-60876-791-5 (E-Book) 1. Climatic changes--Risk management--Europe. 2. Agriculture--Europe--Decision making. 3. Europe--Climate. I. Title. S600.7.C54S83 2009 338.1'4--dc22 2009000131

Published by Nova Science Publishers, Inc.    New York

CONTENTS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Preface

ix

Part I

Policies and Tools

1

Chapter 1

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making Vesselin Alexandrov

3

Chapter 2

Droughts as a Climatic Variability: Actions to Reduce Drought Impacts José Roldán and Enrique Cabrera

23

Chapter 3

Understanding the Interactions between Agricultural Production and Climate Variability C. Rodríguez-Puebla and S. M. Ayuso

35

Chapter 4

ADAGIO – Adaptation of Agriculture in European Regions at Environmental Risk under Climate Change Gerhard Kubu and Josef Eitzinger

41

Chapter 5

WSSTP Vision on the European Agriculture and Water Issues as the Response to Climate-Change Challenge Marek Nawalany

47

Chapter 6

Land-Air Parameterisation Scheme (LAPS): A Tool for Use in Agrometeorological Modelling D. T. Mihailovic and B. Lalic

53

Chapter 7

The ENSEMBLES Climate Change Project Paul van der Linden

Chapter 8

Downscaled Climate Change Scenarios for Spain Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

Chapter 9

Using Crop Modelling as Support for Agricultural Decision-Making under Variable Climate Conditions Josef Eitzinger

65 107

119

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

vi

Contents

Chapter 10

Modeling Water Management Strategies Using the SWAP/WOFOST Model Joop G. Kroes

Chapter 11

Use of DSSAT Models for Climate Change Impact Assessment: Calibration and Validation of CERES-Wheat and CERES-Maize in Spain Ana Iglesias

127

137

Part II

Applying Climate and Crop-Growth Modelling Tools to Support Agricultural Decision-Making

Chapter 12

Opportunities and Challenges of Using Modelling Tools for Agricultural Decision-Making under Climate Change Conditions Angel Utset

Chapter 13

Modelling Climate Change Impacts on Crop Growth and Management in Germany K. C. Kersebaum, W. Mirschel, K. O. Wenkel, R. Manderscheid, H. J. Weigel and C. Nendel

Chapter 14

Modelling of Maize Production and the Impact of Climate Change on Maize Yields in Croatia Višnja Vučetić

Chapter 15

Consequences of Climate Change on Irrigation Water Requirements in Southern Spain J. A. Rodríguez Díaz, E. K. Weatherhead, J. W. Knox and E. Camacho

Chapter 16

Adaptations of Irrigated Cropping Systems of Southern Italy as Affected by Climate Change at Field/Farm Scale Domenico Ventrella, Nicola Losavio, Rita Leogrande, Luisa Giglio, Mirko Castellini, Enza Di Giacomo, Angel Utset, Juan Carlos Martinez, Javier Rojo, Blanca del Rio and Dimos Anastasiou

Chapter 17

Climate Change and Agricultural Risk in Hungary Márta Ladányi

Chapter 18

Assessment of the Impact of Climate Change and Adaptation on Potato Production in Egypt Mahmoud Medany

239

Chapter 19

Vineyard Full Irrigation Requirements under Climate Change Scenarios for Ebro Valley, Spain Jordi Marsal and Angel Utset

255

Chapter 20

Climate Variability and Change over the Balkan Peninsula and Related Impacts on Sunflower Stanislava Radeva and Vesselin Alexandrov

267

163 165

183

195

203

213

227

Contents Chapter 21

Optimizing Irrigation Water Management on the Global Change Context in a Spanish Mediterranean Region José Antonio Rodríguez, Ángel Utset, Carmen Navarro, Juan Carlos Martos and Ana Iglesias

Chapter 22

Simulating Plant Growth, Soil-Water and Nitrogen Dynamics in Maize Crop in Brazil Using DSSAT Model Camilo de Lelis Teixeira de Andrade, Ramon Costa Alvarenga, Israel Alexandre Pereira Filho, João Carlos Ferreira Borges and Cristiano Márcio Roque Silva

Chapter 23

Potential Impact of Climate Change on Agricultural Soils Simulated by Roth-C Model Elena Charro and Amelia Moyano

311

Chapter 24

Simulation of Crop Yield Near Hedgerows under Aspects of a Changing Climate: An Austrian Attempt Thomas Gerersdorfer, Josef Eitzinger and Pablo Rischbeck

321

Chapter 25

Introduction of Crop Modelling Tools into Serbian Crop Production: Calibration and Validation of Models B. Lalic, D. T. Mihailovic and M. Malesevic

331

Index

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

vii

283

303

347

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PREFACE Global climate change will lead to shifts in climate behaviour and cause manifold impacts on ecosystems in the next decades. In particular, climate change will have significant effects on agricultural production, which has been considered as the most weather-dependent among all the human activities. Negative impacts on agricultural production could be avoided or reduced significantly by applying appropriate measures using available models, as well as forecasts and warning systems for decision-making. This will secure sustainable agricultural production in the future as well. There are several climate modeling tools currently available. For the long-term (decades) assessments, Global Circulation Models (GCM) compute various scenarios of the future climate behavior which have been considered sounder enough. For short-term, seasonalforecasts according to El Niño – South Oscillation (ENSO) and North Atlantic Oscillation (NAO) behaviors, as well as other sources of climate variability, are also available. These scenarios and forecasts have been downscaled by dynamical and statistical methods to reflect local climatic conditions. Besides, agricultural impact models such as crop simulation models can effectively estimate crop yields, as well as yield risk, under any climate conditions, even though they need to be carefully validated before being used for decision-making. Unfortunately, climate prediction science and forecast product development are advancing independent of their applications. Furthermore, climate-researchers and agriculture-oriented institutions are typically separated at the highest levels of national governments through over the world, which makes difficult the contacts among them. While experts and researchers at high-level centers in Europe and other places (“developers”) have established significant Know-How and produced relevant of the above-cited tools for such climate-impacts studies; practical experts at local agricultural research centers as well as agricultural advisers and extension services (“users”), those who should apply these tools for agricultural decision-making, are often not aware about the available tools or their access to such tools is quite limited due to several reasons, as financial issues or lack of user-friendly design of tools. A connection is needed between the “developers” and “users”, to improve decision making by better implementing this Know-How and model tools. Furthermore, feedback from low end-users to the tool-provider researchers is a prerequisite for improving these tools for their practical use e.g. by providing background information, setting up the actual input data needs, fitting time and spatial scales as required by specific applications and other similar issues.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

x

Preface

In that context the European Union (EU) Sixth Framework Program supported the AGRIDEMA proposal (www.agridema.org) from January 2005 to June 2007. The AGRIDEMA general objective was to establish initial contacts and to conduct primary collaborations between “developers” and potential “users”, basically researchers and experts at agricultural services. Such contacts were achieved through introductory courses on the modeling tools, followed by “Pilot Applications” of such tools on local conditions. The AGRIDEMA applications were addressed to evaluate climate-change and extreme-events risks in agriculture, using the learned climate and crop-growth modeling tools. Furthermore, the applications pointed out the advantages or constraints of the used modeling tools, their needed improvement as well as potential benefits of the obtained results for agricultural decision-making. According to the available Climate-Change impacts assessments, Mediterranean countries could face the highest negative consequences of global warming within Europe, through water-shortage and crop-water requirements increments. Besides, since climatechange and extreme events effects could be more serious in countries with less-developed agriculture, the EU associated countries from Central and Eastern Europe, with relative reduced technological capacities, would be more affected than Northern-European countries. Therefore, AGRIDEMA focused on Southern, Central and Eastern Europe, as well as on the countries of the Mediterranean area. Nevertheless, the AGRIDEMA courses and activities were internationally open. The AGRIDEMA results have been disseminated among “developers” and “users” communities, as well as within politicians and decision makers. A final workshop was held in June 2007, providing the results of the AGRIDEMA applications as well as promoting exchange among “developers”, “users” and decision-makers. Farmer associations, insurance companies and other stakeholders were also invited. This book comprises the AGRIDEMA contributions of important “developers” of climate and crop-growth modeling tools, that can be potentially used to support agricultural decisionmaking under variable climate conditions. Furthermore, the results obtained through the AGRIDEMA applications are included as separate chapters. The book is aimed to point out the usefulness, potentialities and limitations of the currently-available climate and cropgrowth simulation models. Furthermore, the book will comprise several chapters aimed to exemplify the combined application of the climate and crop-growth tools to specific climatechange risks in the local agriculture of about 10 different countries. By reading this book, researchers and farmer advisers from local agricultural and extension services can effectively realize which practical decisions should be taken for mitigating the possible climate risks on their local conditions. These potential “users” of the modeling tools are continuously facing climate risks, but they usually do not know how to deal with them. Applied agricultural researchers are not usually connected to high-level researches and they need being introduced to the available climate and crop-growth modeling tools, though simple descriptions of the tool background and descriptive applications. The book highlights the potential usefulness of such tools, providing several examples of applications in Climate risks of local agriculture in several countries. The book is divided in two parts. The first part summarizes the available climate and crop-growth modeling tools, as well as the policy context regarding climate risks. Due to the importance of drought risks and water management under future conditions, European and Spanish policies regarding this issue are provided as examples. Besides, the EU proposal

Preface

xi

ENSEMBLES, the largest European attempt to develop sounder Climate forecasts, is especially highlighted. The Spanish downscaling effort, aimed to provide regional ClimateChange scenarios for local applications is also a good example of existing tool. Easily understanding details about the two most used crop-growth models, the DSSAT and the water-oriented SWAP models are also provided. The first part comprises also other contributions regarding policies and tools. The second part is devoted to applications of both, climate and crop modeling tools, to support agricultural decision-making under variable climate conditions, at regional and local scales. As well as AGRIDEMA, this book is aimed to promote the practical use of the available modeling tools, to support agricultural decision-making under variable climate conditions. I expect the readers being able to start simple applications of the shown climate and cropgrowth modeling tools on their local conditions. Addresses and emails directions of both “developers” and “users” are provided. I encourage readers to exchange with us.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Valladolid, Spain, 15 July 2008

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PART I. POLICIES AND TOOLS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 1

CURRENT CLIMATE FORECASTING AS A HELPING TOOL FOR AGRICULTURAL DECISION MAKING Vesselin Alexandrov∗ National Institute of Meteorology and Hydrology, Sofia, Bulgaria

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Agriculture is a very weather-dependant sector; approximately 80% of the variability in agricultural production is due to weather variability; climate variability and the severe weather events that are responsible for natural disasters impact the socio-economic development of many nations. Agriculture is an important sector for the economies of many countries; over the next two decades, the world will need 15-20% more water for agriculture. In this study CLIPS surveys are first discussed. In 2003 a comprehensive CLIPS questionnaire was created and disseminated among WMO RA VI member countries. Here a summary of the respective answers from eastern European countries is given. Climate forecasts: needs, advances, case studies, challenges are then discussed. Case studies are considered as well. For example, the Climate Prediction and Agriculture (CLIMAG) project was initiated following the International Workshop on Climate Prediction and Agriculture held in Geneva in September 1998. CLIMAG is based on the awareness of the adverse impact climate variability has on agriculture, and the premise that advances in climate knowledge and prediction capacity at the seasonal time scale can contribute to adaptive management. CLIMAG Workshop 1999 considered a number of important issues relating to climate prediction applications in agriculture. CLIMAG project should be viewed as a partnership of potential users and researchers with multiple stakeholders. Based on an International Workshop held in Geneva in 2005, reviews the advances made so far in seasonal climate predictions and their applications for management and decision-making in agriculture and identifies the challenges to be addressed in the next 5 to 10 years to further enhance operational applications of climate predictions in agriculture. The following challenges are discussed: improving the model accuracy; establishing a better networking between researchers in climate and agriculture; giving greater priority to extension and communication activities; responding to diverse ∗

[email protected]

4

Vesselin Alexandrov user needs: promoting beneficial use of forecasts; involving the stakeholders in climate prediction applications; rating better institutional and policy environment.

1. INTRODUCTION Agriculture is a very weather-dependant sector: • •

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• •

approximately 80% of the variability in agricultural production is due to weather variability climate variability and the severe weather events that are responsible for natural disasters impact the socio-economic development of many nations agriculture is an important sector for the economies of many countries over the next two decades, the world will need 15-20% more water for agriculture

In 2002 Antonio Divino Moura, Director General at the IRI said "Using climate forecasts to better manage climate-sensitive sectors such as agriculture, health, and water resources is a new frontier, with potentially very significant implications for humankind“ And "Skilful climate forecasts, appropriately communicated and used for decision-making, have the potential to help countries cope with the impacts of climate variability, and also to assist them in adapting to long-term climate change, thereby contributing to poverty reduction and sustainable development" According to Meteo-France, in relation with the predictability the statement « a Thunderstorm will be observed next Sunday over the Toulouse « Météopole » between 15h and 16h » sounds unrealistic, the confidence that one can have in this forecast is very low. However, when one says « a rainy system will cross the Toulouse region Sunday afternoon», it is assumed realistic, one can be quite confident in this forecast. It is considered that when discussing any issue related to climate predictions the respective definitions of meteorological forecasting ranges should be clarified (Table 1). According to the recommendations of the ad hoc ICCD/COP(4)/CST/4 panel on early warning systems (October 2000) it is necessary: to integrate early warning results with the results of other climate prediction systems such as the World Meteorological Organization (WMO).Climate Information and Prediction Services (CLIPS) and CLIVAR (CLImate VARiability and predictability); to encourage the further development and application of seasonal climate forecasting and long-range forecasting as tools for early warning systems; etc. The specific objectives of CLIVAR (www.clivar.org) are: a) to describe and understand the physical processes responsible for climate variability and predictability on seasonal, interannual, decadal, and centennial time-scales, through the collection and analysis of observations and the development and application of models of the coupled climate system, in cooperation with other relevant climate-research and observing programmes; b) to extend the record of climate variability over the time-scales of interest through the assembly of quality-controlled paleoclimatic and instrumental data sets; c) to extend the range and accuracy of seasonal to interannual climate prediction through the development of global coupled predictive models; d) to understand and predict the response of the climate system to

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

5

increases of radiatively active gases and aerosols and to compare these predictions to the observed climate record in order to detect the anthropogenic modification of the natural climate signal. Table 1. Definitions of meteorological forecasting ranges (source: www.wmo.int) 1.

Nowcasting

A description of current weather parameters and 0 -2 hours description of forecasted weather parameters

2.

Very short-range weather forecasting

Up to 12 hours description of weather parameters

3.

Short-range weather forecasting

Beyond 12 hours and up to 72 hours description of weather parameters

4.

Medium-range weather forecasting

Beyond 72 hours and up to 240 hours description of weather parameters

5.

Extended-range weather forecasting

Beyond 10 days and up to 30 days description of weather parameters, usually averaged and expressed as a departure from climate values for that period.

6.

Long-range forecasting*

From 30 days up to two years

6.1

Monthly outlook

Description of averaged weather parameters expressed as a departure (deviation, variation, anomaly) from climate values for that month (not necessarily the coming month).

6.2

Three month or 90 day outlook

Description of averaged weather parameters expressed as a departure from climate values for that 90 day period (not necessarily the coming 90 day period).

6.3

Seasonal outlook

Description of averaged weather parameters expressed as a departure from climate values for that season.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Notes:

*

(1)

In some countries, long-range forecasts are considered to be climate products

(2)

Season has been loosely defined as Dec/Jan/Feb = Winter; Mar/Apr/May = Spring; etc...in the northern hemisphere. In the tropical areas seasons may have different durations. Outlooks spanning several months such as multiseasonal outlooks or tropical rainy season outlooks may be provided.

7.

Climate forecasting

Beyond two years

7.1

Climate variability prediction

Description of the expected climate parameters associated with the variation of inter-annual, decadal and multi-decadal climate anomalies.

7.2

Climate prediction

Description of expected future climate including the effects of both natural and human influences.

In the paper assumed as Seasonal to Interannual Prediction (SIP).

6

Vesselin Alexandrov

2. CLIPS SURVEYS CLIPS (www.wmo.ch/web/wcp/clips2001/html/) is a project of the World Meteorological Organization that deals with the implementation of climate services around the globe. Climate services are any activity that employs and/or applies climate knowledge, climate information and climate predictions to the benefit of individuals, organizations and countries. Increasingly governments, international organizations, companies and individuals are recognizing the impacts that climate has on their activities, whether from long-term climate change or from climate variability on time scales of up to a few seasons or years. A survey undertaken during the year 2000 on behalf of the CLIPS project (Kimura 2001) revealed that about one-third of the WMO member countries already had, or planned to obtain in the near future, the capability to provide some form of operational seasonal to interannual prediction (SIP). Most of the member countries do not have the necessary human and financial resources to develop and issue their own predictions (Kimura 2001, Sivakumar 2006). According to Kimura (2001) the total issuance of official climate forecasts in the WMO RA VI (Europe) was 27, which is higher than the totals of the other regional associations (Table 2). Generally, one-third of the WMO Members already had, or planned to obtain in the near future, the capability to provide some form of operational seasonal to interannual prediction (Kimura, 2001): • • • • •

Most models in use predict only for single countries Rainfall is the most popular predictand, Usually the forecasts are for a single three-month season (or a part of this period) at zero lead Vast majority of cases use empirical models

In 2003 a comprehensive CLIPS questionnaire was created by Gocheva and Heckler (2004) and disseminated among WMO RA VI member countries. Here a summary of the respective answers from eastern European countries is given.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. Issuance of official climate forecasts (Kimura 2001)

Yes (class A) Planned (Class B) No (class C) Total

RA I 15 7 0 22

RA II 9 2 6 17

RA III 4 0 0 4

RA IV 4 0 0 4

RA V 7 0 1 8

RA VI 10 5 12 27

Global 49 14 19 82

There is a wide range of answers to the question “Is SIP currently successful in specified regions and sectors only?” Countries such as Albania and Cyprus do not use SIP and have not any precise opinion about SIP. However, according Armenia, Moldova and Kazakhstan SIP is successful in wide geographical regions. Azerbaijan: it is difficult to say anything about successfulness of SIP; Latvia: it is difficult to point out any geographic region where SIP works better; Bulgaria; Estonia, Slovenia and Cyprus: SIP seems successful for specific

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

7

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

regions and sectors; Croatia, Poland, Romania: successful in ENSO-related regions with insignificant predictability (NAO) in mid-latitudes (Gocheva and Heckler 2004) In respect to the question “Does your NMHS (National Meteorological and Hydrological Service) provide official SIP?” the following countries replied negatively: Albania, Croatia, Cyprus, Estonia, Greece, Lithuania, Slovenia. Bulgaria, Latvia, Serbia and Montenegro provide monthly SIP; Belarus, Armenia, Azerbaijan, Poland - monthly and seasonal SIP; Russia -operational 1-3 month SIP containing both regional and global predictions. It is necessary to point out the Romanian information. The National Meteorological Administration in Romania beyond the one-month weather forecasts, provides: prognostic estimates for the next two months, following the forecasting month; “seasonal supplement”, containing the anomaly notification in the geophysical environment in past season and meteorological outlook for the next season; a bulletin with annual forecasting assessments, elaborated at the beginning of each season and containing estimates of the temperature and precipitation anomalies for the next four seasons (Gocheva and Heckler 2004). Figure 1 shows the major global producers of long range forecasts. The next question from the survey of Gocheva and Heckler (2004) considers this topic: “Does your NMHS use SIP products from global producers?” The responses were as follow: Croatia, Cyprus, Estonia – such products have been not applied yet; Armenia, Azerbaijan, Belarus, Latvia are using ROSHYDROMET products; Slovakia and Greece explore ECMWF (Figure 2 and 3) products only; Bulgaria uses ECMWF, IRI (International Research Institute for Climate and Society), UK Met Office, Météo-France in the terms of development of monthly weather forecast involving local weather and climate archive data downscaling.

Figure 1. Global producers of long range forecasts (source: www.wmo.int).

For example, the National Institute of Meteorology and Hydrology in Sofia, Bulgaria shows on its web page (info.meteo.bg) UK Met Office and IRI products as well as their interpretation for the Balkan peninsula and especially Bulgaria (Figure 4); Lithuania – products from IRI (Figures 5 and 6), World Resource Institute and Swedish Regional Climate Modelling Programme; Poland - ECMWF, IRI, DWD (The German Weather Service); Romania - ECMWF, Met Office, IRI and Japan Meteorological Agency.The agricultural applications of the climate outlooks (Figure. 6) are as follow:

8

Vesselin Alexandrov • • •

Climate Outlooks provide an opportunity to plan mitigation measures before the season begins Climate information for the coming months or season make it possible for planners to more effectively deal with their climate-related issues through improved practices Medium to long-term agro-meteorological information is useful for: o Prevention of damage due to weather and climate conditions on harvest o Market regulation o Building national safety stock o Plan fodder o Mitigation of food crises o Food aid projections

The following critical points, however, need to be taken into account:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• •

Climate prediction is global, but agricultural applications are considerably local The science of climate prediction is relatively new, but farmer’s traditions persist for a long time – sometimes it is difficult to change the farmer’s behaviour

Figure 2. Spatial pattern of correlation between modelled February-April snow cover and NCEP/NESDIS observations; a) shows the correlation for the GloSea model ;b) shows the correlation for the ECMWF S2 model (source: Shongwe et al. 2006).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

9

Figure 3. ECMW products for Europe (source: www.ecmwf.int)

It is considered that the ability to predict climate fluctuations months in advance is improving to the point where there are good prospects for using such forecasts to modify, for example, the management of crops and livestock so as to ameliorate some of the negative impacts of climate variability in some environments (Mason 2001, Hansen 2002, O’Brien and Vogel 2003, Thornton, 2006). Gocheva and Heckler (2004) collected also the answers of the following question: “Do you apply SIP in the management of agricultural production, water resources, etc.?” Countries such as Armenia, Azerbaijan, Belarus, Bulgaria, Kazakhstan, Latvia, Poland, Romania have relatively widely SIP applications in the different sectors of

10

Vesselin Alexandrov

economy (e.g. energy, agriculture, insurance, transport, water resources, tourism, human health, etc.).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. The web page of the Information Center at the National Institute of Meteorology and Hydrology in Sofia, Bulgaria, where links to UK Met Office and IRI information on SIP are posted (source: info.meteo.bg).

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. IRI products for Europe (source: iri.columbia.edu).

Figure 6. IRI climate outlook for Europe, November 2006 – April 2007 (source: iri.columbia.edu).

11

12

Vesselin Alexandrov

Russia, Croatia, Serbia and Montenegro, Slovakia apply SIP partially in some sectors, occasionally. Albania, Cyprus, Greece, Lithuania and Slovenia answered completely negatively. The eastern European countries were spitted to 50:50 in respect to the question “Has your NMHS contracts for regular SIP provision with a specific sector for example, agriculture?” However, 90% of them confirmed availability of user’s requests towards SIP products. There were additional questions within the survey of Gocheva and Heckler (2004), such as: “Is your SIP officially issued by media?”; “Do you develop the theoretical basis of your SIP activities by own research efforts?”; “How do you maintain the theoretical basis of your operational SIP activities?”; “Do you apply downscaling methods for specific sectors/applications/locations?”; “What are the predicted meteorological elements and parameters in your national SIP practice?”, etc. It is necessary to point out that the above summary could be slightly different from the current situations because of some positive changes in the terms of SIP applications during the last few years.

3. CLIMATE FORECASTS: NEEDS, ADVANCES, CASE STUDIES, CHALLENGES 3.1. Needs • • • •

improved information on weather and climate could make the agricultural sector more productive it is important to integrate the issues of climate variability into resource use and development decisions the improved knowledge on choice of policies, practices and technologies will decrease agricultural vulnerability to climate variability advantage should be taken of current data bases, increasing climate knowledge and improved prediction capabilities

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.2. Advances •

•

will come from a better understanding and simulation of teleconnections involving the other ocean basins and from the inclusion of land surface conditions in climate prediction models North Atlantic Oscillation (NAO) also exhibit variability on the seasonal-tointerannual time frame that might represent a potential source of predictability

The principal scientific basis of seasonal forecasting is founded on the premise (e.g. Palmer and Anderson 1994) that lower-boundary forcing, which evolves on a slower timescale than that of the weather systems themselves, can give rise to significant predictability of atmospheric developments. These boundary conditions include sea surface temperature (SST), sea-ice cover and temperature, land-surface temperature and albedo, soil moisture and snow cover, although they are not all believed to be generally of equal

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

13

importance (Sivakumar, 2006). Seasonal climate forecasts are based also on the interactions between ocean and atmosphere as manifested in sea surface temperatures, which can offer some predictability in terms of future temperatures and rainfall amounts (Ziervogel et al. 2005). According to Hansen (2006) forecasts of climate fluctuations with a seasonal lead-time are possible because the atmosphere responds to the more slowly varying ocean and land surfaces, an example being climate fluctuations associated with the El Niño-Southern Oscillation (ENSO) in the tropical Pacific (Mason 2001). Regional forecasts for Europe are dependent on North Atlantic Oscillation events, as well (Figure 7). • •

empirical-statistical methods (e.g. discriminant and canonical correlation analyses; optimal climate normals; analysis of climatic anomalies associated with ENSO) dynamical methods - model-based, using atmospheric GCMs; coupled Atmosphereocean GCMs; and two-tiered models

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Seasonal forecasts can be developed using mathematical models of the climate system. Such dynamical seasonal forecasts are an extension of the numerical methods used to predict the weather a few days ahead. Dynamical models represent the climate system by a set of computer-solved equations, to predict its evolution several months in advance. Several climate prediction centers routinely issue probabilistic seasonal forecasts based on dynamic general circulation models (GCMs) that model the physical processes and dynamic interactions of the global climate system in response to sea and land surface boundary forcing. Probabilistic forecasts are obtained from ensembles of GCM integrations initialized with different atmospheric conditions. In addition to dynamical predictions, empirical seasonal forecasts (Moura and Hastenrath 2004) can also be used in an attempt to find statistical links between current observations and general weather conditions some time in the future.

Figure 7. Statistical forecast for the NAO index (www.metoffice.gov.uk).

14

Vesselin Alexandrov •

Regional Climate Models (RCMs) with far more resolution than is possible using present global models, and that use boundary conditions supplied by a pre-run of a global model

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

A recent trend is to examine the potential use of regional climate models. A sophisticated method of obtaining more localised estimates of climate is to apply numerical regional climate models at high resolution over the region of interest. Regional models have been used in several climate impact studies for many regions of the world, including parts of North America, Asia, Europe, Australia and Southern Africa (Giorgi and Mearns 1999, Mearns et al. 1997). The regional climate models obtain sub-grid scale estimates (sometimes down to 25 km resolution) and are able to account for important local forcing factors, such as surface type and elevation. Particularly, the regional climate model RegCM (Figure 8) was originally developed at the National Center for Atmospheric Research (NCAR), USA and has been mostly applied to studies of regional climate and seasonal predictability around the world. It is further developed by the Physics of Weather and Climate group at the Abdus Salam International Centre for Theoretical Physics (ICTP) in Trieste, Italy.

Figure 8. RegCM3 inputs (source: Pal et al. 2005).

• •

use of multiple models, each running their own ensemble from varying initial conditions, provides an improvement in skill not available from a single model alone multiple model systems have been examined in the USA under the Dynamic Seasonal Prediction projects, internationally under Seasonal forecast Model Intercomparison Project

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making • •

15

the Asia-Pacific Climate Network (APCN) based in, South Korea also uses multiple model inputs in Europe, under the DEMETER (Development of a European Multimodel Ensemble system for seasonal to inTERannual prediction) project, plans are being drawn for an operational system using multiple coupled models

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In Europe the multi-model approach seems to be the most adequate to produce reliable probabilistic climate forecasts (Doblas-Reyes 2006). The advantages of the multi-model approach have been illustrated in, among other research efforts, the DEMETER project. Briefly, the DEMETER system comprises 7 global coupled ocean–atmosphere models. Uncertainties in the initial state were represented through an ensemble of 9 different ocean initial conditions. Atmospheric and land-surface initial conditions are taken directly from the ERA-40 (ECMWF Re-analysis) atmospheric re-analysis (Uppala et al. 2005). The performance of the DEMETER system has been evaluated from a comprehensive set of predictions for past cases, or hindcasts, over a substantial part of the ERA-40 period (1958– 2001). One of the main results of the experiment is that the DEMETER multi-model forecast system provides, on average, more skilful seasonal forecasts than is possible using a singlemodel ensemble system (Doblas-Reyes 2006) The relative merits of the above methods (Figure 9) which are under evaluation within EU-funded project such as ENSEMBLES, EUROCLIM, CECILIA, etc. The ENSEMBLES objectives (www.ensembles-eu.org) are to: run ensembles of different climate models to sample uncertainties; measure variations in reliability between models; produce probabilistic predictions of climate change.

Figure 9. Regional Climate Change Index, based on 20 models and 3 emission scenarios (source: Giorgi 2006).

16

Vesselin Alexandrov

The project will, for the first time, develop a common ensemble climate forecast system for use across a range of time (seasonal, decadal and longer) and spatial scales (global, regional, and local); link these projections to potential impacts: agriculture, health, energy, insurance, ecosystems, etc. One of the EUROCLIM project (euroclim.nr.no) targets is to improve climate models in order to better predict future climate conditions. The main CECILIA project (www.cecilia-eu.org) objectives are as following: producing high resolution (10 km) 30-year time slices over four target areas; comparing model responses with coarser results from existing simulations to assess the gain of a higher resolution; archiving daily data from the simulations in a common database; improving high resolution models for future scenarios (Figure 10).

Figure 10. CECILIA project web page (source: www.cecilia-eu.org).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• •

interpretation and delivery of the climate prediction information promoted through the development of regional climate outlook forums forecasts are now freely transmitted around the globe by the internet

3.3. Case Studies According to Hansen (2002) studies indicated that, while it is still too early to be entirely specific about the potential value of climate predictions for agriculture, there is a reason to be optimistic concerning future opportunities realised by further research. The Climate Prediction and Agriculture (CLIMAG) project was initiated following the International Workshop on Climate Prediction and Agriculture held in Geneva in September

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

17

1998. CLIMAG is based on the awareness of the adverse impact climate variability has on agriculture, and the premise that advances in climate knowledge and prediction capacity at the seasonal time scale can contribute to adaptive management CLIMAG Workshop 1999 (Figure 11) considered a number of important issues relating to climate prediction applications in agriculture. CLIMAG project should be viewed as a partnership of potential users and researchers with multiple stakeholders. Based on an International Workshop held in Geneva in 2005, reviews the advances made so far in seasonal climate predictions and their applications for management and decisionmaking in agriculture and identifies the challenges to be addressed in the next 5 to 10 years to further enhance operational applications of climate predictions in agriculture. The major points of the UK Government Seasonal Weather Forecasting for the Food Chain project are:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• • • •

Field vegetables - most benefit in the frozen produce Sugar beet - scheduling aspects, agrochemical use Tomato - integration of forecasts on a range of scales Optimal benefit through co-ordinated actions throughout the food chain rather than through independent decisions

Figure 11. CLIMAG project books including many case studies across the world.

Some results of cases studies carried out in the Czech Republic and Bulgaria are presented in Figures 12 and 13.

3.4. Challenges • •

improving the model accuracy establishing a better networking between researchers in climate and agriculture

18

Vesselin Alexandrov • • • • •

giving greater priority to extension and communication activities responding to diverse user needs promoting beneficial use of forecasts involving the stakeholders in climate prediction applications creating better institutional and policy environment

DSS PERUN (CZ): probabilistic seasonal crop yield forecasting

Construction of weather series (Trnka et al., 2005)

DSS PERUN (CZ): probabilistic seasonal crop yield forecasting

Running the WOFOST model (Trnka et al., 2005)

DSS PERUN (CZ): probabilistic seasonal crop yield forecasting

Predicted crop yield in Czech Republic (Trnka et al., 2005)

Figure 12. A case study in the Czech Republic.

b) +1/

+6/-4 -4/

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

+11/

-4/ 0/+4

-1/-6

+7/+9

-1/ +4/

+15/+4 -5/

-4/ +1/-4

+17/

+4 0/0

+5/ -1/ +4/-1

-1/

/-2

+2/+11

-3/+1 -9/-14

-2/-2

+3/-1

-6/-5

+12/ 0/+12

Deviation (in days) between the forecasted and observed flowering/maturity dates in Bulgaria, 1995 (Slavov et al., 1998) 1998)

Figure 13. A case study in Bulgaria.

Forecast (in kg/dka kg/dka;; 1dka = 1.101.10-1 ha) of the expected grain yield of winter wheat in Bulgaria, 1995 (Slavov et al., 1998) 1998)

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

19

4. CONCLUDING REMARKS Climate variability and change contribute to the vulnerability of individuals, businesses, communities and regions. This influences decision makers at all levels (policy, businesses and farms), regardless of the level of economic development. When anticipating potential changes in climate such as possible changes in the frequency and/or magnitude of extreme events and other changes in the pattern of climate and system, there is a need for improved seasonal forecasts. It is considered that in the future there would be better characterization of predictability at finer spatial and temporal scales and perhaps challenge the convention of presenting operational forecasts only as seasonal climatic means at an aggregate spatial scale. If improved forecasting is to be harnessed effectively, however, various conditions will need to be met. Seasonal climate forecasts must address a real and perceived need, and they must have value (Hansen 2002). There are also clear needs for appropriate institutional structures that are adapted to manage such processes (Hudson and Vogel 2003, O’Brien and Vogel 2003). According to Hansen et al (2006) seasonal forecasts can be calibrated and evaluated at a local scale, although attempts to quantify the effect on prediction skill have so far been few. Incorporating understanding of fine-scale climatic influences - such as orography, land–water interfaces, or land cover - into either statistical downscaling models or highresolution, regional dynamic climate modeling is likely to further enhance prediction skill at the local scale that is relevant to farm impacts and decisions. Although it is impossible to predict the timing of daily weather events through a season, it is reasonable to assume that the large-scale ocean–atmosphere interactions that give rise to predicable shifts in seasonal means may also influence higher-order statistics of synoptic weather events that are important to agriculture, such as the frequency and persistence of rainfall events, the distribution of dry spell durations, the timing of season onset and the probabilities of intense rainfall events or temperature extremes. For now, the predictability of these higher-order statistics at a seasonal lead-time remains largely unquantified. The following important elements should be emphasized:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• • • •

• •

advances have been done in the last years in developing understanding of climate prediction in relation to agriculture need to further refine and promote the adoption of current climate prediction tools improved climate prediction techniques are growing faster and finding more applications close contacts between climate forecasters, agrometeorologists, agricultural research and extension agencies in developing appropriate products for the user community are needed crop models can be an important tool for decision support making in agriculture bringing science to society – feedbacks from the end user are essential identifying the opportunities for agricultural applications

20

Vesselin Alexandrov

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Doblas-Reyes FJ, Hagedorn R, Palmer TN (2006) Developments in dynamical seasonal forecasting relevant to agricultural management. Climate Research 33: 19-26. Giorgi F (2006) Climate change hot-spots. Geophysical Research Letters 33. Giorgi F, Mearns LO (1999) Introduction to special section: Regional climate modeling revisited. Journal of Geophysical Research, 14(D6): 6335-6352. Gocheva A, Heckler P (2004) Questionnaire on CLIPS activities in the NMHSs of RA VI countries (www.wmo.int). Hansen JW (2002) Realising the potential benefits of climate prediction to agriculture: issues, approaches, challenges. Agric. Syst. 74:309–330. Hansen JW, Challinor A, Ines A, Wheeler T, Moron V (2006) Translating climate forecasts into agricultural terms: advances and challenges. Climate Research 33:27-41. Hudson J, Vogel C (2003) The use of seasonal forecasts by livestock farmers in South Africa. In: O’Brien K, Vogel C (eds) Coping with climate variability: the use of seasonal climate forecasts in southern Africa. Ashgate, Aldershot. Kimura Y (2001) A survey on the present status of climate forecasting Thirteenth Session, 21–30 November 2001, World Meteorological Organisation Commission for Climatology, Geneva. Mason S (2001) El Niño, climate change, and Southern African climate. Environmetrics 12:327–345. Mearns LO, Rosenzweig C, Goldberg R (1997) Mean and variance change in climate scenarios: methods, agricultural applications and measures of uncertainty. Climatic Change 35: 367-396. Moura AD, Hastenrath S (2004) Climate prediction for Brazil’s Nordeste: performance of empirical and numerical modelling methods. J. Clim. 17:2667–2672. O’Brien K, Vogel C (2003) A future for forecasts? In: O’Brien K, Vogel C (eds), Coping with climate variability: the use of seasonal climate forecasts in Southern Africa. Ashgate Press, Aldershot. Pal JS, Giorgi F, Bi X, Elguindi N, Solmon F, Gao X (2005) Regional Climate Modeling and the ICTP Regional Climate Model (RegCM3). Presentation at the AGRIDEMA workshop, Vienna, Austria. Palmer TN, Anderson DLT (1994) The prospects for seasonal forecasting: a review paper. Q J. R. Meteorol. Soc. 120: 755–793. Thornton PK (2006) Ex ante impact assessment and seasonal climate forecasts: status and issues. Climate Research 33: 55-65. Shongwe ME, Ferro CAT, Coelho CAS, van Oldenborgh GJ (2006) Predictability of cold spring seasons in Europe. Journal of Climate (submitted). Sivakumar MVK (2006) Climate prediction and agriculture: current status and future challenges. Climate Research 33: 3-17. United Nations Convention to Combat Desertification (UNCCD) (2006) United Nations Convention to Combat Desertification in those countries experiencing serious drought and/or desertification, particularly in Africa: text with annexes. UNCCD Secretariat, Bonn, Germany.

Current Climate Forecasting as a Helping Tool for Agricultural Decision Making

21

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Uppala S, Kållberg PW, Simmons AJ, Andrae U and 42 others (2005) The ERA-40 reanalysis. Q J R Meteorol. Soc. 131: 2961–3012. Ziervogel G, Bithell M, Washington R, Downing T (2005) Agent-based social simulation: a method for assessing the impact of seasonal climate forecast applications among smallholder farmers. Agric. Syst. 83:1–26.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 2

DROUGHTS AS A CLIMATIC VARIABILITY: ACTIONS TO REDUCE DROUGHT IMPACTS José Roldán∗1 and Enrique Cabrera2 1

Hydraulic Enginnering, University of Cordoba, Cordoba, Spain 2 Technological Institute of Water, Polytechnic University of Valencia, Valencia, Spain

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT In this paper, the authors will discuss in greater depth aspects such as the development of drought impact matrixes, drought impact assessment and strategies designed to mitigate drought. There is a wide range of actions aimed at minimizing the impact of drought. Proactive measures are understood as any action that either contributes to increasing resources (for example, reutilization) or permits water use to be reduced (for example by improving water network performance). On the other hand, reactive measures are measures which are implemented to reduce impact when in the throes of a drought To reduce the impact of water shortage it is useful to previously develop an impact matrix that classifies the consequences of drought and allows decisions to be made in a reasonable manner. In the reactive measures group, we must refer to the decision tree diagram that sets out how to manage a drought according to its evolution based on the established protocol and taking into consideration risk matrixes. It is common to group impacts according to a variety of criteria (economic, environmental and social) that should be envisaged in all sustainable water management plans. Finally, there are other actions that do not directly affect the water balance, and they should not be considered direct measures. Instead, these are indirect measures that serve to complement direct actions as their ultimate objective is none other than to smooth the way for the latter. As a conclusion, drought cannot be managed efficiently without a plan that has been properly developed beforehand. The final goal is to ensure that planning, rather than improvisation, prevails in the event of a drought.

∗

[email protected]

24

José Roldán and Enrique Cabrera

1. INTRODUCTION The Ministry of the Environment deemed it convenient to promote the creation in September 2005 of an Expert Committee. The ultimate goal of this initiative was to produce a document that would serve as a useful reference tool in the near future and that could contribute, in some way, to adapting Spanish water policy to the present circumstances. Yet given the weight of our history, this is a complex task. It was finally decided that the work commissioned by the Ministry should comprise a series of proposals aimed at tackling future droughts in the best manner possible. Eighteen months later, the work undertaken by the Expert Committee came to light in a document titled La Sequía en España. Directrices para Minimizar su Impacto (Drought in Spain. Guidelines for Minimizing Impact) (http://www.forosequia.com/). The document is comprised of 12 papers signed individually by members of the Committee and 28 conclusions agreed upon unanimously by all of them (Cabrera and Babiano, 2007). In this paper, the authors, all of whom form part of the Expert Committee, will discuss in greater depth aspects such as the development of drought impact matrixes, drought impact assessment and strategies designed to mitigate drought.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2. OBJECTIVES In the opinion of the National Drought Policy Commission (NDPC, 2000), all droughts raise, namely, “How much damage is inflicted by drought?”, “Who is most affected by drought?” and “Where is damage inflicted most?” With this aim, the present document is structured in such a way as to bring to light certain initiatives whose results have provided an adequate answer to these three questions. The first question can be answered from impact matrixes that have been developed to facilitate decision-making process during periods of drought (WDCC, 1998; Rossi et al., 2005). The answer to the second question can be found in the Drought Management Plans that have been developed for this purpose (Wilhite et al., 2005), while the third can be answered from our knowledge about available water resources, the uses that must be satisfied, the possibilities for saving water, and finally, the flexibility of the system. However, given that it is not in our interest to present the full range of possible actions in a disorderly manner, let us first structure them accordingly. This task is especially important as the literature dealing with the problem that concerns us here is so vast, in particular that of the United States. Indeed, a quick look at the website of the country’s main point of reference, the National Drought Mitigation Center (http://www.drought.unl.edu/), serves to confirm the substantial interest that strategies aimed at rational drought planning and management have raised. There are also numerous publications dedicated to drought management that have been produced by professional associations (AWWA, 2002) which place great importance on the presentation of practical cases. A vast amount of information can also be found in Europe, albeit far from the enormous number of documents produced in the United States. The majority of European documents (Cabrera and García-Serra, 1999; EA, 2003; Cubillo and Ibáñez, 2003) deal with drought management within the urban framework. Some of the most prominent studies include an

Droughts as a Climatic Variability

25

overview of the work carried out by the International Commission on Irrigation and Drainage (ICID) from 1995 to 1998, which was summarized (DVWK, 1998) by the German National Committee of the ICID. Of all of these documents, the most innovative for the Spanish case, is one in which the impact of droughts is assessed by means of matrixes constructed for the three components that lead to sustainable water policy: social, economic and environmental. Given the restrictive framework of current water policy, this is an action that merits our interest, particularly in the mid and long term. Not only do we require the necessary resources and time to carry an action out, but must also consider if such an action can be applied immediately or not; an essential step towards defining whether actions are direct or indirect. At the same, direct actions can be either of a long term (or proactive) nature or short term (or reactive) nature. Let us first refer to direct long term or proactive actions. These are actions whose results become evident particularly in the mid and long term. Ultimately, they are actions which will permit water policy to be adapted to the current circumstances. Secondly, there are short term or reactive actions which are implemented during a drought and follow the pace of its evolution. To a large degree their efficacy will depend on the proactive actions that have been implemented in advance. Thirdly, there are the so-called supplementary or indirect actions. This type of action is aimed at facilitating the implementation and development of the other groups of measures (proactive or reactive) with which drought is managed in a direct manner. The vision of drought management presented here is broader than that proposed by other authors (Bouvette, 2004) who have classified actions into two large blocks.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. DEVELOPMENTS OF ACTIONS Let us take a look at the figure below (USACE, 1994), which defines drought from a hydrologic viewpoint. While water availability largely depends on the hydrologic year, consumption follows a more uniform pattern. Although the possibilities of increasing water supply are limited in developed countries (including Spain), as the graph shows, the margin of action concerning consumption is much larger and continues to widen. In any case, proactive measures are understood as any action that either contributes to increasing resources (for example, reutilization) or permits water use to be reduced (for example by improving water network performance). Measures of a proactive nature require time and how and to what degree they should be implemented necessitates discussion and negotiation in times of abundance. We should not forget that, in addition to time, all measures require funding and on many occasions, legal reforms. Periods of drought clearly do not favor the proper state of mind needed to introduce the far-reaching proactive measures that Spanish water policy requires. On the other hand, reactive measures are measures which are implemented to reduce impact when in the throes of a drought. In other words, these are strategies which are taken during periods of scarcity, but which should be negotiated by all of the parties involved when the situation is “normal”. To reduce the impact of water shortage it is useful to previously develop an impact matrix that classifies the consequences of drought and allows decisions to be made in a reasonable manner.

26

José Roldán and Enrique Cabrera

One example of an impact matrix for drought management (there are many other proposals) has been developed in detail by the Western Drought Coordination Council (WDCC, 1998):

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Impact

Cost

Equally Distributed?

Growing?

Public Priority?

Equitable Recovery?

Impact Rank

In short, drought impact decision matrixes are little more than assessment tools that account for a variety of criteria (those in the table are some of the most obvious, but not the only ones) concerning the damage that can be inflicted as a result of restrictions. Based on the final ranking in the matrix analysis, it is possible to establish how to ration water in the most convenient way from the viewpoint of the general public (Wilhite et al., 2005). It is common to group impacts according to a variety of criteria (economic, environmental and social) that should be envisaged in all sustainable water management plans. The first of these, albeit not necessarily the most important, are the most evident. Thus, for example, the loss of crops or livestock or the costs incurred when water is restricted are perfectly quantifiable. On the other hand, environmental impacts include, among many other consequences, the loss of animal and plant diversity or the disappearance of wetlands and natural springs. Finally, droughts produce numerous social impacts. An example of this type of impact include the numerous conflicts that arise among users who compete for water in times of drought. A detailed list of such impacts appears in the report titled How to reduce drought risk by the Western Drought Coordination Council (WDCC, 1998). This report also includes tree diagrams to analyze and follow-up on impacts in terms of the different uses. Yet of all these criteria, the economic criterion is the most “measurable” (that which is most quantifiable) and obvious of all. It should come as no surprise, then, that researchers and institutions (Jenkins et al., 2003; USACE, 1994) propose methods to analyze the economic losses deriving from water shortages. There is no question that mitigating the economic impact is an essential part of proper drought management. It is precisely in this line that Spain can and must work; an undertaking that, incidentally, is closely linked to the creation of water banks and markets.

Droughts as a Climatic Variability

27

But the fact that these concepts can run head on into the complex Water Law of 1985whose final text was revised in 2001 (BOE, 2001b) – should escape no one. Article 60 of the text ranks priorities for water use with “irrigation and agricultural uses” falling into second place followed by “industrial uses for the production of electric energy”. Regardless of whether or not the basin’s Hydrologic Plan - and of course a Drought Decree - would permit this ranking to be modified, it is evident that water legislation in Spain needs to be updated. It should be said, however, that the Special Plans for Action in Emergency Situations and Eventual Drought for the various basins have been developed (as to be expected) in accordance with current legislation. It is for this reason that an impact matrix which considers water use from a present-day viewpoint is difficult to envisage at this point in time. In fact, measures are considered reactive precisely because of the fact that they are provisional. But their development, discussion and implementation are conditioned not only by the physical and legal framework which encompasses them, but by the serenity that they demand. When the necessary calm is lacking, it is almost impossible to discuss priorities. When restricting consumption, the ranking will be more efficient the more robust the system. Finally, there is a wide range of actions aimed at minimizing the impact of drought. However, as they do not directly affect the water balance, they should not be considered direct measures. Instead, these are measures that serve to complement direct actions as their ultimate objective is none other than to smooth the way for the latter. Thus, for example, a new economic water policy oriented towards promoting efficiency and therefore giving meaning to a large number of direct actions, indirectly contributes to achieving a balance between resources and uses; an especially difficult balance to reach in times of drought.

4. DIRECT PROACTIVE ACTIONS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The wide array of actions designed to either increase supply capacity or rationalize and control represent a range of possibilities encompassed within what are known as proactive measures. Because these measures are incompatible with improvised actions, they should be implemented in a progressive manner. The principle measures are outlined below according to how they contribute to increasing resources or rationalizing consumption.

4.1. Increasing and Diversifying Supply Most of these actions appear in the Guide for developing special plans of action in emergency situations and eventual drought published by the Ministry of the Environment (MMA, 2005). The most notable of these include: − − − − −

New storage facilities. New surface water extractions, generally through transfers between basins. Centers for the exchange of water rights. Desalination. New groundwater extractions.

28

José Roldán and Enrique Cabrera − − −

−

Recharging aquifers, a simple and efficient way to conjunctively use surface waters and groundwaters that should be highlighted in an explicit manner. Reutilization. Optimize resources through better hydrologic planning and monitoring. Make better use of the enormous possibilities that the conjunctive use of surface waters and groundwaters affords, especially in times of drought. Finally, it is fundamental to manage a system’s resources in preparation for episodes of drought. To do so, it is essential to have an adequate early warning system and drought characterization system as well as using mathematical models that aid in decision-making processes for the management of reservoirs in real time according to the risk of drought. (Andreu et al., 1996; Rossi et al., 2005).

As indicated in the above Guide, many of these actions require an extended period of implementation, large budgets, political negotiation, social acceptation and when appropriate, modifications to the legislation. This is of foremost importance given that these are works which, in addition to affecting hundred of thousands of people, will be carried out over several terms of office from the time they are conceived until they become operational. Indeed, given the recent experiences in Spain which have been the subject of enormous controversy regarding new transfers, in no way can these measures be considered atypical (USACE, 1994). The actions above were ranked beginning with the most difficult (of any kind) to develop and concluding with the easiest to implement. Nevertheless, and as regards their execution, the final decision should, as always, be based on economic, environmental and social criteria. Logically, the work concerning these actions should be oriented towards analyzing the minimum requisites demanded by each of them and from there, assessing their possibilities for development and ultimately, their feasibility. Finally, a comparative analysis of the possibilities afforded by each and every one of the strategies would be of great utility.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.2. Managing Demand This second via, which began to be explored in detail in the developed world some decades ago, has only recently begun to be explored in Spain. Thanks to this new line of study the margin of savings is now quite large. A recent survey (PI, 2003), which serves as a point of reflection, estimates the potential savings in urban and residential water use in California (where for some years programs have been set up to foment efficiency) at a minimum of 33%. Although precaution should be taken when drawing conclusions, the California study demonstrates the high potential for water savings in Spain. While potential savings in urban and residential water use is notable in Spain, it only accounts for a quarter of the demand in the country and in fact, much greater savings can be achieved in irrigation. Indeed, there are clear indications to this effect. When assuming the energy costs for elevation, groundwater irrigation is five times more productive than surface water irrigation in economic terms (Corominas, 2000). In this line, and according to the provisions set out by the Ministry of Agriculture, Fisheries and Foods (MAPA, 2002) in the

Droughts as a Climatic Variability

29

National Irrigation Plan, more than 5000 hm3 of irrigation water can be saved in Spain through programs to improve water consumption in irrigated areas and reduce excess supply. For all of these reasons, a study that quantifies the potential savings in water use in Spain – especially when taking into account the importance given to water savings in irrigation as a result of the implementation of mechanisms to manage demand - would contribute to identifying water deficits in the hydrographic basins with greater precision. Indeed, due to the important role of irrigation, numerous studies have been dedicated to optimizing water use in the countryside (Pereira et al. 2002). This knowledge is essential for managing droughts in a rational manner. Numerous publications (WDCC, 1998; NCDENR, 2003) refer to specific actions that contribute to reducing demand. These are actions to save water which on many occasions (EPA, 2005) need to be adapted to the different uses to which water is put (urban, residential, commercial, agricultural and industrial). They include: − − − −

Measuring all of the uses and resources (including groundwater wells). Improving the performance of water transport systems (canals, irrigation channels, transport and distribution pipelines and even installations inside buildings). Using rainfall (water harvesting). Reutilization of gray waters in homes and industries. Industrial recirculation.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. DIRECT REACTIVE ACTIONS In this second group, we must refer to the decision tree diagram that sets out how to manage a drought according to its evolution based on the established protocol and taking into consideration risk matrixes. These are decision trees or protocols that, in accordance with current legislation (BOE, 2001a) and logic, should take into account at least two levels: that of the hydrographic basins and that of the city. Having reached this point, it is important to highlight that after consulting the vast information that is available on this issue, especially in the United States (drought management plans for thirty different states can be found at http://drought.unl.edu/plan/ stateplans.htm/), we have come to the conclusion that the Drought Management Plans developed in Spain must not only be improved, but follow similar criteria. In Spain, we are all, in some way or another, following in the wake of the Isabel II Canal that has recently updated its plan (Cubillo and Ibáñez, 2003); a plan that is, without a doubt, a first-rate document. But, as I have already said, due to its singular characteristics, the Isabel II Canal is not an example that can be followed by the majority of Spanish cities. This is not the case of England where the Environment Agency (EA, 2003) has recently published the second revision of the drought management plans developed by water companies in the country. Prior to these plans, the EA had provided clear guidelines for drought management. Many other urban contingency plans are available on water company websites, especially those which are publicly owned. This is the case for example of Melbourne Water in Australia (MWC, 2001) or Denver Water in the United States (DW, 2004). The Water Conservation Committee of the AWWA (WCC, 2002) has developed a model drought

30

José Roldán and Enrique Cabrera

management plan. This plan serves as a support tool for the development of drought plans that will facilitate the work of managers in water companies. Considering the above, the excellent and thorough work Managing Water for Drought (USACE, 1994) merits particular attention as it is the result of a thorough study carried out by more than one hundred professionals. The study was conducted in the early nineties and it was the administration’s reaction to the low rainfall recorded in the western United States at the end of the eighties. The report summarizes and discusses all of the factors that must be taken into consideration in the event of a drought. It includes such issues as the makeup of the teams in charge of developing a drought management plan, how to reconcile conflicting interests, methods of measurement that must be taken into account with a view to making objective decisions as well the negative impact of outdated legislation on drought management.

6. SUPPLEMENTARY OR INDIRECT SUPPORT ACTIONS Given that the indirect actions do not directly mitigate the effects of drought - even when they lead to the success of direct actions of both an reactive and proactive kind - we will considered them separately by structuring them into two different blocks according to the type of direct actions that they influence.

6.1. Measures that Support Proactive Actions These measures facilitate the implementation of complex, but necessary direct proactive actions. Among others, these include:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

− − − − − − − − − −

Citizen awareness. The participation of sociologists and communicators that aid in transmitting the message. Media involvement. Foster citizen participation. Adequate economic policies that foment efficiency and flexibility of use. Proactive water policies that monitor both water use and water resources. Adapt the legislation to the current context. Revise historical water rights. Adaptation of the administration. Water management and monitoring needs to be coordinated. Centers for the exchange of water rights. Provide technical assistance to towns and irrigation communities.

The importance of the first four actions in the above list should be underscored. These are actions that are essential to achieving the viability of the other actions. At least these are the conclusions that can be drawn following a brief look at the National Drought Mitigation Center website (http://www.drought.unl.edu/). Indeed, this website provides everything from educational material to specific information for the media. We must bear in mind that

Droughts as a Climatic Variability

31

Mediterranean water culture has very strong roots and the necessary changes can only come about with ample support from society. The remaining actions are those that directly contribute to mitigating the impact of drought. However, it is clear that if we do not prepare the ground beforehand, it will be unfeasible to implement them. It is surprising that for an issue that would initially seem to be a very delicate one in Spain - historical water rights - specific guidelines already exist in other countries (AWWA, 1995) to redirect the situation in a manner coherent with today’s needs.

6.2. Measures to Support Reactive Actions

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

A Drought Management Plan resembles a decision tree diagram which, depending on the circumstances, guides planners in one direction or another. Decision-making is a dynamic process. If the Plan is good, it will be even better if we have the appropriate information to guide us along the right paths within the tree. In short, all of the actions that permit us to remain one step ahead of the problem are included in this section, as well as supplementary actions. Forecasting and characterizing droughts is a research topic that has gone hand in hand with the stochastic analysis of hydrologic series (Yevjevich, 1967; Dracup et al., 1980). Due to the growing impact of water shortages, research continues to be conducted in this line (Salas et al., 2005). However, in order to manage drought in such a way as to mitigate its impact before a drought occurs, and from the viewpoint of needs that must be satisfied, we must assess the water deficit. The balance between availability and need is the origin of other indicators; indicators which activate the various stages of a Drought Management Plan (Fisher and Palmer, 1997). To sum up, when reliable information is available, it is possible to foresee the events and properly manage reactive actions that have been designed in advance for this purpose. The remaining strategies and actions either facilitate the development of drought management plans (conflict management) or promote both the acquisition and dissemination of knowledge. All of these actions would therefore include: − − − − − − −

Meteorological accurate follow-up and adequate data treatment. Meteorological drought indicators. Drought management indicators. Activation thresholds for the different phases of a plan. Conflict resolution. International relations. Technical assistance.

CONCLUSIONS The main conclusion that can be drawn from the above is that a drought cannot be managed efficiently without a plan that has been properly developed beforehand. A plan that takes into account both proactive and reactive measures will permit exceptional measures to

32

José Roldán and Enrique Cabrera

be reduced to a minimum; measures which until now have been applied in a systematic manner (curiously, exceptional is an antonym of systematic) and characterize the actions implemented by water administrations in Spain in times of drought. Thus, the final goal is to ensure that planning, rather than improvisation, prevails in the event of a drought. The economic, social and environmental impacts caused by the increasingly frequent water shortages occurring in the 21st century will continue to worsen. Although it has not been mentioned explicitly in this report, we should not forget the looming threat of climate change, making rational drought management of vital importance to the future. To achieve such an aim, however, we must travel down a very long road that should not demoralize those who walk upon it.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Andreu J., Capilla J. and Sanchis E., 1996. AQUATOOL: A generalized decision support system for water resources planning and operational management. Journal Hydrology, 177:269-291. American Water Works Association, AWWA, 1995. Water rights and allocations for sound resource management. AWWA Mainstream. American Water Works Association, AWWA, 2002. Drought Management Handbook. AWWA Denver, CO, USA. BOE (Official State Bulletin), 2001a. Law 10/2001, of 5 July of the National Hydrologic Plan [BOE no. 161, of 6 July 2001, pp. 24228-24250]. BOE (Official State Bulletin), 2001b. Legislative Royal Decree 1/2001, of 20 July to approve the revised text of the Waters Law [BOE no. 176, of 24-07-2001, pp. 26791-26817]. Bouvette T., 2004. Structural and non structural project components for drought mitigation. Chapter 8 of the Report on Drought and water supply assessment. Colorado Water Conservation Board. State of Colorado. Cabrera E. and Babiano L. (coord.), 2007. La sequía en España. Directrices para Minimizar su Impacto. Comité de Expertos en Sequía del Ministerio de Medio Ambiente, Madrid. http://www.forosequia.com/ Cabrera E. and García-Serra J., editors, 1999. Drought Management Planning in Water Distribution Systems. Kluwer Academic Publishers, Dordrecht, Holland. Corominas J., 2000. El papel económico de las aguas subterráneas en Andalucía. Papeles del Proyecto Aguas Subterráneas. Fundación Marcelino Botín. Cubillo F. and Ibáñez J.C., 2003. Manual de abastecimiento del Canal de Isabel II. Canal de Isabel II. Madrid. Denver Water (DW), 2004. Drought Response Program. http://www.denverwater.org /drought/DRP04_04.html/ Deutscher Verband für Wasserwirtschaft (DVWK), 1998. Cómo establecer una estrategia para la mitigación de sequías. Una guía de la ICID. ATV-DVWK German Association for Water, Waste Water and Waste. Germany. Dracup J.A., Lee K.S., Paulson Jr. E.G., 1980. On the definition of droughts. Water Resources Research, 16(2):297–302.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Droughts as a Climatic Variability

33

Environment Agency (EA), 2003. Review of water company drought plans. Environment Agency, Rio house, Bristol, UK. Environmental Protection Agency (EPA), 2005. Water use efficiency program. Environmental Protection Agency, USA. Fisher S. and Palmer R.N., 1997. Managing water supplies during drought: triggers for operational responses. Water Resources Updated, 3(108):14-31. Jenkins M.W., Lund J.R. and Howitt R.E., 2003. Economic losses for urban water scarcity in California. Journal of the American Water Works Association (AWWA). Melbourne Water Corporation (MWC), 2001. Drought Response Protocol. www.melbournewater.com.au/ Ministerio de Agricultura, Pesca y Alimentación (MAPA), 2002. Plan Nacional de Regadíos. Horizonte 2008. Madrid. Ministerio de Medio Ambiente (MMA), 2005. Guía para la redacción de planes especiales de actuación en situación de alerta y eventual sequía. Ministerio de Medio Ambiente. Madrid. National Drought Policy Commission (NDPC), 2000. Preparing for drought in the 21st century. US Department of Agriculture, USA. NC Department of Environment and Natural Resources (NCDENR), 2003. Water shortage response planning handbook. NC Department of Environment and Natural Resources, Raleigh NC, USA. Pacific Institute (PI), 2003. Waste not, want not. The potential for urban water conservation in California. Pacific Institute, Oakland, California. Pereira L.S., Cordery I. and Iacovides I., 2002. Coping with water scarcity. Technical Documents in Hydrology nº 58. UNESCO. Paris. Rossi G., Cancelliere A. and Giuliano G., 2005. Case Study: Multicriteria Assessment of Drought Mitigation Measures. Journal of Water Resources Planning and Management. Salas J.D., Fu C., Cancelliere A., Dustin D., Bode D., Pineda A., and Vicent E., 2005. Characterizing the severity and risk of drought in the Poudre River, Colorado. Journal of Water Resources, Planning and Management. U.S. Army Corps of Engineers, Institute for Water Resources, (USACE), 1994. Managing Water for Drought. IWR Report 94 – NDS – 8, USACE, USA. Water Conservation Committee (WCC), 2002. Drought Management Planning. Drought Management Handbook, pp. 1-37. American Water Works Association, Denver. Western Drought Coordination Council (WDCC), 1998. How to reduce drought risk. Western Drought Coordination Council, USA. Wilhite D.A., Hayes M.J. and Knutson C.L., 2005. Drought Preparedness Planning: Building Institutional Capacity. Chapter 5 (pp. 93-135) in Drought and Water Crises: Science, Technology and Management issues. Ed. Taylor and Francis. Yevjevich, V.M., 1967. An objective approach to definition and investigation of continental hydrologic droughts. Hydrology Pap. 23, Col. State Univ., Fort Collins, USA.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 3

UNDERSTANDING THE INTERACTIONS BETWEEN AGRICULTURAL PRODUCTION AND CLIMATE VARIABILITY C. Rodríguez-Puebla∗ and S. M. Ayuso Dept. de Física de la Atmósfera, Universidad de Salamanca. Spain

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In this study a statistical model was derived to estimate the response of winter cereal to climate factors. Rotated and unrotated Empirical Orthogonal function analyses were performed to reduce the dimensions of the atmospheric fields and to identify the model predictors. The statistical model included components corresponding to the following climate variables: the Standardized Precipitation Index, the minimum temperature, the Scandinavian teleconnection index, the Southern Oscillation index, the zonal wind at 200 hPa and the North Atlantic Oscillation. The trend observed in some of the large-scale variables, for example, the increase in SLP towards the western Mediterranean could affect the yield negatively.

INTRODUCTION Agricultural losses are very dependent on weather and farmers need climatic and seasonal forecast information that anticipates variations in crop production. In this study we propose a method to translate climate information into production, providing results for winter cereal yield. Winter cereal is a major crop in Spain (about 14·103 tons (t) per year), accounting for approximately 10% of the annual European production (http://www.mapa.es/estadistica/). A number of studies have examined the climate-crop yield relationships. For example, in Australia has been carried out by Nicholls (1997). The interactions between crops and ∗

[email protected]

36

C. Rodríguez-Puebla and S. M. Ayuso

teleconnection indices, such as El Niño/Southern Oscillation (ENSO) or the North Atlantic Oscillation (NAO), were examined by several authors, for example Alexandrov and Hoogenboom 2001 and Tao et al. 2004. General Circulation Models (GCMs) provide new opportunities for investigating the impacts of climate change on agriculture and Hansen and Sivakumar (2006) provided the advances in climate-based crop forecasting. Correlation analysis between Cereal Productivity (hereafter CP) and climate fields allowed us to identify the variables that contribute to CP variations. The climate fields considered were regional over the Iberian peninsula, such as precipitation, drought and temperatures, and large-scale over Atlantic-European area, such as sea-level pressure, moisture fluxes and jet stream. The climatic signals that affect CP were identified by applying empirical orthogonal function (EOF) analysis. An empirical/statistical model was derived to represent the response of CP to climate which considers the above-mentioned fields and some teleconnection indices.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RESULTS Historical agricultural data, including crop production and harvested areas, were provided by the government records from the “Ministerio de Agricultura Pesca y Alimentación” (MAPA). The major winter cereals considered are wheat, barley and rye. The yields at different locations distributed over Spain were aggregated to obtain the CP time series (Figure 1a), which is defined as the production relative to the harvested area. Among the causes of CP variability are the changes in agricultural practices, technological improvement and agricultural policy of the countries, these effects cause a trend that is independent of climate (Lobell and Field 2007). Therefore, a linear fit trend was subtracted from the data and figure 1b shows the time series of the anomalies with respect to the trend, which is going to be analysed. The regional climate precipitation and maximum and minimum temperatures were provided by the Meteorological Institutes of Spain and Portugal. We considered the Standardized Precipitation Index (SPI hereafter) as a measure of drought, which is the departure of precipitation from the normal amount. The SPI was computed by fitting a gamma probability density function to the frequency distribution of precipitation of the monthly precipitation data (Hayes et al. 1999; Vicente-Serrano 2006). The large-scale atmospheric fields used in this study are from the reanalysis data of NCEP/NCAR (Kalnay et al. 1996). The contribution of the teleconnection indices to CP are also considered, the indices having been provided by the Climate Prediction Center (CPC) available online at http://www.cpc.ncep.noaa.gov/products/MD_index.shtml (Northern Hemisphere teleconnection patterns and monthly atmospheric and sea surface temperature indices), and from http://www.cru.uea.ac.uk/cru/data/nao.htm for the case of the North Atlantic Oscillation index (NAO). Rotated and unrotated EOF analyses were performed to reduce the dimensions of the atmospheric fields while retaining signals of associated variance (Jolliffe, 2002). The comparison between correlation maps of CP with climatic fields and the spatial variation modes or EOFs allow us to select the potential predictors that capture the CP response to climate variability (Rodriguez-Puebla et al. 2007).

Understanding the Interactions between Agricultural Production and Climate…

37

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. a) Winter cereal productivity in bars in Spain and linear trend (blue line); b) Winter cereal productivity anomalies with respect to the trend.

Figure 2a shows the correlation pattern between CP and the SPI in May. Larger correlation values were obtained toward the south of the Iberian Peninsula (IP hereafter). This correlation pattern resembles the spatial configuration of the second rotated EOF of SPI variability (figure not shown), which accounts for 21% of the total variance and characterizes drought variability toward the south-eastern part of the IP. The opposite connection between CP and maximum temperature (TX) can be seen in the correlation pattern (Figure 2b), which is linked to the third rotated EOF of the TX (figure no shown), which accounts for 24% of the total TX variance. The correlation pattern between CP and TN in winter (DJF) (Figure 2c) resembles the first rotated EOF of TN (figure no shown), which accounts for 34% of total variance. Figure 3a shows the correlation pattern between CP and SLP, which resembles the third rotated EOF of SLP corresponding to May (REOF3_SLP_My), describing 14% of the total variance (figure no shown). CP shows a significant correlation with the zonal wind at 200 hPa level (JET) in May (Figure 3b) and the correlation pattern resembles the first EOF of the JET (figure no shown), which accounts for 29% of the total variance. It has also been found that the CP is related to the Scandinavian pattern in spring (SCA_MAMy). CP is also correlated with the negative phase of the North Atlantic Oscillation in May (NAO_My). A significant correlation was found between the SOI of May and CP. Table 1 depicted the correlation

38

C. Rodríguez-Puebla and S. M. Ayuso

coefficients between CP and the teleconnection indices and the principal components of the atmospheric variables.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure2. Correlation patterns in % between cereal productivity and: a) Standardized precipitation index (SPI) in May; b) maximum temperature in May; c) minimum temperature in December-JanuaryFebruary (DJF).

Figure3. Correlation patterns in % between cereal productivity and: a) Sea Level Pressure in May; b) zonal wind at 200 hPa level in May (JET).

Understanding the Interactions between Agricultural Production and Climate…

39

Table 1

r CM CL CU

RPC2_SPI _My 0.63 ± 0.08 0.1346 0.0324 0.2368

RPC1_TN_D JF 0.46 ± 0.14 0.1219 0.037 0.2068

PC1_JET_My

NAO_My

0.63 ± 0.11 0.0998 -0.0079 0.2075

-0.43 ± 0.10 -0.1026 -0.21 0.0043

SCAND_MA My 0.48 ± 0.11 0.125 0.036 0.2137

SOI_My 0.28 ± 0.12 0.022 -0.0708 0.1148

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Based on the previous results, a statistical model was proposed to estimate the response of winter CP to climate factors. The statistical model included the Standardized Precipitation Index (RPC2_SPI_My), the minimum temperature (RPC1_TN_DJF), the Scandinavian index (SCA_MAMy), the Southern Oscillation index (SOI_My), the zonal wind at 200 hPa (PC1_JET_My) and the North Atlantic Oscillation (NAO_My). Maximum temperature was not entered into the model because its effect was represented by other variables such as the SPI and JET. The results of the model give a correlation coefficient with the observed CP data of 0.86 ± 0.05 for the entire period. This means that approximately 75% of the year-to-year CP variance is described by the climate predictors. Figure 4 shows the model estimations with its confidence interval (error bars) evaluated by cross-validation. Table 1 depicts the results of the correlation coefficients (r) between each predictor and CP, including also the mean (CM), lower (CL) and upper (CU) model regression coefficients.

Figure 4. Time series of cereal productivity observed in bars, the results of the regression model with a thick line and the confidence interval for the estimations with vertical lines.

CONCLUSIONS Our study provides a description of the relationships between the climatic factors and CP variations that could be used to translate climate information into production for farming decisions. The model considers the benefit effects of abundant precipitation (SPI) and the

40

C. Rodríguez-Puebla and S. M. Ayuso

dynamic aspects of the air mass represented by the JET, NAO and SCA indices, during the maturation stage of cereal. It furthermore contains the favourable effects of warm winters at the beginning of cereal growing. The influence of the SOI on winter cereal yield could be explained by its indirect effects on the undulation of atmospheric circulation. The trend observed in some of the large-scale variables, for example, the increase in SLP towards the western Mediterranean and the decrease in the SCA index could affect the yield negatively.

ACKNOWLEDGMENTS The authors would like to thanks the two anonymous reviewers for their comments and suggestions. Thanks to the Meteorology Institutes of Spain and Portugal for “in situ data” and NCEP/NCAR for reanalysis data, the GrADS software developers and the SCI of the University of Salamanca for the English revision. This work was funded by National Research Project MEC-CGL2005-06600CO3-01/CLI and the Regional Project of Castilla and León SA039/A05 with FEDER European funds.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Alexandrov VA, Hoogenboom G (2001) Climate variation and crop production in Georgia, USA, during the twentieth century. Climate Research 17:33-43. Hansen JW, Sivakumar MVK (2006) Advances in applying climate prediction to agriculture. Climate Research 33:1-2. Hayes MJ, Svoboda MD, Wilhite DA, Vanyarkho OV (1999) Monitoring the 1996 drought using the standardized precipitation index. Bulletin of the American Meteorological Society 80:429-438. Jolliffe I (2002) Principal Component Analysis. Springer. Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, Iredell M, Saha S, White G, Woollen J, Zhu Y, Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mo KC, Ropelewski C, Wang J, Leetmaa A, Reynolds R, Jenne R, Joseph D (1996) The NCEP/NCAR 40-year reanalysis project. Bulletin of the American Meteorological Society 77:437-471. Lobell DB, Field CB (2007) Global scale climate–crop yield relationships and the impacts of recent warming. Environ. Res. Lett. 2, doi:10.1088/1748-9326/2/1/014002. Nicholls N (1997) Increased Australian wheat yield due to recent climate trends. Nature 387:484-485. Rodriguez-Puebla C, Ayuso S.M., Frías M.D., García-Casado L.A. (2007) Effects of climate variations on Winter cereal production in Spain. Clim. Res. (in press). Tao FL, Yokozawa M, Zhang Z, Hayashi Y, Grassl H, Fu CB (2004) Variability in climatology and agricultural production in China in association with the East Asian summer monsoon and El Nino Southern Oscillation. Climate Research 28:23-30. Vicente-Serrano SM (2006) Spatial and temporal analysis of droughts in the Iberian Peninsula (1910-2000). Hydrological Sciences Journal-Journal Des Sciences Hydrologiques 51:8397.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 4

ADAGIO – ADAPTATION OF AGRICULTURE IN EUROPEAN REGIONS AT ENVIRONMENTAL RISK UNDER CLIMATE CHANGE Gerhard Kubu∗ and Josef Eitzinger Institute of Meteorology, University of Natural Resources and Applied Life Sciences, (BOKU), Vienna, Austria

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT ADAGIO is a Specific Support Action (SSA) of the European Sixth Framework Program with 11 partners in 3 regions as Central Europe, Eastern Europe and the Mediterranean Area. Historically, agricultural management options have been improved according to economic and environmental conditions. This has been achieved by adopting new technologies and production strategies. Hence, future agricultural policy should be adapted to the new climate conditions. To increase awareness of these potential consequences regional courses, meetings with local experts and farmers will be launched and published in relevant media. Furthermore, the interaction with stakeholders and agricultural decision-makers will point out which other relevant policies, as CAP, WFD, etc. could interact with the foreseen agricultural climate risks. The activities range from establishing the working groups, recognition of most vulnerable regional issues through climate change, identification of feasible potential adaptation measures up to integration into management strategies and agricultural policies. In order to ensure the state of art and the exchange of Know-How at the European level, thematic groups will be created to be a forum of discussion, exchange of Know-How and results, and methods. Thematic groups are: 1) Adaptation of farm production practices 2) Adaptation to climate-related pest and disease risks 3) Adaptation strategies by changing land-use and crop selection ∗

[email protected]

42

Gerhard Kubu and Josef Eitzinger 4) Implementation of adaptation into management strategies and into agricultural policy The project started on 1st January 2007 and will end after 30 months.

ADAGIO OBJECTIVE AND PROJECT ORGANIZATION The overall objective of ADAGIO is the development and dissemination of recommendations on how to better adapt agriculture to climate change in three European regions considered to be significantly affected by climatic change (representing also different climatic conditions and agricultural systems over Europe). The organization structure of the ADAGIO project is shown in Figure 1.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Organization structure of ADAGIO.

The main specific objects of the ADAGIO project are: −

−

To identify significant vulnerability issues in each partner country and the related potential problems due to climate change (WP2) : o Selection of most vulnerable regional issues (e.g. identifying climatic risks) through climate change (defined e.g. by various agroecosystems) by each partner and identifying related potential problems. o To describe the related agroecosystems regarding their main limitations, observed trends, socio-economic conditions. To identify feasible potential adaptation measures for the selected regional agricultural systems, based on the identified problems (WP3) :

ADAGIO – Adaptation of Agriculture in European Regions…

43

Evaluation and description of feasible potential adaptation measures for the selected regional agroecosystems, based on the identified problems from WP2. o Identify ongoing trends or adaptations in order to evaluate potential future adaptation measures. o Identify and describe uncertainties, cost/benefits, risks, opportunities for cobenefits etc. of potential adaptation measures. To identify and demonstrate dissemination strategies of adaptation measures to decision-makers (WP4) : o Identify, recommend and disseminate strategies of adaptation measures to decision-makers. o Demonstrate dissemination strategies of adaptation measures at the national, regional and international level. o

−

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

A graphical representation of the work packages (only support activities) is shown in Figure 2.

Figure 2. Scheme of specific support activities in ADAGIO.

First General Meeting in Vienna February 2007 The first general meeting was held in Vienna from 14.2.07-16.2.07 where all ADAGIO partners were present. The meeting was combined with an official part where all partners presented the actual situation regarding impact of climate change and potential adaptation measures in agriculture in their countries (a first spotlight which will be further investigated in WP2).

44

Gerhard Kubu and Josef Eitzinger

Moreover 4 invited experts gave a presentation on various views on the problem of adaptation and dissemination of information to the end-users. More than 30 additional participants, mostly students, representatives from extension services and researchers attended this official part of the meeting. During the internal part of the meeting, the partners presented their list of established national contacts and their planned activities during WP 2. Further several financial issues and planned regional and next general meetings were discussed. The ADAGIO logo was discussed as well as the ADAGIO web site, which is already online since March 31st 2007 (www.adagio.org). The official ADAGIO website was established on a server of the University of Natural Resources and Applied Life Sciences, Vienna. The address is: www.adagio-eu.org This website provides with information about the ADAGIO project in general and in detail. The start page is shown in Figure 3.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. ADAGIO start page on www.adagio-eu.org.

The internet site gives detailed project information: The Project Info tells about Specific Support Actions in general, shows the ADAGIO organisation structure, the work packages and planned support activities. All Partners are listed with links to the responsible organisation, links to national ADAGIO websites (as already available) and contact person. Meetings: Detailed information is available about the first general meeting in February 2007 including program, participant list and all presentations for download as pdf file. All planned meetings, general, regional and national, are scheduled and linked with further information as available. A link to the Discussion Forum is placed at the main page. The Restricted Area is reserved to project members and provides project partners with internal documents.

ADAGIO – Adaptation of Agriculture in European Regions…

45

ADAGIO FORUM The ADAGIO discussion forum was at first established on a national server but after massive problems with spammers it was moved to yahoo.com. The address is: http://tech.groups.yahoo.com/group/adagio_forum/ This forum is restricted to registered users only, participation rights are provided by the ADAGIO project coordinator. For registration to the forum contact [email protected] please. Besides a general discussion forum it is designated to the four thematic groups: • • • •

Adaptation of farm production practices Adaptation to climate-related pest and disease risks Adaptation strategies by changing land-use and crop selection Implementation of adaptation into management strategies and into agricultural policy

INVITATION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Dear interested colleagues, the ADAGIO project group invites you to visit our web site www.adagio-eu.org where you will find a lot of useful information, digital documents for download and last but not least – please contact us and take part in our discussion forum. The ADAGIO project started in January 2007 and will end after 30 months on June 30th 2009.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 5

WSSTP VISION ON THE EUROPEAN AGRICULTURE AND WATER ISSUES AS THE RESPONSE TO CLIMATECHANGE CHALLENGE Marek Nawalany∗ Warsaw University of Technology Nowowiejska 20, 01-653 Warsaw, Poland

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The article is an attempt of grasping complexity of the water and agriculture issues on the European scale from the perspective of challenges imposed by climate change. After describing the water situation within the context of agriculture in different climatic zones of Europe, major natural and man-made impacts amplified by climate change are shortly described. This is followed by overview of internal and external factors that govern water-agriculture relationships. Finally, water vision within the agricultural sector based on the Water Supply and Sanitation Technology Platform (WSSTP) findings is presented. Most of the data that are used in the article are derived and cited from the Chapter on Water and Agriculture (Final Report, 2006) produced by the Technical Working Group TWG-4 of which author was a chairman for more than one year.

INTERNAL EUROPEAN WATER ISSUES Governing of water in Europe with respect to food production is an exceptionally complex issue involving a number of spatial and temporal scales, two essentially distinct climatic zones and multitude of organizations interlinked by all kinds of feed-back dependencies. Geographically, water use and water demand varies considerably throughout Europe. The water actors competing for the resource using economic and political means make the overall picture even more convoluted. In total, agriculture accounts for 30 Per cent

∗

[email protected]

48

Marek Nawalany

of the water use in Europe but, as it can be seen from the statistical evidence, the conditions differ widely between countries – see Table 1.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Water use of agriculture in Europe (from Aquastat, 2003) Country

Total water use (km3/year)

Agricultura Agricultural Irrigated l water use water use land (ha) (km3/year) (as % of total water use)

EU-25

241

123

51

12792628

Austria Belgium Cyprus Czech Rep Denmark Estonia Finland France Germany Greece Hungary Ireland Italy Latvia Lithuania Luxemburg Malta Netherlands Poland Portugal Slovakia Slovenia Spain Sweden United Kingdom

2.1 0.24 2.6 1.3 0.16 2.5 40 47 7.8 7.6 1.1 44 0.29 0.27 0.06 7.9 16 11 36 3.0 9.5

0.021 0.17 0.055 0.54 0.0080 0.066 3.9 9.3 6.2 2.5 0.00020 20 0.036 0.018 0.014 2.7 1.4 8.8 24 0.26 0.28

0.99 71 2.1 42 4.9 2.7 10 20 81 32 0.018 45 12 6.6 25 34 8.3 78 68 8.9 2.9

4000 39938 24000 476000 3680 64000 2000000 485000 1422000 210000 2698000 20000 9247 763 565000 100000 632000 174000 2000 3640000 115000 108000

Year of Annual irrigation rain fall data (mm/year)

1998 1994 1998 1998 1995 1998 1998 1998 1998 1998 1998 1995 1995 1990 1998 1998 1998 1998 1998 1998 1998 1998

1110 847 498 677 703 626 537 867 700 652 589 1118 832 641 656 934 383 778 600 855 824 1162 636 624 1220

In the South, agriculture accounts for a much higher than average proportion of the total water use (mostly for irrigation), while the opposite is true for northern countries like for example Sweden and Poland. This means that when discussing water and agriculture in an European perspective we need to keep in mind that we are actually talking about two subcontinents – a southern and a northern one – with profoundly different conditions. From an environmental perspective, agriculture is not only one of the major consumers of water but it also a polluter, emitting a considerable flux of liquid and solid waste into environment. Long term water protection policy of the European Union has intervened by establishing in 2002 the Water Framework Directive (WFD) aimed at achieving good status for all European waters by December 2015. The overall objective of the Directive is to get

WSSTP Vision on the European Agriculture and Water Issues…

49

polluted waters clean again, and ensure clean waters are kept clean. Advocated by WFD is the Integrated Water Resources Management (IWRM) - concept that is expected to encapsulate all the complexities of water issues and become commonly accepted practice in water sector in Europe. In particular, it is envisaged that IWRM will guide decision makers and water users in Europe towards sustainable coexistence of agriculture and water resources. However, there is a factor of economic and political risk in implementing IWRM as the latter was never tried on large scale of region or state. Additionally, it must be realized that water issues in agriculture are also strongly linked with the use of water by other sectors which in many instances have their own water problems unresolved, e.g. old and leaking European sanitary infrastructure.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

EXTERNAL WATER CHALLENGES External challenges emerge and generate a political dilemma of choosing priorities when solving continental and global water related problems at the same time; e.g. European Water Initiative (EWI) supporting Millennium Development Goals (MDG) is oriented towards economically unprivileged countries which are in urgent need to improve their water and sanitary situation. European involvement in the initiative clearly compete with solving the continent own unresolved water problems, especially when agricultural and water interests of new member states are to be taken into account. The future course of water and agriculture in Europe is difficult to predict as the European policies, like CAP and EWI are not only conditioned on the continent internal situation but also meet on the global arena similar initiatives of the USA, Japan and China. An important external political factor that will be influencing the future of the water and the agriculture sectors is the World Trade Organization (WTO). By imposing trade rules and setting conditions of competition between markets on the global scale, WTO will directly affect European trade of agriculture goods and hence, indirectly, water sector in Europe. It can be also envisaged that agricultural subsidies within the European Union will be increasingly questioned in the future, resulting in important changes in the economic activities of the sector in and outside Europe. The issues of subsidies for agriculture will become even more complex in view of the energy market increasing demand for bio-energy, renewable energy the agriculture can offer. Structuring and implementing water policy within the agriculture sector will definitely remain an important factor in European politics for next decades in spite of its complicated links with externalities, like global trade and global politics.

AUTONOMOUS CHANGES One of the major challenges Europe must face is how to make the complex wateragriculture link adaptable in response to the unstoppable autonomous factor - the climate change - caused by cosmic and human factors it intervenes into the whole water cycle within a series of time scales. In particular, the cosmic cause of climatic changes on the Earth results from eccentricity of the Earth orbit around the Sun (period T = 110 000 years), inclination of

50

Marek Nawalany

the Earth rotation axis (period T = 41 000 years) and from precession of the Earth rotation axis (period T= 26 000 years). The human factor is represented by an increase in concentration of CO2 (and other greenhouse gases) in the atmosphere and results from burning fuels mainly by industry and transport. In the years to come it will increase absorption of the Earth long wave radiation within the atmosphere thus causing increase in temperature of the atmosphere. It is envisaged that number of by-effects will follow and act as a feedback thus causing further increase/decrease in temperature, e.g. (+) melting of ice caps at the North and South poles (lowering of the Earth albedo), (+) increase of global evaporation (lowering of long wave radiation transmission through the atmosphere as the result of water vapour increase), (+) increase in cloudiness (increase in backward scattering of the long wave radiation), (-) increase in algae growth in the oceans (decrease of CO2 ). All this will effect water resources: number of droughts and flood events are likely to increase; differences in water resources between Northern and Southern Europe will be widened; Northern Europe is likely to have increasing water abundance whereas southern Europe may face growing water shortage or even water scarcity. Clearly, question how to make the agriculture sector more sustainable needs not only be well posed but above all well answered.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ROLE OF THE WATER SUPPLY AND SANITATION TECHNOLOGY PLATFORM (WSSTP) Newly initiated Water Supply and Sanitation Technology Platform (WSSTP) offers the technology-oriented way of approaching European water problems within the context of natural changes and conditioned by economy, social issues and politics. Long term water vision of WSSTP when considering internal and external factors influencing agricultural and water issues takes as a very basic assumption unavoidable autonomous changes. The WSSTP is aimed towards developing new knowledge and futuristic technologies that may help in making “frog-leap” advance in some water issues, which otherwise cannot be solved with the use of conservative and evolutionary approach. Water in Agriculture is one of the major chapters of the WSSTP Vision. It is expected that solutions found in food production technologies will result in savings of water in terms of its quantity as well as in protecting water resources from agriculture-born pollution. The Chapter on Agriculture contains justification and definitions of research needs, priorities and research goals for years 2010, 2020 and 2030 corresponding to five sub-sectors of the European agriculture: Rain-fed agriculture, Irrigated agriculture, Livestock production, Aquaculture and Greenhouses. There is a number of Drivers that have been identified by WSSTP to allow structuring European (water) Vision and then to propose ways of getting into the Vision. Clearly, although strategies are quite well defined, means and methods necessary to achieve the Vision are still under debate at the level of the European Commission. Below, Drivers identified are shortly listed: • •

Water Quantity Management (supply, storage, drainage; fair distribution when water demand is increasing) Water Quality and Environmental Impact (preventing the degradation of quality of water bodies by reduction of emissions, water treatment etc.)

WSSTP Vision on the European Agriculture and Water Issues… • • • •

51

Cost-effectiveness and affordability (improve the economy of processes by development and introduction of cost-effective and affordable technology) Reliability of processes and equipment (innovations driven by legislation or need for competitiveness) Product Quality (e.g. improvement and monitoring of quality at the point of primary production) Networking and Knowledge Transfer (cooperating in developments, sharing knowledge on an European scale, training of farmers, etc.)

WATER VISION General vision on agriculture developed by WSSTP states: “Agriculture – as an important water user – will be economically viable and competitive in the world market. The sector will become more flexible to accommodate the new demands of the market (including environmental demands) and will improve its sustainability.” . This means in particular that • • • •

Agriculture will produce sufficient, affordable and safe food and other agro products, while achieving sustainability. Agriculture will use water more efficiently, and will make better use of nonconventional resources. Agriculture will increasingly require new technologies, equipments and facilities from the water sector resulting in strong economic stimuli for this sector. Agriculture will integrate environmental protection in production

Chapter Vision on Irrigated Agriculture envisages the following: •

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

• •

Given the future possible climate change, transformations in land use and increase in supplementary irrigation schemes irrigated agriculture will have to face an increasing imbalance between water supply and demand. Therefore, irrigated agriculture will have to be more effective and flexible in its management of water resources. Irrigated agriculture will optimise its productivity with sustainable use of natural resources (e.g. water). Finally, irrigated agriculture will have an important role in nature preservation in rural areas.

HOW THE VISION CAN BE REALIZED – TECHNOLOGICAL PERSPECTIVE Additionally, the Chapter specifies priorities for developing water related technologies and summarizes them into three clusters:

52

Marek Nawalany •

•

•

Technology and tool development for increasing of water use efficiency and related savings (use of non-conventional resources, intelligent irrigation systems, improved integrated water management methods, introduction of crops tolerant to salinity and drought resistance), Technology and tool development for optimal use of inputs for agricultural production, in order to safeguard the environment and improve socio-economic benefits (improved knowledge on biogeochemistry and the fate of nutrients, agrochemicals and other organics, including those from various bio-solids; improved knowledge of scaling processes; closing the mass cycles and reduction of emissions; integral monitoring and application systems), Improved governance, institutional framework and stakeholders participation for a sustainable use of water in agriculture (developing tools for farmers to mitigate and adapt to extreme events, governance tools, knowledge building and transfer).

The WSSTP water Vision is presently followed by formulation of Strategy and search for operational Tools for the Strategy realization.

CLOSING OBSERVATION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Whether the WSSP Vision and the corresponding Strategy will be accepted and ultimately made operational depends not only on proven successes in introducing the envisaged technologies or availability of financial resources but above all on factors like active society participation, understanding the water-agriculture issues by decision makers in both sectors and (good) political will within the top circles of the European politics. The prerequisite for this to happen is strong formulation and dissemination of the specific knowledge on water, agriculture and changing environment contained in the WSSTP Vision followed by its general acceptance together with clearly defined socio-economic constraints.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 6

LAND-AIR PARAMETERISATION SCHEME (LAPS): A TOOL FOR USE IN AGROMETEOROLOGICAL MODELLING D. T. Mihailovic∗ and B. Lalic Faculty of Agriculture, University of Novi Sad, Novi Sad, Serbia

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The land surface scheme LAPS (Land Air Parameterisation Scheme) has been developing since 1996. Its first version was comprehensively elaborated in “Global and Planetary Change” (Mihailovic, 1996). After that, LAPS was intensively tested and calibrated using a vast number of data for comparison versus one point and 3-D simulations. Recently, we have focused on the parameterisation points potentially providing conditions that LAPS can be used in both long-term climate simulations, shortterm weather forecasting applications and also in ecosystem, air pollution, chemical, crop, groundwater, regional-scale hydrological budgets, partitioning of annual carbon fluxes and soil contaminant modelling. Here we shortly elaborate the main features of this scheme that includes modelling the interaction of the land surface and the atmosphere, under processes divided into three sections: subsurface thermal and hydraulic processes, bare soil transfer processes and canopy transfer processes. They are: interaction of vegetation with radiation, evaporation from bare soil, evapotranspiration including transpiration and evaporation of intercepted water and dew, conduction of soil water through the vegetation layer, vertical water movement in the soil, surface and subsurface runoff, heat conduction in the soil and momentum transport within and above the vegetation. This scheme can be used as a useful toll for different purposes in agricultural science and practice.

∗

[email protected]

54

D. T. Mihailovic and B. Lalic

1. INTRODUCTION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In a very near past the atmospheric modelling community was not recognising that some aspects of the atmosphere–ecosystem–ocean system play(ing) a very important role in atmospheric circulations. Moreover, them had been prescribed relatively minor role. Ecosystem and soil processes and their effect on the atmosphere are certainly in this category. Therefore, most mesoscale and global atmospheric models of 20 years ago either ignored or treated in an extremely simple manner interactions of the atmosphere with underlying surface (soil, homogeneous and sparse vegetation, other natural and artificial surfaces). Now, field and modelling studies have demonstrated that these interactions are extremely important in both long-term climate simulations, short-term weather forecasting applications and also in ecosystem, air pollution, chemical, crop, groundwater, regional-scale hydrological budgets, partitioning of annual carbon fluxes and soil contaminant modelling (Dickinson, 1984; Pielke et al. 1998; Chase et al. 1998; Baron et al. 1998; Stohlgren et al. 1998; Mihailovic, 2005). Because the role of these interactions has been recognised, parameterisations of vegetation and soil processes have progressively become more sophisticated over the years in order to treat the complexities of the physical system through soil–vegetation–atmosphere transfer (SVAT) schemes employed in general circulation, mesoscale, and small-scale atmospheric numerical models (Deardorff, 1978; Avissar et al., 1985; Dickinson et al., 1986; Sellers et al., 1986; Mihailovic et al., 1993; Bosilovich and Sun, 1995; Viterbo and Beljaars, 1995; Pleim and Xiu, 1995; Cox et al., 1999; Walko et al., 2000). The Land Air Parameterisation Scheme (LAPS) is one SVAT scheme that is designed at the University of Novi Sad (Serbia) to be a component of any aforementioned model. It has been developing since 1996 through: (i) the scheme background making work (Mihailovic et al., 1993; Mihailovic et al., 1995; Mihailovic, 1996; Mihailovic et al., 2004), (ii) single point validation (Mihailovic and Kallos, 1997; Mihailovic et al., 2000) and (iii) 3-D simulation with the NCEP non-hydrostatic model with the implemented LAPS scheme (Mihailovic, 2003; Mihailovic, 2004). In this chapter we describe a new version of the LAPS scheme that includes refinement of turbulent transfer coefficient for calculation of aerodynamic resistances and air temperature inside tall grass canopies and forest. Section 2 describes the LAPS governing equations (Subsection 2.1) and representation of fluxes (Subsection 2.2). Section 3 considers calculating the turbulent transfer coefficient inside: (i) tall grass (Subsection 3.1) and (ii) forest (Subsection 3.2) canopies. In Section 4 are given concluding remarks.

2. THE STRUCTURE OF THE SCHEME 2.1 Governing Equations The LAPS scheme uses the morphological and physiological characteristics of the vegetation community for deriving the coefficients and resistances that govern all the fluxes between the surface and atmosphere. The model has seven prognostic variables: three temperatures (canopy, soil surface and deep soil);, interception store for canopy; and three soil moisture stores. The prognostic equations for the canopy temperature, Tf, the soil surface temperature, Tg and deep soil temperature, Td are

Land-Air Parameterisation Scheme (LAPS)

Rnf = λE f + H f + C f ∂T f / ∂t

,

Rng = λEg + H g + G + C g ∂Tg / ∂t

55

(1) ,

Rng = λE g + H g + C g 365 / 2∂Td / ∂t

(2) ,

(3)

where Rn is the net radiation, the latent heat of vaporisation, E the evapotranspiration rate, H the sensible heat flux, G the soil heat flux, C the heat capacity and T the surface temperature. The subscripts f and g refer to the upper-level canopy and soil, respectively. The soil heat flux is parameterised using "force-restore" method. The ground heat capacity Cg is parameterised following Zhang and Anthes (1982). he prognostic equations for the water stored on the canopy, wf is

∂w f / ∂t = Pf − Ewf / ρ w

(4)

where w is the density of water, Pf the water amount retained on the canopy, Ewf the evaporation rate of water from the wetted fraction of canopy. When the conditions for dew formation are satisfied, the condensed moisture is added to the interception store, wf. The parameterisation of the soil content is based on the concept of the three-layer model (Mihailovic, 1996). The governing equations take the form

∂ϑ1 / ∂t = {P1 − F1, 2 − ( Eg − Et f ,1 ) / ρ w − R0 − R1 }/ D1

∂ϑ2 / ∂t = {F1, 2 − F2 ,3 − Et f , 2 / ρ w − R2 }/ D2

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

∂ϑ3 / ∂t = {F2 ,3 − F3 − R3 }/ D3

(5) (6) (7)

where ϑi is the volumetric soil water content in the ith layer, P1 the infiltration rate of precipitation into the upper soil moisture store, Di is thickness of the ith soil layer, Fi,i+1 the water flux between i and i+1 soil layer, F3 the gravitational drainage flux from recharge soil water store, Etf,1 and Etf,2 the canopy extraction of soil moisture by transpiration from the rooted first and second soil layers, respectively, R0 the surface runoff and Ri the subsurface runoff from the ith soil layer. Eqs. (1)-(3) are solved by an implicit backward method , while Eqs. (4)-(7) are solved using an explicit time scheme.

2.2. Representation of Energy Fluxes Our treatment of the energy fluxes may be classified as the so-called "resistance" representation. This formulation is often used to describe the energy fluxes in an Ohm′s law analogue form: flux = (potential difference)/resistance.

(8).

56

D. T. Mihailovic and B. Lalic

Potential difference for sensible and latent heat fluxes can be expressed in terms of chosen prognostic variables, atmospheric boundary layer reference temperature and water vapour pressure in the canopy air space. Since new approaches for aerodynamic resistances and “parameter” and “flux” aggregation will be considered in the further text, here we will shortly describe physical background of the energy fluxes used in the model. The LAPS schematic diagram of transfer pathways for latent and sensible heat fluxes is shown in Figure 1.

er

Tr

Atmospheric boundary layer

ra

ra λEf + λEg λEf rb

Hf +Hg e (Tf) rc

Tf

Hf rb

Ta

rd

H

Hg λEg

rd

rplant

r h e (Tg) , αs

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

*

r

Tg

Soil

Figure 1. Schematic diagram of the Land- Air Parameterisation Scheme (LAPS). The transfer pathways for latent sensible heat fluxes are shown on the left- and right-hand sides of the diagram, respectively.

The latent heat flux from canopy vegetation to canopy air space is given by

λE f = [e* (T f ) − ea ][ w f / rb + (1 − ww ) /(rb + rc )]ρc p / γ

,

(9)

where e*(Tf) is the saturation water vapour pressure at the canopy temperature Tf, ea the water vapour pressure of the air at the canopy source height, ww the wetted fraction of canopy, rb the bulk boundary layer resistance for the canopy leaves, rc the bulk stomatal resistance of the canopy leaves, the density of air, cp the specific heat of air at constant pressure and the psychrometric constant.

Land-Air Parameterisation Scheme (LAPS)

57

The sensible heat flux from canopy vegetation to canopy air space is given by

H f = 2(T f − Ta )ρc p / rb

(10)

where Ta is the temperature of air at the canopy source height. The latent and sensible heat fluxes from soil surface are parameterised as

λE g = ρc p (1 − σ c )[α s e* (Tg ) − ea ] /(rsurf + rd ) / γ

(11)

H g = ρc p (T f − Ta ) / rd

(12)

where c is the fractional vegetation cover, e*(Tg) the saturation water vapour pressure at the soil surface temperature Tg, rsurf the bare soil surface resistance and rd the aerodynamic resistance between the soil surface and the canopy source height. Parameter s is calculated according to Mihailovic et al. (1995) as a function of the volumetric soil moisture content of the top soil layer, ϑ1 and field capacity, ϑfc 2

⎧1 - [(ϑ fc − ϑ1) / ϑ fc )] εs = ⎨ 1 ⎩

ϑ1 ≤ ϑ fc ϑ1 > ϑ fc

(13)

The flux Ewf from the wetted portion of foliage, with wetted fractions denoted by wf according to Eq. (9) is

λEwf = [e* (T f ) − ea ](ww / rb ) ρc p / γ

.

(14)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The fraction of the foliage that is wet, ww is parameterised according to Deardorff (1978) and Dickinson (1984). Transpiration occurs only from dry leaf and it is only outward. This physiological process is parameterised with the equation

λEtf = [e* (T f ) − ea ][(1 − ww ) /(rb + rc )]ρc p / γ

(15)

where Etf is the transpiration rate from foliage. Dew formation occurs when e*(Tf)≤e*. In that case the condensed moisture is added to the surface interception store, ww. The transpiration rate is zero under this condition. Air within the canopy has negligible heat capacity, so the sensible heat flux from the canopy, Hf and from the soil surface, Hg must be balanced by the sensible heat flux to the atmosphere, Ht

H t = H f + H g = (Ta − Tr )ρc p / ra

(16)

58

D. T. Mihailovic and B. Lalic

where Tr is the air temperature at the reference level zr within the atmospheric boundary layer and ra the aerodynamic resistance between canopy air space and reference level. Similarly, the canopy air is assumed to have zero capacity for water storage so that the latent heat flux from canopy air space to reference level in atmospheric boundary layer, Et between the latent heat flux from canopy vegetation to canopy air space, Ef and latent heat flux from soil surface to canopy air space, Eg

λEt = λE f + λE g = [(ea − er ) / ra ]ρc p / γ

(17)

where er is the water vapour pressure of the air at the reference level within the atmospheric boundary layer. Diagnostic variables Ta and ea were calculated from Eqs. (16) and (17), using corresponding expressions for Hf, Hg, Ef and Eg

Ta = [2T f / rb + Tg / rd + Tr / ra ] /[2 / rb + 1 / rd + 1 / ra ]

(18)

and

ea = {1 / ra + ε s e* (Tg )(1 − σ c )/ (rsurf + rd ) + e* (T f )[(1 − ww ) /(rb + rc )]} / {1 / ra + (1 − σ c )/ (rsurf + rd ) + [(1 − ww ) /(rb + rc )]}

(19)

3. AERODYNAMIC RESISTANCES INSIDE AND OVER CANOPIES In the model, the vegetation is represented as a block of constant density porous material sandwiched between two constant stress layers, the height of the canopy top, hc and the height of the canopy bottom, hb (Fig. 1) (Sellers et al. 1986; Mihailovic and Kallos 1997). The aerodynamic resistance, ra between zr and the water vapour and sensible heat source height, ha (Sellers et al. 1986) is calculated as hr

zr

ha

hc

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ra = ∫ (1 / K m ) dz + ∫ (1 / K m ) dz

,

(20)

where Km is the turbulent transfer coefficient inside and above the canopy in the intervals (ha, hc) and (hc, zr), respectively. The aerodynamic resistance in canopy air space, rd is obtained as hb

ha

zg

hb

rd = ∫ (1 / K m ) dz + ∫ (1 / K m ) dz (21) where zg is the effective ground roughness length while the area-averaged bulk boundary layer resistance,

rb has the form (Sellers et al. 1986)

Land-Air Parameterisation Scheme (LAPS)

59

hc

1 / rb = ∫ Ld u( z ) /( Cs Ps ) dz ha

,

(22)

where Ld is the area-averaged stem and leaf area density (also called canopy density), which is related to leaf area index (LAI) as LAI = Ld (hc - h ) ; u(z) is the wind speed; Cs the transfer

coefficient (Sellers et al., 1986); and Ps the leaf shelter factor. Eqs. (21)-(22) can be modified to take into account the effects of nonneutrality. According to Sellers et al. (1986), the position of the canopy source height, ha is estimated by obtaining the centre of gravity of the 1 / rb integral. Thus, ha

hc

hc

hb

ha

hb

∫ ( Ld / rb ) dz = ∫ ( Ld / rb ) dz = ( ∫ ( Ld / rb ) dz) / 2 = 1 / rb

.

(23)

We may obtain ha by successive estimation.

3.1. Calculating the Turbulent Transfer Coefficient Inside Tall Grass Canopies The shear stress above the canopy was calculated according to the Monin-Obuhkov theory having the form

τ = ρ {ku r /[(( z − d ) / zo)] + ψ M }2

(24)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

where ur is the wind speed at a reference level, zr within the atmospheric boundary layer, k = 0.41 the von Karman′s constant, d the zero plane displacement height, z0 the canopy roughness length and M(z/L) the stability adjustment function for momentum transport and L Monin-Obuhkov length (Paulson, 1970). The function M(z/L) is given for stable conditions (z/L > 0) by M(z/L) = 4.7 z/L and for unstable (z/L < 0) by

ψ M ( z / L ) = −2 ln[(1 + x) / 2] − ln[(1 + x 2 ) / 2] + 2 tan −1( x) − π / 2 where x = [1 − 15 z / L ]

1/ 4

(24a)

.

The shear stress inside the canopy using the "K-theory" is expressed as

τ = ρ K m ∂u / ∂z

(25)

where Km is the turbulent transfer coefficient which is parameterised as

K m = σ s u( z ) ,

(26)

60

D. T. Mihailovic and B. Lalic

where

s

is a constant. The constant

s

is defined following Mihailovic et al. (2004)

σ s = 2Cdg2 hc /[Cd Ld (hc − hb )] ,

(27)

where Cdg = [ k (ln(hb / z g )] is the leaf drag coefficient estimated from the size of the 2

roughness elements of the ground (Sellers et al., 1986; Mihailovic et al., 2004), Cd the leaf drag coefficient. The wind speed inside the canopy is given by

u ( z ) = u (hc ) exp[− β (1 − z / hc ) / 2] where u(hc) is the wind speed at the canopy top. The extinction factor, morphology and is defined as

β 2 = 2Cd Ld (hc − hb )hc / σ s .

(28)  depends on the plant

(29)

Let us note that the wind speed profile (28) is a particular solution of equation

d u / dz 2 = 2Cd Ld (hc − hb )u 2 /(σ s hc ) that approximates the wind profile inside a dense 2

tall grass canopy fairly well (Brunet et al. 1994; Mihailovic et al., 2004). Between the canopy height and canopy bottom height, the wind profile attenuates exponentially according to Eq. (28), while beneath the canopy bottom height it follows a classical logarithmic profile of the form

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

u ( z ) = u (hc ) exp[− β (1 − z / hc ) / 2] ln( z / z g ) / ln(hb / z g ) .

(30)

In a sparse tall grass canopy (one in which the plant spacing is the order of the canopy height or larger), Km is strongly affected by processes in the environmental space, including the plants and the space above the bare soil fraction. Therefore, Km inside a sparse canopy, can be represented by some combination of turbulent transfer coefficients in that space. If, as a working hypothesis, we assume a linear combination weighted by the fractional vegetation cover, c (a measure of how sparse the tall grass is), then we can define K m = σ cσ s u + ku* (1 − σ c ) z , where u* is the friction velocity above the bare soil fraction and z the vertical coordinate. The wind speed inside the sparse canopy is calculated from the equation (Mihailovic et al., 2004)

a (u , z )d 2 u / dz 2 + b( z )du / dz + c(du / dz ) 2 = gu 2 , where, a (u , z ) = σ cσ s u + ku* (1 − σ c ) z c = σ cσ s

b( z ) = ku* (1 − σ c ) z g = σ c [Cd Ld (hc − hb )] / hc .

(31)

Land-Air Parameterisation Scheme (LAPS)

61

To get the wind profile inside the sparse canopy, we have to solve Eq. (31) numerically, using, for example, the fourth-order Runge-Kutte method (Ayers, 1952).

3.2. Calculating the Turbulent Transfer Coefficient Inside Forest Canopies The turbulent transfer coefficient, Km is calculated directly from the equation following (Mihailovic et al., 2004)

(du / dz ) /(dK m / dz ) + (d 2 u / dz 2 ) K m = σ c Cd Ld (hc − h)u 2 / hc

(32)

that is a differential equation of the first order and first degree, where Km is an unknown function. This equation can be solved for Km after assuming a functional form of solution for wind speed inside the vegetation canopy, containing an attenuating parameter that is obtained iteratively. Solution of this equation can be found if the wind speed is used to be a linear combination of two terms, expressing behaviour of the wind speed over dense and sparse vegetation. Thus,

u ( z ) = σ c u (hc ) exp[−δ (1 − z / hc ) / 2] + (1 − σ c )u*[ln( z / zb ) + ψ M ( z / L)] / k

,

(33)

where is an unknown constant to be determined, u* the friction velocity and zb the roughness length over the non-vegetated surface. The first term in the expression (33) is used to approximate the wind profile inside the vegetation canopy (Brunet et al., 1994; Mihailovic et al., 2004), while the second term simulates the shape of wind profile above bare soil. After we introduce Eq. (32) (33) into Eq. (32), and rearrange, we reach

dK m / dz + a( z ) K m = b( z ) ,

(34)

where

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

a( z ) =

δ 2σ c u (hc ) exp[−δ (1 − z / hc ) / 2] /(4 hc 2 ) + (1 − σ c )u* [−1 / z 2 + ψ M'' ( z / L)] / k δσ c u (hc ) exp[−δ (1 − z / hc ) / 2] /( 2 hc ) + (1 − σ c )u*[1 / z + ψ M' ( z / L)] / k

(35)

and

b( z ) = {σ c u (hc ) exp[ −δ (1 − z / hc ) / 2] + (1 − σ c )u*[ln( z / zb ) + ψ M ( z / L)] / k }2 x

σ c Cd Ld (hc − h) / hc δσ c u (hc ) exp[−δ (1 − z / hc ) / 2] /(2hc ) + (1 − σ c )u*[1 / z + ψ M' ( z / L)] / k here ψ m ( z / L ) = dψ m ( z / L ) / dz and ψ m ( z / L ) = d '

''

ψ m (z / L ) / dz 2 .

2

(36)

62

D. T. Mihailovic and B. Lalic

CONCLUDING REMARKS We have given a detailed description of recently introduced approaches in the parameterisation of land surface processes in the LAPS (Land Air Parameterisation Scheme). Our attention was focused on the approaches sophisticatedly capturing energy transfer processes through and over the heterogeneous underlying surfaces.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Ayers, Jr. F., 1952: Theory and Problems of Differential Equations. Shaum Publishing, New York, 296 pp. Avissar, R., Avissar, P., Mahrer, Y., and Bravdo, B., 1985: A model to simulate response of plant stomata to environmental conditions. Agric. For. Meteor., 64, 127–148. Baron, J. S., Hartman, M. D., Kittel, T. G. F., Band, L. E., Ojima, D. S., and Lammers, R. B., 1998: Effects of land cover, water redistribution, and temperature on ecosystem processes in the South Platte Basin. Ecol. Appl., 8, 1037–1051. Bosilovich, M., and Sun, W., 1995: Formulation and verification of a land surface parameterization for atmospheric models. Bound.- Layer Meteor., 73, 321–341. Brunet, Y., Finnigan, J. J., and Raupach, M. R., 1994: A wind tunnel study of air flow in waving wheat: Single-point velocity statistics. Bound.-Layer Meteor., 70, 95-132. Chase, T. N., Pielke, Sr. R. A., Kittel, T. G. F., Baron, J. S., and Stohlgren, T. J., 1998: Potential impacts on Colorado Rocky Mountain weather and climate due to land use changes on the adjacent Great Plains. J. Geophys. Res., 104, 16 673–16 690. Cox, P.M., Betts, R.A., Bunton, C.B., Essery, R.L.H., Rowntree, P.R. and Smith, J., 1999: The impact of new land surface physics on the GCM simulation of climate and climate sensitivity. Clim. Dyn., 15, 183–203. Deardorff, J. W., 1978: Efficient prediction of ground surface temperature and moisture, with inclusion of a layer of vegetation. J. Geophys. Res., 83 (C4), 1889 –1903. Dickinson, R.E., 1984: Modeling evapotranspiration for three dimensional global climate models. In: J.E. Hansen and T. Takahashi (Eds), Climate Processes and Climate Sensitivity. Am. Geophys. Union, Washington, DC, pp. 58-72. Dickinson, R.E., Henderson-Sellers, A., Kennedy, P. J., and Wilson, M. F., 1986. Biosphere– Atmosphere Transfer Scheme for the NCAR Community Climate Model. NCAR Tech. Rep. NCAR/TN-275 STR, 69 pp. [Available from NCAR, P.O. Box 3000, Boulder, CO 80307-3000.] Mihailovic, D. T., Pielke, R. A. Sr., Rajkovic, B., Lee, T. J., and Jeftic, M., 1993; A resistance representation of schemes for evaporation from bare and partly plant-covered surfaces for use in atmospheric models. J. Appl. Meteor., 32, 1038–1054. Mihailovic, D. T., Rajkovic, B., Lalic, B., and Dekic, Lj., 1995: Schemes for parameterizing evaporation from a non-plant-covered surface and their impact on partitioning the surface energy in land-air exchange parameterization. J. Appl. Meteor., 34, 2462-2475. Mihailovic, D. T., 1996: Description of a land-air parameterization scheme (LAPS). Global Planet. Change, 13, 207-215.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Land-Air Parameterisation Scheme (LAPS)

63

Mihailovic, D. T., and Kallos, G., 1997: A sensitivity study of a coupled soil-vegetation boundary layer scheme for use in atmospheric modeling. Bound.-Layer Meteor., 82, 283315. Mihailovic, D. T., Lee, T. J., Pielke, R. A., Lalic, B., Arsenic, I., Rajkovic, B., and Vidale, P. L, 2000: Comparison of different boundary layer schemes using single point micrometeorological field data. Theor. Appl. Climatol., 67, 135-151. Mihailovic, D. T., 2003: Implementation of Land-Air Parameterization Scheme (LAPS) in a limited area model. Final Report, The New York State Energy Conservation and Development Authority, Albany, NY, 110 pp. Mihailovic, D. T., Alapaty, K., Lalic, B., Arsenic, I., Rajkovic, B., and Malinovic, S., 2004: Turbulent transfer coefficients and calculation of air temperature inside the tall grass canopies in land-atmosphere schemes for environmental modeling. J. Appl. Meteor., 43, 1498-1512. Mihailovic, D. T., 2005. LAPS – land surface scheme for use in crop modelling. Workshop on Introducing Tools for Agricultural Decision-Making under Climate Change Conditions by Connecting Users and Tool-Providers (AGRIDEMA), 21 November – 3 December 2005, Vienna (Austria), http://www.agridema.net, (Invited lecture). Paulson, C.A., 1970: The mathematical representation of wind speed and temperature in the unstable atmospheric surface layer. J. Appl. Meteor., 9, 857-861. Pielke, R. A. Sr., Avissar, R., Raupach, M., Dolman, H., Zeng, X., and Denning, S., 1998: Interactions between the atmosphere and terrestrial ecosystems: Influence on weather and climate. Global Change Biol., 4, 101–115. Pleim, J., and Xiu, A., 1995: Development and testing of a surface flux and planetary boundary layer model for application in mesoscale models. J. Appl. Meteor., 34, 16–32. Sellers, P. J., Mintz, Y., Sud, Y., and Dalcher, A., 1986: A simple biosphere model (SiB) for use within general circulation models. J. Atmos. Sci., 43, 505–531. Stohlgren, T. J., Chase, T. N., Pielke, R. A., Kittel, T. G. F., and Baron, J. S., 1998: Evidence that local land use practices influence regional climate, vegetation, and stream flow patterns in adjacent natural areas. Global Change Biol., 4, 495–504. Viterbo, P., and Beljaars, A., 1995: An improved land-surface parameterization scheme in the ECMWF model and its validation. J. Climate, 8, 2716–2748. Walko, R. L., and Coauthors, 2000: Coupled atmosphere-biophysics-hydrology models for environmental modeling. J. Appl. Meteor., 39, 931-944. Zhang, D., Anthes, R. C., 1982: A high-resolution model of the planetary boundary layer sensitivity tests and comparisons with SESAME-79 data. J. Appl. Meteor., 21, 15941609.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 7

THE ENSEMBLES CLIMATE CHANGE PROJECT Paul van der Linden∗ ENSEMBLES, Hadley Centre, Met Office, United Kingdom

WHAT IS THE ENSEMBLES CLIMATE CHANGE PROJECT?

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Context: Is Climate Changing, If So Why, and Does it Matter? It is now accepted that “warming of the climate system is unequivocal” as seen by the observed rise in global average temperature of 0.76°C from 1850 to 2005 (IPCC 2007a). Changes have also been observed in other components of the climate system (e.g. sea level, glaciers, precipitation) and at different spatial scales (global and regional). The IPCC report also states that “most of the observed increase in global average temperatures since the mid20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations”. Carbon dioxide is the most common of the greenhouse gases and its atmospheric concentration has risen from a pre-industrial level of about 280 ppm to 379 ppm in 2005. Socio-economic projections suggest that the upward trend of carbon dioxide emissions and other greenhouse gases will continue, and thus the atmospheric concentrations of these gases will rise too. The IPCC (IPCC 2007a) projects that due to these greenhouse gases the global average temperature is likely to rise by a further 1.8 to 4.0°C by 2100. Projected impacts from climate change this century include water stress, species extinction, changes in crop productivity, increased storm and flood damage, and increased morbidity and mortality (IPCC 2007b). These impacts will not be spread evenly, but will affect systems, sectors and regions in different ways. All regions of the globe will be affected by climate change impacts including Europe. Many governments are seeking to mitigate and adapt to future projected changes in climate at a national or regional level and through international agreements such as the Kyoto Protocol, part of the United Nations Framework Convention on Climate Change (UNFCCC). This is because of the threat climate change poses to humankind and society. In order to do ∗

[email protected]

66

Paul van der Linden

this governments, their policymakers and society as a whole must have the best available information about future possible climate change, its impacts, the risks they pose, and uncertainties involved when making decisions. The EU set up the ENSEMBLES project to provide this information needed to address the problems caused by climate change.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Overview and Aims of the ENSEMBLES Project The main objective of the ENSEMBLES project is to provide probabilistic estimates of climatic risk through ensemble integrations of Earth system models (ESM). (Hewitt and Griggs, 2004.) The project has developed an ensemble climate forecast system for use across a range of timescales (seasonal, decadal and longer) and spatial scales (global, regional and local). This modelling system is being used to construct scenarios of future climate change to provide a basis for quantitative risk assessment of climate change and climate variability. Emphasis is being placed on changes in extreme events (for example the severity and frequency of heatwaves, drought, forest fires, storminess and flooding), and the effects of high-impact but low-probability events such as a shutdown of the thermohaline circulation in the North Atlantic. Ensembles using multi-model inputs with slightly different starting conditions are used in climate modelling as a way of reducing the uncertainty in the projections. The project validated the ensemble prediction system using high quality gridded data sets for Europe. There has also been quantification and reduction in uncertainty in the representation of physical, chemical, biological, and human-related feedbacks in the Earth system (including water resource, land-use, and air quality issues, and carbon cycle feedbacks); and links the outputs of the ensemble prediction system to a range of applications for different systems and sectors. The 5 year project is funded by the European Commission’s Sixth Framework Programme as an ‘Integrated Project’, and is scheduled for completion in August 2009. It is coordinated by the Hadley Centre at the Met Office in the UK and involves 66 institutes from 19 countries, mainly in Europe. The ENSEMBLES project directly addresses key objectives of the UNFCCC and the IPCC, two of the most important international agencies formulating climate change policy. The relevant objectives of the UNFCCC and IPCC are to: provide the best available scientific information and assessment on climate change and its impacts, to provide input for policy makers concerning the assessment of dangerous anthropogenic interference with the climate system; reduce uncertainties in knowledge of the climate system and the adverse impacts of climate change; promote the development and implementation of education and training programmes; and increase the awareness and public access to information on climate change. The findings of the ENSEMBLES project will therefore be of great practical value to policy makers, stakeholders and the public. The results are being disseminated through web sites and an outreach programme intended to improve the understanding of climate change for wide-ranging audiences, including researchers in areas related to climate change.

Climate Modelling in ENSEMBLES Predicting the future climate is a major challenge due to the complex nature of the Earth system. The only tools for this task are physically-based climate models of the key

The ENSEMBLES Climate Change Project

67

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

components of the Earth system (Figure 1). However, predictions of natural climate variability and the human impact on climate are inherently probabilistic, due to uncertainties in the representation of key processes within models, initial conditions used for the forecasts, and climatic forcing factors such as future concentrations of atmospheric carbon dioxide. Hence, estimates of climatic risk are best made through multiple integrations of models of the Earth system in which the uncertainties are explicitly incorporated by using different representations of processes within a model and different models, slightly varying the initial conditions, and exploring different scenarios of climatic forcing. The ensuing ensemble of results (see Figure 2 for an example) allows quantification of the uncertainty in the climate projections by using statistical techniques across a range of timescales (seasonal, decadal and longer) and spatial scales (global, regional and local). Hindcasts made by the model system for the 20th century were validated against quality controlled, high-resolution gridded datasets for Europe. The model system was used to construct scenarios of future climate change which provided a basis for quantitative risk assessment of climate change and climate variability, and policy relevant information on climate change and its interactions with society. Changes in extreme events (for example the severity and frequency of heatwaves, drought, forest fires, storminess and flooding), and the effects of high-impact but low-probability events such as a shutdown of the thermohaline circulation in the North Atlantic have also been examined. The outputs of the ensemble prediction system are used to drive a wide range of applications, including agriculture, health, energy and water resources. In turn, feedbacks to the climate system from some of these impact areas will also be addressed.

Figure 1. A schematic example of a climate model.

68

Paul van der Linden

Figure 2. An example of the predicted change in summer-average precipitation over Europe from an ensemble of model simulations.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Organisation and Structure of ENSEMBLES The project intends to meet its main objectives by use of ten interlinked Research Themes (RT). Each RT is composed of several Work Packages (WP) which enabling the milestones and deliverables to be achieved within each RT. RT0 provides the overall coordination and management of the project, the research activities are carried out under RT1 to RT7, and the dissemination, education and training activities are carried out under RT8. A diagram showing the structure of the Research Themes and the linkages between them is given in Figure 3 along with a summary of each RTs activities: At the core of the ENSEMBLES Integrated Project is the development of the first global, high resolution, fully comprehensive, ensemble based, modelling system for the prediction of climate change and its impacts. In order to do this the first step was to assemble currently available ESM component modules to provide models for use in the ensemble prediction

The ENSEMBLES Climate Change Project

69

system. The resulting ESMs were then combined into a multi-model ensemble system, with common output. This system was then initialised and pre-production runs at seasonal to decadal and longer timescales performed and evaluated. This work is being carried out in RT1.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Diagram showing the organisation, main activity and relationship between the Research Themes within the project.

The purpose of RT2A is to produce sets of climate simulations with several models and to provide the multi-model results needed in other Research Themes. In the first 2 years of the project the simulations were performed using existing atmosphere-ocean-sea ice models; as they provide state-of-the-art benchmark multi-model simulations. After the completion of the development and building of more comprehensive ESMs, and their pre-validation as provided by RT1, other model components were gradually introduced in the simulations to investigate their impact on climate predictions. The results from are used for validation (RT5), studies of feedbacks in the Earth system (RT4), as well as boundary conditions and forcing fields for regional model simulations/predictions (RT3/RT2B). The simulations cover timescales ranging from seasons, to decades and centuries. RT3 has the responsibility for providing improved climate model tools developed in the context of regional models, first at spatial scales of 50 km at a European-wide scale and later also at as high a resolution as 25km for specified sub-regions. Analogous to RT1, and using boundary conditions from RT2A, RT3 has produced a multi-model ensemble based system for regional climate prediction from decadal to centennial timescales to be applied in RT2B. Along with RT2A, RT2B provides the “model engine” of the ENSEMBLES project. RT2B is constructing probabilistic high-resolution regional climate projections and seasonalto-decadal hindcasts using dynamical and statistical downscaling methods in order to add

70

Paul van der Linden

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

value to the ESM output from RT1 and RT2A and to exploit the full potential of the Regional Climate Models (RCMs) developed in RT3. Output is in formats which are appropriate for input to the RT6 assessments of the impacts of climate change as well as for more general end users and stakeholders. The purpose of RT4 is to advance understanding of the basic science issues at the heart of the ENSEMBLES project. Using the outputs of RT2A and RT2B, the work focuses on the elucidation of the key processes that govern climate variability and change, and that determine the predictability of climate on timescales of seasons, decades and beyond. Particular attention is given to understanding linear and non-linear feedbacks in the climate system that may lead to climate “surprises”, and to understanding the factors that govern the probability of extreme events. The improved scientific knowledge gained in RT4 is fed back into further development of the models used in RT1 and RT3. The development of the ensemble based prediction system and the production of probabilistic high-resolution regional climate scenarios will be of uncertain value without rigorous evaluation. Hence, RT5 is conducting an extensive and independent evaluation of the performance of the ENSEMBLES prediction system developed in RT1 and RT3 and run through the model engines of RT2A and RT2B, against analyses/observations. This includes the production of the high-resolution observational datasets necessary to perform this task. The results of this evaluation is also being fed back into further development of the models used in RT1 and RT3. RT6 uses the output from the ensemble based prediction system, developed in RT1 and RT3 and run through the model engines of RT2A and RT2B, to carry out impact assessment within ENSEMBLES. Its primary objective is to simulate the potential impacts of future climate change during the 21st century on natural systems and human activities at different scales under alternative scenarios of future climate. This will include, for example, the integration of process models of impacts on the natural and managed global environment into ESMs, the results from which will be fed back into the model development in RT1. However, the main output from RT6 is to the wider public and stakeholder community through RT8. RT7 provided RT1 with ensembles of emissions and land-use scenarios with and without mitigation policies, as well as scenarios of adaptive capacity. This is based on existing scenarios and run with existing models. However, the main aim of RT7 is to integrate the human dimension into ESMs. As with RT6, the main output from RT7, besides scientific output, is to the wider public and stakeholder community through RT8.

THE RELEVANCE OF THE ENSEMBLES PROJECT TO AGRICULTURAL RESEARCH Aspects of Climate Change Modelling within ENSEMBLES of Interest to Agricultural Researchers There are many aspects of the ENSEMBLES project of interest to agricultural researchers and crop modellers, some of these derive from the integrated nature of the project (e.g. the probabilistic projections) while others are of a more singular nature e.g., downscaling model outputs.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The ENSEMBLES Climate Change Project

71

Firstly, the ENSEMBLES project is the first to provide probabilistic estimates of climatic risk through ensemble integrations of Earth system models. The ensemble climate forecast system is configured across a range of timescales (seasonal, decadal to centennial) and spatial scales (global, regional and local) and has constructed scenarios of future climate change to provide a basis for quantitative risk assessment of climate change, climate variability and changes in extreme events (for example the severity and frequency of heatwaves, drought and flooding). The project will also use a crop model to produce a projection ensemble (see section 2.4 for detail). Another part of ENSEMBLES of interest to agricultural researchers and crop modellers is the integration of process models of impacts on the natural and managed global environment into the Earth System Models used in the project. Existing ecosystem, crop, and hydrological models are being used in both offline and online mode. In online mode the impact models are integrated into the Earth System Models of ENSEMBLES, so that cause and effect are coupled, and the impacts of climate change feedback to the atmosphere and climate. Within the project is an independent and extensive evaluation of the ENSEMBLES simulation-prediction system against observations/analyses, including the production of a high-resolution observational dataset necessary to perform this task. The evaluation is fully independent and will consider: processes and phenomena, forecast quality, extreme events in observational and RCM data, and impact models when forced with downscaled ERA-40, hindcasts and gridded observational datasets. The analysis on the European region will focus in particular on extreme events both in high-resolution simulations and in the observations. All the output from this part of the project should be of interest to crop modellers. The way in which uncertainty is being treated holistically across the project is of significance to impact modellers using climate projection data output. In practical terms this means that the uncertainties arising from model assumptions and parameterizations are better described because of the multi-model nature of the ensembles. At the RCM level, different GCM forcings are used to give the fullest description of uncertainty to the output. When RCM data sources are used to drive impacts models then bootstrapping is used to estimate uncertainty in the RCM. This means that the uncertainty within the impact model, although dependant on the underlying RCM used, can be quantified. Some impact modelling groups within the project will consider changes across Europe as a whole, while others will focus on case-study regions and specific impact sectors such as agriculture or forestry, or specific events, such as deep cyclones, heavy rainfall and wind storms, or a combination of both, such as crop production and drought risk. Characterisation of climate and weather extremes (for example: rainfall intensity, temperature, wind) are found within the GCM and RCM output data, and further data can be derived from this output (for example: drought, soil moisture, frost days) See Appendix 1 for a list of common climate model variables. This data was produced from the many climate model runs of the different scenarios of projected future climate. Extremes data has also been produced for historic climates (20th century) for verification purposes. Further, a website has been developed for the project where climate data can be produced ‘on demand’ by statistical downscaling. This statistical downscaling data portal was developed by two partners, the University of Cantabria and the EC Joint Research Centre, working on crop yield forecasting. Users and stakeholders were consulted about their needs during the development of this system. These publicly available tools and methods provide a valuable resource for the wider climate, seasonal forecasting and impacts communities. The

72

Paul van der Linden

development of these tools is seen as desirable because they give users the potential to construct their own tailored outputs rather than being reliant on the scenario developers. This is increasingly important, given the growing demand for scenario projections for many diverse impacts applications. Further information about the statistical downscaling data portal can be found in Section 6.

Global Climate Models Eleven European climate centres have contributed GCM experiments to the ENSEMBLES project produced under a range of different scenarios. Simulations include perturbed parameters and also historical forcings. The data produced are available through the project data archive. See Appendix 1 for a summary of all the model runs and their attributes which will be available on completion of the project.

Regional Climate Models

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Ten European climate centres have contributed RCM output to the project under a range of different scenarios and model forcings, the simulations also include historical forcings. See Appendix 1 for a summary of all the model runs and their attributes which will be available on completion of the project. Figure 4 shows the area for the regional modelling.

Figure 4. Common domain for regional climate model simulations.

The ENSEMBLES Climate Change Project

73

The RCM simulations for this common domain either 25 or 50km resolution depending on the realisation, both resolutions taking boundary conditions from the ERA40 (a 40-year European Re-Analysis covering the period mid-1957 to present) dataset. The purpose of these simulations is to assess regional model performance on interannual and seasonal time scales over several decades. The multi-decade period is important in order to simulate climate variability and extremes, as well as longer-term climate variability. The climate centres have also produced transient projections of climate out to 2100 under different scenarios (A2 and B2), again using either 25 or 50km resolution.

Crop Models in ENSEMBLES The ENSEMBLES project is developing and evaluating impact models (e.g. water, crops, health) and integrating them within the prediction system. Impact models are being coupled to probabilistic projections of climate change to develop probabilistic projections of future impacts. There is also the aim of maximizing skill in the impacts models driven by seasonalto-decadal scale forecasting. Application models for a range of sectors will be driven using GCM and RCM output to make predictions at seasonal-to-decadal time scales. The application models can produce probabilistic predictions on seasonal-to-decadal time scales at regional scales. A number of the impact models in the project are specifically crop models, which have either been developed or enhanced for use within the project. Table 1 is a list of the crop models and their application within the project.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. List of the crop models used within the ENSEMBLES project and their main area of application Crop Model GLAM

Institute University of Reading

Contact Tim Wheeler

WOFOST

Agenzia Regionale per la Prevenzione e l’Ambiente dell’Emilia-Romagna, Servizio Idro Meteorologico ARPA-SIM Joint Research Centre of the European Community – Institute for the protection and Security of the Citizen JRC-IPSC

Vittorio Marletto

WOFOST; CGMS

Bettina Baruth

Application within ENSEMBLES Evaluation runs using ERA-40 input. Test runs with GLAM integrated within METO-HC model. Assessment of seasonal predictability of crop yield using ENSEMBLES seasonal hindcasts. Offline assessments of sensitivity to key thresholds using output from ENSEMBLES scenarios. Ensembles of tropical crop yield; cropclimate modelling. Ensembles of regional crop models – northern Italy

Ensembles of crop models – Europe

74

Paul van der Linden Table 1. (Continued) Crop Model Crop suitability; N-leaching DAISY; N-leaching model CERES

Institute Finish Environment Institute SYKE University of Aarhus

LPJ

Department of Agronomy and Land Management, University of Florence DISAT PIK

TRIFFID

Met Office

ORCHIDÉE

CNRS-IPSL

Contact Timothy Carter Jørgen Olesen Marco Bindi

Application within ENSEMBLES Temperate crops/N/Soil C

Wolfgang Cramer Richard Betts Nathalie de NobletDucoudre

Dynamic global vegetation modelling Dynamic global vegetation modelling Dynamic global vegetation modelling

Temperate crops/N/Soil C Mediterranean crops

Climate Datasets and Variables

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The defining parameters (time, scenario, resolution, etc) of the different model runs produced by the different modelling centres are summarised in Tables A.1 to A.5 of Appendix 1. Information about how to access the datasets from these experiments is also given. When a GCM or RCM is run to simulate the physical and chemical processes of the Earth-atmosphere system it employs many variables (typically in excess of 100), which, for key parameters at each model grid point are stored for later analysis. Most of these variables are common between models and a list of the most common for is shown in Tables A.6 of Appendix 1. Subsequent, further variables can also be derived from the set produced by the model for users with specific needs, such as soil parameters for crop modellers. One data RCM simulation plan within the project which is outside the Europe region (the domain for all the other RCM simulations in the project) is the West African domain. This is to support an EU Framework Programme 6 project called AMMA (African Monsoon Multidisciplinary Analysis).

CROP MODELS IN THE ENSEMBLES PROJECT: RESEARCH ACTIVITIES AND RESULT At the time of writing (April 2008) the ENSEMBLES project still has 18 months left to run out of a total of 60. Much of the data and results produced by the whole project so far come from the climate modelling side. The climate modelling results feed into the impacts modelling part of the project, thus many of the results of the impact models and crop models have yet to be seen. The progress made so far with the crop models is in their construction, linking with climate models and calibration.

The ENSEMBLES Climate Change Project

75

Crop Modelling Activities within ENSEMBLES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The work structure of ENSEMBLES is one of objectives and tasks which are tied to Work Packages involving relevant institutes. Performance of Work Packages (WP) is judged against pre-defined standards for Deliverables and Milestones. Some of the work is devoted to running the impact models, whilst some is spend on testing, calibrating and validating the impacts models. Measurement of skill within a number of impact models is being carried out by running the models firstly with appropriately downscaled ERA-40 data and gridded datasets and secondly with fully downscaled and bias corrected probabilistic ensemble hindcasts at seasonal-to-decadal scales. The models include agriculture models (crop yield models for Europe and the tropics, agri-environmental impact models)

Global Changes in Biophysical and Biogeochemical Processes – Integrated Analysis of Impacts and Feedbacks [WP6.1] The work will produce, for the first time, fully integrated European- and global-scale assessments of the impacts of changes in CO2 and climate on vegetation structure, function and productivity, forest and arable crop productivity, terrestrial carbon cycling and freshwater supply and will also consider the potential for feedbacks from these changes to the atmosphere and climate. The work will have two main strands. The first strand will be based on offline simulation using historical climate observations, and externally provided climate change scenarios for the future, to drive integrated models of terrestrial biosphere processes. The second strand will be based on online simulation (hindcasts and future projections), with the impacts incorporated into the ensemble of ESMs being constructed in RT1. The state of the art in modelling the impacts of climate change on ecosystem services in Europe is represented by ATEAM (FP5), which is producing “vulnerability maps” for crop production, forestry, biodiversity and freshwater supply, based on a range of scenarios and climate models (but without accounting for feedbacks). The main impacts model used by ATEAM is the Lund-Potsdam-Jena model (LPJ), which is a leading dynamic vegetation model under continuous development (currently co-ordinated by the University of Bristol). Crops and forest management are represented for European applications. The state of the art in modelling biosphere-atmosphere feedbacks is represented by recent work at the Met Office’s Hadley Centre and CNRS-IPSL. These two groups have pioneered the development of fully coupled, global models of the carbon cycle and climate and have established that the feedback from terrestrial carbon cycling to the rate of increase of atmospheric CO2 has the potential to amplify climate change over the next century, suggesting that it will be important to include biosphere processes in climate models generally. Work at Met Office’s Hadley Centre has also helped to quantify the important biophysical consequences of land use. ENSEMBLES will merge the best of these two research approaches, with a view to producing better-founded, assessments of the impacts of climate change on ecosystems processes at European and global scales, and better-founded projections of climate change taking into account the effects of landuse change and carbon and nitrogen cycle effects. Measurable outputs of WP6.1: Globally applicable offline representations of managed forests and crops (top 15-20 crop types).

76

Paul van der Linden

A comparative assessment of the “offline” and “online” crop models when driven by climatological data.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Linking Impact Models to Probabilistic Projections of Climate Change [WP6.2] This Work Package, focuses on crops, water resources, forests, energy supply, and human health, and seeks to extend previous analyses (see below). The output from the impacts analyses in WP6.2 will be delivered to RT7 for development of scenarios. Impacts of climate change will be represented using response surfaces, based on multiple model simulations across a wide range of plausible future climates, which depict a measure of impact against key climatic variables (e.g., see Figure 5a).

Figure 5. (a) Response surface for spring wheat yields (percent relative to present) to changes in temperature and CO2 concentration in southern Finland, and (b) risk of impact in 2050, based on hypothetical probabilistic climate/CO2 scenarios.

The ENSEMBLES Climate Change Project

77

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Impacts to be studied in this way include: soil moisture, soil temperature, crop productivity, nitrogen use efficiency, nitrogen leaching, soil carbon storage, stream discharge, and water availability. The future climates to be considered will be guided by the range of climate projections for Europe in ENSEMBLES, including the E1 (“450 ppm”) stabilisation scenario. Threshold levels of impact will be defined either based on historical impacts, or according to established operational conditions (e.g. minimum stream discharge for hydroelectric production). They will be selected to illustrate possible levels of tolerance to climate change, the exceedence of which may be regarded as unacceptable by decision makers (Article 2 of the UN Framework Convention on Climate Change). Thresholds will be sector-, system- and region-specific. Each climate impact projection can be located in “climate space” on the impact response surfaces, enabling the following issues to be analysed: (i) the significance of modelled impacts under changing climate in relation to impacts estimated for natural variability alone; (ii) the risk of exceedence of impact thresholds at different times in the future with unmitigated emissions (e.g. see Figure 5b); (iii) the reduction in risk of impact threshold exceedence under stabilisation scenarios; (iv) the levels of impact to be regarded as “unavoidable” in Europe under all estimates of climate change, and for which adaptation will presumably be required; and (v) the types of impact that might be expected in Europe under abrupt, non-linear climate changes. Work is also being done to develop and analyse appropriate methods for evaluating the impacts of changes in the frequency, magnitude and distribution in time of extreme weather events under a changing climate. In northern Europe the extremes of greatest concern will be wind storm and flood, and their impacts on human health and safety, property and forestry. Over southern Europe the emphasis will be on drought and heat stress, and the relationship to forest fire, agriculture, water resources and human health (see Table 2). The Daisy model will be used to analyse sensitivity of crop production, nitrogen use efficiency, nitrogen leaching and soil carbon storage to changes in temperature, rainfall and atmospheric CO2 concentration. The sensitivities will be used to develop response surfaces for the selected indicators. The analysis will be performed for arable crop rotations at selected sites along a North-South and East-West gradient representing catchments in Northern Europe.

Impact Modelling at Seasonal-to-Decadal Time Scales [WP6.3] The current state of the art for quasi-operational impact models running for the forthcoming season is that they make use of the observed climate through the first part of the season or through the preceding season. As an example wheat yield predictions in Europe are made using a dynamical crop growth model which is run to a given date using climate observations and then an end of season yield prediction is made through a statistical model relating crop development phase, at that date, to previous known wheat yields. In this application the lead-time is limited and the prediction become more reliable the later in the season they are produced. Probabilistic information is routinely used in weather risk management models. However this is mostly confined to distributions from climatological records, with occasional input from the mean of seasonal forecasts. The tropical yield crop model GLAM has successfully been used to simulate yields over large

78

Paul van der Linden

areas in India using observed gridded data and reanalysis. Some preliminary work has also been undertaken using the DEMETER ensembles for the period 1987-98. Table 2. Impact models, sectors and scales of analysis to be employed by partners for further details see Appendix 1 Local University of Aarhus Department of Agronomy and Land Management, University of Florence, DISAT Finnish Meteorological Institute National Observatory of Athens Polish Academy of Sciences Swedish Meteorological and Hydrological Institute Finish Environment Institute, SYKE University of East Anglia

Temperate crops/N/Soil C Mediterranean crops

Catchment/ national Temperate crops/N/Soil C Mediterranean crops

Soil water

Wind energy

Europewide

Mediterranean crops; forest fire

Forest damage Forest fire, heat stress Regional flooding

Runoff/stream discharge Temperate crops

Crops/N/ Soil C Human health

Free University of Berlin University of Lund

Kassel University

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CGAM, University of Reading

Extremes

Windstorm; heat stress; arrival patterns of extremes Windstorm Ecosystem damage; flood damage

Water availability/ water quality Crop-climate modelling

During DEMETER (an EU Framework Programme 5 project) a small number of application models were developed or modified to use the seasonal probabilistic forecasts to drive, daily time step, impact model integrations. In DEMETER the range of impact models run, the regions covered and number of ensemble members utilised was limited. Questions that arose during DEMETER regarding downscaling, bias correction and how to interpret the probabilistic impact model output were only partially addressed and these will be more fully investigated within ENSEMBLES. In ENSEMBLES, a number of new impact models will run addressing a much larger potential user community. ENSEMBLES will attempt to answer a number of novel research questions i. How to maximise the integrated model system skill using techniques such as ensemble dressing, ii How to quantify the existing skill within an integrated modelling system, iii. How to estimate the skill required from the driving seasonal forecasts to allow a skilful seasonal impact forecasts.

The ENSEMBLES Climate Change Project

79

The integration of the impacts models within the ENSEMBLES EPS (Ensemble Prediction System) will lead to a number of measurable outputs including as examples. Probabilistic estimates of wheat yield for Germany, Spain, France and Hungary will be driven using ENSEMBLES seasonal hindcasts and will be validated against EUROSTAT crop yield data and compared with the forecasts from the current state of the art MARS Crop Yield Forecasting System. The General Large-Area Model for annual crops (GLAM) will be used to simulate the yield of major legume and cereal crops in India. Observed district-level yields will be used to assess the skill of the model output and examine the issue of spatial scale of predictions. Groundnut will be the first crop to be studied as the district-level yields for this crop are already available. Efforts will also be made to assemble datasets for some of the major tropical cereals.

Crop Modelling Results from ENSEMBLES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

This section contains extracts from internal ENSEMBLES reports (“Deliverables”) describing results to date from crop modelling and supporting activities within the project.

GLAM Tropical Crop Productivity Model [Deliverable 6.2] The GLAM (General Large Area Model for annual crops) model studies of tropical crops provide an important extension to the general ENSEMBLES domain of Europe. These highly detailed experiments examine the crucial influence of short term temperature extremes on the productivity of groundnut, wheat, and rice. There is increasing evidence from crop experiments that short-term climate events of only a few days duration can severely impact crop productivity if they coincide with a sensitive phase of crop growth. One example is the occurrence of high temperatures near to the time of crop flowering (Wheeler et al., 2000, Figure 6). It is now thought that how crops respond when these climate thresholds are exceeded will be a vital part of the impact of climate change on crop productivity in some regions. The response of groundnut, wheat and rice crops to short-term high temperature events has been quantified, and also how this response differs among varieties of groundnut and rice, in controlled environment experiments. The modelling of the impacts of high temperature extremes can now be developed using these observations. A full description of how the impact of high temperature extremes on groundnut crops are simulated in GLAM-HTS is given in Challinor et al. (2007). In brief, the first stage of the simulation of high temperature stress is the identification of episodes of high temperature. This is done by comparing the mean 8am to 2pm (solar time) temperature to a pre-defined critical value and to the development stage of the crop (pre- and post-anthesis). The critical temperature above which pod-set begins to be affected, and the temperature at zero pod-set, are defined and pod-set at intermediate temperatures is determined by linear interpolation. The reduction in the total pod-set is then given for each high temperature episode as a sum of the impact of that episode on each of the days during the flowering, where a flowering distribution prescribes the fraction of total flowers opening on each day. The fraction of yielddetermining pods from the episode with the greatest impact is then used to modify the rate of change of harvest index, and hence crop yield.

80

Paul van der Linden

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 6. The effect of a 1 day high temperature event on the fruit / seed set of groundnut plants grown in controlled environments.

Figure 7. Observed and simulated (a) yields and (b) pod-set as a fraction of the control (no high temperature) plants. The x-axis indicates the timing of the start of the six-day high temperature episode. Simulations using six flowering distributions are shown by the lines, and the observations shown by the symbols. Model simulations were compared with crop responses to high temperature episodes imposed using a range of controlled environment experiments (Challinor et al., 2007). For example, Prasad et al. (1999) reported a significant impact of high temperature on pod-set and pod weight of groundnut. These results are compared with GLAM-HTS run with a range of flowering distributions in Figure 7. The choice of flowering distribution in the model affected the simulations. The impact of the timing of the high temperature episode relative to anthesis on pod-set and yield was captured by the model. Most of the absolute difference between simulated and observed yields were less than 10% of the control yield. Skilful simulation of the effects of short-term temperature extremes also relies on good prediction by the crop model of the effects of mean temperature on the rate of crop development. For example, good simulations of the impacts of temperature extremes at sensitive stages of a crop will be of limited value for prediction if the timing of crop development is not also well simulated. Figure 8 shows the simulated and observed impact of seasonal mean temperature on groundnut crop duration. GLAM-HTS was capable of reproducing the observed impacts of mean temperature on crop duration. The mean simulated duration (across the three simulations) is within 5% of observed values at all three temperatures. The mean simulated change in duration from 20°C to either 24 or 28°C is within 8% of the observed mean change (that is, 5 days). As conclusion, crop response to high temperature extremes has now been incorporated into GLAM to give a GLAM-HTS (high temperature stress) model version. This opens up opportunities to examine how short-term, inter-seasonal variability in temperature will affect crop productivity in current and future climates.

The ENSEMBLES Climate Change Project

81

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 7. Observed and simulated (a) yields and (b) pod-set as a fraction of the control (no high temperature) plants. The x-axis indicates the timing of the start of the six-day high temperature episode. Simulations using six flowering distributions are shown by the lines, and the observations shown by the symbols.

Figure 8. The simulated and observed impact of seasonal mean temperature on total crop duration. Observations are from Nigam et al. (1994).

82

Paul van der Linden

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 9. Use of DEMETER multi-model ensemble for groundnut yield in Gujarat, 1998, from Challinor et al. (2005).

Seasonal-to-decadal Application Models Running as Part of an Integrated Probabilistic ESM Based on DEMETER Hindcasts [Deliverable 6.4] The main purpose of this deliverable was to develop and demonstrate the ability of the seasonal to decadal application groups to run their application models and analysis methodologies with existing FP5 DEMETER EPS seasonal hindcasts. This has allowed partners to build up expertise in running multi-model ensembles and it has acted as a ‘dry run’ to the forthcoming ENSEMBLES seasonal to decadal hindcasts. Further it has started to identify important scientific questions that need to be addressed when integrating application models within seasonal ensemble predictions systems which will be addressed through ongoing cooperative and across research themes research and development activities within ENSEMBLES. In fact the work was much more than a ‘dry run’ and it has begun to produce interesting results from the application runs with the DEMETER hindcasts. This has resulted in a number of publications and presentations, attributed fully or in part to the activities within ENSEMBLES leading to this deliverable. In addition some of the partners have started to work evaluate issues that arise particularly from uncalibrated EPS output. The EU Joint Research Centre (JRC) prepared the necessary framework to run the European crop yield model (Crop Growth Monitoring System) using the bias corrected probabilistic seasonal-to-decadal hindcasts as issued by DEMETER. The crop yield model had to be re-adapted to link to the different sources of data especially taking into account the probabilistic nature of the ensembles input data and in order to produce ensemble crop yield forecasts. Crop yield forecasts were run based on the different data sets after which crossvalidation with the existing output from the crop yield model and with the actual crop yield data was carried out to assess the skill of the ensemble based crop yield forecasts.

The ENSEMBLES Climate Change Project

83

The JRC crop yield model uses individual ensemble members, and the probabilistic crop yield forecasts are created by equally weighting the forecast produced by each ensemble member. The University of Reading model uses the average of the ensemble members and the ensemble spread. The ensemble mean is used to make deterministic estimates of crop yield while the ensemble spread provides information on the probability of crop failure and is also used to construct probability density functions (PDFs) of crop yield. The PDF in Figure 9 was generated by researchers at the University of Reading using DEMETER data. Further information about these results can be seen in Challinor et al. (2005). Agenzia Regionale per la Prevenzione e l’Ambiente dell’Emilia-Romagna, Servizio Idro Meteorologico (ARPA-SIM) has compared the median of the simulated yields with observed yields using the coefficient of determination and the RMSE and bias, see Figure 10. ARPA-SIM is also designing a new set of scores based on the width of the forecast distribution and the proximity of the simulated median to observations. 12000

10000

8000

6000

k /h

4000

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2000

0 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Year anno (1985 is not reported due to severe frost damage) Distribuzione (5,10,25,50,75,50,90,95 %)%) delle di frumento Distribution (5, 10, 25, 50, 75, 90, 95 of rese simulated wheattenero yields(kg//ha) (kg/ha) simulate utilizzando la data serie meteorologica della stazione di Cadriano using local weather (Cadriano, Bologna, Italy) from October to April eand le previsioni stagionali DEMETER (riportate ad June una griglia di 15 X15 m) downscaled Demeter output for May and per i mesi di maggio e giugno.

Rese osservate Observed yields Rese simulate utilizzando la serie meteorologica Simulated yields using local weather data onlydella stazione di Cadriano Figure 10. Comparison of the median of simulated yields with observed yields using downscaled DEMETER output.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

84

Paul van der Linden

Preliminary report on a comparative study of response surface and multiple scenario approaches to assessing risks of impacts using selected impact models [Deliverable 6.7] Two data sources were available for this work: the perturbed parameter simulations of the HadCM3 (Collins et al. 2006; Murphy et al. 2004) and a set of scenarios constructed by resampling an ensemble of AOGCM simulations prepared for the Fourth Assessment Report of the IPCC (Räisänen and Ruokolainen 2006). A data set based on the perturbed physics experiment of the QUMP project was provided by the Hadley Centre. Simulations were performed for a limited ensemble of variants of the HadCM3 model with perturbed input of 31 uncertain parameters. Based on these simulations, an emulator was built which estimated the output at values for 102 grid cells in Europe. For all grid cells, changes in annual temperature and annual precipitation for ten 20 year periods in the 21st century were made available. It was then possible to apply the sensitivity analysis and response surface methods developed by Jones (2000). (See also WP6.2 in Section 3.1 for information about the planned use of response surfaces in this area of the project.). A preliminary analysis of the climate change impact on winter wheat yield and nitrogen leaching was performed using the Danish soil-plant-atmosphere model Daisy. The focus was on developing a method to construct response surfaces and combine them with probabilistic information about future climate. The dynamic model, Daisy, simulates crop growth and water and nitrogen dynamics in soil and plants. The driving variables are air temperature, global radiation, precipitation, vapour pressure, and wind speed. The model was used for initial sensitivity analyses and construction of preliminary impact response surfaces. Model simulations were made with continuously grown winter wheat and the fertilisation rate was calculated from the simulated maximum yield and a regression between maximum yield and optimum N rate. The regression was found from simulations with five soil types, nine European climates, and five N rates (50, 100, 150, 200, and 250 kg N/ha). In this preliminary work two pilot areas were chosen: Denmark and South Germany. The same soil type (sandy loam, 14% clay in top soil) was applied for the two areas. Baseline climate data from the MARS/STAT database from Joint Research Centre (JRC), 1976-2004 were used as driving variables for the model. For the future climate, perturbed physics climate data from 2050-2070 from climate grids from Denmark and South Germany were used as examples. In the Daisy simulations the temperature was modified with steps of 1C from 0 to +6C for both sites. Precipitation was modified with steps of 5%. For Denmark the precipitation was changed from 10% to +25% and for south Germany from 15% to +15%. The changes in temperature and precipitation correspond to the changes found in climate grids covering the two sites for 2050-2070 in the perturbed physics data. The probability of exceeding a threshold of 20% reduction in yield compared to baseline was calculated, as was the probability of exceeding a threshold of nitrogen leaching larger than 25 kg N/ha was calculated. Also seasonal variations were included in the climate data. Simulations with the Daisy model results in response surfaces showing the probability of exceeding a threshold value of 20% decrease in yields compared to baseline as a function of changes in mean yearly temperature and precipitation. The simulations were performed with uniform and seasonal variations in precipitation change for the two sites. The yields are

The ENSEMBLES Climate Change Project

85

affected by the change in temperature but the effect of precipitation is small. Using a uniform or seasonal variation in precipitation change resulted in only minor differences.

Yield Modelling of Mediterranean Crops [Deliverable 6.7] Please see Appendix 2 where the research results to this part of the project are given in full as a case study. Changes in Climate Extremes and Their Relation to Agriculture [Deliverable 6.8] This work was conducted early on in the project primarily as a way of demonstrating the approach to be taken. It uses a single climate model, HadCM3, under an A2 scenario, it does not use a multi-model ensemble. The impacts assessments part of the project which use regional climate multi-model ensemble will use the approach demonstrated here. For brevity only crop based results are shown. Presented here are a series of simple models appropriate for studying impacts of climate extremes. The premise behind each model is that particular sectors of agriculture and farming are vulnerable to climate thresholds, that is, above (or below) a given threshold of temperature, say, a damaging response can be expected in livestock or crops. As might be expected, the thresholds are normally rarely exceeded and can be considered extremes. Note that the data are normally applied to countries. This type of model has several advantages over more “realistic” simulations of impacts: • • • • •

It is very easily adjusted to different areas It is easily understood by non-specialists The thresholds are well-documented in the literature It lends itself naturally to consideration of climate extremes It does not depend on highly detailed information from climate models

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Here we present a few examples to illustrate the method. Note that although the examples focus on the Mediterranean, we have data to apply the models to the whole of Europe (where relevant, i.e. no olive production in Norway!). Information about agricultural labour, irrigation, exports and meat production is used to assess the importance of agriculture to the areas being considered.

Crops Olive Oil Figure 11 shows olive oil exports (1000 tonnes) 1961-2005 for the major producers. All countries show an enormous growth in trade over the period. This is very variable year-onyear, everywhere apart from Italy. Olives are vulnerable to variations in available water with the dependency characterized as follows: • •

Annual rain 400-600 mm gives average yield without irrigation Annual rain 600-800 mm gives high yield without irrigation

86

Paul van der Linden • • •

Even wet winters will have dried out by August-September, so some irrigation may be needed to maintain good yields Where full irrigation is needed, this is throughout the year apart from May-July Harvesting for oil requires less water than harvesting for fruit.

60 N

50 N

500 700

500

40 N

400 350

Spain

400

450

450

650

600

120

Italy

300 250

250

200 300 200 100 0 1960

100

100

50

1965

1970

1975

1980

1985

1990

1995

2000

1965

Greece

150

10

0 1960

1975

1980

1985

1990

1995

2000

2005 100

2005

20

10

0 1960

0

60

40

1970

50

30 N 10 W

Turkey

80

200

150

100

100

240

10 E

1965

0 1960 1970

1975

20 E

1980

1985

1990

1995

2000

20 1965

1970

1975

1980

1985

1990

1995

2000

30 E

40 E

Figure 11. Olive oil exports (1000 tonnes) 1961-2005.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2005

2005

Figure 12. Annual total rainfall (mm) averaged over 20-year periods (HadCM3 A2a scenario).

The ENSEMBLES Climate Change Project

87

To give a preliminary idea of projected changes in rainfall, Figure 15 shows likely changes in rainfall over the region from HadCM3. Apart from Iberia, no major changes in rainfall over olive-growing areas are indicated. However, the final olive oil model needs a much more detailed analysis using the much higher resolution rainfall data coming from the ENSEMBLES models.

Citrus Fruit Exports of citrus fruit (tonnes) for 1961-2004 are shown in Figure 13. Algeria and Italy show declining citrus production over the period, all other countries have an increase. The climate limitations on citrus are: • o o • o • o

Temperature Best growing season when mean daily temperature between 28-30° C mean daily temperature >38° C: shrivels fruit Water Between 900-1200 mm/year Soil temperature Root growth only when soil temp>12° C

60 N

50 N

120

5

1 x 10

100

80

France

60

40

20

0 1960

0

1970

1975

1980

1985

1995

2000

2005

500

Spain

300

1500

200

500 1960

5

100 1960

10 x 10 1965

1970

1975

Italy

400

2000

1000

5

6 x 10

600

35 x 10

2500

1980

700

700

600

5

2 x 10

1990

1995

2000

1965

1970

1975

1980

1985

2005

1990

1995

2000

500

Greece

400

5

4 x 10

2005

100

250

2.8 x 10

200

0 1960

400

Turkey

300

200

5

5

6.8 x 10

600

500

300

1985

300

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1990

5

3000

40 N

700

1965

3500

200

5

100

1 x 10 1965

1970

1975

1980

1985

1990

1995

2000

2005

0 1960

1965

1970

0

1975

1980

1985

1990

1995

2000

2005

Algeria

150

100

50

0 1960

30 N 10 W

1965

0

1970

1975

1980

1985

0

1990

1995

2000

2005

10 E

20 E

30 E

40 E

Figure 13. Exports of citrus fruits (tones) 1961-2004.

The HadCM3 data are not suitable for detailed examination of these models, and will not discriminate accurately mean daily temperatures greater than 38° C. However, using a threshold of daily maximum temperature >38° C, we arrive at Figure 14. Figure 14 suggests that North Africa, the Middle East, Turkey, and southern Spain/Portugal will not be suitable places for growing citrus by the end of the century.

88

Paul van der Linden

Figure 14. HadCM3 (A2a) annual total number of days with Tmax>38°.

Livestock Chickens Figure 15 shows annual chicken production 1961-2005. In all countries except Italy, the number of chickens increased over the period. Note that, to make the time series clearer, the scaling is not the same on all plots. 60 N

5

50 N

2.6

x 10

8

2.4 x 10

2.4 2.2 2 1.8

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1.6

8

1.6 x 10

France

5

1.4

1.7

1.2 1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

4

13

x 10

1.5

7

13 x 10

12 11

x 10

1.6

Italy

8

1.4 x 10

1.4

5

1.3

3

10

40 N

Spain

8 7

1 0.9 1960

5 4

8

3 x 10

2.5

1.1

6

3 1960

x 10

1.2

9

4

1965

1970

1975

1980

7

3 x 10

1965

1970

1975

3.2

8

0.9 x 10

x 10

1985

1990

1995

2000

2005

2.8

1985

1990

1995

2000

2.2

x 10

7

1.6 1.4 1960

8

0.3 x 10

0 1960

1.8

Algeria

8

0.5

Greece

2

12 x 10

12 10

1

2005

4

14

1.5

7

2.4 x 10

2.6 2.4

1980

Turkey

2

3

1965

1970

1975

1980

1985

1990

1995

2000

2005

7

1.4 x 10 1965

1970

1975

1980

1985

1990

1995

2000

2005

6 4 2 0 1960

30 N 10 W

7

1 x 10 1965

0

1970

1975

1980

1985

1990

1995

2000

2005

10 E

Figure 15. Annual chicken production 1961-2005.

20 E

30 E

40 E

The ENSEMBLES Climate Change Project

89

The climate vulnerabilities of chickens are: Temperature >25° C: loss of appetite, lower egg production >30° C: illness Water Up to 0.3 litres/day 3-4 times that on hot days Likely changes in the number of days with daily maximum temperature greater than 25° C are shown in Figure 16. By the end of the century, egg production will have been seriously affected over most of the Mediterranean. Note that, even if the chickens are kept in airconditioned sheds, their profitability will be reduced by the extra energy costs.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 16. Of days per year with daily maximum temperature > 25° C.

The equivalent plot for the number of days with daily maximum temperature greater than 30° C (Figure 17), indicates that many parts of the Mediterranean are likely to suffer sick or dying chickens by the end of the century.

Sheep The annual total numbers of sheep 1961-2005 (Figure 18), shows a varying picture across the region. In France, Italy, and Turkey sheep numbers have declined over recent years, whereas in Spain, Algeria, and Greece numbers have increased.

90

Paul van der Linden

Figure 17. Number of days per year with daily maximum temperature > 30° C.

60 N

4

1.4

50 N

x 10

1.3

1.2

7

1.3 x 10

France

1.1

1

7

0.9 x 10

0.9

0.8 1960

1965

1970

1975

12000

1980

1985

1990

1995

2000

6

11.5 x 10

11500

2005

11000 4

2.6

x 10

10500 4

10000 2.4

2.2

40 N

7

7

2.4 x 10

2.2 x 10

9000 8500

2

8000

1.8

1.6

1.4 1960

6

9400 9200 9000

1965

1970

1975

1980

1985

1990

1995

x 10

4.5

8 x 10

7500 1960

Spain

5

Italy

9500

2000

2005

6

9.3 x 10

8600

7

4

Greece

8800

4.9 x 10

Turkey

3.5

8400

3

7

3.3 x 10

8200 1965

1970

1975

1980

1985

1990

1995

2000

2005

8000

x 10

7400 1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

7.6 x 10

7600

7

1.8 x 10

2.5 1960

6

7800

4

2 1.8

1965

1970

1975

1980

1985

1990

1995

2000

2005

1.6 1.4 1.2

Algeria

1 0.8

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0.6 0.4 0.2 1960

30 N 10 W

7

0.4 x 10 1965

0

1970

1975

1980

1985

1990

1995

2000

2005

10 E

20 E

30 E

Figure 18. Annual total number of sheep 1961-2005.

The climate vulnerabilities of sheep are: Temperature >30° C: decline in milk yield and fat content even with low humidity >32° C: reproduction declines, rams can become impotent for 2-3 months Water 3-6 litres/day in winter, 14 litres/day in summer more on very hot days

40 E

The ENSEMBLES Climate Change Project

91

Figure 19. Number of days per year with daily maximum temperature > 32° C.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The change in the number of days with daily maximum temperature above 30° C is shown in Figure 20. Figure 22 shows the number of days with daily maximum temperature above 32° C. The spatial distribution of increases in the number of days with high temperatures indicate that Spain and Algeria, both with increasing sheep production, can expect reduced milk yield and reproductive problems in their flocks by the end of the century. Note that milk yield is vital in this region since Mediterranean countries have only 11% of the world’s sheep but produce 67% of global sheep dairy produce. Although the above models can only demonstrate the method, and require detailed configuration using data from the resolution Regional Climate Models in ENSEMBLES, some tentative conclusions arise that can be used as a benchmark for comparison with later experiments: Importance of farming to the Mediterranean • Farming is becoming less important as a source of employment • Relative importance of food exports has declined, but is still over 10% of total exports • Meat production is booming everywhere Dependence of farming on climate • Extensive adoption of irrigation has removed much of the direct dependence on weather • But, water for irrigation depends on regular replenishment in winter and introduces a new, deferred, dependence to farmers Olive oil production • Increasing everywhere, presumably because of irrigation

92

Paul van der Linden •

•

Water stresses under climate change are not important until after 2040 when southern Spain, Sicily, and southern Greece will require permanent irrigation to maintain yields. The main cropping areas are, on average, not badly affected in terms of the water requirements of olives. NB this could change with higher resolution model data

Citrus exports • Large increases in the number of days with Tmax>38 deg C over Spain and Turkey (by 2 months or more) mean that the two largest producers could be severely affected Chickens • With a likely doubling in the number of days with Tmax>25 deg C, producers will have to spend on air conditioning to maintain production. • Similarly for days with Tmax>30 deg C. Here to prevent possibly fatal health problems.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Sheep • Increases in the number of days with Tmax>30 deg C, means reduced milk yield, particularly in southern Spain. Solutions involve more robust breeds or shifting production to cooler sites. • Similarly, increases in days with Tmax>32 deg C mean diminished reproduction for longer periods. Southern Spain and Algeria are particularly badly affected. The most important development of the models described here is to repeat the experiments with Regional Climate model data which will give a much more reliable estimate of the distribution of climate extremes. It will also allow the estimation of uncertainty in the projected impacts. In addition to the Mediterranean examples in this report, many of the models are appropriate for the whole of Europe. The complete list of farming parameters that have been modelled includes: wheat, potatoes, chickens, sheep, cattle, pigs, citrus, olives and grapes, and can be extended. In addition, it is important to realise that the thresholds will vary between breeds and varieties, with potential for adaptation using drought resistant crops, for example. To make the modelling as flexible as possible, therefore, it is proposed to develop software to accept user specified thresholds, then run the model. This will run as an MS Windows program on a PC and will not require climate expertise. However, full understanding of the vulnerabilities of crops or livestock will be needed.

SUMMARY OF PROJECT PROGRESS Ensembles of climate models for constructing probabilistic projections have been constructed by other organisations (IPCC and the PRUDENCE project), but not in the fully integrated way that it has been done in ENSEMBLES with impact models.

The ENSEMBLES Climate Change Project

93

One of the strengths of the project is that the verification of models and their output has been rigorous with evaluation against observations at different timescales by hindcasting. Also the way in which climate variability, predictability and extremes have been treated throughout each component of the project (as well as being a Research Theme in its own right) means that it has been treated consistently. Uncertainty is another key issue for climate modellers and those who use their output, and again this has been treated throughout the project in an integrated way – from scenario, to climate model, to impact model, to the feedbacks between these components. Part of the project process has been to consult with stakeholders and users of climate change information to canvass their opinions about ways in which results can be geared to make them more useful and relevant to their needs. The results are then presented in a comprehensive outreach programme including workshops, newsletters, papers, PhD training, symposia presentations and press releases to name a few of the output media. At the time of writing the ENSEMBLES project is 44 months through its 60 month programme and has delivered about two thirds of its agreed output. The final results of the project, which have yet to be delivered, are mostly the outputs from climate change impact models, including crop models. The input which these models need (probabilistic projections from GCM and RCM ensembles) has been produced and is now available along with all the supporting data and information. It is anticipated that the full results will be published in late 2009 in a single volume.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PROJECT LINKS, CONTACTS AND ACKNOWLEDGEMENT The ENSEMBLES project website is www.ensembles-eu.org. On this site can be found project output (papers, results, data) and information about project progress (Deliverables and Milestones) as well as general information (meetings, glossary, links, project documents). There are also links to the websites of the Research Themes and the RT2B regional scenarios portal: grupos.unican.es/ai/meteo/ensembles/ Outreach for the ENSEMBLES project is handled by Research Theme 8 (RT8), who produce newsletters, workshops, PhD training, leaflets, conference papers, book contributions and specialist project results websites. For further information please visit the website for RT8 at: www.unige.ch/climate/Projects/ENSEMBLES-RT8.html Contacting the project can either be directly to the project office at: [email protected] or through the contacts given on the RT websites. Research institutes interested in working more closely with ENSEMBLES can apply for affiliation. There is no cost and affiliates are invited to meetings and receive newsletter and email information. In return we ask affiliates to share with the us how they use the data and information derived from our project. Note: individuals and projects cannot be affiliated, only institutes/organisations, in the case of projects affiliation will be given to the coordinating institute. Please contact the project office using the email above for more information. The project is financially supported by the EU FP6 Integrated Project ENSEMBLES (Contract number 505539).

94

Paul van der Linden

APPENDIX 1. SUMMARY OF ENSEMBLES MODEL RUNS BY RESEARCH THEME This Appendix lists the models and summarises the runs which each Research Theme (RT) is contributing to the project. Access to the data produced by the modelling groups is through each individual RT website. (Note: some data is open and available while other data is only for project partners.) Research Theme 1: http://www.ecmwf.int/research/EU_projects/ENSEMBLES/index.html Table A.1. Summary of the models and runs carried out by the GCM modelling groups in RT1 Group

Model

Short description of model runs carried out

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

OCANAL denotes model integrations providing analyses of ocean observations used to initialise ensemble hindcasts. HIND denotes seasonal to decadal hindcasts carried out to test the multimodel seasonal to decadal prediction system. SCEN denotes climate change scenario simulations of 21st century climate change supporting development of the ensemble-based methodology for centennial climate prediction. CERFACS ARPEGE + ORCA, Variational OCANAL, HIND Initial Cond CNRM ARPEGE + ORCA HIND CNRM ARPEGE + ORCA + ISBA-Ags SCEN ECMWF IFS + ORCA OCANAL, HIND IFM-GEOMAR ECHAM5 + MPI OM OCANAL, HIND INGV ECHAM4.6 + ORCA OCANAL IPSL ORCA OCANAL KNMI ORCA OCANAL METO-HC HadCM3 HIND METO-HC GloSea OCANAL, HIND METO-HC HadCM3 SCEN UOXFDC HadCM3L SCEN DMI ECHAM5+MPIOM, DCM SCEN

Research Theme 2A: http://www.cnrm.meteo.fr/ensembles/ Table A.2. Summary of the models and runs carried out by the GCM modelling groups in RT2A Group

Model

Short description of model runs carried out

CNRM

ARPEGE + ORCA + OPA 8

CERFACS

ARPEGE + PRISM + ORCALIM

Seasonal to decadal hindcasts 20th Century hindcasts and 21st century scenarios Seasonal to decadal hindcasts with variational initial conditions

The ENSEMBLES Climate Change Project

95

Group

Model

Short description of model runs carried out

UiO

CTM

CNRM ECMWF

ARPEGE + OPA 8 IFS + ORCA

MPIMET DMI METO-HC

ECHAM 5 + MPI-OM ECHAM 5 + MPI-OM HadCM3

IPSL INVG NERSC UCL-ASTR UREADMM With METO-HC

IPSL-CM4 ECHAM4.6 + OPA 8.2 + LIM ARPEGE + MICOM IPSL-CM4 UK-HiGEM High resolution version of HadGEM1

Evolution of the chemical composition of the atmosphere 20th Century hindcasts and 21st century scenarios Second stream seasonal to decadal hindcast production 20th Century hindcasts and 21st century scenarios 20th Century hindcasts and 21st century scenarios 20th Century hindcasts and 21st century scenarios seasonal-decadal hindcast production 20th Century hindcasts and 21st century scenarios 20th Century hindcasts and 21st century scenarios 20th Century hindcasts and 21st century scenarios 21st century scenarios climate change scenarios for the 21st Century

Research Theme 2B: http://www.cru.uea.ac.uk/projects/ensemblesrt2b/ This research theme has online data access and a statistical downscaling portal at: http://grupos.unican.es/ai/meteo/ensembles/index.html Table A.3. Summary of statistical downscaling methods to be used in RT2B. NB methods implemented in the ENSEMBLES web-based downscaling service (http://grupos.unican.es/ai/meteo/ensembles/) are not listed. ACC: Anthropogenic Climate Change runs to 2100. s2d: seasonal to decadal hindcasts ENSEM BLES partner ARPASIM

Variables to be downscaled

Method

Daily Precipitation, Tmin, Tmax

Regression, conditioned by circulation - canonical correlation analysis (CCA) Multiple linear regression –Model Output Statistics + BLUE Two-step analogue method

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Daily Precipitation, Tmin, Tmax

FIC

GKSS IAP

IAP

Daily precipitation and temperatures. Wind and humidity will be tested. Marine surface wind Daily temperature (and daily precipitation?) Daily temperature

ENSEMBLES runs to be downscaled ACC GCM runs

Region(s) where downscaling will be applied N-Italy

Stream 1 s2d runs

Italy

ACC GCM runs RT2B RCMs if time

Europe – ENSEMBLES gridded observations Germany/Netherl ands ECAandD European stations ECAandD European stations

Conditional stochastic weather generator Regression, conditioned by circulation

ACC GCM runs

Multilayer perceptron neural network

ACC GCM runs

ACC GCM runs

96

Paul van der Linden Table A.3 (Continued)

ENSEM BLES partner IAP

IAP

KNMI

NIHWM

NMA

UEA

Variables to be downscaled

Method

Precipitation, Tmin and Tmax, solar radiation Daily temperature (and daily precipitation?) Multi-site (sub)daily RCM precipitation (and temperature) Temperature, precipitation, drought indices, river discharge Daily precipitation

Conditional stochastic weather generator

Daily precipitation, Tmax, Tmin, vapour pressure, wind speed, sunshine duration, relative humidity, reference PET

ENSEMBLES runs to be downscaled ACC GCM runs

Region(s) where downscaling will be applied ECAandD European stations

Multiple linear regression

ACC GCM runs

ECAandD European stations

Nearest-neighbour resampling

ACC GCM runs

River Rhine catchment

Conditional stochastic weather generator

ACC GCM runs

Danube basin

Mixture between twostate first order Markov chain and CCA Stochastic weather generator

ACC GCM runs Possibly RCM runs Change factors taken from RT2B RCM runs

Southern Romania 7 mainland European stations, plus 3-4 UK stations.

Research Theme 3: http://ensemblesrt3.dmi.dk/

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table A.4. Proposed simulations with Global models and Regional models. Instead of naming the models we list the name of the partner producing the simulations. Simulations denoted with * are outside of the contractual obligation of the partners. Partners denoted ** are not obliged to do any simulations within RT2B of ENSEMBLES. Simulations beyond 2000 are part of the RT2B contribution to ENSEMBLES, while simulations up till 2000 are for RT3 Global model Regional model METO-HC

METO-HC

MPIMET

1950-2100 (x3)

1950-2100

MPIMET

1950-2100

CNRM DMI ETH KNMI ICTP

1950-2100 1950-2050 1950-2050 1950-2050

IPSL

CNRM

NERSC

CGCM3

Total number 2 (4)

19502050*

2 19502050 19502050*

2 2 1 1 1

The ENSEMBLES Climate Change Project Global model Regional model SMHI

METO-HC

UCLM C4I GKSS**

1950-2050

MPIMET

IPSL

CNRM

1950-2050

NERSC

CGCM3

19502050*

19502050* 19502050*

CHMI**

1

19502050*

1

OURANOS** 4 (6)

6

2

Total number 2 1 1 1

1950-2050

Met.No**

Total (1950-2050)

97

3

2

19502050* 1

1 18 (20)

RT5: http://www.knmi.nl/samenw/ensembles_rt5/index.html This research theme has produced a high resolution dataset of daily climate observations for Europe, available at: http://eca.knmi.nl/download/ensembles/ensembles.php This was prepared from daily station data provided by national meteorological services by gridding (20km) and quality controlling the data for consistency through space and time. The grid and variables are the same as the ENSEMBLES regional climate data projections, and information about uncertainty within the data has also been quantified.

Climate Model Experiment Variables

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table A.5. Typical list of variables describing data output in a seasonal to decadal climate model run, including information about periodicity of data Atmosphere and ocean common variables defined for the s2d experiments. These variables are a minimum common set to ensure that some analyses, downscaling exercises and end-user applications can be carried out for all the coupled models. A longer list is also available upon request from the individual modelling centres. Modellers also provide the land-sea mask and orography fields used in their atmospheric models. Atmosphere (GRIB format) The daily variables are either instantaneous at 00 GMT or accumulated daily over 24 hours starting at 00 GMT. The pressure levels available are 850, 500, 200 and 50 hPa. The data are written in the model grid. Fluxes are downward positive. Maximum and minimum temperature at 2 metres are computed using four values at 6, 12, 18 and 24 GMT. Pressure level fields: geopotential (m**2/s**2) temperature (K) zonal wind (m/s) meridional wind (m/s) specific humidity (kg/kg)

98

Paul van der Linden Table A.5. (Continued) Surface fields: surface temperature (K) snow depth (m of water) surface sensible heat flux (Ws/m**2) surface latent heat flux (Ws/m**2) mean sea level pressure (Pa) total cloud cover ( [0,1]) zonal component of 10m wind (m/s) meridional component of 10m wind (m/s) 2m temperature (K) 2m dewpoint temperature (K) surface downward solar radiation (Ws/m**2) surface downward longwave radiation (Ws/m**2) surface net solar radiation (Ws/m**2) surface net longwave radiation (Ws/m**2) top net solar radiation (Ws/m**2) top net longwave radiation (Ws/m**2) moisture flux from the surface into the atmosphere or evaporation (m of water) 2m Tmax (K) 2m Tmin (K) total precipitation (m of water) vertically integrated volumetric soil water (m**3/m**3) variable units are in the brackets.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Monthly means for all the variables listed above are computed using one value per day (at 00 GMT). However, monthly means for 2m temperature and dewpoint, total cloud cover, mean sea level pressure and 10m wind should be computed using 6-hourly data (6, 12, 18 and 24 GMT) to adequately sample the daily cycle. Monthly means for the 100 hPa level are also expected. The accumulated fields should be converted into fluxes for the monthly means, for which the units are m/s for precipitation and W/m**2 for heat and radiation fields.

APPENDIX 2. CASE STUDY: THE DISAT CROP YIELD MODEL AS DEVELOPED IN ENSEMBLES Yield modelling of Mediterranean crops Marco Bindi, Roberto Ferrise and Marco Moriondo Department of Agronomy and Land Management, University of Firenze (DISAT), Italy

The ENSEMBLES Climate Change Project

99

Introduction The main tasks of DISAT presented in this report were: i) complete the development of a simple statistical model that emulates process-based crop yield models for three selected crops (olive, grapevine and durum wheat); ii) create yield response surfaces for one of the three crops for a pilot study area (i.e. durum wheat), and iii) define critical thresholds of impacts and obtain preliminary estimates of the likelihood of exceeding these thresholds using probabilistic information about future climate.

Development of a Simple Statistical Model to Emulate Process-Based Crop Yield Models An Artificial Neural Network (ANN) was used to develop simple statistical model that emulates process-based crop yield model (i.e. durum wheat).

Wheat Growth Simulation Model SIRIUS is a wheat simulation model that calculates biomass production from PAR (photosynthetically active radiation) and grain growth from simple partitioning rules (Jamieson et al. 1998), and includes soil-water budget and soil-plant nitrogen budget. The model needs, as inputs, daily weather data consisting of minimum and maximum temperature (Tmin and Tmax, respectively), rainfall (P) and global radiation (Rad). SIRIUS was calibrated on 3 sites located in Italy and it was validated on 6 grid points (50 Km x 50 Km) over the Mediterranean Basin.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Scenario Sensitivity Analyses Daily climatic data for temperature, rainfall and global radiation was obtained for about 30 years (1975-2005) for 9 different grid points (50 Km x 50 Km), representatives of the climatic variability over the Mediterranean Basin (Figure 1) (from MARS JRC archive).

Figure 1. The 9 representative sites used to train the ANN.

100

Paul van der Linden

Using these data-sets as a basis, a sensitivity analysis was carried out with respect to changes in mean annual temperatures and annual precipitation. The sensitivity analysis was conducted for precipitation changes between –40% and +20% (with 20% step) and temperature changes between 0°C and +8°C (with 2°C step) applied uniformly across all months of the year. For each of the resulting scenarios, SIRIUS model was run for 4 CO2 concentration levels (from 350 ppm to 650 ppm with 100 ppm step), 3 different soil types (Table 2) and 3 N-rates (110, 170 and 230 Kg N ha-1). For each scenario combination the average grain yield (Mg ha-1) over the 30-years period was calculated from the output of the SIRIUS model. Table 2. Soil types used to train the ANN

Sand % Clay % Soil Water Capacity

Sandy 93 3.6 53

Sandy-loam 72 25.4 115

Loamy 5 17.5 215

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Artificial Neural Network The ANN structure adopted in this study was a multilayer perceptron (MLP) with a feedforward configuration. This structure, selected because of its capacity to solve climatic problems, has been well demonstrated in several previous studies (Gardner and Dorling 1998; Trigo and Palutikof 1999). The multilayer perceptron (MLP) consists of a system of simple interconnected nodes or neurons assembled in several different layers. Each node calculates a linear combination of weighted inputs from the links feeding into it. The summed value is transformed using linear or non linear functions. The output obtained is then passed as an input to other nodes of the next layer. These calculations are repeated until the output layer is reached. More specifically, in this study an ANN-MLP structure with three layers and 20 nodes per layer was selected. A non-linear transfer function (log-sigmoid) was selected for all nodes and layers and a back-propagation algorithm (Rumelhart et al. 1986) was used for training the ANN. The optimal number of hidden nodes (over a range of 5-25 with a 5 node step) and the proper learning rate and momentum were determined by making a sensitivity analyses. Following the approach proposed by Olesen et al. (2007), 5 input variables were used to train the ANN over all 9 grid points: Soil water content (SWC), in mm N level, in Kg ha-1 CO2 concentration, in ppm T(AMJ): mean temperature over the period April-June, in °C P(AMJ): cumulated rainfall over the period April-June, in mm. The trained ANN was tested on the same 9 grid points using the Leave-One-Out Cross Validation test. Results showed a good correlation between ANN estimations and SIRIUS outputs (Figure 2). Pearson’s correlation coefficient was 0.95 and all the points lie close the

The ENSEMBLES Climate Change Project

101

line of 1:1 correspondence. Considering each single site, Pearson’s coefficient ranged from 0.90 (Turkey) to 0.96 (Spain and Italy) (Table 3). 12

1:1

-1

ANN Outputs (Mg ha )

10 8 6 4 2 0 0

2

4

6

8

10

12

SIRIUS outputs (Mg Ha-1)

Figure 2. Comparison between ANN and SIRIUS estimates of crop yields for the 9 sites and scenario combinations.

Table 3. Pearson’s correlation coefficients between ANN and SIRIUS estimates of crop yields for each of the 9 grid cells with all weather scenarios, 3 soils and 3 N-rates

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Grid Cell Identification Number 30090 34058 34080 37043 37063 43031 43046 44056 45061

Latitude 36.64 40.12 39.29 41.12 41.42 42.93 43.93 44.62 45.05

Longitude 27.31 9.29 22.11 0.34 12.30 -7.40 1.81 8.05 11.23

Pearson’s correlation coefficient 0.90 0.94 0.95 0.96 0.95 0.92 0.95 0.93 0.96

Yield Response Surfaces for Durum Wheat for a Pilot Study Area in France The trained ANN was applied on a study area in France (Lon 1.81, Lat 43.94, Alt 210m). Yield response surfaces were estimated by altering baseline climate for temperatures (from 0°C to 8°C) and for precipitation (from -40% to 20%). Two CO2 concentration scenario were considered (350 ppm and 550 ppm). Soil Water Capacity and N fertilization were respectively set to 115 mm and 170 Kg ha-1. From the comparison with SIRIUS outputs, it is concluded that ANN reproduces quite well the trend of crop yield across different scenarios (Figure 3 and 4).

102

Paul van der Linden 8

France

ANN O utpu ts (Mg h a-1 )

7

6

5

4

y = 0.9369x + 0.0207 R2 = 0.9459

3

2 2

3

4

5

6

7

8

SIRIUS Outputs (Mg ha-1)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Comparison between ANN and SIRIUS estimates of crop yields used to draw the response surfaces for a study area.

Figure 4. Response surfaces for 2 different CO2 levels for a pilot study area.

The ENSEMBLES Climate Change Project

103

The Pearson’s correlation coefficient was 0.97 and the slope of the regression line close to unity. Furthermore, bigger shifting of ANN outputs to SIRIUS yield were close to the climate extreme variations (e.g. +5% of yield for ΔT>6.5°C; data not shown).

Critical Impact Thresholds and Preliminary Likelihood Estimates of Their Exceedence Critical thresholds of impacts were determined by: −

−

Calculating, for each grid cell, the distribution of the selected parameter (e.g. average crop yield, phenological phases, etc.) according to the joint distribution of temperature and rainfall changes for present period (1990-2010), Selecting the values that correspond to the 20th percentile of cumulative probability.

An example for the pilot grid cell in France is shown in Figure 5.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. Cumulative distribution of average yield in a pilot study area. Average yield distribution was calculated altering the 30-years baseline climate with ΔT and ΔP joint distribution, from Hadley Centre perturbed physics experiment, for present period (1990-2010).

Estimating Risk Probabilities for Durum Wheat: A Case Study Since the meteorological input variables for the ANN were the absolute values of mean temperature and cumulated precipitation over the period April-June and the data from perturbed physics experiment of Hadley Centre consisted of annual variations of both temperature (ΔT , in °C) and rainfall (ΔP, as %), an interpolative method was adopted to calculate future yields. Data from the Hadley Centre were interpolated with the data of response surfaces by using the bilinear interpolative method of the package “fields” of the statistical software “R” (http://www.r-project.org). The risk probability in each time period was defined as the percentage of perturbed yields that do not overcome the yield threshold (Figure 6). The above procedure was applied to define preliminary risk probabilities in Tuscany. The study area consisted of 8 grid cells in 50 Km x 50 Km spatial resolution (Table 8).

104

Paul van der Linden

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 8. The 8 grid cells representing the study area and the correspondent yield impact threshold Grid Cell Id. Number

Latitude

Longitude

Altitude

Yield Threshold (Mg ha-1)

39061 40061 41062 41061 41060 42062 42061 42060

42.4 42.8 43.2 43.3 43.3 43.7 43.7 43.7

11.2 11.1 11.7 11.2 10.5 11.8 11.2 10.5

14 129 350 324 107 600 229 78

4.79 4.88 6.20 6.16 5.91 6.12 5.78 5.92

a

b

c

d

Figure 6. Graphical representation of the risk probability. Figure was obtained by overlapping response surfaces and joint distribution of T and P changes (red and blue dots) for a grid box in Tuscany for the period 2021-40 (a), 2041-60 (b), 2061-80 (c) and 2081-2100 (d). Red dots are yields lower than the threshold (red line) . The percentage of red dots represents the risk probabilities.

The ENSEMBLES Climate Change Project

105

Response surfaces for each grid point and at different time periods were obtained by altering the baseline climate from MARS JRC and considering a CO2 concentration according to the A1B IPCC scenario (on the basis of which the Hadley data are estimated). Soil properties were from the Eusoils database and 170 Kg N ha-1 was uniformly adopted as nitrogen availability. As mentioned above, yield threshold was calculated, for each grid cell, as the 20th percentile of the cumulative distribution of the average yield for present period (Table 8). Preliminary results are illustrated in Figure 7. Changes in risk probability were calculated as difference between the percentage of yield that not overcome the selected threshold in future and present period. In the next 30 years the risk probability for durum wheat in Tuscany shows an overall and uniform decrease with respect to present time. In average, the absolute risk to not overcome the critical yield is about 7% all over the region. Afterward the risk progressively increases, even though for the great part of the region it keeps below 20%. Maximum risk was estimated in 2060 and 2070 when the greatest reductions in yield were accounted all over the study area. In the last two decades of the century the risk decreases again showing great variability across the region. These dynamics could be explained considering the fertilizing effect of the CO2 increase that will compensate, at least in the first period, for negative effects due to rising temperatures and rainfall reductions. Furthermore, water use efficiency will enhance as a result of the CO2induced stomata closure and consequent reduced transpiration. On average for Tuscany, the preliminary Hadley Centre data indicate temperature increases up to 4°C and rainfall reductions not greater than 8%, at the end of the century. As a consequence, annual rainfall could range from 589 to 820 mm, and assuming that the distributions of precipitations will not change, the minimum crop water demand (~ 500 mm year-1) will be anyway satisfied. In these conditions, it is likely that the reduced yield, as expected as a consequence of reduced growing season due to high temperatures, could be compensated by the increased radiation use efficiency supported by adequate water availability.

Change in risk probabilities (%)

39061

40061

41062

41061 42061

41060 42060

42062

5 0 -5 -10

Periods

Figure 7. Change in risk probability in the 8 grid cells in the next decades.

0 209

208 0

0 207

0 206

0 205

204 0

0 203

0 202

0

-15 201

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

10

106

Paul van der Linden

Finally, must be noted that the reduction in risk probability, as estimated for the end of the century, may probably be ascribed to the greater uncertainty of climate projections (greater dispersion of the cloud as depicted in figure 6d).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Challinor, A.J., Wheeler, T.R., Craufurd, P.Q., Slingo, J.M., Grimes, D.I.F., 2004: Design and optimisation of a large-area process-based model for annual crops. Agricultural and Forest Meteorology 124: 99-120. Challinor, A.J., Wheeler T.R., Craufurd P.Q., Slingo J.M., 2005: Simulation of the impact of high temperature stress on annual crop yields. Agricultural and Forest Meteorology, 135 (1-4): 180-189. Challinor, A. J., J. M. Slingo, T. R. Wheeler and F. J. Doblas-Reyes (2005). Probabilistic hindcasts of crop yield over western India. Tellus 57A 498-512. Challinor, A.J., Wheeler T.R., Craufurd P.Q., et al., 2007: Adaptation of crops to climate change through genotypic responses to mean and extreme temperatures. Agriculture Ecosystems and Environment, Vol 119, Iss. 1-2, 190-204. Collins, M., B. B. B. Booth, G. R. Harris, J. M. Murphy, D. M. H. Sexton, and M. J. Webb, 2006: Towards quantifying uncertainty in transient climate change. Climate Dynamics, 27, 127147. Hewitt, C.D and Griggs, D.J., 2004: Ensembles-based predictions of climate changes and their impacts. Eos, 85, 566. IPCC, 2007a: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, Qin, Manning and Chen (eds.)]. Cambridge University Press, Cambridge, UK. IPCC, 2007b: Climate Change 2007: Impacts Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Parry, Canziani and van der Linden (eds.)]. Cambridge University Press, Cambridge, UK. Murphy, J. M., D. M. H. Sexton, D. N. Barnett, G. S. Jones, M. J. Webb, and coauthors, 2004: Quantification of modelling uncertainties in a large ensemble of climate change simulations. Nature, 430, 768–772. Nigam, S.N., Rao, R.C.N., Wynne, J.C., Williams, J.H., Fitzner, M. and Nagabhushanam, G.V.S. 1994. Effect and interaction of temperature and photoperiod on growth and partitioning in three groundnut (Arachis hypogaea L.) genotypes. Annals of Applied Biology 125: 541-552. Prasad et al. 1999, Sensitivity of peanut to timing of heat stress during reproductive development. Crop Science, 39, 1352-1357. Räisänen, J., and L. Ruokolainen, 2006: Probabilistic forecasts of near-term climate change based on a resampling ensemble technique. Tellus, 58A, 461472. Wheeler T., Peter Q. Craufurd, Richard H. Ellis, John R. Porter, P. V. Vara Prasad, 2000: Temperature variability and the yield of annual crops. Agriculture, Ecosystems and Environment. 82: 159-167.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 8

DOWNSCALED CLIMATE CHANGE SCENARIOS FOR SPAIN Ernesto Rodríguez∗, Eduardo Petisco and Petra Ramos AEMet, National Institute of Meteorology, Spain

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The availability of downscaled regional climate change projections is a pre-requisite for evaluating impacts to climate sensitive socio-economic sectors and ecological systems. The final objective of such evaluation is to design strategies of adaptation to some of the already unavoidable effects of climate change. A process of generation of downscaled climate change scenarios for Spain has been organized in two phases to meet such needs. This paper summarizes the main elements and outcomes of the first phase. Special focus has been placed on the probabilistic treatment of regionalized climate change projections. The probabilistic approach based on an ensemble of projections allows a preliminary estimation of uncertainties coming from different emission scenarios, different coupled atmosphere-ocean global circulation models and different downscaling algorithms. Also the impact models themselves must be based on ensembles of projections providing an evolutionary picture both of climate and impacts.

INTRODUCTION Spain, due to its geographical situation and socio-economic features, is very vulnerable to climate change (IPCC, 2007). Recent studies and surveys have estimated the relevant predicted impacts in different socio-economic sectors and ecological systems (see (Moreno, 2005) and references therein). To cope with such high vulnerability, the Spanish Ministry of Environment issued in 2006 the Spanish National Adaptation Plan to Climate Change, which provides the general reference framework for all the activities related to assessment of impacts, vulnerability and adaptation to climate change. The main aim of this plan is to ∗

[email protected]

108

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

integrate adaptation to climate change into the strategy of the different socio-economic sectors in Spain (MMA, 2006). The Adaptation Plan to Climate Change selected initially a few main priority lines for its first working programme. Among them, the generation of downscaled climate change projections was considered as a core activity for the evaluation of impacts and adaptation to climate change. The Spanish State Agency for Meteorology (AEMet) was the body assigned to coordinate at national level the production of such climate change projections and to make them available to the impacts community. The work of generating climate change projections has been organized in two phases. The first phase, lasting one year, made essentially use of already known empirical downscaling techniques and existing databases. Results (both methods and data) from EU 5th FP projects related with climate change and its regionalization were extensively used and compiled (singularly from PRUDENCE and STARDEX projects). This contribution briefly describes the main outcomes of the first phase, which are thoroughly described in (Brunet et al., 2008). The second phase, starting in 2007 and lasting 5 years, will extensively cover a broader range of downscaling methods, of coupled Atmosphere Ocean Global Circulation Models (AOGCM) and of emission scenarios to provide a probabilistic view of climate evolution. Section 2 provides an overview of observed temperature and precipitation trends over Spain. Section 3 summarizes changes of annual mean temperature and precipitation as obtained directly from global models for the last 30 years of 21st century. The downscaled climate change projections and their uncertainties for Spain are described in Section 4. Uncertainties affecting climate change projections and some general recommendations for their use by the impacts community are briefly described in Sections 5 and 6, respectively. The basic work related with quality control of observational data and with evaluation of global models is not discussed here. Both items are thoroughly discussed in (Brunet et al., 2008)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

OBSERVED TRENDS OF TEMPERATURE AND PRECIPITATION OVER SPAIN According to the Forth Assessment Report (AR4) of the IPCC, the warming of the climate system is unequivocal, as it is evident from observations of increases in global average air and ocean temperatures, widespread melting of snow and ice, and rising global average sea level. Eleven of the last twelve years (1996-2007) rank among the twelve warmest years in the instrumental record of global surface temperature (since 1850). The 100year linear trend (1906-2005) of 0.74 [0.56 to 0.92] °C given in AR4 is larger than the corresponding trend of 0.6 [0.4 to 0.8] °C (1901-2000) given in the Third Assessment Report (TAR) of the IPCC. The temperature increase is widespread over the globe, and is greater at higher northern latitudes. Land regions have warmed faster than the oceans ((IPCC, 2001), (IPCC, 2007)). However, this warming is by no means spatially homogeneous. Brunet et al. (2007), using a selection of 22 stations over the Iberian Peninsula covering the period 18502005, showed unquestionable signs of warming during the instrumental period. The estimated warming ranges from moderate to high in comparison with global average. The temperature evolution over the Iberian Peninsula is consistent with the corresponding global average, showing a remarkable increase since 1973. The most recent warming event (1973-2005)

Downscaled Climate Change Scenarios for Spain

109

shows a rate of change of 0.48 [0.36 to 0.66] ºC per decade, whereas the linear trend for 1901-2005 is 0.13 [0.10 to 0.16] ºC per decade, which is roughly speaking double than the corresponding global average. The temperature increase has been mainly due to the high warming observed in spring and summer seasons. Finally, the mean temperature increase is at a considerable extent caused by the bigger rate of change of maximum temperatures during 1850-2005. The vigorous warming in the equinoctial seasons and winter has been the main contributor to this differential warming. Mean temperature data from another set of 40 selected stations over the Iberian Peninsula also shows an increasing tendency of 3.7ºC/100 years for the period 1980-2006 (López, 2007, personal communication). Precipitation trend over Spain does not show such a well defined behaviour as temperature. The trend towards reduction of precipitation over subtropical latitudes reported in TAR (IPCC, 2001) and AR4 (IPCC, 2007) has no easy verification in the Spanish case due to the complexity of its spatial and seasonal distribution (Castro et al., 2005).

PROJECTIONS FROM GLOBAL MODELS OVER SPAIN

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The exploration of global model outputs over Spain allows a preliminary approximation to regional climate change projections. Using data from some of the global models evaluated in TAR (Mitchell et al., 2002) for the period 2070-2100 with respect to the reference period 1961-1990, figure 1 shows over Spain and for the emission scenario SRES A2 a noticeable increase of temperature ranging between 3 and 9 ºC in summer and between 2 and 5 ºC in winter.

Figure 1. Change of precipitation (%) against change of temperature for the period 2070-2100 with respect to the control period 1961-1990 averaging all grid points over the Iberian Peninsula. Data corresponds to 9 AOGCMs reviewed by TAR (IPCC, 2001). Summer (left) and winter (right) seasons are represented for the A2 SRES emission scenario.

110

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

The corresponding ranges for the emission scenario SRES B2 (not shown here) are 2-6 ºC and 1.5-4 ºC for summer and winter seasons, respectively. The change of precipitation, on the other hand, is hardly significant in winter and predominantly negative in summer, although some models show also positive values.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

DOWNSCALED PROJECTIONS OVER SPAIN The climate projections provided by AOGCMs for a selection of emission scenarios lack of the enough spatial resolution required by most users from the community of impacts and adaptation to climate change. In order to accommodate global projections, using resolutions ranging from 200-300 km, to regional or even local features a variety of downscaling techniques can be applied. These techniques adapt the AOGCMs model outputs to the physiographic characteristics of a certain region or point (Wilby and Wigley, 1997). All downscaling techniques make use of the projections provided by AOGCMs, adding small scale details associated to finer information on orography, land-use, etc. Consequently, downscaled projections inherit problems and weakness from global models. If a global model incorrectly simulates aspects of the large-scale variability relevant to the regional/local climate, then downscaling has no sense. On the other hand, if natural climate variability is reasonably simulated by a global model, it will be very useful to translate the information contained in global patterns to specific points. It must be born in mind that the downscaling process introduces additional uncertainties which should be evaluated and quantified. A variety of downscaling techniques have been here applied (see Brunet et al. 2008), involving both dynamical and statistical methods. All dynamically downscaled projections here applied come from the PRUDENCE Project database, whereas the statistically downscaled projections were specifically computed for this project using 4 AOGCMs and 3 different empirical downscaling algorithms. The whole set of downscaled climate change projections –including both dynamical and statistical methods- allows a crude and preliminary estimation of uncertainties over different Spanish regions, as it is shown in figure 2 for the case of the Spanish region of Castilla y Leon. The panel represents directly the change of mean annual projections with respect to the control period 1961-1990 and the spread around the average projection providing some hint on the uncertainty coming from different global models and downscaling techniques. No attempt has been made to weight different projections. As most PRUDENCE projections were produced using boundaries from the HadAM3 model, the final result is excessively biased toward this model evolution. It should be also noticed that PRUDENCE projections were only available for the period 20702100, which is causing some inflexion around year 2070 for the multi-projection curves. Figure 3 shows the monthly evolution of maximum temperature warming for the particular case of the HadAM3H model using the empirical downscaling method based on analogues and for the A2 SRES emission scenario. It is noticeable the higher increase of temperature corresponding to the summer period and to the more inland regions. The same behaviour is shown by projections based on other AOGCMs and other downscaling techniques.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Downscaled Climate Change Scenarios for Spain

111

Figure 2. Different downscaled projections of change of annual mean maximum (top left) and minimum (middle left) temperature over Spanish region of Castilla y León using different AOGCMs, different downscaling techniques and two SRES emission scenarios (compared to the 1961-1990 base period). Evolution of mean value (bold line) and mean value +/- standard deviation (shadow) for maximum (top right), minimum (top middle) temperature and precipitation (bottom). [A 10 year running mean has been applied to the precipitation curve].

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

112

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

Figure 3. Change of monthly mean maximum temperature for the period 2071-2100 with respect to the reference 1961-1990 as computed from HadAM3H global model and downscaled by an analogue method for the emission scenario SRES A2.

UNCERTAINTIES IN GENERATION OF DOWNSCALED CLIMATE CHANGE PROJECTIONS Uncertainties can be explored using different techniques. These techniques include sensitivity and scenario analysis, as well as a formal probabilistic approach (Katz, 2002). In a traditional sensitivity analysis, the rate of change in the input of a model is determined as a

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Downscaled Climate Change Scenarios for Spain

113

single input is varied by an infinitesimal amount. In a scenario analysis instead of varying only one input at a time, all inputs are simultaneously changed, typically by much more than an infinitesimal amount. This technique is very much used in global climate change studies. The next step is the formal probabilistic approach, where one would have effectively an infinite number of scenarios, each one weighted by its likelihood. A practical realization of the formal probabilistic approach is the Monte-Carlo analysis, where inputs are randomly drawn from probability distributions. As Monte-Carlo simulations involve a huge computational effort to explore the whole range of uncertainties, a practical approach consists of generation of ensembles, frequently spanning a limited set of uncertainties. The regional climate prediction is a problem characterized by inherent uncertainty coming from a variety of sources, affecting hierarchically each step of the generation of downscaled climate change projections (Mitchell and Hulme, 1999). The first step is associated with alternative scenarios of future emissions, their conversion to atmospheric concentration and the radiative effects of these. The second step is related to the simulation by AOGCMs for a given emission scenario. This kind of uncertainty has a global aspect, related to the model global sensitivity to forcing, as well as a regional aspect, more tied to the model simulation of general circulation features. This uncertainty is important both, when AOGCM information is used for impact work without the intermediate step of a regionalization tool, and when AOGCM fields are used to drive a regionalization technique. The final step of uncertainty occurs when the AOGCM data are processed through a downscaling or regionalization method. Figures 4 and 5 show a simple view of the uncertainty range of downscaled climate change projections over the Iberian Peninsula associated to the three mentioned steps: (i) emission scenario, (ii) AOGCM, and (iii) downscaling technique. Using data from the PRUDENCE project referred to two SRES emission scenarios (A2 and B2), two AOGCMs (HadAM3H and ECHAM4/OPY) and two regional climate models (DMI and SMHI), changes of mean temperature and precipitation have been computed for the period 2070-2100 with respect to the control 1961-1990. Figure 4 shows that highest sensitivity for warming corresponds to changes in emission scenarios and in global driving models. Changes in regional models are hardly significant, except perhaps in spring and summer time. Therefore, the uncertainty introduced by the choice of the driving global model and by the choice of the emission scenario is generally larger than by the choice of the regional model, probably due to the strong constraint of using the same global model as boundaries imposed on different regional models. This constraint is even larger in winter time, when large scale phenomena are dominant. Figure 5 shows the corresponding figure for precipitation. Now the uncertainties introduced by the choice of the driving global model, emission scenario and regional mode are now of the same order. Another noticeable feature is the bigger standard deviation between different years in the case of precipitation as compared with temperature. The IPCC explores the uncertainties of climate change evolution coming either from future GHG and aerosol emissions (using a limited set of plausible scenarios) or from AOGCM (using an ensemble of 23 members). The result of this exploration is represented in the iconic figure from the AR4 (IPCC, 2007) (see figure 6).

114

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. Change of annual mean temperature for the period 2071-2100 with respect to control 19611990 over peninsular Spain in winter, spring, summer and fall. The value was estimated from two regional models (DMI and SMHI), two global models (HadAM3H and ECHAM4/OPYC) and two emission scenarios (A2 and B2 from IPCC_SRES). Vertical bars represent +/-1 standard deviation from mean value (marked with a point).

However, the same AR4 recognizes the existence of uncertainties not contemplated in the standard IPCC AR4 climate projection multi-model simulations. This is the case of feedbacks coming from processes, as, e.g., from carbon cycle and vegetation, not fully coupled in the standard AR4 projections, which are responsible of additional sampling of model uncertainties. The use of very large ensembles with multiple options for parameterization schemes and parameter values has proven that the sampling of projections based on the standard multi-model AR4 approach is somehow restrictive (see, e.g., Staniforth et al., 2005) and cannot span the full range of plausible model configurations. Only large ensembles of GCM projections sampling the widest possible range of modelling uncertainties can provide a reliable specification of the spread of possible regional changes. It still remains the crucial problem of knowing how well model uncertainties are sampled with the standard multi-model approach.

Downscaled Climate Change Scenarios for Spain

115

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. The same as figure 4, but for precipitation.

Figure 6. Left panel: Solid lines are multi-model global averages of surface warming (relative to 19801999) for the SRES scenarios A2, A1B and B1, shown as continuations of the 20th century simulations. The orange line is for the experiment where concentrations were held constant at year 2000 values. The bars in the middle of the figure indicate the best estimate (solid line within each bar) and the likely range assessed for the six SRES marker scenarios at 2090-2099 relative to 1980-1999. Right panels: Projected surface temperature changes for the early and late 21st century relative to the period 19801999. The panels show the multi-AOGCM average projections for the A2 (top), A1B (middle) and B1 (bottom) SRES scenarios averaged over decades 2020-2029 (left) and 2090-2099 (right). {IPCC, 2007}.

116

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

The standard AR4 multi-model approach to estimate climate change projections should be considered as an improvement compared with ensembles of a single model sampling only uncertainty in the initial state. However, members of the multi-model ensemble share common systematic errors, and cannot span the full range of possible model configurations due to resource limitations. An additional strength of multi-model ensembles is that each member is subjected to careful testing in order to obtain a plausible and stable control simulation, although it is not guaranteed to identify the optimum location in the model parameter space. The same probabilistic approach should be followed to estimate downscaled projections. In fact, one of the main conclusions of the STARDEX project (see, e.g., (Haylock et al., 2006)) is that no single downscaling method is superior to the others for all regions, variables and seasons. Therefore a variety of downscaling methods should be applied to explore the uncertainties coming from this additional step in the generation of downscaled projections.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RECOMMENDATIONS FOR USAGE OF DOWNSCALED CLIMATE CHANGE PROJECTIONS The previous Section has shown some of the uncertainties attached to the process of estimating downscaled climate change projections. Uncertainties can be explored or even bound by using an ensemble of projections. Ideally, the number of members of an ensemble should be large enough to allow some quantification of the existing uncertainties in climate change projections by probability density functions. Results here presented have only shown sensitivity to changes in emission scenarios, AOGCMs and downscaling methods and, due to the small number of members, do not intend to represent a proper ensemble, exploring optimally the phase space of a climate system. Besides, no attempt was made to assign weight to different members depending on their quality representing climate of a control period. The approach based on ensembles providing an evolutionary picture of climate should be also conducted by the community of impacts. By their part, impact models are also introducing an additional uncertainty in the hierarchy of uncertainties appearing in each step of the process of generating downscaled climate change projections. Therefore, it is by no mean reasonable to study or explore impacts solely based on a single climate projection. The usage of an ensemble, including members spanning different impact models, should be a common tool also for impact studies. When using different members of an ensemble for evaluation of impacts, coincidence of results under a big number of members should be interpreted as a sign of robustness in the conclusions. On the other hand lack of coincidence would be associated with high uncertainty in the conclusions. A final cautionary remark should be made on biases. Results here shown are referred to differences between future scenario projections and control simulations. The climate evolution, mainly provided by AOGCMs, is not free of biases. Also downscaling methods can contribute with additional biases. In fact, it can be rather common that the difference between the observed and control climatology is comparable to the difference between future projection and control (Giorgi and Francisco, 2000). Assuming that biases from simulations are constant in time, absolute values of future projections can be estimated adding to observed climatology the difference between future and control simulations.

Downscaled Climate Change Scenarios for Spain

117

Figure 7. Mean temperature distribution for certain month and location obtained from: (i) daily series of observations in period 1961-90 (blue); (ii) control simulation (AOGCM + downscaling) in period 196190 (green); (iii) future simulation (AOGCM + downscaling) under some emission scenario for period 2071-2100 (red).

Figure 7 shows an example of distribution of downscaled temperature obtained from observations (obs) and control (con) and projection (pr) simulations. Control simulation is conducted for the same period than the observation series. To obtain a distribution projection consistent with observations, it is reasonable to correct this distribution with the bias of the control simulation.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Brunet, M.; Jones, P. D.; Sigró, J.; Saladié, O.; Aguilar, E.; Moberg, A.; Della-Marta, P. M.; Lister, D.; Walther, A.; López, D., 2007: Temporal and spatial temperature variability and change over Spain during 1850-2005. Journal of Geophysical Research, 112, doi:10.1029/2006JD008249. Brunet, M.; Casado, M.J.; Castro, M.; Galán, P.; Lopez, J.A.; Martín, J.M.; Pastor, A.; Petisco, E.; Ramos, P.; Ribalaygua, J.; Rodríguez, E.; Sanz, I.; Torres, L., 2008. Generation of downscaled climate change scenarios for Spain (in Spanish). Ministry of Environment (in press). Castro M., Martín-Vide J. and Alonso S., 2005. The climate in Spain: past, present and scenarios for 21st century (in Spanish). Chapter from the MMA report “Preliminary Evaluation of Impacts in Spain by Climate Change”. Giorgi, F. and Francisco, R., 2000. Evaluating uncertainties in the prediction of regional climate change. Geophysical Research Letters, 27, 1295-1298. Haylock, M.R., G.C. Cawley, C. H arpham, R. Wilby and C. M. Goodess, 2006. Downscaling heavy precipitation over the United Kingdom: a comparison of dynamical and statistical

118

Ernesto Rodríguez, Eduardo Petisco and Petra Ramos

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

methods and their future scenarios. Int. J. Climatol., 26(10), 1397–1415, doi:10.1002/joc.1318. IPCC, 2001. Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). J. T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P. J. van der Linden y D. Xiaosu (Eds.). Cambridge University Press, UK. pp 944. IPCC, 2007. Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. S. Salomon, D. Qin, M.Manning, Z. Chen, M. Marquis, K.B. Averyt, M Tignor and K.L. Miller (Eds.). Cambridge University Press, U.K. and New York, NY, USA, pp 996. Katz, R.W., 2002. Techniques for estimating uncertainty in climate change scenarios and impact studies. Climate Research, 20, 167-185. Mitchell, T.D. , and Hulme, M., 1999. Predicting regional climate change: living with uncertainty. Progress in Physical Geography, 23 (1), 57-78. Mitchell ,T.D., Hulme, M., and New, M., 2002. Climate data for political areas. Area 34:109112. MMA, 2006ª. Plan Nacional de Adaptación al Cambio Climático (PNACC), [http://www.mma.es/portal/secciones/cambio_climatico/areas_tematicas/impactos_cc/pdf /pna_v3.pdf], Oficina Española de Cambio Climático, Ministerio de Medio Ambiente. MMA, 2006b. Primer Programa de Trabajo del PNACC, (2006), [http://www.mma.es/portal/secciones/cambio_climatico/areas_tematicas/impactos_cc/pdf /1_prog_trabajo_v1.pdf],Oficina Española de Cambio Climático, Ministerio de Medio Ambiente. Moreno, J.M. (ed)., 2005. Preliminary Evaluation of Impacts in Spain by Climate Change (in Spanish). Report of Ministry of Environment. Staniforth, D.A, T. Aina, C. Christensen, M. Collins, N. Faull, D. J. Frame1, J. A. Kettleborough, S. Knight, A. Martin, J. M. Murphy, C. Piani, D. Sexton, L. A. Smith, R. A. Spicer, A. J. Thorpe and M. R. Allen, 2005. Uncertainty in predictions of the climate response to rising levels of greenhouse gases conditions. Nature, 433, 403-406. Wilby, R.L. and Wigley, T.M.L., 1997. Downscaling General Circulation Model output: a review of methods and limitations. Progress in Physical Geography, 21, 530-548.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 9

USING CROP MODELLING AS SUPPORT FOR AGRICULTURAL DECISION-MAKING UNDER VARIABLE CLIMATE CONDITIONS Josef Eitzinger∗ Institute of Meteorology, Department of Water, Atmosphere and Environment, University of Natural Natural Resources and Applied Life Sciences, Vienna, Austria

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Global climate change will lead to shifts in climate behaviour and cause manifold impacts on ecosystems in the next decades. In particular, climate change will have significant effects on agricultural production, which has been considered as the most weather-dependent among all the human activities. Negative impacts on agricultural production could be avoided or reduced significantly by applying appropriate adaptation measures (e.g. in farm technology) supported by available impact models, as well as forecasts and warning systems for decision-making. This will secure sustainable agricultural production in the future as well. Agricultural impact models such as crop simulation models can effectively estimate crop yields, as well as assess yield risk and effects of potential adaptation measures, under any climate conditions, even though they need to be carefully validated before being used for decision-making. In view of these potentials, the AGRIDEMA project (www.agridema.org) is one of the efforts to bring relevant model developers and users (applicants) closer together in order to better use these available tools for adaptation measures and related agricultural decision making.

∗

[email protected]

120

Josef Eitzinger

AGRICULTURAL-IMPACT MODELS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Crop production is the primary agricultural sector to be affected by climate change and variability. Therefore all related impact models to be used for decision support in agriculture could be potentially implemented, which can range from complex crop models to more simple approaches such as irrigation scheduling models or drought risk assessment algorithms. Also socio-economic tools (e.g. whole farm models) are important as decision making is strongly related to economic issues. Numerous impact models such as crop growth and irrigation models have been developed. They can use weather data input, such as short term weather forecast, a season’s forecasted weather or climate scenarios to estimate potential or actual growth, development or yield of crops or sropping systems. Despite that many simple models can be implemented, the use of mechanistic or physically-based crop-simulation models has been pointed out as the most reliable approach for agricultural decision-making under variable climate conditions (Hoogenboom, 2000). However, simple approaches may reduce uncertainties if necessary model input parameters of complex models are not available or have bad quality. The optimal choose of a model for a certain purpose regarding uncertainty impacts on the results depends therefore on many factors including avaialable data inputs and the complexity of the simulated problem (Figure 1.). Additionally, the technical infrastructure, user friendly models and also skills of model applicants are critical for use of such tools in general. Other factors concern the continuity in model development, maintainance and support for customers (applicants).

Figure 1. Relationship of model complexity and parameters of uncertainty.

The current situation in model application examples reveal that they are mostly used for research studies rather than for operational purposes and at the end user level. Tubiello and

Using Crop Modelling as Support for Agricultural Decision-Making…

121

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Ewert (2002), for example, summarized several published assessments of CO2-enrichment and climate change effects on crop production and they conclude that the group of models developed in the Netherlands (Van Ittersum et al., 2003) is among the most used in such assessments. However, there are many recognized crop-growth simulation and other impact models developed in the United Kingdom, France, Italy, Germany and several other EU countries. Recent contributions to crop modelling appeared in the European Society of Agronomy web, www.esagr.org. Besides, contacts through related COST actions such as COST718 – “Meteorological Applications for Agriculture” (http://agrometcost.bo.ibimet.cnr.it/) or the new COST 734 (“Impacts of Climate Change and Variability on European Agriculture: CLIVAGRI) are valuable sources of information on modelling tools and their applications in Europe. An overview on available impact models can be also seen e.g. at http://dino.wiz.uni-kassel.de/ecobas.html. A recent survey on crop model applications in Europe by the COST734 action, shows clearly that the number of model applications in research as well as for operational purposes (Figure 2) is strongly related to the relative economic importance of crops as well.

Figure 2. Preliminary COST734 survey of operational crop model applications in Europe by crops.

PROBLEMS AND UNCERTAINTIES OF MODEL APPLICATION The European research funds concerning agricultural climate-change impact assessments have been addressed mainly to theoretical issues rather than to research-results applications; although climate change and particularly its linked climate variability could lead to significant damages and yield losses in the next decades (IPCC, 2001). Despite of the recognised relevance of impact modelling tools for climate risk assessments of crop production, they have only marginally applied for supporting agricultural decision-making within Europe

122

Josef Eitzinger

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

(Figure 2.), because of several reasons, which one of them is related uncertainty in modelling results. Also many research studies are limited to specific aspects such as crop yields or yield risks under defined circumstances such as various climate scenarios, land use and management scenarios (e.g. Audsley et al., 2005; Downing et al., 2000; Parry, 2000). Often it is not feasible to consider all influential environmental factors, crop management or socioeconomic feedbacks because of lack of data, methods and information. Even at the crop production level there can be detected uncertainties, e.g. the potential effect of increasing atmospheric carbon dioxide concentration on crop growth processes and yield (e.g. Wolf et al., 2002; Ewert et al., 2002) and other uncertainties in the representation of reality in the simulation of the soil-crop-atmosphere interactions (Jannssen, 1995). Another well known source of uncertainty are sometimes the significant differences between the model inputs and their regionalization at the farmers field level, for example soil data or the weather input data (Nonhebel, 1993; Trnka et al., 2005) and derived weather data from General Circulation Models (GCM´s). Regional climate scenarios can differ from GCM´s on a regional basis considerably as well. Therefore downscaling methods should be used for regional crop yield simulations, if available. Especially in complex terrains such as in Austria, statistical downscaling showed large regional differences in air temperature and precipitation characteristics. Another problem is that climate scenarios mainly assume only changes of the mean (e.g. of temperature and preciptitation), while a change in climate variability can have additional negative impacts on crop yields as sensitivity analyses by crop models indicate (Semenov and Porter, 1995; Figure 3).

Figure 3. Impact of increasing temperature variability on crop yields (SD=standard deviation) and different GCM´s (AVG= average of 7 GCM´s);(Trnka et al., 2004).

Using Crop Modelling as Support for Agricultural Decision-Making…

123

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

From a practical point of view, often the quality, costs and availability of the essential model input data are an important limiting factor for application of crop models. The conditions vary from country to country, as a recent COST734 survey indicates. However, beside others, most problems are indentified especially for spatial (GIS) applications regarding the soil input data and crop experiment data regarding model validation and calibration (Figure 4).

Figure 4. Preliminary COST734 survey on main limiatations for crop model applications in European countries.

124

Josef Eitzinger

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

THE ROLE OF UNCERTAINTIES IN STUDY RESULTS OF CROP MODEL APPLICATIONS Above mentioned uncertainties play a major role for the basic acceptance of model applications for the decision makers. Related probablities often are not well understood and transferred to the end users. For example, some recent publications on climate change impacts on agriculture include crop management and socio-economic scenarios and feedbacks. In several cases it turned out that socio-economic changes could outrange the pure climate change impact in Europe as described for example in the paper of Holman et al. (2005). He investigated climate and socio-economic impacts and adaptation options, and cross-sectoral interactions between four major sectors driving landscape change (agriculture, biodiversity, coastal zones and water resources) in England. Despite yield changes, cropping is generally insensitive to climate, but very sensitive to socio-economic change. Similar findings for central Europe from the EUproject ACCELERATES are described by Audsley et al. (2005). They found that areas of agricultural land use in central Europe may not change significantly, however significant shifts of cropping pattern were simulated. Socio-economic scenarios generated larger changes than climate scenarios. Purely differences in climate scenarios had marginal effects on agricultural land use in many regions. This study however shows that there are still severe limitations for interpreting the results of such studies in general. These includes the often large spatial scale of soil data, which is much more heterogeneous in reality, the simplifications of applied crop models from specific crops, the uncertainties in socioeconomic scenarios. Often hydrological aspects of river basins are neglected, although they could have a significant impact on groundwater resources, water availability for crops, soil erosion and others under changed climate (e.g. Jasper et al., 2004; Prudhomme, 2003). Current site specific agricultural conditions such as cost related technological aspects of farm management, including the field size or the terrain or also often neglected. Another review of past simulation study results in middle Europe in general shows mainly a strong positive yield effect of increasing atmospheric carbon dioxide (e.g. Alexandrov et al., 2002), while new results from FACE and other field experiments often limit this effect considerably. Past simulation studies therefore potentially overestimated or generalized that effect, which in fact shows high variabilities between cultivars and due to many interactions with other stressors such as fertilization. In european regions, especially southern regions, an strongly increasing number and duration of heat and drought periods could affect crop yields through a decrease of available soil water reserves (Eitzinger et al., 2002), especially under poor soil conditions (such as low soil water storage capacity). Especially these soil conditions are highly variable in the spatial scale. All these factors indicate high spatial variability of impacting factors which creates the precondition to apply crop models only on a regional or local scale to be useable for decision making on a farm or regional scale. However, on a large scale supporting advices for decision making are already well developed such as that negative yield impacts through higher temperatures and shortening of the growing period of specific cultivars could be managed by a change to cultivars which are adapted to higher temperatures or an earlier seeding for spring crops. The introduction of more drought resistant cultivars is another option. Future options in farm technology, crop management, land use and socio-economic environment can therefore

Using Crop Modelling as Support for Agricultural Decision-Making…

125

include important adaptations measures for medium and long term crop production to climate change. New policies must of course be adopted under climate change conditions to secure sustainability of agricultural crop production in general, but practical decision making by means of impact models has to be based on local (regional) inputs, conditions and application.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Alexandrov, V., Eitzinger, J., Cajic, V., Oberforster, M., 2002. Potential impact of climate change on selected agricultural crops in north-eastern Austria. Global Change Biology 8 (4), 372-389. Audsley, E., Pearn, K.R., Simota, C., Cojocaru, G., Koutsidou, E., Rounsevell, M.D.A., Trnka, M., Alexandrov, V., 2006. What can scenario modelling tell us about future European scale land use, and what not? Environmental Science and Policy 9(2) (in press). Downing, T.E., Harrison, P.A., Butterfield, R.E., Lonsdale, K.G., (eds), 2000. Climate Change, Climatic Variability and Agriculture in Europe. An Integrated Assessment, Research Report No. 21, Brussels, Belgium: Commission of the European Union, Contract ENV4-CT95-0154, 445 pp. Eitzinger, J., Stastna, M., Zalud, Z. and M. Dubrovsky. 2002. A simulation study of the effect of soil water balance and water stress on winter wheat production under different climate change scenarios. Agric. Water Manage. 2003. 61:195-217. Ewert, F; Rodriguez, D; Jamieson, P; Semenov, MA; Mitchell, RAC; Goudriaan, J; Porter, JR; Kimball, BA; Pinter, PJ; Manderscheid, R; Weigel, HJ; Fangmeier, A; Fereres, E; Villalobos, F, 2002. Effects of elevated CO2 and drought on wheat: testing crop simulation models for different experimental and climatic conditions. Agriculture Ecosystems and Environment, 93 (1-3): 249-266. Holman, IP., Nicholls, RJ., Berry, PM., Harrison, PA., Audsley, E., Shackley, S., Rounsevell, MDA., 2005. A regional, multi-sectoral and integrated assessment of the impacts of climate and socio-economic change in the UK. Climate Change 71 (1): 43-73. Hoogenboom, G. 2000. Contribution of agrometeorology to the simulation of crop production and its applications. Agric. For. Meteorol. 103:137-157. Intergovernmental Panel on Climate Change (IPCC). 2001. Impacts, Adaptations and mitigation of climate change: Scientific-Technical analysis. Cambridge University Press.,879. Jasper, K., Calanca, P., Gyalistras, D., Fuhrer, J., 2004. Differential impacts of climate change on the hydrology of two alpine river basins. Climate Research 26 (2): 113-129. Jannssen, P.H.M.,1994. Assessing sensitivities and uncertainties in models : a critical evaluation. In : Grasman, J. and G. van Straten (Eds.). Predictability and nonlinear modeling in natural sciences and economics.- Kluwer, Dordrecht, 344-361. Nonhebel, S.,1993. The importance of weather data in crop growth simulation models and assessment of climate change effects, Ph.D. Thesis, Wageningen Agricultural University. Parry M (ed), 2000. Assessment of Potential Effects and Adaptations for Climate Change in Europe. The Europe Acacia Project. Jackson Environment Institute, University of East Anglia, Norwich, UK, 320 pp.

126

Josef Eitzinger

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Prudhomme, C., Jakob, D., Svensson, C., 2003. Uncertainty and climate change impact on the flood regime of small UK catchments. Journal of Hydrology 277 (1-2): 1-23. Semenov, M. A., Porter, J. R., 1995. Climatic variability and the modelling of crop yields. Agric.For.Meteorol. 73, 265-283. Trnka, M., Žalud, Z., Eitzinger, J., Dubrovský, M., 2005: Quantification of uncertainties introduced by selected methods for daily global solar radiation estimation. Agricultural and forest meteorology, Volume 131, Issues 1-2: 54-76. Tubiello, F.N. and F. Ewert. 2002. Simulating the effects of elevated CO2 on crops: approaches and applications for climate change. Eur. J. Agron. 16:1-18. Van Ittersum, M.K., Leffelaar, P.A., van Keulen, H., Kropff, M.J., Bastiaans, L. and J. Goudriaan. 2003. On approaches and applications of the Wageningen crop models. Eur. J. Agron. 18:2021-234. Wolf, J., van Oijen, M., Kempenaar, C., 2002. Analysis of the experimental variability in wheat responses to elevated CO2 and temperature. Agriculture Ecosystems and Environment 93 (1-3): 227-247.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 10

MODELING WATER MANAGEMENT STRATEGIES USING THE SWAP/WOFOST MODEL Joop G. Kroes∗ Centre for Water and Climate, Wageningen UR, 6700 AA, Wageningen, The Netherlands

ABSTRACT SWAP simulates transport of water, solutes and heat in the variably saturated/unsaturated zone. The model has interaction with vegetation development which may be simulated using the generic crop growth model WOFOST. The model has a physical base with flexible boundary conditions, and enables interaction of soil water with groundwater and surface water. Recently the model has been improved and enables applications of the model for different water management strategies such as e.g. evaluation of the impact of climate changes on drainage and irrigation conditions. Model use will be discussed and illustrated with case studies on agricultural water productivity and percolation of pesticides.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

INTRODUCTION From 1950 onwards agricultural research in The Netherlands has developed rapidly with special interest for drainage problems in agricultural land (NHV, 2006). Since the early 1970s agricultural problems in The Netherlands have been studied with models, with an important role for the relation between crop growth, groundwater and surface water. With increasing complexity of problems also the modeling tools became more complex. The SWAP1 model is such a relatively complex tool. It was developed to analyze the impact of the physical environment on plant development with special emphasis on soil physics, climate and water management strategies. To allow evaluations of water management strategies on drainage and ∗ 1

[email protected] SWAP is an acronym for Soil Water Atmosphere Plant

128

Joop G. Kroes

irrigation, much effort has been put in the flexibility of boundary conditions. This flexibility combined with the physical basis of this kind of models allows evaluations of water and solute flow dynamics in areas where both drainage and irrigation take place and where the interaction with groundwater plays an important role. This will be illustrated with two case studies. First a brief model description will be given and some examples of recent model evaluations. This paper will end with a discussion about the use of this kind of models in water management scenarios.

MODEL DESCRIPTION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Development of the SWAP model started some 30 years ago with the release of the agrohydrological model SWATR2 (Feddes et al, 1978). Since then several versions were launched and distributed all over the world. From 1990 onwards a group of researchers from Wageningen University and Research Centre has been involved in the development and application of the model. Solute transport, generic crop growth, soil heterogeneity and heat flow were added (van Dam et al., 1997 and van Dam, 2000), and the model name changed to SWAP (Soil Water, Atmosphere and Plant). The current base version of SWAP was released in 2003 (Kroes and van Dam, 2003), is available on internet (www.swap.alterra.nl) and is distributed as open source; a new base version is expected this year. SWAP simulates transport of water, solutes and heat in the variably saturated/unsaturated zone (Figure 1). The model has interaction with vegetation development which may be simulated using the generic crop growth submodel WOFOST (Ittersum et al., 2003). Crop growth is affected by the actual soil moisture and salinity status on a daily basis.

Figure 1. Hydrological processes in SWAP.

2

SWATR

is an acronym for Soil Water Actual Transpiration Rate.

Modeling Water Management Strategies Using the SWAP/WOFOST Model

129

The SWAP model numerically solves Richards’ equation for water transport, the convection dispersion equation for solute transport and the general heat conduction equation for soil temperatures. A description of theoretical aspects and corresponding mathematical equations of the current base version have been given by Kroes and van Dam (2003). The core of the model is the numerical solution of Richards’ equation: ⎡ ⎛ ∂h ⎞ ⎤ ∂ ⎢ k ( h) ⎜ + 1 ⎟ ⎥ ∂θ ⎝ ∂z ⎠⎦ = ⎣ − S ( h) ∂t ∂z

(1)

where θ is volumetric water content (-), t is time (T), k is hydraulic conductivity (L T-1), h is soil water pressure head (L), z is vertical coordinate (L, positive upwards) and S is a source/sink term (T-1) which accounts for root water extraction, lateral drainage and/or water exchange between soil matrix and macropores. Eq. (1) is solved integrally for both the unsaturated and saturated zone with the gradient of the soil water potential, i.e. ∂ (h + z ) / ∂z to induce soil water movement in both zones.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Eq. (1) is solved numerically, using the Mualem-Van Genuchten functions to describe the relations between θ, h and k. Different groundwater tables may be generated including perched water tables, which may occur above dense layers in the soil profile. With sound numerical solutions the model is able to simulate complex field conditions with strongly variable boundary conditions, being relevant for many climate studies. Recent developments focus on improvements of numerical solutions, macroporous flow, evapotranspiration, interactions with groundwater and surface water (van Dam et al, 2007). The improvements include a full revision of the numerical scheme and improvements on macroporous flow being based on field experiments. Calculations on evapotranspiration are similar to those proposed by FAO (Allen et al, 1998). The interaction between soil water and surface water may be described by i) overland flow caused by infiltration-excess, saturation runoff, run-on, inundation, snowmelt, ii) subsurface flow caused by drainage/infiltration to and from nearby surface waters. Irrigation strategies may be applied with fixed or a scheduled regime for time and depth of the dosage. Much effort has recently been put in evaluations and testing of model concepts, model codes and procedures. An example of one of these evaluations is given further below.

MODEL EVALUATION Sound numerical solutions are very important and therefore much attention has been paid to verification of the code, and more specifically to the testing of numerical implementations. One of those tests will be described hereafter. The benchmark given by Vanderborght et al. (2005) was used to evaluate numerical solutions for soil water pressure head profiles during steady state flow in layered soil profiles. Analogous to Vander-borght et al (2005) three soil profiles with 2 layers were simulated: i)

130

Joop G. Kroes

loam over sand, ii) sand over loam and iii) clay over sand. Results (Figure 2) of recent SWAP versions were satisfactory with low RMSE-values3. RMSE-loam-sand - 0.04 (cm)

0 -60

-40

-20

analytic numeric

-80

-60

-40

Depth (cm)

-20

0

analytic numeric

-100

-100

-80

Depth (cm)

RMSE-sand-loam - 2.57 (cm)

-50

-40

-30

-20

-10

-50

-40

-30

-20

-10

PressureHead (cm)

PressureHead (cm)

0

RMSE-clay-sand - 0.11 (cm)

-40 -60 -100

-80

Depth (cm)

-20

analytic numeric

-50

-40

-30

-20

-10

PressureHead (cm)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 2. Comparison of numerical and analytical solutions for soil pressure heads in 3 different layered soil profiles.

Similar evaluations were carried out for soil temperatures and solute concentrations with comparisons against analytical solutions. A detailed explanation will be given by Groenendijk and Kroes (in preparation). An extensive test protocol ensures the numerical code quality of SWAP. Furthermore, several studies have been carried with different purposes, but practically all include an evaluation of model performance by comparing simulated with measured data. An overview of studies has beengiven by van Dam (2000) on the web site www.swap.alterra.nl and by van Dam et al (2007).

CASE STUDIE 1 – INTERACTION WITH GROUNDWATER FLOW In most irrigated areas a close interaction exists between flow and transport in the vadose zone, groundwater and surface water. SWAP/WOFOST has therefore been applied in such an irrigated area in Haryana, India. The model was used to analyze the impact of different water

3

RMSE (Root Mean Square Error) is a measure of the difference between numerical and analytical values.

Modeling Water Management Strategies Using the SWAP/WOFOST Model

131

management strategies on regional crop water productivity, defined as crop yield per amount of water utilized. SWAP was used to determine water flow, WOFOST to assess crop yield. Water management in the area is complex due to low and erratic rainfall, poor groundwater quality and rising and declining groundwater levels (Figure 3, Singh et al, 2006). It was regarded essential to be able to model percolation as well as capillary rise since the lack or presence of these flows might cause drought and/or salinity stress. The SWAP/WOFOST model was calibrated against data from both extensive research and farmer field experiments for the major crops wheat, cotton and rice in the area (Bessembinder, 2003).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Schematic overview of an area where shallow groundwater levels affect both percolation and capillary rise.

At regional scale distributed modeling was applied using existing databases and satellite images for land use and crop rotations. The regional analysis showed that a combination of improved crop management, irrigation water distribution and seepage reduction will allow higher yields and higher water productivity (van Dam et al, 2006).

CASE STUDIE 2 – PERCOLATION AND CAPILLARY RISE The SWAP model has also been applied to generate hydrological conditions for the PEARL4 model (Leistra et al, 2000), i.e. a one-dimensional numerical model of pesticide behaviour in the soil-plant system. PEARL uses the hydrological and soil schematization of the SWAP model. in the EU review process (FOCUS, 2000) SWAP/PEARL is used as one of the

4

PEARL

is an acronym for Pesticide Emission Assessment at Regional and Local scales

132

Joop G. Kroes

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

FOCUS5 simulation models to analyze scenarios, that are used to calculate the concentrations of plant protection products in groundwater and surface water. Nine realistic worst-case groundwater scenarios have been defined to assess leaching potential of active substances. These scenarios collectively represent agriculture in the EU (Figure 4, see also http://viso.jrc.it/focus/index.html).

Figure 4. Nine representative locations for worst-case scenarios.

Crop information, weather data and soil properties have been defined for scenarios down to the ground waterlevel. Crop information has also been defined for each scenario, including five crops which can be grown across the whole EU, as well as a further twenty, which are particular to specific parts of the EU. Three models were applied in the scenario analysis: PRZM, PELMO and PEARL/SWAP. PRZM and PELMO are capacity models and SWAP/PEARL is a Richard’s model. All models 5

FOCUS:

is an acronym for FOrum for Co-ordination of pesticide fate models and their USe

Modeling Water Management Strategies Using the SWAP/WOFOST Model

133

determined irrigation based on pre-processed scheduled timing and depth of irrigation events. The overall benefit from this study is that it has delivered a consistent process in evaluating leaching potentials across the EU. Differences in results were foreseen and minimized as much as possible given the constraints of the scenario evaluations. Results showed differences in water balance components for root water uptake, overland flow and percolation. The differences are illustrated for the Sevilla case, where winter cereals were simulated for a period of 20 years. The overall percolation predicted from PEARL/SWAP turned out to be equal to or less than that from PRZM and PELMO (Figure 5). The differences were regarded as sufficient given an acceptable degree of uncertainty (FOCUS, 2000).

Figure 5. Simulated percolation from winter cereals at Sevilla over 20 years; negative flow means capillary rise (FOCUS, 2000).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

DISCUSSION It is inevitable that differences in modeled processes will produce different outputs. In many hydrological models for the unsaturated zone a physically based Richard’s approach is common (see e.g. Guilding, 1991, Vanclooster et al, 2004). Integrated and coupled physically based models like SWAP are used in many studies of which two case studies have been presented in this paper. The first case was on irrigated areas with a close interaction between flow and transport in the vadose zone, groundwater and surface water. The second case illustrated the interaction between vadose zone and groundwater. These cases also showed that in areas where root water uptake is influenced by upward flow (capillary rise) it may be of importance to simulate these as process. Guswa et al (2002) made a comparison between a simple and a complex model for soil moisture dynamics. They found that predictions of the two modeling approaches are quite similar if the plant can extract water from locally wet regions to make up for roots in dry portions of the soil columns; if not, the match is poor.

134

Joop G. Kroes

Ines et al (2001) compared SWAP and DSSAT and concluded that the physical basis of SWAP elevates its performance. However they also conclude that data requirements are high and the interface is minimal. Eitzinger et al (2004) compared CERES, WOFOST and SWAP and their ability to simulate soil water dynamics under different soil conditions. They concluded that CERES and SWAP simulated soil water dynamics well in the upper 30 cm of the soil and that in comparable environments their multi-layer approach is preferred. They recommended improvement of evapotranspiration routines and crop parameterization like a more effective rooting zone. Ma et al (2005) studied a coupling of a Richard’s based RWZQM (Root Zone Water Quality Model) with the crop growth part of DSSAT. Coupling was successful and showed for both approaches similar results. The use and application of such a coupling has its largest advantages in the ability to simulate macropore flow, tile drainage flow, pesticide movement and ground water tables. The previous comparisons showed that root water uptake has been critized by some authors. In some cases an improved parameterization will help, whereas in other cases new concepts are required. An example of such a promising new concept has been given by De Jong van Lier (2006) who proposed a numerical root water extraction submodel which was successfully tested and implemented. The use of models in water management strategies is only feasible if the model offers flexible boundary conditions. Interaction with groundwater and surface water is required to simulate drainage and irrigation under different climate conditions. The choice of the application of a complex versus a more detailed model depends on the purpose of the study. In general a simple approach is favourable, because it is easier to achieve input parameters and it facilitates uncertainty analyses. In practice water management situations are complex, cannot be simplified and require appropriate complex tools to analyze different strategies.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Allen, Richard G., Luis S. Pereira, Dirk Raes, Martin Smith, 1998. Crop evapotranspiration Guidelines for computing crop water requirements - FAO Irrigation and drainage paper 56, FAO, Rome. Bessembinder, J.J.E., A.S. Dhindwal, P.A. Leffelaar, T. Ponsioen, and Sher Singh. 2003. Analysis of crop growth. p. 59-83. In J.C. van Dam and R.S. Malik (ed.) Water productivity of irrigated crops in Sirsa district, India. Integration of remote sensing, crop and soil models and geographical information systems. WUR/CCSHAU/IWMI/ WaterWatch, WATPRO final report, Wageningen. de Jong van Lier, Quirijn, Klaas Metselaar and Jos C. van Dam, 2006. Root Water Extraction and Limiting Soil Hydraulic Conditions Estimated by Numerical Simulation. In: Vadose Zone J. 5:1264-1277 (2006) Eitzingera, J. , M. Trnka, b, J. Höschc, Z. Žaludb and M. Dubrovskýd, 2004. Comparison of CERES, WOFOST and SWAP models in simulating soil water content during growing season under different soil conditions. In: Ecological Modelling 171 (2004) 223-246. Feddes, R.A., P.J. Kowalik, and H. Zaradny. 1978. Simulation of field water use and crop yield. Simulation Monographs. Pudoc. Wageningen.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Modeling Water Management Strategies Using the SWAP/WOFOST Model

135

FOCUS, 2000. FOCUS groundwater scenarios in the EU review of active substances - The report of the work of the Groundwater Scenarios Workgroup of FOCUS (FOrum for the Co-ordination of pesticide fate models and their USe), Version 1 of November 2000. EC Document Reference Sanco/321/2000 rev.2, 202pp, http://viso.jrc.it/focus/gw/index.html Groenendijk, P. and J.G. Kroes. in prep. Performance of the numerical implementation of the Richards' equation and associated boundary conditions in the SWAP model. Guilding, 1991. Qualitative Mathematical Analysis of the Richards Equation. In: Transport in Porous Media 5: 651-666, 1991. Guswa, A.J., M.A. Celia and I. Rodriguez-Iturbe, 2002. Models of soil moisture dynamics in ecohydrology: A comparative study. In: Water Resour. Res. 38(9), 1166. Ines, A. V. M.; P. Droogers; I. W. Makin; and A. Das Gupta. 2001. Crop growth and soil water balance modeling to explore water management options. IWMI Working Paper 22. Colombo, Sri Lanka: International Water Management Institute. Ittersum, M.K. van, P.A. Leffelaar, H. van Keulen, M.J. Kropff, L. Bastiaans, and J. Goudriaan. 2003. On approaches and applications of the Wageningen crop models. Europ. J. Agronomy 18: 201-234. Kroes, J.G., and J.C. van Dam (ed.). 2003. Reference manual SWAP version 3.03. Alterra report 773. Wageningen. Ma, L., G. Hoogenboom, L. R. Ahuja, D. C. Nielsen, and J. C. Ascough, 2005. Development and Evaluation of the RZWQM-CROPGRO; Hybrid Model for Soybean Production. In Agron. J. 97:1172–1182 (2005). Published online July 13, 2005. Leistra, M., A.M.A. van der Linden, J.J.T.I.Boesten, A. Tiktak and F. van den Berg, 2000. PEARL model for pesticide behaviour and emissions in soil-plant systems. Description of processes. Alterra report 13, RIVM report 711401009, Alterra, Wageningen, 107 pp. NHV, reprint 2006. Water in The Netherlands, available internet: http://www.nhvsite.info/witn2004.htm (verified on 20070831). Singh, R., J.G. Kroes, J.C. van Dam, and R.A. Feddes. 2006b. Distributed ecohydrological modelling to evaluate the irrigation system performance in Sirsa district. I. Current water management and its productivity. J. Hydrol. 329: 692-713. Vanclooster, M., J. Boesten, A. Tiktak, N. Jarvis, J.G. Kroes, R. Muñoz-Carpena, B.E. Clothier and S.R. Green, 2004. On the use of unsaturated flow and transport models in nutrient and pesticide management, In: Unsaturated-Zone Modeling: Progress, Challenges and Applications, R.A. Feddes; G.H. de Rooij and J.C. van Dam (eds), Wageningen, The Netherlands. Published by Kluwer Academic Publishers and on internet (http://library.wur.nl/frontis/unsaturated/index.html ). Vanderborght, J., R. Kasteel, M. Herbst, M. Javaux, D. Thiery, M. Vanclooster, C. Mouvet, and H. Vereecken. 2005. A Set of Analytical Benchmarks to Test Numerical Models of Flow and Transport in Soils. Vadose Zone J. 4:206–221. van Dam, J.C. 2000. Field scale water flow and solute transport. SWAP model concepts, parameter estimation and case studies. Ph.D. dissertation. Wageningen University, the Netherlands. van Dam, J.C., J. Huygen, J.G. Wesseling, R.A. Feddes, P. Kabat, P.E.V. van Walsum, P. Groenendijk, and C.A. van Diepen. 1997. Theory of SWAP version 2.0. Simulation of water flow, solute transport and plant growth in the Soil-Water-Air-Plant environment. Technical document 45. Alterra. Wageningen.

136

Joop G. Kroes

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

van Dam, J.C., R. Singh, J.J.E. Bessembinder, P.A. Leffelaar, W.G.M. Bastiaanssen, R.K. Jhorar, J.G. Kroes, and P. Droogers. 2006. Assessing options to increase water productivity in irrigated river basins using remote sensing and modeling tools. Water Res. Development 22: 115-133. van Dam, J.C., P. Groenendijk, R.F.A. Hendriks, and J.G. Kroes. 2007 (in prep). Modeling soil moisture flow in the hydrological top system with SWAP. Vadose Zone Journal. accepted for publication.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 11

USE OF DSSAT MODELS FOR CLIMATE CHANGE IMPACT ASSESSMENT: CALIBRATION AND VALIDATION OF CERES-WHEAT AND CERES-MAIZE IN SPAIN Ana Iglesias∗ Univ Politecn Madrid, Dept Agr Econ and Social Sci, E-28040 Madrid, Spain

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The General features of the CERES-WHEAT and CERES-MAIZE models are presented. Presentation comprises general model features as well as the data needed for specific calibrations. A typical calibration of the CERES-WHEAT, conducted in the South of Spain, is shown in order to illustrate the procedure. The phenology coefficients P1V, P1D and P5 were calibrated so the observed and simulated phenological dates were as close as possible. In this WHEAT model, the simulated dates of the phenological stages, and therefore the number of days available for accumulation of grain dry matter, are most sensitive to the photoperiod coefficient (P1D). The sensitivity of the predicted phenology to changes in the vernalization coefficient (P1V), greatly depends on the value of the photoperiod coefficient (P1D). For a particular combination of P1D and P5, the physiological maturity is more sensitive to increases in P1V than the anthesis date. Once the phenology coefficients were calibrated, and therefore the simulated number of days available for grain filling, the yield component coefficients must be adjusted to represent as accurately as possible the actual yield components in the zone. Two validations of the obtained CERES-WHEAT calibrated coefficients are also shown, pointing out the simulations reliability. Similar analysis is provided for the CERES-MAIZE model.

∗

[email protected]

138

Ana Iglesias

CERES-WHEAT MODEL Model Description The CERES-Wheat model (Godwin et al., 1990; Ritchie and Otter, 1985) is a simulation model for maize that describes daily phenological development and growth in response to environmental factors (soils, weather and management). Modelled processes include phenological development, i.e. duration of growth stages, growth of vegetative and reproductive plant parts, extension growth of leaves and stems, senescence of leaves, biomass production and partitioning among plant parts, and root system dynamics. The models include subroutines to simulate soil and crop water balance and nitrogen balance, and they have the capability to simulate the effects of nitrogen deficiency and soil water deficit on photosynthesis and pathways of carbohydrate movement in the plant.

Phenology The primary variable influencing phasic development rate is temperature. The thermal time for each phase is modified by coefficients that characterize the response of different genotypes. The timing of crop phenological stages can be calibrated by modifying the coefficients that characterize vernalization (P1V), photoperiod response (P1D), duration of grain filling (P5) and phillochron interval (PHINT) of a particular variety.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Growth Potential dry matter production is a linear function of intercepted photosynthetically active radiation (PAR). The percentage of incoming PAR intercepted by the canopy is an exponential function of leaf area index (LAI). The dry matter allocation is determined by partitioning coefficient according to phenological stages and water stress. Final grain yield is the product of plant population, kernels per plant and weight of kernel. The number of kernels per plant is a linear function of stem weight and coefficients that accounts for the variation between genotypes of the number of grains per ear (G1) and spike number (G3). The maximum kernel growth rate is an input coefficient depending on the genotype of wheat (G2).

Water Balance The model includes a water balance routine where precipitation is an daily input; runoff is a function of soil type, soil moisture and precipitation; infiltration is precipitation minus runoff; drainage occurs when the soil moisture is greater than the soil water holding capacity of the bottom layer. Potential evaporation is calculated by the Priestley-Taylor relation; total evaporation is a function of potential evaporation, LAI and time as described by Ritchie

Use of DSSAT Models for Climate Change Impact Assessment

139

(1972); and transpiration is modified by LAI, soil evaporation and soil water deficit. Daily soil moisture is calculated as precipitation minus evaporation minus runoff minus drainage.

Input Data The model requires daily weather values of solar radiation, maximum and minimum temperatures and precipitation. Soil information needed includes drainage, runoff, evaporation and radiation reflection coefficients, soil water holding capacity amounts, and rooting preference coefficients foe each soil layer and initial soil water content.

CALIBRATION AND VALIDATION SITES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The field data are from the Agricultural Experimental Station of Tomejil (+37.40oN; .80oW); this Station is part of the Red Andaluza de Experimentacion Agraria (RAEA). It is located in the province of Sevilla at 30 Km from Sevilla (capital). The site represents one of the main agricultural regions of Spain (Valle del Guadalquivir). Wheat is the a main crop in the region (accounts for about 40% of the national wheat production). The cultivars grown are winter wheats that require little vernalization, sown in the late winter and non-irrigated. The experiments are dryland and nitrogen fertilized. Potential production was estimated based on the largest reported production in the area when the water balance for the wheat growing season showed no stress for the crop (RAEA, 1989, 1991). The calibration is based on the 1988-89 field experiments (RAEA, 1989) and the validation on the 1990-91 field experiments (RAEA, 1991). The data for the calibration correspond to field experiments performed during 1987-88 and include: daily weather data (maximum and minimum temperatures, precipitation and solar radiation); soil data; and crop and management data (dates of the main phenological stages, final yield, and fertilizer applications). The model was validated with an independent experimental data set in Tomejil (19901991) and in Las Tiesas (1990-1991).

FIELD EXPERIMENTS The calibration of the CERES-Wheat model in Tomejil is based on field data from 19881989 (RAEA, 1989); the validation in Tomejil is based on field data from 19901991 (RAEA, 1991) and the validation in Las Tiesas (Albacete) is based on field data from 1990-1991 (ITAP, 1991). All experiments were nitrogen fertilized to cover completely crop needs. In Tomejil the experiments are dryland (water-limited production) and in Las Tiesas are dryland and irrigated (water-limited and potential production). Potential production in the southern site (Tomejil) was estimated based on the largest reported production in the area when the water balance for the wheat growing season showed no stress for the crop (RAEA, 1989, 1991).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Field data. Sevilla. 1988-89. Sowing date 7 December 1988 (day 341). Sowing rate 360 seeds m2. Emergence 24 December 1988 (day 358). Dryland. Nitrogen fertilized

OBSERVED DATA Plants m-2 Tillering (day) Stem elongation (day) End spike growth (day)

1 380 25 69 103

2 335 25 73 100

3 320 25 69 98

4 425 25 69 99

Wheat CULTIVARS (*) 5 6 7 270 360 310 25 25 25 73 75 73 100 109 103

8 330 25 73 99

9 425 25 69 98

10 335 25 73 100

11 290 25 69 93

12 295 25 73 99

Anthesis (day) Spikes m-2 Physiological maturity (day) kg ha-1

108 547 147 594 8

108 557 147 589 1

105 672 143 687 7

105 557 143 600 1

107 485 147 579 0

105 640 147 682 1

105 662 148 622 7

108 575 149 603 2

100 570 143 632 1

105 602 148 644 4

OBSERVED DATA Plants m-2 Tillering (day) Stem elongation (day) End spike growth (day) Anthesis (day) Spikes m-2 Physiological maturity (day) kg ha-1

13 280 25 69 97 106 475 144 617 9

14 175 25 69 97 106 417 144 645 0

15 220 25 69 95 101 517 145 696 0

16 475 25 69 103 110 615 149 661 6

Wheat CULTIVARS (*) 17 18 19 325 265 360 25 25 25 69 73 69 95 97 95 101 103 101 432 507 545 145 146 145 607 1 607 1 654 8

20 480 25 69 95 101 542 145 648 8

21 305 25 73 97 103 452 147 643 0

22 325 25 73 93 100 702 142 669 9

23 390 25 73 93 100 502 142 545 0

24 295 25 73 97 106 555 144 551 2

110 697 147 442 8

107 602 147 594 6

(*) Wheat CULTIVARS: ADALID (1), ADONAY (2), ABANTO (3), ALBARES (4), ALCALA (5), ALDEANO (6), ANZA (7), ARGANDA (8), BETRES (9), CARDENO (10), CAJEME (11), CARTAYA (12), JABATO (13), LACHIS (14), MEXA (15), NIVELO (16), PESUDO (17), ROQUEÑO (18), TAURO (19), TRIANA (20), VITRON (21), YECORA (22), RINCONADA (23), CIBELES (24).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. Field data. Sevilla. 1990-91. Sowing date 29 November 1990 (day 333). Sowing rate 450 seeds m-2. Emergence 14 December 1990 (day 348). Dryland. Nitrogen fertilized

OBSERVED DATA Plants m-2 Tillering (day) Stem elongation (day) End spike growth (day)

Wheat CULTIVARS (*) 1 2 3 395 446 424 11 11 10 46 46 46 101 101 101

4 419 8 45 98

5 376 8 46 102

6 355 7 41 98

7 348 9 41 103

8 424 10 46 103

9 389 12 46 101

10 435 14 46 98

11 380 11 46 100

12 420 9 44 103

Anthesis (day) Spikes m-2 Physiological Maturity (day) kg ha-1

107 328 148 551 7

100 512 141 555 2

106 372 147 680 6

99 504 146 574 3

109 324 150 461 4

108 531 149 601 3

107 408 148 586 7

100 496 141 563 1

107 380 148 666 4

108 560 149 607 9

OBSERVED DATA Plants m-2 Tillering (day) Stem elongation (day)

Wheat CULTIVARS (*) 13 14 15 355 383 350 10 11 8 46 46 44

16 395 8 44

17 405 10 46

18 393 8 45

19 445 8 45

20 400 8 45

21 397 9 46

22 364 8 44

23 353 8 44

24 353 8 44

25 353 8 44

107 452 148 545 3

107 460 148 499 1

End spike growth (day)

98

98

95

99

100

93

88

89

93

95

99

99

99

Anthesis (day) Spikes m-2 Physiological Maturity (day) kg ha-1

105 248

105 400

103 288

105 368

107 316

99 372

95 344

95 508

99 440

101 400

108 456

108 320

108 320

146 558 1

146 539 2

144 559 1

146 631 2

148 527 9

140 559 2

136 530 9

136 459 7

140 599 2

142 649 7

149 603 4

149 540 2

149 512 2

(*) Wheat CULTIVARS: ABANTO (1), ADONAIS (2), ALCALA (3), ALDURA (4), AMPUERO (5), ANGRE (6), ANTON (7), ANZA (8), BETRES (9), CAJEME (10), CARTAYA (11), DARTAGNAN (BRIO) (12), DURADERO (13), GRANIZO (14), JABATO (15), MEXA (16), OSONA (17), PESUDO (18), RINCONADA (19), SEVILLANO (20), TAURO (21), TRIANA (22), VALIRA (23), VITRON (24), YECORA (25).

142

Ana Iglesias

Table 3. Wheat field tests in Tomejil (Sevilla) (1988-89; 1990-91). Dryland; Nitrogen fertilized. Average values and standard deviations Observed data Sowing (day) Seeds m-2 Emergence (day) Plants m-2 Tillering (day) Stem elongation (day) End Spike growth (day) Anthesis (day) Spikes m-2 Physiological maturity (day) Grain yield (kg ha-1) Potential grain yield (kg ha-1)

1988-89 341±0 360±0 358±0 332±73 25±0 71±2 98±4 105±3 559±79 146±2 6175±547

1990-91 333±0 450±0 348±0 390±31 9±2 45±1 98±4 104±4 402±80 145±4 5665±559

Table 4. Field data from a wheat irrigation experiment in Las Tiesas (Albacete) (19901991). Nitrogen fertilized (no nitrogen stress). Wheat variety: BETRES Observed data Sowing (day) Seeds m-2 Plants m-2 End spike growth (day) Physiological maturity (day)

Irrigation 349 500 350 136 181

Rainfed 349 400 210 136 181

Grain yield (kg ha-1)

7165

1842

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Weather Single year files, with the extension ". w**" (** indicates the last two digits of the year, i.e. 88 for 1988). The following is an example of the weather file "tome00112.w88" for Tomejil (1988): TOME 37.40 -5.80 12.07 .00 TOME 88 1 4.35 16.1 7.8 4.3 .00 TOME 88 2 7.44 15.6 10.0 .5 .00 TOME 88 3 6.10 13.3 5.6 .0 .00 TOME 88 4 9.07 12.8 2.2 .0 .00 TOME 88 5 3.76 9.4 2.8 3.3 .00 TOME 88 6 9.61 11.7 2.2 .0 .00 TOME 88 7 3.80 8.9 .6 18.3 .00 TOME 88 8 3.85 9.4 5.0 5.3 .00 First line: four letter code for the station, latitude, longitude, PAR conversion. First column: four letter code of the station Second column: last two digits of the year (88 for 1988) Third column: day of year (1 to 365) Fourth column: Solar radiation (MJ m-2 day-1)

Use of DSSAT Models for Climate Change Impact Assessment

143

Fifth column: maximum temperature (oC) Sixth column: minimum temperature (oC) Seventh column: precipitation (mm) Eighth column: no meaning (.00) The following table presents monthly means of temperature, precipitation and solar radiation in Tomejil and Las Tiesas during the years of the field experiments. Table 5. Monthly means of temperature (0C), precipitation (mm) and solar radiation (MJ m-2 day-1) in Tomejil (Sevilla) (+37.40oN; -5.80oW) (1987-91) and Las Tiesas (+38.95oN;-1.85oW) (1990-1991) Site/Year

Month

Temp oC

Precip mm

Solar Rad MJ m-2 day

Tomejil/1987

1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8

8.6 10.8 14.5 16.8 19.7 24.3 26.4 26.7 26.1 17.4 12.5 12.4 8.8 9.8 12.8 12.2 21.3 23.8 27.4 26.3 25.4 18.9 14.5 9.3 9.0 11.4 13.7 13.7 19.4 24.2 29.0 29.2

126.9 92.1 11.7 65.0 4.7 0.3 45.8 31.3 9.0 107.2 86.9 238.7 47.9 101.8 39.4 48.4 14.5 17.5 1.0 0.0 0.0 132.0 118.0 0.0 33.0 64.0 19.0 55.0 18.0 0.0 0.0 0.0

8.4 11.0 16.2 19.8 25.6 28.9 27.0 22.1 19.1 10.9 9.6 6.7 8.3 8.6 16.4 18.4 23.8 26.9 26.1 25.2 17.0 13.4 10.4 6.8 10.3 12.0 13.4 16.0 23.0 24.7 24.9 21.9

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tomejil/1988

Tomejil/1989

144

Ana Iglesias Table 5. (Continued)

Site/Year

Tomejil/1990

Tomejil/1991

Month 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

LasTiesas/1990

LasTiesas/1991

1 2 3

Temp oC 24.7 19.5 13.4 8.7 8.8 9.5 12.7 12.1 21.2 23.7 27.3 26.5 24.7 19.5 12.9 9.3 8.1 7.6 13.4 14.1 17.8 23.6 27.0 28.8 24.7 19.5 13.4 8.7 4.0 9.0 8.8 9.2 14.1 20.3 22.5 23.9 21.0 14.3 8.6 4.1 4.2 5.7 9.5

Precip mm 20.3 43.2 49.5 31.8 47.9 93.7 47.5 48.4 14.5 17.5 1.0 0.0 20.3 90.4 71.5 15.3 12.8 113.5 99.5 40.5 21.0 17.0 0.0 0.0 20.3 43.2 49.5 31.8 30.3 58.0 31.4 67.2 51.6 16.2 31.4 1.8 55.0 36.8 17.2 11.6 8.2 16.1 51.2

Solar Rad MJ m-2 day 17.9 12.8 10.9 8.3 9.0 11.2 15.0 18.0 21.8 23.9 24.7 22.3 17.6 13.4 9.8 7.7 8.8 11.4 15.4 19.0 22.1 25.0 26.0 22.8 18.3 13.8 10.4 8.3 8.1 10.8 12.8 15.2 17.8 23.5 22.6 22.2 14.0 11.8 8.4 6.0 8.6 11.4 15.2

Use of DSSAT Models for Climate Change Impact Assessment Site/Year

Month 4 5 6 7 8 9 10 11 12

Temp oC 9.6 13.0 21.2 23.6 23.7 20.2 12.0 7.9 5.9

Precip mm 94.5 30.1 71.0 48.8 18.2 29.2 30.8 62.6 13.0

145

Solar Rad MJ m-2 day 16.6 20.4 24.2 23.4 21.6 17.0 11.8 9.0 5.9

Soils An accurate description of the soil profile is ESSENTIAL in the case of water-limiting simulations. The characteristics of the soil profile are described in the input file "sprofile.wh2", and include: albedo, soil drainage, limits of water content for each layer (lower limit, drained upper limit, field capacity, etc), pH, organic mater, nitrogen content. The following information can be used as a guideline in the elaboration of the soil input file:

Soil surface Albedo (SALB) Appropriate values for SALB can be obtained from the colour of the soil surface layer as according to the following table. When colour is not known, use a default value of 0.13. If the soil is sandy, slightly higher values may be used (up to 0.17). If there is substantial organic matter present, lower values to 0.10 should be used. Table 6. Values of SALB according to soils colour

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Colour Brown Red Black Grey Yellow

SALB 0.13 0.14 0.09 0.13 0.17

First Stage Evaporation Coefficient (U)

Generally in the range 5 to 12 mm/day. Values are determined from texture of the surface horizon. Typical values are presented in the following table. Table 7. Values of first stage evaporation coefficient of the soil according to soil texture Texture Coarse textured (sandy) Medium Textured (loams) Medium to Heavy Textured soils (30 to 50% clay)

Value 5-8 8 - 11 10 - 12

146

Ana Iglesias

Whole Profile Drainage Rate Coefficient (SWCON) This is a zero to unity number which reflects the rate of drainage from the layer in the profile which most impedes drainage. Suitable values can be obtained from drainage class information used in soil classification is presented in the following table. Table 8. Values of SWCON according to soil drainage class Drainage Class Excessively Somewhat Excessively Well Moderately Well Somewhat poorly Poorly Very Poorly

SWCON 0.8 0.8 0.6 0.4 0.2 0.05 0.005

If drainage information is not available, a value of 0.5 could be used but with caution.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Runoff Curve Number This coefficient which has a value between 60 and 100 is used in runoff calculations. It is based on the USDA Soil Conservation Service Runoff Curve Number technique for estimating runoff. This technique recognizes four soil groups. Soil Group: Description A. Lowest Runoff Potential. Includes deep sands with very little silt and clay, also deep, rapidly permeable loses. B. Moderately Low Runoff Potential. Mostly sandy soils less deep than A, and losses less deep or less aggregate than A, but the group as a whole has above-average infiltration after thorough wetting. C. Moderately High Runoff Potential. Comprises shallow soils and soils containing considerable clay and colloids, though less than those of group D. The group has below-average infiltration after pre-saturation. D. Highest Runoff Potential. Includes mostly clays of high swelling percent but the group also includes some shallow soils with nearly impermeable sub-horizons near the surface. Slope also affects runoff curve number greatly. Given the above soil groups, SWCON can be estimated using slope information presented in the following table. Table 9. Definition of soil groups according to runoff potential Soil Group/ Slope A B C D

0 to 5 % 64 76 84 87

5 to 10 % 68 80 88 91

> 10% 71 83 91 94

Use of DSSAT Models for Climate Change Impact Assessment

147

Lower Limit Volumetric Moisture Content of Layer L (LL) This is the lowest limit to which plants can extract water in a soil layer. The units of measure are volume fraction of soil. The range is 0.02 to 0.50. If this is not known, it can be reliably estimated from soil texture information. The INPUTS program will estimate lower limit from sand, silt clay and organic matter. Further description of lower limit can be found in Ritchie (1981). Drained Upper Limit Moisture Content of Layer L (DUL) This refers to the volumetric moisture content which occurs after a wetted soil drains. This moisture content defines the upper limit of water availability in the soil. It has values in the range 0.10 to 0.60. If values are not known, they can be estimated from soil texture information (see LL). Field Saturated Moisture Content (SAT) This refers to the volumetric moisture content of a soil layer at saturation. Typical values can be estimated from soil texture information (see LL). Rooting Preference Function (WR) The root distribution weighing factor (WR) is used to estimate the relative root growth in all soil layers in which roots actually occur. In deep well-drained soils with no chemical or physical barriers to root growth, the following equation can be used to estimate WR for any soil layer: WR(I) = EXP (-4.*Z(I)/200.) where Z(I) is the depth (cm) to the center of the layer I. In the top soil layer, WR can be set to 1.0. The user should reduce WR(I) to reflect physical or chemical constraints on root growth in certain soil layers. For example, WR(I) could be reduced to half the value estimated from the preceding equation when soil strength or aluminium toxicity produces moderate restrictions in root growth. When these constraints are severe, calculated values of WR(I) can be reduced by 80% to 90%.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Soil pH Soil pH as measured in a 1:1 soil water slurry. Default value is 7.0. Nitrogen Measurements The soil nitrogen concentration (nitrate and ammonium) have to be included if the model is run with the nitrogen balance routine. The % of organic carbon is also necessary in this case because its value initialises the soil organic nitrogen pools. In all our field experiments the nitrogen levels were adequate, no nitrogen stress occurred and therefore the nitrogen balance was NOT be simulated. Initial Soil Water Content It is essential to specify the initial soil water content in water-limiting simulations. This parameter is imputed in FILE 5. In Tomejil the initial soil water in 6 Dec 1988 and 29 Nov

148

Ana Iglesias

1990 was equal to field capacity. The water balance routine of the CERES model was run starting two months before sowing to check if this value was correct. In Las Tiesas the initial soil water content in 15 December 1990 was 70% of the field capacity. Table 10. Description of the soil profile "Tomejil" (order VERTISOL, suborder XERERTS, group CHROMOXERERTS, subgroup ETNIC, series CARMONA; deep clay). See methods for description of the parameters. Values that refer to water content in each layer were calculated from texture data TOMEJIL, DEEP CLAY SALB= .11 U= 10.5 CM LL 0 - 10 .215 10 - 25 .216 25 - 50 .218 50 - 80 .221 80 - 110 .225 110 - 140 .229 140 - 170 .231 170 - 200 .231

SWCON= .40 DUL .361 .361 .361 .361 .360 .360 .360 .360

RUNOFF CURVE NO.= 85 SAT WR .416 1.000 .415 .819 .412 .607 .412 .407 .409 .247 .407 .135 .407 .000 .405 .000

PH 7.9 7.9 7.7 7.7 7.7 7.7 7.7 7.7

Table 11. Description of the soil profile "Las Tiesas". See methods for description of the parameters. Values that refer to water content in each layer were calculated from texture data

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

LAS TIESAS, SHALLOW LOAM-CLAY SALB= .13 U= 8.5 SWCON= .20 CM LL DUL 0 -10 .215 .361 10 -20 .216 .361 20 -30 .218 .361 30 -40 .221 .361 40 -55 .225 .360

RUNOFF CURVE NO.= 84 SAT WR .416 1.000 .415 .819 .414 .607 .412 .407 .409 .247

PH 7.2 7.2 7.2 7.5 7.5

Genetic Coefficients There are a number of coefficients that can be adjusted in the CERES-Wheat model. The "genetic coefficients" describe the phenology and grain yield components of a particular variety, they are located in the file "genetics.wh9"; the calibration of these coefficients is described below. The "phillochron interval" is located in the experimental input file with the extension ".wh8". The phillochron interval (in degree days) is used to determine the rate at which leaves appear. It will also affect the time between terminal spikelet and anthesis. It can vary between 75 and 110 degree days. If experimental data are available they should be used. In other cases a typical value of 95 should be used as general value for most varieties and areas. A value of

Use of DSSAT Models for Climate Change Impact Assessment

149

75 should be used for spring sown wheat in upper latitudes when the mean daily temperature is below 5oC at the time of germination and emergence. A number of coefficients are fixed internally in the CERES-Wheat model (i.e. P2O optimal photoperiod for development = 20 hours) that are in general standard for all wheat varieties. There are six coefficients that need to be adjusted to calibrate the model for each wheat variety in a particular climatic area. These coefficients are scalar values that are converted into physiological meaning values within the model.

P1V - Vernalization Coefficient "Relative amount that development is slowed for each day of unfulfilled vernalization, assuming that 50 days of vernalization is sufficient for all cultivars". This coefficient reflects the differing vernalization requirements of varieties. The input value is a 0 to 9 scalar which is used internally within the model to compute the required number of vernalizing days. The following table can be used as a guide. Table 12. Values of PV1 and genetic material P1V 1

GENETIC MATERIAL No vernalization requirement. True spring wheats (eg. Mexipak,Anza)

3 4

Intermediate types Many winter wheats from Western Europe and the great Plains of North America

6 7 8

Most winter wheats (eg. Arthur, Maris Huntsman) Some wheats from Northern Europe Very Long duration high vernalization materials

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

P1D - Photoperiod Coefficient "Relative amount that development is slowed when plants are grown in a photoperiod 1 hour shorter than the optimum (which is considered 20 hours)." This coefficient is used to describe the sensitivity of varieties to photoperiod. It is input as a scalar value between 1 and 5 which is used internally within the model to scale the rate at which development to terminal spikelet occurs. Use 1 for an insensitive variety and 5 for a highly sensitive variety. P5 - Grain Filling Duration Coefficient "Relative amount of degree days above a base of 1oC that are needed from 20 after anthesis to maturity". This is a 1 to 5 scalar which is used internally within the model to alter the duration from anthesis to physiological maturity. G1 - Kernel number coefficient A scalar value of 1 to 5 that indicates the relative kernel number per unit weight of stem (less leaf blades and sheaths) plus spike at anthesis (kernel number g-1). G2 - Kernel weight coefficient A scalar value from 1 to 5 that indicates the relative kernel filling rate under optimum conditions (mg day-1). G3 - Spike number coefficient

150

Ana Iglesias

A scalar value from 1 to 5 that indicates the relative amounts of non-stressed dry weight of a single stem (less leaf blades and sheaths) and spike when elongation ceases (g).

Calibration in Tomejil, Sevilla (Spain) 1988-89 For the calibration we selected the variety ANZA because it represents the medium cycle wheat varieties grown in the area and it is generally included in all wheat tests performed in Spain, therefore there are many experimental data related to it in other regions. First we calibrated the coefficients related to phenology and then the coefficients related to the grain filling characteristics. Phenology: P1V, P1D and P5 We analyzed the sensitivity of the crop biological responses to changes in the coefficients that relate to phenology. The simulated dates of the phenological stages, and therefore the number of days available for accumulation of grain dry matter, are most sensitive to the photoperiod coefficient (P1D). The sensitivity of the predicted phenology to changes in the vernalization coefficient (P1V), greatly depends on the value of the photoperiod coefficient (P1D). For a particular combination of P1D and P5, the physiological maturity is more sensitive to increases in P1V than the anthesis date. It is important to notice that for certain values of P1D there is an apparent threshold of P1V. The grain filling duration coefficient (P5) does not have any effect on the flowering date, but for values of P5 above 1.5 there is an increase in the number of days between emergence and physiological maturity. Increases in P5 increase the grain filling period. The coefficients P1V, P1D and P5 were calibrated so the observed and simulated phenological dates were as close as possible: P1V = 3.5 P1D = 2.8 P5 = 4.0 The following table compares observed and simulated phenological dates with this combination of coefficients.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 13. Dates of emergence, flowering and physiological maturity observed and simulated for ANZA wheat in Tomejil (1988-1989). 1 = 1 January. Sowing date 7 December = Day 341) 1988-89 Emergence date End of spike growth date Anthesis date Phys. Maturity date Anthesis to maturity (days)

Observed 358 103 107 147 40

Simulated 357 102 107 145 38

Yield components: G1, G2 and G3 Once the phenology coefficients were calibrated, and therefore the simulated number of days available for grain filling, we adjusted the yield component coefficients to represent as accurately as possible the number of spikes m-2, the weight spike-1 (from kernel only) and the

Use of DSSAT Models for Climate Change Impact Assessment

151

final grain yield (kg ha-1). The following table shows some of the combinations tested for the adjustment of these coefficients.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 14. Sensitivity of the final yield and number of spikes to changes in G1, G2 and G3. Waterlimited production. Tomejil 1988-89, sowing 7 December 1988 (day 341), 310 plants m-2, nitrogen non-limiting. Observed data: kg ha-1 = 5964; spikes m-2= 550. P1V=3.5; P1D=2.8; P5=4.0. G1

G2

G3

kg ha-1

spikes m-2

2.9 3.5 3.9 4.5 4.0 4.1 4.2 4.1 4.1 4.1 4.1 4.1 4.1 4.1 4.0 4.2 4.1 4.1 4.1 3.9 3.8 3.9 3.9 3.9 3.9 3.9

3.0 2.7 2.8 1.5 3.0 3.2 3.5 3.5 3.9 3.9 4.1 4.1 4.5 3.0 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 3.9 4.1 3.0 3.5

1.7 4.4 2.3 2.5 2.0 2.3 2.5 2.3 2.3 2.5 2.3 2.5 2.3 2.3 2.3 2.3 3.0 3.5 4.0 2.3 2.3 2.5 3.0 3.0 4.4 4.0

5499 5230 5595 5372 5669 5700 5600 5769 5992 5992 5816 5816 5883 5522 5787 5754 5903 5903 5903 5671 5555 5671 5671 5847 5556 5542

705 333 550 514 616 550 514 550 550 514 550 514 550 550 550 550 445 395 357 550 550 514 445 445 333 357

We selected the following values as representative: G1 = 4.1 G2 = 3.5 G3 = 2.3 The following table shows grain yield and spike number observed and simulated with these coefficients.

152

Ana Iglesias Table 15. Observed and simulated yield and spike number

The following table shows the field data and simulated data for the calibrated wheat in Sevilla. Table 16. Dates of emergence, flowering and physiological maturity, and spike number and final yield observed and simulated for ANZA wheat in Sevilla (1988-1989). 1 = 1 January. Sowing date 7 December = Day 341)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1988-89 Emergence date End of spike growth date Anthesis date Phys. Maturity date Anthesis to maturity (days) Grain yield (kg ha-1) Spike number (spikes m-2)

Observed 358 103 107 147 40 5946 602

Simulated 357 102 107 145 38 5992 550

Potential Production Since there are not experiments with full irrigation in this region, the potential production in Tomejil was estimated from the maximum production reported in the area under meteorological conditions that did not imply water stress during any part of the crop cycle (RAEA, 1991): 8500 to 9000 kg ha-1. Simulated phenological and yield responses as result of changes in the genetic coefficients. The coefficients not shown had their previously adjusted value. Waterlimited production is much more sensitive to changes in the genetic coefficients than potential production. Therefore we suggest that water-limited production should be included in model calibration. It is important to notice that none of the possible combinations of G coefficients in the CERES-model resulted in yields of 9000 kg ha-1 with the meteorological conditions of Tomejil and in the grain filling period fixed by the observations. It seems that the model may limit grain filling at high temperatures in a way that does not represent the observations in the area.

REPRESENTATIVE VARIETIES The coefficients that define a wheat variety in the CERES model only refer to dates of development and accumulation of dry matter; many other variety characteristics are not defined by these coefficients (such as drought resistance, pest and disease resistance, etc). Therefore a particular set of coefficients may be representative of a group of varieties of

Use of DSSAT Models for Climate Change Impact Assessment

153

similar characteristics in a particular geographical area. In particular, the variety BETRES can be defined with the same coefficients than ANZA in the CERES model. The following table shows the coefficients that define a representative wheat variety of medium cycle used in southern and central Spain (ANZA-type). Table 17. Set of genetic coefficients for an ANZA-type variety P1V 3.5

P1D 2.8

P5 4.0

G1 4.1

G2 3.5

G3 2.3

Table 18. Calibration and validation of the CERES-Wheat model in Tomejil, Sevilla

Variable Sowing (DOY) Emergence (DOY) Anthesis (DOY) Phys. Mat. (DOY) Anth. to Phys. Mat. (days) Grain Yield (kg ha-1) Water limited Grain Yield (kg ha-1) Potential Spikes m-2

SEVILLA CALIBRATION (2) RAEA89 (1988-89) OBS SIM 341 341 358 357 107 107 147 145 40 38 5946 5992 8150 7181 542 550

VALIDATION RAEA91 (1990-91) OBS 333 348 108 149 41 6013 8350

SIM 333 347 110 150 40 6769 8258

(1)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Calibrated genetic coefficients for Rothamsted: P1V=6.0; P1D=3.2; P5=7.0; G1=4.7; G2=4.2; G3=3.0. (2) Calibrated genetic coefficients for Sevilla (ANZA VARIETY): P1V=3.5; P1D=2.8; P5=4.0; G1=4.1; G2=3.5; G3=2.3. (*) POTENTIAL PRODUCTION (NO LIMITATIONS OF WATER OR NITROGEN).

Validation in Tomejil, Sevilla The following table shows observed and simulated data from a wheat experiment in Tomejil in 1990-1991. There is a reasonable adjustment between the observed and the simulated data. Validation in Las Tiesas, Albacete The following table shows observed and simulated data from a wheat experiment in Las Tiesas in 1990-1991; this experiment included potential production. The wheat variety used in the experiment was BETRES. In this site, cooler than Tomejil, potential production seems to be more accurately simulated.

154

Ana Iglesias Table 19. Dates of emergence, flowering and physiological maturity observed and simulated in Tomejil (1990-1991) for the ANZA variety. 1= 1 January. Sowing 29 November 1990 (day 333)

1990-91 Emergence date Anthesis date Phys. Maturity date Anthesis to maturity (days) Grain yield (kg ha-1)

Observed 348 108 149 41 6013

Simulated 347 110 150 40 6769

Table 20. Observed and simulated physiological maturity date and final grain yield under waterlimited and potential conditions in Las Tiesas (1990-1991) for BETRES wheat variety. 1= 1 January. Sowing 15 December 1990 (day 349) 1990-91 Phys. Maturity date Potential production (kg ha-1) Water-limited production (kg ha-1)

Observed 179 7165 1848

Simulated 179 7449 2507

Other Calibrated Varieties CERES-Maize Model

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 21. Summary of orientative genetic coefficients calibrated by Ana Iglesias in different sites Genotype Fidel (France) Fidel * Arminda (France) Arminda * Spring N. Europe* Winter N. Plains* Winter S. Plains* Winter Europa* Anza* Sevilla ** Rothamsted **

P1V 4.5 0.5 6.0 6.0 0.5 6.0 4.0 6.0 0.5 3.5 6.0

P1D 3.0 3.5 4.5 2.7 3.5 2.5 3.0 3.5 3.4 2.8 3.2

P5 4.5 2.5 4.5 2.0 2.5 2.0 2.5 4.0 2.0 4.0 7.0

G1 4.5 4.0 4.6 4.3 4.0 4.0 3.0 4.0 3.5 4.1 4.7

G2 1.5 3.0 1.2 4.6 3.0 2.0 3.0 3.0 2.7 3.5 4.2

G3 2.5 2.0 1.7 1.9 2.0 1.5 2.0 2.0 4.4 2.3 3.0

CERES-MAIZE MODEL A crop model with an embedded water-balance model (CERES-Maize, Jones and Kiniry, 1986) was calibrated and validated with experimental field data at two sites that represent contrasting agro-climatic conditions in the Mediterranean Region (Albacete in the Central Plateau and Sevilla in the Guadalquivir Valley (Spain)). The low precipitation during the crop

Use of DSSAT Models for Climate Change Impact Assessment

155

growing season in these regions (less than 100 mm), makes irrigation a necessity (Bignon, 1990; Minguez and Iglesias, 1995). Because evapotranspiration (ET) constitutes an important component of the hydrologic balance and therefore its accurate calculation is essential, the calibration also included the adjustment of the ET calculation of the CERES-Maize model.

Model Description The CERES-Maize model (Jones and Kiniry, 1986) is a simulation model for maize that describe daily phenological development and growth in response to environmental factors (soils, weather and management). Modelled processes include phenological development, i.e. duration of growth stages, growth of vegetative and reproductive plant parts, extension growth of leaves and stems, senescence of leaves, biomass production and partitioning among plant parts, and root system dynamics. The model includes subroutines to simulate the soil and crop water balance and the nitrogen balance, which include the capability to simulate the effects of nitrogen deficiency and soil water deficit on photosynthesis and carbohydrate distribution in the crop.

Development The primary variable influencing phasic development rate is temperature. The thermal time for each phase is modified by coefficients that characterize the response of different genotypes. The timing of crop phenological stages can be calibrated by modifying the coefficients that characterize the duration of the juvenile phase (P1), photoperiod sensitivity (P2), and duration of the reproductive phase (P5).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Dry Matter Production Potential dry matter production is a linear function of intercepted photosynthetically active radiation (PAR). The percentage of incoming PAR intercepted by the canopy is an exponential function of leaf area index (LAI). The dry matter allocation is determined by partitioning coefficients which depend on phenological stage and degree of water stress. Final grain yield is the product of plant population, kernels per plant and kernel weight. The number of kernels per plant is a linear function of stem weight (at anthesis) and coefficients that accounts for the variation between genotypes in potential kernel number (G2) and kernel growth rate (G3). Water Balance Precipitation is a daily input; runoff is a function of soil type, soil moisture and precipitation; infiltration is precipitation minus runoff; drainage occurs when soil moisture is greater than the soil water holding capacity of the bottom layer. Potential evapotranspiration is calculated by the Priestley-Taylor relation. Actual transpiration is modified by LAI, soil evaporation and soil water deficit. Actual evaporation is a function of potential evaporation, LAI and time as described by Ritchie (1972). Daily change in soil moisture is calculated as precipitation minus evaporation minus runoff minus drainage.

156

Ana Iglesias

Carbon Dioxide Sensitivity The CERES-Maize model has been modified to simulate changes in photosynthesis and evapotranspiration caused by higher CO2 levels. These modifications have been based on published experimental results (see Rosenzweig and Iglesias (1994) for a description of the methodology). Input Data The model requires daily values for solar radiation, maximum and minimum temperature and precipitation. Soil data needed are values for the functions of drainage, runoff, evaporation and radiation reflection, soil water holding capacities and rooting preference coefficients for each soil layer, and initial soil water contents.

Site and Field Experimental Data Input data for the calibration and validation process were obtained from published field experiments conducted at the Agricultural Research Stations of Lora del Rio and Montoro (Sevilla, Spain, +37.42oN, -5.88oW, 31m altitude; Aguilar, 1990; Aguilar and Rendon, 1983), and Las Tiesas and Santa Ana (Albacete, Spain, +38.95oN, -1.85oW, 704m altitude, ITAP, 1985-1993). Maize hybrids selected for the calibration represent highly productive simple hybrids grown in the different agricultural regions. Local daily climate data and soil information for the Sevilla site were provided by the Department of Agronomy of the University of Córdoba (Córdoba, Spain) and for the Albacete site by the Instituto Agronomico Tecnico Provincial de Albacete (ITAP, Albacete, Spain). Daily observed crop evapotranspiration data are from to field experiments in Las Tiesas (Albacete, Spain, Martin de Santa Olalla et al., 1990). In all field experiments the crop was irrigated and nitrogenfertilized to cover total crop requirements.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Calibration of Crop Phenology, Biomass and Yield The model was calibrated and validated with independent field data sets (that included yield components, phenology, and crop ET) for maize hybrids of different crop growth duration. The coefficients that define a maize hybrid in the CERES model only refer to rate of development and accumulation of dry matter; many other hybrid characteristics are not defined by these coefficients (such as drought resistance, pest and disease resistance, etc). Therefore one set of coefficients may be representative of a group of hybrids of similar characteristics grown in a particular geographical area. A set of coefficients were estimated for the most widely used hybrids in Spain and other Mediterranean regions. The table below shows the coefficients that define representative maize hybrids of different crop-cycle duration used in southern Europe. The coefficients were first calibrated in relation to phenology based on the thermal integrals of the juvenile period and of the reproductive period. Once the phenology coefficients were calibrated, and therefore the simulated number of days available for grain filling, the yield component coefficients were adjusted to represent as accurately as possible

Use of DSSAT Models for Climate Change Impact Assessment

157

the number of grains ear-1, the final grain yield (t ha-1), and the final biomass (t ha-1). In the experiment crop nitrogen and water requirements were fully covered and pests and diseases were controlled. Nevertheless, these experimental yields are not potential yields, and they include some effect of losses by diseases and suboptimal management. Observed and simulated data are compared (see table). Crop responses to changes in planting date and density under non-limiting conditions were also analyzed. The ability of the CERES-Maize model to simulate grain yields for long cycle hybrids (700 and 800) is proven. The table below shows the agreement between simulated and observed crop data from a second set of field experiments used for validation.

Calibration of the Water Balance As stated above, potential evapotranspiration is calculated in the CERES model with the Priestley-Taylor relation (Priestley and Taylor, 1972). Potential transpiration is directly related to potential evapotranspiration by a coefficient (alpha) which value is fixed to 1.1 in the CERES-Maize v2.1 (Jones and Kiniry, 1986). In many areas of the Mediterranean region, maximum temperatures over 35oC occur in the summer months of July and August. When short cycle maize crop is sown after another crop in late June or July, the crop is subject to high temperatures before reaching full ground cover. These conditions imply that the advective and micro-advective (between rows) processes occur increasing crop ET. Such conditions were not represented in the original ET formulation of the CERES-Maize model, and therefore, simulations of crop ET with the original model underestimated field-observed values. When advective conditions prevail the mentioned coefficient should be higher (Shouse et al., 1980; Rosenberg et al., 1983; Pereira and Villa-Nova, 1992). The coefficent was set at 1.26 when maximum temperatures were below 35oC and 1.45 above 35oC. The result of these changes is a better estimate of total crop ET in Mediterranean conditions.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 22. Values of the calibrated genetic coefficients used as input for the CERESmaize model. P1: Juvenile phase coefficient. P2: Photoperiodism coefficient. P5: Grain filling duration coefficient. G2: Kernel number coefficient. G3: Kernel weight coefficient Hybrid(1) 200 400 600 700 800 (1)

Thermal units(2) 2000 2000-2300 2500 2800 3000

Cycle length (3) 90-100 100-110 120 130 150

P1 200 200 200 220 260

P2 0.30 0.76 0.70 0.52 0.50

P5 600 750 800 910 980

G2 825 650 800 700 600

G3 9.0 9.0 8.0 7.0 8.5

Hybrids used in the calibration and validation are: FURIO (200), DEMAR (400), LUANA (600), AE703 (700), and PRISMA (800). (2) Accumulated total thermal units during the growing cycle (sum of degree days above 8oC). (3) Average duration of the growing cycle (days) in Albacete and Sevilla.

158

Ana Iglesias

Table 23. Calibration: Comparison of phenology and yield data observed and simulated in Lora del Río (1986, sowing day 75) and in Las Tiesas (1993, sowing day 137). Day 1= January 1 LORA DEL RIO Flowering date

LORA DEL RIO Physiological maturity date

Grain filling period (days)

Grain yield (t ha-1)

LAS TIESAS Flowering date Physiological maturity date Grain filling period (days) Grain yield (t ha-1)

HYBRID 600 700 800 HYBRID 600 700 800 600 700 800 600 700 800 HYBRID 200 400 200 400 200 400 200 400

Observed 175 175 179 Observed 223 223 230 48 48 51 13.26 13.43 15.05 Observed 205 207 245 255 40 48 12.86 13.38

Simulated 169 171 177 Simulated 216 224 232 47 53 55 14.90 14.25 15.74 Simulated 201 204 240 255 39 51 12.64 13.51

Table 24. Validation: Comparison of phenology and yield data observed and simulated in Lora del Río (1987, sowing day 63), Montoro (1981, sowing day 65), Santa Ana (1991, sowing day 151) and Las Tiesas (1991, sowing day 126). Day 1= January 1

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

LORA DEL RIO Flowering date

Physiological maturity date

Grain filling period (days)

Grain yield (t ha-1)

MONTORO

HYBRID 600 700 800 600 700 800 600 700 800 600 700 800 HYBRID

Observed 162 166 170 210 220 224 48 54 54 15.05 14.21 15.63 Observed

Simulated 156 158 164 203 212 220 47 54 56 16.21 14.51 16.50 Simulated

Use of DSSAT Models for Climate Change Impact Assessment Flowering date

700 800 700 800 HYBRID 400 HYBRID 600 700 HYBRID 400 400

Grain yield (t ha-1) SANTA ANA Grain yield (t ha-1) LORA DEL RIO

LAS TIESAS Flowering date Grain yield (t ha-1)

168 169 16.36 16.27 Observed 11.54 Observed 13.97 15.07 Observed 217 9.94

159

165 168 16.19 16.76 Simulated 11.20 Simulated 14.84 15.06 Simulated 212 9.83

Table 25. Soils and crop management variables

Site Sevilla Badajoz Albacete Lérida Zamora

Soil sandy loam sandy loam silty loam silty clay sandy loam

Wheat Variety ANZA ANZA ANZA MARIUS MARIUS

Sowing 1 Dec 1 Dec 1 Dec 15 Nov 1 Nov

Maize Hybrid 700-800 700-800 700 700 500-600

Sowing 15 Mar 30 Mar 15 May 15 May 15 May

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Aguilar, M. (1990). Influencia de la densidad de plantas en crecimiento, rendimiento y calidad de grano de tres cultivares de maiz (Zea mais L.), ciclos 600, 700, 800 FAO, en el Valle Medio del Guadalquivir. PhD. Thesis, University of Cordoba. Cordoba (Spain). Aguilar, M. and Rendón, M. (1983). El Cultivo del Maíz en Regadíos de Climas Cálidos. Ministerio de Agricultura, Pesca y Alimentación. HD Num. 1/83. Bignon, J. (1990). Agrometeorology and the physiology of maize. Publication: EUR 13041 EN. Office for Official Publications of the EC, Series: An Agricultural Information System for the EC. Luxembourg. Bignon, J. 1990. Agrometeorology and the physiology of maize. Publication: EUR 13041 EN. Office for Official Publications of the EC, Series: An Agricultural Information System for the EC. Luxembourg. Carter, T.R., Parry, M.L. and Porter, J.R. (1991a). Climatic change and future agroclimatic potential in Europe. Int. J. Climatol., 11, 251-269. Carter, T.R., Porter, J.R. and Parry, M.L. (1991b). Climatic warming and crop potential in Europe: Prospects and uncertainties. Global Environmental Change, 1, 291312. Dale, R.F., Coelho, D.T. and Gallo, K.P. (1980). Prediction of daily green leaf area index for corn. Agron. J., 72, 999-1005. Elena-Rosselló, R., Tella-Ferreiro, G., Allué-Andrade, J.L. and Sanchez-Palomares, O. (1990). Clasificación Biogeoclimática Territorial de España: Definición de Eco-regiones. Ecología, Fuera de Serie N. 1, 59-79.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

160

Ana Iglesias

Font-Tullot, I. (1983). Climatología de España. Instituto Nacional de Meteorología. MOPT. España. Godwin, D., Ritchie, J., Singh, U. and Hunt, L. 1981. A User's Guide to CERES WheatV2.10. International Fertilizer Development Center. Simulation Manual IFDCSM-2. Muscle Shoals, AL. Goudriaan, J. Unsworth, M.H. (1990). Implications of increasing carbon dioxide and climate change for agricultural productivity and water resources. In Kimball, B.A., N.J Rosenberg, L.H. Allen Jr. (eds.) Impact of Carbon dioxide, Trace Gases, and Climate Change on Global Agriculture. ASA Special Publication Number 53. American Society of Agronomy, Inc. pp. 71-82. Iglesias, A. and Mínguez, M.I. (1995). Perspectives for maize production in Spain under climate change. In L. Harper, S. Hollinger, J. Jones and C. Rosenzweig. Climate Change and Agriculture. ASA Special Publication. American Society of Agronomy. Madison, WI. Imerson, A., Dumont, H. and Sekliziotis, S. (1987). Impact Analysis of Climatic Change in the Mediterranean Region. Volume F: European Workshop on Interrelated Bioclimatic and Land Use Changes. Noordwijkerhout, The Netherlands, October 1987. ITAP (1985-1993). Boletines Monográficos de Resultados de los Ensayos de Variedades de Cereales. Instituto Técnico Agrónomico Provincial, S.A. Albacete. ITAP. 1991. Resultados de los Ensayos de Variedades de Cereales. Instituto Tecnico Agronomico Provincial. Albacete. Boletin Monografico N. 11. Jones, C.A. and Kiniry, J.R. (eds.) (1986). CERES-Maize: A Simulation Model of Maize Growth and Development. Texas AandM University Press. College Station, TX. 194 pp. Kenny, G.J. and Harrison, P.A. (1992). Thermal and moisture limits of grain maize in Europe: model testing and sensitivity to climate change. Climate research, 2, 113-129. Körner, C. (1990). CO2 fertilization: the great uncertainty in future vegetation development. In: A. Salomon and D. Reidel. Global Vegetation Change. Hingham, Mass. Le Houerou, H.N. (1990). Global change: vegetation, ecosystems and land use in the Mediterranean basin by the mid twenty-firs century. Israel J. Bot., 39, 481-508. López-Bellido, L. (1991). Cultivos Herbaceos, Vol. I: Cereales. Mundi-Prensa, Madrid. MAPA. 1993. Manual de Estadística Agraria. Secretaría General Técnica. Servicio de Publicaciones del Ministerio de Agricultura, Pesca y Alimentación de España. Madrid, Spain. Martín de Santa Olalla, de Juan Valero, F.A. and Tarjuelo Martin-Benito, J.M. (1990). Respuesta al Agua en Cebada, Girasol y Maíz. Instituto Técnico Agronómico Provincial, S.A. y Universidad de Castilla-La Mancha. Albacete. Pereira, A.R. and Villa-Nova, N.A. (1992). Analysis of the Priestley-Taylor parameter. Agric. For. Meteorol., 61, 1-9. Priestley, C.H.B. and Taylor, R.J. (1972). On the assessment of surface heat flux and evaporation using large scale parameters. Monthly Weather Review, 100, 8192. RAEA (1989). Variedades de Trigos Campaña 88/89. Red Andaluza de Experimentacion Agraria. Junta de Andalucia, Consejeria de Agricultura y Pesca, Direccion General de Investigacion y Extension Agrarias. Sevilla. RAEA (1991) Variedades de Trigos Campaña 90/91. Red Andaluza de Experimentacion Agraria. Junta de Andalucia, Consejeria de Agricultura y Pesca, Direccion General de Investigacion y Extension Agrarias. Sevilla.

Use of DSSAT Models for Climate Change Impact Assessment

161

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Richardson, C.W. and Wright, D.A. (1984). WGEN: A Model for Generating Daily Weather Variables. ARS-8. U.S. Department of Agriculture, Agricultural Research Service. Washington, DC. 83 pp. Ritchie, J. and Otter, S. (1985). Description of and performance of CERES-Wheat: A useroriented wheat yield model. In Willis, W.O. (ed). ARS Wheat Yield Project. Department of Agriculture, Agricultural Research Service. ARS-38. Washington, DC. pp. 159-175. Ritchie, J.T. (1972). Model for predicting evaporation from a row crop with incomplete cover. Water Resources Research, 8, 1204-1213. Ritchie, J.T. 1981. Soil Water Availability. Plant and Soil 58, 327. Rosenberg, N.J., Blad, B.L. and Verma, S.B. (1983). Microclimate, The Biological Environment. John Wiley and Sons. New York. Rosenzweig, C. and Iglesias, A. (eds). (1994). Implications of climate change for international agriculture: Crop modeling study. U.S. Environmental Protection Agency. Washington DC. Santer, B. (1985). The use of General Circulation Models in climate impact analysis- a preliminary study of the impacts of a CO2- induced climatic change on Western European Agriculture. Climatic Change, 7, 71-93. Shouse, P., Jury, W.A. and Stolzy, L.H. (1980). Use of deterministic and empirical models to predict potential evapotranspiration in an advective environment. Agronomy Journal, 72, 994-998. Watts, W.R. (1974). Leaf Extension in Zea mays. III. Field measurements of leaf extension in response to temperature and leaf water potential. J. Exp. Bot., 25, 1085-1096.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PART II. APPLYING CLIMATE AND CROP-GROWTH MODELLING TOOLS TO SUPPORT AGRICULTURAL DECISION-MAKING

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 12

OPPORTUNITIES AND CHALLENGES OF USING MODELLING TOOLS FOR AGRICULTURAL DECISION-MAKING UNDER CLIMATE CHANGE CONDITIONS Angel Utset1∗ Agrarian Technological Institute of Castilla y León (ITACyL), Valladolid, Spain

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CLIMATE SCENARIOS, SEASONAL FORECASTS AND DOWNSCALING ISSUES Complex Models regarding general atmospheric circulation (GCM) have been developed to predict the future earth climate. Those models are able to simulate the energy and mass exchanges between the atmosphere and the earth surface, according to several man-due scenarios of greenhouse gases emissions (IPCC, 2007). The HadCM3 model, developed by the United-Kingdom Meteorological Office, and the German ECHAM4 model were considered in the IPCC (2007) report, among other non-European GCM’s. On the other hand, seasonal time-scale climate predictions are now made routinely at a number of operational meteorological centers around the world, using comprehensive coupled models of the atmosphere, oceans, and land surface (Stockdale et al. 1998; Mason et al. 1999; Kanamitsu et al. 2002; Alves et al. 2002; Palmer et al., 2004). Particularly, GCM were integrated over 4-month time scales with prescribed observed sea surface temperatures (SSTs) within the PROVOST project (Palmer et al., 2004). Single model and multi-model ensembles were treated as potential forecasts. A key result was that probability scores based on the full multi-model ensemble were generally higher than those from any of the singlemodel ensembles (Palmer et al., 2004). Based on PROVOST results, the Development of a European Multi-model Ensemble System for Seasonal to Inter-annual Prediction project (DEMETER) was conceived, and funded under the European Union 5th Framework Environment Programme (Palmer et al., 1 ∗

Current position: AMBCLIM, Environment and Climate Consultancy, Zaragoza, Spain [email protected]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

166

Angel Utset

2004). The principal aim of DEMETER was to advance the concept of multi-model ensemble prediction by installing a number of state-of-the-art global coupled ocean–atmosphere models on a single supercomputer, and to produce a series of 6-month multi-model ensemble hindcasts with common archiving and common diagnostic software. As a result of DEMETER, real-time multi-model ensemble seasonal global predictions are now routinely made at the European Centre for Medium-Range Weather Forecasts (ECMWF). Palmer et al. (2004) showed some DEMETER applications. Results indicate that the multi-model ensemble is a viable pragmatic approach to the problem of representing model uncertainty in seasonal-to-inter-annual prediction, and will lead to a more reliable forecasting system than that based on any one single model (Palmer et al., 2004). On the other hand, Doblas-Reyes et al. (2006), pointed out the potential of DEMETER predictions of seasonal climate fluctuations to crop yield forecasting and other agricultural applications. They recommend a probabilistic approach at all stages of the forecasting process. The ENSEMBLES EU-funded proposal (Hewitt, 2005) is an important recent effort to improve the skill of seasonal forecasts and to make them available to stakeholders. The ENSEMBLES proposal uses the collective expertise of 66 institutes to produce a reliable quantitative risk assessment of long-term climate change and its impacts. Particular emphasis is given to probable future changes in climate extremes, including storminess, intense rainfall, prolonged drought, and potential climate ‘shocks’ such as failure of the Gulf Stream. To focus on the practical concerns of stakeholders and policy makers, ENSEMBLES considers impacts on timeframes ranging from seasonal to decadal and longer, at global, regional, and local spatial scales. Several useful tools to assess climate-change impacts on agriculture have been developed during the last years. GCM are among such tools. However, GCM estimations of temperature, precipitation and other meteorological variables are usually made for large areas. For instance, Guereña et al. (2001) showed that those estimations are not very useful to Spanish agricultural climate-change impact assessments, due to the notable topographical changes within the Peninsula for relative small distances. Therefore, a “downscaling” of GCM outputs is absolutely needed before using their estimations for agricultural applications. Wilby and Wigley (2001) summarized the available downscaling techniques; which can be classified as statistical, dynamical and weather generators. A dynamical downscaling method is to apply numerical regional climate models at high resolution over the region of interest. Regional models have been used in several climate impact studies for many regions of the world, including parts of North America, Asia, Europe, Australia and Southern Africa (e.g Giorgi and Mearns, 1999; Kattenberg et al., 1996; Mearns et al., 1997). The regional climate models obtain sub-grid scale estimates (sometimes down to 25 km resolution) and are able to account for important local forcing factors, such as surface type and elevation. Particularly, the regional climate model RegCM was originally developed at the National Center for Atmospheric Research (NCAR), USA and has been mostly applied to studies of regional climate and seasonal predictability around the world. It is further developed by the Physics of Weather and Climate group at the Abdus Salam International Centre for Theoretical Physics (ICTP) in Trieste, Italy. The PRUDENCE Regional Models Experiment has been developed in Europe under the EU Framework Research Program (Christensen and Christensen, 2007). PRUDENCE project provides a series of high-resolution regional climate change scenarios for a large range of climatic variables for Europe for the period 2071-2100 using four high resolution GCMs and eight RCMs.

Opportunities and Challenges of Using Modelling Tools…

167

Wilby and Wigley (2001) classified statistical downscaling in regression methods and weather-pattern approaches. The regression method uses statistical linear or non-linear relationships between sub-grid scale parameters and coarse resolution predictor variables. Wilby and Wigley (2001) included Artificial Neural Network within the regression-type statistical downscaling. On the other hand, weather-pattern based approaches involve grouping meteorological data according to a given classification scheme. Classification procedures include principal components, canonical correlation analyses, fuzzy rules, correlation-based pattern recognition techniques and analogue procedures; among others (Wilby and Wigley, 2001). Theoretically, dynamical downscaling methods are better than simple statistical methods since they are based on physical laws. However, statistical downscaling method are less computational exigent and can give good results if the relationships between predictand and predictors are stationary (Wilby and Wigley, 2001).

WEATHER GENERATORS Weather generators have been very used in agriculture climate-change impact assessments (Hoogenboom, 2000; Sivakumar, 2001). A weather generator produces synthetic daily time series of climatic variables statistically equivalent to the recorded historical series, as well as daily site-specific climate scenarios that could be based on regional GCM results (Semenov and Jamieson, 2001). The weather generator usually mimics correctly the mean values of the climatic variables, although underestimates their variability (Mavromatis and Jones, 1998; Semenov and Jamieson, 2001; Wilby and Wigley, 2001). Different weather generators are available, but according to Wilby and Wigley (2001), the US-made and the UK-made WGEN and LARS-WG are the most widely used.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SIMULATING CROP-GROWTH AND CROP WATER-USE On the other hand, many crop simulation models have been appeared in the last 20 years. Those models are able to estimate crop water-use and growth under any weather and cop management conditions. Those models, combined with downscaled GCM scenarios, can be a reliable approach to support decision-making under climate change conditions (Hoogenboom, 2000). Despite many models are available, Mechanistic models i.e. those based in the physical laws of the soil-water-plant-atmosphere continuum, are the most suitable to climatechange impact assessments (Eatherall, 1997), since the laws are, in principle, valid for al climatic conditions. According to Tubiello and Ewert (2002), more than 40 assessments of climate-change impact on agriculture have been published up to now. They pointed out that generally models provided accurate results, compared to actual data. The most used models in such assessments are DSSAT (Jones et al., 2003) and those developed in Wageningen (Van Ittersum et al., 2003). As pointed out above, modelling tools appeared in the eighties, due to computer availability, aimed to simulate crop growth and final yields. Numerous crop growth models have been developed since them. The models can use weather data input, such as short term

168

Angel Utset

weather forecast, a season’s forecasted weather or climate scenarios to estimate potential or actual growth, development or yield. Historical-production records are useful for assessing the impacts of climate variability on crop yields, but cannot reveal crop response under alternative management strategies, which can be done through modelling simulations. Bastiaansen et al. (2004) provided an update revision of the modelling applications to irrigation assessments. They pointed out the opportunities lying in such modelling approaches to irrigation and drainage assessments, with more than 40 examples. The simulation examples comprises assessing irrigation supply needs, as well as irrigation designing, scheduling, management and performance; salt-affected soils due to irrigation, groundwater recharge and estimating soil losses, among others. Models have been usually classified as empirical, functional and mechanistics (Connolly, 1998; Bastiaansen et al., 2004). Mechanistic models, i.e. those based in the physical laws of the soil-water-plan-atmosphere are more suitable to assess climate-change effects on agriculture than empirical models (Eatherall, 1997; Hoogenboom, 2000), since the theoretical mechanistic-model backgrounds is still valid under these new conditions.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CROP-GROWTH ORIENTED SIMULATION MODELS According to Stockle et al. (2003), the initial crop-growth simulation models, mainly theoretical approaches, appeared in the 1970s (de Wit et al., 1970; Arkin et al., 1976). Applications oriented models appeared during the 1980s (Wilkerson et al., 1983; Swaney et al., 1983). Models such as SUCROS and others associated with the Dutch ‘School of de Wit’ (Bouman et al., 1996); as well as those produced in the US as the CERES (Ritchie, 1998) and CROPGRO (Boote et al., 1998) families of models had a significant impact on the crop modelling community (Stockle et al., 2003). As pointed out by Brisson et al (2003), the rest of the crop-growth simulating models, although different; generally follow similar guidelines than the originally produced models. Tubiello and Ewert (2002) reported the most used crop-growth models regarding climatechange impact assessments in agriculture. Furthermore, Alexandrov (2002) provided a complete summary of the crop models that have been used in Europe. At the world level, the DSSAT and the Wageningen models have been the most used (Tubiello and Ewert, 2002). Same conclusion can be drawn in Europe, although CROPSYST model has been significantly used as well (Alexandrov, 2002). The Decision Support System for Agrotechnology Transfer (DSSAT) was originally developed by an international network of scientists, cooperating in the International Benchmark Sites Network for Agrotechnology Transfer project (IBSNAT, 1993; Tsuji et al., 1998; Uehara, 1998; Jones et al., 1998), to facilitate the application of crop models in a systems approach to agronomic research (Jones et al., 2003). DSSAT is a microcomputer software package that contains crop-soil simulation models, data bases for weather, soil, and crops, and strategy evaluation programs integrated with a ‘shell’ program which is the main user interface (Jones et al., 1998). DSSAT originally comprises the CERES models for maize (Jones and Kiniry, 1986) and wheat (Ritchie and Otter, 1985), as well as the SOYGRO soybean (Wilkerson et al., 1983) and PNUTGRO peanut (Boote et al., 1986) models, among others (Jones et al., 2003).

Opportunities and Challenges of Using Modelling Tools…

169

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The decision to make these models compatible led to the design of the DSSAT and the ultimate development of compatible models for additional crops, such as potato, rice, dry beans, sunflower, and sugarcane (Hoogenboom et al., 1994; Jones et al., 1998; Hoogenboom et al., 1999; Jones et al., 2003). According to Hoogenboom et al. (1999), the DSSAT Cropping System Model (CSM) simulates growth and development of a crop over time, as well as the soil water, carbon and nitrogen processes and management practices. As pointed out by Van Ittersum et al. (2003), the Wageningen group has a long tradition in developing and applying crop models in its agroecological research program, based on the pioneering work of C.T. de Wit. In the 1960s and 1970s the main aim of these modelling activities was to obtain understanding at the crop scale based on the underlying processes. De Wit and co-workers at the Department of Theoretical Production Ecology of Wageningen University, and the DLO Research Institute for Agrobiology and Soil Fertility developed the model BACROS and evaluated components of the model (such as canopy photosynthesis) with especially designed equipment and field experiments (De Wit et al., 1978; Goudriaan, 1977; Van Keulen, 1975; Penning de Vries et al., 1974). These modelling approaches have served as the basis and inspiration for modelling groups around the world (Stockle et al., 2003). In the 1980s a wide range of scientists in Wageningen became involved in the development and application of crop models. The generic crop model SUCROS for the potential production situation was developed (Van Keulen et al., 1982; Van Laar et al., 1997), which formed the basis of most recent Wageningen crop models such as WOFOST (Van Keulen and Wolf, 1986), MACROS (Penning de Vries et al., 1989), and ORYZA (Bouman et al., 2001). In the 1990s the Wageningen group focused more on applications in research, agronomic practice and policy making (Van Ittersum et al., 2003). Numerous crop-growth simulation models are now available, with different objectives, and many new models are still appearing. Actually, there is no universal model and it is necessary to adapt system definition, simulated processes and model formalisations to specific environments or to new problems (Van Ittersum et al., 2003). Particularly, the European Society of Agronomy (ESA), has a special session dedicated to such modelling tools. Besides, special numbers of the European Journal of Agronomy have been dedicated to promote such models. The CROPSYST model (Stockle et al., 2003), the French model STICS (Brisson et al., 2003) and the Australian model APSIM (Keating et al., 2003) are examples of such other models.

AGROHYDROLOGICAL MODELS The IPCC report of working group II (IPCC, 2007) pointed out that one of the main Climate Change effects that can be expected in European agriculture regards the water deficit in Southern Europe. Hence, water-related assessments should be made, based on the available simulation tools. According to Bastiaansen et al. (2004), concerning suitability to describe irrigation and drainage processes, models can be classified as bucket, pseudo-dynamics, Richards-equation based, SVAT models, multidimensional and crop-production models. However, at the plot and field scale only the bucket, the crop-oriented models and those based on the Richards equation have been significantly used. The Richards equation describes the

170

Angel Utset

vertical movement of water within the soil profile and its solutions can, at least theoretically, provide the water distribution under certain initial and border conditions (Kutilek and Nielsen, 1994). Therefore, concerning irrigation studies at field and plot scales, the most important mechanistic modelling approaches are those mainly aimed to simulate crop-growth and those addressed to physically-based simulation of soil-water movement, through numerical solutions of the Richards equation. These models have been called agrohydrological models, because they combine agricultural and hydrological issues (Van Dam, 2000). Agrohydrological or water-oriented models were significantly developed during the last years (Bastiaansen et al., 2004). The models SWAP (Van Dam et al., 1997), DRAINMOD (Kandil et al., 1995), WAVE (Vanclooster et al., 1994), ISAREG (Teixeira and Pereira, 1992) and HYDRUS (Vogel et al., 1996) can be considered as agrohydrological models, among others. The unsaturated zone, i.e. the zone between the soil surface and the groundwater, is a complicated system governed by highly non-linear processes and interactions. Flow processes can alternatively be described by means of physical-mathematical models. According to Bastiaansen et al. (2004), unsaturated-zone models can be used to simulate the timing of irrigations and irrigation depths, drain spacing and drain depth, and system behaviour and response. The models have increased our understanding of irrigation and drainage processes in the context of soil–plant–atmosphere systems. Progress in modelling can be attributed to merging separated theories of infiltration, plant growth, evapotranspiration and flow to drain pipes into a single numerical code (Bastiaansen et al., 2004). According to Van Ittersum et al. (2003) agrohydrological models are more suitable to irrigation and water-use assessments than crop-growth oriented models, although both approaches have been used. Applying climate and crop-growth simulation tools to support agricultural decisionmaking. Despite the manifold papers addressed to estimate climate-change and climate-variability effects on agriculture, appeared in the last years, few cases can be referred where such tools have been effectively used to provide recommendations to farmers and stakeholders. Some of the current actions regarding such applications are outlined below.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PREVIOUS EXPERIENCE Probably, the most important contributions to the use of the available simulation tools to support agricultural decision-making are CLIMAG (Climate Prediction and Agriculture) activities. The CLIMAG workshops were held in Geneva 1999 and 2005, sponsored by WMO and IRI (Sivakumar, 2001; Sivakumar and Hansen, 2007). The CLIMAG proceedings remain as significant guidelines for using such tools in future studies. However, the assessments of climate-impacts on agriculture made in the framework of CLIMAG were only specially-funding applications. The CLIMAG assessments did not yield to sustainable applications of the simulation tools in the targeted countries. One of the most important successful applications of climate information, combined to crop-growth simulation models, have been made in Australia (Meinke et al, 2001; Hammer et

Opportunities and Challenges of Using Modelling Tools…

171

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

al., 2001; Meinke et al., 2006). Rainfall and many other meteorological variables strongly depend on El Niño behaviour. Since ENSO occurrence can be forecasted several months in advance, this information can be used to support agricultural decision-making, through cropgrowth modelling exercises (Hammer et al., 2001). A “participatory process” (Meinke et al., 2001) comprising researchers, stakeholders and extension facilities has been pointed out as a way to provide sustainable and sounder support to Australian agriculture. Another important application of climate information and crop-growth models can be found in the South East Climate Consortium (SECC), as described by Hoogenboom et al. (2007). Seasonal forecasts at the county level, using ENSO phases, are combined with DSSAT modelling results in order to provide estimations of final yields, water and fertilizer requirements and several other outputs very useful to farmers. The forecast and the whole system are only reliable in El Niño years, although current researches aim to enlarge the system reliability to other years (Baigorria, 2007). El Niño signal is not very strong over Europe, which limits the applicability of ENSObased forecasts. Marletto et al. (2005) showed that WOFOST simulations of winter wheat yields, based on the available seasonal forecasts, departed significantly from the recorded data. However, using spatial and temporal aggregated data, as usually done by JRC when providing recommendations to the EU Commission of Agriculture, might be not the right approach to capture the relationships between weather variables and crop growing. Hansen et al. (2006) provided an insight view of current advances and challenges while translating climate forecasts into reliable agricultural decision-making. They describe several methods used up to now to spatially and temporal downscale the forecasts, recommending methods comparisons and evaluations at local scales. Likewise, Alexandrov (2007) provided an update revision of current state on applying climate scenarios and seasonal forecasts to support agricultural decision-making. Alexandrov (2007) points out that improved climate prediction techniques are growing faster and finding more applications; hence close contacts between climate forecasters, agrometeorologists, agricultural research and extension agencies in developing appropriate products for the user community are needed. Furthermore, feedbacks from end users are essential identifying the opportunities for agricultural applications (Alexandrov, 2007).

THE ROLE OF LOCAL AGRICULTURAL RESEARCH AND EXTENSION SERVICES As pointed out above, negative climate-variability impacts could be reduced by following adaptation options, which can be obtained from crop-model simulations combined with climate scenarios. The usefulness of such simulation tools has been proved in manifold papers, usually produced in Universities or similar centres. However, despite of the considerable public concern about climate change, European stakeholders and farmers are not yet using these scientific results for agricultural decision-making. Actually, the most reliable climate-change mitigation options depend on each specific situation. While experts and researchers at high-level centres in Europe and other places (“developers”) have established significant Know-How and produced relevant of the abovecited tools for such climate-impacts studies; practical experts at local agricultural-research or

172

Angel Utset

extension centres as well as agricultural advisers (“users”), those who should apply these tools for agricultural decision-making, are often not aware about the available tools or their access to such tools is quite limited due to several reasons, as financial issues or lack of userfriendly design of tools. A connection is needed between the “developers” and “users”, to improve decision making by better implementing the climate and crop-growth simulation model tools. Furthermore, feedback from low end-users to the tool-provider researchers is a prerequisite for improving these tools for their practical use e.g. by providing background information, setting up the actual input data needs, fitting time and spatial scales as required by specific applications and other similar issues.

THE AGRIDEMA PROPOSAL In that context AGRIDEMA, a Specific Support Action (SSA), has been funded by the EU Sixth Framework Program from January 2005 to June 2007. The SSA aims to promote a research network, linking European modelling tool-providers and developers with the potential users of their research results (Utset et al., 2007). AGRIDEMA general objective is to establish initial contacts and to conduct primary collaborations between “developers” and potential “users”, basically researchers and experts at agricultural services.

GENERAL DESCRIPTION AGRIDEMA comprises the following specific objectives: −

−

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

− − −

To identify European human resources that developed, improved and tested simulating tools such as GCM, seasonal forecasts, regional downscaling techniques and agricultural-impact simulation models; inviting them to participate in the SSA proposal activities for implementing their tools and Know-How. To identify and to invite attending the SSA activities to potential users of the European-provided modelling tools. To conduct short courses, where the invited “developers” will present the particularities of their developed or validated tools to the invited “users”. To perform pilot collaborations between “developers” and “users”. To disseminate the obtained results and to build up a wider consortium, comprising both, the “developers” of the simulating tools and the potential “users” of such tools (e.g. experts from regional agricultural-oriented research centres, advisers and farmers).

According to these objectives, several tasks or “work packages” were scheduled. The tasks can be seen in Figure 1. Following the AGRIDEMA timetable, three Workpackages were finished during the 1st period, i.e. “Identifying and Contacting developers”, “Identifying and contacting users” and “Courses on climate and crop-growth modelling tools”.

Opportunities and Challenges of Using Modelling Tools…

173

IDENTIFYING AND CONTACTING DEVELOPERS The AGRIDEMA consortium created an initial list of which developers should be contacted. The list was based mainly the partners experience and previous contact. The use of European-made simulation tools was encouraged. The following Table shows the simulation tools that were contacted by the AGRIDEMA consortium, as well as the corresponding European institution and relevant person. Table 1. Models, institutions and “developers” included in the AGRIDEMA activities Model REGCM3 LARS-WG, SIRIUS MetandRoll Statistical downscaling DEMETER LAPS SWAP WOFOST PERUN ROIMPEL CROPSYST STICS DSSAT

Institution ICTP, Italy Rothamsted Experimental Station, UK Inst. Atmos. Physics, Czech Republic BOKU, Austria ECMWF, UK Agrometeorological Institute Novi Sad, Serbia Wageningen Agricultural University, The Netherlands Wageningen Agricultural University, The Netherlands Mendel University Brno, Czech Republic Romanian Foundation on Global Change, Romania ISCI, Italy INRA, France University of Madrid, Spain

Contact J. Pal M. Semenov M. Dubrovsky H. Formayer F. Doblas-Reyes D. Michailovic J. Kroes K. Van Diepen M. Trnka C. Simota M. Donatelli F. Ruget A. Iglesias

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Several work agreements were achieved between the AGRIDEMA consortium and most of the contacted “developers”, pointing out the “developers” participation in the AGRIDEMA courses, as well as their future support of the “Pilot Assessments” to be conducted by the “users” in the framework of AGRIDEMA.

IDENTIFYING AND CONTACTING USERS Mediterranean countries could face the highest negative consequences of global warming within Europe, through water-shortage and crop-water requirements increments. Besides, since climate-change and extreme events effects could be more serious in countries with lessdeveloped agriculture, the EU associated countries from Central and Eastern Europe, with relative reduced technological capacities, would be more affected than Northern-European countries. Therefore, AGRIDEMA focuses on “users” coming from Southern, Central and Eastern Europe, as well as from the countries of the Mediterranean area. The members of the AGRIDEMA consortium released a call to “users” applicants since April 2005. Relevant institutions were contacted, according to AGRIDEMA partner’s

174

Angel Utset

experience, as well as official centres depending on the Countries’ Ministries of Agriculture or similar institutions. The call was published also using all the available means, including email lists and internet facilities. As pointed out by the three AGRIDEMA partners during their first meeting, the basic criteria for selecting the “users” institutions to be involved in the proposal were: i.

To be able to communicate in English and to be able to work with data management software (Windows, Excel, etc.). ii. To be involved with local agricultural decision-making, advising and farming. iii. To be aware about the potential benefits of agricultural decision modelling tools, being able to identify which agricultural management options should be change and how to optimise management and reduce climate risk of local agricultural production. iv. To have available data for the training course and for the potential conducting the SSA pilot assessments (crop growth and yields, meteorological variables, soil properties, irrigation and crop management scheduling, etc).

Additionally, users conducting PhD studies in the same subjects of AGRIDEMA activities will be especially considered for invitation.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

THE AGRIDEMA COURSES ON CLIMATE AND CROP-GROWTH SIMULATION TOOLS The Courses were held in Vienna in November-December 2005, as scheduled in the proposal. Since many applications to the AGRIDEMA courses were received from “users” out of the targeted countries, the AGRIDEMA partners decided to include these applicants also without any course fee, if they were able to support their trips and lodging expenses. Finally, 44 “users” were present in total, from more than 15 different countries. Sixteen “users” were fully supported by AGRIDEMA and other eight were partially supported by the SSA. A picture showing all the participants in the AGRIDEMA courses is depicted in Figure 2. Institutions from several countries decided to support additional participants in AGRIDEMA courses. Particularly, the participation of five Spanish researchers was supported by the INIA AC05-008 complementary action. Besides, several students and researchers from BOKU, Austria, were in the courses too. The AGRIDEMA web page (www.agridema.com) shows all the details of Courses held in Vienna; as the lectures program, time schedule, invited developers, participant users, etc. The courses on Climate tools comprise lectures on climate change scenarios, dynamical and statistical downscaling, as well as weather generators. The details and work-performance of quite known crop-growth models as SIRIUS, DSSAT, CROPSYST and WOFOST; among others, were shown too.

Opportunities and Challenges of Using Modelling Tools…

175

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

THE AGRIDEMA PILOT ASSESSMENTS The AGRIDEMA pilot assessments were basically applications conducted by some of the “users” that attended the AGRIDEMA courses on climate and crop-growth simulation tools that were held in Vienna ending 2005. Assessments were made using existing data and were addressed to relevant issues concerning climate risks and agricultural decision-making in their respective countries and institutions. The following issues were considered in the Assessments propositions: Local modelling comparisons and validations using available data will be encouraged as the agreement subjects. All the collaborations must identify clearly the potential benefits of these modelling applications for local agricultural decision-making. Particularly, those applications which include farmers from regional medium and small enterprises, as potential users of the tested tools, will be better considered for funding. Educational outputs, such as Ph. D. studies connected to the work agreements are highly desirable. Only original and different propositions can be supported. The selection of the pilot assessments was based on the propositions that the “users” made at the end of the AGRIDEMA courses. The selection was geographically made. The AGRIDEMA consortium partners considered the available budget and the agreements among them during their first meeting. Eight assessments were selected in the Mediterranean area, although only six were funded (those not leaded by ITACyL). Five assessments were conducted in Central Europe and three in Eastern Europe. The SSA coordinator gave priority to cooperation and exchange among the Mediterranean “users”, as well as funding dissemination activities, rather than to support ITACyL researches that have other funding possibilities. Furthermore, ITACyL received additional funds from the Spanish government to strength the cooperation with the Spanish institutions working in AGRIDEMA Pilot Assessments (Complementary Action CGL2006-26211-E) as well as with the Mediterranean countries involved (International Complementary Action PCI2005-A7-0105). The Pilot assessments were conducted from March to October 2006. The AGRIDEMA support was based on agreements to be signed between each “user” conducting Pilot assessments and the corresponding partner of the AGRIDEMA consortium. The Pilot assessments information can be seen in the AGRIDEMA web page: www.agridema.org The reports of the AGRIDEMA Pilot assessments, which can be downloaded from the AGRIDEMA web, comprise an excellent collection of many different applications from several European and Mediterranean countries, all of them addressed to local potential climate-risks. The strategic AGRIDEMA goal, which aims to promote a research network, linking European modelling tool-providers and developers with the potential users of their research results, has been already partially fulfilled.

176

Angel Utset

THE AGRIDEMA PILOT-ASSESSMENTS RESULTS According to the AGRIDEMA goals, the Pilot assessments should comprise the two following tasks:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

− −

Downscaling the provided GCM-outputs and/or seasonal-forecasts. Simulating impacts (such as crop growth and yield, drought stress level etc.) under the locally-obtained climate scenarios, evaluating several management options, which might mitigate the probable climate impact.

Scientific quality of the assessments is quite variable, which can be expected from the wide irregularity of AGRIDEMA applicants. Furthermore, the applicability of the obtained results varies among the assessments as well. Generally, those assessments conducted by researchers located in agricultural stations are closer to stakeholder goals, but unclear regarding climate analysis. Objectives were too ambitious sometimes and could not be fulfilled with a simple application. Moreover, the use of the climate and crop modelling tools shown in the AGRIDEMA courses was mainly related to the user’ previous experience and not to the tool reliability for the expected goal. Comparisons between modelling tools were not conducted in most of the cases. In spite of all the above, the AGRIDEMA results constitutes perhaps the most important collection of independent climate-change agricultural assessments that have been made in Europe. The downscaling method was clearly identified in all the AGRIDEMA Pilot-Assessment reports. However, the GCM data source is not always pointed out. In some cases, the assessment was very simple due to the lack of data or relevant knowledge on the crop model. The most used GCM outputs were the Canadian CCCMa (4 times) and HADCM3 (3 times). The ECHAM and CSIRO GCM outputs were used in one assessment, where a GCM output comparison was performed. The LARS-WG appeared in 8 of the AGRIDEMA pilot assessment, which makes this weather generator as the most frequently used climate-tool in the framework of AGRIDEMA. According to Wilby and Wigley (2001), LARS-WG is one of the most used weather generators. The AGRIDEMA results confirm this conclusion. The MetandRoll weather generator was found in 3 assessments, including one comparison with LARS-WG. The Regional ReGCM3 model was used in 2 assessments and the MAGICC model in one. Regarding the crop models considered in the Pilot assessments, the Wageningen model WOFOST was the most used, but in its SWAP (3 times) and PERUN (3 times) versions. DSSAT models were employed in 4 Pilot Assessments, while CROPSYST was used 3 times, ROIMPEL was considered 2 times and SIRIUS was used in one assessment, which performed a model comparison with PERUN. The frequency of using crop models in the AGRIDEMA Pilot assessments is similar to that found in climate-change impact assessments all over the world, reported by Tubiello and Ewert (2002) as well as in Europe, as pointed out by Alexandrov (2002). Introducing Climate and Crop-growth simulation tool to support agricultural decisionmaking: The “users” point of view. “Users” and “Developers” were invited to the final AGRIDEMA workshop, which was held in Valladolid, Spain, middle 2007. The results of the assessments report were presented,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Opportunities and Challenges of Using Modelling Tools…

177

focusing on the “users” points of view regarding the limitations of the available climate and crop-growth modeling tools. Hence, the “developers” received a feedback on how to improve the corresponding tools. Furthermore, the “developers” also pointed out the current development of the tools. Some representatives of farmer organizations, insurance companies and policy makers were also present. AGRIDEMA interactions between “Users” and “developers” yielded some interesting results. Regarding the GCM outputs, “users” complained on the data format and the time scale. Only the Canadian CCCMa model provides daily data in an easily-converted format, through a Web service. This became such model as the most used in the AGRIDEMA framework. Besides, “Users” request to the national meteorological services to provide statistical (and/or) dynamical downscaled data of the most relevant GCM and emission scenarios. Such data can be used at each country in climate-change agricultural applications. Some of the “Users” and particularly the farmer representatives argue about the utility of RCM data, since the 2070-2100 seems to be extremely far for practical medium-term assessments. Farmers are mainly interested on seasonal or short-term applications. Furthermore, market prices, CAP, WFD and European or national policies can significantly influence farmer decision, besides of climate conditions. Particularly, CAP cross-compliance and the rural development funds can be an important instrument to introduce and evaluate climate-change adaptation measures in the European agriculture. Concerning the weather generators, “Users” from the Mediterranean region pointed out that the main current approach, based on generating the variables needed for Priestly and Taylor evapotranspiration approach might not be useful. The Penman-Monteith approach has been largely recognized as the most adequate in dry conditions. “Users” took note about the facilities provided through the EU proposal ENSEMBLES. The availability of downscaled data from seasonal forecast and decadal scenarios could be an important encouragement for climate-risk agricultural assessments. According to the AGRIDEMA results, DSSAT, WOFOST and CROPSYST are the most relevant crop-growth simulation models that are being used in Europe for climate-change risk assessments. The utility of crop models to support agricultural decision-making has been recognized. However, the “cascade approach” considered in many models to simulate soil water balance might be not adequate. This approach ignores capillary rising, which might be important in rainfed or deficit irrigation crop systems. The AGRIDEMA participants strongly encouraged Universities and Educational politicians to held courses of current climate and crop-growth simulation tools. These tools are still unknown by most of their potential “Users”, which has been considered as the main current limitation to introduce them in practice. Besides, they encourage also conducting demonstration proposals, addressed to calibrate and validate the simulation tools in several farm conditions. FP7 cooperation program aims to increase private investment rates in R+D in Europe. The demonstration proposals, funded by FP7, could count on farmers and agribusiness since they are interested in adopting reliable measures in order to reduce climate risks. The participation of agricultural applied-research or extension services in those proposals is crucial.

178

Angel Utset

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Alexandrov, V. 2002. Summarizing Crop Growth Simulation Models in Europe with Potential for Operational Assessment of Crop Status and Yield Prognosis. In: Dunkel, Z., V.Alexandrov, Z. Gat, R. Guerreiro, A. Kleschenko and Y. Ozalp (eds.), Report of the RA VI Working Group on Agricultural Meteorology. CAgM Report No.89, WMO/TD No.1113, Geneva, Switzerland, pp. 119-214. Alexandrov, V. 2007. Current Climate Forecasting as a Helping Tool for Agricultural Decision Making. AGRIDEMA, FP6-EC contract No 003944, Deliver 8, 11 pp. Alves, O., G. Wang, A. Zhong, N. Smith, G. Warren, A. Marshall, F. Tzeitkin, and A. Schiller. 2002. POAMA: Bureau of Meteorology operational coupled model seasonal forecast system. Proc. ECMWF Workshop on the Role of the Upper Ocean in Medium and Extended Range Forecasting, Reading, United Kingdom, ECMWF, 22–32 Arkin, G.F., Vanderlip, R.L., Ritchie, J.T., 1976. A dynamic grain sorghum growth model. Trans. ASAE 19, 622-626, 630. Baigorria, G.A. 2007. When there is no El Niño: Approaches for crop yield forecasting. Bastiaansen, W.G.M., Allen, R.G., Droogers, P., D’Urso, G. and P. Steduto. Inserting man’s irrigation and drainage wisdom into soil water flow models and bringing it back out: How far we progressed?. In R.A. Feddes, G.H. de Rooij and J.C. Van Dam (eds). Unsaturatedzone modelling: Progress, challenges and applications. Kluwer Academic Publishers, Wageningen. Boote, K.J., Jones, J.W., Hoogenboom, G., Pickering, N.B., 1998. The CROPGRO model for grain legumes. In: Tsuji, G.Y., Hoogenboom, G., Thornton, P.K. (Eds.), Understanding Options for Agricultural Production. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp.99-128. Boote, K.J., Jones, J.W., Mishoe, J.W., Wilkerson, G.G., 1986. Modeling growth and yield of groundnut. Agrometeorology of Groundnut: Proceedings of an International Symposium, ICRISAT Sahelian Center, Niamey, Niger. 21-26 Aug, 1985, ICRISAT, Patancheru, A.P. 502 324, India, pp. 243-254. Bouman, B.A.M., Kropff, M.J., Tuong, T.P., Wopereis, M.C.S., Ten Berge, H.F.M., Van Laar, H.H., 2001. ORYZA2000: Modeling Lowland Rice (ISBN 971-22-0171-6). International Rice Research Institute/Wageningen University and Research Centre, Los Banos (Philippines)/ Wageningen, pp. 235. Bouman, B.A.M., van Keulen, H., van Laar, H.H., Rabbinge, R., 1996. The ‘School of de Wit’ crop growth simulation models: a pedigree and historical overview. Agric. Syst. 52, 171-198. Christensen, J.H. and O.B. Christensen. 2007. A summary of the PRUDENCE model projections of changes in European climate by the end of this century. Climatic Change, 81: 7-30. Connolly, R.D., 1998. Modelling effects of soil structure on the water balance of soil-crop systems: A review. Soil Till. Res. 48:1-19. de Wit, C.T., Brouwer, R., Penning de Vries, F.W.T., 1970. The simulation of photosynthetic systems. In: Setlik, I. (Ed.), Prediction and measurement of photosynthetic productivity. Proceeding IBP/PP Technical Meeting Trebon 1969. Pudoc, Wageningen, The Netherlands, pp. 47-50.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Opportunities and Challenges of Using Modelling Tools…

179

de Wit, C.T., et al., 1978. Simulation of Assimilation, Respiration and Transpiration of Crops (Simulation Monographs). Pudoc, Wageningen, The Netherlands, p. 141. Doblas-Reyes, F.J., Hagedorn, R. and T.N. Palmer. 2006. Developments in dynamical seasonal forecasting relevant to agricultural management. Climate Res. 33:19-26. Eatherall, A. 1997. Modelling climate impacts on ecosystems using linked models and a GIS. Climatic change 35: 17-34. Giorgi, F. and L.O.Mearns. 1999. Introduction to special section: Regional climate modeling revisited. Journal of Geophysical Research, 14(D6): 6335-6352. Goudriaan, J., 1977. Crop Micrometeorology: A Simulation Study. Simulation Monographs. Pudoc, Wageningen, The Netherlands, p. 257. Guereña, A., Ruiz-Ramos, M., Díaz-Ambrona, C.H., Conde, J.R. and M.I. Mínguez. 2001. Assessment of climate change and agriculture in Spain using climate models. Agron. J. 93:237-249. Hammer, G.L., Hansen, J.W., Philips, J.G., Mjelde, J.W., Hill, H., Love, A. and A. Potgieter. 2001. Advances in application of climate prediction in agriculture. Agricultural Systems. 70:515-553. Hansen, J.W., Challinor, A., Ines, A., Wheeler, T. and Moron, V. 2006. Translating climate forecasts into agricultural terms: advances and challenges. Climate Res. 33:27-41. Hewitt, C.D., 2005: The ENSEMBLES Project: Providing ensemble-based predictions of climate changes and their impacts. EGGS newsletter, 13, 22-25. Hoogenboom, G. 2000 Contribution of agrometeorology to the simulation of crop production and its applications. Agric. For. Meteorol. 103:137-157. Hoogenboom, G., Fraisse, C.W., Jones, J.W., Ingran, K.T., O’Brien, J., Bellow, J.G., Zierden, D., Stooksbury, D.E., Paz, J.O., Garcia, A., Guerra, L.C., Letson, D., Breuer, N.E., Cabrera, V.E., Hatch, L.U. and C. Roncoli. 2007. Climate-based agricultural risk management tools for Florida, Georgia and Alabama, USA. Sivakumar, M.V.K., Hansen, J. (Eds.) Climate Prediction and Agriculture, Adavances and Challenges Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 11-13 May 2005, Washington D.C. USA, International START Secretariat, pp 273-278. Hoogenboom, G., Jones, J.W., Wilkens, P.W., Batchelor, W.D., Bowen, W.T., Hunt, L.A., Pickering, N.B., Singh, U., Godwin, D.C., Baer, B., Boote, K.J., Ritchie, J.T., White, J.W., 1994. Crop models. In: Tsuji, G.Y., Uehara, G., Balas, S. (Eds.), DSSAT Version 3, vol. 2. University of Hawaii, Honolulu, HI, pp. 95-244. Hoogenboom, G., Wilkens, P.W., Thornton, P.K., Jones, J.W., Hunt, L.A., Imamura, D.T., 1999. Decision support system for agrotechnology transfer v3.5. In: Hoogenboom, G.,Wilkens, P.W., Tsuji, G.Y. (Eds.), DSSAT version 3, vol. 4 (ISBN 1-886684-04-9). University of Hawaii, Honolulu, HI, pp. 1-36. International Benchmark Sites Network for Agrotechnology Transfer. 1993. The IBSNAT Decade. Department of Agronomy and Soil Science, College of Tropical Agriculture and Human Resources, University of Hawaii, Honoluly, Hawaii. IPCC, 2007: Summary for Policymakers. In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

180

Angel Utset

Jones, C.A., Kiniry, J.R., 1986. CERES-Maize: A Simulation Model of Maize Growth and Development. Texas A&M University Press, College Station, Texas. Jones, J. W., . Tsuji, G.Y, Hoogenboom G., Hunt, L.A., Thornton, P.K., Wilkens, P.W., Imamura, D.T., Bowen, W.T., Singh, U. 1998. Decision support system for agrotechnology transfer: DSSAT v3. En: G. Tsuji, G. Hoogenboom and P. Thornton (Eds), Understanding options for agricultural production, Kluwer Academic Publishers, Dordrecht, pp 157-178. Jones, J.W., Hoogenboom, G., C.H. Porter, K.J. Boote, Batchelor, W.D., Hunt, L.A., Wilkens, P.W., Singh, U., Gijsman, A.J. and J.T. Ritchie. 2003. The DSSAT cropping system model. Eur. J. Agron. 18:235-265. Kanamitsu, M., and Coauthors, 2002: NCEP dynamical seasonal forecast system 2000. Bull. Amer. Meteor. Soc., 83, 1019–1037. Kandil, H.M., Skaggs, R.W., Abdel-Dayem, S.A. 1995. DRAINMOD-S: water management model for irrigated arid lands, crop yield and applications. Irrigation and Drainage systems 9:239-258. Kattenberg, A., F. Giorgi, H. Grassl, G.A. Meehl, J.F.B. Mitchell, R.J. Stouffer, T. Tokioka, A.J. Weaver, and T.M.L. Wigley. 1996: Climate models - projections of future climate. In: Climate Change 1995. The Science of Climate Change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change. [Houghton, J.T., L.G.M. Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.)]. Cambridge University Press, Cambridge, pp. 285-357. Keating, B.A., Carberry, P.S., Hammer, G.L., Probert, M.E., Robertson, M.J., Holzworth, D., Huth, N.I., Hargreaves, J.N.G., Meinke, H., Hochman, Z., McLean, G., Verburg, K., Snow, V., Dimes, J.P., Silburn, M., Wang, E., Brown, S., Bristow, K.L., Asseng, S., Chapman, S., McCown, R.L., Freebairn, D.M., Smith, C.J., 2003. An overview of APSIM, a model designed for farming systems simulation. Eur. J. Agron., 18, 267-288. Kutilek, M., Nielsen, D., 1994. Soil Hydrology. Cremlingen-Destedt, Catena Verlag, 370pp. Marletto, V., Zinoniu, F., Criscuolo, L., Fontana, G., Marchesi, S., Morgillo, A., Van Soetandale, M., Ceotto, E. And U. Andersen. 2005. Evaluation of downscaled DEMETER multi-model ensemble seasonal hindcasts in a northern Italy by means of a model of wheat growth and soil water balance. Tellus, 57:488-497. Mason, S. J., L. Goddard, N. E. Graham, E. Yulaeva, L. Sun, and P. A. Arkin, 1999: The IRI seasonal climate prediction system and the 1997/98 El Niño event. Bull. Amer. Meteor. Soc., 80, 1853–1873. Mavromatis, T., Jones, P.D. 1998. Comparison of climate scenario construction methodologies for impact assessment studies. Agric. For. Meteor. 91:51-67. Mearns, L.O., C. Rosenzweig, and R. Goldberg. 1997: Mean and variance change in climate scenarios: methods, agricultural applications, and measures of uncertainty. Climatic Change, 35, 367-396. Meinke, H., Baethgen, W.E., Carberry, P.S., Donatelli, M., Hammer, G.L., Selvaraju, R. and C.O. Stockle. 2001. Increasing profits and reducing risks in crop production using participatory systems simulation approaches. Agricultural Systems. 70:493-513. Meinke, H., Nelson, R., Kokic, P., Stone, R., Selvaraju, R. and W. Baethgen. 2006. Actionable climate knowledge: from analysis to synthesis. Climate Res. 33:101-110. Palmer, T. N., Alessandri, A., Andersen, U., Cantelaube, P., Davey, P., Délécluse, P., Déqué, M., Díez, E., Doblas-Reyes, F.J., Feddersen, H., Graham, R., Gualdi, S., Guérémy, J.F.,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Opportunities and Challenges of Using Modelling Tools…

181

Hagedorn, R., Hoshen, M., Keenlyside, N., Latif, M., Lazar,A., Maisonnave, E., Marletto, V., Morse, A.P., Orfila, B., Rogel,P., Terres,J.M. and M. C. Thomson. Development of a european multimodel ensemble system for seasonal-to-interannual prediction (DEMETER). American Meteorological Society, p 853-872. Penning de Vries, F.W.T., Brunsting, A.H.M., Van Laar, H.H., 1974. Products, requirements and efficiency of biosynthesis: a quantitative approach. Journal of Theoretical Biology 45, 339-377. Penning de Vries, F.W.T., D.M. Jansen, H.F.M. Ten berge and A.H. Bakema, 1989. Simulation of eco-physiological processes of growth of several annual crops. Simulation Monographs, PUDOC-IRRI, Wageningen, the Netherlands. Ritchie J. and S. Otler, 1985. Description and performance of CERES wheat: a user-oriented wheat yield model. En: W. Willis (Ed) ARS. Wheat yield Project. U.S. Dept. of Agric. Res. Serv. ARS-38, 217pp. Ritchie, J.T. 1998. Soil water balance and plant water stress. In: G. Tsuji, G. Hoogenboom and P. Thornton (Eds), Understanding options for agricultural production, Kluwer Academic Publishers, Dordrecht, pp 41-54. Semenov, M.A. and P.D. Jamieson. 2001. Using weather generators in crop modelling. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat, 322 pp. Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat, 322 pp. Sivakumar, M.V.K. and J. Hansen (Eds). 2007. Climate Prediction and Agriculture, Adavances and Challenges Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 11-13 May 2005, Washington D.C. USA, International START Secretariat, 288 pp. Stockdale, T. N., D. L. T. Anderson, J. O. S. Alves, and M. A. Balmaseda. 1998. Global seasonal rainfall forecasts using a coupled ocean–atmosphere model. Nature, 392, 370– 373. Stockle, C.O., Donatelli, M. and R. Nelson. 2003. CropSyst, a cropping systems simulation model. Eur. J. Agron. 18:289-307. Swaney, D.P., Jones, J.W., Boggess, W.G., Wilkerson, C.G., Mishoe, J.W. 1983. Real-time irrigation decision analysis using simulation. Trans. ASAE 26, 562-568. Teixeira, J.L. and Pereira, L.S. 1992. ISAREG: an irrigation scheduling simulation model. ICID Bulletin 41:29-48. Tsuji, G. Hoogenboom and P. Thornton. 1998. Understanding options for agricultural production, Kluwer Academic Publishers, Dordrecht. Tubiello, F.N., Ewert, F. 2002. Simulating the effects of elevated CO2 on crops: approaches and applications for climate change. European J. Agron. 18:57-74. Uehara, G. 1998. Synthesis. In: Tsuji, G.Y., Hoogenboom, G., Thornton, P.K. (Eds.), Understanding Options For Agricultural Production. Kluwer Academic Publishers, Dordrecht, The Netherlands, pp. 389-392. using Global and Regional Circulation Models. In Proceedings of II International Workshop on Climate Change and its impact on agriculture, Viçosa, Brasil.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

182

Angel Utset

Utset, A., Eitzinger, J. and V. Alexandrov. 2007. AGRIDEMA: An EU-funded effort to promote the use of climate and crop simulation models in agricultural decision-making. Sivakumar, M.V.K., Hansen, J. (Eds.) Climate Prediction and Agriculture, Advances and Challenges Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 11-13 May 2005, Washington D.C. USA, International START Secretariat, pp 259-264. Van Dam, J. C., Huygen, J., Wesseling, J. G., Feddes, R. A., Kabat, P., van Walsum, P. E. V., Groenendijk, P., van Diepen, C. A., 1997. Theory of SWAP version 2.0. Report 71. Technical Document 45, Wageningen, 167 pp. Van Dam, J.C. 2000. Field-scale water flow and solute transport. SWAP model concepts, parameter estimation and case studies. Doctoral Thesis, Wageningen University. ISBN 90-5808-256-3, 167 pp. Van Ittersum, M.K., Leffelaar, P.A., van Keulen, H., Kropff, M.J., Bastiaans, L. and J. Goudriaan. 2003. On approaches and applications of the Wageningen crop models. Eur. J. Agron. 18:201-234. Van Keulen, H. 1975. Simulation of water use and herbage growth in arid regions. Simulation Monographs. Pudoc, Wageningen, The Netherlands, p. 184. Van Keulen, H., 1982. Crop production under semi-arid conditions, as determined by nitrogen and moisture availability. In: Penning de Vries, F.W.T., Van Laar, H.H. (Eds.), Simulation of Plant Growth and Crop Production. Simulation Monographs. Pudoc, Wageningen, The Netherlands, pp. 234-249. Van Keulen, H., Wolf, J. 1986. Modelling of agricultural production: weather, soils and crops. Pudoc Wageningen Simulation Monographs. Van Laar, H.H., Goudriaan, J., Van Keulen, H., 1997. SUCROS97: Simulation of crop growth for potential and water-limited production situations. Quantitative Approaches in Systems Analysis, No. 14. C.T. de Wit Graduate School for Production Ecology and Resource Conservation, Wageningen, The Netherlands, pp. 52. Vanclooster, M., Viane, P., Diels, J. 1994 WAVE: a mathematical model for simulating water and agrochemicals in the soil and vadose environment. Reference and User’s Manual (release 2.0). Katholieke Universiteit Leuven, Leuven. Vogel, T., K. Huang, R. Zhang, and M. Th. van Genuchten. 1996. The HYDRUS code for simulating one-dimensional water flow, solute transport, and heat movement in variablysaturated media, Version 5.0, Research Report No 140, U.S. Salinity laboratory, USDA, ARS, Riverside, CA. Wilby, R.L. and T.M.L. Wigley. Down-scaling general circulation issues in climate prediction. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat. p 39-68. Wilkerson, G.G., Mishoe, J.W., Jones, J.W., Boggess, W.G., Swaney, D.P., 1983. Withinseason decision making for pest control in soybeans. ASAE Paper No. 83-4044, St. Joseph, MI.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 13

MODELLING CLIMATE CHANGE IMPACTS ON CROP GROWTH AND MANAGEMENT IN GERMANY K. C. Kersebaum∗a, W. Mirschela, K. O. Wenkela, R. Manderscheidb, H. J. Weigel b and C. Nendel a a

Leibniz-Centre of Agricultural Landscape Research, Institute of Landscape Systems Analysis, Eberswalder Str. 84 D-15374 Müncheberg, Germany b Federal Agricultural Research Centre, Institute of Agroecology, Bundesallee 50, D- 38116 Braunschweig, Germany.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Climate change will affect crop growth and consequently crop management. We used two different models to i) simulate the effect of elevated CO2 on crop growth and yield formation of winter wheat, winter barley and sugar beets using data of the German FACE (Free Air Cabon Dioxide Enrichment) experiment with the AGROSIM model family and ii) to estimate the effects of climate change on crop growth of winter wheat and maize and the resulting consequences for water and nitrogen management using the HERMES model in combination with downscaled climate change scenarios of selected weather stations in east Germany. For i) model results and measured data were compared regarding crop biomass, yield and water consumption. An increase of biomass and yield under elevated CO2 was well reflected by the model AGROSIM. However, the observed differences in the water consumption will require further modifications of the model approach. For ii) the different response at different locations (climate and soil) on crop growth and yield formation were analyzed. Different water (irrigation) and nitrogen (fertilization) management options were evaluated and nitrogen response curves were generated under rain fed and irrigated conditions. Model results indicated that with decreasing summer precipitation the nitrogen use efficiency decreases.

∗

[email protected]

184

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

INTRODUCTION Climate change is very likely to affect future crop growth and consequently crop management. In detail, there are many climatic factors affecting agro-ecosystems and thereby influencing crop yield in different ways. Rising atmospheric CO2 levels act like a natural carbon fertilization similar to current practice in glasshouse vegetable production, thereby enhancing crop growth in general terms. Increasing temperatures in winter and spring will lead to season shift with earlier and faster crop ontogenetic development, while higher temperatures in summer will most likely affect photosynthesis negatively and thus obstructing biomass production. Absence of summer rainfall will necessitate irrigation in today rain-fed agricultural production systems or at least lead to frequently occurring summer dry spells with decreasing yields. The combination of the named climatic factors may also not only act upon crop yield but also on crop quality, with the potential of affecting farm income more severely than just yield loss. Simulation models that are able to assess climate impact on crop growth, yield and farm economy today still lack complete feedback structures. Only single aspects can be investigated. However, modelling these single aspects already increases knowledge on what to expect from climate change, if interpreted carefully and in the context of the model’s abilities. In order to assess direction and magnitude of the impact of some aspects of climate change we used two different simulation models to i) simulate the effect of elevated CO2 on crop growth and yield formation of winter wheat, winter barley and sugar beets using data of the German FACE (Free Air Carbon Dioxide Enrichment) experiment with the AGROSIM model family and ii) to estimate the effects of future temperature and moisture regimes on crop growth of winter wheat and maize and the resulting consequences for water and nitrogen management using the HERMES model in combination with downscaled climate change scenarios of selected weather stations in east Germany.

MATERIAL AND METHODS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

AGROSIM The agro-ecosystem model family AGROSIM (AGRO-ecosystem SIMulation) was developed at the Institute of Landscape Systems Analysis of the Leibniz-Centre for Agricultural Landscape Research (ZALF) and parameterized and validated for the agriculturally used moraine landscapes of Northeast Germany. The AGROSIM simulation model focuses on crop growth processes. It enables the investigation of consequences of (i) different crop management strategies at farm level as well as of (ii) different forms of land use and of (iii) climate change effects at regional level on biomass production and crop yield. Until now, models for winter wheat (AGROSIM-WW), winter barley (AGROSIM-WG), winter rye (AGROSIM-WR), sugar beet (AGROSIM-ZR), and different catch crops (AGROSIM-ZF) are available, all validated for Northeast Germany weather and soil conditions. The individual AGROSIM models are based on the same modelling philosophy, such as a consistent modular structure and the use of rate equations to describe process dynamics between sowing and harvest on daily basis. Implemented interactions between

Modelling of Maize Production and the Impact of Climate Change…

185

climate, biomass increase, ontogenesis, and soil processes make the models sensitive to weather, site, and management effects, assuming homogeneous crop stands. The models require meteorological standard data as driving forces and easy-available regionalised input parameters. The general structure of all process relations for winter cereals is illustrated in Figure 1. Within AGROSIM, state variables and rates, such as actual and potential evapotranspiration, soil water content in different soil layers, soil nitrogen availability, mineralization, nitrogen and water uptake, percolation, soil temperature, ontogenesis, root, aboveground, green, dead and grain biomass, respiration, rate of grain filling, etc., are obtained as outputs with daily time resolution.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Model scheme of AGROSIM used for winter cereals.

In the AGROSIM models the influence of atmospheric CO2 content is coupled to assimilation and respiration processes only. The algorithm for describing the influence of atmospheric CO2 concentration on crop photosynthesis is based on a Michaelis-Menten equation and is expressed according to a reference concentration of 350 ppm (Mirschel and Wenkel, 1998): CO 2 − C 0 k 1 + CO 2 − C 0 f(CO 2 ) = ; 350 − C 0 k 1 + 350 − C 0

k 1 = 220 + 0.395 ⋅ QP;

C 0 = 80 − 0.09 ⋅ QP

(1)

with CO2 being the atmospheric CO2 concentration (ppm) and k1 and C0 being the halfsaturation and the compensation CO2 content, respectively. Both parameters proved to be dependent on the global radiation QP (W m-2; Hoffmann, 1993). The AGROSIM models are

186

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

described in more detail by Mirschel et al. (2001, 2007) for winter cereals and by Mirschel et al. (2002) for sugar beet.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

HERMES The HERMES model is a one-dimensional, multi-year and multi-crop model for the simulation of water and nitrogen dynamics in agro-ecosystems across rotations with a daily time step (Kersebaum and Beblik, 2001). It includes a capacity water balance model including the simulation of evapotranspiration, water fluxes and capillary rise, a module for nitrogen cycling and transformations (net-mineralization, denitrification) and convectivedispersive transport and a module for crop growth, phenological development and N-uptake. The model has been described in detail by Kersebaum (1995) and Kersebaum and Beblik (2001). The model simulates nitrogen net mineralisation from two pools of potentially decomposable nitrogen (Richter et al., 1982) which are derived from soil organic matter and amounts of crop residues related to the yield of the previous crop. A default percentage of 13% of the total soil organic nitrogen is set as initial value of the slowly decomposable nitrogen pool (Nuske, 1983). Daily mineralisation is simulated depending on temperature and soil moisture (Myers et al., 1982). Denitrification is modelled for the top soil (0-30 cm) using a Michaelis-Menten kinetic dependent on nitrate content modified by reduction functions depending on water filled pore space, and temperature (Richter and Söndgerath unpublished, cited in Schneider, 1991). The submodel for crop growth was built on the basis of the SUCROS model (van Keulen et al., 1982). The simulation of daily net dry matter production as a result of photosynthesis and respiration is driven by global radiation and temperature. Dry matter production is partitioned depending on crop development stages calculated from a thermal sum (°C days) which can be modified specifically for each stage by day length and vernalisation if applicable for a specific crop. The yield is estimated at harvest from the weight of the storage organ and nitrogen recycling through crop residues is calculated automatically from the simulated crop nitrogen uptake minus the nitrogen exported at harvest with yield and removed by-products (straw, leaves etc.). Crop growth is limited by water and nitrogen stress. Drought stress is indicated by the ratio between actual and potential transpiration. Temporary limitation of soil air by water logging is considered through reducing transpiration and photosynthesis according to Supit et al. (1994). Water and nitrogen uptake is calculated from potential evaporation and crop nitrogen status, depending on the simulated root distribution and water and nitrogen availability in different soil layers (Kersebaum 1995). An empirical exponential distribution function (Gerwitz and Page, 1974) is used calculate the root length density from the simulated root biomass. N uptake simulation considers convective as well as diffusive transport to the roots. The concept of critical N concentration in plants as a function of crop developmental stage (Kersebaum and Beblik 2001) or as a function of crop biomass (Greenwood et al., 1990) is applied to assess the impact of N shortage. The model is used to evaluate different management options in arable crop production from field to catchment scale (e.g., Kersebaum 2007, Kersebaum et al. 2003) and to derive

Modelling of Maize Production and the Impact of Climate Change…

187

nitrogen fertilization recommendations for farmers on a field and sub-field scale (e.g., Kersebaum et al. 2005).

THE FACE-EXPERIMENT The FACE-experiment set up at the Institute of Agro-ecology of the Federal Agricultural Research Centre Braunschweig as the principal experiment of the Braunschweig Carbon Project (Weigel and Dämmgen, 2000) is the basis for parameterization and verification of the model approach for considering the fertilization effect of increased atmospheric CO2 used in the AGROSIM models. In this experiment, six CO2 gassing circles with a diameter of 20 m each are installed on a 24ha field (see Figure 3). Inside the circles the aerial CO2 concentration can be increased, considering the natural ambient CO2 concentration, wind speed and wind direction for automatic control of the experimental concentration level. The circles surround 315 m2 cropped area. Treatments were: CO2 levels of 550 ppm (elevated level, gassing current conditions), 375 ppm (ambient level, gassing current conditions) and 375 ppm (ambient level, natural current conditions) as well as two N fertiliser regimes (current practice and 50% of current practice). The CO2 used for gassing is marked with a stable carbon isotope signature. The FACE experiment was carried out for two crop rotation periods from 2000 to 2005, including winter barley (variety Theresa), sugar beet (variety Wiebke) and winter wheat (variety Batis). Sufficient water supply and crop protection was secured during the whole growing season. Consistent data on soil processes, crop growth and site meteorology was collected. A detailed description of the FACE experiment is given by Weigel et al. (2001) and first crop-related results are presented by Weigel et al. (2005).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

CLIMATE CHANGE SCENARIO SIMULATION To simulate the effects of potential climate change effects on wheat yield and nitrogen management generated time series of weather data from a statistical downscaling procedure from Gerstengabe et al. (1999) were used. Using the ECHAM4 scenario SRES A1B (moderate CO2 emission according to IPCC (2001)) as a boundary condition, historical time series from stations of the meteorological observation network of the German Weather Service (DWD) were processed by a special cluster analysis procedure by Gerstengabe et al. (2003) to generate future weather time series for the years 2005 – 2055 for eastern Germany. For selected stations model calculations with HERMES were performed with defined soil profiles typical for each region. The weather data set was divided into four periods of 20 year length: 1961-1981, 1982-2002, 2005-2025 and 2035-2055. Each period was simulated as a crop rotation with winter wheat monoculture assuming a nitrogen fertilization of 200 kg N ha-1 to avoid nitrogen stress. Table 1 gives the climatic characteristics of each period and the assumed soil for each site. To estimate the effect of climate change on nitrogen management, a virtual fertilizer experiment was performed for the periods 1961-1981 and 2035-2055 for the location Kemlitz in east Brandenburg, which is expected to have the most dramatic reduction in annual

188

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

precipitation in Brandenburg. For this, the total amount of nitrogen fertilizer for winter wheat was altered in the simulation between 20 and 200 kg N ha-1 in 20 kg steps. Table 1. Water balance characteristic of climate change at selected sites in East Germany and assumed typical soils Location

period

Angermünde

Müncheberg

Kemlitz

1961-1981 1982-2002 2005-2025 2035-2055 1961-1981 1982-2002 2005-2025 2035-2055 1961-1981 1982-2002 2005-2025 2035-2055

winter 222 203 207 185 215 245 238 200 272 232 210 186

Precipitation summer 314 305 285 241 309 307 302 243 370 306 271 233

ETP

soil

659 672 682 717 654 680 685 722 650 675 684 722

Loamy sand

Sand

Loamy sand

RESULTS For CO2-level impact modelling, the results of the upgraded AGROSIM models were compared to measured data of the Braunschweig Carbon Project FACE experiments regarding crop biomass, yield and water consumption. The general increase of biomass and yield under elevated CO2 was well reflected by the model AGROSIM (Figure 2).

Sugar beet -1

ha

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

16

Tuber , ,

550 ppmv 375 ppmv

12

Leaf

8 4 0

150

200

250

Days after 1. January 2001

300

Modelling of Maize Production and the Impact of Climate Change…

189

Winter barley -1

t ha

, ,

20

550 ppmv 375 ppmv

Above-ground 15

10

Grain 5

Root 0

300

350

400

450

500

550

600

Days after 1. January 2002

Figure 2. Comparison between simulated (AGROSIM) and observed (FACE experiment) biomass increase of different plant organs during growth period of sugar beet (left) and of winter barley (right).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

However, the model persistently under-predicts of the observed data, both for cereals and for sugar beet (Table 2). In the experiment, also a decreased crop water use (about 25 mm less during the growing season) was observed under elevated CO2 environment, combined with a higher surface temperature of the crop (Figure 3).

Figure 3. IR exposure of the FACE winter wheat crop 2005.

190

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

Table 2. Upgraded AGROSIM model over-predicts biomass production of cereals and sugar beet as observed in the Braunschweig Carbon Project FACE experiments (Sufficient N supply, elevated atmospheric CO2 level of 550 ppm)

Winter barley (N+, 550 ppm)

Biomass - Experiment dt ha-1 above-ground: 183 grain: 104

Biomass - Model dt ha-1 above-ground: 211 grain: 108

Overprediction % 15,1 3,3

2001

Sugar beet (N+, 550 ppm)

tuber: 163 leaf: 84

tuber: 187 leaf: 99

14,7 17,8

2001/2002

Winter wheat (N+, 550 ppm)

above-ground: 177 grain: 57

above-ground: 195 grain: 59

9,9 3,6

2002/2003

Winter barley (N+, 547 ppm)

above-ground: 148 grain: 69

above-ground: 161 grain: 68

8,7 -1,4

2004/2005

Winter wheat (N+, 550 ppm)

above-ground: 198 grain: 98

above-ground: 208 grain: 101

5.1 3.1

Year

Crop

1999/2000

Table 6. Development of yields and different soil processes as simulated by the HERMES model under historic and predicted temperature and moisture regime at different sites located in the Federal State of Brandenburg, Germany Site (Period)

Grain yield average/cv t ha-1/ %

N Leaching

Leachate

Denitrification

kg N ha-1 y-1

mm y-1

Kg N ha-1 y-1

Net N Mineralisation kg N ha-1 y-1

1961-1981

7.1/31.0

76

51

1.45

62

1982-2002

7.9/20.8

47

15

1.62

67

2005-2025

6.5/19.0

20

-1

1.21

49

2035-2055

5.6/26.9

2

-5

0.79

44

1961-1981

5.7/34.3

77

67

0.03

56

1982-2002

6.8/30.4

57

70

0.03

61

2005-2025

5.9/22.6

56

52

0.03

45

2035-2055

5.1/23.7

23

-1

0.02

44

1961-1981

8.1/23.7

82

126

2.23

69

1982-2002

7.4/22.9

69

41

1.69

69

2005-2025

5.9/23.7

50

11

1.36

49

2035-2055

5.3/35.1

1

-5

1.05

47

Angermünde

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Müncheberg

Kemlitz

Modelling of Maize Production and the Impact of Climate Change…

191

For temperature and moisture regime impact the response on crop growth and yield formation was analyzed for different locations (climate and soil) with HERMES. Different nitrogen fertilization management options were evaluated and nitrogen response curves were generated under rain-fed and irrigated conditions. Yield simulations show decreasing yield expectations for all sites. However, variability is affected quite differently, as it increases for the buffered soils of Angermünde and Kemlitz and rather decreases for the sandy soils of Müncheberg. The model results for Kemlitz showed that although the yield level is expected to decrease significantly with decreasing summer precipitation, the N fertiliser amount to achieve maximum yield decreases only slightly (Figure 4). This indicates that the nitrogen use efficiency of winter wheat is likely to decrease. This results in higher residual soil mineral nitrogen contents after harvest, which can often be observed after dry growing seasons (Kersebaum, 2000). Under past and present conditions, residual nitrogen from overfertilization is often leached out of the root zone during the following winter period (Figure 4). However, a closer look on different soil processes simulated by the model (Table 3) reveals that decreasing summer precipitation consequently leads to a lower amount of leachate and leaching N. Therefore, under the conditions of the period 2035-2055, almost no leaching was calculated for Kemlitz and Angermünde. At Müncheberg, despite of a slightly negative water flux, nitrogen was leached out during winter to deeper layers, but did not rise up during summer with capillary rise. Furthermore, it indicates that processes influenced by soil moisture, such as N mineralisation and denitrification, are also expected to decrease at all sites due to dryer conditions.

80

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

-1

-1

Grain yield [dt ha ] / N-leaching [kg N ha ]

90

70 60 50 40 30 20

yield 1961-1982

10

yield 2035-2055 N-leaching 1961-1981

0 20

40

60

80

100

120

140

160

180

200

220

-1

N-fertilization [kg N ha ]

Figure 4. Winter wheat N response curves and system N leaching under historic temperature and moisture conditions (1961-1982) and under predicted temperature and moisture conditions (20352055). Arrows indicate economically optimal fertiliser amount. Almost no leaching was estimated for 2035-2055.

192

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

DISCUSSION The simulation of the FACE data with the AGROSIM model shows a general ability of the selected approach to explain the observed increased biomass production under an elevated CO2 environment. However, the persistent under-prediction of the data by the model indicates that the selected approach may be inadequate or the parameterisation insufficient. Observed decreased crop water use and higher crop surface temperatures indicate that the crop is able to save water by closing its stomata more frequently. The CO2 supply may be much more efficient at higher concentrations as the crop demand may be covered by stomata gas exchange in shorter time. This observed dependency of crop water consumption on the atmospheric CO2 environment will require further modifications of the model approach. The decrease of fertilizer demand being much lower than expected from decreasing yield levels may be explained by a simultaneously decreasing natural N supply by the soil under predicted summer dry spells. Since soil N mineralisation being a soil moisture dependent process, decreasing soil moisture will result in decreasing N mineralisation. The resulting gap in natural N supply of the crop during growing season obviously needs to be covered by a higher amount of fertilizer, as indicated by the simulation. Modelling future climate impact on crop yield and management bears a lot of uncertainty if certain aspects are omitted. For the discussion of the presented results this must be understood and considered when concluding from the results. The AGROSIM simulation showed two things: at first the integration of a CO2 response algorithm helped to describe potential biomass production increase under future climate, for which distinctively elevated CO2 levels are almost surely predicted. Secondly, it was also made clear that the response algorithm must be more detailed than the one used in this investigation and be based on up-todate and transferable experiments. The simple Michaelis-Menten approach may turn out to be sufficient when better parameterised but a leaf-level approach might even do better. The general effect of an elevated CO2 environment as observed in the FACE experiments also shows that yield predictions based on simple temperature and moisture assumptions can not be sufficiently exact. Indeed, both these factors are expected to be the most important affecting yield and therefore the direction of future development indicated by the results can be seen as somehow resilient. However, the unconsidered factors may still noticeably alter the predictions. Especially the expected season shift and the consequently adapted crop management strategies (i.e. crop rotations, soil tillage, irrigation) as well as the consequences of climate change for product quality bear the potential of completely overrule the findings of simulation models considering only changes in temperature and moisture regimes. The hazard of an increased frequency of catastrophic weather events on crop yield is also barely considered, although such an event can erase the farm income for the whole season, irrespective of its date of occurrence in the season. For all these reasons it is emphasised here that current simulation models are able to predict the impact of some aspects of climate change on crop growth, yield and farm net returns. However, the simultaneous impact of all aspects of climate change, including their interactions and interdependences, can not be modelled with currently available tools. For this reason, conclusions for climate change impact on agro-ecosystems based on results of current model outputs must be drawn with care.

Modelling of Maize Production and the Impact of Climate Change…

193

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Gerstengarbe, F.-W., F. Badeck, F. Hattermann, V. Krysanova, W. Lahmer, P. Lasch, M. Stock, F. Suckow, F. Wechsung, P.C. Werner (2003): Studie zur klimatischen Entwicklung im Land Brandenburg bis 2055 und deren Auswirkungen auf den Wasserhaushalt, die Forst- und Landwirtschaft sowie die Ableitung erster Perspektiven. PIK-Report 83, ISSN 1436-0179 Gerstengarbe, F.-W., P.C. Werner, K. Fraedrich (1999): "Applying non-hierarchical cluster analysis algorithms to climate classification: some problems and their solution", Theor. Appl. Climatol., 64 (3-4), 143-150. Gerwitz A. and E.R. Page (1974): An empirical mathematical model to describe plant root systems. J. Appl. Ecol. 11, 773-781. Greenwood D.J., G. Lemaire, G. Gosse, P. Cruz, A. Draycott and J.J. Neeteson (1990): Decline of percentage N of C3 and C4 crops with increasing plant mass. Ann. Bot. 66, 425-436. Hoffmann, F. (1993): Die CERES-Modell – Übersicht, Weiterentwicklung und Erfahrungen. Schriftenreihe Agrarinformatik 24, 139-150. IPCC (2001): "Climate change 2000, Summary for policy makers", Cambridge University Press, Cambridge,UK. Kersebaum, K.C. (1995): Application of a simple management model to simulate water and nitrogen dynamics. Ecological Modelling 81, 145 - 156. Kersebaum, K.C. (2000): Model based evaluation of land use and management strategies in a nitrate polluted drinking water catchment in North-Germany. In: R. Lal (ed.): Integrated Watershed Management in the Global Environment. CRC Press, Boca Raton, 223 - 238. Kersebaum, K.C. (2007): Modelling nitrogen dynamics in soil-crop systems with HERMES. Nutrient Cycling in Agroecosystems, 77, 39-52. Kersebaum, K.C. and A.J. Beblik (2001): Performance of a nitrogen dynamics model applied to evaluate agricultural management practices. In: Shaffer, M. et al. (Eds.): Modeling carbon and nitrogen dynamics for soil management. Lewis Publishers, Boca Raton, 549 569. Kersebaum K.C., K. Lorenz, H.I. Reuter, J. Schwarz, M. Wegehenkel and O. Wendroth (2005): Operational use of agro-meteorological data and GIS to derive site specific nitrogen fertilizer recommendations based on the simulation of soil and crop growth processes. Phys. Chem. Earth 30 (1-3), 59 – 67. Kersebaum, K.C., J. Steidl, O. Bauer and H.-P. Piorr (2003): Modelling scenarios to assess the effects of different agricultural management and land use options to reduce diffuse nitrogen pollution into the river Elbe. Phys. Chem. Earth, 28 (12/13), 537 - 545. van Keulen, H., F.W.T. Penning de Vries and E.M. Drees (1982): A summary model for crop growth. In: Penning de Vries, F.W.T. and Laar H.H. van (Eds.) Simulation of plant growth and crop production. PUDOC, Wageningen, The Netherlands, 87-97. Mirschel, W., H. Förkel and U. Franko (2002): Modulares dynamisches Wachstumsmodell für Zuckerrüben als integrativer Bestandteil von komplexen agrarökologischen Simulationsmodellen. In: Gnauck, A. (Hrsg.): Systemtheorie und Modellierung von Ökosystemen. Umweltwissenschaften, (UmweltWissenschaften), Physika-Verlag (Unternehmen des Springer-Verlages) Heidelberg, 2002, 136 – 156.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

194

K. C. Kersebaum, W. Mirschel, K. O. Wenkel et al.

Mirschel, W., A. Schultz, A.and K.-O. Wenkel (2001): Assessing the Impact of Land Use Intensity and Climate Change on Ontogenesis, Biomass Production, and Yield of Northeast German Agro-landscapes. In: Tenhunen, J.D.; Lenz, R.; Hantschel (eds.): Ecosystem Approaches to Landscape Management in Central Europe. Ecological Studies, Vol. 147, Springer-Verlag Berlin Heidelberg, New York, 2001, 299-313. Mirschel, W. and K.-O. Wenkel (1998): Estimation of Consequences of Climate Changes Using the Agroecosystem Model Family AGROSIM. In: Dalezios, N.R. (ed.): COST 77,79,711- International Symposium on Applied Agrometeorology and Agroclimatology (Volos, Greece, 24 to 26 April 1996), Proceedings, Office for Official Publication of the European Commission (Luxembourg), 1998 (EUR 18328 EN), 67-72. Mirschel, W. and K.-O. Wenkel (2007): Modelling soil-crop interactions with AGROSIM model family. In: Kersebaum,K.C.; Hecker, J.-M.; Mirschel, W.; Wegehenkel, M. (eds.): Modelling water and nutrient dynamics in soil-crop systems (Proceedings of a workshop on “Modelling water and nutrient dynamics in soil-crop systems” held on 14-16 June in Müncheberg, Germany), Springer, Dordrecht, 59-73. Myers, R. J. K., C.A. Campbell and K.L. Weier (1982): Quantitative relationship between net nitrogen mineralization and moisture content of soils. Can. J. Soil Sci. 62, 111 - 124. Nuske, A. (1983): Ein Modell für die Stickstoff-Dynamik von Acker-Lößböden im Winterhalbjahr-Messungen und Simulationen. Ph.D. Thesis, University of Hannover, Hannover, Germany, 164 p. Richter, J., A. Nuske, W. Habenicht and J. Bauer (1982): Optimized N-mineralization parameters of loess soils from incubation experiments. Plant Soil 68, 379-388. Schneider, U. (1991): Messungen von Denitrifikations- und Nitratauswaschungsverlusten in einem landwirtschaftlich genutzten Wassereinzugsgebiet. Ph.D. Thesis, University of Bonn, Bonn, Germany, 86 p. Supit, I., A.A.Hooijer, C.A. van Diepen (Eds.) 1994. System description of the WOFOST 6.0 crop simulation model implemented in CGMS. Vol. 1: Theory and Algorithms. EC Publication EUR 15956, Luxemburg. Weigel, H.J. and U. Dämmgen (2000): The Braunschweig Carbon Projekt: atmospheric flux monitoring and free air carbon dioxide enrichment (FACE). J. Appl. Bot. 74, 55-60. Weigel, H.J., U. Dämmgen, C. Frühauf, S. Burkart and R. Manderscheid (2001): Zwischen Himmel und Erde–Dem Kohlenstoff aus der Atmosphäre auf der Spur. ForschungsReport-Zeitschrift des Senats der Bundesforschungsanstalten, Bundesministerium für Verbraucherschutz, Ernährung und Landwirtschaft, 1/2001, 1418. Weigel, H.J., R. Manderscheid, A. Pacholski, S. Burkart and G. Jansen (2005): Mehr CO2 in der Atmosphäre: Prima Klima für die Landwirtschaft?. ForschungsReport - Zeitschrift des Senats der Bundesforschungsanstalten, Bundesministerium für Verbraucherschutz, Ernährung und Landwirtschaft, 1/2005, 14-17.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 14

MODELLING OF MAIZE PRODUCTION AND THE IMPACT OF CLIMATE CHANGE ON MAIZE YIELDS IN CROATIA Višnja Vučetić∗ Meteorological and Hydrological Service, Grič 3, 10000 Zagreb, Croatia

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The effect of climate change on maize growth and productivity in the central part of Croatia has been researched using the crop CERES-Maize model. The Zagreb Maksimir meteorological data during the period 1949–2004 and pedological, physiological and genetic data obtained in the field maize experiment in Zagreb 1999 have been used. In order to estimate the intensity of the regional impact of climate change on maize production, a synthetic meteorological series was created by the stochastic weather generator MetandRoll for different climate change scenarios. The CERES-Maize model was run with meteorological series representing the present climate and synthetic meteorological series representing the changed climate. All climate change scenarios during the 21st century, including only the climate change effect, projected a shorter growing season of 34-44 days and a reduction in maize yields of 8-15%.

1. INTRODUCTION The weather impact on crop growth, development and yield can be the best represented by agrometeorological (crop-weather) models. One of the most used crop model is DSSAT programme (Decision Support System for Agrotechnology Transfer, Tsuji and Balas, 1993, Hoogenboom et al., 1995) which includes: cereals and maize, leguminous plants and root and tuber crops. Each crop group has its own basic simulation model, which is then adapted to a particular crop. The most widely used are the simulation models for maize and wheat under the common name of CERES (Crop-Environment Resource Synthesis). ∗

[email protected]

196

Višnja Vučetić

Maize and winter wheat being the most important agricultural crops in Croatia, the CERES model for maize has been applied (Vučetić (2006) presented the preliminary results) to investigate the impact of climatic changes on biomass development and maize yield. Dubrovský from the Academy of Sciences of the Czech Republic collaborated in the Croatian Pilot Assessment in the frame of the AGRIDEMA project as provider who developed the stochastic weather generator MetandRoll (Dubrovský, 1996a, 1996b, 1996c, 1997 and 2004) and prepared the climate change scenarios by the pattern scaling technique using the different global climate models.

2. THE DSSAT SIMULATION USING HISTORICAL DATA The DSSAT programme - CERES model for maize (Jones and Kiniry, 1986, Ritchie et al., 1990, Hunkar, 1994), besides simulating maize growth, development and yield also assesses the commencement of the phenological phases, soil water balance and soil nitrogen transformation. The meteorological data used by the model are daily values of maximum and minimum air temperature, precipitation amount and global solar radiation. As the field maize experiment in 1999 was carried out at the farm of the Zagreb University Faculty of Agriculture, the meteorological data used in the analysis were taken from the nearest meteorological station, Zagreb Maksimir (1949-2004), located at about 650 m from the field experiment site. Samples of the vertical pedological profile were taken for the chemical and physical analysis of the soil about two weeks before sowing, according to IBSNAT recommendations (1990a and 1990b). The predicted values of the CERES model derived from long-term series of meteorological data (1949–2004) and from pedological, physiological and cultivation data measured during the 1999 field experiment are very similar to the observed values: beginning of silking and physiological maturity, kernel mass and maximum leaf area index LAI (Table 1). The model underestimated the 1999 yield and biomass per hectare, the grain N (%) and the total N uptake.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Comparison of the predicted values according to the CERES-Maize model and the observed values collected during the field experiment and from the pedological profile at the farm of the Faculty of Agriculture at Zagreb in 1999 Descriptions Silking date Physiological maturity date Grain yield (kg/ha) Kernel mass (g) Maximum LAI (m2/m2) Biomass (kg/ha) N grain (%) Total N uptake (kg/ha)

Predicted 201 261 9773 0.380 4.7 17121 1.03 118.0

Observed 200 258 13095 0.347 4.8 22389 1.30 177.0

Modelling of Maize Production and the Impact of Climate Change…

197

Vegetation period in 1999 was extremenly warm but the precipitation amount was at an average. The model underestimated the maize productivity due to too warm condition during the vegetation period. A good assessment is a deviation of predicted and observed variables to 20% (Alexandrov et al., 2001). As the main goal was to investigate the impact of weather conditions on maize yield during the long period, the CERES model was also run with the same input values of plant and soil characteristics as for 1999 year but with the varied daily meteorological data from year to year during the period 1949–2004. Thus, 56-year time series were estimated for the beginning of silking and physiological maturity dates, grain yield, kernel mass, biomass, maximum LAI, grain N (%) and total N uptake. The linear trends of particular maize parameters and the non-parametric Mann-Kendall test (Mitchell et al., 1966, Sneyers, 1990) indicated a significantly earlier start (2 days/10 years) for silking and 5 days/10 years for physilogical maturity (Figure 1).

Zagreb Maksimir (1949-2004) 2-Nov

MATURITY DATE

2-Oct

1-Sep

y = -0.5212x + 291

-5.1 day/10 years SILKING DATE

1-Aug

y = -0.167x + 211 -1.6 day/10 years 1-Jul 1949 1955 1961 1967 1973 1979 1985 1991 1997 2003 year

21000

Zagreb Maksimir (1949-2004)

19000

BIOMASS

17000 kg/ha

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

y = 8.4105x + 17890

83 kgha-1/10 years

15000 GRAIN YIELD

13000 11000 9000 y = -4.8324x + 11445

-48 kgha-1/10 years

7000 1949 1955 1961 1967 1973 1979 1985 1991 1997 2003 year

Figure 1. Predicted time series and linear trends of the commencement of silking and maize physiological maturity (days), grain yield and biomass (kg/ha) according to the CERES-Maize model for Zagreb Maksimir in the period 1949–2004. Linear trends significant at the 0.05 level are bolded.

198

Višnja Vučetić

The linear trend analysis showed a slight decrease in maize yield (48 kg/ha) but an increase in biomass by 83 kgha-1/10 years. This significant earlier beginning of silking and physiological maturity started in 1995. It is result of the significant positive linear trend in air temperature especially Tmin (0.4°C/10 years) which started in the late 1980s.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. THE DSSAT SIMULATION USING DIFFERENT CLIMATE CHANGE SCENARIOS In order to estimate the intensity of regional impact of climate change on maize production in Croatia, the synthetic meteorological series was created by the stochastic weather generator MetandRoll for present climate and different climate change scenarios. Dubrovsky designed the weather generator (WG) MetandRoll to provide the synthetic meteorological series of daily data: global solar radiation (SRAD), maximum (Tmax) and minimum (Tmin) temperatures and precipitation amount (PREC) for the CERES-Maize crop model. The WG parameters derived from the long time series (1949–2004) for the Zagreb Maksimir station to generate synthetic meteorological series representing the present climate. The results of validation presented that the WG MetandRoll well preserves some features of the stochastic structures of daily meteorological series. However, the discrepancies were found in reproducing the shape of the distributions of SRAD and PREC. Fortunately, some discrepancies in the cold season are not influence on the maize production. Further investigation involved modification of the WG parameters in accordance with the climate change scenarios and the generation of synthetic meteorological series representing the changed climate. The climate change scenarios were prepared by the pattern scaling technique using the following global climate models (GCM): ECHAM4/OPYC3, HadCM3 and CSIRO-Mk2. As a validation analysis (Dubrovský et al., 2005) showed that these GCMs were a good choice for a representative set of climate changes scenarios in the Czech Republic, the same GCMs were used for Croatia. The changes in global mean temperature (ΔTG) were estimated by the 1-dimensional climate model MAGICC for different emission scenarios and climate sensitivities (IPCC, 1997 and 1999). The range of the CO2 concentration in the newer emission scenarios (SRES A1, A2, B1 and B2) is from 548 ppm for SRES B1 to 826 ppm for SRES A2, where 333 ppm is the baseline CO2 level. According to the IPCC proposal, the range values of ΔTG relate to climate sensitivities within 1.5 to 4.5°C. The values of the scaling factor for the middle emission scenario combined with intermediate climate sensitivity (ΔTG = 2.5°C) at the end of 21st century were used. This intermediate scaling factor was obtained as an average from (emission scenario SRES B2 + middle climatic sensitivity) and (emission scenario SRES A1 + middle climatic sensitivity). When the three climate change scenarios had been prepared, the WG MetandRoll was applied to generate a 99-years synthetic meteorological series representing the changed climate. After that, the CERES-Maize model was run with this synthetic meteorological series. In the future climate these scenarios at the end of the 21st century projected an increase: in SRAD 3–7%, in Tmin around 3°C and in Tmax 3–4°C and a decrease in PREC for 8% except HadCM3 showed the increase 2% in the central part of Croatia. All transient climate change scenarios during the 21st century, including only the climate change effect, projected a shorter

Modelling of Maize Production and the Impact of Climate Change…

199

growing season and a reduction in maize yields. The main conclusion is that the maize vegetation period, including only the climate change effect, in central part of Croatia would be 34 days shorter for CSIRO-Mk2, 43 days for ECHAM4/OPYC3 and 44 days for HadCM3, which would result in 10%, 8% and 15% smaller yields for maize, respectively, at the end of the 21st century (Table 2). Bacsi and Hunkar (1994) obtained a similar maize result for Hungary when including only the climate change effect. Research in Slovenia shows that on the assumption of a 2°C increase in temperature the cultivation area should be raised to a higher altitude (Kajfež-Bogataj, 1993, 1996 and 1998). It has been predicted that the maize yield at 500 m under present conditions would correspond to the yield at 900 m at the end of 2100. Table 2. Predicted mean (MEAN) values of particular maize parameters according to the CERES-Maize model based on the Zagreb Maksimir meteorological data for the 1949–2004 period and the synthetic meteorological series for various climate scenarios which were prepared by the global climatic models: ECHAM, HadCM and CSIRO for Zagreb at the end of 21st century. STD is the standard deviation Physiolog maturity date Zagreb Maksimir (1949–2004) MEAN 25-Jul 3-Oct STD 6 17 ECHAM 4/OPYC3 MEAN 8-Jul 21-Aug STD 3 3 HadCM3 MEAN 8-Jul 20-Aug STD 3 3 CSIRO-Mk2 MEAN 12-Jul 27-Aug STD 3 3

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Silking date

Biomass (kg/ha)

Grain yield (kg/ha)

Grain N (%)

Kernel mass (g)

Max. LAI (m2/m2)

18130 962

11307 977

0.91 0.06

0.36 0.02

4.5 0.2

18503 689

10415 899

1.07 0.07

0.29 0.01

4.7 0.1

17737 1079

9646 1217

1.01 0.08

0.27 0.02

4.7 0.2

17963 761

10146 993

1.01 0.07

0.29 0.02

4.7 0.2

As carbon dioxide in the atmosphere directly affects plants by increasing photosynthesis and decreasing transpiration the both effects (direct CO2 effect and indirect climate change effect) showed a 17–18% increase in stressed maize yield and 5–14% in potential maize yield in the Czech Republic (Dubrovský et al., 2000, Žalud and Dubrovský, 2002). It is also shown that the increase in the simulated yields of other agricultural crops (e.g. winter wheat, soybean etc.) for the 21st century was primarily because of the beneficial influence of the direct CO2 effect (Alexandrov et al., 2002).

CONCLUSION The CERES-Maize model results using different climate scenarios for the central part of Croatia describe the vulnerability of agroecological systems affected by possible climatic

200

Višnja Vučetić

changes, including only the indirect CO2 effect. Further investigations are to estimate the climate change impact on maize productivity using a different climate scenarios: low/high emission scenario combined with low/high climate sensitivity (increase in the global mean temperature is 1.5°C/4.5°C) during the 21st century and to simulate and compare the direct CO2 effect (through the increased fertilization effect of ambient CO2) and the indirect CO2 effect (through changed weather) on maize yields.

ACKNOWLEDGMENTS This research has been carried out as a part of the the project Climate variations and changes and response in affected system of the Ministry of Science, Education and Sport of Republic of Croatia and the AGRIDEMA project Introducing tools for agricultural decisionmaking under climate change conditions by connecting users and tool-providers. I thank the Faculty of Agronomy of Zagreb University and the Geophysical Department of the Faculty of Natural Sciences of the Zagreb University for their help. I also wish to thank Lučka KajfežBogataj, Marta Hunkar, Gordon Y. Tsuji, James R. Kiniry, Josef Eitzinger and Martin Dubrovský for their support.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Alexandrov, V., J. Eitzinger, V., and M. Oberfoster, 2001: Adaptation of crop-weather models in Austria and Bulgaria, Proceedings of 5th European Conference on Applications of Meteorology, Budapest, Hungary, 24–28 September 2001, 22 pp (CD-Rom). Alexandrov, V., J. Eitzinger, V., Čajić, V. and M. Oberfoster, 2002: Pontential impact of climate change on selected agricultural crops in north-eastern Austria, Global Change Biology, 8, 372–389. Bacsi, Zs. and M. Hunkar, 1994: Assessment of the impacts of climate change on the yields of winter wheat and maize using crop models, Időjárás, 98, 2, 119–134. Dubrovský, M., 1996a: MetandRoll: the stochastic generator of daily weather series for the crop growth model (in Czech), Meteorological Bulletin, 49, 97–105. Dubrovský, M., 1996b: Validation of the stochastic weather generator MetandRoll (in Czech), Meteorological Bulletin, 49, 129–138. Dubrovský, M., 1996c: MetandRoll: the weather generator for the crop growth model, Regional workshop on Climate variability and climate change vulnerability and adaptation, Prague, Czech Republic, 1995, Institute of Atmospheric Physics - U.S. Country Studies Program, Washington, D.C., 285–292. Dubrovský, M., 1997: Creating weather serious with use of the weather generator, Environmetrics, 8, 409–424. Dubrovský, M., Žalud Z. and Šta’stná M., 2000: Sensitivity of CERES-Maize yields to statistical structure of daily weather series, Climatic Change, 46, 447–472. Dubrovský, M., Buchtele J. and Z. Žalud, 2004: High-frequency and low-frequency variability in stochastic daily weather generator and its effect on agricultural and hydrologic modelling, Climatic Change, 63, 145–179.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Modelling of Maize Production and the Impact of Climate Change…

201

Dubrovský, M., Nemesova, I. and J. Kalvova, 2005: Uncertainties in climate change scenarios for the Czech Republic, Climate Research, 29, 139–156. Hoogenboom, G., G:Y: Tsuji, N. B. Pickering, R. B. Curry, J. W. Jones, U. Singh and D. C: Godwin, 1995: Decision support system to study climate change impacts on crop production, Climate change and agriculture: Analysis of potentional international impacts, ASA, Special publication, 59, 51–75. Hunkar, M., 1994: Validation of crop simulation model CERES-Maize, Időjárás, 98, 1, 37–46. IBSNAT, 1990a: Field and laboratory methods for the collection of the IBSNAT, Technical Report 2, 67 pp. IBSNAT, 1990b: Documentation for IBSNAT, Crop model input and output files, Technical Report 5, 61 pp. IPCC, 1997: An introduction to simple climate models used in the IPCC, Second Assessment Report, IPCC Tech Paper 2, Geneva. IPCC, 1999: Guidelines on the use of scenario data for climate impact and adaptation assessment, 69 pp. Jones, C. A. and J.R. Kiniry, 1986: CERES-Maize, A simulation model of maize growth and development, Texas, University press, College station, 193 pp. Kajfež-Bogataj, L, 1993: Impacts on future climate change on spring barley and maize yield in Slovenia, Journal of Agricultural Meteorology, 48, 627–630. Kajfež-Bogataj, L, 1996: Effects of climate warning on CERES-Maize field in Slovenia: Sensitivity study, Res. Report, Biotechnical Faculty of the University in Ljubljana, 67, 11–18. Kajfež-Bogataj, L, 1998: Potential crop shifts to higher altitude in Slovenia due to climatic change, Proceedings of Agriculture and forestry – adaptability to climate change, Zagreb, Croatia, 19–20 May 1998, 143–152. Mitchell, J.M. Jr., B. Dzerdzeevskii, H. Flohn, W.L. Hofmeyr, H. H. Lamb, K. H. Rao and C.-C. Wallen, 1966: Climatic change, WMO Tech. Note 79, Geneva, 58–75. Ritchie, J., Singh, U., D. Godwin, and L. Hunt, 1990: A users guide to CERES-Maize – V2.10, 86. pp. Sneyers, R., 1990: On the statistical analysis of series of observations, WMO Tech. Note, 143, 1–15. Tsuji, G. and S. Balas (eds), 1993: The IBSNAT decade, University of Hawaii, Honolulu, 178 pp. Vučetić, V., 2006: Impact of climate change on the maize productivity in Croatia, Proceedings of Abstracts of 6thEMS/6th ECAC, Ljubljana, Slovenia, 4–8 September 2006, (CD-Rom). Žalud Z. and Dubrovský M., 2002: Modelling climate change impacts on maize growth and development in the Czech Republic, Theoretical and Applied Climatology, 72, 85–102.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 15

CONSEQUENCES OF CLIMATE CHANGE ON IRRIGATION WATER REQUIREMENTS IN SOUTHERN SPAIN J. A. Rodríguez Díaz1, 2∗, E. K. Weatherhead2, J. W. Knox2 and E. Camacho1 1

Department of Agronomy. University of Cordoba. Spain 2 Centre for Water Science, Cranfield University. UK

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Irrigated production in Southern Spain, in particular the Guadalquivir River basin, has grown significantly over the last decade. As a consequence, water resources are under severe pressure, with an increasing deficit between available supplies and water demand. Substantial investments have been made in conversion from open channels to pressurized networks and replacing surface irrigation by more efficient irrigation methods such as trickle. However, climate change is a new threat which could severely aggravate the situation. In this study, the impacts of climate change on irrigation water demand have been estimated for 14 of the most representative irrigation districts in the Guadalquivir river basin The results show a typical increase of 15% - 20% in seasonal irrigation demand by 2050., because of higher evapotranspiration rates and lower precipitation, with a longer irrigation season. The peak demands are also expected to increase, by around 11 -12 %, with implications for the capacity of existing systems.

INTRODUCTION AND BACKGROUND The Guadalquivir river basin has an estimated irrigated area of 714015 ha, which represents 25.5% of the total cropped area. This irrigated-cropping area ratio is twice that for Spain as a whole (12.5%). Irrigation allows farmers to assure their summer production and to ∗

Corresponding author: Juan Antonio Rodríguez Díaz. Department of Agronomy. University of Cordoba. Campus Rabanales. 14071 Córdoba. Spain. [email protected]

204

J. A. Rodríguez Díaz, E. K. Weatherhead, J. W. Knox et al.

produce high-value crops which would otherwise be impossible to cultivate. Thus, in Guadalquivir, one irrigated hectare has a yield six times bigger than in rainfed conditions and it is four times more profitable for farmers. Also the contribution to employment from one irrigated hectare is estimated to be 3.5 times higher than from one un-irrigated hectare (Berbel and Gutiérrez, 2004). As a result, irrigated agriculture has become a substantial wealth generator and an important element of the region’s rural based economy. However, irrigated agriculture consumes 80% of the total available water resources locally. The water sources used for irrigation abstraction are split between surface water (80.9%), groundwater (18%) and wastewater (1.1%) (Aquavir, 2005). A wide range of crops are cultivated in the basin. In terms of irrigated area, olive trees constitute the most important (42.3%), followed by extensive field crops (39.4%), mainly cotton, sugar beet, rice, wheat, maize and sunflower. The areas cropped with outdoor vegetables (9.6%), fruit trees (3.8%) and citruses (2.7%) are much lower. However, in terms of productivity (gross value per unit of water applied), field crops have been estimated to generate 0.21 € per m3, compared to 0.81 € per m3 for tree crops and 1.60 € per m3 for outdoor vegetables (Consejería de Agricultura y Pesca, 1999). The economic importance of these ‘other’ crops is therefore much higher than their relative irrigated areas alone might suggest. The recent growth in irrigated area in the basin has been dramatic. Since 1900, the irrigated area has increased by 500%, from 142900 ha in 1904 to 714015 ha in 2004 (Camacho, 2005). The increase has been particularly rapid in the last decade. Due to this expansion, the demand for water has also increased considerably. As a consequence, there is now significant pressure on local water resources; by 1998 it was estimated that there was a 13% deficit between available water supplies and water demand (Ministerio de Medio Ambiente, 1998). The rising demand for irrigation water coinciding with a series of concurrent dry years with reduced recharge has undoubtedly increased this water deficit. At present water resources management is based on the Basin Hydrological Plan (BHP) (CHG and MIMAM, 1995). This document stated the concept of supply warranty for each user according to its priority. Supply warranty is defined as the probability of demand to be satisfied. This is a key concept in water resources systems exploitation. Particularly, the BHP considers that the irrigation demand is satisfied if the supplied water varies from 100% down to 75% of the full water right.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The BHP recommendations to assure irrigation demand are: Maximum annual deficit between 20-40% of full water rights Maximum two years aggregated deficit between 30-60% of full water rights Maximum ten years aggregated deficit between 40-80% of full water rights However, in times of water scarcity the assigned water can be reduced below the original concession. Table 1 shows the overall reductions in water assigned to the irrigation districts in the case study river basin over the last 22 irrigation seasons. In only 8 of 22 years (36%) have farmers actually received their full allocation. In most years, the water assigned was considerably lower, with adverse impacts on crop production. In 5 years (23%) the assigned water was reported to be a third or less of the concession (Camacho, 2005).

Consequences of Climate Change on Irrigation Water Requirements…

205

Table 1. Summary of ratios of assigned water to concession, number of equivalent irrigations, and likely impacts on crop production between, 1982 and 2004 (Camacho, 2005) Ratio of assigned water to concession (%) 80-100 60-80 35-60 0-35

Number of irrigation seasons 8 7 2 5

36 32 9 23

Impacts on crop production Low Medium High Very high

Total

22

100

-

%

Water scarcity in the river basin varies mainly depending on the weather in each year. Figure 1 shows the average volume of water (m3 ha-1) assigned to 30 irrigation districts selected as being representative in terms of crops grown and irrigation systems used. The volumes of water assigned were severely reduced between 1992 and 1996 when Spain was in a protracted drought. Indeed, in 1995 farmers could not irrigate at all (Camacho, 2005; Camacho and Rodríguez Díaz, 2006). Due to concerns regarding long-term water scarcity, and in order to conserve available supplies, some of the largest irrigation districts are in the process of system modernisation. The old open channel networks are being replaced by ‘on-demand’ pressurized networks. The primary aim of these investments is to achieve more efficient conveyance and use of water. As a consequence, nearly half (45%) of the total irrigated area relies on micro (trickle) irrigation, which is now the most common application method in the basin (Aquavir, 2005). This is in contrast to 15 years ago when surface irrigation was predominant (61%) and trickle (12%) was still regarded as a specialist technique. 8000 7000 6000

3

m /ha

4000 3000 2000 1000

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

0 1982

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5000

Year

Figure 1. Reported average volumes (m3 ha-1) of assigned water (1982 to 2004) for 30 irrigation districts in the Guadalquiver river basin (Camacho, 2005).

206

J. A. Rodríguez Díaz, E. K. Weatherhead, J. W. Knox et al.

However, irrigation in the Guadalquivir river basin will have to face another threat in the next century, from climate change. The predicted impacts are expected to be particularly severe in southern Spain. The latest climate change predictions for Spain suggest a significant increase in temperature (up by 0.4 degrees per decade in winter and 0.7 degrees per decade in summer) and a reduction and changed annual distribution of rainfall during the 21st century (Moreno, 2005). These changes in climate will have significant impacts on irrigation water demand. The situation will be even more dramatic considering that some authors predict significant decreases in water resources for the basin because of the new precipitations regime (Ayala, 2002). In this study, the predicted impacts of climate change on the local agroclimate and on irrigation water demand in the Guadalquivir basin have been modelled and mapped. The outputs provide a basis for identifying adaptation options that might be required by farmers in the region in order to respond to changing water availability.

METHODOLOGY Climate Change Scenarios and Modelling

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In this study, climate change data derived from the HadCM3 global climate model have been used. The HadCM3 model is a coupled atmosphere-ocean general circulation model (AOGCM) developed at the Hadley Centre for Climate Prediction and Research (Gordon et al., 2000; Pope et al., 2000). This model has a spatial resolution of 2.5° x 3.75° (latitude by longitude); this produces a grid of 72 x 96 cells, resulting in a grid resolution of 280 km x 320 km across Spain. For this study, the SRES scenarios A2 and B2 have been used, respectively representing strong economic values (A2) and strong environmental values (B2), both under increasing regionalization (Nakicenovic et al., 2000). The changes predicted by the HadCM3 model must be applied to a ‘baseline’. In this research, changes have been applied to a baseline climatology developed by the International Water Management Institute (IWMI). This is a 10’ latitude/longitude dataset containing mean monthly surface climate over global land areas, excluding Antarctica, for the period 1961 to 1990 (New et al., 2002). The scenarios have then applied to this baseline to generate perturbed climatic datasets for future defined time periods.

Agroclimatic Indicator To correlate irrigation need with climate the agroclimatic indicator PSMD “Potential Soil Moisture Deficit” was used. PSMD relates irrigation demand to the main climate drivers: rainfall and evapotranspiration. Previous authors have shown that there is a correlation between the value of the indicator and irrigation water demand (e.g. Knox et al., 1997). To estimate PSMD, a monthly time-step water balance model was used, working from rainfall and reference evapotranspiration (ETo) data. The PSMD for each grid pixel at the end of each month is calculated from:

Consequences of Climate Change on Irrigation Water Requirements… PSMDi = PSMDi-1 + EToi – Pi

207 [1]

where; PSMDi = potential soil moisture deficit at the end of month i, mm PSMDi-1 = potential soil moisture deficit at the end of month i-1, mm EToi = potential evapotranspiration in month i, mm Pi = precipitation in month i, mm At the start of the irrigation season the PSMD is assumed to be zero. If PSMDi-1 as calculate above is less than zero, for example after heavy rainfall, the previous soil moisture deficit is assumed to have been filled, with any excess precipitation lost as runoff or deep percolation, and the PSMDi is reset to zero. PSMD were calculated for all the pixels in Spain and results were mapped using a GIS.

Crop Modelling

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Fourteen irrigation districts, typical in terms of crop, location and irrigation systems, were selected for study, accounting a total irrigated surface of more than 145000 ha (approximately 20% of the total irrigated area in the basin). Their distribution along the basin is shown in Figure 2. For each irrigation district, the typical cropping patterns were calculated using a six year dataset of crops and irrigated areas. Using these defined cropping patterns and the derived climate datasets for the baseline and each SRES scenario (2050_A2 and 2050_B2), the crop water requirements, seasonal irrigations needs and peak water demand in each irrigation district were simulated using a computer model termed CROPWAT (Clarke et al., 1998).

Figure 2. Spatial distribution of irrigated areas and the selected irrigation districts in the Guadalquiver river basin.

208

J. A. Rodríguez Díaz, E. K. Weatherhead, J. W. Knox et al.

RESULTS Mapping Climate Change Impacts

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The climate change predictions for Spain suggest increases in temperature in both summer and winter. Average annual precipitation is expected to reduce, but the main change will be an altered seasonal distribution, with more rainfall in winter and less in summer. The average evapotranspiration in the country may increase by 8 % for 2050s and around 20 % for 2080s.

Figure 3. Predicted changes in agroclimate (PSMD, mm) from the baseline (1961-1990) to the 2050s for selected SRES emissions scenario (B2 and A2).

Consequences of Climate Change on Irrigation Water Requirements…

209

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Taking the outputs climate change model HadCM3 and applying the predicted percentage changes to the baseline, the future climate scenarios have been mapped for Spain using the agroclimatic indicator PSMD described above. The large climate datasets from the climate change model have thus been reduced into a single number for each single pixel which can then be easily related to irrigation water needs. The PSMD values for the entire country have been imported into a GIS to map the spatial distribution of PSMD as shown in Figure 12. The model predicts an average increase of PSMD values of more than 20 % for 2050 and around 50 % for 2080. In the particular case of Guadalquivir river basin, the agroclimate maps show that the PSMD will reach significantly higher values throughout the entire basin (the average PSMD values for the IPCC scenarios A2 and B2 will increase by 16% and 12%, respectively.). The impacts are slightly greater under the 2050_A2 scenario than the 2050_B2 scenario. The central areas of the basin, where most of the irrigation districts are concentrated, could potentially have extremely high PSMD values, ranging between 1050 and 1300 mm, compared to 800 to 900 mm for the baseline (1961-90). For 2080, PSMD in most of the basin could reach up to the range 1300 and 1450 mm. Figure 4 shows the ranked evolution of PSMD for a single point in the south of the country in the last 25 years and the predicted average values for 2050s and 2080s scenarios. Of significant relevance to water resource planning and irrigation management is that these values for the 2050s are similar to the most extreme driest years experienced in the last 25 years, such as 1995 and 2005. These recent ‘actual’ dry years could therefore be used as reference or benchmark for growers and the water regulatory authorities when considering adaptation strategies or assessing the likely impacts of ‘future’ climate change on water resources for irrigation. For the 2080s, the average PSMD values could be even much higher than the values of the indicator in the driest years registered in the past (Rodríguez Díaz et al., 2006). Of course, there will still be variability between years superimposed on these average values, and about half the years will be even drier.

Figure 4. Ranked PSMD in the last 25 years for a single point and predicted average values for the future.

210

J. A. Rodríguez Díaz, E. K. Weatherhead, J. W. Knox et al.

Impacts on Crop Water Requirements Crop water requirements were calculated for each irrigation district according to their typical cropping patterns the present and for both future scenarios, A2 and B2. The modelled crop water requirements vary between districts but the average predicted increases are 9.4% and 8.3% (2050_A2 and 2050_B2), respectively. However, the average increase in irrigation needs is approximately double that (19.3% in 2050_A2 and 16.3% in 2050_B2). This is due both to the reduction in rainfall and to its changed distribution over the year (Rodríguez Díaz et al., 2007). The modelled peak monthly irrigation needs are also strongly impacted by climate change. The average increases are 12% and 11% (2050_A2 and 2050_B2); but in some districts values of up to 17% have been predicted. For the existing irrigation networks, which were designed for much lower and shorter peak demands, these predicted changes would cause major problems in meeting higher peak flows.

CONCLUSIONS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Despite the investments in modernisation of the irrigation districts, the growth of the irrigated surface in the Guadalquivir river basin during the last decade (60% from 1995) and the present water deficit already makes the present situation difficult to sustain. In only 8 of the last 22 years could farmers use their full water rights because of the water scarcity. Due to climate change, additional increases of around 20% and 16 % (scenarios 2050_A2 and 2050_B2) in the irrigation needs are predicted by the 2050s. The irrigation seasons are also predicted to be longer than at present due to the lower rainfall between April to June. The increases are particularly severe in the west of the basin where most irrigated agriculture is currently located. Irrigation systems will have to be designed for longer and higher peaks (11% - 12 %) in irrigation water demand. That may cause problems in some of the existing networks. These changes are likely to be co-incident with significant reductions in water availability, also due to climate change. Reduction in irrigated area may be inevitable, with implications for the regional economy and employment. Under these conditions of increasing water scarcity, promoting more efficient use of the water will become even more important.

REFERENCES Aquavir (2005) Superficies de los cultivos de regadío y sus necesidades de riego, en la demarcación de la Confederación Hidrográfica del Guadalquivir. CHG. Spain. Ayala FJ (2002) Notas sobre los impactos físicos previsibles del cambio climático sobre los lagos y humedales españoles. III Congreso Ibérico sobre gestión y planificación de aguas. Sevilla. Spain. Berbel J, Gutiérrez C (2004) I Estudio de Sostenibilidad del regadío del Guadalquivir. FERAGUA. Spain.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Consequences of Climate Change on Irrigation Water Requirements…

211

Camacho E (2005) Análisis de la eficiencia y el ahorro de agua en el regadío de la cuenca del Guadalquivir. Inversiones en la modernización de regadíos. FERAGUA. Spain. Camacho E, Rodríguez Díaz JA (2006) Informe anual del sector agrario de Andalucía. Chapter: Situación actual y futura del regadío en la cuenca del Guadalquivir: La modernización y los efectos del cambio climático. Analistas Económicos de Andalucía. Spain. Confederación Hidrográfica del Guadalquivir (CHG) and Ministerio de Medio Ambiente (MIMAM) (1995): “Plan Hidrológico del Guadalquivir”. Consejería de Agricultura y Pesca (1999) Inventario y Caracterización de los Regadíos de Andalucía. Junta de Andalucía. Spain. Clarke D, Smith M, El-Askari K (1998) CropWat for Windows: User Guide. FAO, Rome. Gordon C, Cooper C, Senior CA, Banks H, Gregory JM, Johns TC, Mitchell JFB, Wood RA (2000) The simulation of SST, sea ice extents and ocean heat transports in a version of the Hadley Centre coupled model without flux adjustments. Climate Dynamics 16: 147168. Knox JW, Weatherhead EK, Bradley RI (1997) Mapping the total volumetric irrigation water requirements in England and Wales. Agricultural Water Management 33(1): 1-18. Ministerio de Medio Ambiente (1998) Libro blanco del agua en España. Spain. Moreno JM (2005) Evaluación preliminar de los impactos en España por efecto del cambio climático. Ministerio de Medio Ambiente. Spain. Nakicenovic N, Alcamo J, Davis G, de Vries B, Fenhann J, Gaffin S, Gregory K, Grübler A, Jung TY, Kram T, La Rovere EL, Michaelis L, Mori S, Morita T, Pepper W, Pitcher H, Price L, Raihi K, Roehrl A, Rogner H, Sankovski A, Schlesinger M, Shukla P, Smith S, Swart R, van Rooijen S, Victor N, Dadi Z (2000) IPCC Special Report on Emissions Scenarios Cambridge. University Press. Netherlands. New M, Lister D, Hulme M, Makin I (2002) A high resolution data set of surface climate over global land areas. Climate Research 21, 1-25. Pope VD, Gallani ML, Rowntree PR, Stratton RA (2000) The impact of new physical parametrizations in the Hadley Centre climate model: HadAM3. Climate Dynamics. 16: 123-146. Rodríguez Díaz JA, Knox JW, Weatherhead EK (2006) Uso del indicador agroclimático PSMD para la representación y evaluación del cambio climático en España. Ingeniería del agua. 13(4): 311-319. Rodríguez Díaz JA, Weatherhead EK, Knox JW, Camacho E (2007) Climate change impacts on irrigation water requirements in the Guadalquivir river basin in Spain. Regional Environmental Change. DOI 10.1007/s10113-007-0035-3.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 16

ADAPTATIONS OF IRRIGATED CROPPING SYSTEMS OF SOUTHERN ITALY AS AFFECTED BY CLIMATE CHANGE AT FIELD/FARM SCALE Domenico Ventrella1, Nicola Losavio1, Rita Leogrande1, Luisa Giglio1, Mirko Castellini1, Enza Di Giacomo1, Angel Utset2, Juan Carlos Martinez2, Javier Rojo2, Blanca del Rio2 and Dimos Anastasiou3 1

2

CRA-Istituto Sperimentlae Agronomico, Bari, Italy Instituto Tecnologico Agrario de Castilla y Leon, ITACyL, Spain 3 Fthiotida Development Agency, FDA, Greece

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Precipitation and temperature in the Mediterranean area in the last century show significant trends. Some authors found a reduction for winter precipitation. However, sub-regional variability is high and trends in many regions are not statistically significant in view of the large variability. The forecasted climate changes require a re-evaluation of the Mediterranean cropping system in order to maximize the use efficiency of exogenous resources. Because of high evaporative demand of the atmosphere the irrigation management is one of the most important factors to preserve high yield level and soil fertility. In particular, climate warming could have a substantial impact on some agronomical practices as the choice of the crops and/or varieties to be included in the cropping systems, the sowing time, conservative tillage for reducing soil evaporation and runoff and increasing water infiltration. Concerning the irrigation management, the scheduling (how much and when irrigate), the irrigation method choice and the soil water status monitoring are the most important factor to be optimized in order to obtain high value of water use efficiency (WUE). In this AGRIDEMA pilot assessment the model SWAP (Soil-Water-Atmosphere- Plant: Van Dam, J.C., Huygen, J., Wesselling, J.G., Feddes, R.A., Kabat, P., van Walsum, P.E.V., Groenendijk, P., van Diepen, C.A. 1997. Theory of SWAP version 2.0. Simulation of water flow, solute transport and plant growth in the Soil-Water-Atmosphere-Plant environment. Report 71, Alterra, Wageningen, The Netherlands, 167 pp.) was used in order to individuate integrated approaches for

214

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al. optimizing water resources use by maximising the cropping systems WUE, for two “real” cropping systems: sorghum and a winter vegetable cultivation. For this purpose several irrigation strategies were defined including different scheduling and salinity levels of water irrigation: (i) a fix watertable at 6 m depth (DG) and (ii) a fluctuant and shallow watertable from 0 to 3 m depth (SG). The data-set used for the calibration and validation was collected from a long-period research carried out in an experimental farm located in Metaponto (Italy) within the ionical coastal zone in Southern Italy. The two bottom boundary conditions influenced significantly the seasonal fluxes of the soil water balance. The irrigation amounts were higher for DG than SG scenarios that was characterized by an higher upward flux from the watertable and, above all, by an higher soil water content at the crop sowing time. The reference Winter crop cultivation was carried out in a private farm close to the experimental farm of CRA-SCA on sand soil from October to March. For this pilot assessment we used a input data-set concerning crop growth, phenology, morphology and soil hydraulic parameters that comes from our previous research in order to parameterize and SWAP for this cropping system. Moreover an intensive geostatistical study of physical soil properties was applied on two 10-ha fields integrated with a characterization of the hydraulic properties. The pilot assessment for this cultivations included the following steps: (i) sensitivity analysis of soil hydraulic parameters for soil-crop water fluxes; (ii) evaluation of agronomical practices to alleviate the climatic change.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

BACKGROUND The “Mediterranean” climate is characterized by mild and wet winters and hot-dry summers. This may occur on the West side of continents between about 30° and 40° latitude. However, the presence of a relatively large mass of water is unique to the actual Mediterranean region. Precipitation and temperature in the Mediterranean area in the last century show significant trends. Some authors found a reduction for winter precipitation. However, subregional variability is high and trends in many regions are not statistically significant in view of the large variability. During winter, the prevailing winds are from North (Tramontana) and North-West (Mistral), bringing winter rain with snow and gales. During summer, heavy thunderstorms bring rain that rapidly evaporates. The Scirocco (South-East) is a hot wind that in summer moves from Africa to the whole country. With its hot-dry summers and cool-wet winters, Southern Italy experiences a Mediterranean climate. Altitude, distance from the sea and aspect give local weather variations with the coastal zones kept warm by the sea influence. The forecasted climate changes require a re-evaluation of the Mediterranean cropping system in order to maximize the use efficiency of exogenous resources. Because of high evaporative demand of the atmosphere the irrigation management is one of the most important factors to preserve high yield level and soil fertility. In particular, climate warming could have a substantial impact on some agronomical practices as the choice of the crops and/or varieties to be included in the cropping systems, the sowing time, conservative tillage for reducing soil evaporation and runoff and increasing water infiltration. Concerning the irrigation management, the scheduling (how much and when irrigate), the irrigation method

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Adaptations of Irrigated Cropping Systems of Southern Italy…

215

choice and the soil water status monitoring are the most important factor to be optimized in order to obtain high value of water use efficiency (WUE). The impact of climatic change on agricultural activity will depend also on the relationships in the continuum “soil-plant-atmosphere” and this continuum has to be included in the analysis for forecasting purposes and the combination of effects concerning increase of CO2 and temperature increase, change in evaporation and change rainfall variables. In this AGRIDEMA pilot assessment we used the the model SWAP (Soil-WaterAtmosphere- Plant: Van Dam, J.C., Huygen, J., Wesselling, J.G., Feddes, R.A., Kabat, P., van Walsum, P.E.V., Groenendijk, P., van Diepen, C.A. 1997. Theory of SWAP version 2.0. Simulation of water flow, solute transport and plant growth in the Soil-Water-AtmospherePlant environment. Report 71, Alterra, Wageningen, The Netherlands, 167 pp.). SWAP is a physically based model that resolves the Richards equation to describe soil water fluxes and solute transport in saturated/unsaturated soil media. The upper boundary is located just abve the crop, while the lower bondary in the top growndwater system. Into the soil the main water flow processes are vertical, allowing a one-dimensional model description. The soil column is divided in compartments were the transport and water balance equation is resolved. The time step is the day. The irrigation can be prescribed at fixed time or scheduled according to a several criteria. These possibilities make the model very useful to analyse different agronomic scenarios for cropping systems carried out in environment characterized by limited water resources and high evaporative demand. Concerning the crop component, SWAP may simulate a one-year rotation with up to three crops. The crop growth can be simulated with a detailed module (WOFOST model) based on the absorption of radiation energy by the canopy as function of incoming photosynthetic active radiation and crop leaf area. The potential photosintetesis rate is reduced due to water and/or salinity stress. The dry matter production is partitioned among roots, leaves, stems and storage organ. This partition can be a function of the crop development. Because of vary processes involved in the crop growth, the WOFOST module require many parameters that can be directly measured or calibrated. On other possibility to describe the crop growth come from a simple module based on the leaf area index or soil cover fraction, rooting depth and crop height that are directly prescribed as data input by the users. This module is particularly useful when the crop growth data are insufficient or the crop is characterized by a particular shape. This is the case of the lettuce (Image 1) with many leaves gathered in a compact spherical structure that also at the end of the crop cycle doesn’t cover completely the soil. For this reason it is difficult to describe the time evolution of soil cover fraction utilizing relationships based on leaf area index as other crops like corn, wheat, sorghum, etc. Because the Wofost module was projected for these crops we preferred to use the simple module for the Lettuce crop. The Objective of Pilot Assessment was individuate integrated approaches for optimizing water resources use by maximising the cropping systems WUE, approaches that could be utilized by political stakeholders in land planning activity. Relations with other national projects: − −

PRAECO: Agronomic improvements of cropping systems at farm scale. CLIMESCO: Evolution of cropping systems as affected by climate change

216

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al. −

AQUATER: Decision support systems to manage water resources at irrigation district level in Southern Italy using remote sensing information

Image 1. The Lettuce field.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

GENERATION OF CLIMATIC DATA The climatic scenario were obtained from CGM output and downscaled to local condition using the weather generators LARS-WG. Such generated data have been also compared with our measured ones collected since 1981 in an experimental farm of CRA-ISA located in Metaponto (MT) near to the Jonic sea about global radiation, minimum and maximum temperatures, precipitation, air humidity and wind speed in order to predict the potential evapotranspiration. In order to simplify the generation of climatic data, we used the simple evapotranspiration equation of for which the only parameters of Temperatue (minimum and maximum) and global radiation are necessary for estimating the refercence daily evapotranspiration. For a limited period a previous comparison with Penman-Montheith method was carried out and we adopted a linear regression in order to minimize the typical underestimation of the equation of Priestley-Taylor. In order to optimize the irrigation management by highlighting the relationships among the water fluxes in the continuum groundwater-soil-plant-atmosphere, the agronomical scenario analysis have been carried out utilizing SWAP model.

Adaptations of Irrigated Cropping Systems of Southern Italy…

217

The Pilot Assessment was carried out for two “real” cropping systems: sorghum and a winter vegetable cultivation.

SORGHUM CULTIVATION We used a complete input data-set concerning crop growth, phenology, morphology and soil hydraulic parameters that comes from our previous research with the principal goals to calibrate and validate the crop growth for SWAP/WOFOST model. The sorghum was cultivated for 6 years on a clay soil in an experimental farm of CRA-ISA located in Metaponto (MT) near to the Jonic sea. The calibration and validation gave acceptable results despite an underestimation of Leaf area Index (LAI) in the central part of crop cycle. However we considered this underestimation no very determinant because it occurred when the measured and estimated LAI values were, in any case, larger than 3-4. In fact when the LAI is more than 2-2.5 the soil evaporation is not relevant and the potential transpiration is equal to total potential evapotranspiration. For this crop the detailed module of crop growth was used. The following scenarios were considered: Scenario 0: reference scenario with the past conditions of climate and water availability Scenario 1: forecasted climate + no-limited irrigation water availability Scenario 2: forecasted climate + limited water availability Scenario 3: forecasted climate + limited availability of good water with the addition of corresponding amount of saline water Such scenarios were evaluated utilizing the following indicators of sustainability: crop yield, harvest Index, potential and actual evapotranspiration, deep percolation, runoff, salt accumulation etc.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

WINTER VEGETABLE CULTIVATION The reference crop cultivation was carried out in a private farm close to the experimental farm of CRA-USA. The soil is sand and the crop is cultivated from October to February with consecutive sowing time and it is irrigated with micro-sprinkler method. For this pilot assessment we used a input data-set concerning crop growth, phenology, morphology and soil hydraulic parameters that comes from our previous research in order to calibrate and validate SWAP for this crops. Moreover an intensive geostatistical study of physical soil properties was applied on two 10-ha fields integrated with a characterization of the hydraulic properties. For this crop we choosed to use the same model SWAP but the simple growth subroutine. This decision was taken because of the particular morphological characteristics of lettuce plant which is very different from the typical crop for which the detailed module of WOFOST was projected. In partical the key parameter to describe the crop growth and to estimate the

218

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al.

partition of evapotranspiraation in potential values of evaporation and transpiration, is the soil cover fraction as measured as orthographic projection of plant leaves taking in account the distance between the rows and between the plants. The pilot assessment for this cultivations included the following steps: − − − −

Sensitivity analysis of soil hydraulic parameters, root depth and water stress function parameters; Impact of hydraulic parameter variability on soil-crop water fluxes (transpiration, evaporation, drainage and productivity) Calibration and validation Evaluation of agronomical practices to alleviate the climatic change: sowing time, irrigation scheduling

The main results obtained in this pilot assessment are described in the following paragraphs: 1) Analysis of water flux and solute transport for a clay soil cultivated with grain Sorghum under different groundwater conditions in Southern Italy The aim of this work is to apply SWAP/WOFOST model for simulating water flow and solute transport into the soil and to evaluate different irrigation strategies in order to optimize the use of water resources characterized by different salinity levels and to preserve the soil to salt accumulation. For this purpose several irrigation strategies were defined including different scheduling and salinity levels of water irrigation. SWAP, with the detailed module of crop growth (WOFOST model), was parameterized and tested by using Sorghum crop (for seed production) and soil data set collected in agronomic researches carried out at the experimental farm of the CRA-Istituto Sperimentale Agronomico, situated in the coastal area of the Basilicata region in Southern Italy (Metaponto: lat. 40° 24’ N and long. 16° 48’ E). The soil is classified as a Typic Epiaquerts with average clay and silt contents of 60 and 36%, respectively. In this study SWAP was applied for a period of ten years with a unique initialization comparing the factors described in Table 1.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Agronomic factors involved in the pilot assessment Irrigation scheduling criteria Bottom Boundary conditions

Irrigation salinity

Depletion of readily available (R_#) soil water with depletable fractions of 0.5 and 0.7. Two groundwater conditions, typical in the coastal zone of Metaponto: (i) a constant watertable at depth of 6 m (Deep) and (ii) a shallow groundwater (Shallow) remaining nearly constant at a depth of 1.8-2 m during the summer months and rising to a depth of 0.2 0.4 m during the winter period (measured data) Two solute concentrations for the irrigation water: 1.26 and 2.56 mg cm-3 corresponding to 2 and 4 dS m-1

Adaptations of Irrigated Cropping Systems of Southern Italy…

219

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RESULTS The bottom boundary conditions and the different criteria of irrigation scheduling showed significant effects on computed annual irrigation depths due to different dynamics of soil water fluxes that took place in the soil root zone and at interface soil/atmosphere. The figure 1 reports the water budget integrated at crop cycle scale for the two irrigation scheduling criteria under deep and shallow groundwater condition. For R_05 and R_0.7 no significant differences were detected. However, compared to Shallow groundwater, under Deep condition more irrigation water was used. This result was mainly determined by (i) higher soil evaporation and (ii) lower soil water depletion. Under Shallow groundwater condition, our analysis showed the importance of this component of soil budget for root water uptake because of the dynamics of watertable during the winter months that caused higher soil water content at sowing time. The Figure 2 shows the impact of saline irrigations on Sorghum productivity. In general the irrigation salinity determined an higher variability with yield reductions no significant for the lower solute concentration. However under salinity conditions, determined by the highest saline treatment, the average yield reduction was 23% and 33% under Deep and Shallow groundwater condition, respectively. The temporal evolution of solute concentration adopting the highest solute concentration is shown in figure 3. The cyclic behaviour with high peaks due to crop irrigations and salinity conditions alleviated by the following winter precipitations is evident but this effect was markedly different for the two groundwater conditions with the higher concentrations under Shallow scenario from the end of 1995 to the 2002. For this scenario the salt irrigation input was smaller than Deep groundwater but the more conservative condition regarding the soil moisture caused more a more restrictive environment for the sorghum growth. Finally the Figure 4 reports the salt leaching below the root layer in autumn and winter periods with the highest values in 1996 and 1997 under Deep groundwater condition, values that explain the better productive performances under this scenario with saline irrigation water. The performance and utility of SWAP/WOFOST model was assessed by simulating water flow and solute transport in a fine-textured soil subject to different boundary conditions typical for Southern Italy. In this context we have applied the model in order to optimize several agronomic factors and to understand the dynamics of water flux and solute transport in an agro-ecological system funded on Sorghum cultivation. The indications obtained for a period of 10 years have allowed to quantify the contribution of the shallow groundwater to root water uptake. In this condition the higher soil water content at sowing time decreases the irrigation requirements but this bottom boundary condition can determine smaller salt leaching below the root layer with negative consequences when saline water are used as irrigation and/or the crop is particularly sensitive to salt stress. In such case it is necessary to adopt irrigation strategies to reduce the salt input into the soil by alternating, if possible, waters of different salinity level during the crop cycle or in the crop rotation. Under deep groundwater condition, the winter precipitation of Southern Italy can ensure a adequate leaching but because of the low hydraulic conductivity of clay soil, it is necessary to monitor the soil salinity when saline irrigation water are used.

220

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. The water budget.

Figure 2. The impact of salinity.

2) Impact of soil hydraulic properties on water fluxes for a typical winter cropping system of Southern Italy: a sensitivity analysis using SWAP model The modelling of soil water fluxes is a fundamental task for evaluating the agricultural cropping systems from environmental and economic points of view. The soil water fluxes as drainage and evaporation are difficult to measure. Indeed it is possible to measure the plant transpiration, but the measurements are expensive and time-consuming. For these reasons the

Adaptations of Irrigated Cropping Systems of Southern Italy…

221

model applications are become very useful for evaluating different irrigation management strategies with the objective to increase the water use efficiency (WUE). It is evident that this issue is particularly important for the regions where the scarce water resources are often the limitant factor for agricultural productivity.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. The temporal evolution of salt concentration in the root layer.

Figure 4. The winter leaching (mg cm-2) below the root layer.

222

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SWAP (Soil-Water-Atmosphere-Plant) is a physically-based model that simulates transport of water, solutes and heat in variably saturated soils. The program is designed to simulate the transport processes at field scale level and during whole growing seasons. We present the results of a sensitivity analysis carried out for a “real” lettuce cropping system typical in the Southern Italy during the winter season. The analysis was performed by using SWAP with the main objective to evaluate the sensitivity of the main soil/plant/atmosphere fluxes (transpiration, evaporation and drainage) to the hydraulic parameters included in the of Mualem-van Genuchten soil hydraulic functions: saturated hydraulic conductivity (Ks), saturated and residual soil water content ( s and r, respectively) and the fitting parameters ( and n). This research activity is part of an extensive study aimed at optimizating the irrigation management at farm scale (Image 1 and 2). Numerical runs were generated for three climatic data sets - wet, normal and dry crop seasons - based on data recorded by a meteorological station located in Metaponto which is situated in the coastal area of the Basilicata region in Southern Italy (lat. 40° 24’ N and long. 16° 48’ E). The simulations were generated for a 70-cm-depth profile considered homogeneous for hydraulic properties. All the crop parameters and the hydraulic properties were derived and/or measured from a farm field concerning a lettuce crop cultivated on a sandy soil from November 2004 to February 2005 in an area of 20 ha. For the sensitivity analysis we have generated a reference run for all three different climatic conditions using soil hydraulic parameters averaged for 20 sampling location. For each location, undisturbed soil cores (5 cm in height × 8 cm in diameter) were collected and the drying part of water retention function and the Ks were determined by means a hanging water column apparatus and the constant head method, respectively. For each location, the van Genuchten equation (van Genuchten, 1980) was fitted to the measured retention data by non linear regression analysis using the RETC model.

Image 2. Localization and map of Lettuce field.

Adaptations of Irrigated Cropping Systems of Southern Italy…

223

Table 2. Statistics of estimated soil retention parameters

θs cm3 cm-3 0.385 0.314 0.496 14.2

Mean Min. Max. VC (%)

Ks cm d-1 821 278 1859 55.7

α cm-1 0.030 0.011 0.085 50.0

n 2.278 1.401 3.103 19.8

For modeling the water fluxes in the soil-plant-atmosphere system, the agrohydrological SWAP model (van Dam et al., 1997) was used with the crop growth simulated with the simple module. Successively, the averaged soil parameters (Ks, s, and n) were independently perturbed by 5% of the nominal values to quantify the sensitivity of cumulated outputs (actual evaporation and transpiration and drainage water flux at the bottom boundary) to the changed hydraulic parameters. Sensitivity coefficients (SC) were calculated with the following equation:

SC J ( p j ) = 100

[J ( p

j

+ Δp j ) − J ( p j )

]

[1]

J(pj)

and represent the relative changes in percentage in cumulative actual flux J (evaporation, transpiration and drainage) corresponding to a change of 0.05) in parameter pj. In order to test a linearity of output variables additional changes between -25 and +25% were performed on the soil hydraulic parameters.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RESULTS The variability in soil hydraulic properties measured within the experimental field is shown in Table 3 that reports several simple statistics of the van Genuchten parameters. The Ks and parameters are characterized by the highest variability (variation coefficient, VC, greater than 50%) compared to s and n (VC lower than 20%). The parameter values are typical of sandy soils with large Ks, n greater than 1.2 and small s. Table 3. Results of sensitivity analysis in terms of coefficients calculated with Eq. 1 for actual fluxes of transpiration (Ta), evaporation (Ea) and drainage (Dr) Parameter θs Ks n

Ta 3.12 -0.62 -0.31 2.81

Normal year Ea Dr 1.47 -1.64 0.21 0.13 -1.47 -0.31 -1.26 -25.94

Ta 3.22 -0.36 -0.36 -5.74

Dry year Ea 1.05 0 -1.47 -5.25

Dr -3.41 0.68 -0.68 4.44

Ta 0 0 0 0

Wet year Ea Dr 0.30 -0.65 0 0 -0.91 -0.07 -3.35 0.72

224

Domenico Ventrella, Nicola Losavio, Rita Leogrande et al.

The Figure 5 reports the corresponding 20 soil retention functions that show a large variability with the maximum values of volumetric water contents ranging from 0.3 to 0.4 and an high water depletion when the soil matric potential decreases from -10 to -1000 cm.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. Spatial variability of soil retention functions.

Figure 6. Impact of perturbations in θ

s (th)

and n for actual transpiration, evaporation and drainage

Results of the sensitivity analysis are summarized in Table 3 in terms of sensitivity coefficients (SC) for the three climatic conditions. According to Hupet et al. (2004) a higher sensitivity was found in the normal and dry year for s and the shape parameter n. In these years, these effects were relevant for all three fluxes. The SCs of s were positive for transpiration and evaporation while the SCs of n were positive just for transpiration in the normal year. Moreover the drainage variations were inversely related to the parameter perturbations. The sensitivity of the hydraulic parameters was much less relevant for the wet year above all for transpiration and drainage fluxes while a little influence was detected for

Adaptations of Irrigated Cropping Systems of Southern Italy…

225

the actual soil evaporation. In the Figure 6, we present the results obtained when the s and n parameters are perturbed from -25% to +25%. It’s evident that the model response can not be constant but non-linear (e.g. for n concerning the transpiration and the evaporation) because of non-linearity of Richards equation. In conclusion, the impact of hydraulic properties on transpiration, evaporation and drainage was investigated. In general the water fluxes were found to be most sensitive to s and n changes while the effect of Ks and was rather weak. The impact of the hydraulic parameter variations was generally higher in dry climatic conditions with the model responses constant for some parameters and non-linear for others.

REFERENCES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Hupet, F., van Dam, J.C., Vanclooster M. 2004. Impact of within-field variability in soil hydraulic properties on traspiratioon fluxes and crop yields: a numerical study. Vadose Zone Journal. 3:1367-1379. Van Dam, J.C., Huygen, J., Wesselling, J.G., Feddes, R.A., Kabat, P., van Walsum, P.E.V., Groenendijk, P., van Diepen, C.A. 1997. Theory of SWAP version 2.0. Simulation of water flow, solute transport and plant growth in the Soil-Water-Atmosphere-Plant environment. Report 71, Alterra, Wageningen, The Netherlands, 167 pp.van Genuchten, M.Th. 1980. Soil Sci. Soc. Am. J. 44:892-898. van Genchten , M. Th. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44: 892-898.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 17

CLIMATE CHANGE AND AGRICULTURAL RISK IN HUNGARY Márta Ladányi∗ Corvinus University of Budapest, Villányi út 29. H-1118, Hungary

ABSTRACT

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In this paper first we introduce some methods as the most essential tools of risk assessment and prove that the risk of crop as well as vine production increased in Hungary in the past few decades and discuss the possible reasons. We have set a list of the main weather indicators that are mostly responsible for the production fluctuation. The indicators, of course, have been drawn up specifically for the considered plant types (e.g. corn, wheat, vine etc.). We introduce a data management software which can be used as a tool to reveal how the trends as well as appearance of extreme values of the examined indicators have been changing with time. Learning the trends through some historical case studies, there are strong grounds for suspecting that the anomalies of the indicators have been becoming more and more frequent. Moreover, based on the most widely accepted GCM scenarios we point out that such kind of extremes are expected to become further more frequent in the next decades. We analyse how and in what extend this change is going to proceed.

INTRODUCTION It is never more urgent to understand climate change impacts on risk in agriculture as the observed trends have undoubtedly great influence on production and quality. Though climate change in Hungary tends to be more and more similar to Mediterranean, anomalies are going to be increasingly frequent. Therefore risk is especially meaningful as it has considerably increased in the last few decades. ∗

[email protected]

228

Márta Ladányi

Unambiguous symptoms of climate change in Hungary in the past few years highlighted the vulnerability of production. Risk caused by climate change should be managed with coordinated adaptive strategies. Researches on impacts and adaptation possibilities have to support the decision makers in policy as well as in agricultural industries with information and plans.

DATA AND DATA MANAGEMENT METHODS We used historical weather data with minimum, maximum and average temperature, precipitation and sunshine hours daily data as well as production data for several Hungarian regions for corn, wheat and vine. For the scenarios approach we applied GCMs of GFDL (Geophysical Fluid Dynamics Laboratory, USA) which were downscaled to Hungary and refer to about 2030 and 2060 as well as CRU control data base with reference time series 1961-90, PRUDENCE monthly data with reference time series 2071-2100 (based on Hadley Center A2, B2 scenarios) and finally, PRUDENCE control data (reference time series 196190). The resolution of the downscaled CRU and PRUDENCE data is 5 km. (Christensen, 2005, New et al., 1999, Bartholy et al., 2007a,b,c). During our work we usually faced data-problems: the available historical data were often too short, not detailed enough or not relevant. To avoid the problem of data absence or shortage we used some very effective data-management methods which are known in the literature such as: − − − − −

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

−

Phillips-method for making the data temporary and spatially relevant (Phillips, 1971), Delphi-method for the synthesis of experts’ opinion (Linstone and Turoff, 2002), Smoothing methods for the distribution determination, Triangle-method for the probability assessment (Hardaker et al., 2004) Rank correlation or Fackler stochastic dependency methods to determinate the dependency of the risk variables (Fackler, 1991). Weather generator C2W (Bürger, 1997)

There were available historical and control weather data for the baseline period 1961-90. Moreover, there were available GCM downscaled scenarios with reference period 2071-2100. For the period in between, however, we needed a weather generator the weather data of which can somehow connect the two periods. We do not say that the climate changes proceeds in this way, this was just an approximated approach for the time in the near future. We needed it because stakeholders can be persuaded only if we draw a possible scheme of near future as well. To this, we used C2W which is aimed at disaggregating climatological means and anomalies into realistic weather processes. The weather parameters were calculated by a multiple linear fit connecting the baseline and the far future parameters. The parameters were then normalized by the probit normalization method. Then a first order autoregressive model was fitted to the normalized parameters. By way of Monte-Carlo simulations it was assured that the means of the simulated data converge statistically to the given parameters. The

Climate Change and Agricultural Risk in Hungary

229

simulated data, however, are not suitable for extreme weather event approach but they are suitable for trend analysis.

Data Management System AGRO-MET (Szenteleki et al., 2007b) In order to collect, organize, manage and search databases for CC research in a handsome and friendly way, we used a special data management system which was developed by Szenteleki et al. at Corvinus University of Budapest. The system has the capacity to filter and aggregate data from different perspectives. Applying the so-called Climate Profile Indicator Module we can create weather indicators by defining lower and upper boundary conditions regarding to daily as well as monthly data. It is also possible to combine temperature (minimum, average, maximum) as well as precipitation and radiation data of any time period. In the case of daily data, the system of conditions can be set up by day, while for making parameters for longer time periods, linear interpolation can be applied. One can define conditions for plant production demands. The software can survey and evaluate historical as well as GCM data to monitor temperature and precipitation characteristics and helps to decide whether the examined variables (of the examined time series) indicate sufficient or not sufficient conditions, according to the profile indicators.

Risk Analysis Concepts and Methods Certainty equivalent, utility function and risk aversion. The certainty equivalent (CE), i.e., the minimum selling price for a payoff distribution, depends on the decision maker’s (DM) personal attitude toward risk. A utility function, U, can be used to represent a DM's attitude toward risk. Assuming that we express utility as a function of total return t we require that the utility function should be invariant to positive linear transformation. Absolute risk aversion ra can be defined as

ra : t a −

U ( 2) (t ) . Quite typical and the most widely applied utility function is the soU (1) (t )

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

called negative exponential utility function: U : t a U (t ) = 1 − exp(−ct ) with constant ra . U can be approximated by the Taylor expansion at the mean of the distribution of t

U (t ) ≈ U ( E ) + U (1) ( E )(t − E ) + U ( 2 ) ( E )(t − E ) 2 / 2 . In the case when the distribution of t is normal, the so-called Freund-approximation for CE can be used as: CE = E − 0.5ra (t )V (Freund, 1956). Efficiency Criterion (E-V-efficiency) E-V efficiency criterion is used when we intend to find the set in which the expected total return is maximized for different levels of variance of total return (Anderson, 1977, Hardaker

230

Márta Ladányi

et al., 2004) and normality is supposed. However, the method is quite robust to violation of normality. Nevertheless, E-V efficiency criteria usually leaves some alternatives unranked.

Stochastic Ddominance (1st and 2nd degree) Stochastic dominance involves comparing points on more distributions. Namely, if x

F1 ( x) ≤ F2 ( x) (1st degree) or

x

∫ F (t )dt ≤ ∫ F (t )dt 1

−∞

2

(2nd degree) then the first

−∞

alternative is better than the second one. Its disadvantage is that we often cannot rank all the distribution functions (Drynan, 1986).

Trends in Hungarian Climate According to the downscaled Prudence A2/B2 scenarios we can expect the following changes in Hungary after 2070: −

− − −

Increasing mean temperature. The rate of increase is the greatest in summer (A2: 4.55.1 °C or B2: 3.7-4.2 °C). Consequently, we can expect warmer and longer growing seasons and season’s shift. Increasing winter-spring (A2: 0-37%, B2: 3-27%) while decreasing summer-autumn precipitation (A2: 3-33%, B2: 0-20%). More frequent and serious anomalies (drought, heat wave (Table 1), storm, wind, flood, hail). The number of winter/frost days decreases at more than 60% (Table 1), nevertheless, late frost risk increases because of early vegetation period start.

Longer Extreme Periods Increasing uncertainty due to delayed ecological and economical feedback. Extreme positive and negative production, nevertheless, in both cases with great economical risk.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 1. Expected changes of extreme temperature indices after 2071 in Hungary (Pongrácz and Bartholy, 2007) Extreme indices Number of days when the daily maximum temperature is above 25 °C Number of days when the daily maximum temperature is above 30 °C Number of days when the daily maximum temperature is above 35 °C Number of days when the daily maximum temperature is under 0 °C Number of days when the daily minimum temperature is under -10 °C Number of days when the daily minimum temperature is under 0 °C Number of days when the daily minimum temperature is above 20 °C

The rate of change +39% +91% +250% -75% -83% -65% +625%

Climate Change and Agricultural Risk in Hungary

231

RESULTS Production Risk Increase Approach The historical production data were corrected with the help of Phillips-method. We investigated the risk of corn and wheat production with three times twenty years (1951-70, 1961-80, 1971-90). Three efficiency criteria were considered, namely: the E-V-efficiency, the stochastic dominance based on subjective distribution functions (which were calculated based on historical data and experts’ opinion), as well as the criterion based on the utility function. The risk increase was in several cases evident, however, in some cases there was no ranking between the time series. The risk increase was finally proved by a recently simplified variant of the general stochastic dominance criterion (Hardaker et al., 2004). The utility function was calculated depending on the degree of absolute risk aversion ra

∫

as: U ( x, ra ) = U (t , ra ) ⋅ f (t )dt . The certainty equivalent CE was defined as: −1

CE ( x, ra ) = U ( x, ra ) . Under the assumption of a negative exponential utility function, U can be approximated by:

⎡ (exp(− ra x i ) − exp(− ra x i +1 )) ⎤ U ( x, ra ) = ∑ (Fi +1 − Fi )⎢1 − ⎥ ra ( x i +1 − x i ) i ⎣ ⎦ Certainty equivalent CE can be obtained as the inverse of U by the formula:

CE =

− ln[1 − U ( x, ra )] . If we represent the graph of certainty equivalent CE depending ra

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

on the absolute risk aversion ra , the highest curve indicates the less risky time series. With the above listed data management and risk methods we gained the following results: For corn and wheat yield we recognized that beside the yield loss caused by the Hungarian political situation at the end of the eighties, the deviation of the yield started to become greater yet at the beginning of the eighties. (There was a heavy yield loss in 1990 caused by several extreme weather events.) With the help of the above graphic representation we proved that the risk of corn and wheat production increased between 1951 and 1990 in all examined Hungarian counties, independently to the rate of risk aversion (Figure 1). In some regions the rate of increase became even quicker (Ladányi and Erdélyi, 2005). We proved that in every examined area the risk of vine production increased between 1964 and 2000, independently to the rate of risk aversion. In some regions the rate of increase became even quicker (Figure 2). Though some of the examined regions are quite different from each other with respect to their climate, terrain as well as vine production structure, the fact of risk increase can undoubtedly be proved for all of them (Ladányi et al., 2007b). In order to see the trends, the time interval 1964-2000 was split into four parts: 1964-76, 197082, 1976-88 and 1988-2000 and the risk of vine production was elicited for each interval: TS1: 1964-76 no symptoms of CC TS2: 1970-82 extreme high/low production occurs rarely;

232

Márta Ladányi TS3: 1976-88 yield loss occurs frequently; TS4: 1988-2000 serious extreme events, warming. Gy-M-S_5170

CE (kg/ha) 4900

Gy-M-S_6180

4700

Gy-M-S_7190

4500 4300 4100 3900 0

0.001

0.002

0.003

0.004

0.005

0.006

r_a Gy-M-S_5170

CE (kg/ha) 4000

Gy-M-S_6180 Gy-M-S_7190

3500 3000 2500 2000 0

0.001

0.002

0.003

0.004

0.005

0.006

r_a

Figure 1. Stochastic efficiency for corn (left) and wheat (right) production in Győr-Moson-Sopron county with respect to 1951-1970, 1961-1980, 1971-1990. If we represent the graph of certainty equivalent depending on the absolute risk aversion, the highest curve indicates the less risky time series.

Győr-Moson-Sopron CE (kg/ha) Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

6000 5500

CE_6476

5000

CE_7082

4500

CE_7688

4000

CE_8800

3500 3000 r_a

2500 0

0,001

0,002

0,003

0,004

0,005

0,006

Figure 2. Stochastic efficiency for vine production in Győr-Moson-Sopron county with respect to 196476, 1970-82, 1976-88, 1988-2000. If we represent the graph of certainty equivalent depending on the absolute risk aversion, the highest curve indicates the less risky time series.

Climate Change and Agricultural Risk in Hungary

233

Table 2. Plant specific weather indicators for corn and wheat (left, Karl et al., 1999, Peterson et al., 2001, Ángyán and Menyhért, 1987) and vine (right, Salonius, 2002, Huglin, 1986, Gladstones, 1992, 2000, Amerine and Winkler, 1944, Dry and Smart, 1988, Happ, 1999, Nicholas et al., 1994, Bruggen and Semenov, 1999, Salinary,et al., 2006, Allen CG, 2005) Number of frost free days days (FFD) Apr-Sept monthly average temperature (°C) Average growing season temperature (°C) Effective heat sum between Apr. and Sept. (°C) Aridity index (°C /mm) Apr-Sept monthly precipitation (mm) Growing season precipitation sum (Apr-Oct, mm) Winter precipitation sum (Oct-March, mm) Growing season length (days) Growing season starting/ending date Apr-Sept monthly sunshine hours Growing season sunshine hours sum (Apr-Oct)

Degree days (DD) (°C) Huglin index (°C) Winkler index (°C) Biologically effective DD (°C) Mean July temperature (°C) Harvest maximum temperature (°C) Spring frost index (°C) Summer rainfall (mm) Annual rainfall (mm) No. of growing season rain days Ripening month rainfall (mm)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

PLANT SPECIFIC WEATHER INDICATOR APPROACH Based on international literature and Hungarian experts' opinion as well as case studies we have drawn up a list of the most important, plant specific weather indicators for corn, wheat and vine (Table 2). We do not state that production (quality and quantity) is exactly determined by these factors and there can not be other important indicators. We just think that agricultural production depends on these factors very much and there can be detected a very strong, mainly very obvious connection between years with high/low amount and/or quality of production and the values of the weather indicators in the historical data. Therefore, if we learn the indicators more or less, we can conclude the production as well as the expected risk. (Ladányi et al., 2007a, Szenteleki et al., 2007b). With the help of AGRO-MET data management software we investigated the trends and the appearance of extreme values of the outlined indicators. In Figure 3, as an example, we can see how Huglin and Winkler indices are expected to change with time in Sopron area. It is no doubt that the anomalies of the indicators have been becoming more and more frequent. Moreover, based on GCM scenarios we have pointed out that such kind of extremes are expected to become further more frequent in the next decades.

CONCLUSIONS What GCM’s Predict on Impacts Changes in phenological timing (for crops the sowing date, for all plants the ripening dates shift to earlier and the length of the phenophases become shorter; in vine production we expect earlier and likely higher sugar ripeness as well as not sufficient flavouring).

234

Márta Ladányi

Water related challenges that influence timing and availability (the frequency of extreme weather events, such as droughts and floods are more probable). With increasing solar radiation maturation and sugar accumulation of vine are positively effected while the danger of sunburn increases as well. Changes in viability of some varieties (e.g. a region may potentially be shifted into another maturity type, Jones, 2006). WSopCRU_A2

hours 3000

HSopCRU_A2

hours 3500 3000

2500

2500

2000

2000

1500

1500

1000

1000

500

500

0

0

1980

2000

2020

2040

2060

2080

1980

2000

2020

2040

2060

year

WSopCRU_B2

hours

year

3000

hours 3500

2500

3000

HSopCRU_B2

2500

2000

2000

1500

1500

1000

1000

500

500

0

0

1980

2000

2020

2040

2060

1980

2080

2000

2020

2040

2060

WSopPRU_A2

hours 3500

2080 year

year

HSopPRU_A2

hours 4000

3000

3500

2500

3000 2500

2000

2000

1500

1500

1000

1000

500

500

0

0

1980

2000

2020

2040

2060

2080

1980

2000

2020

2040

2060

year

WSopPRU_B2

hours

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2080

year

3000

hours 3500

2500

3000

HSopPRU_B2

2500

2000

2000

1500

1500

1000

1000

500

500

0 1980

2080

0 2000

2020

2040

2060

2080 year

1980

2000

2020

2040

2060

2080 year

Figure 3. Winkler (left) and Huglin (right) indices calculated for Sopron for 1991- 2070. The weather data are interpolated by C2W weather generator between 1. CRU and Prudence Hadley Center (HC) A2, 2. CRU and HC B2, 3. Prudence Controll and HC A2, 4. Prudence Controll and HC B2.

Climate Change and Agricultural Risk in Hungary

235

Changes in the spread of pests and diseases. Warming up may result changes in the phenology of some pests. E.g. grape moth in Hungary may have one extra generation which can danger the maturation period. Increased fluctuation of vine production quality and increasing economic risk (IPCC, 2007). E.g. higher daily temperature can be advantageous for vine during ripening but can have negative effect between anthesis and maturation. High night temperature during veraison is disadvantageous for wine quality. Extreme weather events, especially when cumulated, are expected to have serious negative impacts in Hungarian viticulture. (Hajdu and Botos, 2005).

OUTLOOK With extreme event approach, in the future we intend to analyse in details how and in what extend climate change in Hungary may harm crop and vine production. As a consequence, we are going to show the connection between the occurrences of extreme events in weather and production fluctuation in quality and quantity. Using historical data and GCM scenarios we have to detect the most possible reasons of risk increase in the past as exactly as can be and forecast the potential main points of future risk. This approach makes us possible to respond to climate change with a National Climate Change Adaptation Programme which is yet under construction.

ACKNOWLEDGMENT Our work was supported by NKFP-B3-2006-0014, OTKA T042583 tender and the Jedlik Ányos NKFP6-00079/2005 program.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Allen Consulting Group: (2005) Climate Change, Risk and Vulnerability. Promoting an efficient adaptation response in Australia. Final Report, March 2005. Amerine M. A., Winkler A. J. (1944) Composition and Quality of Musts and Wines of California Grapes. Hilgardia, 15(6), 493-674. Anderson, J. R., Dillon, J. L. and Hardaker, J. B. (1977) Agricultural Decision Analysis. Iowa State University Press, Ames. Ángyán, J. and Menyhért, Z. (1987) Agro-ecological effects on crop production (in Hungarian) GATE-KSZE 1987. Bartholy, J., Pongrácz, R. (2007a): Regional analysis of extreme temperature and precipitation indices for the Carpathian Basin from 1946 to 2001.Global and Planetary Change 57: 83–95. Bartholy J., Pongrácz R., Gelybó Gy., Szintai B., Szabó P., Torma Cs., Hunyady A., Kardos P. (2007b): Expected regional climate change in the Carpathian Basin using different climate model outputs. Geophysical Research Abstracts, Vol. 9, CD-ROM. EGU General Assembly 2007. EGU2007-A-04602. Vienna, Austria, 15-20 April 2007.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

236

Márta Ladányi

Bartholy J., Pongrácz R., Torma Cs., Hunyady A. (2007c): Regional climate change expected in Eastern/Central Europe. 87th AMS Annual Meeting (AMS Forum: Climate Variations and Change Manifested by Changes in Weather). San Antonio, TX, 14-18 January 2007. Bruggen van A. H. C., Semenov A. M. (1999) A new approach to the search for indicators of root disease suppression. Australasian Plant Pathology. 28, 4-10. Bürger, G. (1997) On the disaggregation of climatological means and anomalies. Climate Research, Vol. 8. pp.183-194. Christensen, J. H. (2005): Prediction of Regional scenarios and Uncertainties for Defining Dry, P., and R. E. Smart. (1988). The grapegrowing regions of Australia. Viticulture. Volume 1. Resources, Coombe, B. G. and P. Dry, eds., Winetitles, Adelaide. Drynan, R. G. (1986) A Note on Optimal Rules for Stochastic Efficiency Analysis. Australian Journal of Agricultural Economics 30, pp. 53-62. Fackler, P. L. (1991) Modeling Interdependencies: an Approach to Simulation and Elicitation. American Journal of Agricultural Economics 73, pp. 1091-1098. Freund, R. J. (1956) The introduction of risk into a programming model. Econometria 24, 253-161. Gladstones, J (1992) Viticulture and Environment Winetitles Adelaide. Gladstones, J. (2000) Past and Future Climatic Indices for Viticulture. Paper presented at 5th International Symposium for Cool Climate Viticulture and Oenology, Melbourne, Australia, January 16-20. Hajdu, E. and Botos, E. (2005) Climate change impact on vine production (in Hungarian). AGRO 21, Vol. 45, pp. 198-204. Happ, E. (1999) Indices for exploring the relationship between temperature and grape and wine flavour. The Australian and New Zealand Wine Industry Journal 14(4): 68-75. Hardaker, J. B., Huirne, R. B. M., Anderson, J. R., Lien, G. (2004) Coping with Risk in Agriculture. 2nd edn. CABI Publishing, Wallingford-Cambridge. Huglin, P. (1986) Biologie et ecologie de la vigne. Payot Lausanne, Paris. IPCC Report on Climate Change 2007. Working Group II: Impacts, Adaptation and Vulnerability. Jones, G. (2006) Climate change and wine: Observation, Impacts and future implications. Wine Industry Journal, 21/4. 21-26. Karl, T.R., N. Nicholls, and A. Ghazi, 1999: CLIVAR/GCOS/WMO workshop on indices and indicators for climate extremes: Workshop summary. Climatic Change, 42, 3-7. Ladányi, M. and Erdélyi, É (2005) The increase of risk in maize production detected by a new stochastic efficiency method. Agrárinformatika 2005, Debrecen, 2005, pp. 1-6. Ladányi, M., Erdélyi, É. and Szenteleki, K. (2007a) Weather indicators and their tendencies caused by climate change. OIV Conference, Budapest, Hungary. Ladányi, M., Erdélyi, É.and Szenteleki, K. (2007b) The increase of the risk in Hungarian vine production due to climate change. Efita Conference, Glasgow, Scotland, 2007. Linstone, H. A. and Turoff, M. (eds) (2002) The Delphi Method: Techniques and Applications. New Jersey Institute of Technology, Newark, New Jersey. New, M., Hulme, M., Jones P. (1999): Representing twentieth-century space-time climate variability. Part I: Development of a 1961-90 mean monthly terrestrial climatology. – Journal of Climate 12: 829–856. Nicholas, P., P. Magarey, and M. Wachtel. (1994) Diseases and Pests. Grape Production Series, Winetitles, Adelaide, 106.

Climate Change and Agricultural Risk in Hungary

237

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Peterson, T.C., and Coauthors: Report on the Activities of the Working Group on Climate Change Detection and Related Rapporteurs 1998-2001. WMO, Rep. WCDMP-47, WMO-TD 1071, Geneve, Switzerland, 143pp. Phillips, J. B. (1971) Statistical Methods in Systems Analysis In: Dent, J. B. and Anderson, J. R. (eds) Systems Analysis in Agricultural Management. Wiley, Sydney, pp. 34-52. Pongrácz R., Bartholy J. (2007): Detected trends in extreme temperature and precipitation indices in the Central/Eastern European region. 87th AMS Annual Meeting (AMS Forum: Climate Variations and Change Manifested by Changes in Weather). San Antonio, TX, 14-18 January 2007. Salinari, F., Giosuè, S., Tubiello, F. N., Rettori, A., Rossi, V., Spanna, F., Rosenzweig, C., and Gullino, M. L. (2006) Downy mildew (Plasmopara viticola) epidemics on grapevine under climate change. Global Change Biology 12 (7), 1299–1307. Salonius, P. (2002) A New Climate Index for Grape Growing in Short Season Areas, Minnesota Grape Growers Association Annual Report, 2002. Szenteleki, K., Botos, E. P., Szabó, A., Horváth, Cs., Martinovich, L. and Katona, Z. (2007a) Definition of the ecological facilities, ecological indicators and quality of products int he Hungarian vine and wine sector using updated GIS. Efita Conference, Glasgow, Scotland, 2007. Szenteleki, K., Ladányi, M., Szabó, É., Horváth, L., Hufnagel, L. and Révész, A. (2007b) A climate research database management software. Efita Conference 2007, Glasgow.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 18

ASSESSMENT OF THE IMPACT OF CLIMATE CHANGE AND ADAPTATION ON POTATO PRODUCTION IN EGYPT Mahmoud Medany∗ Agricultural Research Center, Ministry of Agriculture and Land Reclamation, Egypt

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT This work was conducted in order to assess the impact of climate change on potato yield and to investigate the possible options for overcoming the negative impacts. A field experiment was carried out at El-Beheira Governorate in season 2005/2006; and included three levels of irrigation (80, 100 and 120 %) from the potential evapotranspiration. The DSSAT (Decision Support System for Agrotechnology Transfer) model was run with weather data, soil data and experimental data in order to predict potato tuber yield. Predicted and measured yields were compared and the results showed a non-significant deference between them. In conclusion, DSSAT was able to simulate potato crop parameters under current conditions with a difference from 0.01 to 0.08 % compared to the actual yield. The potential impact of climate change on potato production was evaluated by simulating different planting dates in Second cultivation (January 1st; January 15th and January 30th) and first cultivation (September 30th October15th and October 30th), irrigation levels (80, 100 and 120 %) on potato production with climate change scenarios by the years 2025s, 2050s, 2075s and 2100s compared with that predicted under the current conditions of 2005. Using the future climate data a yield loss of -3.98, -1.41, 0.16, 0.75 % was projected for second cultivation, while a yield increase of 17.0, 35.9, 44.6, 45.5 % was projected for the first cultivation at time series of 2025s, 2050s, 2075s and 2100s, respectively. The negative impact was decreased when planting date of second cultivation was changed from January 15th to January 1st (from -10.5, -7.2, -5.9, -4.9 to 5.6, 9.9, 13.2, 15.0 for the time series 2025s, 2050s, 2075s and 2100s, respectively). A tuber fresh yield under climate change scenarios (A1, A2, B1, and B2) was increased in different planting dates, irrigation levels in first cultivation. Water level ∗

[email protected]

240

Mahmoud Medany of 100% from potential evapotranspiration gave the highest tuber yield at difference planting dates with climate change Scenarios. Water level of 80% irrigation level treatment had the highest water use efficiency with different planting dates in both cultivations. The difficulties associated with this assessment are mainly related to the limited utilization of other crop models rather than DSSAT. In addition, reliable data for validation are rather limited and further work is required to cover major strategic crops in relation to limited irrigation water.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

INTRODUCTION Potato (Solanum tuberosum L.) is one of the major crops in Egypt. The national potato crop production does meet both the current local demand, and the export market. The rapid growth of the country population, the economic stress of reliance on food imports, and the limited area for agriculture require Egyptians to find new ways to increase agriculture productivity in general and food crops in specific. Potato is a major industrial crop in Egypt and is one of the main food crops grown mainly in delta, and middle Egypt. The production of potato is concentrated in Beheira, Menufia, Gharbia, Giza, Dakahlia and Minia. The selection of high yielding, the optimization of water levels, sowing dates to improve potato crop production under current and future climate change have to be evaluated. The choice of the best ones can conserve the agricultural Resources by sowing dates potato crop in the best time for sustainable agriculture in Egypt. In this connection, the best sowing time can improve productivity. Sinha et al. (1988) in India found that the best sowing date for screening cultivars against wilt would be around June 20 as maximum. Disease incidence is likely in the crop sown during this period. Garcia (1987) in Cuba grew sweet potato in January, and July. He found delaying from January to July reduced that yield. Early sowing Potato growers believe that early sowing produces a good yield before blight defoliation begins. There is evidence that foliar blight increases in later crops as compared to early ones. However, early sowing cannot be relied upon for blight control in years with early monsoons (Maung et al. 2001). Wolfe et al (1982) in field trials on deep yolo loam soils, found that seasonal evapotranspiration of potato ranged from 31.6 to 61.0 cm for Kennebee cv and from 33.1 to 63.0 for White Rose cv. Water stress reduced yield production due to reduced leaf area and leaf formation duration. Yamaguchi et al. (1964) found that yield, specific gravity and starch content of Russet Burbank cv and White Rose cv tubers were higher, and the sugar content lower when grown at soil temperatures between 15 and 24°C, than when grown at higher temperatures. The DSSAT family of models was used extensively to simulate potato growth and yield (Tsuji et al., 1994). Sepp and Tooming (1991) show that potato productivity is most sensitive to changes in spring water storage in soils, shifts in potato sowing time, and growth of precipitation. The key problem of mankind’s response to climate change is the adaptation of agriculture to the changed agroclimatic conditions and resources. Tooming (1988) recommend that plant cultivation be established according to the principle of maximum plant productivity. Generally speaking, plant cultivation in accordance with the principle of maximum plant productivity will utilize natural resources i.e. soil and climate not only with maximal

Assessment of the Impact of Climate Change and Adaptation…

241

productivity in the existing environment but also with maximal efficiency. Natural plants and plant communities are systems that have adapted to the existing climate and all environmental conditions during a long evolutionary process. Their structure and functions are harmonically related and well adapted to the climate and environmental conditions. Field crops have been developed by human activity over a long period of time. Accordance of plant demands with the given climatic and environmental conditions for agricultural crops is the most important precondition for high productivity if climatic changes as projected by atmospheric scientists (IPCC, 2001) adversely affected crop production, Egypt would have to increase its reliance on costly food imports. The rising trend of the global atmospheric carbon dioxide concentration (CO2) is well established. Estimates of future increases range from 45% to 115 % above the pre- industrial levels (near 280 ppm) by the year 2040 (Pearman, 1988). This increase of CO2 is expected to induce a change in climate, which its magnitude is still uncertain (Bolin, et al., 1986; IPCC, 2001). Assessments of the impact of CO2 induced change on agricultural productivity are needed for both scientific and policy making purposes. The complexity of climate – crop production interaction makes simulation a useful and probably, the only practical approach available for making the needed assessments (Claudio, et al., 1992). In Egypt MAGICC and SCENGEN Climate Scenarios Generator Models was used to create climate change data in vulnerability and adaptation assessments Eid et al. (2001). Water demand for irrigation is expected to increase in all countries of North Africa and it is important to define adaptation strategies that take into account the possible deficit of water for irrigation in the future (Eid et al., 1995; Iglesias et al., 2003). The present study aimed to assess the impact of climate change on potato yield and to investigate the possible option for overcoming the possible negative impacts.

WORK PROCEDURE

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1. Current Experiment Studies Field experiment was carried out at El-Bossily region, El- Beheira Governorate, Egypt, during growing seasons of 2005/2006 to study the effect of water levels on potato yield (Solanum tuberosum L.) Valour cultivar.. In addition, validation predicted yield by DSSAT (SUBSTOR-Potato model) was compared with actual data and impact of climate change scenarios on production and water needs of potato under different planting dates. The treatments comprised of three water levels (80, 100 and 120 % of the amount of water calculated according to class A pan equation). Date of plating was October 15, 2005. All other agriculture practices of cultivation were performed as recommended by Ministry of Agriculture. Chemical and physical properties of the soil of the experiment were analyzed before cultivation; the results are tabulated in Table (1). The permanent wilting point (PWP) and field capacity (FC) of the trial soil were determined according to Israelsen and Hansen (1962). Plot area was 150 m2 (15 m length x 10 m width). Plant distances were 30 cm apart. A distance of 2 m was left between each two irrigation treatments. The total amount of irrigation water was calculated by class A Pan equation. The normal agro-meteorological data of class A Pan evaporation for El-Bosaily region was obtained and expressed as mm/day. Drip irrigation was used the cultivation season.

242

Mahmoud Medany

Table 1. Chemical and physical properties of the soil of the experiment analyzed before cultivation Chemical properties Ec m/moh

pH

3.00

7.89

Ca++ meq/l 30

K+ meq/l 1.66

HCO3meq/l 2.5

Silt %

Mg++ Na+ meq/l meq/l 10 14.26 Physical properties Texture FC %

Sand %

Clay %

95.31

4.29

PWP %

Bulk density g/cm3

0.36

Sandy

5.65

1.435

16.77

Cl meq/l 12.6

The total amount of drip irrigation was applied by water flow-meter for each treatment (EC of water irrigation 0.8 dS/m). Irrigation treatments were named low, medium and high respectively. The EC of the irrigation ranged from 0.7 to 0.9 dS/m. Potato tubers were harvested during February, and total yield per feddan was estimated for each treatments. Water Use Efficiency (WUE) was calculated according to the following equation: WUE [kg/m

3

]=

Total yield [kg/fed] Total applied water [m 3 /fed]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2. Crop Model Validation for Current Climate Field data was used by SUBSTOR-Potato through DSSAT model to simulate and predict potato yield. The experiment data were prepared on the basis of IBSNAT data set (1988). The required climatic data for El- Beheira Governorate (Latitude 31.24, Longitude 30.24) were obtained from the Central Laboratory for Agricultural Climate, A.R.C. Egypt. The physical properties of the soil of the experiment of site ;tuber yield potato (kg/hectare) were recorded. Genetic coefficients allow a single potato crop growth model to predict differences in development, growth, and yield among cultivars when planted in the certain environment. The genetic coefficients can be divided into those relate to development and both vegetative and reproductive growth. Definitions of the coefficients. Genetic coefficients were calculated by run The SUBSTOR-Potato model with weather data and experimental data for valour varietiy to calculate the genetic coefficient by using sub model (GENCALC) in order to predict potato growth and yield. The comparison between actual data and predicted data was done through SUBSTORPotato model under DSSAT interface in three steps, i.e. retrieval data (converting data to SUBSTOR-Potato model), validation data (comparing between predicted and observed data) and run the model DSSAT provides validation of the crop models that allows users to compare simulated outcomes with observed results. Necessary files were prepared as required. Evaluation of applying SUBSTOR-Potato model: Calculating the difference percentage between predicted and observed data, Correlation coefficient and Paired T-test.

Assessment of the Impact of Climate Change and Adaptation…

243

3. Climate Data for Future (Time Series 2025s, 2050s, 2075s and 2100s) Climate change scenarios for site were assessed according to future conditions derived from MAGICC/SCENGEN software of the University of East Anglia (UK). In this the study four scenarios of climatec data were used i.e. A1, A2, B1 and B2. The principle of MAGICC/SCENGEN is to allow the user to explore the consequences of a medium range of future emissions scenarios. The user selects two such scenarios from a library of possibilities. The reason for two scenarios is, primarily, to be able to compare a no action scenario with an action or policy scenario. Thus, in MAGICC/SCENGEN, the two emissions scenarios are referred to as a 'reference' scenario and a 'policy' scenario (Wigley et al., 2000). The data which generated from MAGICC/SCENGEN are represented in 4 Scenarios (A1, A2, B1, and B2). These scenarios are described by IPCC 2001 as follows: The A1 scenario describes a future world of very rapid economic growth, global population that peaks in mid-century and declines thereafter, and the rapid introduction of new and more efficient technologies; The A2 scenario describes a very heterogeneous world. The underlying theme is self-reliance and preservation of local identities. There are increasing in population. Economic development is primarily regionally oriented and per capita economic growth and technological change more fragmented and slower than other scenarios; The B1 scenario describes an approximate world with the same global population that peaks in mid-century and declines thereafter, as in the A1 storyline, but with rapid change in economic structures toward a service and information economy, with reductions in material intensity and the introduction of clean and resourceefficient technologies. The emphasis is on global solutions to economic, social and environmental sustainability, including improved equity, but without additional climate initiatives; The B2 scenario describes a world in which the emphasis is on local solutions to economic, social and environmental sustainability. Increasing global population is at a rate lower than A2, intermediate levels of economic development, and less rapid and more diverse technological change than in the B1 and A1 storylines. While the scenario is also oriented towards environmental protection and social equity, it focuses on local and regional levels.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4. Crop Simulation The potato simulation model that used was the SUBSTOR-Potato model. The model simulates crop responses to changes in climate, management variables, soils and different levels of CO2 in the atmosphere. The software used to run the programs, which was developed by the DSSAT and includes database management, crop model and application programs (Tsuji, et al, 1994). Potential changes in potato physiological responses (yield) were estimated using the SUBSTOR-Potato model under different climate scenarios. The model simulates physiological crop responses (water balance, phonology and growth throughout the season) on a daily basis to the major climate factors (daily solar radiation, maximum and minimum temperature and precipitation), edaphic factors and management (cultivar, planting date, plant population, row spacing and sowing depth).

244

Mahmoud Medany

5. Options To Mitigation Negative Impacts in Production Studies on simuilation of different sowing dates and water levels carried out using the following method with DSSAT model. Simulation runs on different sowing dates and water levels through DSSAT model on potato at El- Beheira governorate were carried out. The simulation study was carried for the validation test as well as planting dates at 15 day intervals in Second cultivation (January 1st; 15 January and 30th January 2005) and first cultivation (September 30th, October15th, and October 30th 2005) in order to valdiate applying model under deferance planting date with recommendation tepmerature (the optimum potato planting date can be determined depended on the ongoing weather data minimum temperature for 8-12°C ongoing days). Studies on simuilation of different sowing dates (Second cultivation and first cultivation ) in order to impact of climate change on production planting dates and water.

RESULTS OF THE ASSESSMENT 1. Current Experiment Studies The effect of different irrigation levels on potato tuber yield illustrated in Table (2). Regarding the effect of different irrigation treatments, data show that using 100% irrigation level increased potato yield followed by 80 % treatment. The lowest yield was obtained by 120% irrigation level treatment. Relevant to the effect of different irrigation levels on water use efficiency, data in Table (2) also show that increasing irrigation quantity led to decrease water use efficiency for all irrigation treatments. The highest water use efficiency (WUE) obtained when 80% irrigation levels was used. The results of this study generally agreed with the observations that increase in water level above 100% irrigation level led to decrease WUE Norwood (2000), Shani and Dudley (2001) and Erdem et al, (2006).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. Effect of different water application on water use efficiency as well as actual and estimated potato yield Tuber fresh yield (kg/ha) Water treatments

80% 100% 120% Mean

3295 4119 4943 4119

WUE Kg/m3 water

6.61 6.02 3.8 5.32

Correlation coefficient=0.998. T-value=-1.76. P-value =0.135.

Actual

Estimated

Yield Change %

21352 24623 17340 21105

21770 24800 18760 21105

-0.02 -0.01 -0.08 -0.036

Assessment of the Impact of Climate Change and Adaptation…

245

2. Crop Model Validation for Current Climate The comparison between observed and predicted data for tuber fresh yield (kg/ha) in the three water levels at El-bossily region, El- Beheira Governorate is presented in Table (2). It was noticed that the output data from the SUBSTOR-Potato model (predicted data) were in harmony with the observed data for fresh tuber yield. Regarding the effect of different irrigation treatments, data showed that using 100% irrigation level increased potato tuber fresh yield significantly followed by 80 % treatment. The lowest yield was obtained by 120% irrigation level treatment. The same data were obtained from predicted model, Differences in tuber fresh yield (kg/ha) due to water levels in both results from observed and predicted data, 100% irrigation level gave the highest value for yield production as compared with other water levels in observed (24623 kg/ha) and predicted data (24800 kg/ha). The difference percentage between observed and predicted data was from -0.02 to -0.08 % , the average of difference percentage was -0.036%. Value of correlation coefficient was significant (p-value 0.000); this means the same trend was found in predicted and observed data. Paired T-test value was not significant (pvalue 0.135); this means no difference between observed and predicted data. Results of the validation experment indicat that the SUBSTOR- Potato crop model can be used successful in Egypt.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3. Effect of Sowing Dates and Water Levels on Simulated Potato Tuber Yield Simulation results of tuber fresh yield (kg/ha) as affected by different sowing dates and water levels are shown in Table (3). Results show that; delaying sowing date from September 30th to October 30th reduced gradually tuber fresh yield in first cultivation ; delaying sowing date from January 15th to January 30th reduced tuber fresh yield in Second cultivation . These results are in agreement with Ali (1993). The general trends detected from the overall avarages of simulated tuber yield indicate that potato crop have to be sown in September 30th in first cultivation and January 15th in Second cultivation to obtain the maximum tuber yield and saving irrigation water at the same time. This results is in agreement with that obtained by Ainer et al. (1993) and Ali (1993). Relevant to the effect of different irrigation levels on water use efficiency (WUE), data in Table (3) showed that increasing irrigation quantity led to decrease water use efficiency for all irrigation treatments. The highest WUE obtained by 80% irrigation levels. The highest WUE obtained by 80% irrigation level followed by 100% irrigation level. The lowest tuber yield was obtained by 120% irrigation level. The results of this study general agreed with the observations that increase water level above 100% irrigation level led to decrease WUE Norwood (2000), Shani and Dudley (2001) and Erdem et al, (2006). In general, 80% irrigation level was the best aimed at maximum WUE in this study. This recommendation is slightly different in irrigation from our recommendation aiming at optimum tuber yield obtained by 100% irrigation level. The adoption of 80 % irrigation level will be superior to 100% irrigation level.

246

Mahmoud Medany

Table 3. Effect of different sowing dates on potato production and WUE under current climate conditions Water treatments Planting dates water consumption m3/ha 80% 3564 September 30th, 100% 4455 2005 120% 5346 80% 3295 First October 15th, 100% 4119 cultivation 2005 120% 4943 80% 2868 October 30th, 100% 3585 2005 120% 4302 80% 3234 100% 4042 January 1st, 2005 120% 4850 80% 3684 Second January 15th, 100% 4605 cultivation 2005 120% 5526 80% 4278 January 30th, 100% 5348 2005 120% 6417 80% 3487 Water treatments 100% 4359 120% 5231 Mean 4359

Tuber fresh yield kg/ha) WUE Kg/m3 water 9.46 7.37 5.59 6.61 6.02 3.80 6.90 5.35 4.01 7.68 6.28 5.54 6.86 5.63 4.71 5.63 4.65 3.90 7.19 5.88 4.59 5.89

Current 33730 32810 29860 21770 24800 18760 19800 19180 17250 24830 25390 26890 25280 25940 26030 24080 24860 25020 24915 25497 23968 24793

Mean 32133

21777

18743 25703

25750

24653 24793

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4. Potato Planting Prediction Potato is cultivated during three seasons: fall, winter and spring/summer (El-Bedewy R. and A. Sharara, 1990). The main problem in potato production during Second cultivation (local multiplication potato seeds in Egypt) is the germination decline percentage. The minimum temperature after planting is the major factor for successes germination. The weather analysis results indicated that, the potato planting threshold is between 7.5-8 °C as minimum temperature under the Egyptian weather conditions. Riha et. al., (1996) found that the suitable minimum temperature for success germination is 7-10°C. The above mentioned analysis indicate that, the optimum planting dates in the study region (Bosaily, north Egypt) are 1st to 4th January 2005, 11th to 16th January 2005 and 27th January 2005 to 3rd February 2005, In this connection, the best sowing time can improve productivity. The observed and simulated fresh tuber yield value was closed with planting

Assessment of the Impact of Climate Change and Adaptation…

247

prediction results. Generally, it could be concluded that SUBSTOR-Potato model could be used predict tuber yield of potato under North Egypt conditions in future to make different strategy.

Threshold

Suitable periods for potato planting. Un-suitable periods for potato planting.

Figure 1. Minimum air temperature during the groing season in comparison with recommended and minmum temperature threshold planting dates.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. Effects of Climate Change and Water Levels on Potato Production 5.1. Potato Second Cultivation The potential impact of climatic changes on tuber fresh yield potato was evaluated by simulating different planting dates (Second cultivation ), irrigation requements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) medium refer to aerosol levels by the year 2025, 2050, 2075 and 2100 compared with that predicted under the current conditions 2005 (Table 4 and 5). The tuber fresh yield potato differed according to water levels, planting dates and climate change Scenarios. The difference percentage between current predicted (2005) and predicted under climate change senirios was from -16.1 to 26.5 % , the average of difference percentage was -3.98, -1.41, 0.16 and 0.75 % by the year 2025, 2050, 2075 and 2100 Respectively. The predicted temperature increases affected crop production negatively. The B1 scenario resulted in simulated tuber yield reductions from 0.1% to -4.2 % less than that predicted under the current conditions. By studying the decreasing effect of the climatic changes on the crop using some alternative ways include changing of planting date. The results show that early planting date (January 1st) tends to positively increase yield compared with the current conditions and under the changed climatic conditions. Potato adaptation would do little to counterbalance the negative temperature effects seen in our simulations. Current Egyptian potato production is limited to cultivars that need a period of cold weather for tuber initiation. The only viable strategy to reduce yield losses would be a change in planting dates, to allow for increased storage of carbohydrates and sufficient time for leaf area development prior to tuber initiation. These results are in agreement with Medany (2001) he found that the Change of planting dates under changed climatic conditions improved simulated yield on maize.

248

Mahmoud Medany Table 4. Interaction of the effect of different planting dates (Second cultivation ), irrigation requirements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) for the years 2025s, 2050s, 2075s and 2100s

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Tuber fresh yield (kg/ha) 2005

2025s A1

Current

Estimated

24830 26890 25390 25280 26030 25940 24080 25020 24860

27320 27450 27050 23160 23350 22380 20430 24710 23770 2025s A2 27270 27240 26980 23340 23140 23080 20410 24610 23740 2025s B1 27030 27030 26770 23290 23160 22730 20390 24670 23700 2025s B2 27130 27110 26840 23000 23210 22740 20320

24830 26890 25390 25280 26030 25940 24080 25020 24860 2005 24830 26890 25390 25280 26030 25940 24080 25020 24860 24830 26890 25390 25280 26030 25940 24080

2050sA1 difference % 10 2.1 6.5 -8.4 -10.3 -13.7 -15.2 -1.2 -4.4 9.8 1.3 6.3 -7.7 -11.1 -11 -15.2 -1.6 -4.5 8.9 0.5 5.4 -7.9 -11 -12.4 -15.3 -1.4 -4.7 9.3 0.8 5.7 -9 -10.8 -12.3 -15.6

Estimated 28590 28850 28190 23780 24620 23330 20480 24610 23870 2050s A2 28590 28850 28190 23780 24620 23330 20480 24610 23870 2050s B1 27920 28080 27620 23500 24250 22880 20600 24400 23790 2050s B2 27680 28060 28060 24020 24770 23760 20460

2075s A1 difference % 15.1 7.3 11 -5.9 -5.4 -10.1 -15 -1.6 -4 15.1 7.3 11 -5.9 -5.4 -10.1 -15 -1.6 -4 12.4 4.4 8.8 -7 -6.8 -11.8 -14.5 -2.5 -4.3 11.5 4.4 10.5 -5 -4.8 -8.4 -15

Estimated 29350 29520 29170 24140 24940 23710 20350 24900 23840 2075s A2 29970 29770 29570 24920 24260 24630 20240 24860 23050 2075s B1 28800 28550 28190 24730 23700 23440 20800 24500 23830 2075s B2 28710 28910 28310 23830 24970 23540 20740

2100sA1 difference % 18.2 9.8 14.9 -4.5 -4.2 -8.6 -15.5 -0.5 -4.1 20.7 10.7 16.5 -1.4 -6.8 -5.1 -15.9 -0.6 -7.3 16 6.2 11 -2.2 -9 -9.6 -13.6 -2.1 -4.1 6.8 16.4 11.5 -8.5 -1.2 -9.3 -13.9

Estimated 29130 29710 29350 24120 24880 23970 20210 24790 23210 2100s A2 31410 30950 30040 25420 24480 24860 19610 24270 22430 2100s B1 29040 28620 28690 25060 23980 23790 21350 24840 23370 2100s B2 28950 29400 29080 24180 25100 23760 20830

Assessment of the Impact of Climate Change and Adaptation…

249

Table 5. Single effect of different planting dates (Second cultivation ), irrigation requirements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) for the years 2025s, 2050s, 2075s and 2100s Tuber fresh yield (kg/ha) 2005

2025s

2050s

Current

Estimated

Difference %

Estimated

80% 100% 120%

24730 25980 25397

23591 25033 24458

-4.7 -3.7 -3.6

January 1st January 15th January 30th A1 A2 B1 B2

25703 25750 24653 25369 25369 25369 25369 25369

27102 23048 22931 24402 24423 24308 24308 24360

5.6 -10.5 -7.1 -3.8 -3.8 -4.2 -4.2 -3.98

2075s

2100s

Treatments

Irrigation Planting dates Climate change scenarios

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Mean

24157 25863 25060

Difference % -2.4 -0.5 -1.3

28223 23887 22970 25147 25147 24782 25031 25027

9.9 -7.2 -6.9 -0.9 -0.9 -2.4 -1.4 -1.41

EstiMated 24715 26137 25428

DifferRence % -1.2 1.5 0.1

EstiMated 24943 26313 25487

29068 24234 22977 25547 25697 25171 25291 25426

13.2 -5.9 -6.9 0.6 1.2 -0.8 -0.4 0.16

29531 24467 22744 25486 25941 25416 25480 25581

5.2. Potato First Cultivation The potential impact of climatic changes on tuber fresh yield potato was evaluated by simulating different planting dates (first cultivation ), irrigation requements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) medium refer to aerosol levels by the year 2025, 2050, 2075 and 2100 compared with that predicted under the current conditions 2005 (Table 6 and 7). The results indicated that the Climate change Scenarios (A1, A2, B1, and B2) increased tuber fresh yield in different planting dates (first cultivation ), irrigation levele on simulated potato production by the year 2025, 2050, 2075 and 2100 compared with that predicted under the current conditions 2005. The predicted temperature increases affected crop production positively in first cultivation potato. The A1 scenario resulted in simulated yield increased from 18.6 to 47.5 % by the year 2025 and 2100 Respectively. Results show that; delaying sowing date from September 30th to October 30th reduced gradually tuber fresh yield in first cultivation 2005 and difference under climate change. The sowing date in september 1st gave the highest tuber yield. Relevant to the effect of different irrigation levels on tuber yield the results showed that the water level 100% gave the highest tuber yield at diferance planting dates with climate change Scenarios (A1, A2, B1, and B2) by the year 2025, 2050, 2075 and 2100 compared with that predicted under the current conditions 2005. These results are in agreement with Karing et al. (1999) they found that the potato yields under climate change senirios (HADCM2 and ECHAM3TR) increased by about 6 to 8%. The yield increase was larger (from 10 to 16%) on coastal islands and in North Estonia.

250

Mahmoud Medany

Table 6. Interaction of the effect of different planting dates (first cultivation ), irrigation requements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) for the years 2025s, 2050s, 2075s and 2100s .

Tuber fresh yield (kg/ha) Water Req.

80% 100% 120% 80% 100% 120% 80% 100% 120%

Current 32810 33730 29860 21770 24800 18760 19180 19800 17250 2005 32810 33730 29860 21770 24800 18760 19180 19800 17250

80% 100% 120% 80% 100% 120% 80% 100% 120%

32810 33730 29860 21770 24800 18760 19180 19800 17250

80% 100% 120% 80% 100% 120% 80% 100% 120%

32810 33730 29860 21770 24800 18760 19180 19800 17250 25369

80% 100% 120% 80% 100% 120% 80% 100% 120%

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2005

2025s A1 Estimated 39170 40480 35780 29840 30230 24690 24050 24720 21820 2025s A2 40140 38830 35500 29510 30140 24380 23560 23720 21480 2025s B1 39530 38140 35200 28860 29920 23950 23930 23240 21090 2025s B2 38580 39910 35670 29260 30120 24170 23580 24270 21440 29692

2050s A1 Differrence % 23.4 16.1 19.8 37.1 21.9 31.6 21.5 28.9 26.5 22.3 15.1 18.9 35.6 21.5 30 22.8 19.8 24.5 20.5 13.1 17.9 32.6 20.6 27.7 24.8 17.4 22.3 21.6 14.4 19.5 34.4 21.5 28.8 26.5 19.1 24.3 17

Estimated 41480 44280 38470 31460 33640 28030 27150 27990 25200 2050s A2 41480 44280 38470 31460 33640 28030 27990 27150 25200 2050s B1 40420 42330 37780 29580 31930 26420 25240 26100 23240 2050s B2 40700 43580 37820 30870 33050 27340 26280 27120 24520 32492

2075s A1 Differrence % 35 23 28.8 44.5 35.6 49.4 37.1 45.9 46.1 35 23 28.8 54.5 26.9 49.4 45.9 37.1 46.1 29 19.8 26.5 46.7 19.3 40.8 36.1 27.5 34.7 32.8 20.7 26.7 51.8 24.5 45.7 41.4 32.7 42.1 35.9

Estimated 43640 46040 40790 34850 35930 30060 29260 30020 26530 2075s A2 43020 44820 41680 37250 35440 31210 31200 30560 27270 2075s B1 44220 41480 38760 33420 32780 27800 27840 26700 25180 2075s B2 42000 44760 39120 34130 33360 28450 28430 27250 25330 34461

2100s A1 Differrence % 40.3 29.4 36.6 65 40.5 60.2 47.8 56.5 53.8 31.1 32.9 39.6 71.1 42.9 66.4 62.7 54.3 58.1 34.8 23 29.8 53.5 32.2 48.2 45.2 34.8 46 24.5 36.4 31 56.8 34.5 51.7 48.2 37.6 46.8 44.6

Estimated 44630 45950 41430 30900 35980 31130 29140 30870 26770 2100s A2 48360 47300 44370 39500 36180 32450 32730 31830 28630 2100s B1 43950 42490 39880 33960 33920 29000 28760 27150 26060 2100s B2 4350 44850 40820 34950 34940 29740 29780 28240 26480 34374

Assessment of the Impact of Climate Change and Adaptation…

251

Table 7. Single effect of different planting dates (first cultivation ), irrigation requements level on simulated potato production with climate change Scenarios (A1, A2, B1, and B2) for the years 2025s, 2050s, 2075s and 2100s Tuber fresh yield (kg/ha) 2005

2025s

2050s

2075s

2100s

Treatments Current

Irrigation Planting dates Climate change scenario Mean

Estimated

Differrence %

Estimated

Differrence %

Estimated

Differrence %

Estimated

Differrence %

80% 100%

24587 26110

30834 31143

25.4 19.3

32843 34591

40.8 28

35772 35762

48.4 37.9

33418 36642

51.8 31.8

120%

21957

27098

23.4

30043

38.8

31848

47.3

33063

52.9

30-Sep

32133

38078

18.5

40924

27.4

42528

32.4

40698

27.3

15-Oct

21777

27923

28.2

30454

40.8

32890

51.9

33554

55

30-Oct

18743

23075

23.1

26098

39.4

27964

49.3

28870

54.2

A1

25369

30087

18.6

33078

30.4

35236

38.9

35200

47.5

A2

25369

29696

17.1

33078

30.4

35828

41.2

37928

42.8

B1

25369

29318

15.6

31449

24.0

33131

30.6

33908

46.9

B2

25369

29667

16.9

32364

27.6

33648

32.6

30461

44.6

25369

29692

17

32492

35.9

34461

44.6

34374

45.5

CONCLUSIONS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

For the overall results, it could be concluded that DSSAT can be used successfully to prediect potato tuber yield in Egypt. Using the future climate data a yield reduction from 1.41 to -3.98% was projected for the second cultivation that starts in Egypt in January 1st for the years till 2100. A yield increase was projected for the first cultivation that starts in Egypt in September 30th. The negative impact of yield reduction was decreased when planting date of second cultivation was changed from January 15th to January 1st. Water level 100% gave the highest tuber yield at difference planting dates with all climate change Scenarios. Water level 80% irrigation level gave the highest water use efficiency with the different planting dates in first and second cultivation followed by 100% irrigation level.

REFERENCES Ainer, N. G. ; R. A. El-Bedewy and A. N. Khater (1993) Comparison of sprinkler, furrow and drip irrigation systems for potato yield and quaility on vertissols. Zagazig J. Agric. Res. pp. 1663-1671. Ali, M. A. (1993). Physiological studies on water requirments of potato plant. PhD thesis. Fac. Agric. Moshtohor, Zagazig Univ. Bolin, B; B. R.. Doos; J. Jager and R. A., Warrick (eds), (1986) The Greenhouse Effect: Climatic Changes and Ecosystems. John Wiley and Sons, New York, pp.363-92.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

252

Mahmoud Medany

Claudio O; J. Stockle; R. Williams, Norman J.R. and C.Allan Jones(1992)A method for estimating and climatic effects of rising atmospheric carbon dioxide on growth and yield of crops:part 1-modification of the EPIC model for climate change analysis. Agricultural systems. pp 225-238. Eid, H.M., A .A. Rayan., K. A. Mohamed, M. M. A. El-Refaie, M. M. Attia, H. A. Awad, K.M. R. Yousef, and M. M. A. El-Koliey. (1995). Impact of climate change on yield and water requirements of some major crops. On- Farm Irrigation and Agroclimatology. pp.492 - 507. Eid, H.M.; Samia M. El- Marsafawy, and N. G. Ainer (2001) Using MAGICC and SCENGEN Climate Scenarios Generator Models in vulnerability and Adaptation Assessments. Meteorology and Environmental Issues. Conf. March, Egypt. El-Bedewy R. and A. Sharara, (1990) Potato production in Egypt. CIHEAM-CIP Potato Workshop of the Mediterranean Region 10–13 September 1990., Zaragoza, Spain. 7 p. Erdem Y.; S. Seshril, T. Erdem and D. Kenar. (2006) Determination of crop water stress index for irrigation scheduling of bean (Phaseolus vulgaris L.). Turk J Agric. 30: 195202. Garcia, E, G. (1987) Effect of duration of cultivation , planting date and depth of soil preparation on yields of sweet potato. Documentos- De- Cienias Agropecuarias de la Habana, San Jose De Las Lajas, Havaan, Cuba. IBSNAT (International Benchmark Sites Network for Agrotechnology Transfer, 1988). Experimental design and data collection procedures for IBSNAT. The minimum data set for systems analysis and crop simulation. Third Ed. Technical Rep.1 Honolulu, HI, USA; IBSNAT. 74 pp. Iglesias, A; M.N. Ward; M. Menendez, and C. Rosenzweig. (2003) Water Availability for Agriculture Under Climate Change: Understanding Adaptation Strategies in the Mediterranean. In: C. Giupponi, an M. Shechter (eds.). Climate Change and the Mediterranean: Socioeconomic Perspectives of Impacts, Vulnerability and Adaptation. Edward Elgar Publishers. IPCC (Intergovernmental Panel on Climate Change) (2001). The Third Assessment Report (TAR): Climate Change 2001 The Scientific Basis. Cambridge University Press for the Intergovernmental Panel on Climate Change. Israelsen, O. W. and V. E. Hansen .1962. Irrigation Principles and Practices, third edition, John Wiley and Sons, Inc New York.. Karing, Peeter, Ain Kallis, and Heino Tooming. 1999. "Adaptation principles of agriculture to climate change," Climate Research, Vol. 12, No. 2-3, pp. 175-183. Maung M. M., Thein Su and San Thein. (2001). Effects of sowing time, mound burning, practice and timing of fungicide applications in the management of potato late blight in Myanmar. Journal of Agricultural University of Hebei, China Vol. 24 (2): 25–31. Proceedings of the GILB East and Southest Asia Linkage Group workshop on late blight, 16–20 August, Baoding Hebei, China. Medany, M. A. (2001). The Impact of climate change on production of differerent cultivars of maize (Zea mays L.). Meteor. Res. Bull. Vol 16. 194-209. Norwood, C.A. (2000) Water use and yield of limited-irrigated and dryland corn. Soil Sci. Soc. Am. J. 64:365–370.

Assessment of the Impact of Climate Change and Adaptation…

253

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Pearman, G. I., 1988. Greenhouse gases: Evidence for atmospheric changes and anthropogenic cases. In planning for climate change, ed. G.I.pearman. CSIRO, Div. Atmospheric Research. E. J. Brill, New York. Riha S. J; Wilks, D. S, Simoens P. (1996) Impact of temperature and precipitation variability on crop model predictions. Clim. Change 32:293–311. Sepp J. V and Tooming H. G (1991) Productivity resources of potato. Gidrometeoizdat, Leningrad (in Russian). Shani, U. and Dudley, L. M. (2001). Field Studies of Crop Response to Water and Salt Stress. Soil Sci. Soc. Am. J. 65:1522–1528. Sinha, S. K., Rai, K., and Ahsan, M. M. (1988). Effect of date of sowing on the incidence of potato wilt at ranchi. Bacterial- Wilt- Newsletter, ACIAR. Tooming H (1988) Principle of maximum plant productivity. In: Kull K, Tiivel T (eds) Lectures in theoretical biology. Acad Sci Estonia, Tartu, p 129–137. Tsuji, G. Y.; J. W. Jone; G. Uehara and S., Balas, (1994). Decision Support Systems for AgroTechnology Transfer. Version 3 vol .2. IBSNAT. Wigley, T.M.L.; S.C.B. Raper; M. Hulme and S.J. Smith (2000). The Magicc/Scengen Climate Scenario Generator Version 2.4: Technical Manual. C.F: http://www. cru.uea.ac.uk. Wolfe, D. W., Freses, S., Voss, R. E. and Timm H. (1982). Growth and yield response of two potato cultivars levels of applied water. American potato journal (10) 490 (En) California Univ., Davis, CA 95616, USA. (Field Crop Abs. 1983: 36 (6) 4854). Yamaguchi, M., H. Timm, and A.R. Spurr. (1964) Effects of soil temperature on growth and nutrition of potato plants and tuberization, composition, and periderm structure of Tubers. Proc. Am. Soc. Hortic. Sci. 84:412-423.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 19

VINEYARD FULL IRRIGATION REQUIREMENTS UNDER CLIMATE CHANGE SCENARIOS FOR EBRO VALLEY, SPAIN Jordi Marsal1 and Angel Utset2 1

Institut de Recerca i Tecnologia Agroalimentaries –IRTA. Centre UdL-IRTA. Tecnologia del Reg. Av Rovira Roure, 191. 25198 Lleida. Spain 2 Technological Agrarian Institute of Castilla y León (ITACyL), Spain

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The possible effect of climate change during the forthcoming 20 years on full irrigation vineyard requirements was simulated by combining global circulation models and crop models. 2010-2030 climate change scenarios were obtained from the CGCM2 model outputs provided by the Canadian Center for Climate Modeling and Analysis according to the IPCC SRES A2 scenario for greenhouse gases emissions. A historical 42 year series of weather data from Lleida (north east corner of the Iberian Peninsula) was used in combination with the LARS-WG weather generator to generate 100 realizations of local weather data corresponding to 2010, 2015, 2020 and 2025. Climate change scenarios were produced by perturbing the weather generator according to the CGCM2 results corresponding to the study site. CropSyst was used to simulate vineyard water balance. Crop water requirements were simulated from an automatic calculation implemented in the model which derived the amount of applied water that would maintain soil at field capacity while minimizing drainage according to a daily routine calculation. Data from a vineyard irrigation experiment which considers fully irrigated and deficit irrigated vines were used to adjust the crop model parameters. For the adjustment it was considered vineyard water requirements from FAO-56 method, field measures on soil texture and apparent density, crop growth and plant water stress development, as well as vine rooting depth. Further adjustments were done to simulate soil evaporation from a drip irrigation system. For the model validation, simulations on berry yield were compared to experimental data for both irrigation treatments but across three different soil types differing in soil depth (from 0.5 to 0.9 m.). A reasonable yield prediction was found according to a Willmot concordance index of 0.95. The weather generated indicated an increase in average air temperature of 1.5ºC from 2005 to 2025

256

Jordi Marsal and Angel Utset period. Annual Evapotranspiration increased by a maximum 43 mm and rainfall decreased around 100 mm for the same period. Irrigation requirements had increased by 50 mm from 2005 to 2015. Crop development (bud-break and harvest) was advanced by two weeks due to increased temperatures. Crop coefficients simulated by CropSyst mimicked well those used by FAO-56 during 2005. However, the 2025 scenario suggested substantial increases in crop coefficients early spring, and also earlier decreases because of the advancement in the time of harvest. It is therefore concluded that the most important impact of climate change would be more on the irrigation scheduling adjustments to variations in crop development than on possible annual increases in crop water demand.

INTRODUCTION During the last decades there have been a large number of published studies describing how plants can respond to short tem air elevated CO2. This bountiful literature has served to parameterized plant responses, and it has recently been implemented in various crop models so that the effects of climate change can be simulated by computer. However there is little information on how irrigation districts will have to handle these changes and most managers and policy makers are still thinking if this will be needed at all. However, it looks rather straight forward that with the rising temperatures, evapotranspiration will have to increase, and that possible changes in rainfall during the crop growing season will have the last word on what has to come. AGRIDEMA project was born from the complexity of possible outcomes with global weather changes and with the idea to provide some sort of assessment by using the available mathematical tools that could form a partnership between climate and crop models. The use of the two types of model combination is used for the prediction of the impacts of climate change on irrigation management in the Mediterranean region. This pilot project studies the particular case of irrigated vineyards in the northern east corner of the Iberian Peninsula (Ebro basin).

MATERIALS AND METHODS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Future Climate Scenarios The 2010-2030 Climate Change scenarios were taken from the CGCM2 model outputs, provided by the Canadian Centre for Climate Modelling and Analysis (Flato et al., 2000; Flato and Boer, 2001). The IPCC SRES A2 scenario for greenhouse gases emissions (IPCC, 2001) was considered. The CGCM2 model was preferred to others available, because it provides free internet access to daily simulation data in a text format. Hence, this model is more suitable for simple agricultural applications anywhere. According to Merrit et al (2006), results considering CGCM2 are similar than those obtained through other general circulation models. A historical meteorological series of Lleida (INM, Spanish Insituto Nacional de Meteorología), Spain (41.38 N, 0.35 E), comprising daily data from 1956 to 1998 of maximum and minimum temperatures, sunshine hours and precipitation; was used in

Vineyard Full Irrigation Requirements under Climate Change Scenarios…

257

combination with the LARS-WG weather generator (Semenov and Barrow, 2002) to generate 100 realizations of local weather corresponding to 2010, 2015 and 2025. A weather generator produces synthetic daily time series of climatic variables statistically equivalent to the recorded historical series, as well as daily site-specific climate scenarios that could be based on regional GCM results (Semenov and Jamieson, 2001). Different weather generators are available, but according to Wilby and Wigley (2001), the US-made and the UK-made WGEN and LARS-WG are the most widely used. Besides, LARS-WG results are as accurate as those obtained with WGEN and other weather generators (Mavromatis and Jones, 1998; Semenov et al., 1998; Mavromatis and Hansen, 2001). The climate change scenarios were obtained perturbing the weather generator according to the CGCM2 results corresponding to the study site, i.e., Northeast of Iberian Peninsula. The relative change in wet and dry series lengths, as affected by global change, was done following the approach recommended by Semenov and Barrow (2002), based on the daily CGCM2 outputs for each ten-year range. The relative changes in temperature standard deviations, as well as relative changes in mean temperature, precipitation amount and solar radiation were obtained from the CGCM2 daily estimations, as suggested by Semenov and Barrow (2002). The weather data generated required further normalization in order to be used for irrigation purposes. Data from conventional weather stations such as those of the INM are not suitable for the calculation of the evapotranspiration model most widespread and accurate for farming irrigation scheduling (Penman-Monteith model, ET-PM). To calculate ET-PM, weather stations must accomplish with specific requirements on site location and grass cover conditions (agro-meteorological stations). Typically, evapotranspiration estimates from agrometeorological stations are substantially lower from those obtained by conventional stations and other ET models. Unfortunately, agro-meteorological stations began to be used in Spain after the decade of the eighties, and thus their data series is not long enough to be used for generating climate scenarios. The required normalization was done by linearly relating temperature data from a ten year series (1990-2000) between the INM weather station and the agro-meteorological station that was located in the same experimental fields where the experimental data was gathered for the simulations.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The Crop Model The simulation model CropSyst (Stockle and Nelson, 1998) in its recently released version (4.04.14) permits to simulate crop water balance in vineyards on a single season basis and with a daily calculation routine. Outcomes of crop water use starts from bud break and finalize at harvest. Estimates of crop water consumption before and after this period were yearly completed by the FAO-56 method (Allen et al., 1998). CropSyst simulates soil water balance, crop phenology, vine water uptake, canopy intercepted, leaf development, reproductive growth and final harvest. The model considers several management options for irrigation, including the automatic calculation of applied water so that soil is maintained at field capacity while minimizing drainage. The latter option was adopted for the genuine CropSyst determination of vineyard irrigation requirements under future climate scenarios. This evaluation was also compared to a real case in which the FAO-56 was used. Elevated atmospheric CO2 effects on growth and crop water use were simulated by using CO2 sub

258

Jordi Marsal and Angel Utset

model built in CropSyst and compared to a non CO2 increased scenario, since long term crop responses to elevated CO2 are not fully tested. The predicted CO2 atmospheric concentration values were as in other reports, i.e, 554 ppm at 2050 with a baseline of 334 ppm at 1960 (Richter and Semenov, 2005). CO2 concentration at any considered data were inferred from the latter interval, ie., 443, 430, 418, 406 ppm, for 2010, 2015, 2015 and 2025, respectively.

Experimental Data

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

For parameterization and validation of the model, data from an irrigation experimental field were gathered during 2003. In a 10 year old ‘Tempranillo’ (Vitis vinifera L.) vineyard from Raïmat winery (Lleida, Spain) two different irrigation treatments were established. The treatments consisted in: 1) Control fully irrigated treatment and 2) SSDI (seasonal sustained deficit irrigation) in which vines were irrigated at 50% of Control vines. Treatments were replicated 4 times in soils differing in soil depth. Replications were arranged according to an experimental design with 4 completely randomized blocks. Each block-replication consisted of two rows of 12 vines surrounded by two more rows equally irrigated that functioned as a vine guards. Full irrigation requirements were determined according to FAO-56 methodology. At the end of the experiment, soil water holding capacities were determined by opening trenches at every site and pressure release curves were determined for every horizon according to Richter technique (in a Richard membrane). It was then observed that the irrigation treatments affected root distribution with depth, SSDI vines having more superficial root distribution than Control vines. Vine intercepted radiation was determined at peculiar days by measuring with a AccuPAR Ceptometer (Decagon Devices Inc., Pullman, WA, USA). Vine water status was weekly evaluated by measuring midday stem water potential with a pressure chamber (Model 3005; Soil Moisture Equipment, Santa Barbara, CA, USA.) according to Shackel et al (1997) methodology. At harvest (second week of September) grape production was manually harvested per each vine individually. Percentage dry matter was calculated by drying 1 kg of grapes per replication in an oven at 70ºC. This was used to convert yield fresh grape mass to yield dry mass.

Model Parameterization Soil hydraulic properties required to run the CropSyst were estimated from field measurements of soil texture and apparent soil density. For this purpose, the equations provided by Saxton et al. (1986) which are implemented in CropSyst were used (Table 1). As regard to irrigation types, the model only considers flooding irrigation. Therefore the model assumes all applied water is spread throughout the entire planting space. Then, simulated soil evaporation results in substantially higher estimates than what would occur in using sprinkler irrigation systems. To correct this effect, a top soil layer of 0.005 m thickness was added so that soil evaporation was reduced to a 1/3 of crop evapotranspiration (Girona et al,. 2002). Cascade method for soil water transport was preferred to finite differences method since the

Vineyard Full Irrigation Requirements under Climate Change Scenarios…

259

cascade calculations in the specific case of this study provided estimates better correlated to seasonal stress development under deficit irrigation conditions. Table 1. Soil characteristics used in the simulations (20-35 cm) Soil parameter Replication-Blok

Control R-1 R-2

R-3

R-4

SSDI R-1

R-2

R-3

R-4

Effective soil depth (m) Apparent density (Mg/m2) Field capacity (m3/m3) Permanent wilt point (m3/m3)

1.5 1.39 0.283 0.113

0.7 1.41 0.286 0.106

0.55 1.39 0.286 0.112

1.0 1.40 0.280 0.115

0.9 1.43 0.286 0.106

0.6 1.37 0.280 0.133

0.5 1.38 0.283 0.135

0.85 1.37 0.284 0.135

Average soil texture was silty-loam.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. Crop parameters used in the simulations Parameter Measured or field estimated Specific leaf area Stem/leaf partition coefficient Extinction coefficient for solar radiation Begin flowering Begin initial fruit growth Begin rapid fruit growth Physiological maturity Fraction of total solids Thermal time to bud break Leaf water potential at the onset of stomatal closure Wilting leaf water potential

12.00 2.00 0.35 352 723 1627 2407 0.30 500 -700 -1600

Manual CropSyst Unstressed light above ground biomass Root length per unit root mass Surface root density Curvature of root density distribution ET crop coefficient at full canopy

0.003 90 4 1.2 0.8

Adjusted to experimental data Actual to potential transpiration ratio that limits leaf area growth Actual to potential transpiration ratio that limits root growth Above ground biomass transpiration coefficient Maximum fruit load Fraction of above biomass apportioned to fruit after flowering Fraction of biomass apportioned to fruit during accelerated fruit growth Translocation yield factor Leaf duration

Value

0.85 0.5 8 26250 0.9 1 0.1 5000

260

Jordi Marsal and Angel Utset

Irrigation treatments also required crop parameterization since field observations indicated that 50% deficit irrigation SSDI produced the majority of roots at the first soil top 0.25 m, whereas those of the Control were down to 0.4 m (we called this depth, observed rooting depth). This should be implemented by adjusting curvature root density distribution. Unfortunately this feature was not fully operative in the version 4.04.12. Irrigation treatment effects on rooting depth were then accounted by modifying maximum rooting depth parameter. Maximum rooting depth was calculated by adding half the distance between effective soil depth and observed rooting depth to the observed rooting depth. For the model parameterization Penman-Monteith reference evapotranspiration (ET-PM) was used (Allen et al., 1998). ET-PM was calculated from an automated weather station which had all sensors and requirements for the calculation of ET-PM. The station belongs to the Catalan network of agro-meteorological stations (XAC) and is located 400 m from the experimental site. Since climate scenarios did not provide forecasts on humidity nor wind speed, simulations for the climate scenarios were performed using Priestley-Taylor (ET-PT). The CropSyst values for the aridity factor and Priestley-Taylor constant were the default values. These represented adequately ET-PM values for the experimental site. Crop phenology was adjusted by supplying degree days for each crop developmental phase as it occurred during 2003 season. Crop parameters for grape vines were adjusted by comparing yield simulations outputs simultaneously for SSDI and Control treatments with experimental data corresponding to the deepest block –replication conditions (> 1 m deep) (Table 2).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Model Validation For the model validation, berry yield were compared to experimental data from different soil types and irrigation treatments. This was preferred to compare experimental data from different years under fully irrigated conditions because the former provided more of a clear signal at a field level. Seasonal trends on soil water content were also compared to midday stem water potential for congruence in water stress development. Although the latter parameters do not have to evolve exactly in the same way, some sort of agreement is expected. Simulations from the parameterized model were compared to 2003 yield experimental data which corresponded to the other 3 shallower sites that were not used for model parameterization (soil depths spanning from 0.9 to 0.5 m depth). Data from both treatments were considered, Control and SSDI. For the model capacity forecast, the Concordance Index as proposed by Willmot (1982) was used.

RESULTS AND DISCUSSION The validation procedure revealed an index of concordance of Willmot d=0.95, what indicates a reasonable yield prediction from changes in irrigation and soil depth (Figure 1). The outputs of CGCM2 model for the northeast Spanish quadrant are summarized in Figure 2 which predicts an increase in average annual temperature up to 4ºC by the end of the 21th century.

Vineyard Full Irrigation Requirements under Climate Change Scenarios…

261

15

Simulated yield (kg·ha-1)

y = 0.7489x + 2.3143 R2 = 0.9009 10

5

0 0

5

10

15

Observed yield (kg·ha-1)

Figure 1. Yield relationship between observed values, and simulated estimates from CropSyst. Different observations corresponded to different irrigation treatments across different soil depths. Willmot concordance index was 0.95.

Average Temperature (ºC)

19 18

y = 0.0445x - 75.718 R2 = 0.7784

17 16 15 14 13 12 2000

2020

2040

2060

2080

2100

2120

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Year

Figure 2. Average air temperature forecast as predicted by CGCM2 model from the Canadian Centre for Climate Modelling and Analysis corresponding to the Northeast quadrant of the Iberian Peninsula.

Climate scenarios for the period object of study downscaled for Lleida and normalized for the agro-meteorological station used in this study predicted a progressive increase in annual average temperature from 2005 until 2020 of a maximum of 1.5 ºC (Figure 3A). The level off in temperature for 2025 might not imply a significant change in the pattern as it was only related to the natural year-to-year variability within a longer term framework of steady increasing in air temperature (Figure 2 and Figure 3A). Reference Evapotranspiration (ETHargraves model calibrated for the ET-Penman Monteith) compared to 2005 increased up to an annual maximum of 43 mm in 2020, which represents only a 4% increase from 2005 (Figure 3B).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

262

Jordi Marsal and Angel Utset

Figure 3. Yearly patterns of weather parameters (average temperature standard weather station and average temperature for agro-meteorological weather station) (A) and rainfall and evapotranspiration (B).

Annual precipitation seemed to decrease from the 1995-2005 average value of 395 mm to as little as 290 mm at the end of the considered period (2025). Global radiation values seemed to remained steady around the 15.3 MJ/m-2 throughout the simulated period (Figure 3B). The predicted increase in temperature had a dramatic effect on the simulated time of bud break and harvest. Both, bud break and harvest, were advanced by about 2 weeks by 2025, whereas the period spanned between this two events were reduced by only 3 days (Figure 4). Crop water use increased slightly from 2005 initial values to any of the years contemplated in the future scenario, though increases accounted by no more than annual 30 mm (Figure 5A). The effect of no considering CO2 increases on plant physiological feedbacks only represented a tiny decrease of 8 mm on annual crop water use (Figure 5A).

Vineyard Full Irrigation Requirements under Climate Change Scenarios…

2025

Year

2020 Harvest

Bud-break 2015

2010

2005 50

62

74

86

98

110 122 134 146 158 170 182 194 206 218 230 242 254

Day of the year

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. Evolution in crop development events (bud-break and harvest) throughout the considered climate change scenario (2010-2025).

Figure 5. Evolution in crop water use and irrigation requirements for increased and unchanged CO2 throughout the considered climate change period (2010-2025).

263

264

Jordi Marsal and Angel Utset

Irrigation requirements, revealed a maximum increase in annual water demand of 50 mm (2005 vs. 2015). Irrigation requirements resulted more variable than crop water use because the latter includes both crop water use and effective rainfall, and rainfall was a lot more variable than any other weather parameter (Figure 5B). For instance, the lower requirements for 2020 accounted for more rainfall occurring during the crop developmental season than in any other of the years. Crop coefficients used to irrigate commercial orchards (Girona et al., 2006) were very well mimicked by CropSyst simulation during 2005 until the event of commercial summer pruning which is seldom considered for irrigation adjustment in commercial practices (Figure 6). But perhaps most important effect of rising temperatures, rather than increasing annual crop water requirements, is the way in which irrigation will have to be adapted through the season. The 2025 scenario, suggested substantial increases in crop coefficients early spring by 0.2 as a result of hastened crop development by increased temperatures (Figure 6). Later on the season (during June) predictions in crop coefficients were less different than what is actually being used in to irrigate vines (Figure 6). Another factor that will be probably modified is the moment of application of summer pruning. This will probably have to be applied earlier. In summary, the effect of global warming on vine grape irrigation requirements will not represent a heavy burden on the irrigation districts (8% increase), because the increase in crop water demand during certain periods of the year (i.e. early spring) will be compensated by decreases in potential consumption after summer pruning being applied and the earliest event of harvest. However, crop coefficients used for irrigation scheduling will require substantial adjustment since vine growth will occur sooner and in a faster fashion early spring, which is a period very sensitive to water stress for grape production.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Crop Coefficient - Kc

1

Pruning

2005-Simulated

0.8

2025 Simulated FAO-2005

0.6 0.4 Harvest

0.2 0 50

100

150

200

250

300

350

Day of the year Figure 6. Seasonal patterns of crop coefficients (Kc) as traditionally used in field experiments (FAO2005), and those simulated by CropSyst for 2005 and 2025. Arrows signals indicate when commercial summer pruning and harvest occurs.

Vineyard Full Irrigation Requirements under Climate Change Scenarios…

265

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Allen, R.W., Pereira, L.S., Raes, D., Smith, M. 1998. Crop evapotranspiration: Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper No 56. 300 pp. Flato, G.M. and G.J. Boer, 2001: Warming Asymmetry in Climate Change Simulations. Geophys. Res. Lett., 28, 195-198. Flato, G.M., Boer, G.J., Lee, W.G., McFarlane, N.A., Ramsden, D., Reader, M.C., and Weaver, A.J., 2000: The Canadian Centre for Climate Modelling and Analysis Global Coupled Model and its Climate. Climate Dynamics, 16, 451-467. Girona, J., Mata, M., del Campo, J., Arbones, A., Bartra, E., Marsal, J. 2006. The use of midday leaf water potential for scheduling deficit irrigation in vineyards. Irri. Sci.24:115127. Girona, J., Mata, M., Fereres, E., Goldhamer, D.A., Cohen, M. 2002. Evapotranspiration and soil water dynamics of peach trees under water deficits. Agric. Wat. Manag. 54:107-122. IPCC, 2001: Climate Change 2001: The Scientific Basis. J.T. Houghton et al. (eds.), Cambridge University Press, 881pp. Mavromatis, T., Hansen, J.W. 2001. Interannual variability characteristics and simulated crop response of four stochastic weather generators. Agric. For. Meteorol. 109:283-296. Mavromatis, T., Jones, P.D. 1998. Comparison of climate scenario construction methodologies for impact assessment studies. Agric. For. Meteor. 91:51-67. Merritt, W. S., Younes, A., Barton, M., Taylor, B., Cohen, S., Neilsen, D. 2006. Hydrologic response to scenarios of climate change in sub watersheds of the Okanagan basin, British Columbia. J. Hydrol. 326:79-108. Richter, G.M. and Semenov, M.A. 2005. Modelling impacts of climate change on wheat yields in England and Wales - assessing drought risks. Agricultural Systems. 84, 77-97 Saxton, K.E., Rawls, W.J., Romberger, J.S., Papendick, R.I. 1986. Estimating generalized soil water characteristics from texture. Soil Sci. Soc. Amer. J. 50(4):1031-1036. Semenov, M.A. and P.D. Jamieson. 2001. Using weather generators in crop modelling. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat, 322 pp. Semenov, M.A., Barrow, E.M. 2002. LARS-WG. A stochastic weather generator for use in climate impact studies. User Manual. Rothamstead Research, Hertfordshire, 27 pp. Semenov, M.A., Brooks, R.J., Barrow, E.M., Richardson, C.W. 1998. Comparison of the WGEN and LARS-WG stochastic weather generators in diverse climates. Climate Research. 10:95-107. Stökle, C.O., Donatelli, M., Nelson, R. 2003. CropSyst, a cropping systems simulation model. Europ. J. Agronomy. 18:289-307. Wilby, R.L. and T.M.L. Wigley. Down-scaling general circulation issues in climate prediction. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat. p 39-68. Willmott, C.J. 1982. Some comments on the evaluation of model performance. Bulletin American Meteorological Society. 63(11): 1309-1313.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 20

CLIMATE VARIABILITY AND CHANGE OVER THE BALKAN PENINSULA AND RELATED IMPACTS ON SUNFLOWER Stanislava Radeva∗ and Vesselin Alexandrov National Institute of Meteorology and Hydrology, BAS 66 Tzarigradsko Shose, BG – 1784, Sofia, Bulgaria

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT During the last few years climate variability and change as well as related impacts on crops in Europe, including also southeastern Europe, were studied by many researchers. However, the previous impact studies were more focused on popular crops such winter wheat, soybean, etc, but for sunflower there are no or very limited results. That is why, the major goal of the study pilot assessment was to assess climate variability and change over the Balkan Peninsula and climate variability and change impacts on spring crops such as sunflower. The final specific objectives included: (a) analysis of climate variability and change over the Balkan Peninsula during the 20th century; (b) creating climate change scenarios for the 21st century at the region of interest; (c) assessment of sunflower vulnerability and eventual adaptation measures to the projected climate change scenarios; (d) dissemination of the results by contacting decision- and policy-makers from various organizations such as Ministry of Agriculture, Academy of Agriculture, private companies working on decision making in agriculture, farmer unions, separate farmers as well as media. The second half of the 20th century is characterized by a decrease of annual air temperature till the end of the 1970s in the whole region of the Balkan Peninsula. Since the beginning of the 1980s a warming trend has been observed. The anomalies of annual precipitation show the following significant drought periods on the Balkan Peninsula: 1940s, the second half of the 1950s and especially during the last two decades The relation between the NAO index and the winter precipitation anomalies was estimated for selected sites in order to confirm or reject this finding.HadCM3 climate change scenarios were created for every used weather stations from selected areas in Bulgaria. The A2 and ∗

[email protected]

268

Stanislava Radeva and Vesselin Alexandrov B2 HadCM3 climate change scenarios were developed for the selected sites in southeast Bulgaria and northwest Turkey. Precipitation in the considered region in south-eastern Bulgaria and northwest Turkey is expected to decrease during the 21st century, especially throughout the warm half of the year. As a result of expected warming crop-growing duration of sunflower over the Balkan Peninsula is projected to decrease, especially at the end of the 21st century. The yield changes in the selected region show different trends depending on the latitude, altitude, soil properties as well as the time slices during the current century. The sowing dates of spring crops in Bulgaria could shift under the GCM climate change scenarios in order to reduce the yield loss caused by an increase in temperature. The selection of an earlier sowing date for sunflower will probably be the appropriate response to offset the negative effect of a potential increase in temperature. This change in planting date will allow for the crop to develop during a period of the year with lower temperatures, thereby decreasing developmental rates and increasing the growth duration, especially the grain filling period. The simulated results depicted show that the sowing date of sunflower, for example, in northeast Bulgaria should occur at least two weeks earlier in the 2080s under the HadCM3 scenario, relative to the current climate conditions. The dissemination of the results obtained are discussed at the end of this paper.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1. INTRODUCTION The Earth's climate has exhibited marked variations and changes, with time scales ranging from millions of years down to one or two years. Over periods of a several years, fluctuations in global surface temperatures of a few tenths of a degree are common. Some of these are related to the El Niño - Southern Oscillation (ENSO), North Atlantic Oscillation (NAO) and other natural phenomenon such as major volcanic eruptions. It is necessary to emphasize that the global climate change is caused by a both natural and human factors. The natural factors influence on the global atmospheric circulations, oceanic streamflows, criosphere, etc., and therefore on climate and it’s variability and change. The difference between the long time ago and near past is the risk of relatively fast climate change due to the human activities. Most of the scientists consider that the emissions of the carbon dioxide (СО2) and the other so called “greenhouse gases”, emitted in the atmosphere because of human activities such as industry and agriculture, may cause irreversible climate change. The СО2 concentration in the atmosphere has increased by 31% since preindustrial times (i.e. about 1750) as a result of burning fuel, afforestation, etc. Increases in greenhouse gas concentrations since 1750 have led to a positive radiative forcing of climate, tending to warm the surface and to produce changes of climate. Since 1976, the global average temperature has risen at a rate approximately three times faster than the 20th century-scale trend (near 0.6oC). It is quite possible that this warming suggests a discernible human influence on global climate. The last 10 years (1996-2005), with the exception of 1996, are the warmest years on record. The global mean surface temperature in 2005 is currently estimated to be +0.48oC above the 1961-1990 annual average. 1998 and 2005 are considered the warmest years in the temperature record since 1861. In recent decades, a growing number of extreme events, such as severe storms, intensive precipitation, floods and droughts, is also observed.

Climate Variability and Change over the Balkan Peninsula…

269

It is observed that the mean surface temperature over the Balkan Peninsula has being rising. In recent decades, a growing number of extreme events, such as severe floods and droughts is also observed in that region. There is no doubt that the question of regional climate variability and change is a major and important environmental issue facing southeastern Europe, including the Balkan Peninsula at the beginning of the 21st century.

2. GOAL AND OBJECTIVES During the last few years climate variability and change as well as related impacts on crops in Europe, including also southeastern Europe, were studied by many researchers. However, the previous impact studies were more focused on popular crops such winter wheat, soybean, etc, but for sunflower (which is also an important crop for the region of the Balkan Peninsula, at least in some areas) there are no or very limited results. That is why, the major goal of the proposed pilot assessment was to assess climate variability and change over the Balkan Peninsula and climate variability and change impacts on spring crops such as sunflower. The final specific objectives included: (a) analysis of climate variability and change over the Balkan Peninsula during the e 20th century; (b) creating climate change scenarios for the 21st century at the region of interest; (c) assessment of sunflower vulnerability and eventual adaptation measures to the projected climate change scenarios; (d) dissemination of the results by contacting decision- and policy-makers from various organizations such as Ministry of Agriculture, Academy of Agriculture, private companies working on decision making in agriculture, farmer unions, separate farmers as well as media

3. DATA AND MODELS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1. Data Gridded (0.5°N x 0.5°E) monthly values (1901-1995) of average air temperature and precipitation for the regions of the Balkan Peninsula and east Austria were used. These data were provided by the Climate Research Unit (CRU) in the United Kingdom. Additional gridded weather data such as NCEP/NCAR surface annual precipitation rate were collected for the Black Sea domain and applied within the study. The NAO is the dominant mode of winter climate variability in the North Atlantic region ranging from central North America to Europe and much into Northern Asia. The NAO is a large scale seesaw in atmospheric mass between the subtropical high and the polar low. The corresponding index varies from year to year, but also exhibits a tendency to remain in one phase for intervals lasting several years. The positive NAO index phase shows a stronger than usual subtropical high pressure center and a deeper than normal Icelandic low. The increased pressure difference results in more and stronger winter storms crossing the Atlantic Ocean on a more northerly track. This results in warm and wet winters in north Europe but in dry weather conditions in south Europe.The negative NAO index phase shows a weak subtropical high and a weak Icelandic low. The reduced pressure gradient results in fewer and weaker winter storms crossing on a more west-east pathway. They bring moist air into the

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

270

Stanislava Radeva and Vesselin Alexandrov

Mediterranean and cold air to northern Europe. The NAO index was considered in this study attempting to find out its potential relations to the weather conditions in a selected region of southeast Bulgaria and northeast Turkey. Six sites from southeast Bulgaria (Haskovo, Elhovo and Svilengrad) and northwestern Turkey (Edirne, Lüleburgaz and Ipsala) were also considered. These sites are located closely at the hydrological basin of Maritza river. This river has an important influence on the water supply in this part of the Balkan Peninsula. In respect to the selected sites monthly data of observed average air temperature and precipitation for the 20th centuries were completed from different sources. All observed Bulgarian weather data, applied in this study, were provided by the meteorological network of the National Institute of Meteorology and Hydrology. The weather data from the Turkish territory of the Balkan Peninsula were collected by the Istanbul Technical University. Thirty years of current climatic data are normally used in developing a baseline climate scenario. A 30-year period is considered adequate to include a good representation of wet, dry, warm, or cool periods. Selecting a recent 30-year period is preferred because it not only represents the current climate, but also, in most cases, has the most accurate data (e.g. ANL, 1994). The so called “current climate” was therefore based on the period 1961-1990, according to the recommendations by the World Meteorological Organization (WMO). Most climate change studies use estimates of regional climate change from global circulation models (GCMs) (e.g. IPCC, 1997; Tegart et al., 1990; Watson et al., 1996). The major advantage of using GCMs as the basis for creating climate change scenarios is that they are the main tool that estimates changes in climate due to increased greenhouse gases for a large number of climate variables in a physically consistent manner. The following set of GCM output was applied in the study – this from the HadCM3 runs. HadCM3 is a coupled atmosphere-ocean GCM developed at the Hadley Centre, United Kingdom and described by Gordon et al. (1999). It has a stable control climatology and does not use flux adjustment. The HadCM3 GCM dataset (for the time slices 2011-2020, 20412050 and 2071-2080) has a high spatial resolution - 10’ latitude/longitude, which approximates to about 18 km x 18 km at the equator, with the east-west dimension decreasing to ~16 and ~9 km at 30 and 60° N and S, respectively. The high spatial resolution of the climate surfaces was obtained by applying the regional climate model RegCM3 (see below) and interpolation techniques such as thin-plate smoothing splines (e.g. New et al., 2002) with elevation, latitude and longitude as predictors. By 2100 the world will have changed in ways that are difficult to imagine – as difficult as it would have been at the end of the 19th century to imagine the changes of the 100 years since. Each storyline assumes a distinctly different direction for future developments, such that the four storylines differ in increasingly irreversible ways. Together they describe divergent futures that encompass a significant portion of the underlying uncertainties in the main driving forces. They cover a wide range of key “future” characteristics such as demographic change, economic development, and technological change. For this reason, their plausibility or feasibility should not be considered solely on the basis of an extrapolation of current economic, technological, and social trends. Four qualitative storylines yield four sets of scenarios called “families”: A1, A2, B1, and B2 (IPCC SRES, 1999). Most of the recent climate change impact studies are focusing on the A2 storyline. The A2 storyline and scenario family describes a very heterogeneous world. The underlying theme is self-reliance and preservation of local identities. Fertility patterns across regions converge very slowly, which

Climate Variability and Change over the Balkan Peninsula…

271

results in continuously increasing global population. Economic development is primarily regionally oriented and per capita economic growth and technological changes are more fragmented and slower than in other storylines. This study partially assumed also the B2 storyline. The B2 storyline and scenario family describes a world in which the emphasis is on local solutions to economic, social, and environmental sustainability. It is a world with continuously increasing global population at a rate lower than A2, intermediate levels of economic development, and less rapid and more diverse technological change than in the B1 and A1 storylines. While the B2 scenario is also oriented toward environmental protection and social equity, it focuses on local and regional levels.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1. Methods The simulation model RoIMPEL was developed and linked with the GIS based soil/terrain information and GCM derived grid values of weather/climate variables to evaluate the water-, temperature-, nitrogen-limited crop yields, sowing and maturity days and the number of workable days. The outputs of the simulation model are spatially-distributed using the link with the soil/terrain GIS. RoIMPEL is a modular simulation model, using easy-tomap soil and weather data. RoIMPEL is, therefore, appropriate for GIS-based regional and sub-regional land use evaluation projects. Various practices for nitrogen and water management are considered by specifying easy-toderive parameters through external files. RoIMPEL also derives workability day statistics to be used for the optimisation of machinery and labour at the farm level. The minimum requirements for soils data are the soil texture and organic matter classes. The minimum weather data needed by the model are monthly values of the average daily air temperature and the monthly-cumulated rainfall. Hence, RoIMPEL is applicable to climate change analyses where the perturbations of the climate parameters are scaled from GCMs on a monthly basis. The soil is considered as a single reservoir partially filled with water. The zero level of the reservoir corresponds to the total soil water content at the wilting point for a soil layer corresponding to the maximum root front depth. The maximum volume of the reservoir is the maximum soil available water. Therefore, the actual water volume in the reservoir is the actual soil available water. The reservoir is filled with water from rainfall and discharged by crop transpiration. Should the water in the reservoir exceed the maximum reservoir volume, a second reservoir starts to fill. The water in this reservoir is the soil drainable water. The second reservoir is also filled by rainfall, and discharged by drainage flow and evaporation and a threshold corresponds to the wet water content limit for workability. If this threshold is passed the soil is not workable. The dynamics of the water budget elements (evaporation, transpiration, drainage) are computed using the Thornthwaite-Mathers approach if the soil water content is less than the maximum available water, and the travel time approach for drainage flow calculations for soil water contents greater than the maximum available water. The algorithm requires the sharing of total actual evapotranspiration between evaporation and crop transpiration. Ritchie’s formula is used for this partition. Thus, the dynamics of the leaf area index (LAI) is the central driving process for soil water dynamics during the vegetation period and for biomass calculations. The dynamics of LAI is computed using the maximum LAI and a built-in standard analytical function describing the relative LAI as related to the values of the

272

Stanislava Radeva and Vesselin Alexandrov

development stage. Maximum LAI is estimated using an iterative technique matching the values of total cumulated crop transpiration with the water supply during the vegetation period. Therefore, an overall water balance is achieved. The following soil nitrogen processes are simulated: mineralisation, immobilisation, nitrate leaching and nitrogen crop uptake. Active and stable pools of soil nitrogen are considered. The dynamics of the crop residue and its associated nitrogen pool is considered in detail; computing the decay rates for carbohydrate, lignin and cellulose-like materials. The concentration of nitrate in soil solution is 12 calculated for each day of simulation. Nitrogen from mineral fertiliser is applied automatically following criteria described in the management section. RoIMPEL dynamically calculates the variables with a time step of 1 day as it has functions to derive daily weather data (temperature, rainfall, solar radiation) from monthly values. A screening of soil/climate conditions to evaluate the land suitability for a given crop is first performed. For suitable land, the daily dynamics of the crop development stages up to harvest, and of water-, temperature-, and nitrogen-stresses are the main crop processes simulated in RoIMPEL for each crop. The accumulation of biomass is based on the radiation use efficiency and the net photosynthetically active radiation. The radiation use efficiency is CO2 concentration sensitive. The radiation-potential daily biomass increase is corrected by temperature, water and nitrogen stresses. Additional penalties on crop yields are included through alarm criteria (unfavourable weather parameters during the most sensitive development stages) based on crop specific physiology. The soil geographical data base of Europe at the scale 1:1,000,000 was also applied. The information within the database is referenced to soil typological units. In addition to the soil names, the STUs are described by variables (attributes) specifying the nature and properties of the soils, e.g. texture, water regime, stoniness, etc. Averaged crop and agrotechnological data from the selected countries in East Europe were also obtained from different sources.

4. RESULTS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

4.1. Climate Variability and Change during the 20th Century A minimum of the annual average air temperature in the Balkan Peninsula appeared in the first decade during the last century. After that average air temperature increased. In the earlier 1940s there was also a cold spell. The second half of the 20th century is characterized by a decrease of annual air temperature till the end of the 1970s in the whole region of the Balkan Peninsula (Figure 1a). Since the beginning of the 1980s a warming trend has been observed. 1994 was the warmest year during the investigated period 1901-1995. The anomalies of annual precipitation show the following significant drought periods on the Balkan Peninsula: 1940s, the second half of the 1950s and especially during the last two decades (Figure 1b). The anomalies of the NCEP/NCAR surface annual precipitation rate over the Balkan Peninsula region for the last decade (relative to 1961-2000), show that most of the Balkan countries were affected by the drought conditions in the 1990s. The summer drought in 2000 was one of the most severe ones during the 20th century (Figure 2).

Climate Variability and Change over the Balkan Peninsula…

273

a

b

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Long-term anomalies of average annual air temperature (a) and annual precipitation (b) over the Balkan Peninsula, relative to 1961-1990.

Figure 2. Accumulated precipitation (in mm) in July (a) and August (b) 2000 over the Balkan Peninsula.

The relation between the NAO index and the winter precipitation anomalies was estimated for selected sites in order to confirm or reject this finding (Figures 3 and 4).

274

Stanislava Radeva and Vesselin Alexandrov

Figure 3. NAO index, December-March, 1864-2002.

Precipitation anomalies (%)

90

a)

60 30 0 -30 -60 -90 1960

Precipitationanomalies (% )

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

90

1970

1980 Year

1990

2000

1970

1980 Year

1990

2000

b)

60 30 0 -30 -60 -90 1960

Figure 4. Anomalies of winter precipitation in Elhovo (a: southeast Bulgaria) and Edrine (b: northwest Turkey), relative to 1961-1990; red curve – 5-year running average; blue line – linear fit.

Climate Variability and Change over the Balkan Peninsula…

275

The obtained correlation coefficient varies at the most tested sites between minus 0.43 (Svilengrad, southeast Bulgaria) and minus 0.51 (Lulerbargaz, northwest Turkey) ÷ minus 0.54 (Elhovo, southeast Bulgaria). The lowest correlation coefficient was calculated for Haskovo (southeast Bulgaria) – minus 0.34. It was also found that precipitation in January, at the region of interest, is following in most years the NAO index anomalies for the same month.

4.2. Climate Change Scenarios HadCM3 climate change scenarios were created for every used weather stations from selected areas in Bulgaria. Figure 5 shows the monthly climate values of air temperature and precipitation in Novachene (north Bulgaria) under the HaDCM3 climate change scenarios for the years 2020, 2050 and 2080. 9.0

a)

2020 2050

8.0

2080

7.0

Δ T (o C )

6.0 5.0 4.0 3.0 2.0 1.0 0.0 1

2

3

4

5

6

7

8

9

10

11

12

8

9

10

11

12

Month

30

b)

20 10

Δ P (% )

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0 -10 -20 -30 -40 -50 -60 1

2

3

4

5

6

7

Month

Figure 5. Monthly climate values of air temperature (a) and precipitation (b) in Novachene (north Bulgaria) under the HaDCM3 climate change scenarios for 2020, 2050 and 2080.

276

Stanislava Radeva and Vesselin Alexandrov

It could be seen that the newer HadCM3 model simulates higher increases for monthly air temperature in comparison to the previous HadCM2 ones. Even air temperatures in July and August are projected to be in 2080 near 8oC higher than air temperatures, relative to the period 1961-1990 (Figure 5). Simulated HadCM3 precipitation has a similar direction for the 21st century as for the HadCM2 and ECHAM4 models – a decreasing one. Monthly precipitation in Novachene from May to September is projected to be about 50% reduced in 2080. Only precipitation in February and March as well as December is expected to increase during the 21st century. The A2 and B2 HadCM3 climate change scenarios were developed for the selected sites in southeast Bulgaria and northwest Turkey (Figures 6 and 7). The simulated increase of monthly air temperatures in Elhovo and Ipsala is less than 2oC for the year 2020. Even the HadCM3 climate model simulated a slight decrease of air temperature in northwestern Turkey in November for the above slice. Under the B2 emission scenario the UK model projected higher warming in 2020 than the simulated warming under the A2 emission scenario. However the A2 and B2 rates of warming for 2080 are different. Significant warming was projected for the summer months in 2080. 2020

8.0

2050

8.0

2080

a)

b)

6.0 ΔT (oC)

ΔT (oC)

6.0 4.0

4.0 2.0

2.0

0.0

0.0 1

2

3

4

5

6

7

8

1

9 10 11 12

2

3

4

5

8.0

8.0

c)

7

8

9 10 11 12

d)

6.0 ΔT (oC)

ΔT (oC)

6.0

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

6

Month

Month

4.0 2.0

4.0 2.0

0.0

0.0

1

2

3

4

5

6

7

Month

8

9 10 11 12

1

2

3

4

5

6

7

8

9 10 11 12

Month

Figure 6. Monthly changes of air temperature in Elhovo (a,c: southeast Bulgaria) and Ipsala (b,d: northwest Turkey) under the A2 (a,b) and B2 (c, d) HadCM3 climate change scenarios for 2020, 2050 and 2080, relative to the period 1961-1990.

Climate Variability and Change over the Balkan Peninsula…

277

2020 2050

a)

20

2080

0

0

-20

-20

ΔP (%)

ΔP (%)

20

-40 -60

b)

-40 -60

-80

-80

1

2

3

4

5

6

7

8

9 10 11 12

1

2

3

4

5

Month 20

c)

0

0

-20

-20

ΔP (%)

ΔP (%)

20

6

7

8

9 10 11 12

8

9 10 11 12

Month

-40 -60

d)

-40 -60

-80

-80 1

2

3

4

5

6

7

8

9 10 11 12

Month

1

2

3

4

5

6

7

Month

Figure 7. Monthly changes of precipitation in Elhovo (a,c: southeast Bulgaria) and Ipsala (b,d: northwest Turkey) under the A2 (a,b) and B2 (c, d) HadCM3 climate change scenarios for 2020, 2050 and 2080, relative to the period 1961-1990.

Air temperatures in July and August were simulated to increase by 7-8oC at the end of the 21 century under the A2 emission scenario. Precipitation in the considered region in southeastern Bulgaria and northwest Turkey is expected to decrease during the 21st century, especially throughout the warm half of the year. In fact the HadCM3 climate model simulated a slight increase in southeast Bulgaria during the current century (2020, 2050, 2080) only in March. The most significant monthly reduction in precipitation (higher than 70%) in southeast Bulgaria was projected for July, 2080, while for northwest Turkey May is expected to be most affected by the precipitation lowering. The simulated precipitation reductions in 2080 are higher under the A2 emission scenario than the respective precipitation decreases under the B2 emission scenario.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

st

4.3. Climate Change Impacts on Sunflower As a result of expected warming crop-growing duration of sunflower over the Balkan Peninsula is projected to decrease, especially at the end of the 21st century (Figure 8). The yield changes in the selected region show different trends depending on the latitude, altitude, soil properties as well as the time slices during the current century (Figures 9-11).

278

Stanislava Radeva and Vesselin Alexandrov

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. HadCM3 B2 changes (in days) in sunflower growing duration in the Balkan Peninsula for 2071- 2080, relative to current climate; RoIMPEL model.

Figure 9. HadCM3 B2 changes (in %) in sunflower yield in the Balkan Peninsula for 2011-2020, relative to current climate; RоIMPEL model.

The sowing dates of spring crops in Bulgaria could shift under the GCM climate change scenarios in order to reduce the yield loss caused by an increase in temperature. The selection of an earlier sowing date for sunflower will probably be the appropriate response to offset the negative effect of a potential increase in temperature. This change in planting date will allow for the crop to develop during a period of the year with lower temperatures, thereby decreasing developmental rates and increasing the growth duration, especially the grain filling period.

Climate Variability and Change over the Balkan Peninsula…

279

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 10. HadCM3 B2 changes (in %) in sunflower yield in the Balkan Peninsula for 2041-2050, relative to current climate; RоIMPEL model.

Figure 11. HadCM3 A2 changes (in %) in sunflower yield in the Balkan Peninsula for 2071-2080, relative to current climate; RоIMPEL model.

The simulated results depicted show that the sowing date of sunflower, for example, in northeast Bulgaria should occur at least two weeks earlier in the 2080s under the HadCM3 scenario, relative to the current climate conditions.

280

Stanislava Radeva and Vesselin Alexandrov

It should be noted, however, that although changes in sowing date are a no-cost decision that can be taken at the farm-level, a large shift in sowing dates probably would interfere with the agrotechnological management of other crops, grown during the remainder of the year. Another option for adaptation is to use different cultivars. There is an opportunity for cultivation of more productive, later or earlier-maturing, disease- and pest-tolerant cultivars. Switching from sunflower cultivars a long to a short or very short growing season projected an additional decrease of final yield under a potential warming in Bulgaria. However, using hybrids with a medium growing season, would be beneficial for sunflower productivity. Technological innovations, including the development of new crop hybrids and cultivars that may be bred to better match the changing climate, are considered as a promising adaptation strategy. However, the cost of these innovations is still unclear.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

5. DISSEMINATION OF THE RESULTS The above results were partly or fully disseminated through: (a)workshop on COST 734 Impacts of Climate Change and Variability on European Agriculture, 16-17 November 2006, Brussels, Belgium; (b) campaign for urgent measures to prevent climate change caused by human activities in relation to the United Nations Conference on Climate Change: participation in the press-conference held on 2 November 2006 in the EU Information Center in Sofia; (c) ENSEMBLES Workshop organized by Research Theme Climatic change and impacts in Eastern and Central Europe, 13- 15 September, 2006 Poïana Braşov, Romania, Poster: Radeva, S., Valkov, N., Alexandrov, V., 2006. Climate change impact on sunflower in the Balkan Peninsula; (d) 6th EMS Annual Meeting and the 5th European Conference on Applied Climatology, 4-8 September, Slovenia Abstract: Radeva, S., Valkov, N., Alexandrov, V., 2006. Climate change impact on sunflower in the Balkan Peninsula; (e) International Conference "Living with Climate Variability and Change: Understanding the Uncertainties and Managing the Risks", July 2006, Finland. Poster: Alexandrov, V., S. Radeva and N. Valkov, 2006. Vulnerability and adaptation of crops under climate change in Central an Eastern European countries; (f) international workshop on climate change impacts in Central and Eastern Europe and vulnerability assessment, 12-14 June , 2006, Bucharest, Romania; (g) meeting of the Union of the Physicists, June 2006; (h) International conference on water observation and information system for decision support, 23-26 May 2006, Ohrid, Macedonia; (j) Seminar and discussion on climate change issues on the occasion of the World Meteorological Day, 23 March 2006, at the NIMH-BAS; (k) Workshops with colleagues from the Ministry of Agriculture and Ministry of Environment, April and November, 2006; (l) National courses on agrometeorology, 08-12 May 2006 27 November -1 December 2006, Sofia, NIMH-BAS; (m) Iterviews for radio "Shumen" and “Blagoevgrad”, May and November 2006

6. LIMITATIONS The following major limitation was encountered during the project implementation: several attempts were done to establish stable contacts with Bulgarian farmers, but only a few were really successful. The majority of the farmers across the country still do not pay

Climate Variability and Change over the Balkan Peninsula…

281

significant attention to the question on climate variability and change as well as related impacts on agriculture. It is considered that educational activities in the field should be applied in the near future in order to bring the scientific results to the common people from the villages.

ACKNOWLEDGMENTS Acknowledgements are expressed to Dr.Catalin Simota and Dr.Jeremy Pal – model developers for their help, useful suggestions and encouragement.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Argonne National Laboratory (ANL), 1994. Guidance for Vulnerability and Adaptation assessments. U.S. Country Studies Program, Washington D.C., USA. Gordon, C., Cooper, C., Senior, C.A., Banks, H., Gregory, J.M., Johns, T.C., Mitchell, J.F.B. and Wood, R.A., 1999. The simulation of SST, sea ice extents and ocean heat transports in a version of the Hadley Centre coupled model without flux adjustments. Climate Dynamics. Intergovernmental Panel on Climate Change (IPCC), 1998. The Regional Impacts of Climate Change: An Assessments of Vulnerability, IPCC. Intergovernmental Panel on Climate Change Data Distribution Centre (IPCC DDC) 1999. Data/Information supplied by the IPCC Data Distribution Centre for climate change and related scenarios for impact assessments, Version 1.0, April 1999, Norwich, UK. Intergovernmental Panel on Climate Change Special Report on Emission Scenarios (IPCC SRES), 1999. Emission Scenarios, IPCC. Intergovernmental Panel on Climate Change – Task Group on Scenarios for Climate Impact Assessment (IPCC-TGCIA), 1999. Guidelines on the Use of Scenario Data for Climate Impact and Adaptation Assessment, IPCC, 69 p. New, M., Lister, D., Hulme, M. and Makin, I., 2002. A high-resolution data set of surface climate over global land areas. Climate Research, 21: 1–25. Tegart, W.J., Sheldon, G.W. and Griffiths, D.C., 1990. Climate Change – The IPCC Impact Assessment, WMO/UNEP Intergovernmental Panel on Climate Change, Australian Government Publishing Service, Canberra, Australia. Watson, R., Zinyowera, M. and Moss, R. (eds.), 1996. Climate Change 1995 - Impacts, Adaptation and Mitigation of Climate Change. Contribution of WG II to the Second Assessment Report of the IPCC, Cambridge Univ. Press.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 21

OPTIMIZING IRRIGATION WATER MANAGEMENT ON THE GLOBAL CHANGE CONTEXT IN A SPANISH MEDITERRANEAN REGION José Antonio Rodríguez1∗, Ángel Utset2, Carmen Navarro1, Juan Carlos Martos1 and Ana Iglesias3 1

Empresa Pública Desarrollo Agrario y Pesquero. Junta de Andalucía 2 Instituto Tecnológico Agrario de Castilla y León 3 Universidad Politécnica de Madrid

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The aim of this work was to analyse the impact of different irrigation modernization strategies on the management, the productivity and the economic efficiency of irrigated agricultural production at farm scale, within the context of the Water Framework Directive, the Common Agricultural Policy Reform and the climate change. Maize was selected as representative crop of the irrigated agriculture in the central part of the Guadalquivir Valley, which is quite sensitive to the changes derived from the European policies. Optimal strategies of irrigation water management were simulated under three hypothetical solutions of irrigation modernization in the context of three combined scenarios of both, water and agricultural policies, together with a climate change scenario. DSSAT model outputs combined with a seasonal model for irrigation management optimisation at farm level were used for these purposes. Output results showed that the maize crop was more cost-effective and economic efficient within the reference situation of agricultural production and irrigation systems based on high levels of modernization. Under these circumstances, optimal irrigation management was addressed to maximize the crop production. However, a remarkable loss of profitability and economic efficiency was detected in the context of agricultural and water policies induced by the Global Sustainability and World Markets scenarios combined with a climate change scenario. Under this situation, the more suitable solutions were the ones based in medium levels of irrigation modernization.

∗

[email protected]

284

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

1. INTRODUCTION Among the main factors which are becoming important in the process of decision making of solutions for irrigation modernization, the combination of environmental and economic forces arising from the Water Framework Directive (WFD), the Common Agricultural Policy Reform (CAP) and the climatic change can be identified. The modification of the CAP can affect the producers in a negative way as shortage of direct subsidies makes agrarian income decrease in a so government supported agriculture as European agriculture. Transposition of WDF establishes an increase of irrigation costs by the combination of more restrictive environment rules together with the application of policies of water charges based in partial or total cost recovery of water services. On the other hand, different studies have forecasted a rise in temperatures in Spain between 2.5 and 4.8 ºC, a significant reduction of rainfall and an increase of frequency of extreme climatic events (Iglesias et al., 2000; Guereña et al. 2001). In order to stand these new circumstances in the best way, the adoption of solutions for irrigation modernization which can be more effective and productive, as well as, the application of different strategies of irrigation water management become necessary, so that economic proficiency of farm agricultural production is not severely affected. The objective of this work was to analyse the impact of different irrigation modernization strategies on the management, the productivity and the economic efficiency of irrigation at farm level, within the context of Water Framework Directive, the Common Agricultural Policy Reform and the climate change. The case study chosen in this study was the irrigated agriculture in the central part of the Guadalquivir Valley, which is characterized by a prevalence of surface irrigation practices and a remarkable trend of transformation into pressurized systems. Among the great variety of crops of the zone, maize was selected because it is very sensitive to forces for change coming from the context mentioned before.

2. MATERIAL AND METHODS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.1. Crop Model Parameterization The crop simulation model used in this study is included in DSSAT 4.0 (Jones et al., 2003). The decision support system for agrotechnology transfer (DSSAT) is a collection of independent programs that operate together. In our study we used the CERES-Maize (Jones and Kiniry, 1986) which is a simulation model for maize that describe daily phenological development and growth in response to environmental factors (soils, weather and management). Modeled processes include phonological development, i.e. duration of growth stages, growth of vegetative and reproductive plant parts, extension growth of leaves and stems, senescence of leaves, biomass production and partitioning among plant parts, and root system dynamics. The model includes subroutines to simulate the soil and crop water balance and the nitrogen balance, which include the capability to simulate the effects of nitrogen deficiency and soil water deficit on photosynthesis and carbohydrate distribution in the crop. CERES-Maize model calibration and validation was based on Iglesias (1994). The input data for the calibration and validation process were obtained from published field experiments conducted at the Agricultural Research Stations of Lora del Rio and Montoro, Sevilla, Spain

Optimizing Irrigation Water Management on the Global Change Context…

285

(37.42º N, 5.88º W, 31 m altitude). Maize hybrids selected for the calibration represent highly productive simple hybrids grown in the agricultural region. Table 1. Values of the calibrated genetic coefficients used as input for the CERES-Maize model Hybrid 700 800

Thermal Units1 2800 3000

Cycle Length2 130 150

P13 220 260

P24 0.52 0.5

P55 910 980

G26 700 600

G37 7.0 8.5

1 Accumulated total thermal units during the growing cycle (sum of degree days above 8ºC). 2 Average duration of the growing cycle (days) in Sevilla. 3 P1: Juvenile phase coefficient. 4 P2: Photoperiodism coefficient. 5 P5: Grain filling duration coefficient. 6 G2: Kernel number coefficient. 7 G3: Kernel weight coefficient.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Iglesias (1994) calibrated and validated the model with independent field data sets for maize hybrids of different crop growth duration. The coefficients were first calibrated in relation to phenology based on the thermal integrals of the juvenile period and of the reproductive period. Once the phenology coefficients were calibrated, and therefore the simulated number of days available for grain filling, the yield component coefficients were adjusted to represent as accurately as possible the number of grains ear, the final grain yield, and the final biomass. The values of the calibrated genetic coefficients used in this study are shown in Table (1). We particularly used the data corresponding to Hybrid 800.

2.1.1. Climatic Data Daily climate variables (maximum and minimum temperature, precipitation and daily hours of sunshine) corresponding to Aeropuerto Cordoba meteorological station (37.50º N, 4.51º W, 92 m altitude, Cordoba Spain) were provided by the Instituto Nacional de Meteorología for a time series of 44 years (1961-2004). Daily solar radiation was calculated using the approach described in Rietveld (1978). According to the climate variables available for this study, the reference evapotranspiration was calculated in the CERES model with the Priestley-Taylor relation. Potential transpiration is directly related to potential evapotranspiration by a coefficient (alpha) which value is fixed in the model. In many areas of the Mediterranean region, maximum temperatures over 35ºC occur in the summer months of July and August. These conditions imply that the advective and micro-advective (between rows) processes occur increasing crop ET. Such conditions were not represented in the original ET formulation of the CERES-Maize model, and therefore, simulations of crop ET with the original model underestimated field-observed values. When advective conditions prevail, the mentioned coefficient should be higher (Shouse et al., 1980; Pereira and Villa-Nova, 1992). Therefore, the coefficient alpha was set at 1.26 when maximum temperatures were below 35ºC and 1.45 above 35ºC, according to suggestions of Iglesias (1994).

286

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

2.1.2. Soil Data Soil data needed in DSSAT model are values for the functions of drainage, runoff, evaporation and radiation reflection, soil water holding capacities and rooting preference coefficients for each soil layer. The characteristics of the soil (Chromoxerert Typic according to USDA classification) used in the simulations are show in Table 2. Table 2. Soil data used in the simulations with DSSAT crop model Horizon A B C D Horizon A B C D

Depth (cm) 0-25 25-35 35-70 70-120 Depth (cm) 0-25 25-35 35-70 70-120

pH (H2O) 7.50 7.60 7.90 7.70 Bulk Density (gr/cm3) 1.64 1.63 1.71 1.69

EC1 (mS/cm) 0.70 1.10 1.20 2.50 FC2 (%Weight) 30.50 28.70 29.50 27.40

Sand (%) 30.00 29.00 29.00 27.00

Silt Clay (%) (%) 17.00 53.00 20.00 49.00 20.00 51.00 24.00 40.00 PWP3 (%Weight) 21.6 19.9 19.9 18.6

1: Electric Conductivity. 2: Field Capacity. 3: Permanent Wilting Point.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.2. Irrigation Optimization Model and its Integration with DSSAT Model In order to determine the optimal strategies of irrigation water management, a seasonal model for irrigation management optimisation at farm level was developed. In this model several sub-models were integrated, making possible to simulate the temporal evolution of soil water content (soil water balance), the irrigation performance (irrigation hydraulic) and the crop production (multiplicative production function). Economic optimisation was approached with the Dynamic Programming method, with an objective function focussed to maximize the net profits obtained by subsidies and markets of the agricultural production. Integration between DSSAT model and the irrigation optimisation model was undertaken, indirectly, by three processes: The irrigation optimisation model used the daily values of reference evapotranspiration calculated in DSSAT for each climatic data series (historical time series data and forecasted time series data by a climate change scenario). Moreover, other data of interest for the irrigation optimisation model were estimated, such as maximum root depth and duration of crop growth stages. Daily evolution of basal crop coefficient was determined as the relation between potential transpiration and reference evapotranspiration calculated by DSSAT model. Basal crop coefficients were used directly in the irrigation optimisation model. The multiplicative production function used in the irrigation optimisation model was calibrated with the outputs of DSSAT. To make this possible, successive DSSAT simulations were run, by introducing different water stress levels in every phase of crop development.

Optimizing Irrigation Water Management on the Global Change Context…

287

Simulation with DSSAT were made with mean daily values of the historical climatic data as well as with mean daily values forecasted for the year 2020 by a climate change scenario. Calibration of the production function proposed by Doorenbos and Kassam (1979) (equation 1) was carried out by determining the values of yield response factors Kyi through lineal regression between the relative yield and the relative evapotranspiration obtained by DSSAT for each individual period of crop growth. Additionally, Readily Available Water (RAW) for each crop development phase were obtained.

⎛ Ya ⎞ ⎞ ⎛ ⎜1 − ⎟ = Kyi ⎜1 − ETai ⎟ ⎟ ⎜ ⎜ Y ⎟ p ⎠ ⎝ ETpi ⎠ ⎝

(1)

where ETai is the actual evapotranspiration, ETpi is evapotranspiration for non-limiting water conditions during the ith stage of growth, Ya is the actual harvested yield and Yp is the maximum crop yield under given management conditions that can be obtained when water is not limiting.

2.2.1. Soil Water Balance Model The model proposed by Allen et al. (1998) was used to calculate a daily water balance in the soil-plant-atmosphere complex. Potential and actual evapotranspiration were estimated by the method of dual crop coefficients, taking into account the water stress conditions. As well as in DSSAT model, effective precipitation was estimated by the SCS method (Soil Conservation Service, 1972), modified by Williams et al. (1984) in order to consider soil stratification and soil water content at the time when rainfall occurs. Most of input data to estimate water requirements of maize were obtained from the DSSAT simulations, as it has been explained in the previous section. Other data used in this study are exposed as follows:

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Maize potential grain yield (Kg/ha) : 18000 Sowing date: 1st April Furrow spacing (m): 0.75

2.2.2. Farm Irrigation Model A mathematical model was developed in order to simulate all phases (advance, storage, depletion and recession) of furrow irrigation with free runoff. For the simulation of the advance phase, a solution developed by Valiantzas (1997a, b) was used, whereas for the calculation of the depletion and recession phases a simple solution to the volume balance equation proposed by Camacho et al. (1999) was adopted. The model took into account the influence of the furrow wetted perimeter on the infiltration parameters by the procedure formulated by Rodríguez (2003). The infiltration process was described by the Kostiavo’s equation, while hydraulic resistance was expressed in terms of the Manning’s equation with a roughness coefficient of 0.04 and a furrow longitudinal slope of 0.002 m/m. The parameters used to determined infiltration in the soil were the following ones: K = 0.006 m3/m/mina and a= 0.4. The geometric and hydraulic parameters of the furrow cross-section were described by

288

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

a potential equation (Walker, 1989), whose parameters were p1= 0.2727 and p2 = 2.6737. For all the scenarios, the same furrow length of 300 m was assumed. For drip irrigation system modelling, an Application Efficiency of 90%, a drip discharge of 2.3 L/h and a density of 6666 drips/ha were assumed.

2.2.3. Crop Production Model The Jensen’s model (Jensen, 1968) was used to estimate the actual crop yield. This model establishes the relationship between relative yield and the relative evapotranspiration and it integrates the effect of all the water stress throughout the growing season as follows (equation 2): N ⎛ ETai Ya = ∏ ⎜⎜ Yp i =1 ⎝ ETp i

⎞ ⎟⎟ ⎠

λi

(2)

where λi is the sensitivity index of crop to water stress during the ith stage of growth and N is the number of growth stages. The Jensen’s model (Equation 2) was used at time steps less than a growth stage, for example an irrigation interval, through the cumulative procedure presented by Tsakiris (1982). In order to relating the yield response factors Kyi, calibrated from DSSAT results, to the sensitivity index of Jensen’s model, a polynomial function proposed by Kipkorir y Raes (2002) was used.

2.2.4. Economic Optimization Model Dynamic Programming was implemented as the method for economic optimisation, in which each irrigation event was considered as a stage of the process. The states (or decisions) were established with the dates of the irrigation events, the furrow inflow rates or the fractions of the net irrigation depth to apply in drip systems. As objective function, the maximization of net profit of agricultural production was defined, so that the maximum value of equation (3) should be found:

[

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Bn = [(Ya ⋅ Pc ) + S ] − (Va ⋅ Ca ) + (Ya ⋅ C p ) + Cm + C f

]

(3)

where Bn is the net profit (€/ha); Ya is the actual crop yield (Kg/ha); Pc is the price of production commercialisation (€/Kg); S is the subsidy obtained (€/ha); Va is the volume of water applied by irrigation (m3/ha); Ca is the cost of irrigation water (€/m3); Cp is the cost related to crop production (€/kg); Cm is the labour cost for an irrigation practice and Cf is the fixed cost (sum of production and irrigation costs) (€/ha).

2.3. Climate Change Scenario The climate change scenario forecasted for the year 2020 were taken from the CGCM2 model outputs, provided by the Canadian Centre for Climate Modelling and Analysis (Flato

Optimizing Irrigation Water Management on the Global Change Context…

289

et al., 2000; Flato and Boer, 2001). The IPCC SRES A2 scenario for greenhouse gases emissions (IPCC, 2001) was considered. In order to downscaling the forecasted climatic data, the outputs of a weather generator were perturbing according to the CGCM2 results corresponding to the study site, i.e., Southwest of Iberian Peninsula. The historical weather data were used in combination with the LARS-WG weather generator (Semenov and Barrow, 2002) to generate 100 realizations of local weather corresponding to the year 2020. The relative change in wet and dry series lengths, temperature standard deviations, mean temperature, precipitation depth and solar radiation as affected by global change, was done following the approach recommended by Semenov and Barrow (2002), based on the daily CGCM2 outputs for each ten-year range.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.4. Agricultural and Water Policies Combined Scenarios Three combined scenarios of both agricultural and water policies were defined by introducing certain modifications to the original proposals of the WADI Project scenarios (Berbel et al., 2003): Baseline Scenario (BS): Agenda 2000 CAP reform, moderate intervention regime, prices support, export subsidies and existing water policy. Climate corresponding to historical time series data. Global Sustainability (GS): CAP midterm review, medium to low intervention regime, market orientation with targeted sustainability compliance requirements and programs, WFD application and cost recovery of water services. Climate corresponding to forecasted time series data by the climate change scenario. World Markets (WM): World agricultural markets without CAP, free trade, no intervention regime and unrestricted water markets drivers for water abstraction. In contrast to the original WADI proposal, a single farm payment independent from production as support to farmers income, was also considered (Morris y Twite, 2002; European Commission, 2003). Climate corresponding to forecasted time series data by the climate change scenario. For each of the above scenarios, commercialisation prices and subsidy for maize crop production were estimated (Table 3). Likewise, the actual costs of crop production were modified in order to make them fit GS and WM scenarios. The cost variation percentages defined in WADI project were used, but they were adjusted to the conditions for agricultural production of the Guadalquivir Valley (Berbel et al., 2004). All the commercialisation prices, subsidies and production costs, which were estimated for GS and WM scenarios were projected to the year 2020. Actual maize production costs were extracted from different reports made by the Prospective Unit of the Agricultural and Fisheries Ministry of the Regional Government of Andalusia. These costs do not include the charges related to irrigation, which are defined later in the formulation of the irrigation modernization scenarios. The production costs used in this study were: Production cost for drip irrigation (€/ha): 958.70 Production cost for furrow irrigation (€/ha): 965.53 Cost related to maize production (€/Kg): 0.024

290

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

Table 3. Market prices and subventions of maize in each agricultural and water policies combined scenario Crop

Maize

Baseline Scenario Price1 Subvention2 (€/Kg) (€/ha) 0.16 449

Global Sustainability Price3 Single Payment4 (€/Kg) (€/ha) 0.152 397

World Markets Price5 Single Payment (€/Kg) (€/ha) 0.096 397

1

Intervention price of 101.35 €/ton. Support system for producers of certain arable crops (Regulation CE 1251/1999). 3 CAP midterm review (Regulation CE 1782/2003 ). 4 According to Regulation CE 1782/2003. 5 Forecasted price of maize world market (European Commission, 2003). 2

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

2.5. Irrigation Modernization Scenarios In relation to different levels of investment, three scenarios for irrigation modernization were identified. The characteristics of these scenarios are: Basic Level of Irrigation Modernization: The irrigation network is composed of lined channels with few manual elements of flow control, low flexibility in water supply service and furrow irrigation system at farm level. Medium Level of Irrigation Modernization: The main and secondary irrigation networks are composed of lined channels, the tertiary and quaternary networks are composed of low pressure pipelines system with enough elements of flow control with local mechanisms for automation, high flexibility in water supply service and modern furrow irrigation system at farm level. High Level of Irrigation Modernization: The irrigation network is composed by pressurized pipelines system and elements of flow control, high level of automation, water supply on user demand and drip irrigation system at farm level. With the purpose of estimating the cost related to irrigation for each modernization scenario, seven Water Users Associations with different levels of investment in irrigation infrastructure and representative of the Guadalquivir Valley were analysed (Genil-Cabra, Fuente Palmera, MI del Genil, Pantano del Guadalmellato, Sector B-XII del Bajo Guadalquivir, CB Almonazar, Builtrago y Espartales y MD del Río Bembézar). The main characteristics of the Water User Associations were extracted from a study of Characterization of Irrigation Districts in Andalusia (Consejería de Agricultura y Pesca, 1999). These Associations were classified according to the three levels of modernization described before, for which averaged weighted costs and the more representative characteristics were obtained (Table 4). Finally, the costs related to irrigation were increased in accordance with to the Consumer Prices Index evolution since 1998, and they were modified according to the variation percentages defined in the GS and WM scenarios.

2.6. Economic and Irrigation Productivity Indicators Net Profit of agricultural production (€/ha), Irrigation Productivity (Kg/m3) and Economic Efficiency (€/m3) were used as the indicators to evaluate the performance of each

Optimizing Irrigation Water Management on the Global Change Context…

291

scenario defined in this study. The Irrigation Productivity express the actual crop yield per gross irrigation volume applied, while the Economic Efficiency states the relationship between the Net Profit of agricultural production and the gross irrigation volume. Table 4. Characteristics and irrigation cost for each irrigation modernization scenario Policies Scenario

BS

1

GS2

WM3

Level of Irrigation Modernization Basic Medium High Basic Medium High Basic Medium High

Min/Max Irrigation Interval (days) 20/25 5/30 1/30 20/25 5/30 1/30 20/25 5/30 1/30

Charge by Irrigation Area (€/ha) 138.07 245.61 148.88 84.77 175.61 103.42 ----------

Charge by Irrigation Volume (€/m3) ------0.027 0.013 0.018 0.085 0.022 0.034 0.106

Irrigated Area in a Day (ha) 3 6 40 3 6 40 3 6 40

Labor Cost

(€/day/ha)4 13.02 6.51 0.98 17.26 8.63 1.29 14.98 7.49 1.12

1

In this scenario the charge by irrigation volume only includes the cost of the energy consumed in the high level of modernization. 2 In this scenario the charge by irrigation area includes operation and maintenance actions and the administrative costs, while the charge by irrigation volume includes the charge of the Administrative Water Manager (Confederación Hidrográfica) and the cost of the energy consumed. 3 In this scenario all the concepts are paid in terms of volume of water consumed. 4 Daily labor cost of irrigation practice was set to 39.07 €.

3. RESULTS AND DISCUSSION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

3.1. Climate Change and Crop Water Requirements Despite other global circulation models, CGCM2 provides free internet access to daily simulation data in a text format. Hence, this model is more suitable for simple agricultural applications anywhere. According to Merrit et al (2006), results considering CGCM2 are similar than those obtained through other general circulation models. On the other hand, a weather generator produces synthetic daily time series of climatic variables statistically equivalent to the recorded historical series, as well as daily site-specific climate scenarios that could be based on regional GCM results (Semenov and Jamieson, 2001). Different weather generators are available, but according to Wilby and Wigley (2001), the US-made and the UK-made WGEN and LARS-WG are the most widely used. Besides, LARS-WG results are as accurate as those obtained with WGEN and other weather generators (Mavromatis and Jones, 1998; Semenov et al., 1998; Mavromatis and Hansen, 2001). Therefore, the CGCM2 model outputs downscaled with LARS-WG weather generator, turned out to be a practical and functional approach for climate change forecasting in this agricultural application.

292

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

Climate change scenario used in this study predicted a highest increase of 4.5 ºC and 4.1 ºC for maximum and minimum daily temperature respectively (Figure 1a). Annual precipitation seemed to decrease about 38 mm at the end of the considered period (year 2020) (Figure 1b). However, there were no significant differences in solar radiation throughout the analyzed period. This results are consistent with the conclusions obtained in other studies for the south region of Spain (Iglesias et al., 2000; Guereña et al. 2001). A close look to Figure 1(b) shows that the shortage of annual precipitation forecasted for 2020 was, basically, a direct consequence of the winter rainfall decrease. As a matter of fact, during the maize growing period (from April to August) the forecasted precipitation increased up lightly in 2.2 mm with respect to the historical average. 45

Forecasted Weather Data Year: 2020 100 Realizations

Mean Daily Temperature (ºC)

40

Historical Weather Data Years: 1961-2004 44 Years

35 30

Maximum Temperature

25 20 15

Minimum Temperature

10 5

80 10 0 12 0 14 0 16 0 18 0 20 0 22 0 24 0 26 0 28 0 30 0 32 0 34 0 36 0 38 0

60

40

0

20

0

a

DOY Accumulated Mean Daily Precipitation (mm)

550 500 450

Forecasted Weather Data Year: 2020 100 Realizations

400 350 300 250 200

Historical Weather Data Years: 1961-2004 44 Years

150 100 50

80 10 0 12 0 14 0 16 0 18 0 20 0 22 0 24 0 26 0 28 0 30 0 32 0 34 0 36 0 38 0

60

40

20

0 0

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

600

DOY

b

Figure 1. Historical and forecasted weather data corresponding to Aeropuerto Cordoba meteorological station (Cordoba, Spain). (a) Maximum and minimum temperatures. (b) Accumulate annual precipitation.

Optimizing Irrigation Water Management on the Global Change Context…

293

On the other hand, the biggest increase in the temperatures forecasted for 2020 took place from June to September, coinciding with the maximum crop water requirements season. Reference evapotranspiration (ETo) and maize potential transpiration (Tp) increased up to a maximum daily value of 0.9 mm/day and 1.0 mm/day respectively in 2020 (Figure 2a and 2b), which represents only a 3.6% and 4.8% annual increment from the historical reference period. 9.00

Forecasted Weather Data Year: 2020 100 Realizations

8.00

Historical Weather Data Years: 1961-2004 44 Years

ETo (mm/d)

7.00 6.00 5.00 4.00 3.00 2.00 1.00 0

20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380

DOY

a

9.00

Forecasted Weather Data Year: 2020 100 Realizations

8.00 7.00

Tp (mm/d)

6.00 5.00 4.00 3.00

Historical Weather Data Years: 1961-2004 44 Years

2.00 1.00 0.00

90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

DOY

b

10.00

Net Irrigation Requirements (mm/d)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

9.00 8.00

Forecasted Weather Data Year: 2020 100 Realizations

7.00 6.00 5.00 4.00 3.00

Historical Weather Data Years: 1961-2004 44 Years

2.00 1.00 0.00

90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

DOY

c

Figure 2. Historical and forecasted Reference Evapotranspiration (a), Maize Potential Transpiration (b) and Net Irrigation Requirements (c), corresponding to climatic data of Aeropuerto Cordoba meteorological station (Cordoba, Spain).

294

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

From these graphics (Figure 2a and 2b) it was possible to estimate the daily evolution of basal crop coefficient of maize as an input to the irrigation optimization model, while we kept consistency with the DSSAT results. The net irrigation requirement is the sum of the water amount that had to be added during the simulation period to avoid crop water stress. The net requirement does not consider extra water that has to be applied to the field to account for conveyance losses or the uneven distribution of irrigation water on the field. Therefore, the forecasted increment of crop evapotranspiration evenly increased up the net irrigation requirements in 40.2 mm respect to the historical irrigation water demand (Figure 2c). As we mentioned before, the predicted increment of net irrigation requirements of maize was caused by the ascent of temperatures, because the precipitation fallen over the crop development period was practically the same in both, the historical and the forecasted climatic conditions.

3.2. Crop Production Function Figure 3 shows the relation between relative yield (1-Y1/Yp) and relative evapotranspiration (1-ETa/ETp), as well as, the average crop stage length and average Readily Available Water (RAW) obtained by DSSAT for each individual growing period of maize. 0.275 0.25

y = 0.389x R2 = 0.857 Initial Stage

0.225 0.2

Stage Length= 27 days RAW=0.46

(1-Ya/Yp)

0.175 0.15 0.125 0.1

Historical Weather Data Forecast Weather Data

0.075 0.05 0.025 0 0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

0.45 0.5 0.55 0.6 0.65 0.7

(1-ETa/ETp) 0.275

a

y = 0.502x R2 = 0.843 Crop Development Stage

0.25 0.225

Stage Length= 38 days RAW=0.53

0.2

(1-Ya/Yp)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0.3

0.175 0.15 0.125 0.1

Historical Weather Data Forecast Weather Data

0.075 0.05 0.025 0 0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7

(1-ETa/ETp)

b

Optimizing Irrigation Water Management on the Global Change Context…

295

0.8 0.75 0.7

y = 1.286x R2 = 0.842 Mid-Season Stage

0.65 0.6 0.55

Stage Length= 53 days RAW=0.61

(1-Ya/Yp)

0.5 0.45 0.4 0.35 0.3 0.25

Historical Weather Data Forecast Weather Data

0.2 0.15 0.1 0.05 0 0

0.05

0.1

0.15

0.2

0.25 0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

(1-ETa/ETp) 0.325

c

0.3

y = 0.509x R2 = 0.767 Late Stage

0.275 0.25

(1-Ya/Yp)

0.225

Stage Length= 32 days RAW=0.77

0.2 0.175 0.15 0.125 0.1 0.075

Historical Weather Data Forecast Weather Data

0.05 0.025 0 0

0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7

(1-ETa/ETp)

d

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Relationships between relative yield decrease (1 - Ya/Ym) and relative evapotranspiration deficit for the individual growth periods of maize obtained from DSSAT model. (a) Initial stage. (b) Crop development stage. (c) Mid-season stage. (d) Late stage.

In all cases, a significant linear regression between the relative yield and the relative evapotranspiration was obtained. However, major variations between the regression coefficients obtained with the historical climate data and the forecasted climatic data were not detected. That is why the same values of Kyi, crop stage length and RAW were used in both climatic conditions. According to the production model of Doorenbos and Kassam (1979), Kyi values were obtained as the slope of the regression line in each growth stage of maize crop. Figure 3 shows that both, crop development and late stages, exhibited a similar response to water stress (Kyi=0.5), the mid-season stage was more sensitive with a Kyi of 1.3, while the initial stage was the one which can stand the highest water deficits with the least impact on final production, with a Kyi of 0.4.

296

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

3.3. Net Profit, Economic Efficiency and Irrigation Productivity Table 5 contains the economic indicators obtained for each level of irrigation modernization in the context of the agricultural and water policies scenarios. The output results of Table 5 belong, for all the cases, to an intermediate result between the biggest Net Profit solution and the maximal Economic Efficiency solution, as most representative one of the farm income according to Wilchens (2002) and Ortega et al. (2004) recommendations. Table 5. Economic indicators obtained for each level of irrigation modernization combined with agricultural and water policies scenarios

Crop Maize

Maize

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Maize

Level of Irrigation Scenario Modernization BS Basic GS WM BS Medium GS WM BS High GS WM

Net Profit (€/ha) 1034.80 512.80 73.50 1620.50 1020.50 293.30 1754.50 631.60 16.00

Economic Efficiency (€/m3) 0.087 0.048 0.007 0.146 0.112 0.033 0.241 0.084 0.003

Irrigation Productivity (Kg/m3) 1.099 1.223 1.253 1.627 1.964 2.002 2.792 2.808 2.831

As it was expected, all solutions obtained for the GS and WM scenarios bring about a remarkable loss in profitability and economic efficiency with respect to the reference situation (BS Scenario). At the same time, a certain trend to increase the irrigation productivity can be detected, though this trend is more clear in the solutions with basic and medium levels of irrigation modernization than in those ones based on high levels of modernization (Table 5). Although the irrigation solutions based on high levels of modernization reached the highest irrigation productivities in all scenarios, their increment gradients were lower than the rest of solutions (Figure 4a). For these solutions, a decrease of the irrigation water volume resulted in an important loss of crop production. On the other hand, the irrigation solutions based on medium levels of modernization reached the highest increment gradients of productivity, approaching the theoretical upper limit. These solutions were able of assimilating a remarkable reduction of irrigation water volume with a minimal impact on maize production. The maize resulted more profitable and economically efficient within the reference framework of agricultural production (BS Scenario) and irrigation systems based on high levels of modernization. However, this situation drastically changed within the context of GS and WM scenarios (Table 5). Within BS scenario, the irrigation management strategies obtained for the solutions with high levels of modernization were already focused on minimizing the costs, in order to compensate the high operation fees that these systems have to pay out. Therefore, these solutions were not able of absorbing the increments of production

Optimizing Irrigation Water Management on the Global Change Context…

297

costs and water demands, as well as, the decrease of market prices and subsidies that GS and WM scenarios imposed, without a notable loss in farm income. 28 Basic Level of Irrigation Modernization

Increment of Irrigation Productivity (%)

26

Medium Level of Irrigation Modernization 24

High Level of Irrigation Modernization

22

Theorical Upper Limit

20 18 16 14 12 10 8 6 4 2 0 0

2

4

6

8

10

12

14

16

18

20

22

a

Irrigation Volume Depletion (%) 100 Basic Level of Irrigation Modernization

Loss of Economic Efficiency (%)

90

Medium Level of Irrigation Modernization High Level of Irrigation Modernization

80 70 60 50 40 30 20 10 0 0

10

20

30

40

50

60

Loss of Net Profit (%)

70

80

90

100

b

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 4. Percentage variations of the economic indicators with respect to reference situation (BS scenario) for each irrigation modernization scenario. (a) Increment of Irrigation Productivity. (b) Loss of Economic Efficiency.

From Figure 4(b) it can be noticed that the loss in economic efficiency was practically proportional to the loss in profitability in the solutions based on high levels of irrigation modernization. However, the loss in efficiency was less than proportional to the decrease in profitability in the systems with medium levels of modernization. This means that the irrigation systems based on medium levels of modernization were more elastic and, consequently, they were able of assimilating the new paradigm that transposition of the Water Framework Directive, the revision of the CAP and the climate change supposed. On the other hand, the solutions with low levels of irrigation modernization cannot reached enough elasticity because they offered a water supply service with very low flexibility.

298

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

3.4. Irrigation Management In Table 6 it can be observed that optimal irrigation management in the reference scenario was directed to maximize the crop production (ETa/ETp=1) (Figure 5a). However, the optimal strategy of the irrigation management in the GS and WM scenarios was focused to minimize the water volume, although this it implied certain loss of crop production. Evidently, solutions with basic levels of irrigation modernization were not able to reach the maximum crop production even in the baseline scenario, given the low flexibility of their irrigation services. Optimal irrigation management recommended for the GS and WM scenarios was focus to reduce the irrigation water volume and to diminish as well, the impact on the crop production. As we mentioned before, this strategy comes near to the water management practiced in the baseline scenario for irrigation systems with high levels of modernization. Consequently, theses irrigation solutions have less margin of adaptation to the new conditions that the GS, WM and the climate change scenarios imposed, being reflected in a remarkable loss of profitability and economic efficiency. The irrigation management recommended in the GS and WM scenarios for the irrigation solutions with basic levels of modernization, characterized by low flexibility in the water supply services, was addressed to reduce of the number of irrigation events and the induction of a limiting water regime, even more restricted that the practiced one in the reference scenario (Table 6). Under these conditions, the economic benefits of maize production probably will not be attractive anymore for the farmers.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Table 6. Optimal irrigation management parameters for each level of irrigation modernization combined with agricultural and water policies scenarios

Level of Irrigation Scenario Crop Modernization BS Maize Basic GS WM BS Maize Medium GS WM BS Maize High GS WM

Irrigation ETa/ETp Volume (m3/ha) 0.894 11932.80 0.885 10709.20 0.885 10709.20 1.000 11061.30 0.951 8733.20 0.942 8520.00 1.000 6447.40 0.983 6404.40 0.972 6311.50

Number of Irrigation Events 7 6 6 9 9 9 57 57 57

Mean Net Irrigation Depth (mm) 88.20 90.40 90.40 64.00 62.00 60.90 11.70 11.00 10.50

Mean Application Efficiency (%) 51.70 50.60 50.60 52.00 65.10 67.40 90.00 90.00 90.00

Optimal irrigation management recommended for the solutions sustained in medium levels of modernization under the context of GS and WM scenarios was directed to increase up the efficiency of the irrigation events, as well as, the promotion of water stress in the less sensible growth stages of the crop (Figure 5). This strategy minimized the irrigation water volume and diminished as well, the loss of crop production. As it can be appraised in Table 5,

Optimizing Irrigation Water Management on the Global Change Context…

299

the profitability of the maize obtained in these solutions could still be economically interesting at least in the GS scenario context. 160 Rain

Baseline Scenario

Irrigation

140

Soil Water Depletion Readily Available Water

Depth (mm)

120 100 80 60 40 20

23 2

22 2

21 2

20 2

19 2

18 2

17 2

16 2

15 2

14 2

13 2

12 2

11 2

10 2

92

0

DOY 160

Rain

a

Global Sustainability Scenario

Irrigation

140

Soil Water Depletion Readily Available Water

120

Depth (mm)

100 80 60 40 20

DOY

23 2

22 2

21 2

20 2

19 2

18 2

17 2

16 2

15 2

14 2

13 2

12 2

11 2

10 2

92

0

b

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 5. Optimal irrigation management for maize in a system with medium level of irrigation modernization. (a) Baseline Scenario. (b) Global Sustainability Scenario.

The induction of water stress in all growth stages of the crop was the only option that irrigation solutions with high levels of modernization managed for assimilating the new exigencies that GS and WM scenarios supposed (Table 6). These solutions did not have sufficient elasticity to surely compensate the increments of production costs and water demands that GS and WM scenarios required, bringing an important decreased of the economic profitability of the maize and, surely, an abandonment of their commercial production under these circumstances.

CONCLUSIONS Climate change scenario used in this study predicted a highest increase of 4.5 ºC and 4.1 ºC for maximum and minimum daily temperature respectively, and a reduction of annual precipitation of 38 mm. Reference evapotranspiration and maize potential transpiration increased up to a 3.6% and 4.8% from the historical reference period. The forecasted increment of temperatures evenly increased the net irrigation requirements in 40.2 mm.

300

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

Maize crop was more cost-effective and economic efficient within the reference context of agricultural production and irrigation systems based on high levels of modernization. Under these conditions, optimal irrigation management was addressed to maximize the crop production. Maize production experimented a remarkable loss of profitability and economic efficiency in the context of agricultural and water policies induced by the Global Sustainability and World Markets scenarios. However, the irrigation systems based on medium levels of modernization were able of assimilating the new paradigm that transposition of the Water Framework Directive, the revision of the CAP and the climate change supposed. The profitability of the irrigated maize under these circumstances could still be economically interesting. Irrigation solutions based on high levels of modernization did not have sufficient elasticity to compensate the increments of production costs and water demands that Global Sustainability and World Markets scenarios required, being reflected in a remarkable loss of profitability and economic efficiency of maize production.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Allen, R. G.; L.S. Pereira; D. Raes and M. Smith. 1998. Crop evapotranspiration. Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56. Food and Agriculture Organization of the United Nations, Rome. 300 pp. Berbel, J.; M.J. López and C. Gutiérrez. 2004. Country summary: Spain. The case of Guadalquivir basin. Proceeding Future for Irrigation. A quantitative workshop. http://www.uco.es/investiga/grupos/wadi/. Córdoba. Spain. Berbel, J.; J. Morris; D. Boymanns; G. Bazzani; V. Gallerani; D. Viaggi; C. Twite; K. Weatherhead and K. Vasileiou. 2003. Future of Institutional framework related with irrigated system in Europe: combination of agricultural and water policies scenarios. WADI Project. Deliverable D5. 39 pp. http://www.uco.es/investiga/grupos/wadi/. Camacho, E; A.J. Clemmens and T. Strelkoff. 1999. Nuevo modelo de riego por surcos: I. Diseño. Riegos y Drenajes XXI, 105: 28-33. Consejería de Agricultura y Pesca. 1999. Caracterización de Comunidades de Regantes. Acciones relacionadas con la optimización del uso y gestión del agua de riego. Empresa Pública Desarrollo Agrario y Pesquero. Sevilla. 85 pp. Doorenbos, J. and A.H. Kassam. 1979. Yield response to water. FAO Irrigation and Drainage Paper 33. Food and Agriculture Organization of the United Nations, Rome. 193 pp. European Commission. 2003. Prospects for agricultural markets in the European Union. Reform of the common agricultural policy medium-term prospects for agricultural markets and income in the European Union 2003-2010. Brussels. 53 pp. Flato, G.M.; G.J. Boer; W.G. Lee; N.A. McFarlane; D. Ramsden; M.C. Reader; and A.J. Weaver. 2000. The Canadian Centre for Climate Modelling and Analysis Global Coupled Model and its Climate. Climate Dynamics, 16: 451-467. Flato, G.M. and G.J. Boer. 2001. Warming Asymmetry in Climate Change Simulations. Geophys. Res. Lett., 28: 195-198.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Optimizing Irrigation Water Management on the Global Change Context…

301

Guereña, A.; M. Ruiz-Ramos, C. H. Díaz-Ambrona, J. R. Conde and M. I. Mínguez. 2001. Assessment of Climate Change and Agriculture in Spain Using Climate Models. Agron. J. 93:237–249. Iglesias, A. 1994. Climate change impact simulation and management strategies of two references crops: Wheat and maize. Alternatives analysis. (In Spanish, with English Summary). Ph.D. thesis. Univ. Politecnica de Madrid, Madrid, Spain Iglesias, A. and C. Rosenzweig; D. Pereira. 2000. Agricultural impacts of climate change in Spain: developing tools for a spatial analysis. Global Environnemental Change 10: 6980. IPCC. 2001. Climate Change 2001: The Scientific Basis. J.T. Houghton et al. (eds.), Cambridge University Press, 881pp. Jensen, M. E. 1968. Water consumption by agricultural plants. Chap.1 in Water Deficits and Plant Growth. Vol. II, T. T. Kozlowski, ed. Academic Press, New York: 1-22. Jones, C.A. and J.R. Kiniry. 1986. CERES-Maize: A Simulation Model of Maize Growth and Development. Texas AandM University Press, College Station, Texas. Jones, J.W.; G. Hoogenboom; C.H. Porter; K.J. Boote; W.D. Batchelor; L.A. Hunt; P.W. Wilkens; U. Singh; A.J. Gijsman and J.T. Ritchie. 2003. The DSSAT cropping system model. Europ. J. Agronomy 18: 235-265. Kipkorir, E.C. and D. Raes. 2002. Transformation of yield response factor into Jensen’s sensitivity index. Irrigation and Drainage Systems 16: 47-52. Mavromatis, T. and J.W. Hansen. 2001. Interannual variability characteristics and simulated crop response of four stochastic weather generators. Agric. For. Meteorol. 109: 283-296. Mavromatis, T. and P.D. Jones. 1998. Comparison of climate scenario construction methodologies for impact assessment studies. Agric. For. Meteor. 91: 51-67. Merritt, W. S.; A. Younes; M. Barton; B. Taylor; S. Cohen; and D. Neilsen. 2006. Hydrologic response to scenarios of climate change in sub watersheds of the Okanagan basin, British Columbia. J. Hydrol. 326:79-108. Ortega, J.F.; J.A de Juan; J.M. Tarjuelo and E. López. 2004. MOPECO: an economic optimization model for irrigation water management. Irrigation Science. 23: 61-75. Pereira, A.R. and N.A. Villa-Nova. 1992. Analysis of the Priestley-Taylor parameter. Agric. For. Meteorol., 61: 1-9. Rietveld, M.R. 1978. A new method for the estimating the regression coefficients in the formula relating solar radiation to sunshine. Agricultural and Forest Meteorology 19: 243-252. Rodríguez, J.A. 2003. Estimation of advance and infiltration equations in furrow irrigation for untested discharges. Agricultural Water Management. 60: 227-239. Semenov, M.A. and E.M. Barrow. 2002. LARS-WG. A stochastic weather generator for use in climate impact studies. User Manual. Rothamstead Research, Hertfordshire, 27 pp. Semenov, M.A.; R.J. Brooks; E.M. Barrow and C.W. Richardson. 1998. Comparison of the WGEN and LARS-WG stochastic weather generators in diverse climates. Climate Research. 10: 95-107. Semenov, M.A. and P.D. Jamieson. 2001. Using weather generators in crop modelling. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat, 322 pp.

302

José Antonio Rodríguez, Ángel Utset, Carmen Navarro et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Shouse, P.; W.A. Jury and L.H. Stolzy. 1980. Use of deterministic and empirical models to predict potential evapotranspiration in an advective environment. Agronomy Journal, 72: 994-998. Soil Conservations Service (SCS). 1972. National Engineering Handbook, Hydrology Section 4, Chapters 4-10. Tsakiris, G. P. 1982. A method for applying crop sensitivity factors in irrigation scheduling. Agricultural Water Management. 5: 335-343. Valiantzas, J.D. 1997a. Surface irrigation advance equation: Variation of subsurface shape factor. Journal of Irrigation and Drainage Engineering, ASCE. 123(4): 300-306. Valiantzas, J.D. 1997b. Volume balance irrigation advance equation: Variation of surface shape factor. Journal of Irrigation and Drainage Engineering, ASCE. 123(4): 307-312. Walker, W.R. 1989. Guidelines for designing and evaluating surface irrigation system. FAO Irrigation and Drainage Paper No. 45, Rome, 137 pp. Wichelns, D. 2002. An economic perspective on the potential gains from improvements in irrigation water management. Agricultural Water Management. 52: 233-248. Wilby, R.L. and T.M.L. Wigley. 2001. Down-scaling general circulation issues in climate prediction. In Sivakumar, M.V.K. (Ed.) Climate Prediction and Agriculture, Proceedings of the START/WMO International Workshop held in Geneva, Switzerland, 27-29 September 1999, Washington D.C. USA, International START Secretariat. p 39-68. Williams, J.R.; C.A. Jones and P.T. Dyke. 1984. A modeling approach to determining the relationships between erosion and soil productivity. Transactions of the ASAE 27: 129144.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 22

SIMULATING PLANT GROWTH, SOIL-WATER AND NITROGEN DYNAMICS IN MAIZE CROP IN BRAZIL USING DSSAT MODEL Camilo de Lelis Teixeira de Andrade∗1, Ramon Costa Alvarenga1, Israel Alexandre Pereira Filho1, João Carlos Ferreira Borges2 and Cristiano Márcio Roque Silva3 1

Embrapa Maize and Sorghum, CP 151, 35701-970, Sete Lagoas, Minas Gerais, Brazil 2 Universidade Federal Rural do Pernambuco, Garanhuns, Pernambuco, Brasil 3 CNEC Community School, Sete Lagoas, MG, Brazil,

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Modeling is an important tool to evaluate water and nutrient stresses effects on crop productivities, in order to establish recommendations for long-term sustainable production systems. Practical utilization of models requires, however, their calibration and evaluation, especially for tropical ecosystems and no-tillage conditions. The objective of this work was to evaluate the Ceres-maize model capability to simulate maize crop growth, nitrogen extraction and soil-water dynamics under no-tillage compared to conventional tillage soil management. The model was calibrated for the maize triple-cross BRS 3060, using a data set collected at a field trial specially carried out for that purpose. Data to evaluate the model were collected at a long-term experiment used to compare no-tillage with conventional tillage management systems. Soil-water was monitored with gravimetric method from 0 to 20 cm and with neutron probe from 20 to 100 cm of the soil profile. Plant samples were collected along maize cycle to determine biomass, leaf area and nitrogen concentration. Grain productivity was also measured at physiological maturity. The model simulated reasonably well biomass production for year 2000, although it underestimated it in 1999, especially for intermediate crop stages. Leaf area index was simulated well at the disk plow soil management system in both ∗

[email protected]

304

Camilo de Lelis Teixeira de Andrade, Ramon Costa Alvarenga et al. yeas, but was underestimated at the no-tillage system. Crop grain productivity was systematically overestimated. Ceres-maize overestimated soil-water storage in 1999 but simulated it well up to 100 days after planting in 2000. No differences were observed between no-tillage and disk plow management systems regarding soil-water storage measured or simulated. Nitrogen leaf concentration was simulated reasonably well on both cropping systems in 2000. In 1999, the model overestimated N concentration at the beginning of the cycle and underestimated it at the end. Maize tops nitrogen extraction was well simulate in 1999 but overestimated in 2000. Overall, model calibration needs to be refined in order to make it simulate accurately the major maize crop growth processes, although it can be used for regional simulations.

INTRODUCTION Modeling can be used to evaluate the effects of water and nitrogen stresses on crop productivity, in order to access long term sustainability of production systems (Thorton et al., 1991). Several models, with different purposes, were developed in the last years. The package DSSAT (Decision Support System for Agrotechnology Transfer) is one of those models. It can simulate major growth processes of various crops (Jones et al., 1998) and soil-water and nitrogen dynamics (Pang et al., 1998; Garrison et al., 1999). Ceres-maize is one the DSSAT’s models that simulates maize crop growth. Practical utilization of this kind of model requires, however, calibration and evaluation for local and regional conditions (Boote et al., 1996). Ceres-maize model has already been calibrated for local conditions (Sans et al., 1994) and utilized to analyze maize irrigation and nitrogen fertilization strategies. However, the study was based on a low potential yield cultivar (Freitas et al., 1998). This work aimed at evaluating Ceres-maize model capability to simulate growth, grain yield and soil-water and nitrogen dynamics on a maize crop, grown under no-tillage and disk plow soil management systems.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

MATERIAL AND METHODS Ceres-maize model, version 3.5, was calibrated using data from trials carried out at Embrapa Maize and Sorghum Experimental Station in Sete Lagoas, MG, Brazil, specifically for that purpose. The triple cross hybrid BRS 3060 was grown for two years and managed to express its yield potential. Biomass production and leaf area were monitored along crop cycle. Flowering and physiological maturity dates were registered, along with grain yield and average grain weight. Data from two years were used to generate Ceres-maize genetic coefficients. Data from an independent experiment, in which no-tillage and disk plow soil management systems have being compared for 14 years, were used to, independently, evaluate the model. The same hybrid was planted in years 1999 and 2000, following a maizedry bean cropping sequence scheme, under irrigation. Fertilization was based on soil analysis and on expected grain yield. Irrigation was applied to supplement rainfall and was not very well managed, so that some water stress might have occurred along maize cycle. Irrigation depths were measured

Simulating Plant Growth, Soil-Water and Nitrogen Dynamics in Maize Crop…

305

with catch cans. Neutron probe and soil sampling with gravimetric method were used to monitor soil-water dynamics. Biomass, leaf area and plant nitrogen concentration were also measured along crop cycle. Yield was computed after physiological maturity. Basic soil parameters (Andrade et al., 2002), weather data and initial soil nitrogen concentrations were provided to the model, as well as, the genetic coefficients, previously obtained. Maize growth was simulated and compared with measured values, for two study years.

RESULTS AND DISCUSSION In 1999 Ceres-maize overestimated water storage in the two layers of soil profile (Figure 1A). This is an indication that some adjustments need to be made on soil upper and lower limits of available water. In year 2000, the model simulated very well soil-water storage up to 100 days after planting, underestimating it after that (Figure 1B). It seems that the model is not simulating properly soil-water fluxes, especially capillary rise, since deviations were larger for layer 0 to 110 cm as compared to 0 to 50 cm. In spite of deep large water table in most parts of the “Cerrado”(Savana) region of central Brazil, capillary rise in a typical Oxisols can be significant when there is a water shortage after a rainy season or after cutting irrigation (Andrade et al, 1988). It was not observed differences on soil-water storage for the two soil management systems. 450

1999

Storage 0-110 cm

450

A

2000

Storage 0-110 cm

350 300 250

Storage 0-50 cm 200 150

350 300 250

Storage 0-50 cm

200 150

0

20

40

No-Till Obs

60

Dis Plow Sim

80

100

120

Disk Plow Obs

140

160

100 180

0

No-Till Sim

No-Till Obs

20

60

40

Days After Planting (day)

34

54

73

84

95

1999 115

C

120

Disk Plow Obs

140

160

2000

Days After Planting (day) 0

133

180

10

20

30

40

50

60

70

80

90

D

100 110 120

80

0

80

70

20

70

40

60

40

60

60

50

60

50

80

40

80

40

100

30

100

30

120

20

120

20

140

10

140

10

160

0

160

0

16

34

54

73

84

95

Days After Planting (day)

115

133

Precipitation (mm)

0 20

Irrigation (mm)

Precipitation (mm)

16

100

Days After Planting (day)

Days After Planting (day) 0

Dis Plow Sim

80

Irrigation (mm)

No-Till Sim

100

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

B

400 Soil-Water Storage (mm)

Soil-Water Storage (mm)

400

0 0

10

20

30

40

50

60

70

80

90

100 110 120

Days After Planting (day)

Figure 1. Soil-water storage, precipitation and irrigation along maize cycle, grown under notillage and disk plow soil management systems, for two years.

306

Camilo de Lelis Teixeira de Andrade, Ramon Costa Alvarenga et al.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Although rainfall amounts were large in 1999, a bad irrigation management during dry spells could not properly supply crop water requirement (Figure 1A). One can note that soilwater storage in layer 0 to 50 cm of soil profile, where most of maize active rooting system is concentrated, dropped to very low values, making the crop suffer a strong water stress in both no-tillage and disk plow soil management systems. This is one of the reasons for the low biomass values and leaf area indices (LAI) observed in this year (Figure 2A e 2B), what lead to low grain yields (Figure 3).

Figure 2. Measured and estimated biomass (A) and leaf area index - LAI (B) for maize, grown under no-tillage and disk plow soil management systems, for two years.

Simulating Plant Growth, Soil-Water and Nitrogen Dynamics in Maize Crop…

307

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Ceres-maize simulated reasonably well biomass production for both soil management systems, in 2000 (Figure 2A). In 1999, the model underestimated biomass production, especially during intermediate crop stages. It is possible that the model is penalizing too much maize biomass production, as a consequence of water stress the crop suffered in that year (Figure 1A). Leaf area index was simulated well by the model for the disk plow soil management system, in both years. However, LAI was underestimated in the two years for no-tillage system (Figure 2B). Ceres-maize overestimated grain yield for both soil management systems in the two years, indicating that some other factor, not accounted for in the model affected maize crop. Yield varied from 3.5 to 7.0 tons of grain dry matter per hectar (Figure 3). Yield differences can be due a combination of factors like crop diseases, nitrogen leaching, water stress and plant population that was 52 thousands per hectar in 1999 and of 76 thousands per hectar in 2000.

Figure 3. Estimated and measured grain yield for maize, grown under no-tillage and disk plow soil management systems, for two years.

Leaves nitrogen concentration was simulated reasonably well by the model for both soil management systems, in 2000. In 1999, the model overestimated N concentration at the beginning of the cycle and underestimated it by 80 to 100 days after planting (Figure 4A). Nitrogen extracted by crop tops was very well simulated by the model in 1999, but it was overestimated in 2000 (Figure 4B). When the model simulated well biomass production (Figure 2A), it overestimated crop tops nitrogen extraction (Figure 4B), indicating that Ceresmaize is not properly simulating nitrogen concentration of the various maize plant components. It will be necessary some adjustments on model algorithm parameters related to the plant photoassimilates partitioning.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

308

Camilo de Lelis Teixeira de Andrade, Ramon Costa Alvarenga et al.

Figure 4. Measured and estimated leaf nitrogen concentration (A) and plant tops nitrogen extraction (B) for maize crop, grown under no-tillage and disk plow soil management systems, for two years.

CONCLUSIONS An overall analysis of results indicates that some kind of calibration refinement is still necessary in order to use Ceres-maize model at a farm level. However, as a tool for climate risk assessment, crop zoning and yield forecast, in a regional scale, it is accurate enough to be used.

Simulating Plant Growth, Soil-Water and Nitrogen Dynamics in Maize Crop…

309

REFERENCES

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Andrade, C. L. T., Sediyama, G. C., Sans, L. M. A., Ferreira, P. A. Balanço hídrico num latossolo vermelho-escuro alico cultivado com milho com irrigação por aspersão. Revista Ceres, v.35, p.89 - 104, 1988. Andrade, C. L. T.; Alvarenga, R. C.; Freitas, K. E. D. Impacto do manejo em alguns atributos e na dinâmica da água no solo. In: Congresso Nacional de Milho e Sorgo, 24, 2002, Florianópolis, SC. CD... Boote, K. J.; Jones, J. W.; Pickering, N. B. Potential uses and limitations of crop models. Agronomy Journal, 88:704-716, 1996. Freitas, P. S. L.; Leite, C. A. M.; Mantovani, E. C.; Mantovani, B. H. M. Análise econômica de lâminas de água e doses de nitrogênio para a cultura de milho utilizando o modelo Ceres-Maize. In: Congresso Brasileiro de Engenharia Agrícola, 27, 1998, Poços de Caldas, MG. Anais... Garrison, M. V.; Batchelor, W. D.; Kanwar, R. S.; Ritchie, J. T. Evaluation of the CERESMaize water and nitrogen balances under tile-drained conditions. Agricultural Systems, 62(3):189-200, 1999. Jones, J. W.; Tsuji, G. Y.; Hoogenboom, G.; Hunt, L. A.; Thorthon, P. K.; Wilkens, P. W.; Imamura, D. T.; Bowen. W. T.; Singh, U. Decision support system for agrotechnology transfer: DSSAT v3. In: Tsuji, G. Y.; Hoogenboom, G.; Thorton, P. K. Understanding options for agricultural production, Kluwer Acad. Pub., 1998. Pang, X. P.; Gupta, S. C.; Moncrief, J. F.; Rosen, C. J.; Cheng, H. H. Evaluation of nitrate leaching potential in Minnesota glacial outwash soils using the CERES-Maize model. J. Environ. Qual, 27:75-85, 1998. Sans, L. M. A.; Mantovani, B. H. M.; Baethgen, W. Calibração do modelo de simulação Ceres-Maize para as condições do Brasil central. In: Congresso Nacional de Milho e Sorgo, 20, 1994, Goiania, GO. Anais. Thorton, P. K.; Baanante, C. A.; Singh. U. Uso de modelos de simulacion en la evaluacion de sistemas productivos sustentables. In: Sustentabilidad de las rotaciones cultivo-pastura en el cono sur, 1991, Montevideo, Uruguay. Anais.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 23

POTENTIAL IMPACT OF CLIMATE CHANGE ON AGRICULTURAL SOILS SIMULATED BY ROTH-C MODEL Elena Charro∗1 and Amelia Moyano2 1

Dpto. Ciencias Agroforestales, Escuela Técnica Superior de Ingenierías Agrarias. Universidad de Valladolid. Avda. de Madrid. 34004. Palencia 2 Dpto. Producción Vegetal y Recursos Forestales, Escuela Técnica de Ingenierías Agrarias. Campus Duques de Soria, Universidad de Valladolid. 42004. Soria

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The recent agreement at Bali, based on the Kyoto Protocol, has set targets to reduce the emissions of greenhouse gases to the atmosphere. One method to stabilise atmospheric CO2 concentrations is to sequester carbon in terrestrial ecosystems. Soil conditions may have important effects on the size of the organic carbon pool in the soil. On one side, given the increasing levels of atmospheric CO2 and the continous global warming, there is an urgent need to assess the feasibility of managing ecosystems to sequester and store carbon. On the other side, given that the temperatures are increasing, the turnover of soil organic matter is higher and this feature cause a reduction of soil organic carbon (SOC). Indeed, land management in agriculture can be a source of greenhouse gas emissions: reduce the tillage and the fertilizer-N applications are some of the activities to consider. But there are also some agricultural practises that can help to store carbon in soil, as organic fertilization. In the present work, model simulations were made of the changes in SOC over 100 yr and 40 yr under three different scenarious in order to examine the effects of changes on SOC in agricultural soils in North-Central Spanish plateau (Soria province). The RothC model is also used here to estimate the impact of the global warming on the agricultural SOC. Results show that soils that have been abandoned suffer a loss of storaged carbon, and also the interest of promoting organic agriculture to avoid the degradation of the soils. On the other hand, the increase ∗

[email protected]

312

Elena Charro and Amelia Moyano of the temperature due, of climate change, was predicted to reduce SOC concentrations around 5% in only 40 years.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

INTRODUCTION Soil is important for sequestering atmospheric CO2. Soil conditions may have important effects on the size of the organic carbon pool in the soil. Soil organic matter is a key component of any terrestrial ecosystem, and any variation in its abundance and composition has important effects on many of the processes that occur within that system (Batjes 1996, 1999). With the intense focus on the increasing levels of atmospheric CO2 and its relevance for global climate change, there is an urgent need to assess the feasibility of managing ecosystems to sequester and store carbon. The forest ecosystem is one of several ecosystems being considered for storing additional carbon, since it is unique in having large above- and belowground stocks of carbon (Johnson and Kern 2002). The Kyoto Protocol is an international agreement under which developed countries agreed to reduce their overall greenhouse gas emissions to at least 5% below 1990 levels in the commitment period 2008–2012. Article 3 of the Kyoto Protocol provides strategies for reducing emissions and removing greenhouse gases using carbon sinks, in response to direct human-induced land use changes and forestry activities. The Kyoto Protocol calls for net greenhouse gas emissions and carbon sequestration to be measured as changes in carbon stocks, and also calls for establishing carbon stock baselines. Consequently, the magnitude of changes in above- and belowground carbon pools needs to be assessed (Johnson and Kern 2002). Accurate evaluation of the carbon pools and changes in them due to land use change have been discussed actively since the Kyoto Protocol was formulated (Watson et al. 2000, Lal et al. 2001, Kimble et al. 2002). In order to estimate the change in the carbon stocks of soils, it is first necessary to establish a baseline for carbon stocks. The size of the stock of soil organic carbon (SOC) at regional, national, or global levels is an essential information for discussing changes in carbon content or fluxes at each scale. However, problems arising from soil sampling, soil variability, and soil depth make it difficult to estimate the soil carbon pool (Swift, 2001). Most estimates of SOC stock are based on extrapolations of the mean soil carbon content for broad categories of soil or vegetation types (Post et al. 1982, Sombroek et al. 1993 and Kern 1994). While significant uncertainties exist with respect to both, the estimates of the mean SOC content and the estimates of area for each category (Davidson and Lefebvre 1993), regional studies are necessary to refine global estimations obtained by the aggregation of regional estimates, mainly at a country scale (Bernoux et al. 2002). Furthermore, reliable national estimates of organic carbon stocks are needed for international acceptance (Watson et al. 2000). Forest and agricultural ecosystems can store carbon (C) in soil. However, the lack of data and the few long term experiments make it difficult to estimate the potential of land use to sequester C. Testing and validating the existing and widely used models of organic matter in particular cases will help to determine the potential for C storage in a specific area. Moreover, testing the sensitivity of the models to different land use situations will determine the usefulness of the models for managing land.

Potential Impact of Climate Change on Agricultural Soils…

313

Smith et al. (1997), by testing the performance of nine soil organic matter models, using several sets of data, found that RothC was among the best. In this work, we selected the RothC model (Coleman and Jenkinson, 1996) to test their ability to simulate the accretion of organic matter in arable soils. The RothC model, originally developed and parameterised to model turnover of organic carbon in arable surface soils from Rothamsted long-term field experiments (Jenkinson 1990), has been applied to predict potential response of soil carbon storage to changes in climate and atmospheric CO2 (Jenkinson et al 1991, Polglase and Wang 1992, King et al 1997), and also it has been used worldwide in various ecosystems and climates. RothC is solely concerned with soil processes and does not contain a submodel for plant production, as does other models, like Century (Parton et al 1987). Roth C model’s main advantage is that it runs on data that are readily available (Smith et al 1997). Roth C and Century are two of the most widely used soil organic matter models (Kamoni et al 2007). Althougth these two models were in the origin parameterisated for UK and USA respectively, they have been recently used in environmental conditions so different as Africa (Kamoni et al 2007), South-America (Cerri et al 2003, 2007), or Mediterranean latitudes as Spain (Romanya et al 2000). The present work is the first done in the central Spanish plateau, and one of the few ones done in Mediterranean soils.

MATERIALS AND METHODS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Description of the Study Area We conducted our study in arable topsoils from “La Estacada” Field Experiment located in San Esteban de Gormaz in the Soria province, in north central Spain. The experimental site is situated at an altitude of 860m, and between 3º 12´ W and 41º 34´ N. The area has a continental semi-arid climate, with an average annual precipitation of 38 mm, and the minimum and maximum mean annual temperature of 2.9 and 20ºC, respectively. Mineral layers of the soil were sampled in 2005 to a depth of 35 cm, being the average bulk density of 1.3 g/cm3. Chemical analysis was carried out on air dried, 2mm sieved samples. Organic C was analysed using the Walkley and Black (1934) method. According to the Soil Taxonomy (USDA 1999), the soils are classified as Haploxerlf calcic and with a sequence Ap, Bt and Ck. Other characteristics of this mineral soil important for the modelling procedure are the percentage of clay (28.47%), as well as pH (8.15). The Total Organic Carbon (TOC) site is 0.34 tC/ha. In general, these soils are poor in organic matter, that is usual in the Spanish plateau of central Spain, given that these soils have been used for agriculture since long time ago. The site was fallout during 4 years before starting the trial. Manure treatments started in 1996. From 1996 to 2006 the experiment had nine treatments (combination of organic agriculture, chemistry agriculture and no-treatment) comprising three crop rotations (cereal/legume rotation) and three fertility managements (NPK-fertilization, no-fertilization and organic fertilization). The organic annual addition had been 2500 kg/ha, which is an important factor to take into account for modelling.

314

Elena Charro and Amelia Moyano

Modelling RothC-26.3 (Jenkinson 1990, Coleman and Jenkinson 1995, 1996) is written using Fortran77 language. It is a multicompartmental model and was originally parametrized for temperate agricultural soils, but is nowaways one of the most popular models used both in agricultural and forest soils. The Rothamsted model describes the turnover of organic carbon in soil. It uses monthly input data and is sensitive to soil type, temperature, moisture and plant cover. RothC uses a monthly time step with monthly maximum and minimum temperatures and monthly precipitation data that modify decomposition of each organic matter compartment. For RothC simulations we used a monthly series of meteorological data from the meteorological station of the province where the experimental field is located, and have been calculated according the data registered in the last 30 years. In this model, soil organic C is split into four active fractions and one small inert organic matter (IOM) fraction. The four active fractions are decomposable plant material (DPM), resistant plant material (RPM), microbial biomass carbon (BIO), and humified organic matter (HUM). Each fraction decomposes by a first order processes with its own characteristic rate. The IOM fraction is resistant to decomposition.

RESULTS AND DISCUSSION Modelling Soil Carbon Dynamics Before fitting the model to the data from the arable site, it was necessary to run RothC to generate the C content in the soil at the starting point in 1996. At that time, the organic carbon content of the forest soil was assumed to be in steady state. To do this, the program needs the IOM and the annual input of C. The IOM content is usually very small and can be known using radiocarbon techniques. But when these data are no available, it is possible to use the equation proposed by Falloon et al. (1998): IOM = 0.049 TOC1.139

(1)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

where IOM and TOC are in units of tC/ha. Since the radiocarbon age of the soil had not been measured on any of the forest soils, IOM can be calculated from equation (1), and its value is 0.014 tC/ha. The percentage on this fraction of TOC has also been estimated, being this of 4% of the organic C in the 0-35 cm layer. The percentage of IOM into the total C is similar to those 6.6% and 8%, evaluated and measured, respectively by other authors (Cerri et al. 2003) in their works. On the other hand, Falloon et al. (2000) conclude in their work that the error estimating the size of the IOM pool is less than that introduced by those methods previously used for estimating regional C sequestration potential. The input of farmyard manure put on the soil is treated by the model in a different way than the inputs of fresh plant residues. Given the percentage of organic carbon in the compost,

Potential Impact of Climate Change on Agricultural Soils…

315

a content of 0.6425 tC/ha is put in February on the arable soils under organic fertilization treatment. The annual input of C (from plant debris, roots, etc.) is also needed to run the program. The input of plant residues is also taken into account in these soils under organic fertilization treatment. However, the soils with no treatment has this contribution only. This is a very difficult parameter to evaluate given that many factors that have to be considered, as the production and decomposition of the plant residues, and it has to be considered different distributed monthly (because it has an influence on the activity of the roots), mainly. The annual input was iteratively adjusted. Jenkinson et al (1992) specified the value of DPM/RPM ratio for agricultural crops as 1.44. The RothC model has only been used to predict turnover of carbon in the first 35 cm of the mineral horizon of the soil. Although this is where SOC is most concentrated and subject to change, it is possible that deeper soil horizons have the capacity to sequester carbon too. Indeed, carbon at this depth is older than that near the surface, indicating that it has a greater resistance to decomposition or that the environment at depth is less favourable for microbial decomposition processes (Swift 2001).

Effect of Organic Fertilization

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In order to compare how the organic carbon in the soil varies with different treatments, a simulation is done for the next 100 years. For this period, the meteorological conditions are constant along the years. In Figures 1 and 2, the simulation is done for the soil under three treatments (no-fertilization, fallow and organic fertilization) for the period ranged from 1997 to 2096.

Figure 1. Changes in soil organic C modeled using RothC for the soil under fallow (x), soil under nofertilization with (.) and without (+) 1 year of fallow every 3 years with crops during a century (19962096).

316

Elena Charro and Amelia Moyano

Figure 2. Changes in soil organic C modeled using RothC for the soil under organic-fertilization with(-) and without (.) 1 year of fallow every 3 years with crops during a century (1996-2096).

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

One of the treatment is that where the soil has no-fertilization but it is receiving the input of the residues of plants. This soil presents the same C content during all the years, in other words, the C stock is kept constant. When the organic fertilization is applied on the soil, this soil increases its C content. However, for the soil under fallow during all the period simulated, a decrease of organic C is observed. Furthermore, the lost of C experimented by the soil under this condition is approximately of a 50% in 40 years, which is comparable with the experimental result in the INRA, at Versailles in France. On the other hand, if the inclusion of one year fallow for the soil after 3 years with crops is considered, the simulation shows the following: every year of no crop on the soil is equivalent to loss organic C content for all the scenarios (Figures 1 and 2). In Table 1, data for the total organic C predicted using RothC is showed. The soil under fallow is going to lose around 85% of the initial C during one century. Table 1. Total organic carbon in the soil under different treatments predicted at the end of 100 years (2096)

Treatment Fallow No-fertilization No-fertilization + Fallow Organic-fertilization Organic-fertilization + Fallow

C-total (tC/ha) 0.0518 0.34 0.2596 6.3137 4.6072

Potential Impact of Climate Change on Agricultural Soils…

317

If the rotation of one year fallow is included in the soil under no-fertilization or in the soil under organic-treatment, the organic matter loss is 27%, in both cases, at the end of the century. Finally, the slopes of the curves in Figures 1 and 2 at the beginning of the simulation and at the end of it, are the rates of C accumulation at that year, and is evident how the system is getting the equilibrium, given that the C accumulation rate is decreasing with time.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Changes in soil organic C modeled using RothC for the soil under: (.) fallow, (x) fallow and the effect of the Global Climate Change, (*) no-fertilization, and (+) no-fertilization and the effect of Global Climate Change, during 40 years.

Figure 4. Changes in soil organic C modeled using RothC for the soil under organic-fertilization considering Global Climate Change (diamond) or not (square), during 40 years.

318

Elena Charro and Amelia Moyano

Effect of Climate Change In order to study the effect of the Global Climate Change (GCC) (IPCC 2002), new simulations have been done. Now, temperature, rainfall and evapotranspiration are calculated for every 10 years taking into account the prediction for the Soria province at 2100 (Smith 2004). In Figures 3 and 4 the simulation along 40 years from 1996 is showed. For all the soils, independent of the treatment on it, a lost of organic matter is observed. Quantitatively, Table 2 collects the predicted C content in 40 years. When the GCC is considered, 3.4 % is lost in a soil under permanent fallow, 5.1% in a soil under no-treatment and 5.8% in the soil under organic fertilization. This feature is based on the greater mineralization of SOC as temperatures increase. Table 2. Total organic carbon in the soil under different treatments and taking into account the Global Climate Change (GCC) predicted at the end of 40 years (2036)

Treatment Fallow Fallow + GCC No-fertilization No-fertilization + GCC Organic-fertilization Organic-fertilization + GCC

C-total (tC/ha) 0.1399 0.1351 0.34 0.3206 4.2669 4.0178

CONCLUSIONS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

RothC model was shown to be a useful tool for predicting changes in soil C stocks under different agricultural treatments. The soils that are fallow during a period of time are going to lose organic C, and in consequence, they are going to be degradated. Also, the soils with no fertilization are going to lose organic C given that the climate is changing. However, the soils receiving organic manure are going to increase their organic carbon, even if the Global Climate change is considered. Nevertheless, all the rotations that include one year of soil under fallow are going to have a negative effect in the storage of the C.

ACKNOWLEDGMENTS This work was supported by the Spanish Junta de Castilla y León (Projects VA094A06 and VA014A07) and Ministerio de Educación y Ciencia (Project CTM2006-02249/TECNO).

REFERENCES Batjes N.H. (1996) Total carbon and nitrogen in the soils of the world. Eur. J. Soil Sci. 47: 151–163.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Potential Impact of Climate Change on Agricultural Soils…

319

Batjes, N.H. (1999). Management options for reducing CO2- concentrations in the atmosphere by increasing carbon sequestration in the soil. ISRIC. Wageningen, The Netherlands. 114 pp. Bernoux M., Carvalho M.C.S., Volkoff B., Cerri C.C. (2002) Brazil's soil carbon stock. Soil Sci. Soc. Am. J. 66: 888–896. Cerri C.E.P., Coleman K., Jenkinson D.S., Bernoux M., Victoria R., Cerri C.C. (2003) Modelling Soil carbon from forest and pasture ecosystems of Amazon, Brazil. Soil Science Soc. Am. J. 67: 1879-1887. Cerri C.E.P., Easter M., Paustian K., Killian K., Coleman K., Bernoux M., Falloon P., Powlson D.S., Batjes N., Milne E., Cerri C.C. (2007) Simulating SOC changes in 11 land use change chronosequences from the Brazilian Amazon with RothC and Century models. Agriculture, Ecosystems and Environment 122 : 46-57. Coleman K., Jenkinson D.S. (1995) RothC-26.3 – a model for the turnover of carbon in soil: Model description and users guide. Lawes Agricultural Trust, Harpenden. Coleman K., Jenkinson D.S. (1996) RothC-26.3 – a model for the turnover of carbon in soil: In: Powlson D.S., Smith P., Smith J.U. (Eds.), Evaluation of soil Organic Matter Models using Existing, Long-Term Datasets, NATO ASI Series I, Vol. 38 Springer-Verlag, Berling, pp.237-246. Davidson E.A., Lefebvre P.A. (1993) Estimating regional carbon stock and spatially covarying edaphic factors using soil maps at three scales. Biogeochemistry 22:107–131. Falloon P., Smith P., Coleman K., Marshall S. (1998) Estimating the size of the inert organic matter pool for use in the Rothamsted carbon model. Soil Biology and Biochemistry 30, 1207-1211. Falloon P., Smith P., Coleman K., Marshall S. (2000) How important is inert organic matter for predictive soil carbon modelling using the Rothamsted carbon model?. Soil Biology and Bichemistry 32: 433-436. Jenkinson D.S. (1990) The turnover of organic carbon and nitrogen in soil. Philosophical transactions of the Royal Society, B. 329: 361-368. Jenkinson D.S., Adams D.E., Wild A. (1991) Model estimates of CO2 emissions from soil in response to global warming. Nature 351: 304-306. Jenkinson D.S., Harkness D.D., Vanee E.D., Adams D.E., Harrison A.F. (1992) Calculating net primary production ansd annual input of organic matter to soil from the amount and radiocarbon content of soil organic matter. Soil Biol. Biochem. 24: 295-308. Johnson M.G., Kern J.S. (2002) Quantifying the organic carbon held in forested soils of the United States and Puerto Rico. In: J.M. Kimble, L.S. Heath, R.A. Birdsey and R. Lal, Editors. The Potential of U.S. Forest Soils to Sequester Carbon and Mitigate the Greenhouse Effect, Lewis Publishers, Boca Raton, FL, pp. 47–72. Kamoni P.T., Gicheru P.T., Wokabi S.M., Easter M., Milne E., Coleman K., Falloon P., Paustian K., Killian K., Kihanda F.H. (2007) Evaluation of two soil carbon models using two Kenyan long term experimental datsets. Agriculture, Ecosystems and Environment 122: 95-104. Kern J.S., (1994) Spatial patterns of soil organic carbon in the contiguous United States. Soil Sci. Soc. Am. J. 58: 439–455. Kimble J.M., Heath L.S., Birdsey R.A., Lal R. (2002) The Potential of U.S. Forest Soils to Sequester Carbon and Mitigate the Greenhouse Effect. , Lewis Publishers, Boca Raton, FL.

320

Elena Charro and Amelia Moyano

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

King A.W., Post W.M., Wullscheleger S.D. (1997) The potential response of terrestrial carbon storage to changes in climate and atmospheric CO2. Climatic Change 35: 199227. Lal P., Kimble J.M., Follet R.F., Stewart B.A. (2001) Assessment Methods for Soil Carbon. Lewis Publishers, Boca Raton, FL. Parton W.J., Schimel D.S., Cole C.V., Ojima D.S. (1987) Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Sci. Soc. Am. J. 51: 1173-1179. Polglase P.J.,Wang Y.P. (1992) Potential CO2-enhanced carbon storage by the terrestrial biosphere. Australian Journal of Botany 40: 641-656. Post W.M., Emanuel W.R., Zinke P.J., Stangenberger A.G., (1982) Soil carbon pools and world life zones. Nature 298 8: 156–159. Romanyá J.; Cortina, J.; Fallon, P.; Coleman, K., Smith, P. (2000): Modelling soil organic matter after planting fast-growing Pinus radiata on Mediterranean agricultural soils. Eur. J. Soil Sci. 51:627-641. Smith P. 2004. Private communication. Smith P., Smith J. U., Powlson D. S., McGill W. B., Arah J. R. M., Chertov O. G., Coleman K., Franko U., Frolking S., Jenkinson D. S., Jensen L. S., Kelly R. H., Klein-Gunnewiek H., Komarov A. S., Li C., Molina J. A. E., Mueller T., Parton W. J., Thornley J. H. M., Whitmore A. P. (1997) A comparison of the performance of nine soil organic matter models using datasets from seven long-term experiments. Geoderma 81 : 153-225. Sombroek W.G., Nachtergaele F.O., Hebel A., (1993) Amounts, dynamics and sequestering of carbon in tropical and subtropical soils. AMBIO 22: 417–426. Swift R.S. (2001) Sequestration of carbon by soil. Soil Sci. 166: 858–871. USDA 1999. Soil Taxonomy. A basic system of soil Classification for making and interpreting soil surveys. In: Agriculture Handbook 436. Walkley A., Black I.A., (1934) An examination of degtjareff method for determining soil organic acid matter and a proposed chromic acid titration method. Soil Sci. 37: 29-38. Watson R.T., Noble I.R., Bolin B., Ravindranath N.H., Verardo D.J., Dokken D.J. (2000) Land Use, Land-use Change, and Forestry. Cambridge Univ. Press, Cambridge

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 24

SIMULATION OF CROP YIELD NEAR HEDGEROWS UNDER ASPECTS OF A CHANGING CLIMATE: AN AUSTRIAN ATTEMPT Thomas Gerersdorfer∗, Josef Eitzinger and Pablo Rischbeck Institute of Meteorology; Department of Water, Atmosphere and Environment; University of Natural Resources and Applied Life Sciences, Vienna (BOKU)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT Hedgerows play an important role for agricultural production as their impacts on field crops are manifold (Cleugh, 1998; Mayus et al., 1999; McAneney et al., 1990). One effect is the modification of the microclimate of neighbouring fields, especially caused by the well known effect of wind speed reduction. In semiarid regions the effect on evapotranspiration is crucial and can have a significant effect on crop water balance, drought damage and crop yields. As hedgerows are increasingly introduced in ecological farming systems in Austria, an interdisciplinary project investigates the specific microclimatological effects of a selected hedgerow on field evapotranspiration and other water balance parameters in the semiarid region of North-East Austria. In the view of climate change and increasing potential evapotranspiration rates through global warming and extreme hot weather conditions as already observed during the past decades, a comparison of observed results with simulated yields of crop is undertaken by means of DSSAT. Furthermore varying (micro-)climatic parameters are used to simulate possible impacts of climate change on crop yields in this water-limited semiarid region.

∗

Peter Jordan Str. 82, A-1190 Vienna; [email protected]

322

Thomas Gerersdorfer, Josef Eitzinger and Pablo Rischbeck

INTRODUCTION

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The site of the measurements (latitude 48°12'N, longitude 16°34'E and altitude 150 m above sea level) is located in Marchfeld, within a main agricultural crop production region eastern of Vienna. This region is characterized by flat area and strong winds. The hedgerow (Figure 1) is relatively dense and consists of shrubs and trees with a mean height of 8 m and a width of 6 m. The orientation of the hedgerow is nearly from North to South, the main wind direction is from north-west. Transect measurements of microclimatic parameters near a hedgerow were carried out during the vegetation periods of 2003, 2004 and 2005 on a crop field. The continuous measurements were undertaken by automatic stations. Horizontal-vertical wind profiles at two sites and horizontal profiles of precipitation, global radiation, dew occurrence and evapotranspiration were measured. Beside net radiation, air temperature and air humidity, evapotranspiration was measured directly by evaporimeters at canopy height of 0.8 meters (potential ET).

Figure 1. Location of the transect measurements on the east side of the hedgerow.

METHOD This work uses the DSSAT Software (Decision Support System for Agrotechnology Transfer, Version 4) to simulate impacts of a hedgerow on field crops. Measurements (continuous and transect measurements) were carried out by the Institute of Meteorology (BOKU-Met), crop yield data were provided by the Division of Organic Farming (BOKU).

Simulation of Crop Yield Near Hedgerows under Aspects of a Changing Climate

323

RESULTS Yield data of wheat show a decrease with increasing distance from the hedgerow (Figure 2 and Table 1). This effect reaches to a distance which is about the 10fold of the height of the hedgerow (80 m) and strongly depends on the wind direction and location of the measurements (west or east side of the hedgerow, luv or leeward).

Figure 2. Crop yields in the year 2005 at the measurement site, east side of the hedgerow (Division of Organic Farming, BOKU).

Table 1. Averages of crop yields in the year 2005 at the measurement site (Division of Organic Farming BOKU)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

distance [m]

yield dm [kg/m²] yield dm [kg/ha] averaged averaged

8m

0.322

3220

16 m

0.273

2730

24 m

0.250

2500

40 m

0.227

2270

80 m

0.227

2270

The shape of the hedgerow shows a complex impact on the wind field near the hedgerow. A reduction in horizontal wind speed could be observed till approximately 80 m distance leeward from the hedgerow where wind speed was strongly reduced. Wind speed in luv was

324

Thomas Gerersdorfer, Josef Eitzinger and Pablo Rischbeck

reduced significantly less. The reduction of wind speed on the east side of the hedgerow (leeward) at a distance of 8 m was at least 50 % or even more compared to the wind speed in 80 m distance from the hedgerow at the same side. Snowfalls and prevailing westerly winds at the beginning of March 2005 resulted in snow banks. The melting water equivalent of about 158 mm was averaged for a distance of 12 m to the hedgerow (Figure 3). The simulations show the positive effect nearby the hedgerow on soil water content for about 2 months and crop yield, respectively.

snow cover on 10 March 2005 90 80 70 60 50 [cm] 40 30 20 10

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0

hedge northern

southern

Figure 3. Snow banks in March 2005 on the east side of the hedgerow at the measurement site and snow profiles.

SIMULATIONS Several simulations were conducted to investigate the impact of wind and/or precipitation on crop yield and soil water content at different distances from the hedgerow. The impacts on soil water content are presented in Figure 4 to Figure 8. Simulation S1 represents the “open field conditions” at a distance of 80 m of the hedgerow where the influence of the hedgerow

Simulation of Crop Yield Near Hedgerows under Aspects of a Changing Climate

325

is nearly negligible whereas the other simulations (S2, S3, S4, S5) consider a reduction in wind speed close by the hedgerow and/or additional “precipitation” due to melting of snow banks. Layer 1 corresponds to the soil layer 0-5 cm, layer 2 to 5-15 cm and layer 3 to 15-30 cm soil depth. 0.6

0.5

soil water

0.4

0.3

0.2

0.1

0 50

100

150

200

250

300

350

Days after Start of Simulation

S1

SWC layer 1 (CAPO/kal)

SWC layer 2 (CAPO/kal)

SWC layer 3 (CAPO/kal)

Figure 4. Simulation S1: open field conditions, unaffected by the hedgerow.

0.6

0.5

soil water

0.4

0.3

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0.2

0.1

0 50

S2

100

150

200

250

300

Days after Start of Simulation SWC layer 1 (CAPO/kal)

SWC layer 2 (CAPO/kal)

SWC layer 3 (CAPO/kal)

Figure 5. Simulation S2: reduction of wind speed of 50 % caused by the shading effect of the hedgerow.

350

326

Thomas Gerersdorfer, Josef Eitzinger and Pablo Rischbeck 0.6

0.5

soil water

0.4

0.3

0.2

0.1

0 50

100

S3

150

200

250

300

350

Days after Start of Simulation SWC layer 1 (CAPO/kal)

SWC layer 2 (CAPO/kal)

SWC layer 3 (CAPO/kal)

Figure 6. Simulation S3: reduction of wind speed of 75 % caused by the shading effect of the hedgerow.

0.6

0.5

soil water [%]

0.4

0.3

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

0.2

0.1

0 50

100

S4

150

200

250

300

350

Days after Start of Simulation SWC layer 1 (CAPO/kal)

SWC layer 2 (CAPO/kal)

SWC layer 3 (CAPO/kal)

Figure 7. Simulation S4, regarding melting of snow banks nearby the hedgerow.

A survey of the simulations is given in Table 2. Simulated yield (S1, 2193 kg/ha) at a distance of 80m shows little difference to actual yield (2270 kg/ha). Regarding the maximum of actual yield at the distance of 8 m (3220 kg/ha) the simulations show a range from 2983

Simulation of Crop Yield Near Hedgerows under Aspects of a Changing Climate

327

kg/ha to 3653 kg/ha depending on the reduction of wind speed and snow melting respectively. Simulation S4, considering only melting of snow banks is close to actual yield (3048 kg/ha versus 3220 ka/ha). The underestimate might be a result of disregarding the positive influence of wind speed reduction by the shading effect of the hedgerow. 0.6

0.5

soil water [%]

0.4

0.3

0.2

0.1

0 50

100

S5

150

200

250

300

350

Days after Start of Simulation SWC layer 1 (CAPO/kal)

SWC layer 2 (CAPO/kal)

SWC layer 3 (CAPO/kal)

Figure 8. Simulation S5, regarding a reduction of wind speed of 50 % and melting of snow banks nearby the hedgerow.

Table 2. Survey of the conducted DSSAT Simulations Simulations

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

distance

conditions

seasonal crop yield "precipitation" [kg/ha] [mm]

seasonal evapotranspiration [mm]

H2O stress

S1

80 m

open field (without any influence of the hedgerow)

2193

348

345

1.56

S2

8 m reduction in wind speed of 50 %

2983

348

348

1.25

S3

8 m reduction in wind speed of 75 %

3653

348

348

0.79

S4

8 m regarding snow melting

3048

498

412

0.91

S5

8m

3054

498

413

0.69

regarding snow melting and reduction in wind speed of 50 %

Additionally, water stress (total sum) was simulated as well for the same period. As expected, water stress shows the highest value for simulation S1. Simulation S5 – taking a (realistic) wind speed reduction of 50 % and melting of snow into account – shows the lowest value. The shading effect of the hedgerow and therefore a reduction of wind speed combined with high soil water contents at the beginning of the vegetation period results in highest

328

Thomas Gerersdorfer, Josef Eitzinger and Pablo Rischbeck

yields. By comparing simulation S4 and S5 in terms of water stress the important - positive shading effect of the hedgerow is obvious.

CONCLUSIONS AND OUTLOOK The simulations of crop yield for 2005 near a hedgerow in Rutzendorf show high similarities to the undertaken measurements at field level in crop yield itself and in its decrease with increasing distance up to about the tenfold height of the hedgerow. The simulated crop yield (2193 kg/ha) at a distance of 80 m from the hedgerow is comparable to the results of the field measurements (2270 kg/ha). Taking the shading effect of the hedgerow and thus a wind speed reduction of 50% into account, similar rates of yield prevail at a distance of 8 m (3220 kg/ha versus simulated 2983 kg/ha). Combined with melting snow banks on a small band at the lee side of the hedgerow – as seen 2005 - crop yield simulation of 3054 kg/ha is very close to the actual yield. Due to reduced wind speed the reduced actual evapotranspiration produces minor (and later) water stress (Table 2). The simulated soil water content is very similar for S1 (field conditions at 80 m distance) and S2 (wind speed reduction of 50 %) but compared with S5 there is a great shift due to melting snow banks in March. Admittedly, the effect of melting snow banks was limited only to a few meters distance from the hedgerow and for a few days but the simulation shows a positive effect on soil water content for about two months. For a simulation of crop yield near hedgerows the soil water balance and the snow cover thickness in winter near the hedgerow play an important role. These effects have to be further considered for an absolute assessment on the impact of the hedgerow. This work is the first attempt to simulate crop yields on the basis of microclimatic transect measurements near a hedgerow in a semi arid region of Austria. Simulation outputs for yield show good results. Nevertheless some parameters (soil layers, soil cultivation) could be simulated in more detail. In the view of climate change and increasing potential evapotranspiration rates through warming, the knowledge on water saving methods in agriculture will become a more and more important issue.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

ACKNOWLEDGMENTS We thank Andreas Surböck and Markus Heinzinger (Division of Organic Farming, BOKU) for providing crop yield data.

REFERENCES Cleugh, H.A., 1998: Effect of windbreaks on airflow, microclimates and crop yields.Agroforestry Systems 41, 55-84. Mayus, M., Van Keulen, H., Stroosnijder, L., 1999: A model of tree-crop competition for windbreak systems in the Sahel: description and evaluation.- Agroforestry Systems 43, 183-201.

Simulation of Crop Yield Near Hedgerows under Aspects of a Changing Climate

329

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

McAneney, K.J., Salinger, M.J., Porteous, A.S., Barber, R.F., 1990: Modification of an orchard climate with increasing shelter-belt height. Agricultural and Forest Meteorology 49, 177-189.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In: Climate Variability, Modeling Tools… Editor: Angel Utset

ISBN: 978-1-60692-703-8 © 2009 Nova Science Publishers, Inc.

Chapter 25

INTRODUCTION OF CROP MODELLING TOOLS INTO SERBIAN CROP PRODUCTION: CALIBRATION AND VALIDATION OF MODELS B. Lalic∗1, D. T. Mihailovic1 and M. Malesevic2 1

Faculty of Agriculture, University of Novi Sad, Novi Sad, Serbia 2 Institute for Field and Vegetable Crops, Novi Sad, Serbia

ABSTRACT Although crop models represent a power tool of modern agriculture, they are commonly used at Serbian agricultural faculties, institutes and advisory services. This work is the first step in an attempt to introduce crop modelling tools into our agricultural research and practice. Using the SIRIUS and PERUN (WOFOST) models, phenology dynamic, biomass and grain yield were calculated for the winter wheat cultivar Anastasia. The results obtained were compared with the observed values for the 2001-2005 vegetation periods.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

1. INTRODUCTION The first attempt to simulate plant response to atmospheric conditions was made by Reamur in the XVIII century (Reamur, 1735). It was an example of regression analysis application in the prediction of growth stage appearance based on accumulated temperatures. Even though this statistical crop modelling technique was tremendously improved, its main limitations related to location/variety linking are still present. Numerous scientists invested large efforts during the last 50 years to develop dynamical, process-oriented models. The first results of the state-of-the-art crop models testing came at the end of the 1980s with models like CERES-Wheat (Ritchie and Otter, 1985), SOYGRO (Jones et al., 1988), SIRIUS (Jamieson et al., 1998), WOFOST (Boogaard et al., 1998; Supit et al., 1994), CROPSYST ∗

[email protected]

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

332

B. Lalic, D. T. Mihailovic and M. Malesevic

(Stöckle et al., 1992), APSIM (Asseng et al., 2000), DSSAT (Jones et al., 2001; Jones et al., 2003) and many others. The years that followed brought many studies devoted to calibration and validation of models in different agroecological conditions and various crops (for example, Alexandrov et al., 2002; Basso et al., 2001; Bloch et al, 1997; Brisson et al., 2006; Donatelli et al., 1997; Eitzinger et al., 2000; Faber et al., 1996; Jamieson et al., 1998b). However, crop model introduction in operational use takes time. Only a limited number of models have been operationally used (WOFOST, STICKS, CERES). However, calibration and application of all other models have been depending on individual enthusiasm or current research “main stream” (healthy food, sustainable development, climate change), at national and international levels, providing conditions and stable founding for further work. In different regions and countries, the application of crop models is related to the (a) economically most important crops and the level of agricultural research theory and (b) practice development. Unfortunately, the application of dynamic crop models in Serbian crop production cannot be treated from this point of view. Until the end of the 1980s, the former Yugoslavia, of which Serbia was a part, was a mid-developed European country with significant contributions in plant genetics, crop breeding and crop management resulting in creation of new varietes and hybrids (for example, Borojević, 1958; Borojević et al., 1980; Kojic and Ivanovic, 1986; Skorić, 1992). However, during the last decade of the XX century, due to sanctions of the UN, Serbian scientists were completely isolated from all sources of information (journals, books, projects, workshops), including those dealing with the latest trends in agriculture research. The rapid opening of the country after the year 2000 enabled Serbian scientists to establish intensive communication and exchange of scientific results with the global scientific community. Regarding the application of crop modelling tools, participation of specialists from the Novi Sad and Belgrade faculties of agriculture and the agrometeorological division of the RHMS of Serbia in the AGRIDEMA (FP6-2003-Global-2-003944) Workshop in Vienna (21st November – 2nd December 2005) was of high importance. Devoted to connecting model developers and users, this workshop offered the participants an opportunity to gain practical knowledge related to crop models use and their application in crop forecasting and assessment of climate change impact on agricultural production. This paper represents an overview of research results of the AGRIDEMA pilot project entitled: “Introducing crop modelling tools into Serbian crop production”, which was carried out at the Faculty of Agriculture of the University of Novi Sad (Serbia) during the year 2006. The main goals of this assessment study were the calibration and validation of the SIRIUS and PERUN (WOFOST) models using results of field experiments carried out with Serbian crop varieties. These goals were achieved through the following activities: a) gathering of crop, weather, soil and management input and output data; b) calibration of both models c) validation of models comparing simulated and observed values of outputs. Section 2 provides a detailed description of the activities and results obtained. Concluding comments and further plans are elaborated in Section 3.

Introduction of Crop Modelling Tools into Serbian Crop Production

333

2. PILOT ASSESSMENT OVERVIEW A) Gathering The Input Data The first and highly important step in running crop models is to provide input data. An overview of weather, soil, crop and management data necessary to run the SIRIUS and WOFOST crop models is presented in Tables 1 and 2. Table 1. Input weather, soil, crop and management data in the SIRIUS crop model (asteriks indicates the thermal time, min. – minimum value, max. – maximum value, RUE – radiation use efficiency, PAR – intensity of photosynthetic active radiation) Weather data min. air temperature max. air temperature

Soil data saturation moisture content

precipitation

lower limit

radiation

percolation coefficient

wind speed

mineralisation constant

potential max.0

vapour pressure

organic N

phyllochron in degree days

*

t_t from sowing to emergence

Management sowing date

*

drained upper limit

min. mineral N content of soil constant for denitrification pulse amount of inorganic N

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Crop data

t_t from anthesis to beginning of grain fill t_t* from beginning to end of grain fill t_t* from end of grain fill to harvest maturity

soil name cultivar RUE or CO2 concentration No. of management applications irrigation date and amount

min. possible leaf number

initial moisture deficit

absolute max. leaf number

amount of inorganic N

daylength response in leaves per hour of daylength response of vernalisation rate to temperature vernalisation rate at 0 oC PAR extinction coefficient max. protein conc. in unlimited growth conditions

proportion of N in the top (33%) and mid (33%) soil fertilization date and amount

Obviously, it is rather hard to satisfy the high demands regarding input parameters, especially WOFOST’s. For the purposes of this pilot assessment, all crop, management and site information for one typical variety of winter wheat (Anastasia) were taken from the field experiments database of the Institute for Field and Vegetable Crops (Novi Sad, Serbia). Experimental data were conducted during vegetation periods on the basis of 10- and/or 20plant samples. For each plant, measurements of root, steam, leaf, tailings and grain mass were performed. On the basis of these data and assuming 550 plants per m2, grain yield and biomass were calculated and expressed as kg ha-1.

334

B. Lalic, D. T. Mihailovic and M. Malesevic Table 2. Input weather, soil, crop and management data in WOFOST crop model

Weather data

Soil data

min. air temp. vol. soil moisture content max. air temp. precipitation radiation wind speed

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

vapour pressure

soil moisture content at wilting point soil moisture content at field capacity soil moisture content at saturation critical soil air content for aeration hydraulic conductivity of saturated soil max. percolation rate root zone max. percolation rate subsoil 1st topsoil seepage parameter deep seedbed 2nd topsoil seepage parameter deep seedbed 1st topsoil seepage parameter shallow seedbed 2nd topsoil seepage parameter shallow seedbed required moisture deficit deep seedbed

Crop data lower threshold temp. for emergence max. eff. temp. for emergence

Management sowing date soil name

temp. sum from sowing to emergence cultivar pre-anthesis development depends on No. of production temp., daylength or both levels type of water limited optimum daylength for development crop growth type and amount of critical daylength fertilizer applied temp. sum from emergence to anthesis temp. sum from anthesis to maturity daily increase in temp. sum as function of average temp. development stage at harvest initial total crop dry weight LAI at emergence max. relative increase in LAI specific leaf, pod. and steam area life span of leaves growing at 35 oC lower threshold temp. for ageing of leaves extinction coeff. for diffuse visible light light-use effic. single leaf efficiency of conversion of assimilates into leaves, storage org., roots and stems relative maintenance respiration rate for leaves, storage org., roots and stems fraction of total dry matter to roots and storage organs death rates water use rooting max. and min. concentrations of N, P, and K in storage and vegetative organs

For validation purposes, the values of biomass, grain mass and grain yield and the date of growing stage appearance for the 2001-2005 vegetation periods were used.

Introduction of Crop Modelling Tools into Serbian Crop Production

335

B) SIRIUS Model Calibration and Validation For the purpose of SIRIUS crop model calibration, the following data were assimilated: exact dates of phenology phases, biomass, and grain mass and grain yield. For each growing season during the 2001-2005 period, dates of the following phenology phases were calculated for the winter wheat Anastazija: emergence, anthesis, start of grain fill, end of grain fill, and maturity. The results were compared with the observed dates of growth stage appearance and presented in Figure 1. Even a short inspection of this figure shows that there is a perfect match between the calculated and observed dates. This fact is a very important one, since accurate forecasting of phenology dynamic can improve efficiency of farm management operations.

Phenology observed (DOY)

350

E - Emergence A - Anthesis B - Begin of grain fill N - End of grain fill M - Maturity

300 250 200 150 100 50 0 0

50

100 150 200 250 300 350

Phenology calculated (DOY)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. Comparison of the growth stage appearance dates calculated using the SIRIUS crop model and those observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Results obtained for winter wheat biomass development (Figure 2) show non significant deviation of calculated vs. observed biomass, but there is a systematic rush-in of simulated values. It could be addressed to, at least two possible sources of the problem emphasised: (1) error in parameterisation of radiation can cause overestimation of PAR and/or (2) overestimation of PAR interception coefficient can cause overestimation of intercepted PAR. Since biomass production in SIRIUS is related to intercepted PAR, one or both errors can produce a rush-in of simulated values. However, deviation of simulated values decreases during the vegetation period and causes a rather small variation from the observed biomass at maturity (Figure 3). In contrast to biomass development, the calculated timing and rate of grain yield development has a relatively good correspondence to the observed values (Figure 4). The only exception is the year 2001, in which, according to SIRIUS, grain formation started 20 days earlier.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

336

B. Lalic, D. T. Mihailovic and M. Malesevic

Figure 2. Comparison of biomass development calculated using SIRIUS crop model and that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Introduction of Crop Modelling Tools into Serbian Crop Production

337

Observed biomass (kg/ha)

20000

15000

10000

5000 5000

10000

15000

Calculated biomass (kg/ha)

20000

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 3. Comparison of biomass at maturity calculated using SIRIUS crop model and that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Figure 4 (Continued).

338

B. Lalic, D. T. Mihailovic and M. Malesevic

Figure 4. Comparison of the grain yield development calculated using SIRIUS crop model and that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Unfortunately, in all cases maturity occurs 10 – 15 days earlier producing a halt in grain yield development and causing simulation of significantly smaller values of yield at maturity than what was observed (Figure 5). Most likely, the causes of this halt in yield development are inaccurate values of parameters describing different phases of maturity.

Figure 5. Comparison of grain yield at maturity calculated using SIRIUS crop model and that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

In contrast to many other models, SIRIUS has the advantage of being able to calculate grain mass at maturity and give information about single plant quality (Figure 6). From Figure 6, it is obvious that only in the case of one year (2004) there was a 10 mg (25%) difference between the calculated and observed grain mass values.

Introduction of Crop Modelling Tools into Serbian Crop Production

339

Figure 6. Comparison of grain mass at maturity calculated using SIRIUS crop model and observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

In all other cases, the difference was remarkably smaller. Since SIRIUS proved superior in forecasting plant phenology dynamics, it could be a useful tool in forecasting optimal sowing date. If we suppose the same timing and amount of management operations, for each year during the analysed period, grain yield and grain mass at maturity can be calculated for different sowing dates (Figures 7a and 7b). A very simple inspection of both figures leads leads to the conclusion that larger grain yield and mass were achieved under the observed conditions if sowing took place during the first decade of October.

a Figure 7. (Continued).

340

B. Lalic, D. T. Mihailovic and M. Malesevic

b Figure 7. Sensitivity of grain yield (a) and grain mass (b) to sowing date calculated using SIRIUS crop model for Anastasia winter wheat during 2001-2005 vegetation periods.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

C) WOFOST Model Calibration and Validation PERUN is a complex model consisting of the MetandRoll weather generator (Dubrovsky, 1996; Dubrovsky, 1997) and WOFOST, a highly sophisticated crop model with numerous outputs. Unfortunately, at Serbian agricultural extension services and research institutes it is not common practice to collect on a regular basis all the quantities comprised by WOFOST outputs. Due to this reason, calibration and further validation of the model outputs were reduced to growth stages, biomass and grain yield. Grain mass is not one of the calculated variables, because WOFOST is a model that focuses on the canopy as a whole instead of on individual plants. Moreover, there are three regimes for which all crop model outputs could be provided: a) potential growth, b) water and O2 limited crop production and c) nitrogenlimited crop production. In this study, biomass and grain yield development were calculated for two extreme cases a) and b). In the phenology dynamic part, WOFOST calculates only emergence and anthesis appearance. From Figure 8 is seen that calculated dates for emergence are accurate. However, there is a systematic deviation between observed and anthesis dates calculated using WOFOST. In conclusion, it seems that WOFOST is not the best choice for phenology forecasting purposes. The results obtained for biomass and grain yield development are shown in Figures 9 and 11. An inspection of these figures indicates significant deviation between the observed and calculated biomass values and almost perfect correspondence in the case of grain yield development for all vegetation periods. Similarity of the results presented in Figures 2 and 9 suggests that either the same parameter for both models is set on an inadequate value or the parameterisation of a variable important for biomass production is inappropriate. Comparison for both regimes of biomass at maturity with the observed data (Figure 10) shows systematic overestimation of calculated values in all cases.

Introduction of Crop Modelling Tools into Serbian Crop Production

341

Phenology calculated (DOY)

350 300 250 200 150 100 100

150

200

250

300

Phenology observed (DOY)

350

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Comparison of anthesis (A) and emergence (E) apperance dates calculated using WOFOST model and those observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Figure 9.(Continued).

342

B. Lalic, D. T. Mihailovic and M. Malesevic

Figure 9. Biomass developments calculated using WOFOST model for Anastasia winter wheat for 2001-2005 vegetation periods.

On the other hand, the simulation of grain development during the growing season is more accurate, although maturity in the model happens much earlier than in reality. Even though it implies overestimation of calculated grain yield at maturity in comparison with the observed values; WOFOST appears as a powerful tool in simulating potential grain yield development.

Observed biomass (kg/ha)

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

25000

20000

15000

10000

5000 5000

10000

15000

20000

Calculated biomass (kg/ha)

25000

Figure 10. Comparison of biomass at maturity calculated using the WOFOST model with that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Introduction of Crop Modelling Tools into Serbian Crop Production

343

Figure 11. Grain yield development calculated using WOFOST model for Anastasia winter wheat for 2001-2005 vegetation periods.

344

B. Lalic, D. T. Mihailovic and M. Malesevic

Observed yield (kg/ha)

10000 9000 8000 7000 6000 5000 5000

6000

7000

8000

9000

Calculated yield (kg/ha)

10000

Figure 12. Comparison between grain yield at maturity calculated using the WOFOST model and that observed for winter wheat Anastazija during 2001-2005 vegetation periods.

CONCLUDING REMARKS

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

The first step in the validation and calibration of crop models for NS varieties of winter wheat was put into the progress. SIRIUS exerted superiority in phenology dynamic forecasting, both models gave good results in grain yield development forecasting, but biomass development was not satisfactory. It is reasonable to assume that the observed rushin of simulated values for both models is caused by an error in the parameterisation of PAR or/and by setting a higher value for the PAR interception coefficient. However, taking into account all the above mentioned results, it seems that the best practical solution is to use SIRIUS for phenology development forecasting while for improving the calculation of grain yield and biomass the good choice are both SIRIUS and WOFOST crop models.. Further work in this field should be organised in two directions: (1) assessment of climate change impact on crop production using crop models and (2) calibration and validation of models for wider spectra of crop varieties and for other agroecological regions in Serbia.

ACKNOWLEDGMENTS This study was supported by the Serbian Ministry of Science under grant OI141035 and two projects from Sixth Framework Programme (FP6): ADAGIO ("Adaptation of agriculture in European regions at environmental risk under climate change"; SSPE-CT-2006-044210) and RRP CMEP (“Reinforcement of the research potential in Center for Meteorology and Environmental Predictions”; INCO-WBC/SSA-3-2005-043670).

Introduction of Crop Modelling Tools into Serbian Crop Production

345

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

REFERENCES Asseng, S., van Keulen, H., Stol, W., 2000: Performance and application of the APSIM Nwheat model in the Netherlands, European Journal of Agronomy, 12, 37- 54. Alexandrov, V., Eitzinger, J., Cajic, V., Oberforster, M., 2002: Potential impact of climate change on selected agricultural crops in north-eastern Austria, Global Change Biology, 8 (4), 372-389. Basso, B., Ritchie, J.T., Pierce, F.J., Braga, R.P., Jones, J.W., 2001: Spatial validation of crop models for precision agriculture, Agricultural Systems, 68, No. 2, 97-112. Błoch, Z., Faber, A., Demidowicz, G., Kamasa, J., 1997: Zastosowanie modelu WOFOST do symulacji wzrostu i plonowania ziemniaków uprawianych w Polsce. I. Weryfikacja modelu (Growth and yield simulations of potato cultivated in Poland using WOFOST – model verification), Frag. Agron., 2 (54), 96-101. Borojević, S., 1958: Smer oplemenjivanja pšenice, (Wheat breading policy), Vojvodina Agriculture, Novi Sad, 6 (3), 187-191. Borojević, S., Ločniškar, F., Avramov, L., Mišić, P., 1980: Dostignuti nivo i dalji pravci razvoja i primene rezultata nauke u poljoprivedi, U: Agroindustrijski kompleks Jugoslavije, (Achieved level and further directions of development and application of scientific results in agriculture, In: Agro industry of Yugoslavia), Privredni pregled, 143176. Boogaard, H.L., van Diepen, C.A., Rötter, R.P., Cabrera, J.M.C.A., van Laar., H.H., 1998: User’s Guide for the WOFOST 7.1 Crop Growth Simulation Model and WOFOST Control Center 1.5. DLO-Winand Staring Centre, Wageningen, Technical Document 52. Brisson N., Corre-Hellou G., Dibet A., Launay M., Crozat Y., 2006: Evaluation of the STICS crop model within the EU INTERCROP project. Grain Legumes, 45, 10-12. Donatelli, M., Stockle, C., Ceotto, E., Rinaldi, M., 1997: Evaluation of CropSyst for Cropping Systems at two location of northern and southern Italy, European Journal of Agronomy, 6, 35-45. Dubrovsky, M., 1996: MetandRoll: the stochastic generator of daily weather series for the crop growth model, Meteorologicke Zpravy, 49, 97-105. Dubrovsky, M., 1997: Creating Daily Weather Series With Use of the Weather Generator, Environmetrics, 8, 409-424. Eitzinger, J., Žalud, Z., Diepen van, C.A., Trnka, M., Semerádová, D., Dubrovský, M., Oberforster, M., 2000: Calibration and evaluation of the WOFOST model for winter wheat. 8th International Poster day ' Transport of Water, Chemicals and Energy in the System Soil-Crop Canopy-Atmosphere' 16.11.2000, Bratislava. (CD version, ISBN 80968480-0-3, Institute of Hydrology, Slovak Academy of Sciences). Faber, A., Błoch, Z., Nieróbca, A., Demidowicz, G., Kaczyński, L., 1996: Symulacja wzrostu i plonowania pszenicy ozimej uprawianej w Polsce przy użyciu modelu wzrostu WOFOST. II Weryfikacja Modelu. (Growth and yield simulations of winter wheat cultivated in Poland using WOFOST). Fragm. Agron., 4, 51-58. Jamieson P.D., Semenov, M.A., Brooking, I.R., Francis, G.S., 1998: Sirius: a mechanistic model of wheat response to environmental variation, Europ. J. Agronomy, 8, 161-179. Jamieson, P.D., Porter, J.R., Goudrian, J., Ritchie, J.T., van Keulen, H., Stol, W., 1998b: A comparison of the models AFRCWHEAT2, CERES-Wheat, Sirius, SUCROS2 and

346

B. Lalic, D. T. Mihailovic and M. Malesevic

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

SWHEAT with measurements from wheat grown under drought. Field Crops Research, 55, 23-44. Jones, J.W., Boote, K.J., Jagtap, S.S., Hoobenboom, G., Wilkerson, G.G., 1988: SOYGRO v.5.41: soybean crop growth simulation model, Florida Agricultural Experiment Station Journal, 8304, 1-53. Jones, J.W., Keating, B.A., Porter, C.H., 2001: Approaches to modular model development, Agricultural System, 70, 421-443. Jones, J.W., Hoogenboom, G., Porter, C.H., Boote, K.J., Batchelor, W.D., Hunt, L.A., Wilkens, P.W., Singh, U., Gijsman, A.J., Ritchie, J.T., 2003: The DSSAT cropping system model, European Journal of Agronomy, 18 (3-4), 235-265. Kojic, L, Ivanovic, M., 1986: Dugoročni programi oplemenjivanja pšenice, U: Genetika i oplemenjivanje kukuruza – Dostignuća i nove mogućnosti, Long term programs of maize breeding, In: Genetics and maize breeding-Achievements and new possibilities, Belgrade, 11.IX, 57-75 Reamur, R.A.F., 1735: Observation du thermometre, faites to Paris pendant l’année 1735, comparées avec celles qui ont été faites sous la ligne, to the l’Isle of France, to Alger et en quelques-unites of in the l’Amérique isles, Mém. Acad. give Sci. (Paris), 545. Ritchie, J.T., Otter, S., 1985: Description and performance of CERES-Wheat, A user-oriented wheat yield model ARS wheat yield project, ARS, 38, 159-176. Skorić, D., 1992: Achievements and future directions of sunflower breeding, Field Crops Res., 30, 231-270. Stöckle, C., Martin, S., Cambell, G., 1992: A model to assess environmental impact of cropping systems, Amer. Soc. of Agr. Eng., 92, 2041. Supit, I., Hooijer, A.A., van Diepen, C.A. (eds.), 1994: System Description of the WOFOST 6.0 Crop Simulation Model Implemented in CGMS. Vol. 1: Theory and Algorithms. Catno: CL-NA-15956-EN-C. EUR 15956, Office for Official Publications of the European Communities, Luxembourg.

INDEX

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

A absorption, 50, 215 accounting, 35, 75, 207 accuracy, 3, 4, 17 acid, 320 Adams, 319 adaptability, 201 adaptation, 41, 42, 43, 44, 45, 77, 92, 107, 110, 119, 124, 171, 177, 200, 201, 206, 209, 228, 235, 240, 247, 267, 269, 280, 298 adjustment, 59, 151, 153, 154, 255, 264, 270 administration, 30 administrative, 291 aerosol, 113, 247, 249 aerosols, 5 affiliates, 93 Africa, 14, 20, 87, 166, 214, 241, 313 afternoon, 4 Ag, 160 age, 7, 10, 16, 174, 175, 314 ageing, 334 aggregation, 56, 312 agrarian, 284 agricultural, ix, x, xi, 3, 4, 7, 8, 9, 12, 19, 20, 27, 29, 36, 40, 41, 42, 45, 47, 49, 50, 52, 53, 70, 71, 85, 119, 120, 121, 124, 125, 127, 139, 156, 160, 166, 170, 171, 172, 174, 175, 176, 177, 179, 180, 181, 182, 184, 193, 196, 199, 200, 215, 220, 228, 233, 240, 241, 256, 283, 284, 285, 286, 288, 289, 290, 291, 296, 298, 300, 301, 309, 311, 312, 314, 315, 318, 320, 321, 322, 331, 332, 340, 345 agricultural crop, 125, 196, 199, 200, 220, 241, 315, 322, 345 agricultural market, 289, 300 agricultural sector, 12, 47, 120

agriculture, ix, x, 3, 4, 10, 12, 16, 17, 19, 20, 36, 40, 42, 43, 47, 48, 49, 50, 51, 52, 67, 71, 75, 77, 85, 120, 124, 132, 161, 166, 167, 168, 169, 170, 171, 173, 177, 179, 181, 201, 204, 210, 227, 240, 241, 252, 267, 268, 269, 281, 283, 284, 311, 313, 328, 331, 332, 344, 345 AGRIDEMA, x, xi, 20, 63, 119, 172, 173, 174, 175, 176, 177, 178, 181, 196, 200, 213, 215, 256, 332 agrochemicals, 52, 182 aid, 28, 30 air, 40, 50, 53, 54, 56, 57, 58, 62, 63, 66, 84, 89, 92, 108, 122, 186, 194, 196, 198, 216, 247, 255, 256, 261, 267, 269, 270, 271, 272, 273, 275, 276, 313, 322, 333, 334 air quality, 66 Alabama, 179 Albania, 6, 7, 12 algae, 50 Algeria, 87, 89, 91, 92 algorithm, 100, 185, 192, 271, 307 alpha, 157, 285 alternative, 70, 113, 168, 230, 247 alternatives, 230 aluminium, 147 Amazon, 319 ammonium, 147 anthropogenic, 5, 65, 66, 253 application, x, 4, 52, 63, 73, 77, 78, 82, 120, 123, 125, 128, 134, 168, 169, 171, 176, 179, 205, 243, 244, 264, 284, 289, 291, 331, 332, 345 aquifers, 28 arid, 180, 182, 328 Armenia, 6, 7, 9 Army, 33 Army Corps of Engineers, 33 Asia, 14, 15, 166, 252, 269 Asian, 40

348

Index

assessment, 26, 32, 66, 76, 107, 125, 160, 176, 197, 201, 213, 215, 217, 218, 228, 240, 256, 267, 269, 280, 328, 332, 333, 344 assessment tools, 26 assimilation, 185 assumptions, 71, 192 Athens, 78 Atlantic, 36, 269 Atlantic Ocean, 269 atmosphere, 13, 15, 19, 50, 53, 54, 57, 63, 69, 71, 74, 75, 84, 95, 98, 107, 122, 165, 166, 167, 168, 170, 181, 199, 206, 213, 214, 215, 216, 219, 222, 223, 243, 268, 270, 287, 311, 319 Australia, 14, 29, 35, 166, 170, 235, 236, 281 Austria, 20, 41, 48, 63, 119, 122, 125, 173, 174, 200, 235, 269, 321, 328, 345 automation, 290 autoregressive model, 228 availability, 12, 25, 31, 52, 77, 78, 105, 107, 123, 124, 147, 167, 177, 182, 185, 186, 206, 210, 217, 234 averaging, 109 awareness, 3, 17, 30, 41, 66 Azerbaijan, 6, 7, 9

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

B background information, ix, 172 Bali, 311 banks, 26, 324, 325, 326, 327, 328 barley, 36, 183, 184, 187, 189, 190, 201 barriers, 147 behavior, ix Belarus, 7, 9 Belgium, 48, 125, 280 benchmark, 69, 91, 129, 209 benefits, x, 20, 43, 52, 174, 175, 298 bias, 75, 78, 82, 83, 117 biodiversity, 75, 124 biological responses, 150 biomass, 99, 138, 155, 156, 183, 184, 186, 188, 189, 190, 192, 196, 197, 198, 259, 271, 272, 284, 285, 303, 306, 307, 314, 331, 333, 334, 335, 336, 337, 340, 342, 344 biophysics, 63 biosphere, 63, 75, 320 biosynthesis, 181 Black Sea, 269 blocks, 25, 30, 258 boundary conditions, 12, 14, 69, 73, 127, 128, 129, 134, 135, 214, 219, 229 Brazil, vii, 20, 303, 304, 305, 319 Brazilian, 319

breeding, 332, 346 British Columbia, 265, 301 Brno, 173 Brussels, 125, 280, 300 buildings, 29 Bulgaria, 3, 6, 7, 9, 10, 17, 18, 200, 267, 270, 274, 275, 276, 277, 278, 279, 280 burning, 50, 252, 268 by-products, 186

C calibration, 74, 123, 137, 139, 148, 150, 152, 154, 156, 157, 214, 217, 284, 303, 304, 308, 332, 335, 340, 344 canals, 29 CAP, 41, 49, 177, 284, 289, 290, 297, 300 capillary, 131, 133, 177, 186, 191, 305 carbohydrate, 138, 155, 272, 284 carbohydrates, 247 carbon, 53, 54, 65, 66, 67, 75, 77, 114, 122, 124, 147, 155, 160, 169, 184, 187, 188, 190, 193, 194, 199, 241, 252, 268, 311, 312, 313, 314, 315, 316, 318, 319, 320 carbon cycling, 75 carbon dioxide, 65, 67, 122, 124, 160, 194, 199, 241, 252, 268 Carpathian, 235 case study, 18, 85, 204, 284 catchments, 77, 126 cattle, 92 cell, 103, 105 cellulose, 272 Central Europe, 41, 175, 194, 236, 280 cereals, 36, 79, 133, 185, 186, 189, 190, 195 CERES, vi, 74, 134, 137, 138, 139, 148, 149, 152, 153, 154, 155, 156, 157, 159, 160, 161, 168, 179, 181, 193, 195, 196, 197, 198, 199, 200, 201, 284, 285, 301, 309, 331, 345, 346 changing environment, 52 channels, 29, 203, 290 chicken, 88 chickens, 88, 89, 92 China, 40, 49, 252 circulation, 13, 40, 54, 63, 66, 67, 95, 107, 113, 165, 182, 206, 255, 256, 265, 270, 291, 302 citrus, 87, 92 classes, 271 classical, 60 classification, 146, 167, 193, 286 clay, 84, 130, 145, 146, 147, 148, 159, 217, 218, 219, 313 climate extremes, 85, 92, 166, 236

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Index climate warming, 213, 214 Climatic change, 159, 179, 201, 280 climatic factors, 39, 184 climatology, 40, 116, 206, 236, 270 closure, 105, 259 cluster analysis, 187, 193 clusters, 51 CO2, 50, 75, 76, 77, 100, 101, 102, 105, 121, 125, 126, 155, 160, 161, 181, 183, 184, 185, 187, 188, 189, 190, 192, 194, 198, 199, 200, 215, 241, 243, 256, 257, 262, 263, 272, 311, 312, 313, 319, 320, 333 coastal zone, 124, 214, 218 codes, 129 College Station, 160, 179, 301 colloids, 146 Colorado, 32, 33, 62 common agricultural policy, 300 communication, 3, 18, 109, 320, 332 communities, 9, 30, 71, 241 community, 19, 54, 70, 78, 108, 110, 116, 168, 171, 332 compensation, 185 competition, 49, 328 competitiveness, 51 complement, 23, 27 complexity, 47, 109, 120, 127, 241, 256 compliance, 177, 289 components, 25, 32, 35, 38, 65, 67, 69, 93, 133, 137, 148, 150, 156, 167, 169, 307 composition, 95, 253, 312 compost, 314 computing, 134, 265, 272, 300 concentration, 50, 65, 76, 77, 100, 101, 105, 113, 122, 147, 185, 186, 187, 198, 219, 221, 241, 258, 268, 272, 303, 305, 307, 308, 333 concordance, 255, 260, 261 conditioning, 92 conduction, 53, 129 conductivity, 129, 219, 222, 225, 334 confidence, 4, 39 confidence interval, 39 configuration, 91, 100 conflict, 31 Congress, iv congruence, 260 conservation, 33 constraints, x, 52, 133, 147 construction, 74, 84, 180, 235, 265, 301 consulting, 29 consumers, 48 consumption, 25, 27, 29, 183, 188, 192, 246, 257, 264, 301

349

contaminant, 53, 54 contingency, 29 continuity, 120 contracts, 12 control, 27, 80, 81, 109, 110, 113, 114, 116, 117, 182, 187, 228, 240, 270, 290 convection, 129 convective, 186 conversion, 113, 142, 203, 334 corn, 159, 215, 227, 228, 231, 232, 233, 252 correlation, 8, 13, 36, 37, 39, 95, 100, 101, 103, 167, 206, 228, 245, 275 correlation analysis, 95 correlation coefficient, 38, 39, 100, 101, 103, 245, 275 cost-effective, 51, 283, 300 costs, 26, 28, 89, 123, 284, 288, 289, 290, 291, 296, 299, 300 cotton, 131, 204 counterbalance, 247 coupling, 134 covering, 73, 84, 108 critical points, 8 critical temperature, 79 critical value, 79 Croatia, 7, 12, 195, 196, 198, 199, 200, 201 crop models, 19, 73, 74, 76, 93, 120, 122, 123, 124, 126, 135, 168, 169, 176, 177, 182, 200, 240, 242, 255, 256, 309, 331, 332, 333, 344, 345 crop production, 35, 36, 40, 71, 75, 77, 121, 122, 125, 179, 180, 186, 193, 201, 204, 205, 235, 240, 241, 247, 249, 283, 286, 288, 289, 296, 298, 300, 322, 332, 340, 344 crop residues, 186 crop rotations, 77, 131, 192, 313 crops, 9, 26, 35, 52, 73, 74, 75, 76, 78, 79, 85, 92, 98, 99, 106, 120, 121, 124, 125, 126, 131, 132, 134, 168, 169, 181, 182, 184, 193, 195, 196, 199, 200, 204, 205, 207, 213, 214, 215, 217, 233, 240, 241, 252, 267, 268, 269, 278, 280, 284, 290, 301, 304, 315, 316, 332, 345 cross-validation, 39, 82 Cuba, 240, 252 cultivation, 196, 199, 214, 217, 219, 239, 240, 241, 242, 244, 245, 246, 247, 248, 249, 250, 251, 252, 280, 328 culture, 31 current limit, 177 customers, 120 cycles, 52 cycling, 75, 186 cyclones, 71 Cyprus, 6, 7, 12, 48

350

Index

Czech Republic, 17, 18, 173, 196, 198, 199, 200, 201

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

D dairy, 91 danger, 234, 235 data collection, 252 data set, 4, 66, 82, 84, 139, 156, 187, 211, 218, 222, 242, 252, 281, 285, 303 database, 16, 84, 105, 110, 237, 243, 272, 333 database management, 237, 243 death, 334 death rate, 334 decay, 272 decision makers, x, 19, 49, 52, 77, 124, 228 decision making, ix, 119, 120, 124, 172, 182, 267, 269, 284 Decision Support Systems, 253 decision trees, 29 decision-making process, 24, 28 decisions, x, 12, 17, 19, 23, 25, 30, 39, 66, 288 decomposition, 314, 315 deficiency, 138, 155, 284 deficit, 31, 138, 139, 155, 169, 177, 203, 204, 207, 210, 241, 255, 258, 259, 260, 265, 284, 295, 333, 334 deficits, 29, 265, 295 definition, 32, 33, 169 degradation, 50, 311 delivery, 16 Delphi, 228, 236 demographic change, 270 denitrification, 186, 191, 333 Denmark, 48, 84 density, 55, 56, 58, 59, 156, 186, 242, 255, 258, 259, 260, 288, 313 Department of Agriculture, 33, 160, 161 developed countries, 25, 312 deviation, 5, 197, 231, 335, 340 dew, 53, 55, 322 diamond, 317 direct action, 23, 25, 27, 30 direct measure, 23, 27 discharges, 301 diseases, 156 dispersion, 106, 129 displacement, 59 disseminate, 43, 172 distribution, 19, 29, 36, 50, 77, 79, 80, 83, 91, 92, 103, 104, 105, 109, 117, 131, 147, 155, 170, 186, 206, 207, 208, 209, 210, 228, 229, 230, 231, 258, 259, 260, 284, 294 distribution function, 186, 230, 231

diversity, 26 division, 332 dominance, 230, 231 dosage, 129 download, 44, 45, 97 drainage, 50, 55, 127, 129, 134, 138, 139, 145, 146, 155, 156, 168, 169, 170, 178, 218, 220, 222, 223, 224, 225, 255, 257, 271, 286 drinking, 193 drinking water, 193 drought, 20, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 36, 37, 40, 52, 66, 67, 71, 77, 92, 96, 120, 124, 125, 131, 152, 156, 166, 176, 186, 205, 230, 265, 267, 272, 321, 346 droughts, 24, 25, 26, 29, 31, 32, 33, 40, 50, 234, 268, 269 dry matter, 137, 138, 150, 152, 155, 156, 186, 215, 258, 307, 334 drying, 222, 258 duration, 79, 80, 81, 96, 124, 138, 149, 150, 155, 156, 157, 240, 252, 259, 268, 277, 278, 284, 285, 286

E early warning, 4, 28 earth, 165 East Asia, 40 Eastern Europe, x, 41, 173, 175, 237, 280 ecological, 107, 219, 230, 235, 237, 321 ecological indicators, 237 ecological systems, 107 ecology, 187 economic change, 124, 125 economic development, 3, 4, 19, 243, 270 economic efficiency, 283, 284, 296, 297, 298, 300 economic growth, 243, 271 economic indicator, 296, 297 economic losses, 26 economics, 125 ecosystem, 53, 54, 62, 71, 75, 184, 312 ecosystems, ix, 16, 63, 75, 119, 160, 179, 184, 186, 192, 303, 311, 312, 313, 319 efficiency criteria, 230, 231 egg, 89 Egypt, vi, 239, 240, 241, 242, 245, 246, 251, 252 elaboration, 145 elasticity, 297, 299, 300 electric energy, 27 elongation, 140, 141, 142, 150 email, 93, 174 emission, 15, 107, 108, 109, 110, 111, 112, 113, 114, 116, 117, 177, 187, 198, 200, 276, 277

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Index employment, 91, 204, 210 encouragement, 177, 281 end-users, ix, 44, 172 energy, 10, 16, 27, 28, 49, 55, 56, 62, 67, 76, 78, 89, 165, 215, 291 energy supply, 76 energy transfer, 62 engines, 70 England, 29, 124, 211, 265 enthusiasm, 332 environment, 4, 7, 18, 48, 52, 70, 71, 79, 80, 124, 135, 161, 182, 189, 192, 213, 215, 219, 225, 241, 242, 284, 302, 315 environmental conditions, 41, 62, 241, 313 environmental factors, 122, 138, 155, 284 environmental impact, 26, 32, 75, 346 environmental protection, 51, 243, 271 Environmental Protection Agency (EPA), 29, 33, 161 environmental sustainability, 243, 271 epidemics, 237 equilibrium, 317 equity, 243, 271 erosion, 302 estimating, 116, 118, 146, 168, 216, 252, 290, 301, 314 Estonia, 6, 7, 48, 249, 253 Europe, 6, 9, 11, 13, 14, 15, 20, 24, 42, 47, 48, 49, 65, 66, 67, 68, 71, 73, 74, 75, 77, 78, 79, 84, 85, 92, 95, 97, 121, 124, 125, 149, 154, 156, 159, 160, 166, 168, 169, 171, 173, 176, 177, 178, 267, 269, 272, 280, 300 European Commission, 50, 66, 194, 289, 290, 300 European Community, 73 European Union (EU), x, 15, 48, 49, 66, 74, 78, 82, 93, 94, 108, 121, 124, 125, 131, 132, 133, 135, 165, 166, 171, 172, 173, 177, 181, 280, 300, 345 EUROSTAT, 79 evaporation, 50, 53, 55, 62, 98, 138, 139, 145, 155, 156, 160, 161, 186, 213, 214, 215, 217, 218, 219, 220, 222, 223, 224, 225, 241, 255, 258, 271, 286 evapotranspiration (ET), 53, 55, 62, 129, 134, 154, 155, 156, 157, 161, 170, 177, 185, 186, 203, 206, 207, 208, 216, 217, 239, 240, 256, 257, 258, 259, 260, 261, 262, 265, 271, 285, 286, 287, 288, 293, 294, 295, 299, 300, 302, 318, 321, 322, 328 evolution, 13, 23, 25, 29, 108, 110, 113, 116, 209, 215, 219, 221, 286, 290, 294 evolutionary process, 241 excess supply, 29 execution, 28 experimental design, 258 expertise, 82, 92, 166

351

exploitation, 204 export subsidies, 289 exports, 85, 86, 91, 92 exposure, 189 externalities, 49 extinction, 60, 65, 333, 334 extraction, 55, 129, 134, 303, 307, 308 extrapolation, 270

F failure, 83, 166 family, 183, 184, 194, 240, 270 FAO, 129, 134, 159, 211, 255, 257, 258, 264, 265, 300, 302 farmers, 20, 21, 35, 41, 51, 52, 91, 122, 170, 171, 172, 175, 177, 187, 203, 204, 205, 206, 210, 267, 269, 280, 289, 298 farming, 39, 85, 91, 92, 174, 180, 257, 321 farms, 19 FDA, 213 feedback, 50, 71, 75, 172, 177, 184, 230 feed-back, 47 feeding, 100 fees, 296 fertiliser, 187, 191, 272 fertility, 213, 214, 313 fertilization, 101, 124, 160, 183, 184, 187, 191, 200, 304, 311, 313, 315, 316, 317, 318, 333 fertilizer, 139, 171, 187, 192, 193, 311, 334 field crops, 204, 321, 322 field trials, 240 financial resources, 6, 52 finite differences, 258 Finland, 48, 76, 280 fires, 66, 67, 78 flexibility, 24, 30, 128, 290, 297, 298 flood, 50, 65, 77, 78, 126, 230 flooding, 66, 67, 71, 78, 258 flow, 62, 63, 128, 129, 130, 133, 134, 135, 136, 170, 178, 182, 213, 215, 218, 219, 225, 242, 271, 290 fluctuant, 214 fluctuations, 9, 13, 166, 268 focusing, 177, 270 food, 8, 17, 47, 50, 51, 91, 240, 241, 332 food production, 47, 50 forecasting, 4, 5, 7, 12, 19, 20, 36, 53, 54, 71, 73, 166, 178, 179, 215, 291, 332, 335, 339, 340, 344 forest ecosystem, 312 forest fire, 66, 67, 77, 78 forest fires, 66, 67 forest management, 75 forestry, 71, 75, 77, 201, 312

352

Index

forests, 75, 76 France, 4, 7, 48, 79, 89, 101, 103, 121, 154, 173, 316, 346 free trade, 289 frequency distribution, 36 freshwater, 75 friction, 60, 61 frog, 50 frost, 71, 230, 233 fruits, 87 fuel, 268 functional approach, 291 funding, 25, 170, 175 funds, 40, 121, 175, 177 fungicide, 252 futures, 270

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

G gas, 192, 268, 311, 312 gas exchange, 192 gases, 5, 65, 253 generation, 107, 108, 113, 116, 198, 216, 235 generators, 166, 167, 174, 176, 177, 181, 216, 257, 265, 291, 301 genetics, 148, 332 Geneva, 3, 16, 17, 20, 170, 178, 179, 181, 182, 201, 265, 301, 302 genotype, 138 genotypes, 106, 138, 155 geophysical, 7 Georgia, 40, 179 Germany, vi, 20, 32, 48, 79, 84, 95, 121, 183, 184, 187, 188, 190, 193, 194 germination, 149, 246 GIS, 123, 179, 193, 207, 209, 237, 271 glaciers, 65 global climate change (GCC), 113, 268, 312, 318 global trade, 49 global warming, x, 173, 264, 311, 319, 321 goals, 50, 176, 217, 332 governance, 52 government, ix, 6, 36, 65, 175, 284 grain, 99, 100, 137, 138, 142, 148, 150, 151, 152, 154, 155, 156, 160, 178, 185, 190, 196, 197, 218, 268, 278, 285, 287, 304, 306, 307, 331, 333, 334, 335, 338, 339, 340, 342, 344 grains, 138, 156, 285 grapes, 92, 258 graph, 25, 231, 232 grass, 54, 60, 63, 257 grasslands, 320 gravity, 59, 240

Greece, 7, 12, 48, 89, 92, 194, 213 greed, 312 greenhouse, 50, 65, 118, 165, 251, 253, 255, 256, 268, 270, 289, 311, 312, 319 greenhouse gas, 50, 65, 118, 165, 255, 256, 268, 270, 289, 311, 312 greenhouse gases, 50, 65, 118, 165, 255, 256, 268, 270, 289, 311, 312 grid resolution, 206 grids, 84 ground water, 132, 134 groundwater, 27, 28, 29, 53, 54, 124, 127, 129, 130, 131, 132, 133, 134, 135, 168, 170, 204, 216, 218, 219 grouping, 167 groups, 25, 41, 45, 71, 75, 82, 94, 146, 169 growth rate, 138, 155 guidelines, 29, 31, 168, 170 Gujarat, 82

H hanging, 222 harm, 235, 245 harmony, 245 harvest, 8, 79, 184, 186, 191, 217, 256, 257, 258, 262, 263, 264, 272, 333, 334 harvesting, 29, 86 Hawaii, 179, 201 health, 4, 10, 16, 67, 73, 76, 77, 78, 92 health problems, 92 heart, 70 heat, 53, 55, 56, 57, 58, 77, 78, 98, 106, 124, 127, 128, 129, 160, 182, 211, 222, 230, 233, 281 heat capacity, 55, 57 height, 56, 57, 58, 59, 60, 215, 222, 322, 323, 328, 329 hemisphere, 5 heterogeneity, 128 heterogeneous, 62, 124, 243, 270 high pressure, 269 high resolution, 14, 16, 68, 97, 166, 211 high temperature, 79, 80, 81, 91, 105, 106, 152, 157 high-level, ix, x, 171 Holland, 32 horizon, 145, 258, 315 human, ix, 5, 6, 10, 49, 66, 67, 70, 76, 77, 119, 172, 241, 268, 280, 312 human activity, 241 human resources, 172 humidity, 90, 95, 96, 97, 216, 260, 322 Hungarian, 228, 230, 231, 233, 235, 236, 237

Index Hungary, vi, 48, 79, 199, 200, 227, 228, 230, 235, 236 hybrid, 156, 304 hybrids, 156, 157, 280, 285, 332 hydrologic, 25, 28, 31, 33, 154, 200 hydrological, 53, 54, 71, 124, 131, 133, 136, 170, 270 hydrology, 63, 125 hypothesis, 60

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

I Iberian Peninsula, 37, 40, 108, 109, 113, 255, 256, 257, 261, 289 ice, 50, 108 ice caps, 50 identification, 41, 79 images, 131 impact assessment, 20, 23, 24, 70, 121, 166, 167, 168, 176, 180, 265, 281, 301 implementation, 6, 25, 27, 28, 29, 30, 66, 129, 135, 280 imports, 240, 241 in situ, 40 incidence, 240, 253 inclusion, 12, 62, 316 income, 184, 192, 284, 289, 296, 297, 300 incubation, 194 India, 78, 79, 106, 130, 134, 178, 240 indication, 305 indicators, 31, 77, 217, 227, 229, 233, 236, 290, 296 indices, 36, 38, 40, 96, 230, 233, 234, 235, 236, 237, 306 indirect effect, 40 indirect measure, 23 induction, 298, 299 industrial, 27, 29, 65, 240, 241 industry, 50, 268, 345 inert, 314, 319 infinite, 113 information economy, 243 Information System, 159 information systems, 134 infrastructure, 49, 120, 290 initial state, 15, 116 initiation, 247 inorganic, 333 insight, 171 inspection, 335, 339, 340 inspiration, 169 institutions, ix, 26, 173, 174, 175 insurance, x, 16, 177 insurance companies, x, 177

353

integration, 17, 41, 70, 71, 79, 192 interaction, 41, 53, 106, 127, 128, 129, 130, 133, 241 interactions, 13, 19, 35, 54, 67, 122, 124, 129, 170, 177, 184, 192, 194 interdisciplinary, 321 interface, 134, 168, 219, 242 interference, 66 Intergovernmental Panel on Climate Change (IPCC), 65, 66, 84, 92, 105, 106, 107, 108, 109, 113, 114, 115, 118, 121, 125, 165, 169, 179, 180, 187, 193, 198, 201, 209, 211, 235, 236, 241, 243, 252, 255, 256, 265, 270, 281, 289, 301, 318 internet, 16, 44, 128, 135, 174, 256, 291 interval, 138, 148, 231, 258, 288 intervention, 289 investment, 177, 290 Ireland, 48 isotope, 187 Israel, vii, 160, 303 Italy, 14, 48, 73, 85, 87, 88, 89, 95, 98, 99, 101, 121, 166, 173, 180, 213, 214, 216, 218, 219, 220, 222, 345

J Japan, 7, 49 Jordan, 321 justification, 50

K Kazakhstan, 6, 9 kernel, 138, 149, 150, 155, 196, 197 Kyoto Protocol, 65, 311, 312

L labor, 291 laboratory method, 201 labour, 85, 271, 288 land, 12, 13, 15, 19, 41, 45, 48, 51, 53, 62, 63, 66, 70, 75, 97, 110, 122, 124, 125, 127, 131, 160, 165, 184, 193, 206, 211, 215, 271, 272, 281, 311, 312, 319 land use, 51, 62, 63, 75, 122, 124, 125, 131, 160, 184, 193, 271, 312, 319 landscapes, 184, 194 land-use, 41, 45, 66, 70, 110 language, 314 large-scale, 19, 35, 36, 40, 110 Latvia, 6, 7, 9, 48 law, 55

354

Index

laws, 167, 168 leachate, 191 leaching, 74, 77, 84, 132, 133, 191, 219, 221, 272, 307, 309 leaf blades, 149, 150 learning, 100 legislation, 27, 28, 29, 30, 51 legume, 79, 313 legumes, 178 Legumes, 345 Leibniz, 184 lettuce, 215, 217, 222 life span, 334 lignin, 272 likelihood, 99, 113 limitation, 186, 280 limitations, x, 42, 87, 116, 118, 124, 177, 309, 331 linear, 36, 37, 60, 61, 70, 77, 79, 95, 96, 100, 108, 138, 155, 167, 170, 197, 198, 216, 222, 225, 228, 229, 274, 295 linear function, 100, 138, 155 linear regression, 95, 96, 216, 222, 295 links, 10, 13, 44, 49, 66, 93, 100 Lithuania, 7, 12, 48 livestock, 9, 20, 26, 50, 85, 88, 92 localised, 14 location, 116, 117, 187, 207, 222, 257, 323, 331, 345 logging, 186 long period, 197, 241 loss of appetite, 89 losses, 33, 35, 146, 156, 168, 294 Luxembourg, 159, 194, 346 Luxemburg, 48, 194 lying, 168

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

M Macedonia, 280 machinery, 271 macropores, 129 Madison, 160 magnetic, iv maintenance, 291, 334 maize, 138, 155, 156, 157, 159, 160, 168, 183, 184, 195, 196, 197, 198, 199, 200, 201, 204, 236, 247, 252, 283, 284, 285, 287, 289, 290, 292, 293, 294, 295, 296, 298, 299, 300, 301, 303, 304, 305, 306, 307, 308, 346 Malta, 48 management practices, 169, 193 manifold, ix, 119, 170, 171, 321 man-made, 47 manure, 314, 318

market, 49, 51, 177, 240, 289, 290, 297 market prices, 177, 297 markets, 26, 49, 286, 289, 300 Markov chain, 96 mask, 97 matrix, 23, 25, 26, 27, 129 maturation, 40, 234, 235 measurement, 30, 178, 323, 324 measures, ix, 8, 20, 23, 25, 27, 28, 30, 31, 41, 42, 43, 119, 125, 177, 180, 255, 267, 269, 280 meat, 85 media, 12, 30, 41, 93, 182, 215, 267, 269 median, 83 Mediterranean, vii, x, 31, 35, 40, 41, 74, 78, 85, 89, 91, 92, 98, 99, 154, 156, 157, 160, 173, 175, 177, 213, 214, 227, 252, 256, 270, 283, 285, 313, 320 Mediterranean climate, 214 Mediterranean countries, x, 91, 173, 175 melting, 50, 108, 324, 325, 326, 327, 328 meteorological, 4, 5, 7, 8, 12, 97, 103, 152, 165, 166, 167, 171, 174, 177, 185, 187, 193, 195, 196, 197, 198, 199, 222, 241, 256, 257, 260, 261, 262, 270, 285, 292, 293, 314, 315 microbial, 314, 315 microclimate, 321 micrometeorological, 63 Middle East, 87 milk, 90, 91, 92 Millennium, 49 Millennium Development Goals, 49 mineralization, 185, 186, 194, 318 Ministry of Environment, 107, 117, 118, 280 Minnesota, 237, 309 missions, 65, 165, 255, 289, 319 model system, 14, 67, 78 modeling, ix, x, xi, 19, 20, 63, 125, 127, 131, 133, 135, 136, 161, 177, 179, 223, 302 modernisation, 205, 210 modernization, 283, 284, 289, 290, 291, 296, 297, 298, 299, 300 modules, 68 moisture, 12, 36, 54, 55, 57, 62, 71, 77, 98, 128, 133, 135, 136, 138, 147, 155, 160, 182, 184, 186, 190, 191, 192, 194, 207, 219, 314, 333, 334 moisture content, 57, 147, 194, 333, 334 Moldova, 6 momentum, 53, 59, 100 monsoon, 40 Monte-Carlo, 113, 228 Monte-Carlo simulation, 113, 228 Montenegro, 7, 12 morbidity, 65 morphological, 54, 217

Index morphology, 60, 214, 217 mortality, 65 movement, 53, 129, 134, 138, 170, 182 multidimensional, 169 multiplication, 246 Myanmar, 252

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

N naming, 96 NATO, 319 natural, 3, 4, 5, 26, 47, 50, 51, 54, 63, 67, 70, 71, 77, 110, 125, 184, 187, 192, 240, 261, 268 natural disasters, 3, 4 natural resources, 51, 240 natural science, 125 natural sciences, 125 negative consequences, x, 173, 219 negotiation, 25, 28 Netherlands, 48, 95, 121, 127, 135, 160, 173, 178, 179, 181, 182, 193, 211, 213, 215, 225, 319, 345 network, 23, 25, 95, 168, 172, 175, 187, 260, 270, 290 networking, 3, 17 neural network, 95 neurons, 100 New Jersey, 236 New York, iii, iv, 62, 63, 118, 161, 179, 194, 251, 252, 253, 301 New Zealand, 236 newsletters, 93 Nielsen, 135, 170, 180 Niger, 178 nitrate, 147, 186, 193, 272, 309 nitrogen, 75, 77, 84, 99, 105, 138, 139, 142, 145, 147, 151, 155, 156, 169, 182, 183, 184, 185, 186, 187, 191, 193, 194, 196, 271, 272, 284, 303, 304, 305, 307, 308, 309, 318, 319, 340 N-mineralization, 194 nodes, 100 non-linearity, 225 normal, 25, 36, 222, 224, 229, 241, 269 normalization, 228, 257 North Africa, 87, 241 North America, 14, 149, 166, 269 North Atlantic, ix, 12, 13, 35, 36, 37, 39, 66, 67, 268, 269 Northeast, 184, 194, 257, 261 Northern Hemisphere, 36 Norway, 85 nutrient, 135, 194, 303 nutrients, 52 nutrition, 253

355

O obligation, 96 observations, 4, 8, 13, 70, 71, 75, 77, 79, 80, 81, 83, 93, 94, 95, 97, 108, 117, 152, 201, 244, 245, 260, 261 oceans, 50, 108, 165 oil, 77, 85, 86, 87, 91, 139, 146, 156, 185, 219, 241, 312 oil production, 91 oils, 138, 155, 284 olive, 85, 87, 99, 204 olive oil, 85, 87 olives, 92 online, 36, 44, 71, 75, 76, 95, 135 ontogenesis, 185 optimization, 240, 294, 301 orbit, 49 order statistic, 19 organ, 186, 215 organic, 145, 147, 186, 271, 311, 312, 313, 314, 315, 316, 317, 318, 319, 320, 333 organic C, 314, 316, 318 organic matter, 145, 147, 186, 271, 311, 312, 313, 314, 317, 318, 319, 320 orientation, 289, 322

P Pacific, 13, 15, 33 Pap, 33 parameter, 56, 61, 84, 103, 114, 116, 135, 147, 160, 182, 217, 218, 223, 224, 225, 259, 260, 264, 301, 315, 334, 340 parameter estimation, 135, 182 Paris, 33, 236, 346 partition, 215, 218, 259, 271 partnership, 3, 17, 256 pasture, 319 pathways, 56, 138 pattern recognition, 167 peak demand, 203, 210 pedigree, 178 penalties, 272 per capita, 243, 271 percentile, 103, 105 percolation, 127, 131, 133, 185, 207, 217, 333, 334 periodicity, 97 permit, 25, 27, 31 personal communication, 109 perturbations, 224, 271 pest control, 182

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

356

Index

pesticide, 131, 132, 134, 135 pesticides, 127 pests, 156, 235 pH, 145, 147, 242, 286, 313 phase space, 116 Philippines, 178 philosophy, 184 phonological, 284 phonology, 243 photoperiod, 106, 137, 138, 149, 150, 155 photosynthesis, 138, 155, 169, 184, 185, 186, 199, 284 photosynthetic, 178, 215, 333 photosynthetic systems, 178 physical environment, 127 physical properties, 241, 242 physics, 62, 84, 103, 127 physiological, 54, 57, 137, 149, 150, 152, 153, 154, 181, 195, 196, 197, 198, 243, 262, 303, 304, 305 physiology, 159, 272 pigs, 92 pilot study, 99, 102, 103 pipelines, 29, 290 planetary, 63 planning, 23, 24, 28, 32, 33, 209, 215, 253 plants, 60, 80, 81, 84, 147, 149, 151, 186, 195, 199, 218, 233, 241, 253, 256, 301, 316, 333, 340 plausibility, 270 play, 54, 124, 321, 328 Poland, 7, 9, 47, 48, 345 policy makers, 66, 166, 177, 193, 256 policy making, 169, 241 policymakers, 66 politicians, x, 177 politics, 49, 50, 52 pollution, 50, 53, 54, 193 polynomial, 288 pools, 147, 186, 272, 312, 320 poor, 124, 131, 133, 313 population, 138, 155, 240, 243, 271, 307 pore, 186 porous, 58 Portugal, 36, 40, 48, 87 potato, 169, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 345 potatoes, 92 poverty, 4 poverty reduction, 4 power, 331 practical knowledge, 332 pragmatic, 166 precipitation, 7, 36, 38, 39, 40, 55, 65, 68, 84, 95, 96, 98, 100, 101, 103, 108, 109, 110, 111, 113, 115,

117, 122, 138, 139, 143, 154, 155, 156, 166, 183, 188, 191, 196, 197, 198, 203, 207, 208, 213, 214, 216, 219, 228, 229, 230, 233, 235, 237, 240, 243, 253, 256, 257, 262, 267, 268, 269, 270, 272, 273, 274, 275, 276, 277, 285, 287, 289, 292, 294, 299, 305, 313, 314, 322, 324, 333, 334 predictability, 4, 7, 12, 14, 19, 70, 73, 93, 166 prediction, ix, 3, 4, 5, 6, 8, 12, 13, 15, 16, 17, 18, 19, 20, 40, 62, 66, 67, 68, 69, 70, 71, 73, 77, 80, 94, 113, 117, 166, 171, 179, 180, 181, 182, 190, 192, 247, 255, 256, 260, 265, 302, 318, 331 prediction models, 12 predictive model, 4 predictive models, 4 predictor variables, 167 predictors, 35, 36, 39, 167, 270 preference, 139, 156, 286 pressure, 36, 56, 57, 58, 84, 96, 97, 98, 129, 130, 203, 204, 258, 269, 290, 333, 334 prices, 177, 289, 290, 297 PRISM, 94 private, 177, 214, 217, 267, 269 private investment, 177 proactive, 25, 27, 30, 31 probability, 36, 66, 67, 70, 83, 84, 103, 104, 105, 106, 113, 116, 165, 204, 228 probability density function, 36, 83, 116 probability distribution, 113 probe, 303, 305 producers, 7, 85, 92, 284, 290 production costs, 289, 297, 299, 300 production function, 286 productivity, 37, 38, 39, 51, 65, 75, 77, 79, 80, 127, 131, 134, 135, 136, 160, 178, 195, 197, 200, 201, 204, 218, 219, 221, 240, 246, 253, 280, 283, 284, 296, 302, 303, 304 profit, 288 profitability, 89, 283, 296, 297, 298, 299, 300 profits, 180, 286 program, 33, 44, 92, 147, 168, 169, 174, 177, 222, 235, 314, 315 programming, 236 propagation, 100 property, iv, 77 protection, 48, 51, 73, 132, 187, 243, 271 protein, 333 protocol, 23, 29, 130 protocols, 29 pruning, 264 pseudo, 169 public, 26, 66, 70, 171 Puerto Rico, 319 pulse, 333

Index P-value, 244

Q quality control, 67, 97, 108 questionnaire, 3, 6

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

R radiation, 50, 53, 55, 84, 96, 98, 99, 105, 126, 138, 139, 142, 143, 155, 156, 185, 186, 196, 198, 215, 216, 229, 234, 243, 257, 258, 259, 262, 272, 285, 286, 289, 292, 301, 322, 333, 334, 335 radio, 280 rain, 48, 85, 183, 184, 191, 214, 233 rainfall, 13, 19, 29, 30, 71, 77, 86, 87, 99, 100, 103, 105, 131, 166, 181, 184, 206, 207, 208, 210, 215, 233, 256, 262, 264, 271, 272, 284, 287, 292, 304, 306, 318 range, 4, 5, 6, 7, 16, 17, 23, 24, 27, 41, 66, 67, 71, 72, 73, 75, 76, 77, 78, 80, 100, 105, 108, 113, 114, 115, 116, 120, 145, 147, 166, 169, 198, 204, 209, 241, 243, 257, 270, 289, 326 rat, 27 reading, x real time, 28 reality, 122, 124, 342 recession, 287 recognition, 41, 167 reconcile, 30 recovery, 284, 289 recycling, 186 redistribution, 62 reference frame, 107, 296 reflection, 28, 139, 156, 286 reforms, 25 regional, 6, 7, 14, 16, 19, 36, 41, 42, 43, 44, 53, 54, 63, 65, 66, 67, 69, 70, 71, 72, 73, 85, 93, 97, 107, 109, 110, 113, 114, 117, 118, 122, 124, 125, 131, 166, 167, 172, 175, 184, 195, 198, 210, 213, 214, 235, 243, 257, 269, 270, 271, 291, 304, 308, 312, 314, 319 regression, 39, 84, 95, 96, 103, 167, 216, 222, 287, 295, 301, 331 regression analysis, 222, 331 regression line, 103, 295 regression method, 167 regular, 12, 91, 340 regulation, 8 relationship, 69, 77, 194, 236, 261, 288, 291 relationships, 35, 39, 40, 47, 167, 171, 215, 216, 302 relevance, 121, 209, 312

357

reliability, 15, 51, 137, 171, 176 remote sensing, 134, 136, 216 renewable energy, 49 replication, 258, 260 reproduction, 90, 92 research and development, 82 reserves, 124 reservoir, 271 reservoirs, 28 residential, 28, 29 residues, 186, 314, 316 resistance, 52, 55, 56, 57, 58, 62, 152, 156, 287, 315 resolution, 14, 16, 31, 63, 67, 68, 69, 70, 71, 73, 74, 87, 91, 92, 95, 97, 103, 110, 166, 167, 185, 206, 211, 228, 270, 281 resource management, 32 resources, 4, 6, 10, 23, 24, 25, 27, 28, 29, 30, 32, 49, 50, 51, 52, 67, 76, 77, 124, 160, 172, 203, 204, 206, 209, 213, 214, 215, 216, 218, 221, 240, 253 respiration, 185, 186, 334 retention, 222, 223, 224 returns, 192 rice, 79, 131, 169, 204 ripeness, 233 risk, 23, 26, 28, 29, 33, 49, 66, 67, 71, 76, 77, 103, 104, 105, 106, 119, 120, 121, 166, 174, 177, 179, 227, 228, 229, 230, 231, 232, 233, 235, 236, 268, 308, 344 risk assessment, 66, 67, 71, 120, 121, 166, 177, 227, 308 risk aversion, 229, 231, 232 risk management, 77, 179 risks, x, 41, 42, 43, 45, 66, 84, 122, 175, 177, 180, 265 river basins, 124, 125, 136 rivers, 289 robustness, 116 Romania, 7, 9, 96, 173, 280 Rome, 134, 211, 300, 302 rotation axis, 50 rotations, 186, 318 roughness, 58, 59, 60, 61, 287 routines, 134 Royal Society, 319 runoff, 53, 55, 129, 138, 139, 146, 155, 156, 207, 213, 214, 217, 286, 287 rural, 51, 177, 204 rural areas, 51 rural development, 177 Russia, 7, 12 Russian, 253 rye, 36, 184

358

Index

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

S safeguard, 52 safety, 8, 77 saline, 217, 219 salinity, 52, 128, 131, 214, 215, 218, 219, 220 salt, 168, 217, 218, 219, 221 sample, 15, 98 sampling, 114, 116, 222, 305, 312 sanctions, 332 sand, 130, 147, 188, 214, 217 satellite, 131 saturation, 56, 57, 129, 146, 147, 185, 333, 334 savings, 28, 29, 50, 52 scalar, 149, 150 scaling, 52, 88, 182, 196, 198, 265, 302 scarcity, 25, 33, 50, 204, 205, 210 scattering, 50 scheduling, 17, 120, 168, 174, 181, 213, 214, 218, 219, 252, 256, 257, 264, 265, 302 scientific community, 332 scientific knowledge, 70 scores, 83, 165 sea ice, 69, 211, 281 sea level, 65, 98, 108, 322 sea-ice, 12 sea-level, 36 search, 52, 229, 236 seasonal variations, 84 seed, 80, 218 seeding, 124 seeds, 140, 141, 246 selecting, 174 semiarid, 321 semi-arid, 182, 313 senescence, 138, 155, 284 sensing, 134, 136, 216 sensitivity, 62, 63, 73, 77, 84, 100, 112, 113, 116, 122, 137, 149, 150, 155, 160, 198, 200, 214, 220, 222, 223, 224, 288, 301, 302, 312 sensors, 260 Serbia, 7, 12, 53, 54, 173, 331, 332, 333, 344 series, 24, 31, 36, 39, 49, 85, 88, 117, 148, 166, 167, 187, 195, 196, 197, 198, 199, 200, 201, 204, 228, 229, 231, 232, 239, 255, 256, 257, 285, 286, 289, 291, 314, 345 services, iv, ix, x, 6, 44, 75, 97, 172, 177, 284, 289, 298, 331, 340 severity, 33, 66, 67, 71 shape, 61, 198, 215, 224, 302, 323 sharing, 51, 271 shear, 59 sheep, 89, 90, 91, 92

shelter, 59, 329 shocks, 166 shortage, x, 23, 25, 33, 50, 173, 186, 228, 284, 292, 305 short-range, 5 short-term, 53, 54, 79, 80, 177 shrubs, 322 sigmoid, 100 sign, 116 signals, 36, 264 signs, 108 simulation, ix, x, 12, 21, 54, 62, 71, 74, 75, 79, 80, 99, 113, 116, 117, 119, 120, 121, 122, 124, 125, 132, 138, 155, 167, 168, 169, 170, 171, 172, 173, 175, 176, 177, 178, 179, 180, 181, 184, 186, 188, 192, 193, 194, 195, 201, 211, 241, 243, 244, 252, 256, 257, 264, 265, 271, 272, 281, 284, 287, 291, 294, 301, 315, 316, 317, 318, 327, 328, 338, 342, 346 singular, 29, 70 sites, 66, 77, 84, 92, 99, 101, 154, 188, 190, 191, 260, 267, 270, 273, 275, 276, 322 skills, 120 Slovakia, 7, 12, 48 Slovenia, 6, 7, 12, 48, 199, 201, 280 smoothing, 270 social impacts, 26 sociologists, 30 software, 40, 92, 103, 166, 168, 174, 227, 229, 233, 237, 243 soil analysis, 304 soil erosion, 124 soil organic C (SOC), 311, 312, 314, 315, 316, 317, 318, 319 soils, 101, 138, 145, 146, 147, 155, 168, 182, 188, 191, 194, 222, 223, 225, 240, 243, 258, 271, 272, 284, 309, 311, 312, 313, 314, 315, 318, 319, 320 solar, 79, 96, 98, 126, 139, 143, 156, 196, 198, 234, 243, 257, 259, 272, 285, 289, 292, 301 solid waste, 48 Sorghum, 217, 218, 219, 303, 304 sounds, 4 South Africa, 20 South Korea, 15 soybean, 168, 199, 267, 269, 346 soybeans, 182 space-time, 236 Spain, 23, 24, 25, 26, 27, 28, 29, 31, 32, 35, 36, 37, 40, 48, 79, 87, 89, 91, 92, 101, 107, 108, 109, 110, 114, 117, 118, 137, 139, 150, 153, 154, 156, 159, 160, 165, 173, 176, 179, 203, 205, 206, 207, 208, 209, 210, 211, 213, 252, 255, 256, 257, 258, 284, 285, 292, 293, 300, 301, 313

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

Index spatial, ix, 16, 19, 36, 37, 47, 65, 66, 67, 69, 71, 79, 91, 103, 109, 110, 117, 123, 124, 166, 171, 172, 206, 209, 270, 301 spatial analysis, 301 species, 65 specific gravity, 240 specific heat, 56 specific knowledge, 52 speed, 59, 60, 61, 63, 84, 96, 187, 216, 260, 321, 323, 325, 326, 327, 328, 333, 334 springs, 26 Sri Lanka, 135 stability, 59 stages, 31, 80, 137, 138, 139, 150, 155, 166, 186, 272, 284, 286, 288, 295, 298, 299, 303, 307, 340 stakeholder, 70, 176 stakeholders, x, 3, 17, 18, 41, 52, 66, 70, 71, 93, 166, 170, 171, 215, 228 standard deviation (STD), 111, 113, 114, 122, 142, 199, 257, 289 standards, 75 starch, 240 statistical analysis, 201 statistics, 19, 62, 223, 271 steady state, 129, 314 stochastic, 31, 95, 96, 195, 196, 198, 200, 228, 230, 231, 232, 236, 265, 301, 345 stock, 8, 312, 316, 319 storage, 27, 50, 58, 77, 124, 186, 215, 240, 247, 287, 304, 305, 306, 312, 313, 318, 320, 334 storms, 71, 268, 269 strategies, 23, 24, 25, 28, 31, 41, 42, 43, 45, 50, 107, 127, 129, 131, 134, 168, 184, 192, 193, 209, 214, 218, 219, 221, 228, 241, 283, 284, 286, 296, 301, 304, 312 stratification, 287 strength, 116, 147, 175 stress, 58, 59, 65, 77, 78, 79, 80, 106, 125, 131, 138, 139, 142, 147, 152, 155, 176, 181, 186, 187, 215, 218, 219, 240, 252, 255, 259, 260, 264, 286, 287, 288, 294, 295, 298, 299, 304, 306, 307, 327, 328 stress level, 176, 286 stressors, 124 structuring, 30, 50 students, 44, 174 subjective, 231 subsidies, 49, 284, 286, 289, 297 subsidy, 288, 289 substances, 132, 135 subsurface flow, 129 sugar, 183, 184, 186, 187, 189, 190, 204, 233, 234, 240 sugar beet, 183, 184, 186, 187, 189, 190, 204

359

sugarcane, 169 summer, 40, 68, 90, 109, 110, 113, 114, 157, 183, 184, 188, 191, 192, 203, 206, 208, 214, 218, 230, 246, 264, 272, 276, 285 sunflower, 169, 204, 267, 268, 269, 277, 278, 279, 280, 346 superiority, 344 supply, 25, 27, 32, 50, 51, 75, 168, 187, 190, 192, 204, 270, 272, 290, 297, 298, 306 suppression, 236 surface energy, 62 surface layer, 63, 145 surface water, 27, 28, 127, 129, 130, 132, 133, 134, 204 surprise, 26 sustainability, 51, 125, 217, 243, 271, 289, 304 sustainable development, 4, 332 Sweden, 48 swelling, 146 Switzerland, 178, 179, 181, 182, 237, 265, 301, 302 symbols, 80, 81 symptoms, 228, 231 synthesis, 180, 228 systems, 4, 12, 29, 42, 52, 65, 66, 70, 82, 107, 119, 120, 134, 135, 168, 170, 177, 178, 180, 181, 184, 193, 194, 199, 203, 204, 205, 207, 210, 213, 214, 215, 216, 217, 220, 241, 251, 252, 258, 265, 283, 284, 288, 296, 297, 298, 300, 303, 304, 305, 306, 307, 308, 321, 328, 346

T targets, 16, 311 Taylor expansion, 229 technical assistance, 30 technological change, 243, 270 technology, 50, 51, 119, 124 temporal, 19, 40, 47, 171, 219, 221, 286 territory, 270 Texas, 160, 179, 201, 301 theoretical biology, 253 thinking, 256 Thomson, 180 threat, 32, 65, 203, 206 threshold, 77, 84, 85, 87, 103, 104, 105, 150, 246, 247, 271, 334 thresholds, 31, 73, 77, 79, 85, 92, 99, 103 time frame, 12 time periods, 105, 206, 229 time resolution, 185 time series, 36, 88, 167, 187, 197, 198, 228, 229, 231, 232, 239, 257, 285, 286, 289, 291 timetable, 172

360

Index

timing, 19, 80, 81, 106, 133, 138, 155, 170, 233, 234, 252, 335, 339 titration, 320 tolerance, 77 topsoil, 334 tourism, 10 toxicity, 147 trade, 49, 85, 289 tradition, 169 training, 51, 66, 68, 93, 100, 174 transactions, 319 transfer, 52, 53, 54, 56, 58, 59, 60, 61, 62, 63, 100, 179, 180, 284, 309 transformation, 196, 229, 284 transformations, 51, 186 transmission, 50 transpiration, 53, 55, 57, 105, 139, 155, 157, 186, 199, 217, 218, 220, 222, 223, 224, 225, 259, 271, 285, 286, 293, 299 transport, 10, 29, 50, 53, 59, 127, 128, 129, 130, 133, 135, 182, 186, 213, 215, 218, 219, 222, 225, 258 transport processes, 222 travel, 32, 271 travel time, 271 trees, 29, 204, 265, 322 trial, 241, 303, 313 triggers, 33 tropical areas, 5 T-test, 242, 245 tubers, 240, 242 turbulent, 54, 58, 59, 60, 61, 63 Turkey, 87, 89, 92, 101, 268, 270, 274, 275, 276, 277 turnover, 311, 313, 314, 315, 319 Tuscany, 103, 104, 105

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

U U.S. Department of Agriculture (USDA), 146, 160, 182, 286, 313, 320 uncertainty, 20, 66, 67, 71, 92, 97, 106, 110, 113, 116, 118, 120, 122, 133, 134, 160, 166, 180, 192, 230 UNEP, 281 UNESCO, 33 UNFCCC, 65, 66 uniform, 25, 84, 105 unions, 267, 269 United Kingdom (UK), 7, 10, 17, 33, 48, 65, 66, 95, 96, 106, 118, 117, 121, 125, 126, 167, 173, 178, 179, 193, 203, 243, 257, 269, 270, 276, 281, 291, 313 United Nations (UN), 20, 65, 77, 280, 300, 332 United States, 24, 29, 30, 319

Uruguay, 309

V Valencia, 23 validation, 54, 63, 69, 123, 139, 153, 156, 157, 198, 214, 217, 218, 240, 241, 242, 244, 245, 255, 258, 260, 284, 332, 334, 340, 344, 345 values, 5, 37, 80, 84, 97, 103, 110, 114, 115, 116, 130, 139, 142, 145, 146, 147, 149, 150, 151, 156, 157, 167, 196, 197, 198, 199, 206, 209, 210, 217, 218, 219, 223, 224, 227, 233, 258, 260, 261, 262, 269, 271, 272, 275, 285, 286, 287, 295, 305, 306, 331, 332, 334, 335, 338, 340, 342, 344 variability, ix, 3, 4, 5, 6, 9, 12, 17, 19, 20, 36, 37, 66, 67, 70, 71, 73, 77, 80, 93, 99, 105, 106, 110, 117, 120, 121, 122, 124, 126, 167, 168, 170, 171, 191, 200, 209, 213, 214, 218, 219, 223, 224, 225, 236, 253, 261, 265, 267, 268, 269, 281, 301, 312 variables, 35, 36, 38, 39, 40, 54, 56, 58, 71, 74, 76, 84, 97, 98, 100, 103, 116, 159, 166, 167, 171, 174, 177, 185, 197, 215, 223, 228, 229, 243, 257, 270, 271, 272, 285, 291, 340 variance, 20, 36, 37, 39, 180, 229 variation, 5, 36, 40, 85, 138, 155, 223, 289, 290, 312, 335, 345 vegetables, 17, 204 vegetation, 53, 54, 56, 57, 58, 60, 61, 62, 63, 74, 75, 114, 127, 128, 160, 197, 199, 230, 271, 312, 322, 327, 331, 333, 334, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344 velocity, 60, 61, 62 Victoria, 319 vineyard, 255, 257, 258 visible, 334 vision, 25, 47, 50, 51 vulnerability, 12, 19, 42, 75, 107, 199, 200, 228, 241, 252, 267, 269, 280

W Wales, 211, 265 warning systems, ix, 4, 119 Warsaw, 47 wastewater, 204 water policy, 24, 25, 27, 49, 289 water quality, 78 water resources, 4, 9, 24, 30, 32, 49, 50, 51, 67, 76, 77, 124, 160, 203, 204, 206, 209, 214, 215, 216, 218, 221 water rights, 27, 30, 31, 204, 210 water supplies, 33, 204

Index

Copyright © 2009. Nova Science Publishers, Incorporated. All rights reserved.

water table, 129, 305 water vapour, 50, 56, 57, 58 watersheds, 265, 301 watertable, 214, 218, 219 weakness, 110 wealth, 204 web, 6, 7, 10, 16, 44, 45, 66, 95, 121, 130, 174, 175 web sites, 66 web-based, 95 websites, 29, 44, 93 wells, 29 West Africa, 74 Western Europe, 149, 161 wetlands, 26 wetting, 146 wheat, 36, 40, 62, 76, 77, 79, 84, 92, 99, 105, 125, 126, 131, 138, 139, 142, 149, 150, 152, 153, 154, 161, 168, 171, 180, 181, 183, 184, 187, 188, 189, 190, 191, 195, 196, 199, 200, 204, 215, 227, 228, 231, 232, 233, 265, 267, 269, 323, 331, 333, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346 wind, 35, 37, 38, 39, 59, 60, 61, 62, 63, 71, 77, 84, 95, 96, 97, 98, 187, 214, 216, 230, 260, 321, 322, 323, 324, 325, 326, 327, 328, 333, 334 wine, 235, 236, 237

361

winter, 35, 36, 37, 39, 40, 84, 90, 91, 109, 110, 113, 114, 125, 133, 139, 149, 171, 183, 184, 185, 186, 187, 188, 189, 191, 196, 199, 200, 206, 208, 213, 214, 217, 218, 219, 220, 221, 222, 230, 246, 267, 269, 273, 274, 292, 328, 331, 333, 335, 336, 337, 338, 339, 340, 341, 342, 343, 344, 345 wisdom, 178 workability, 271 workers, 169 working groups, 41 World Trade Organization (WTO), 49 writing, 74, 93

Y yield loss, 121, 184, 231, 232, 239, 247, 268, 278 Yugoslavia, 332, 345

Z Zea mays, 161, 252 zoning, 308