Coastal storms : processes and impacts 9781118937075, 1118937074, 9781118937099, 1118937090

Citation preview

Coastal Storms

Hydrometeorological Extreme Events Published Titles in the Series Hydrometeorological Hazards: Interfacing Science and Policy Edited by Philippe Quevauviller Forthcoming Titles in the Series Flash Floods Early Warning Systems: Policy and Practice by Daniel Sempere-Torres

Coastal Storms Processes and Impacts

Edited by

Paolo Ciavola University of Ferrara Giovanni Coco University of Auckland

This edition first published 2017 © 2017 John Wiley & Sons Ltd All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions. The right of Paolo Ciavola and Giovanni Coco to be identified as the editors of the editorial material in this work has been asserted in accordance with law. Registered Offices John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial Office 111 River Street, Hoboken, NJ 07030, USA 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com. Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears in standard print versions of this book may not be available in other formats. Limit of Liability/Disclaimer of Warranty The publisher and the authors make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for every situation. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of experimental reagents, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each chemical, piece of equipment, reagent, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. The fact that an organization or website is referred to in this work as a citation and/or potential source of further information does not mean that the author or the publisher endorses the information the organization or website may provide or recommendations it may make. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this works was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising here from. Library of Congress Cataloging-in-Publication data applied for ISBN: 9781118937105 Cover Design: Wiley Cover Image: © DonatellaTandelli/Gettyimages Set in 10/12pt, TimesTenLTStd by SPi Global, Chennai, India.

10 9 8 7 6 5 4 3 2 1

Contents

List of Contributors

xi

Series Foreword

xv

Introduction Acknowledgments

xvii xix

1

Coastal Storm Definition Mitchell Harley

1.1

Introduction 1.1.1 The challenge of defining coastal storms 1.1.2 A general coastal storm definition 1.1.3 Approaches to assessing coastal storminess Synoptic systems and coastal storms 1.2.1 Tropical cyclones 1.2.2 Extra-tropical cyclones 1.2.3 Storm surge Statistical approaches to identifying coastal storms 1.3.1 Coastal storm events from wave time-series 1.3.2 Coastal storm events from water-level time-series 1.3.3 Indicators of coastal storm severity Conclusion References

1 4 7 8 9 9 10 11 12 12 15 16 18 19

2

Hydrodynamics Under Storm Conditions Xavier Bertin, Maitane Olabarrieta and Robert McCall

23

2.1 2.2

General introduction Storm surges 2.2.1 Introduction 2.2.2 Governing equations Hydrodynamics of the surf zone during storms 2.3.1 Introduction 2.3.2 Longshore currents

23 23 23 24 31 31 31

1.2

1.3

1.4

2.3

1

vi

CONTENTS

2.3.3 Bed return flows 2.3.4 Infragravity waves 2.3.5 Swash zone dynamics Conclusions and future challenges Acknowledgements References

32 33 35 38 38 39

3

Sediment Transport Under Storm Conditions on Sandy Beaches Troels Aagaard and Aart Kroon

45

3.1 3.2 3.3 3.4 3.5 3.6

Introduction Morphological consequences of coastal storms Sediment transport processes during storms Observations of sediment transport on the upper shoreface during storm events Observations of sediment transport on the lower shoreface during storm events Conclusions Acknowledgements References

45 46 48 53 58 60 60 60

4

Examples of Storm Impacts on Barrier Islands Nathaniel Plant, Kara Doran and Hilary Stockdon

65

4.1 4.2 4.3 4.4 4.5

Introduction Barrier island response to storms Quantifying the changes due to specific storms Resilience Summary Acknowledgements References

65 66 70 75 76 77 77

5

Storm Impacts on the Morphology and Sedimentology of Open-coast Tidal Flats Ping Wang and Jun Cheng

81

5.1 5.2 5.3 5.4

Introduction Sedimentologic characteristics Erosion-deposition processes and morphodynamics of open-coast tidal flat Conclusions References

81 83 88 96 96

6

Storm Impacts on Cliffed Coastlines Sue Brooks and Tom Spencer

99

6.1 6.2 6.3 6.4 6.5

Introduction Methodologies and their application Storminess and the cliff record Case study: Soft rock cliff geology and responses to storms Modelling shoreline retreat for cliffed coasts and the incorporation of storminess Future storm impacts on clifflines under accelerated sea-level rise and changing storminess

2.4

6.6

99 104 106 110 115 117

CONTENTS

vii

6.7

Conclusions Acknowledgements References

119 119 119

7

Storms in Coral Reefs Ana Vila-Concejo and Paul Kench

127

7.1 7.2

Introduction Geomorphic units of reefs 7.2.1 Reefs as ecomorphodynamic structures 7.2.2 Unique interactions of storm waves with coral reefs Storms on the forereef: Role of spurs and grooves 7.3.1 Destructive effects of storms in the forereef and spur and groove 7.3.2 Constructive effects of storms in the forereef Storms on the reef flats: Development of rubble flats and rubble spits 7.4.1 Waves on the reef flats 7.4.2 Destructive effects of storms on reef flats 7.4.3 Constructive effects of storms on reef flats Storms on the backreef: Sand aprons, reef islands and beaches 7.5.1 Sand aprons 7.5.2 Reef islands Conclusion Acknowledgements References

127 129 130 132 134 135 136 136 136 136 137 139 139 139 145 145 145

8

Storm Clustering and Beach Response Nadia Sénéchal, Bruno Castelle and Karin R. Bryan

151

8.1 8.2

Introduction Storm clustering: Genesis and definitions 8.2.1 Genesis 8.2.2 Definitions Approaches used to assess storm clustering impact on coasts 8.3.1 Data collection 8.3.2 Numerical models Beach response to storm cluster 8.4.1 Bar dynamics under storm clustering 8.4.2 Morphological feedback 8.4.3 The dynamic equilibrium concept 8.4.4 Water level 8.4.5 Recovery periods Conclusions References

151 153 153 154 156 156 157 159 159 160 162 164 165 167 167

Overwash Processes: Lessons from Fieldwork and Laboratory Experiments Ana Matias and Gerhard Masselink

175

Introduction 9.1.1 Overwash definition 9.1.2 Occurrence of overwash 9.1.3 Relevance of overwash

175 175 177 180

7.3

7.4

7.5

7.6

8.3

8.4

8.5

9

9.1

viii

9.2

9.3

9.4

9.5

10

CONTENTS

Methods to study overwash processes 9.2.1 Fieldwork measurements 9.2.2 Laboratory experiments Hydrodynamic processes during overwash 9.3.1 Oceanographic conditions 9.3.2 Hydraulics of overwash flows Morpho-sedimentary dynamics by overwash processes 9.4.1 Morphological changes induced by overwash 9.4.2 Morphodynamic processes during overwash Conclusion Acknowledgements References

180 180 181 183 183 183 185 185 187 189 190 190

Modeling the Morphological Impacts of Coastal Storms Ap van Dongeren, Dano Roelvink, Robert McCall, Kees Nederhoff and Arnold van Rooijen

195

10.1

Introduction 10.1.1 Empirical models 10.1.2 Process-based models 10.1.3 Process-model applications 10.1.4 Operational models 10.2 Outlook Acknowledgements References

11

Preparing for the Impact of Coastal Storms: A Coastal Manager-oriented Approach José Jiménez, Clara Armaroli and Eva Bosom

11.1 11.2

195 196 197 201 209 209 210 210

217

Introduction Coastal vulnerability assessment framework 11.2.1 General framework 11.2.2 How to characterize storm-induced hazards 11.2.3 How to measure the vulnerability 11.2.4 How to select the probability to be analyzed 11.2.5 The Catalonia coastal vulnerability assessment framework 11.3 Coastal early warning systems 11.3.1 Generalities 11.3.2 Coastal EWSs 11.3.3 The Emilia-Romagna coastal early warning system 11.4 Conclusion Acknowledgements References

217 219 219 219 221 222 223 227 227 228 228 234 235 235

12

Assessing Storm Erosion Hazards Roshanka Ranasinghe and David Callaghan

241

12.1 12.2

Introduction The diagnostic conundrum

241 242

CONTENTS

12.3

Quantifying storm erosion volumes for coastal management/planning 12.3.1 Coastal profile model application with Extrapolated Wave Exceedance Characteristics (EWEC) 12.3.2 Coastal profile model application with the Synthetic Design Storm (SDS) approach 12.3.3 The Joint Probability Method (JPM) approach 12.3.4 Corbella and Stretch (CS) approach 12.4 Application of storm erosion volume estimates in coastal management/planning 12.5 Conclusions and recommendations Acknowledgments References

ix

243 243 245 246 248 250 251 254 254

Conclusions and Future Perspectives

257

Index

259

List of Contributors

Troels Aagaard Department of Geoscience and Natural Resources, University of Copenhagen, Copenhagen, Denmark Clara Armaroli Department of Physics and Earth Sciences, University of Ferrara, Ferrara, Italy Xavier Bertin UMR 7266 LIENSs, CNRS-Université de La Rochelle, La Rochelle, France Eva Bosom Laboratori d’Enginyeria Marítima, Universitat Politécnica de Catalunya Barcelona Tech, Barcelona, Spain Sue Brooks Department of Geography, Environment and Development Studies Birkbeck, University of London, UK Karin R. Bryan School of Science, University of Waikato, Hamilton, New Zealand David Callaghan School of Civil Engineering, University of Queensland, Brisbane, Australia Bruno Castelle Univ. Bordeaux, UMR EPOC, Pessac, France CNRS, UMR EPOC, Pessac, France Jun Cheng School of Geosciences, University of South Florida, Tampa, USA Paolo Ciavola Dipartimento di Fisica e Scienze della Terra, Università di Ferrara, Ferrara, Italy

xii

LIST OF CONTRIBUTORS

Giovanni Coco School of Environment, University of Auckland, Auckland, New Zealand Ap van Dongeren Deltares, Delft, The Netherlands Kara Doran US Geological Survey, Saint Petersburg, Florida, USA Mitchell Harley Water Research Laboratory, School of Civil and Environmental Engineering, UNSW Sydney, Manly Vale, NSW, Australia Jose Jimenez Laboratori d’Enginyeria Marítima, Universitat Politécnica de Catalunya Barcelona Tech, Barcelona, Spain Paul Kench School of Environment, University of Auckland, New Zealand Aart Kroon Department of Geoscience and Natural Resources, University of Copenhagen, Copenhagen, Denmark Gerhard Masselink School of Marine Science and Engineering, University of Plymouth, Plymouth, UK Ana Matias Centro de Investigação Marinha e Ambiental (CIMA), Universidade do Algarve, Portugal Robert McCall Deltares, Delft, The Netherlands Melisa Menéndez Grupo de Clima Marino y Cambio Climático IH Cantabria, Universidad de Cantabria Avda, Santander, Spain Kees Nederhoff Deltares, Delft, The Netherlands Maitane Olabarrieta Civil and Coastal Engineering Department, ESSIE, University of Florida, Gainesville, Florida, USA Nathaniel Plant US Geological Survey, Saint Petersburg, Florida, USA

LIST OF CONTRIBUTORS

xiii

Roshanka Ranasinghe UNESCO-IHE Institute, Delft, The Netherlands Dano Roelvink Deltares, Delft, The Netherlands UNESCO-IHE Institute, Delft, The Netherlands Arnold van Rooijen Deltares, Delft, The Netherlands University of Western Australia, Crawley, WA, Australia Nadia Sénéchal Univ. Bordeaux, UMR EPOC, Pessac, France CNRS, UMR EPOC, Pessac, France Tom Spencer Cambridge Coastal Research Unit, University of Cambridge, UK Hilary Stockdon US Geological Survey, Saint Petersburg, Florida, USA Ana Vila-Concejo Geocoastal Research Group, School of Geosciences, The University of Sydney, NSW, Australia Ping Wang School of Geosciences, University of South Florida, Tampa, USA

Series Foreword

The increasing frequency and severity of hydrometeorological extreme events are reported in many studies and surveys, including the 5th IPCC Assessment Report. This report and other sources highlight the increasing probability that these events are partly driven by climate change, while other causes are linked to the increased exposure and vulnerability of societies in exposed areas (which are not only due to climate change but also to mismanagement of risks and ‘lost memories’ about them). Efforts are ongoing to enhance today’s forecasting, prediction and early warning capabilities in order to improve the assessment of vulnerability and risks and develop adequate prevention, mitigation and preparedness measures. The Book Series on ‘Hydrometeorological Extreme Events’ has the ambition to gather available knowledge in this area, taking stock of research and policy developments at an international level. While individual publications exist on specific hazards, the proposed series is the first of its kind to propose an enlarged coverage of various extreme events that are generally studied by different (not necessarily interconnected) research teams. The series encompasses several volumes dealing with various aspects of hydrometeorological extreme events, primarily discussing science – policy interfacing issues, and developing specific discussions about floods, coastal storms (including storm surges), droughts, resilience and adaptation. While the books are looking at the crisis management cycle as a whole, the focus of the discussions is generally oriented towards the knowledge base of the different events, prevention and preparedness, early warning, and improved prediction systems. The involvement of internationally renowned scientists (from different horizons and disciplines) behind the knowledge base of hydrometeorological events makes this series unique in this respect. The overall series will provide a multidisciplinary description of various scientific and policy features concerning hydrometeorological extreme events, as written by authors from different countries, making it a truly international book series. The book, Prevention of Hydrometeorological Extreme Events – Interfacing Sciences and Policies is the first book of this series; it has been written by policy-makers and scientific experts in the field. It offers the reader an overview of EU international policies, discussions on science – policy interfacing, and a snapshot of the knowledge base of various types of events which are developed in separate volumes of the series. Philippe Quevauviller Series Editor

Introduction Paolo Ciavola and Giovanni Coco

Coastal storms can be one of the most destructive natural hazards. In coastal cities, they can disrupt activities and affect large parts of the population; they can also cause major economic damage and often pose a threat to human lives. The problem of understanding the physical processes operating during a storm and predicting their impact is relevant for scientists and has clear societal implications. Here, we focus on some specific aspects of coastal storms, from inundation to the morphological changes along the coastline. An understanding of this is becoming increasingly relevant because of the ongoing climatic changes and the ever increasing population pressure along coastlines. We have tried to provide a textbook that we hope will be useful to advanced undergraduate and graduate students in variety of fields, ranging from ocean sciences to geomorphology, coastal engineering and geophysics. We decided to split the book in two parts. In the first part, we asked authors to provide a general overview of the present understanding of storms. In the second part we looked more closely at how storms impact different natural systems. In the first part, the definition of a ‘storm’ is addressed (Chapter 1) and detailed reviews of processes controlling hydrodynamics (Chapter 2), sediment transport (Chapter 3) and overwash processes (Chapter 4) under storm conditions are provided. The reader is then ready to tackle an understanding of how storms impact a variety of geomorphic landscapes, from barrier islands (Chapter 5) to cliffed coastlines (Chapter 6), tidal flats (Chapter 7) and coral reefs (Chapter 8). We also decided that a specific chapter should be dedicated to the role of storm clustering (Chapter 9) and to the most up-to-date advances on the numerical modelling of storm dynamics and effects (Chapter 10). The final chapters focus on the societal aspects of storms and show how to develop frameworks to assess hazards (Chapter 11) and risk management (Chapter 12). We asked some of the most well-known scientists in the field to help us provide this overview on coastal storms by writing individual chapters. On several occasions the chapters report knowledge gained by the authors during years of research on their topic of expertise, developed with financial support from research agencies in Europe, USA, Australia and New Zealand. We are hugely indebted to the authors, it has been a privilege to share their passion for research and their effort to promote science. Finally, while reading the chapters, it will appear evident that there are still many poorly understood issues that require attention. Research on this topic is still constrained

xviii

INTRODUCTION

by a limited understanding of the analogies between theoretical process and natural system behaviour during extreme forcing. Field measurements still remain scarce, as acquisition of pre- and post-storm datasets requires quick and costly deployment of state-of-the-art equipment. We hope this book will stimulate scientists to advance knowledge on coastal storms and contribute to a better planning of measures to increase resilience of coastal communities.

Acknowledgments

This book is the result of many years of research and fruitful collaborations with scientists worldwide. None of this could have happened without funding from a number of agencies in a variety of countries. Here, we wish to specifically acknowledge the role of the EU in promoting research on coastal storm processes and impacts during the Seventh Research Framework (e.g. MICORE and RISC-KIT projects). The ongoing support from the EU-RISC-KIT Project-grant 603458 (Paolo Ciavola) and the MBIE-GNS Hazard Platform (Giovanni Coco) is gratefully acknowledged. Paolo Ciavola would also like to express gratitude to his wife, Clara, and his son, Leonardo, for putting up with late nights trying to bridge the time zone with New Zealand. Together we have seen plenty of beaches around the world and not many mountains. Giovanni Coco would like to thank Mattia for reminding him that beaches are for fun, not for work. Thanks to Jennifer Montano for spotting the final typos. Finally a little story about the editors of this book. We both grew up in Catania in Italy, on the foothills of the largest volcano in Europe. We both went to scientific high schools there, only a few kilometres from each other. But we never met at the time when we were living on the island. Once again, we studied different topics in different places in Italy and abroad, and destiny did not bring us together. Then life took us around the world and finally science enabled us to meet and discover another fellow who liked the idea of a book on storm processes. Some would call this serendipity, but without doubt it shows us the power of research in bringing people together, whatever their country, belief and personal opinions are.

1 Coastal Storm Definition Mitchell Harley Water Research Laboratory, School of Civil and Environmental Engineering, UNSW Sydney, Manly Vale, NSW, Australia

1.1

Introduction

Storms represent nature in one of its most energetic and violent states. The word “storm” is synonymous with images of destruction – strong winds lashing at trees and buildings, intense precipitation flooding towns or dumping meters of snow, large seas eroding beaches and coastal properties, and rapid surges in ocean levels inundating entire islands and vast lowland areas. At the same time, storms are essential to human life and an integral part of the global weather and natural ecosystems. Storms help break droughts by delivering much needed water to drought-stricken areas, thereby recharging reservoirs, river systems and underground aquifers. Many ecosystems are also reliant on the episodic arrival of large storms for their rejuvenation after extended periods of calm, stable conditions (e.g. the flushing of hypersaline lagoons due to hurricanes, Tunnell, 2002). Globally, storms rank as one of the deadliest of all natural hazards (International Federation of Red Cross and Red Crescent Societies, 2014). In the decade spanning the years 2004–2013, storms were responsible for over 180,000 deaths worldwide – second in terms of lives lost only to those of earthquakes and tsunamis (Table 1.1). Flooding, including marine flooding as a result of waves and storm surge, were meanwhile responsible for over 60,000 deaths worldwide and rank fourth on this list. In the United States, storms have contributed to the vast majority of monetary losses resulting from natural hazards over the last half century. Hurricanes and tropical storms alone have overwhelmingly been the most costly of all natural hazards, having resulted in a total of US$ 267 billion in monetary losses between the years 1960 and 2014 (Figure 1.1). Severe weather, flooding, tornadoes and miscellaneous coastal hazards (loosely defined as hazards including rip currents, coastal flooding, coastal erosion, strong winds, etc.) have also caused combined losses of US$ 364 billion (Hazards and Vulnerability Research Institute, 2015). There are few regions more vulnerable to storms than the narrow ribbon of the Earth’s surface that constitutes the coastal zone. Situated at the interface between land and large water bodies such as oceans, seas and lakes, the coastal zone is a region in constant flux as consolidated and unconsolidated sediments are constantly shaped and re-shaped by Earth’s forces. As these forces – winds, waves and currents – interact with Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

2

COASTAL STORMS: PROCESSES AND IMPACTS

Table 1.1 Total number of people killed globally by natural disasters between 2004 and 2013 according to disaster type. Rank 1 2 3 4 5 6 7 8 9

Disaster Type

Total number of people killed

Earthquakes/tsunamis Storms Extreme temperatures Floods∗ Mass movement: wet Forest/scrub fires Droughts/food insecurity Volcanic eruptions Mass movement: dry Total

650,321 183,457 72,088 63,207 8,739 705 384 363 273 979,537

∗ includes wave and surge events (Source: International Federation of Red Cross and Red Crescent Societies, 2014, p. 226)

U.S. Hazard Losses 1960–2014 (US$ 2014 billion) $267

$167 $136 $70

$58

al oa

st

er C

re ve

Se

ug

w

an

d

ea

th

he

at

er th ea

ht

te in W

ro

nd

sl

id

G

e/

eo

Av

rw

al

ph

an

ys

ild

ic

ch

e

al

e fir

o W

ad

$3

La

ic op Tr

e/ an

rn

or al

st

di oo Fl

ric

$11

H

ur

To

ng

e ak qu rth

m

$22

Ea

$58

$48

D

$57

Figure 1.1 Total hazard losses in the United States (1960–2014) by hazard type (Source: Hazards and Vulnerability Research Institute, 2015).

coastal sediments, energy is dissipated to such a degree that under normal everyday conditions, their short-term effects on the adjacent coastal hinterland are minimal. During destructive storm conditions, however, the elevated energy and/or water levels may well be beyond the capacity of the coastal zone to dissipate, potentially exposing the backshore and coastal hinterland to unusually large forces and hazardous conditions. Given the low-lying nature as well as the sheer density of people living close to the coast (with an estimated 23% of the world’s population and population densities greater than three times the global average, Small & Nicholls, 2003), the exposure to elevated

CH1 COASTAL STORM DEFINITION

3

water levels, waves and currents that may occur during storm conditions can have devastating effects. Some historical examples of extreme storms striking the coast include the 1900 hurricane in Galveston, Texas that claimed the lives of an estimated 8000–12,000 people and is recognized even today as the deadliest natural disaster in the United States’ history (Blake & Gibney, 2011). In 1953, a large storm surge in the North Sea inundated tens of thousands of hectares of coastal hinterland in the Netherlands, Belgium and the United Kingdom and claimed over 2500 lives. In Bangladesh, the Bhola cyclone of 1970 is considered one of the worst natural disasters of all time, generating a 10 m storm surge that killed up to 500,000 people and left a huge toll on the country’s population and economy. Such devastation was repeated in the same region 21 years later, when another tropical cyclone caused a surge that extended 160 km inland and resulted in 138,000 deaths (Haque, 1997). In more recent years, coastal storms have received considerable attention as access to news and information via the Internet has grown exponentially and the world has become more aware of the dangers associated with climate change. A particularly significant event that has remained in the conscience of many people was that of Hurricane Katrina that struck the Louisiana coastline in 2005. Hurricane Katrina demonstrated that even in an age of significant advancements in scientific understanding, technology and computer forecasts, nations can still be caught off-guard by the arrival of coastal storms. Hurricane Katrina also highlighted that when coastal storms do occur, it is often the most vulnerable people of a society that are affected the most (Laska & Morrow, 2006). Some other recent examples of coastal storms include Cyclone Sidr in Bangladesh (2007), the Xynthia cyclone in France (2010), Hurricane Sandy in the Caribbean, New Jersey and New York (2012), Typhoon Haiyan in the Philippines (2013), the 2013/2014 winter storms in the United Kingdom and Tropical Cyclone Pam in Vanuatu (2015). Figure 1.2 indicates a rare occurrence of three concurrent tropical cyclones close to the coastline that was observed in southern hemisphere waters in March 2015. Considering their destructiveness and relevance to today’s world, surprisingly few books have dealt specifically with the subject of coastal storms and no overarching definition presently exists to assist in their identification. Indeed a degree of confusion surrounding the use of the term coastal storm is evident. An inspection of Table 1.1, for example, indicates that coastal storms fall into the category of both storms and floods, but are not recognized as a category on their own. This is in spite of the fact that the processes governing the formation and development of coastal storms are very different from those of, for instance, river floods. Figure 1.1, meanwhile, highlights the variety of ways in which coastal storms are classified in the commonly-used SHELDUS database for US disaster statistics, with hurricane/tropical storms and coastal hazards treated separately. As this chapter discusses, the lack of clarity when it comes to defining coastal storms stems from the complexities surrounding the ways in which storm energy is generated, transported and interacts with the coastline. A robust definition of a coastal storm is, however, necessary if we want to answer important societal questions, such as: • • •

How vulnerable are coastal communities and ecosystems to coastal storms? Are coastal storms becoming more frequent or increasing in magnitude? What influence is climate change having on coastal storms?

4

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 1.2 A composite image taken from the NASA of three tropical cyclones occurring simultaneously in the southern hemisphere in March, 2015. Tropical Cyclone Pam to the right of the image struck the island of Vanuatu and is considered one of the worst natural disasters in the island’s history (Source: NASA Earth Observatory: http://earthobservatory.nasa.gov/).

• How near to the coast can we safely build infrastructure away from the influence of coastal storms? • How can we design coastal structures to withstand coastal storm forces? This chapter begins by first summarizing the challenges of defining coastal storms. These challenges are then taken into consideration to form a general qualitative coastal storm definition that can be applied to all coastlines. Section 1.2 follows by describing the most common synoptic conditions associated with coastal storms. Section 1.3 then presents the various approaches taken to identify coastal storm events from observational records and summarizes ways of quantifying coastal storm severity.

1.1.1 The challenge of defining coastal storms The term storm is defined as: “a disturbance of the atmosphere marked by winds and usually by rain, snow, hail, sleet, or thunder and lightning” (Merriam-Webster, 2015). When it comes to defining coastal storms, however, a simple application of this definition to the coastal zone is not sufficient. Although storms and coastal storms go hand in hand – indeed a coastal storm cannot take place without some sort of atmospheric disturbance occurring somewhere – there are several important features about coastal storms that make them unique from other storm types (thunderstorms, and snowstorms, for instance) and hence particularly challenging to define and categorize. These features include:

CH1 COASTAL STORM DEFINITION

• • • •

5

The location of the atmospheric disturbance relative to the impact area The diversity of environments in which coastal storms occur The ways in which these diverse environments respond to the same environmental forcing The timing and duration of the storm

By their very nature, coastal storms must comprise a maritime component, such as waves, currents and/or water levels. One of the first inroads into defining storms in a maritime setting dates back to the works of the British Royal Navy officer, Sir Francis Beaufort, in the early 1800s. Since the methods of reporting weather and sea conditions in ship logs were very subjective at the time, Beaufort proposed a system of standardizing these observations by using consistent language according to a thirteen-point scale. This system has since evolved to what is now known as the Beaufort Scale, and describes the relationship between wind speed and the associated sea state on the open ocean. Storms according to the modern version of the Beaufort Scale have a ranking (or Beaufort number) of 10 or more, which translates to wind speeds of at least 24.5 m/s (approximately 88 km/h). In terms of sea state, the Beaufort Scale specifically identifies storm conditions using the following terminology: “Very high waves with overhanging crests; sea-surface takes on white appearance as foam in great patches is blown in very dense streaks; rolling of sea is heavy and visibility is reduced” (Wright et al., 1999). While the Beaufort Scale serves as a practical guide for use in weather forecasts and seafaring, its application to storms on the coastline is somewhat limited. This is because sea states in this scale are based on the idealized concept of the “fully-developed sea”, which is the equilibrium sea state that develops when winds blow on the open ocean according to a number of conditions: (1) a constant wind speed; (2) a constant wind direction; (3) over a sufficiently-long fetch; and (4) for a sufficiently-long period of time. In reality these conditions are almost never met and, close to the coastline, wind and wave fields become increasingly influenced by the coastal boundary. Ultimately this means that coastal waves are usually much smaller and steeper than those found on the open ocean. The modification of storm waves as they move from deep to shallow water is described in detail in Chapter 2. Another limitation of the Beaufort Scale is that it focuses on the local sea conditions and does not include the influence of swell – waves that have moved outside their area of wave generation and travel freely across the ocean surface. To take into account the additional influences of swell waves, Sir Percy Douglas devised a scale in the 1920s of describing the overall state of the sea. This scale, which is known nowadays as the Douglas Scale, is based on a two-digit code system. The first digit describes the local sea state from 0 to 9 (with 0 being the least energetic and 9 the most energetic sea state) and the second digit describing the swell conditions, also from 0 to 9. The nature of swell waves means that it is possible to stand on the coastline and observe large waves arriving from a distant storm even when there are clear blue skies in the area (as illustrated in Figure 1.3). In terms of coastal storms, this means that the local atmospheric conditions do not necessarily influence the storm itself. As sea and swell waves approach the shore, the type of coastal environment plays a major role in determining how the coastline responds. For coastlines usually susceptible to low-energy wave conditions (such as those found in bays and estuarine environments), even relatively small waves can induce significant changes to the coastal

6

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 1.3 Large swell waves arriving on a coastline during a sunny day with clear skies (photo: A.D. Short).

environment. In Delaware Bay on the northeast seaboard of the USA, for example, the average height of waves reaching the shore is in the order of just 0.1 m. In such a low energy environment, wave heights of a mere 0.5 m can for all intents and purposes be considered “storm waves” that leave a long-lasting signature on the beach profile (Jackson et al., 2002). At the other end of the spectrum, high-energy coastlines such as those found in southwest Tasmania (south of mainland Australia) are exposed to year-round significant wave heights in excess of 2.5 m (Hemer et al., 2008). On these types of coastlines, the same 0.5 m wave from the previous example is easily absorbed by the coastal environment and has negligible impact at the shoreline. Hence the magnitude of the waves relative to the modal or equilibrium wave conditions helps determine the magnitude of coastal response. While up until now we have focused on wave conditions, another critical component to coastal storms is the water level at the coastal boundary. The total water level (TWL) represents the sum of both astronomical tides and non-tidal residuals, also known as tidal anomalies. Non-tidal residuals can occur on the coastline due to a variety of factors, such as storm surge (discussed in detail in Chapter 2), wave set-up, basin seiching, complex tide-surge interactions and freshwater input. The TWL can be expressed as a function of time according to the equation: TWL (t) = Z𝟎 (t) + T (t) + R (t)

(1.1)

CH1 COASTAL STORM DEFINITION

7

where Z0 (t) is the mean sea-level (which varies over longer time-scales), T (t) is the tidal component and R (t) represents non-tidal residuals (Pugh, 1987). The TWL is a critical factor for erosion of barrier islands (discussed in Chapter 4) as well as for overwash and inundation on low-lying coasts (discussed in Chapter 9). Finally, the timing and duration of a storm event is also an important factor contributing to the identification of a coastal storm. Depending on the stage of the tidal cycle in which a storm occurs (i.e. high vs low tides, spring vs neap tides), tidal anomalies can result in either exceptionally high total water levels, or levels typical of everyday tidal fluctuations. Likewise, large waves that occur at high tides are more likely to cause impacts than those at low tide. Storm duration is also important as it determines the timescales over which sediment can be transported from its pre-storm location, as well as increasing the probability that a storm occurs at higher tide levels.

1.1.2 A general coastal storm definition By taking the above factors into consideration, we can now define a coastal storm in a broad sense as a: “meteorologically-induced disturbance to the local maritime conditions (i.e. waves and/or water levels) that has the potential to significantly alter the underlying morphology and expose the backshore to waves, currents and/or inundation”. They are usually associated with the passage of cyclonic systems such as tropical or extra-tropical cyclones (discussed in section 1.2), which can strike the coastline directly or track at a sufficient distance from the coastline to influence the local maritime conditions. Coastal storms may also (but not necessarily) coincide with strong winds and/or precipitation that, in conjunction with the anomalous maritime conditions, can contribute to a storm’s severity. The disturbance to the local maritime conditions that occurs during a coastal storm must be of sufficient magnitude that the underlying morphology (sandbars, coral reefs, etc.) can be significantly altered from its modal or everyday form. This means that in the absence of human interventions (e.g. the construction of breakwaters, temporary protection barriers, etc.), or abnormal antecedent conditions, the morphology is transformed in such a way that a period of recovery ensues. This recovery period, whereby the storm-altered morphology returns to a modal form associated with non-storm conditions, usually transpires over a time-scale larger than the storm itself. In the case of storm clusters (see Chapter 8), the recovery period may be unusually prolonged as storms continue to interrupt or reverse the recovery process. For extreme coastal storms, it is possible that the coastal zone may never recover and a new equilibrium state associated with these higher-energy conditions is reached. Alternatively or in conjunction with the above impacts to the underlying morphology, the coastal backshore – which under normal conditions is protected from waves, currents and inundation – can in the event of a coastal storm (and without human intervention), be suddenly exposed to these processes. This exposure may occur over a brief period of time during the actual coastal storm event, or in the case of particularly severe events, remain for many weeks and months after the event itself has subsided. Depending on factors such as local topography and the severity of the event, this exposure may be sufficient to erode or overtop backshore dunes and inundate the adjacent coastal hinterland.

8

COASTAL STORMS: PROCESSES AND IMPACTS

1.1.3 Approaches to assessing coastal storminess A key question to understanding the degree of exposure to coastal storms for a particular coastal region is to examine temporal patterns and trends of storm arrivals – in other words, the storminess of a coastal region. Assessments of coastal storminess may include studies such as: • The frequency of coastal storms arriving over a given year • The timing of storm arrivals (e.g. regular/irregular intervals, storm clusters, seasonal patterns) • Teleconnections with large-scale climate patterns (e.g. the El Niño/Southern Oscillation, the North Atlantic Oscillation) • Directional shifts in coastal storms • Trends in coastal storm extremes • The effects of climate change The various ways taken to assess the storminess of a coastal region can be summarized into two main approaches: (1) a synoptic climatological approach and (2) a statistical approach. The synoptic climatological approach to assessing coastal storminess involves the pairing of regional synoptic information such as storm tracks and sea level pressure data with coastal-based observations (instrumental records, hindcast data, historical newspaper reports, etc.) for the particular coastal region of interest. As outlined by Yarnal (1993), depending on which set of paired information is analyzed first, this approach can either be described as a circulation-to-environment approach or an environment-to-circulation approach. In the circulation-to-environment approach, a synoptic classification of the region is first undertaken independently of the coastal response. This information is in turn compared to the coastal-based observations, in order to gain an understanding of the relationships between regional-scale storminess and localized coastal response. In the environment-to-circulation method, the coastal-based observations are analyzed first and only the synoptic configurations associated with the extreme local maritime disturbances are determined. Some examples of synoptic climatological approaches to assessing coastal storminess include that of Mather et al. (1964) for the eastern seaboard of the United States, Short and Trenaman (1992) for the southeast coastline of Australia, Betts et al. (2010) for the Atlantic coast of Northern France and Lionello et al. (2012) for the Adriatic coastline of Northern Italy. The statistical approach to assessing coastal storminess, on the other hand, focuses purely on the coastal-based observational data and uses statistical methods to separate individual storm events from quiescent or non-storm periods. Statistical methods used to separate these periods are described in detail for wave and water-level time-series datasets in section 1.3. While the statistical approach does not provide an overall understanding of the meteorological processes associated with coastal storms (as the synoptic climatological approach does), it is objective and capable of quantifying coastal storm variability across entire regions and basins. Some examples of the statistical approach include coastal storminess assessments for the coastline of Sydney, Australia (Harley et al., 2010), the Gulf of Cadiz in Spain/Portugal (Plomaritis et al., 2015) and for the entire Southern European region (Cid et al., 2015).

CH1 COASTAL STORM DEFINITION

1.2

9

Synoptic systems and coastal storms

A fundamental aspect with regards to coastal storms that distinguishes them from other extreme events (e.g. tsunamis) is that they are ultimately generated by atmospheric disturbances over a water body. Although these disturbances may occur in a variety of forms, two main synoptic systems are responsible for the vast majority of coastal storms worldwide – tropical cyclones and extra-tropical cyclones. Depending on a number of factors, related to both the cyclonic system as well as the coastal setting in which it occurs, these systems may also generate storm surge.

1.2.1 Tropical cyclones There are few things more awe-inspiring and destructive than that of a tropical cyclone (TC) impacting on a coastal area. Tropical cyclones are intense low-pressure systems that consist of strong winds spiraling around a warm central core. They go by a variety of names according to their wind strength and the region of the world in which they occur. A TC with sustained one-minute wind speeds at the surface between 17 and 32 m/s is known as a tropical storm, whereas those with sustained wind speeds in excess of 32 m/s are known as hurricanes in the north Atlantic and northeastern Pacific, typhoons in the northwestern Pacific and tropical cyclones elsewhere. The general term tropical cyclone is typically used collectively to represent all of these cyclonic system types. Tropical cyclones derive their energy primarily from the evaporation of water off the sea surface and hence are formed in areas where the water temperature is sufficiently high (typically greater than 26.5∘ C). Other conditions that favor TC growth are a sufficiently strong Coriolis force (i.e. at latitudes of 5∘ of greater) and weak vertical wind shears between the lower and upper troposphere (Wallace & Hobbs, 2006). Structurally, a TC cell consists of a central core approximately 30–50 km in diameter, known as the eye, within which the surface air is calm and has minimal cloud cover. Adjacent to the eye is the eye wall – a circular rim of intense thunderstorms where the strongest surface winds of the entire storm cell are situated. Moving further away from the center, the outer structure consists of a series of rain bands (smaller cells of converging and diverging air) and the surface winds gradually diminish in strength until they reach ambient conditions. A typical diameter for the entire TC cell is in the order of 650 km. Figure 1.4 illustrates the spatial distribution of TC tracks worldwide according to their severity, based on the Saffir-Simpson scale (a measure of TC severity discussed in section 1.3.3). On average there are 80–90 TCs per year globally (Marks, 2003), most of which are concentrated in several regions of pronounced TC activity. The most active region in terms of annual rates of TC formation is the northwestern Pacific, which includes the coastlines of Japan, Taiwan and the Philippines and has an average of 27 TCs per year. Other active regions include the northeastern Pacific (17 TCs/year), the southwestern Indian region (12 TCs/year) and the northwestern Pacific and the Australian/southeast Indian regions (each with 10 TCs/year). The path which TCs take is steered by a combination of external factors as well as internal dynamics of the TC itself (Ahrens, 2000). This path is difficult to predict, but, in the absence of environmental steering, tends to move in a polewards and westwards direction (Marks, 2003). A TC is said to make landfall when the eye of the TC crosses a

10

COASTAL STORMS: PROCESSES AND IMPACTS

Saffir-Simpson Hurricane Scale: Tropical depression

Tropical strom

Hurricane category 1

Hurricane category 2

Hurricane category 3

Hurricane category 4

Hurricane category 5

Figure 1.4 Spatial distribution of tropical cyclone storm tracks (1946–2006) and their intensities according to the Saffir-Simpson Hurricane Scale (Source: Radical cartography: www.radical cartography.net).

coastal boundary. In this case the system loses its main source of energy and, in combination with the increased friction over land, the system rapidly dies out. Tropical cyclones may also die out when they move polewards into cooler mid-latitude waters. If a coastal location is subjected directly to the maximum radial winds of the cyclone (but the TC hasn’t necessarily made landfall), the location it is said to have taken a direct hit.

1.2.2 Extra-tropical cyclones Extra-tropical cyclones (ETCs, or mid-latitude cyclones) reflect a broad class of cold-core cyclones, including lows, depressions and frontal systems that draw their energy from temperature gradients in the atmosphere (such as those that occur between warm and cold air masses). They form predominantly in the mid-latitudes between 30∘ and 60∘ and, unless interrupted by other synoptic systems, tend to follow a zonal west-east path across the globe. The strongest ETCs usually occur during the winter months when atmospheric temperature differences are most pronounced (May et al., 2013). In comparison to TCs, ETCs are usually much larger (with length scales in the order of 2000 km) and slower-moving systems. Extra-tropical cyclones are also much more widespread and frequent than TCs, with 234 ETCs forming on average over the northern hemisphere winter (Gulev et al., 2001) and some 2500–2900 ETCs forming annually in the southern hemisphere (Simmonds & Keay, 1999). While ETCs typically have lower surface wind speeds than TCs, their slow-moving nature and size means that they are capable of affecting vast swathes of coastline and linger offshore for extended periods of time. Extra-tropical cyclones are subsequently

CH1 COASTAL STORM DEFINITION

11

capable of coastal impacts comparable or even greater in severity than those of TCs (Zhang et al., 2000). Some common examples of coastal storms generated by ETCs are nor’easters that occur on the eastern seaboard of the USA (e.g. Dolan & Davis, 1992), winter windstorms in Europe (e.g. Kolen et al., 2013) and east-coast lows in southeastern Australia (e.g. Browning & Goodwin, 2013). Figure 1.5 presents an example of the erosive potential of extra-tropical cyclones for a storm that struck the SE Australian coastline in June 2007. This event, which is known as the Pasha Bulker storm due to the 40,000 ton bulk carrier that ran aground in the resulting storm conditions, was an east-coast low that caused onshore wind gusts of up to 37 m/s and waves reaching a significant wave height of 6.9 m. The coastal impacts from this storm were significant: daily shoreline measurements from a video monitoring station installed permanently at Narrabeen-Collaroy Beach (lower-left panel, Figure 1.5) indicate that the average width of the beach retreated 29 m during the storm event, as sand was rapidly removed from the subaerial beach and deposited offshore. The ensuing recovery period for this event took approximately ten months (Phillips et al., 2015).

1.2.3 Storm surge Storm surge refers to the sudden increase in water levels associated with certain coastal storms that can have catastrophic consequences for low-lying coastlines. This rapid

Figure 1.5 Pre- and post-storm shoreline measurements following the Pasha Bulker storm, an extra-tropical cyclone that struck the coastline of southeastern Australia in June 2007. A video monitoring station installed permanently at Narrabeen-Collaroy Beach (see Harley et al., 2011) measured 29 m of rapid retreat in beach width as a result of the storm.

12

COASTAL STORMS: PROCESSES AND IMPACTS

onset of high water levels is considered the most destructive component of TCs and, to a lesser extent, ETCs. The degree of storm surge resulting from a coastal storm, however, is a complex process and depends on interactions between meteorological factors and the coastal setting in which they occur. Meteorological factors include the radial wind speeds of the cyclone, the cyclone’s central pressure and the forward speed of the cyclone system. Influences of the coastal setting, meanwhile, include the angle of cyclone approach relative to the coastline, the width and slope of the coastal shelf, as well as local features (National Weather Service, 2013). Storm surge is discussed in extensive detail in Chapter 2.

1.3 Statistical approaches to identifying coastal storms Statistical approaches to identifying coastal storms involve the analysis of wave or water-level time-series from an appropriate maritime location close to the site of interest. For wave-dominated coastlines, this analysis is typically conducted on time-series of the significant wave height. For coastal sites where meteorologically-driven increases in the water level beyond the range of usual tidal variability (i.e. greater than mean high water spring) are more significant, coastal storms can, meanwhile, be defined from measured water-level time-series. The various approaches to identifying storm events from these two data types are outlined below.

1.3.1 Coastal storm events from wave time-series As long-term wave measurement and hindcast datasets become increasingly available, a common means of identifying coastal storm events for a particular coastal location is through statistical analysis of the significant wave height (Hsig ) time-series. The identification of coastal storms from Hsig time-series is usually undertaken through the application of the so-called peaks-over-threshold (POT) method. The POT method has its origins in extreme value analysis, where it is used as a robust way of extracting data subsets for estimating the return values of environmental variables, for example the 50-year design wave for coastal structure design. As the name suggests, the POT method obtains a set of peak values from data clusters above a certain threshold level. In extreme value analysis, these data clusters and associated peak values are generally used to fit a Poisson – Generalized Pareto distribution. For coastal storm identification approaches, meanwhile, these data clusters represent the actual storm events. As shown in Figure 1.6, storm events can be identified by the POT method through the specification of three parameters: 1. 2. 3.

The storm threshold (Hthresh ) The minimum storm duration (D) The meteorological independence criterion (I)

The storm threshold is defined as the critical value that separates storm waves from non-storm waves for a particular coastal site. The storm duration, meanwhile, is defined by the length of time between an up-crossing and subsequent down-crossing of the

CH1 COASTAL STORM DEFINITION

13

P

Hsig D

I

Hthresh

t

Figure 1.6 The Peaks-Over-Threshold (POT) method for defining individual storm events from a significant wave-height time-series. P denotes the peak significant wave height of the storm, D the storm duration, I the meteorological independence criterion and Hthresh the threshold signignificant wave height. Individual storm events classified by this method are shaded gray.

storm threshold. Since wave heights can fluctuate for a brief period of time above (and below) this threshold, a minimum storm duration D is set to include only storm events of a significant duration. The final parameter is the meteorological independence criterion, which is a value that restricts the period of time between individual storm events and hence ensures that they are generated by independent synoptic systems such as a particular tropical or extra-tropical cyclone. The meteorological independence criterion also ensures that brief crossings below the storm threshold during a single storm event are included within the same event. Table 1.2 presents an overview of different values adopted of the three parameters discussed above for different coastal settings worldwide. From this table it is clear that the chosen values of these three parameters fluctuate greatly from site to site and that no standardized method currently exists to aid in their selection. Despite the absence of a standard method, some general guidelines can be ascertained. In terms of the storm threshold (the most critical of the three parameters), for reasons discussed in section 1.1.1, Hthresh is strongly related to the modal wave conditions of the site (represented in Table 1.2 by the average significant wave height). While statisticians argue that the threshold should be set according to a goodness of fit to the Generalised Pareto distribution (e.g. Mazas & Hamm, 2011), a more pragmatic approach related specifically to coastal storm analyses is to simply set the threshold according to the 95th percentile of the significant wave height dataset. This approach intrinsically takes into account the modal wave conditions and has been applied to wave-dominated coastlines in both the United Kingdom (e.g. Masselink et al., 2014) and France (e.g. Castelle at al., 2015). In terms of values for the minimum duration and meteorological independence criterion, an understanding of the local setting and regional meteorology is necessary. Tropical cyclones, for instance, are generally faster-moving systems and hence may afford a smaller time period (e.g. 12 hours) to differentiate individual coastal storm events. Slower-moving extra-tropical cyclones, on the other hand, may warrant a larger time gap (e.g. 24–72 hours) to distinguish between events. For complex cases such as a TC transitioning into an ETC, a careful (i.e. manual) selection of the meteorological independence criterion may be required.

14

COASTAL STORMS: PROCESSES AND IMPACTS

Table 1.2 Overview of different storm classifications based on significant wave heights and the peaks-over-threshold method Site

Lake Huron, Canada

Average significant wave height 0.4 m (summer) 1.0 m (winter)

Wave height threshold (Hthresh )

Minimum storm Meteorological duration independence (D) criterion (I)

2.0 m

None specified

2.5 m

None specified 6 hours (primary) 72 hours (secondary) None specified

East Coast, USA New South Wales, Australia

1.6 m

3.0 m (primary) 2.0 m (secondary)

Perth, Australia

1.6 m (summer) 2.7 m (winter)

4.0 m (primary) 2.0 m (secondary) 3.5 m

None specified

3.0 m

None specified

2.8 m (primary) 1.7 m (secondary) 2.0 m (primary) 1.5 m (secondary) 3.9 m (primary) 2.2 m (secondary) 1.5 m

None specified

1.5 m

Durban, 1.7 m South Africa Algarve, 0.9 m Portugal Perranporth, 1.4 m UK Catalonia, Spain

0.8 m

Gironde, France

1.4 m

Emilia0.4 m Romagna, Italy Cadiz, Spain 1.0 m

Reference

None specified Houser & Greenwood (2005) None specified Dolan & Davis (1992) 24 hours Shand et al. (2010)

None specified Lemm et al. (1999) 2 weeks

Corbella & Stretch (2012) 30 hours Almeida et al. (2012) None specified Masselink et al. (2014)

6 hours

72 hours

Mendoza et al. (2011)

None specified

None specified Castelle et al. (2015)

6 hours

None specified Armaroli et al. (2012)

None specified

None specified Plomaritis et al. (2015)

Table 1.2 also indicates a two-threshold approach to coastal storm identification undertaken at several sites. In most of these cases, the upper wave height threshold represents the primary threshold used for storm identification as depicted by the POT method in Figure 1.6. A secondary lower threshold is then used to calculate the start time (i.e. up-crossing of the lower threshold) and end time (i.e. down-crossing of the lower threshold) of the storm and hence the storm duration. Masselink et al. (2014) define this lower threshold for identifying the start duration as the 75th percentile of the significant wave height data. A secondary lower threshold may also be used to further refine the meteorological independence criterion (e.g. Mendoza et al., 2011),

CH1 COASTAL STORM DEFINITION

15

or to include coastal storm events of particularly long duration but not necessarily of high peak wave heights (e.g. Shand et al., 2010).

1.3.2 Coastal storm events from water-level time-series The approach taken to identifying coastal storm events from water-level time-series is similar to that taken for wave time-series discussed above. A key question for this type of storm identification, however, is whether or not to base classifications on the total water levels (TWL, Equation 1.1), or to eliminate tidal variability and use the non-tidal residuals (R, Equation 1.1). The answer to this question lies in the purpose of the coastal storm assessment being undertaken. For studies focused on coastal storm impacts, or for communicating coastal storm hazards to the wider community (e.g. for issuing alerts within the context of a coastal storm early warning system), it is the TWL that ultimately dictates the exposure of the backshore and hinterland to inundation. Hence it follows that coastal storms be identified based on the TWL in these scenarios. This is the case for low-lying coastlines such as Venice in Northern Italy, where the commonly-occurring acqua alta (Italian for “high water” – a form of coastal storm caused by strong SE winds blowing along the length of the Adriatic Sea) is defined by TWL exceedances above a local threshold TWLthresh . These thresholds are directly linked to local features that represent measures of coastal storm exposure – in the case of Venice, the TWL in which a significant part of the city is inundated (Massalin et al., 2007). By the same measure, Aagaard et al. (2007) identified storm events for the Danish coastline based on a TWLthresh of 2.4 m above a local datum, which coincides with the approximate elevation of the dune toe of this region. For studies where an understanding of coastal storminess variability over time is the main goal, a more appropriate identifier of coastal storms is given by the exceedance of R above a certain threshold Rthresh . By defining coastal storm events based on R (as opposed to the TWL), the influence of the tidal signal on long-term storminess trends (after taking into consideration complex tide-surge interactions) is removed and the focus is solely on meteorological effects. Table 1.3 indicates a number of approaches taken to define Rthresh for coastal storm identification at various sites worldwide. For the eastern seaboard of the USA, Zhang et al. (2000) define Rthresh as equal to two standard deviations of R. Similar to the coastal storm identification based on wave height data, a meteorological independence criterion of 12 hours was applied to discern individual events. This period was selected to remove the influences of free oscillations in the water level following the storm surge, which had residence times of approximately this period. Bromirski and Flick (2008) meanwhile define Rthresh as the 98th percentile of the low-pass filtered R time-series, which was applied to remove high-frequency variability in the non-tidal residuals. A minimum storm duration of six hours was also adopted in this analysis to ensure only events of an adequate persistence were identified. To highlight the differences in the storm occurrence depending on the type of water-level threshold adopted, Figure 1.7 presents a hypothetical scenario whereby a coastal storm has been identified from the same water-level time-series using both the TWL and R thresholds. As indicated by the shaded regions in the upper and lower panels, both the timing and duration of the storm differ significantly between the two threshold types. In the case of defining the event based on Rthresh (upper panel), the storm remains above the selected 1 m threshold for a period of 11 hours. The fact that

16

COASTAL STORMS: PROCESSES AND IMPACTS

Table 1.3 Site-specific classifications of coastal storms based on water-level information. Site

Spring tidal range

Threshold

Meteorological independence criterion (I)

3m

R > 98th percentile

None specified

1.2 m

R > 2 x standard deviation R > 0.25 m

None specified

TWL > 1.1 m (linked to % flooding of Venice) TWL > 2.4 m (typical dune toe elevation)

None specified

San Francisco, USA North Carolina, USA East coast, UK

3.6–6.2 m

Venice, Italy

1.1 m

Skallingen, Denmark

1.8 m

12 hours

None specified

Reference

Bromirski & Flick (2008) Zhang et al. (2000) Horsburgh & Wilson (2007) Massalin et al. (2007)

Aagaard et al. (2007)

this event commenced at a period of relatively low tide, however, meant that, when considering only the TWL, the event only became a storm according to this definition (TWLthresh = 2 m) following the onset of high tide some 6 hours later. The duration of the storm in the lower panel is also considerably shorter.

1.3.3 Indicators of coastal storm severity Following the initial identification of a coastal storm event from the methods described above, a subsequent step is to understand the potential severity of the storm itself. In terms of TCs, severity is commonly communicated in the form of the familiar Saffir-Simpson hurricane wind scale. This hurricane scale ranks cyclone severity into one of five categories based on the maximum one-minute sustained wind speeds. A Category 1 cyclone according to this scale therefore has a maximum one-minute sustained wind speed between 33 m/s and 42 m/s, whereas a Category 5 cyclone has a maximum wind speed in excess of 70 m/s. As noted by Fritz et al. (2007), based on observations of hurricane impacts at New Orleans, however, this scale only has limited application in terms of understanding potential storm surge and can in fact mislead the public into a false sense of security about hurricane impacts. Hurricane Katrina that struck New Orleans in 2005, for instance, was only classified as a Category 3 cyclone at landfall according to this scale, but produced peak surges in excess of 10 m and tremendous destruction. In contrast, Hurricane Camille that hit the same coastal location in 1969 was classified as a Category 5 cyclone at landfall, but resulted in more moderate storm surge and impacts. Another limitation of the Saffir-Simpson scale is that it is only relevant to TCs and cannot be applied to other synoptic systems. As a first-pass assessment of coastal storm severity with regards to the local maritime conditions, a common indicator is the return period of the peak wave or water level (as

Non-tidal residulas (m)

CH1 COASTAL STORM DEFINITION

17

2 1.5 Rthresh 1 0.5 0 01/12/09

02/12/09 Date (day/month/year)

03/12/09

5 Total water level Astronomical tide

Water level (m)

4 3

TWLthresh

2 1 0 –1 –2 01/12/09

02/12/09 Date (day/month/year)

03/12/09

Figure 1.7 Comparison between coastal storm definitions based on a non-tidal residual threshold Rthresh (upper panel) and a threshold of the total water-level TWLthresh (lower threshold). Shaded regions highlight the different coastal storm periods identified according to these two definitions.

depicted by P in Figure 1.6). This assessment entails an extreme value analysis of the historical data and means that coastal storm severity can be communicated in terms of its average recurrence interval or annual exceedance probability. A limitation of using only peak values to classify storm severity, however, is that it does not take into account the duration of the event or its joint occurrence with elevated water levels. To address the limitation of storm severity and event duration, Dolan & Davis (1992) developed a storm intensity classification specifically for longer-duration ETC storms on the East coast of the USA. The duration of the storm event was taken into account in this severity classification by integrating the wave energy over the whole event, as given by the equation: t2

E=

∫t1

Hsig 2 dt

Hsig ≥ Hthresh

(1.2)

where E is the storm energy content and t1 and t2 correspond to the start and end times of the storm event determined by the up and down-crossings of the storm threshold Hthresh . Storm severity levels were subsequently developed based on ranges of these E values, using a five-category system analogous to the Saffir-Simpson hurricane wind scale. Note that this system was developed for the East coast of the USA only and is therefore site-specific. Mendoza et al. (2011) applied this same classification methodology to the coastline of Catalonia in Spain and resulted in different ranges of E for each storm class (Table 1.4). These differences can be attributed to the different storm

18

COASTAL STORMS: PROCESSES AND IMPACTS

Table 1.4 Coastal storm severity classifications for the coastlines of the East Coast, USA and Catalonia, Spain based on the total storm energy methodology of Dolan and Davis (1992). Storm class

I II III IV V

Description

Weak Moderate Significant Severe Extreme

Storm Energy Range (m2 h) East Coast, USA (Dolan and Davis, 1992)

Catalonia, Spain (Mendoza et al., 2011)

2323

24–250 251–500 501–700 701–1200 >1200

thresholds selected for the two sites (see Table 1.3), as well as local storm wave characteristics. Zhang et al. (2000), meanwhile, adopted a similar intensity index for storm surge events on the East Coast of the USA by instead integrating the non-tidal residual time-series over the duration of the surge event. Further refinements to coastal storm severity indicators have been undertaken by combining the effects of both wave and water levels during a coastal storm event. Kreibel et al. (1997) developed a Risk Index RI for ETCs on the East Coast of the USA based on the wave height, storm duration and storm surge level. This index is given by: (1.3) RI = SP(D∕12)0.3 where S is the surge elevation in feet, P is the peak wave height in feet and D is the event duration in hours. As shown in Equation 1.3, this index scales the event duration by 12 hours to obtain the number of tidal cycles during the storm event. In order to achieve an index range between 0 and 5 (i.e. similar to that of the Saffir-Simpson and Dolan and Davis scales discussed above), the index is then normalized by that of the most severe storm on record (RI = 400 for a severe ETC in March, 1962) and multiplied by a factor of five.

1.4 Conclusion Coastal storms are a particularly challenging phenomenon to define and have historically been misrepresented in both the literature and various natural hazard evaluation studies. This chapter has outlined the numerous issues associated with defining coastal storms, which include the location of the atmospheric disturbance relative to the impact area, the diversity of coastal environments and ways in which they respond to maritime forcing, as well as storm timing. Based on these considerations, a general coastal storm definition has been established for all coastal environments as a meteorologically-induced disturbance to the local maritime conditions (i.e. waves and/or water levels) that has the potential to significantly alter the underlying morphology and expose the backshore to waves, currents and/or inundation. Statistical approaches to identifying coastal storms involve the establishment of a site-specific storm threshold that discriminates storms from non-storm periods. For

CH1 COASTAL STORM DEFINITION

19

coastal storms based on wave height information, this storm threshold is strongly related to the modal wave conditions of the site. For those based on water-level data, meanwhile, a consideration of whether the total water level or non-tidal residuals is more appropriate to define storms is required. Once the coastal storm event has been identified, the potential storm severity can then be gauged through consideration of the peak storm wave height, the duration of the event and water-level variability.

References Aagaard, T., Orford, J. & Murray, A.S. (2007) Environmental controls on coastal dune formation: Skallingen Spit, Denmark. Geomorphology, 83, 29–47. Ahrens, C.D. (2000) Meteorology Today: An introduction to Weather, Climate, and the Environment, Brooks/Cole Publishing, Pacific Grove, USA, sixth edition. Almeida, L.P., Vousdoukas, M.V., Ferreira, O., Rodrigues, B.A. & Matias, A. (2012) Thresholds for storm impacts on an exposed sandy coastal area in southern Portugal. Geomorphology, 143–144, 3–12. Armaroli, C., Ciavola, P., Perini, L., Calabrese, L., Lorito, S., Valentini, A. et al. (2012) Critical storm thresholds for significant morphological changes and damage along the Emilia-Romagna coastline, Italy. Geomorphology, 143–144, 34–51. Betts, N.L., Orford, J.D., White, D. & Graham, C.J. (2004) Storminess and surges in the South-Western Approaches of the eastern North Atlantic: The synoptic climatology of recent extreme coastal storms. Marine Geology, 210, 227–246. Blake, E.S. & Gibney, E.J. (2011) The deadliest, costliest and most intense United States tropical cyclones from 1851 to 2010 (and other frequently requested hurricane facts). NOAA Technical Memorandum NWS NHC, 6, 1–47. Bromirski, P.D. & Flick, PD. (2008) Storm surge in the San Francisco Bay/Delta and nearby coastal locations. Shore and Beach, 76 (3), 29–37. Browning, S.A. & Goodwin, I.D. (2013) Large-scale influences on the evolution of winter subtropical maritime cyclones affecting Australia’s east coast. Monthly Weather Review, 141, 2416–2431. Castelle, B., Marieu, V., Bujan, S., Splinter, K.D., Robinet, A., Senechal, N. et al. (2015) Impact of the winter 2013–2014 series of severe Western Europe storms on a double-barred sandy coast: Beach and dune erosion and megacusp embayments. Geomorphology, 238, 135–148. Cid A., Menéndez, M., Castanedo, S., Abascal, A.J., Méndez, F.J., & Medina, R. (2016) Long-term changes in the frequency, intensity and duration of extreme storm surge events in southern Europe. Climate Dynamics, 46(5), 1503–1516. Corbella S.& Stretch, D.D. (2012) Multivariate return periods of sea storms for coastal erosion risk assessment. Nat. Hazards Earth Syst. Sci., 12, 2699–2708. Dolan, R. & Davis, R.E. (1992) An intensity scale for Atlantic coast northeast storms. Journal of Coastal Research, 8 (4), 840–853. Fritz, H.M., Blount, C., Sokoloski, J., Singleton, J., Fuggle, A., McAdoo, B.G. et al. (2007) Hurricane Katrina storm surge distribution and field observations on the Mississippi Barrier Islands. Estuarine, Coastal and Shelf Science, 74, 12–20. Gulev, S.K., Zolina, O. & Grigoriev, S. (2001) Extratropical cyclone variability in the northern hemisphere winter from NCEP/NCar reanalysis data. Climate Dynamics, 17, 795–809. Haque, C.E. (1997) Atmospheric hazards preparedness in Bangladesh: A study of warning, adjustments and recovery from the April 1991 cyclone. Natural Hazards, 16, 181–202.

20

COASTAL STORMS: PROCESSES AND IMPACTS

Harley, M.D., Turner, I.L., Short, A.D. & Ranasinghe, R. (2010) Interannual variability and controls of the Sydney wave climate. International Journal of Climatology, 30, 1322–1335. Harley, M.D., Turner, I.L., Short, A.D. & Ranasinghe, R. (2011) Assessment and integration of conventional, RTK-GPS and image-derived beach survey methods for daily to decadal coastal monitoring. Coastal Engineering, 58, 194–205. Hazards and Vulnerability Research Institute (2015) 1960-2014 US Hazards Losses, University of South Carolina. Available from: http://hvri.geog.sc.edu/SHELDUS/docs/Summary_1960_ 2014.pdf (6 October, 2015). Hemer, M.A., Simmonds, I. & Keay, K. (2008) A classification of wave generation characteristics during large wave events on the Southern Australian margin. Continental Shelf Research, 634–652. Horsburgh, K.J. & Wilson, C. (2007) Tide-surge interactions and its role in the distribution of surge residuals in the North Sea,. J. Geophysical Research-Oceans, 112, C08003, doi:10.1029/2006JC004033. Houser, C. & Greenwood, B. (2005) Profile response of a lacustrine multiple barred nearshore to a sequence of storm events. Geomorphology, 1–4, 118–137. International Federation of Red Cross and Red Crescent Societies (2014) World Disasters Report 2014. Available from: http://www.ifrc.org/world-disasters-report-2014 (6 October, 2015). Jackson, N.L., Nordstrom, K.F., Eliot, I. & Masselink, G. (2002) “Low energy” sandy beaches in marine and estuarine environments: A review. Geomorphology, 48 (1–3), 147–162. Kolen, B., Slomp, R. & Jonkman, S. (2013) The impacts of storm Xynthia February 27–28, 2010 in France: Lessons for flood risk management. Journal of Flood Risk Management, 6, 261–278. Kriebel, D., Dalrymple, R., Pratt, A. & Sakovich, V. (1997) A shoreline risk index for Northeasters. Proc. Conf. Natural Disaster Reduction, ASCE, 251–252. Laska, S. & Morrow, B.H. (2006) Social vulnerabilities and Hurricane Katrina: An unnatural disaster in New Orleans. Marine Technology Society Journal, 40 (4), 16–26. Lemm, A.J., Hegge, B.J. & Masselink, G. (1999) Offshore wave climate, Perth (Western Australia), 1994–96. Mar. Freshwater Res., 50, 95–102. Lionello, P., Cavaleri, L., Nissen, K.M., Pino, C., Raicich, F., & Ulbrich, U. (2012) Severe marine storms in the Northern Adriatic: Characteristics and trends, Physics and Chemistry of the Earth, 40–41, 93–105. Marks, F.D. (2003) Hurricanes. In: J.R. Holton, J.A. Curry & J.A. Pyle (Eds) Encyclopedia of Atmospheric Sciences. Elsevier, pp. 942–966. Massalin, A., Zampato, L. & Canestrelli, P. (2007) Data monitoring and sea level forecasting in the Venice lagoon: The ICPSM’s activity. Bolletino di Geofisica Teorica ed Applicata, 48, 241–257. Masselink, G., Austin, M., Scott, T., Poate, T. & Russell, P. (2014) Role of wave forcing, storms and NAO in outer bar dynamics on a high-energy, macro-tidal beach. Geomorphology, 226, 76–93. Mather, J.R., Adams III, H. & Yoshioka, G.A. (1964) Coastal storms of the Eastern United States. Journal of Applied Meteorology, 3, 693–706. May, S.M., Engel, M., Brill, D., Squire, P., Scheffers, A. & Kelletat, D. (2013) Coastal hazards from tropical cyclones and extratropical winter storms based on Holocene storm chronologies. In: C.W. Finkl (Ed.) Coastal Hazards. Springer, pp. 557–585. Mazas, F. & Hamm, L. (2011) A multi-distribution approach to POT methods for determining extreme wave heights. Coastal Engineering, 58 (5), 385–394. Mendoza, E.T., Jimenez, J.A. & Mateo, J. (2011) A coastal storm intensity scale for the Catalan sea (NW Mediterranean). Nat. Hazards Earth Syst. Sci., 11, 2453–2462.

CH1 COASTAL STORM DEFINITION

21

Merriam-Webster (2015) Storm. Accessed from: www.merriam-webster.com (6 October, 2015). National Weather Service (2015) What is storm surge? National Oceanic and Atmospheric Administration, http://www.nws.noaa.gov/om/hurricane/resources/surge_intro.pdf> (6 October, 2015). Phillips, M.S., Turner, I.L., Cox, R.J., Splinter, K.D. & Harley, M.D. (2015) Will the sand come back? Observations and characteristics of beach recovery, 22nd Australasian Conference on Coastal and Ocean Engineering, Engineers Australia, Auckland NZ , 15–18 September. Plomaritis, T.A., Benavente, J., Laiz, I. & Del Rio, L. (2015) Variability in storm climate along the Gulf of Cadiz: The role of large scale atmospheric forcing and implications to coastal hazards. Climate Dynamics, 445 (9), 2499–2514. Pugh, D.T, (1987) Tides, surges and mean sea-level: A handbook for engineers and scientists. John Wiley & Sons Ltd, Chichester, UK. Shand, T.D., Goodwin, I.D., Mole, M.A., Carley, J.T., Coghlan, I.R. & Harley, M.D. (2010) NSW coastal inundation hazard study: Coastal storms and extreme waves. Water Research Laboratory Technical Report, UNSW, Australia. Short, A.D. & Trenamen, N.L. (1992) Wave climate of the Sydney region, an energetic and highly variable ocean wave regime. Aust. J. Mar. Freshwater Res., 43, 765–791. Simmonds, I. & Keay, K. (2000) Mean Southern Hemisphere extratropical cyclone behavior in the 40-year NCEP-NCAR Reanalysis. Journal of Climate, 13 (5), 873–885. Small, C. & Nicholls, R.J. (2003) A global analysis of human settlement in coastal zones. Journal of Coastal Research, 19 (3), 584–599. Tunnel, J.W. (2002) Geography, Climate and Hydrography. In: J.W. Tunnel & F.W. Judd (Eds) The Laguna Madre of Texas and Tamaulipas. Texas A&M University Press, pp. 7–27. Wallace, J.M. & Hobbs, P.V. (2006) Atmospheric science: An introductory survey, Elsevier, second edition. Wright, J.D., Colling, A., and Park, D. (1999) Waves, tides and shallow-water processes. Butterworth-Heinemann Oxford, second edition. Yarnal, B. (1993) Synoptic climatology in environmental analysis: A primer. Belhaven Press, London. Zhang, K., Douglas, B. & Leatherman, S. (2000) Do storms cause long-term erosion along the US east barrier coast? Journal of Geology, 110 (4), 493–502.

2 Hydrodynamics Under Storm Conditions Xavier Bertin1 , Maitane Olabarrieta2 and Robert McCall3 1 UMR

7266 LIENSs, CNRS-Université de La Rochelle, La Rochelle, France and Coastal Engineering Department, ESSIE, University of Florida, Gainesville, Florida, USA 3 Deltares, Delft, The Netherlands 2 Civil

2.1

General introduction

Storms usually induce a surge that causes the water level to reach a higher position compared to normal conditions. In addition, wave energy is usually higher than normal under storm conditions so that currents and subsequent sediment transport can induce dramatic morphological changes, including severe erosion. This chapter aims to present the main processes that control water level variations and hydrodynamic circulation under storm conditions. The section following this introduction presents the main mechanisms that control storm surges: atmospheric pressure gradients, surface stress and short-wave dissipation. The third section presents the mechanisms that drive the hydrodynamic circulation along the coast: longshore currents, bed return flows, infragravity waves and swash dynamics. This chapter concludes with some perspectives and challenges.

2.2

Storm surges

2.2.1 Introduction Storm surges correspond to non-astronomic variations in the ocean free-surface driven by meteo-oceanic forcing. Storm surges larger than 9 m were reported in the Gulf of Mexico during hurricane Katrina in 2005 (Blake, 2007; Dietrich et al., 2010) and in the Bay of Bengal. These have long been known as the regions around the world most exposed to major storm surges and coastal flooding. Recently “super” storm Sandy (2012) in the New York area and typhoon Haiyan (2013) in the Philippines drove 3–6 m storm surges, and evidenced that other regions in the world are exposed to extreme storm surges and coastal flooding. Besides being located along the track of major storms, these vulnerable regions systematically exhibit low-lying coastal zones Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

24

COASTAL STORMS: PROCESSES AND IMPACTS

bordered by shallow waters and/or extensive continental shelves (i.e. shelf width in the range of 150–500 km). The relationship between the height of the storm surge and the bathymetry is explained through the governing shallow water equations, where the wind effect is inversely proportional to the water depth. As a consequence, in shallow waters, the wind effect is often dominant over the well-known “inverse barometer effect” (Doodson, 1924) caused by atmospheric pressure gradients (Rego & Li, 2010). In the nearshore, wave dissipation induces radiation stress gradients (Longuet-Higgins & Stewart, 1964) that drive a setup that can easily reach several tens of centimeters during storms and therefore contribute to the storm surge. In volcanic islands, where continental shelves are almost absent, the wind contribution is often very weak, and the wave setup can become dominant over the other contributions (e.g. Kennedy et al., 2012).

2.2.2 Governing equations In the absence of density gradients and/or stratification, the coastal circulation is usually well described by the depth-integrated shallow water equations (Saint-Venant equations). This simplification is only more questionable for estuaries when the freshwater is not well mixed, in larger water depths where the fast moving upper layer of the ocean is poorly represented by depth-integrated velocities or in surf zones when a bed return flow is present. Including the source terms associated with storm surges and waves, the governing equations read: 𝜕𝜁 𝜕 (h + 𝜁 ) u 𝜕 (h + 𝜁 ) v + + =0 𝜕t 𝜕x 𝜕y

) 𝜕Sxx 𝜕Syx + 𝜕x 𝜕y ( ) 𝜕Syy 𝜕Syx 𝜏sy − 𝜏by 1 + − + 𝜌w ⋅ (𝜁 + h) 𝜌w ⋅ (𝜁 + h) 𝜕y 𝜕x

𝜏 − 𝜏bx 𝜕𝜁 Du 1 𝜕Patm 1 = fv − g − + sx − Dt 𝜕x 𝜌w 𝜕x 𝜌w ⋅ (𝜁 + h) 𝜌w ⋅ (𝜁 + h) 𝜕𝜁 Dv 1 𝜕Patm = −fu − g − Dt 𝜕y 𝜌w 𝜕y

(

(2.1)

where ζ is the free surface elevation, u and v are the horizontal component of the depth-integrated velocity, h is the water depth below mean water level, 𝜌w is water density, g is the acceleration of the gravity, f is the Coriolis parameter, Patm is the sea-level atmospheric pressure, 𝜏b and 𝜏s are the bottom and surface stress, respectively. Finally, Sxx , Sxy , Syx and Syy are the wave radiation stresses, which correspond to the flux of momentum associated with the waves. Considering an axis perpendicular to a uniform shoreline, under steady state, and considering that the cross-shore component of the depth-integrated velocity u is nil, the source terms responsible for the storm surge are balanced by the barotropic pressure gradient: ( ) 𝜕Sxx 𝜏sx 𝜕𝜁 1 𝜕Patm 1 =− + − (2.2) g⋅ 𝜕x 𝜌w 𝜕x 𝜌w ⋅ (𝜁 + h) 𝜌w ⋅ (𝜁 + h) 𝜕x The three terms on the right hand side of Equation (2.2) correspond to the atmospheric pressure gradient, the wind induced surface stress and the shore normal component of wave forces. These are the main driving processes for storm surges, which will be described in the following subsections.

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

25

2.2.2.1 Atmospheric pressure The effect of atmospheric pressure gradients on sea-level variations was recognized very early and referred to as “inverse barometric effect” (e.g. Doodson, 1924). The associated empirical rule states that a variation of 1 mBar with respect to the mean sea-level atmospheric pressure (1013 mBar) causes a 1 cm variation in sea level. Although very simple, this empirical rule holds as long as changes in atmospheric pressure are slow enough, so that the dynamics associated with sea-level adjustment is negligible. For extreme tropical hurricanes, where the minimum sea-level pressure can reach 900 mBar or less, the contribution of this effect in the storm surge can thus exceed 1.0 m. Given that the term corresponding to the atmospheric pressure gradient is independent from the water depth, its contribution to the total storm surge is often dominant in deep water or at a volcanic island where the continental shelf is restricted or absent. Moving atmospheric pressure disturbances associated with fronts (such as squalls) and atmospheric gravity waves can also trigger sea level oscillations with periods from a few minutes to a few hours. These waves, known as meteo-tsunamis, are caused by multi-resonant processes (e.g. Proudman, 1929; Greenspan, 1956) and the shelf (Monserrat et al., 2006), and harbor resonance and their effects can be as severe as those caused by tsunamis and become catastrophic in some regions (Rabinovich, 2009). 2.2.2.2 Surface stress For a long time, it has been common practice to compute the wind surface stress 𝜏s based on bulk formula: 2 𝜏s = 𝜌a ⋅ Cd ⋅ U10

(2.3)

where 𝜌a is the air density, U10 the 10 m wind speed and Cd is a drag coefficient corresponding to the sea roughness that increases linearly with the wind speed for low to moderate winds (e.g. Pond and Pickard, 1998). However, although the simplicity of such formulae could make them attractive for direct implementation in storm surge models, it has two main shortcomings. First, several studies relying on field and laboratory measurements suggested that under extreme winds, the sea roughness and therefore the drag coefficient, could reach a maximum or even decrease (Figure 2.1) due to wave-induced streaks of foam and sprays for winds larger than 35–40 m/s (Powell et al., 2003; Takagaki et al., 2012). Second, for a given wind speed, a significant scatter exists and Cd could vary by 30% or even more (Figure 2.1). This scatter is partly explained by the fact that the sea roughness does not only depend on the wind speed but can also be impacted by the sea state. Based on the pioneer work of Charnock (1955), Stewart (1974) proposed that for a given wind speed the sea roughness should also depend on the wave age, which is defined as the ratio between the wave phase speed and the friction velocity. The dependence of the surface stress to the sea state was then corroborated in many studies (Mastenbroek et al., 1993; Moon, 2005; Brown & Wolf, 2009; Bertin et al., 2012; Olabarrieta et al., 2012). The importance of this phenomenon was recently demonstrated by Bertin et al. (2015), who performed a comparison between two storms that recently hit the central part of the Bay of Biscay. Despite displaying comparable wind fields in the study area, these two storms induced very different storm surges and sea states. The former storm, Xynthia (27–28 of February 2010), was characterized by large (up to 7 m significant wave height Hs) and short-period waves and induced an

26

COASTAL STORMS: PROCESSES AND IMPACTS

4

x 10−3 Donelan et al. (2004) Hawkins & Rubsam (1968) Powell et al. (2003) Takagaki et al. (2012)

3.5

3

Cd (-)

2.5

2

1.5

1

0.5

0 0

10

20

40 30 10 m wind speed (m.s−1)

50

60

70

Figure 2.1 Sea-surface drag coefficient as a function of wind speed, based on the dataset of Donelan et al. (2004), Hawkins & Rubsam (1968), Powell et al. (2003) and Takagaki et al. (2012).

exceptional storm surge for the study area, locally larger than 1.6 m. The second storm, Joachim (15–16 of December 2011), was characterized by very large (up to Hs > 10 m) and long-period waves, but only induced a storm surge almost two times lower. The analysis of modeling results revealed that the very large differences in surges induced by these two storms originated from major differences in sea states (Figure 2.2). During Xynthia the wave spectra time series (Figure 2.2, top left) showed that, at the beginning of the storm, most of the energy was found in the range 0.10–0.15 Hz, with levels reaching 70 m2 Hz−1 . This spectral distribution strongly differs for Joachim, where maximum energy in this frequency band reaches 30 m2 Hz−1 , although the total energy was much larger. Such conditions during Xynthia are representative of a very young sea state, characterized by steep waves, which enhanced the surface stress up to a factor of two (Figure 2.2, middle row). These results suggest that the bulk formula may only perform when sea states are mature.

2.2.2.3 Ekman transport Due to the rotation of the Earth, the wind-driven flow is deviated to the right in the northern hemisphere, and to the left in the southern hemisphere, with respect to the direction of the wind. At the surface, this deviation is theoretically 45∘ under steady state, although several studies documented smaller values in the range 20–30∘ (e.g. Holmedal & Myrhaug, 2013). When going down in the water column, the current velocity decreases until the so-called Ekman layer is reached, while the current direction also shifts to the right (respectively to the left). Under steady state and in deep water, the net water transport is oriented at 90∘ to the right of the wind direction (respectively to the left), although this value tends to decrease in shallow water due to the increase in bottom shear stress. When a storm approaches the coast, the

A-Xynthia

100 Frequency (Hz)

B-Joachim

120

0.25 0.2

80 0.15

60 40

0.1 20 0.05

0

0.25

100 80 0.15

2 1 0

Storm surge (m)

1.5 1

20

4

0 Bulk formula Wave-dependent

3 2 1 0

Observed surge Wave-dependent Bulk formula

1.5

Observed surge Wave-dependent Bulk formula

1

0.5

0.5

0 27/02/2010

40

0.05 5

3

60

0.1

Surface stress (Pa)

Surface stress (Pa)

4

120

0.2

5 Bulk formula Wave-dependent

27

Variance density (m2.Hz−1)

HYDRODYNAMICS UNDER STORM CONDITIONS

Variance density (m2.Hz−1)

CH2

0 28/02/2010

01/03/2010

15/12/2011

16/12/2011

17/12/2011

Figure 2.2 Time series of wave energy spectra (top row), surface stress (middle row) and storm surge (bottom row) in La Rochelle (Bay of Biscay, France) during Xynthia (left) and Joachim (right). Adapted from Bertin et al. 2015.

28

COASTAL STORMS: PROCESSES AND IMPACTS

alongshore component of wind stress drives an “Ekman setup” (Kennedy et al., 2011) at the coast located to the right hand side of the wind (to left hand side in the southern hemisphere). Note that this process is distinct from the effect of wind speed asymmetry in hurricanes, which can also lead to increased surge levels (Xie et al., 2011). As an important consequence of the Ekman setup, the coastal region located to the right side of the storm track usually suffers larger storm surge and damages than the coastal region located to the left side. This phenomenon was illustrated by Kennedy et al. (2011), who reported a large water level anomaly along the coasts of Louisiana and Texas 12–24 h before the landfall of Hurricane Ike (2008) and attributed the origin of these forerunner anomalies to an Ekman setup. The storm surge associated with Ike was hindcast using the modeling system SELFE (Zhang & Baptista, 2008), with an unstructured grid covering the Gulf of Mexico (Figure 2.3a) and forced with fields of wind speed and sea-level pressure originating from the CFSR reanalysis (Saha et al., 2010). The wind speed data show that 15 h before landfall, a band of wind ranging from 20 to 30 m/s developed parallel to the shore over almost 500 km (Figure 2.3c). According to the Ekman theory, this band of high winds parallel to the coast drives a transport towards the coast, which results in the development of a storm surge before the landfall of the hurricane. This phenomenon is confirmed in both the model simulation results and the tide gauges at Gavelston and Freshwater, where this forerunner surge is shown to reach 1.5–2.0 m 15 h before landfall (Figure 2.4a, c and d). In order to quantify the importance of the Ekman transport, the model was run without the Coriolis term and subsequent Ekman forcing. In the Coriolis-free simulation, the forerunner surge no longer develops and the water level is up to two meters lower than measured before landfall. The surge peak is also underestimated by almost 1.0 m. These results demonstrate that the wind direction, which is partly controlled by the track of the storm, should be considered very important.

2.2.2.4 Short wave dissipation In the nearshore, wave dissipation related to depth-limited breaking induces radiation stress gradients, which drive currents as well as a setup along the coast. Under big waves, this setup can easily reach several tens of cm and therefore significantly contribute to the storm surge. This contribution can even be dominant over other forcing terms at volcanic islands and coastal zones bordered by narrow continental shelves (e.g. Kennedy et al., 2012). Recent studies also showed that wave setup can propagate outside surf zones and thereby contribute to the storm surge in areas sheltered from wave breaking, such as coastal lagoons (e.g. Bertin et al., 2009; Dodet et al., 2013), estuaries (Olabarrieta et al., 2011; Arnaud & Bertin, 2014; Bertin et al., 2015) and tropical lagoons bordered by coral reefs (Aucan et al., 2012). Arnaud and Bertin (2014) numerically investigated the contribution of wave setup in the southern part of the Bay of Biscay during storm Klaus (January 2009) characterized by an offshore significant wave height exceeding 13 m. These authors showed that wave setup reaching almost 1.0 m developed along the coast exposed to these massive waves (not shown), which is in line with current theories. More surprisingly, these authors also showed that a wave setup of the order of 0.4 to 0.5 m propagated inside the Arcachon Lagoon and the Adour Estuary. The comparison of the observed surge to simulations with and without wave forces reveals that it is possible to reproduce the storm surge only when wave forces are taken into account. More specifically, accounting for wave forces reduced the Root Mean Squared Error (hereafter RMSE) by a factor 3 and

CH2

31

HYDRODYNAMICS UNDER STORM CONDITIONS

29

>1000 (a)

30 800

28

600

27 400

26

Bathymetry (m)

39

25 200 24 23

0 −94

−92

−90 1020 1010

28

1000 990

26

980 970

24 −95

−90

−85

960

−86

−88

1030

(b)

Sea-level pressure (hPa)

30

−96

30

−84

−82 40

(c)

30 28 20 26 10 24 −95

−90

−85

Wind speed (m/s)

−98

0

Figure 2.3 (a) Simplified bathymetry of the Gulf of Mexico superimposed with the track of Ike and the location of the tide gauges used in this study; (b) sea-level pressure; and (c) 10 m wind speed 15 h before landfall, showing wind speed in the range 20–30 m/s parallel to the shore.

30

3

(a) Model with Coriolis

29.5

2.5

29

2

28.5

1.5

28

1

27.5

0.5

27

−0.5

26 −98

−97

−96

−95

−94

−93

−92

−91

−90

1.5 1

28

0.5

27.5

0

27

−0.5

26.5

−1

26 −98

−97

−96

−95

−94

−93

−92

−91

−90

−1

5 Data Model with Coriolis Model without Coriolis 15 h before landfall

RMSE = 0.43 m, BIAS = −0.29 m RMSE = 0.89 m, BIAS = −0.66 m

3

RMSE = 0.14 m, BIAS = −0.10 m RMSE = 0.84 m, BIAS = −0.59 m

2 1 0 -1 10/09

Data Model with Coriolis Model without Coriolis 15 h before landfall

(d) Gavelston Entrance 4

Storm surge (m)

(c) Freshwater Canal Lock

Storm surge (m)

2

28.5

5 4

2.5

29

0

26.5

3

(b) Model without Coriolis

29.5

Storm surge (m)

30

COASTAL STORMS: PROCESSES AND IMPACTS

Storm surge (m)

30

3 2 1 0

11/09

12/09

13/09

14/09

15/09

-1 10/09

11/09

12/09

13/09

14/09

15/09

Figure 2.4 Modeled storm surge 15 h before landfall with (a) and without (b) Coriolis term. Observed (black) and modeled storm surge with (blue) and without (red) Coriolis term at Gavelston (c) and Freshwater (d).

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

2

(a)-Arcachon

RMSE = 0.13 m BIAS = −0.06 m

Storm surge (m)

1.5

31

Data Baseline model Model without wave forces

RMSE = 0.33 m BIAS = −0.28 m

1 0.5 0 2

(b)-Bayonne

RMSE = 0.099 m BIAS = −0.012 m

Storm surge (m)

1.5

Data Baseline model Model without wave forces

RMSE = 0.26 m BIAS = −0.23 m

1 0.5 0 2009/01/23

2009/01/24

2009/01/25

2009/01/26

Figure 2.5 Observed (black circles) and modeled storm surge with (blue) and without (red) wave forces at Arcachon (top) and Bayonne (bottom). Adapted from Arnaud & Bertin (2014).

almost cancelled the bias (Figure 2.5). In Bayonne, this remotely-generated wave setup can represent up to 50% of the total storm surge. This study shows that the storm surges deduced from the tide gauges located in these areas sheltered from waves are not only induced by atmospheric pressure gradients and wind effects but also result from remote wave breaking.

2.3

Hydrodynamics of the surf zone during storms

2.3.1 Introduction During storms, wave energy is usually higher than average, which induces changes in the hydrodynamic circulations compared to normal conditions. This section summarizes the main processes that drive the circulation in surf zones, including longshore currents, bed return flows, infragravity and swash.

2.3.2 Longshore currents When waves approach the shore at an angle, depth-induced breaking drives a longshore current that can easily exceed 1 m/s. Under steady state, considering a longshore uniform beach topography and neglecting horizontal mixing and Earth’s rotation, Equation 2.1 simplifies to: 𝜕Sxy = 𝜏by (2.4) 𝜕x

32

COASTAL STORMS: PROCESSES AND IMPACTS

In this equation, the longshore flow is governed by a balance between bottom shear stress and the cross-shore gradient of the longshore component of the wave momentum flux Sxy given by: H 2 Cg cos 𝜃 sin 𝜃 (2.5) Sxy = 𝜌g s 16 C Where Hs is the significant wave height, Cg and C are the wave group and phase velocity and 𝜃 is the wave angle. Considering that the angle between the wave direction and the currents is close to 90∘ , one can linearize the bottom stress and express it as a function of the mean longshore current V and the maximum wave orbital velocity U0 : 2 𝜏by = 𝜌Cd VU0 𝜋

(2.6)

Where Cd is a friction coefficient that can be computed using the methods of Jonsson (1966) or Swart (1974). Combining Equations 2.4–2.6, one can observe that the longshore current V increases with Hs and 𝜃, theoretically until a maximum value of 45∘ . The beach gradient is also of considerable importance, as it controls the rate of wave dissipation. As a consequence, for given wave conditions, steep beaches will induce larger gradient of wave radiation stress and larger longshore currents compared to gently sloping beaches.

2.3.3 Bed return flows As seen in section 2.4, wave dissipation in the surf zone causes the divergence of the wave momentum flux, which results in the development of a setup that reaches its maximum along the shoreline. The resulting inclination of the free surface induces an offshore-directed barotropic pressure gradient that balances the depth-integrated wave forces: 𝜕Sxx 𝜕𝜁 1 = (2.7) g 𝜕x 𝜌 (h + 𝜁 ) 𝜕x However, this equilibrium is locally imbalanced because wave forces are stronger in the upper part of the water column, while the barotropic pressure gradient is depth uniform (Garcez-Faria et al., 2000). This disequilibrium drives an offshore directed flow in the lower part of the water column, referred to as undertow. In addition to this first mechanism, the fact that wave orbits are not closed results in a net forward motion of water particles in a wave, as first demonstrated by Stokes (1847). This so-called Stokes drift is larger at the surface, which drives a return flow to ensure mass conservation. This return flow also contributes to the undertow. During storms, undertows commonly reach 0.5 m/s−1 and can peak to more than 1 m/s (e.g. Bertin et al., 2008), although field measurements are scarce in the literature under storm conditions. The first consequence is that undertow can transport considerable quantities of sediment offshore and contribute to the beach erosion, as detailed in the next chapter. A second consequence is that undertow can modify water levels and reinforce wave setup through the associated onshore-directed bed shear stress. Using the water level measurements of the Sandy

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

33

Duck experiment (November 1997), Apotsos et al. (2007) showed that wave setup predictions are very underestimated under energetic wave conditions if this process is not taken into account.

2.3.4 Infragravity waves Inside the surf zone, part of the wave enegy is observed at frequencies lower than the incident peak period. These long waves, known as infragravity waves (IG waves hereafter), have frequencies between 1/300 and 1/30 Hz and result from the groupiness of the incident gravity wave field. Munk (1949) and Tucker (1950), independently, were the first to report low frequency wave motions outside the surf zone and relate them to the wave groupiness. The observed long waves, originally known as “surf beat”, were believed to be generated in the surf zone and reflected back to the open ocean. An example of a signal of free surface elevation showing wave groupiness is depicted in Figure 2.6. These wave groups, through the radiation stress gradients, force a low frequency bound wave (red line). This signal was measured during the Duck85 experiment, outside the surf zone at a water depth of 5 m (station H5 in Figure 2.7a). The energy transfer from gravity to IG waves within the surf zone is shown in Figure 2.7b. In this specific example, outside the surf zone (station H4) the gravity band was dominant and was characterized by a swell condition (Hrms = 0.42 m and Tp = 12 s). This energy band was damped within the surf zone (e.g. station R3) while the incoming IG waves gained energy. At station R3 the reflected IG wave energy was lower than the incident energy and at station H4 the IG waves were not detected, which shows the relevance of IG energy dissipation within the surf zone. Based on the concept of radiation stress, Longuet-Higgins and Stewart (1962, 1964) explained that the surf beat may result from the release of the bound long wave associated to the wave groups. As the wave groups approach the coast, the gravity waves shoal until they become unstable and break, destroying the wave structure and releasing a free long wave in the surf zone. This explanation is supported qualitatively by field measurements (e.g. Guza et al., 1984; Masselink, 1995), several numerical studies (e.g. List, 1992; van Dongeren et al., 2002) and laboratory data (e.g. Janssen et al., 2003). 0.4

m

0.2 0 −0.2 −0.4

0

50

100

150

200

250

300

Time. sec

Figure 2.6 Sea-surface elevation, wave-group envelope and associated bound wave (computed with the method of Guza et al., 1985) at station H5 located (Figure 2.7a) outside the surf zone, 9 September (1500 EST), see List (1992). The black line represents the measured free surface elevation, the blue line the wave envelope and the red line the bound wave.

34

COASTAL STORMS: PROCESSES AND IMPACTS

(a) Beach profile and instrument location (Duck85 field experiment) 1 R2

Water depth, m

0

R1

H1

R3

VS

R4

H4

H3

H2

−1

H5

−2 −3 −4 −5 −6

Instrument position 0

50

150 200 Offshore distance, m

100

250

300

350

(b) Directional wave spectrum inside (station R3, left panel) and outside (station H4, right panel) the surfzone. direction ( degrees) / frequency (Hz)

STATION R3

0

0.2

330

0.035 STATION R4

0.2

0.1

60

0.025

0.035 30 0.03

0.15

0.03

0.15 300

0 330

30

0.1

300

60

0.02 90

270

0.02 270

90 0.015

0.015 240

120

0.01 0.005

210

150 180 m2s

/ deg

0.025

0.05

0.05

0

240

120

0.01 0.005

210

150 180

0

m2s / deg

Figure 2.7 (a) Beach profile and location of the pressure and velocity sensors during Duck85 field experiment. (b) Example of the directional wave energy transformation within the surf zone. Directional wave energy spectra at sensor R4 located thin the surf zone (left) and at sensor H3 located outside the surf zone (right). Velocity and pressure measurements, from which the directional wave spectra has been computed, completed during the Duck85 field experiment, 9 September (1500 EST).

An alternative mechanism for the generation of IG waves was presented by Symonds et al. (1982), who considered the temporal variation of the breakpoint almost acting as a wave maker, generating surf beat both seaward and shoreward. The latter waves are reflected at the shoreline and combined with other outgoing long waves. The moving breakpoint mechanism can also be thought as “dynamic setup” in the surf zone. The largest waves within a wave group produce higher gravity wave energy dissipation within the surf zone, producing higher setups. This variability introduces oscillations in the setup at the wave group frequency. Laboratory experiments have confirmed that long waves generated by the moving breakpoint dominate the low frequency wave field on a relatively steep uniform slope beaches (Kostense, 1984; Baldock et al., 2000; Baldock & Huntley, 2002) and on barred beaches (Baldock et al., 2004; Pomeroy et al., 2012). Schäffer (1993) combined the two mechanisms in an analytical model and simulated the experiment of Kostense (1984) with good qualitative agreement. Regarding the relative importance of the two mechanisms, Battjes et al. (2004) proposed that surf beat depends on a normalized slope parameter, that is, in the steep-slope regime the breakpoint generation is more effective, while in the mild-slope regime the released bound long wave is more likely responsible. The same conclusion was

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

35

derived from the laboratory experiments by Dong et al. (2009), who confirmed that the normalized slope presented by Battjes et al. (2004) was a prospective index for the effectiveness of the breakpoint mechanism. Independent of the generation mechanism, the wave groupiness produces incoming infragravity waves within the surf zone. Depending on their period, these waves can be dissipated and/or reflected in the shore and, depending on their direction, frequency and beach slope, can propagate back to the ocean as leaky waves or get trapped in the coast by refraction (Herbers et al., 1995). The latter can become progressive or standing edge waves, depending on the coastal morphology. IG waves can often dominate water velocity and sea surface displacement fields in shallow waters (Huntley et al., 1977; Guza & Thorton, 1982; Holman & Bowen, 1984; Guza et al., 1985; Henderson & Bowen, 2002). One of the most critical factors for potential beach and dune erosion, overwash, and breaching is the total run-up. Its amplitude is controlled by a range of factors, the most significant being the astronomic tide, storm surge, incident waves and IG waves. Under storm conditions on dissipative beaches, the run-up component due to incident waves gets saturated and the IG wave component becomes dominant (Guza & Thornton, 1982; Holman & Sallenger, 1985; Ruessink et al., 1998; Ruggiero et al., 2004; Stockdon et al., 2006; Senechal et al., 2011). Implications for sediment transport are discussed in Chapter 3. Other coastal areas where IG waves might dominate over gravity waves include coral reefs (e.g. Sheremet et al., 2011; Pomorey et al., 2012; van Dongeren et al., 2013; also see Chapter 10) and wave dominated tidal inlets (Bertin and Olabarrieta, 2016). IG waves associated with regular swell wave groups can also produce seiches in harbors (e.g. Bowers, 1977), and consequently affect the operability of mooring lines, the stability of moored ships, produce interruptions in berthing operations resulting in harbor downtime (e.g. McComb et al., 2005; van der Mollen et al., 2006). Field observations (e.g. Okihiro & Guza, 1996; Lopez et al., 2012; Thotagamuwage & Pattiaratchi, 2014) have further confirmed the strong correlation between infragravity waves inside harbors and the swell wave energy outside the harbor.

2.3.5 Swash zone dynamics The swash zone is generally defined as the region of the beach between the wave run-down and run-up limit, which is intermittently wet and dry (e.g. Masselink & Puleo, 2006) and constitutes the boundary between the sea and land. As the tide rises and falls, the region of the swash zone will transverse, and extend beyond, the whole of the intertidal beach. The swash zone represents a highly complex and dynamic region, where strong, unsteady flows lead to high rates of morphological change. Waves that approach the coast shoal, refract and eventually break in the surf zone, as discussed above. However, depending on the degree of surf zone saturation, a certain amount of wave energy will propagate into the swash zone, where this energy leads to wave run-up on the foreshore. During storms wave run-up can lead to the overtopping of coastal structures and flooding of coastal infrastructure, as well as causing dunes to erode and become overwashed (Sallenger, 2000; Ruggiero et al., 2001). From the point of view of coastal storm impacts and coastal safety assessment, it is therefore essential to be able to correctly estimate wave run-up during storm events.

36

COASTAL STORMS: PROCESSES AND IMPACTS

The majority of the predictive wave run-up formulations developed in the previR ous century are Iribarren-type parameterizations of the form H = a𝜉0b + c, based on the work of Hunt (1959) and Battjes (1974), where R is the wave run-up height above still-water level, H is the incident wave height, and a, b (O(1)) and c (O(0.1)) are fitting coefficients with varying values found at different field sites and in laboratory data. While this type of run-up equation is found to correspond reasonably well to wave run-up on coastal defense structures and field observations on steep beaches, wave run-up on flatter and dissipative beaches appears less dependent on the Iribarren parameter and to scale with the deep √ water significant wave height Hs,0 (Ruessink et al., 1998; Ruggiero et al., 2001), or Hs,0 ∕L0 (Nielsen & Hanslow, 1991). In order to address this difference in a practical manner, Stockdon et al. (2006) developed an empirical run-up equation based on separate contributions of the incident-band and infragravity band to wave run-up at the shore. The run-up equation was developed and validated using data from dissipative and reflective sandy beaches in the USA and the Netherlands. While separate optimal formulations can be found for steep and mild sloping beaches, Stockdon et al. also developed a general expression for application on all beaches:

R2%

⎛ ⎜ √ = 1.1 ⎜0.35𝛽f Hs,0 Lp,0 + ⎜ ⎜ ⎝

√ ( )⎞ Hs,0 Lp,0 0.563𝛽f2 + 0.004 ⎟ ⎟ ⎟ 2 ⎟ ⎠

(2.8)

where R2% is the run-up height exceeded by 2% of swash events, 𝛽f is the beach slope measured over the portion of the beach where run-up occurs, Hs,0 is the deep water significant wave height and Lp,0 is the deep water wave length of the peak period wave. The wave run-up equations discussed above all show a dependence between the incident wave height and the total run-up elevation above the still water level. However, in analogy with the surf zone, saturation can occur in the swash zone when incident wave bores interact with the run-up or run-down of previous waves. The measure for swash-swash interactions, and hence swash saturation, is described by the ratio between the incident wave period T and the natural swash period Ts . Through equation of the theoretical run-up potential of Baldock and Holmes (1999) and the empirical run-up equation of Hunt (1959), Brocchini and Baldock (2008) estimate the ratio between the incident wave period and the natural swash period as a function of the offshore wave height H0 : 1 ( ) 14 ( K2 H ) 4 Ts 2 0 (2.9) =2 T 𝜋 gT 2 𝛽 2 where K is a constant in the range of 0.6–0.8 and swash saturation occurs when Ts > T. The equation shows that the swash zone is saturated on most natural beaches, with the exception of long-period swell on steep beaches (Guza & Thornton, 1982; Brocchini & Baldock, 2008). The importance of swash saturation during storms is that once the swash is saturated, an increase in the incident wave height can lead to additional mean water level setup in the swash zone, but not to increased variance of the run-up at the incident wave frequency. This concept is depicted in Figure 2.8, where in a series of numerical

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

37

Wave height, wave run-up and significant swash amplitude

Offshore significant wave height (m) 2% Wave run-up exceedence (m) Significant swash amplitude (m)

1.6 Hm,0

1.4

H2% Sa

1.2 1 0.8 0.6 0.4 0.2 0

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

Ts/Tp (-)

Figure 2.8 Example of incident wave height, wave run-up and significant swash amplitude of a JONSWAP spectrum with constant peak period, as a function of the ratio between the swash period and the peak period. Results are computed using the wave-resolving model XBeach-G (McCall et al., 2014) on a 1/20 slope.

simulations the swash period is increased with respect to the incident wave period through an increase in the offshore significant wave height (Equation 2.9). The results of the simulation show that although the 2% wave run-up exceedence R2% does increase with increasing offshore wave heights Hm,0 , the significant swash amplitude around the peak frequency Sa does not increase once Ts ∕Tp >∼ 0.9. Partly due to swash zone saturation, swash motions are typically dominated by low-frequency components (Brocchini & Baldock, 2008). On dissipative beaches, the primary generation mechanism for swash motions is low-frequency (infragravity) waves generated outside the swash zone (Guza & Thornton, 1982). On reflective beaches, high frequency wave energy will enter the swash zone, generating high frequency swash motions. However, the persistence of wave groupiness on reflective beaches, as well as the damping of high-frequency motions through swash saturation, generally lead to a dominance of low-frequency motions over the incident high-frequency wave band motions in the swash zone (Mase, 1995; Baldock et al., 1997). While low-frequency components originating outside the swash zone dominate motions in the swash, the swash zone itself is generally not a source of low-frequency wave energy. Watson et al. (1994) showed that for the swash zone to generate low-frequency wave energy, the ratio between the wave group period Tg and the natural swash period Ts should be close to unity. Since Ts is in the order of 1–3 on most natural beaches, such low frequency wave generation in the swash will only occur for very short wave groups (1–3 waves per group; Brocchini & Baldock, 2008). The swash zone is in the interface between the sea and the land, and hence also an area where interaction occurs between the sea and groundwater in the beach. Infiltration into the permeable beach has long been known to affect swash hydrodynamics (Bagnold, 1940; Grant, 1948) through weakening of the backwash with respect to the uprush, leading to so-called swash asymmetry. While this asymmetry

38

COASTAL STORMS: PROCESSES AND IMPACTS

occurs on all permeable beaches to a certain extent, Masselink & Li (2001) showed through numerical model simulations that swash asymmetry through infiltration losses becomes important to beach morphology on beaches with a median beach grain size D50 exceeding 1.5 mm (very coarse sand). Infiltration in the upper beach and on the barrier crest also increases coastal safety against flooding on coarse sand and gravel barriers during storms by significantly reducing overtopping and overwash rates (McCall et al., 2012). Infiltration and exfiltration further affect flow in the swash through drawing streamlines closer to the bed during infiltration, thereby increasing the effective bed shear stress, and pushing streamlines farther from the bed during exfiltration, thereby reducing the effective bed shear stress (Nielsen, 1992; Conley & Inman, 1994). The net effect of this ventilated boundary layer is to enhance the bed shear stress during the uprush phase and reduce the bed shear stress during the backwash phase by approximately 5% on sandy beaches (Butt et al., 2001) and by as much as 20–40% on permeable gravel beaches (Masselink & Turner, 2012).

2.4 Conclusions and future challenges At the beginning of this chapter, we saw that storm surges were controlled mainly by atmospheric pressure gradients, wind-induced surface stress and wave forces, although other phenomenon such as resonance may play a major role locally. While the role of atmospheric pressure gradients in now well understood and taken into account in storm surge models, the parameterization of the surface stress, its dependence on the sea state under multi-modal spectra or extreme wind is still a matter of debate and requires further research. Also, we showed that wave setup could propagate outside surf zones and therefore significantly contribute to storm surges. While wave setup is usually well predicted under low to moderate energy, the parameterization of bottom stress and turbulence may become critical under storms, as basic models ignoring these processes usually strongly underestimate wave setup. In the nearshore, we also saw that IG waves can develop, driven by the release of a bound wave that develops in the shoreface or by breakpoint movements associated with wave groups. Although the mechanisms associated with IG wave generation are well established and supported by observations, the fate of IG waves once released in the surf zone is still a matter of debate. Understanding their dissipation or trapping by refraction leading edge waves will require additional research. Finally, the effect of surf zone dynamics, including infragravity waves, and beach face morphology on storm wave run-up in the swash zone is only partially understood and captured by current empirical and numerical models. Further investigation of the hydrodynamics, as well as the morphodynamics, in the swash zone during high-energy events will provide insights for more accurate predictions of coastal safety against flooding.

Acknowledgements This study is a contribution of the European FP7 project Risc-Kit (Grant Agreement n∘ 603458) and also benefited from the project “Submersion” carried out at UMR LIENSs from 2011 to 2013.

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

39

References Apotsos, A., Raubenheimer, B., Elgar, S., Guza, R.T. & Smith, J.A. (2007) Effects of wave rollers and bottom stress on wave setup. Journal of Geophysical Research, 112, C02003, doi:10.1029/2006JC003549. Arnaud, G. & Bertin, X. (2014) Contribution du setup induit par les vagues dans la surcote associée à la tempête klaus. XIII èmes Journées Nationales Génie Côtier _ Génie Civil, pp. 859–867. Aucan, J., Hoeke, R. & Merrifield, M.A. (2012) Wave-driven sea level anomalies at the midway tide gauge as an index of North Pacific storminess over the past 60 years. Geophysical Research Letters, 39 (17), L17603. Bagnold, R. (1940) Beach formation by waves; some model-experiments in a wave tank. Journal of the ICE, 15, 27–52. Baldock, T. & Holmes, P. (1999) Simulation and prediction of swash oscillations on a steep beach. Coastal Engineering, 36 (3), 219–242. URL http://www.sciencedirect.com/science/article/pii/ S0378383999000113. Baldock, T.E & Huntley, D.A. (2002) Long–wave forcing by the breaking of random gravity waves on a beach. Proc. R. Soc. London, Ser. A, 458 (2025), 2177–2201. Baldock, T.E., Holmes, P. & Horn, D. (1997) Low frequency swash motion induced by wave grouping. Coastal Engineering, 32, 197–222. Baldock, T.E., Huntley, D.A., Bird, P.A.D., O’Hare, T.J. & Bullock, G.N. (2000) Breakpoint generated surf beat induced by bichromatic wave groups. Coastal Eng., 39, 213–242, doi:10.1016/S0378-3839(99)00061-7. Baldock, T.E., O’Hare, T.J. & Huntley, D.A. (2004) Long wave forcing on a barred beach. J. Fluid Mech., 503, 321–343. Battjes, J.A. (1974) Surf similarity. In: 14th International Conference on Coastal Engineering. Vol. Volume I, Chapter 26. Copenhagen, pp. 467–479. Battjes, J.A., Bakkenes, H.J., Janssen, T.T. & van Dongeren, A.R. (2004) Shoaling of subharmonic gravity waves, J. Geophys. Res., 109, C02009, doi:10.1029/2003JC001863. Bertin, X., Bruneau, N., Breilh, J.F., Fortunato, A.B. & Karpytchev, M. (2012) Importance of wave age and resonance in storm surges: The case Xynthia, Bay of Biscay. Ocean Modelling, 42 (2012), 16–30. Bertin, X., Castelle, B., Chaumillon, E., Butel, R. & Quique, R, (2008) Estimation and inter-annual variability of the longshore transport at a high-energy dissipative beach: The St Trojan beach, SW Oléron Island, France. Continental Shelf Research, 28, 1316–1332. Bertin X., Fortunato, A.B. & Oliveira A. (2009) A modeling-based analysis of processes driving wave-dominated inlets. Continental Shelf Research, 29, 819–834. Bertin, X., Li, K., Roland, A. & Bidlot, J.R. (2015) The contributions of short-waves in storm surges: Two case studies in the Bay of Biscay. Continental Shelf Research, 96, 1–15. Bertin, X. & Olabarrieta, M. (2016) Relevance of infragravity waves in a wave-dominated inlet. Journal of Geophysical Research: Oceans 121, doi:10.1002/2015JC011444. Blake, E.S. (2007) The deadliest, costliest and most intense United States tropical cyclones from 1851 to 2006 (and other frequently requested hurricane facts). NOAA Technical Memorandum NWS TPC 5. Bowers, E.C. (1977) Harbour resonance due to set-down beneath wave groups. J. Fluid Mech., 79, 71–92. Brocchini, M. & Baldock, T.E. (2008) Recent advances in modeling swash zone dynamics: Influence of surf-swash interaction on nearshore hydrodynamics and morphodynamics. Reviews of Geophysics, 46 (3). URL 10.1029/2006RG000215.

40

COASTAL STORMS: PROCESSES AND IMPACTS

Brown, J.M. & Wolf, J. (2009) Coupled wave and surge modelling for the eastern Irish Sea and implications for model wind-stress. Continental Shelf Research, 29 (10), 1329–1342. Butt, T., Russell, P. & Turner, I. (2001) The influence of swash infiltration-exfiltration on beach face sediment transport: Onshore or offshore? Coastal Engineering, 42 (1), pp. 35–52, ISSN 0378-3839, 10.1016/S0378-3839(00)00046-6. Charnock, H. (1955) Wind stress on a water surface. Quarterly Journal of the Royal Meteorological Society, 81, 639–640. Conley, D.C. & Inman, D.L. (1994) Ventilated oscillatory boundary layers. Journal of Fluid Mechanics 273, 261–284. Dietrich, J.C., Bunya, S., Westerink, J.J., Ebersole, B.A., Smith, J.M., Atkinson, J.H., et al. (2010) A high-resolution coupled riverine flow, tide, wind, wind wave, and storm surge model for southern Louisiana and Mississippi. Part II: Synoptic description and analysis of hHurricanes Katrina and Rita. Monthly Weather Review, 138 (2), 378–404. Dodet, G., Bertin, X., Bruneau, N., Fortunato, A.B., Nahon, A. & Roland, A. (2013) Wave-current interactions in a wave-dominated tidal inlet. Journal of Geophysical Research, 118 (C3), 1587–1605., doi: 10.1016/j.coastaleng.2009.08.006. Donelan, M.A., Haus, B.K., Reul, N., Plant, W.J., Stiassnie, M., Graber, H.C., et al. (2004) On the limiting aerodynamic roughness of the ocean in very strong winds. Geophys. Res. Lett., 31, L18306, doi:10.1029/2004GL019460. Dong, G., Ma, X., Xu, J., Ma, Y. & Wang, G. (2009) Experimental study of the transformation of bound long waves over a mild slope with ambient currents. Coastal Engineering, 56, 1035–1042. van Dongeren, A.R., Bakkenes, ,H.J. & Janssen, T.T. (2002) Generation of long waves by short wave groups. In: J. Smith McKee (Ed.) Coastal Engineering. World Sci., River Edge, N.J. van Dongeren, A.R., Lowe, R., Pomeroy, A., Trang, D.M., Roelvink, D., Symonds, G., et al. (2013) Numerical modeling of low-frequency wave dynamics over a fringing coral reef. Coastal. Engineering, 73, 178e190. Doodson, A.T. (1924) Meteorological perturbations of sea-level and tides. Geophysical Journal International, 1, 124–147. Elfrink, B. & Baldock, T. (2002) Hydrodynamics and sediment transport in the swash zone: A review and perspectives. Coastal Engineering, 45, 149–167. Garcez-Faria, A.F., Thornton, E.B., Lippmann, T.C. & Stanton, T.P. (2000) Undertows over a barred beach. Journal of Geophysical Research, 105, (C7), 16999–17010. Grant, U. (1948) Influence of the water table on beach aggradation and degradation. Journal of Marine Research, 7, 655–660. Greenspan, H.P. (1956) The generation of edge waves by moving pressure distributions. J Fluid Mech, 1 (6), 574–592. Guza, R.T. & Thornton, E.B. (1982) Swash oscillations on a natural beach. Journal of Geophysical Research, 87 (C1), 483–491. Guza, R.T. & Thornton, E.B. (1985) Velocity Moments in Nearshore. Journal of Waterway Port Coastal and Ocean Engineering-Asce, 111, 235–256. Guza, R.T., Thornton, E.B. & Holman, R.A. (1984) Swash on steep and shallow beaches. In: Proceedings of the Coastal Engineering Conference, 1984, edited by B.L. Edge, pp. 708–723, American Society of Civil Engineers. Reston, Va. Guza, R.T., Thornton, E.B. & Holman, R.A.(1985) Swash on steep and shallow beaches. In: Proceedings of the 19th Conference on Coastal Engineering, pp. 708–723, American Society of Civil Engineers. New York. Hawkins, H.F. & Rubsam, D.T. (1968) Hurricane Hilda, 1964 II. Structure and budgets of the hurricane on October 1, 1964. Monthly. Weather Review, 96 (9), 617–636, doi:10.1175/1520-0493.

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

41

Henderson, S.M. & Bowen, A.J. (2002) Observations of surf beat forcing and dissipation, J. Geophys. Res., 107 (C11), 3193, doi:10.1029/ 2000JC000498. Herbers, T.H.C., Elgar, S., Guza, R.T. & Reilly, W.C.O. (1995) Infragravity frequency (0.005–0.05Hz) motions onthe shelf. Part II: Free waves. Journal of Physical Oceanography, 25 (6), 1063–1079, doi:10.1175/1520-0485(1995)0252.0.CO;2. Holman, R. A. & Bowen, A. (1984) Longshore structure of infragravity wave motions. J. Geophys. Res., 89, 6446–6452, doi:10.1029/ JC089iC04p06446. Holman, R.A. & Sallenger, A.H. (1985) Setup and swash on a natural beach, J. Geophys. Res., 90, 945–953, doi:10.1029/JC090iC01p00945. Holmedal, L.E. & Myrhaug, D. (2013) Combined tidal and wind driven flows and bedload transport over a flat bottom. Ocean Modelling, 68, 37–56. Hunt, I. (1959) Design of seawalls and breakwaters. In: Proc. Am. Soc. Civ. Eng., J. Waterw. Harbors Div., 85, 123–152. Huntley, D.A., Guza, R.T. & Bowen, A.J. (1977) A universal form for shoreline run-up spectra?, J. Geophys. Res., 82 (C18), 2577–2581. Janssen, T.T., Battjes, J.A. & van Dongeren, A.R. (2003) Long waves induced by short-wave groups over a sloping bottom, J. Geophys. Res., 108 (C8), 3252, doi:10.1029/2002JC001515. Jonsson, I.G. (1966) Wave boundary layers and friction factors. Proceedings of the 10th Conference on Coastal Engineering, 1, 127–148. Kennedy, A.B., Gravois, U., Zachry, B.C., Westerink, J.J., Hope, M.E., Dietrich, J.C., et al. (2011) Origin of the Hurricane Ike forerunner surge. Geophysical Research Letter, 28 (8), L08608. Kennedy, A.B., Westerink, J.J., Smith, J.M., Hope, M.E., Hartman, M., Taflanidis, A.A., et al. (2012) Tropical cyclone inundation potential on the Hawaiian Islands of Oahu and Kauai. Ocean Modelling, 52–53, 54–68. doi:10.1016/j.ocemod.2012.04.009. Kostense, J.K. (1984) Measurements of surf beat and set-down beneath wave groups. Proc. 19th Int. Conf. Coastal Eng. ASCE, Houston, pp. 724–740. List, J.H. (1992) A model for the generation of two-dimensional surfbeat. J. Geophys. Res., 97, 5623–5635. Longuet-Higgins, M.S. & Stewart, R.W. (1962) Radiation stress and mass transport in gravity waves, with application to ‘surf beats’. J. Fluid Mech., 13, 481–504. Longuet-Higgins, M.S., Stewart, R.W. (1964) Radiation stresses in water waves: A Physical discussion, with applications. Deep-Sea Research, 11, 529–562. López, M., Iglesias, G. & Kobayashi, N. (2012) Long period oscillations and tidal level in the Porto Ferrol. Appl. Ocean Res., 38, 126–134. Mase, H. (1995) Frequency down-shift of swash oscillations compared to incident waves. Journal of Hydraulic Research, 33 (3), 397–411. URL 10.1080/00221689509498580. Masselink, G. (1995) Group bound long waves as a source of infragravity energy in the surf zone. Cont. Shelf Res., 15, 1525–1547. Masselink, G. & Li, L. (2001) The role of swash infiltration in determining the beachface gradient: A numerical study. Marine Geology, 176, 139–156. URL http://www.sciencedirect.com/ science/article/pii/S002532270100161X. Masselink, G. & Puleo, J.A. (2006) Swash-zone morphodynamics. Continental Shelf Research, 26, 661–680. Masselink, G. & Turner, I.L. (2012) Large-scale laboratory investigation into the effect of varying back-barrier lagoon water levels on gravel beach morphology and swash zone sediment transport. Coastal Engineering, 63, 23–38. ISSN 0378-3839, 10.1016/j.coastaleng.2011.12.007.

42

COASTAL STORMS: PROCESSES AND IMPACTS

Mastenbroek, C., Burgers, G. & Janssen, P.A.E.M. (1993) The dynamical coupling of a wave model and a storm surge model through the Atmospheric Boundary Layer. Journal of physical Oceanography, 23, 1856–1866. McCall, R., Masselink, G., Poate, T., Roelvink, J.A., Almeida, L.P., Davidson, M., et al. (2014) Modelling storm hydrodynamics on gravel beaches with XBeach-G. Coastal Engineering, 91, 231–250. ISSN 0378-3839, 10.1016/j.coastaleng.2014.06.007. McCall, R., Masselink, G., Roelvink, J., Russell, P., Davidson, M. & Poate, T. (2012) Modeling overwash and infiltration on gravel barriers. In: Proceedings of the 33rd International Conference on Coastal Engineering. Santander, Spain. McComb, P., Gorman, R. & Goring, D. (2005) Forecasting infragravity wave energy within a harbour. In: B.L.E.J.C. Santas (Ed.), Proceedings of the Fifth International Symposium on Ocean Wave Measurement andAnalysis (WAVES). IAHR Secretariat, Madrid, Spain. 3–7 July. van der Molen, W., Monárdez-Santander, P. & van Dongeren, A.R. (2006) Numerical simulation of long-period waves and ship motions inTomakomai Port, Japan. Coast. Eng. J., 48 (1), 59–79. Monserrat, S., Vilibic, I. & Rabinovich, A.B. (2006) Meteotsunamis: Atmospherically induced destructive ocean waves in the tsunami frequency band. Nat Hazards Earth Syst Sci, 6 (6), 1035–1051. Moon, I.J. (2005) Impact of a coupled ocean wave-tide-circulation system on coastal modelling. Ocean Modelling, 8, 203–236. Munk, W.H. (1949) Surf beats. Eos Trans AGU, 30, 849–854. Nielsen, P. (1992) Coastal bottom boundary layers and sediment transport. Vol. 4 of Advanced Series on Ocean Engineering. World Scientific, Singapore. Nielsen, P. & Hanslow, D.J. (1991) Wave run-up distributions on natural beaches. Journal of Coastal Research, 1139–1152. Okihiro, M. & Guza, R.T. (1996) Observations of seiche forcing and amplification in three small harbours. J. Waterw. Port Coast. Ocean Eng., 122 (5), 232–238. Olabarrieta, M., Warner, J.C., Armstrong, B., Zambon, J.B. & He, R. (2012) Ocean-atmosphere dynamics during Hurricane Ida and Nor’Ida: an application of the coupled ocean-atmosphere wave sediment transport (COAWST) modeling system. Ocean Modelling, 43–44, 112–137. Pomeroy, A., Lowe, R., Symonds, G., van Dongeren, A. & Moore, C. (2012) The dynamics of infragravity wave transformation over a fringing reef. J. Geophys. Res., 117, C11022, doi:10.1029/2012JC008310. Pond, S. & Pickard, G.L. (1998) Introductory Dynamical Oceanography. Butterworth-Heinmann. Powell, K.A. (1990) Predicting short term profile response for shingle beaches. Tech. rep., HR Wallingford SR report 219. Powell, M.D., Vickery, P.J. & Reinhold, T.A. (2003) Reduced drag coefficient for high wind speeds in tropical cyclones. Nature, 422, 279–283. Proudman, J. (1929) The effects on the sea of changes in atmospheric pressure. Geophys Suppl Mon Not R Astron Soc, 2, 197–209. Rabinovich, A.B. (2009) Seiches and harbour oscillations. In: Y.C. Kim (Ed.) Handbook of coastal and ocean engineering. World Scientific, Singapore, pp. 193–236. Rego, J.L. & Li, C. (2010) Nonlinear terms in storm surge predictions: effect of tide and shelf geometry with case study from Hurricane Rita. Journal of Geophysical Research, 115, C06020. Roland, A., Zhang, Y., Wang, H.V., Meng, Y., Teng, Y.C., Maderich, V., et al. (2012) A fully coupled 3D wave-current interaction model on unstructured grids. Journal of Geophysical Research, 117: doi:10.1029/2012JC007952. Ruessink, B.G., Kleinhans, M.G. & van den Beukel, P.G.L. (1998) Observations of swash under highly dissipative conditions. Journal of Geophysical Research: Oceans, 103 (C2), 3111–3118. URL 10.1029/97JC02791

CH2 HYDRODYNAMICS UNDER STORM CONDITIONS

43

Ruggiero, P., Holman, R.A. & Beach, R.A. (2004) Wave run-up on a high-energy dissipative beach. J. Geophys. Res., 109, C06025, doi:10.1029/2003JC002160. Ruggiero, P., Komar, P.D., McDougal, W.G., Marra, J.J. & Beach, R.A. (2001) Wave run-up, extreme water levels and the erosion of properties backing beaches. Journal of Coastal Research, 407–419. Saha, S., Moorthi, S., Pan, H.-L., Wu, X., Wang, J. & Nadiga, S. (2010) The NCEP climate fore-cast system reanalysis. Bull. Am. Meteorol. Soc., 91, 1015–1057. Sallenger, A. (2000) Storm impact scale for barrier islands. Journal of Coastal Research, 16 (3), 890–895. Schäffer, H.A. & Svendsen, I.A. (1988) Surf beat generation on a mild slope. In: Coastal Engineering, pp. 1058–1072. Am. Soc. of Civ. Eng., Reston, Va. Schäffer, H.A., Madsen, P.A. & Deigaard, R. (1993) A Boussinesq model for waves breaking in shallow water. Coastal Engineering, 20 (3–4), 185–202. Senechal, N., Coco, G., Bryan, K.R. & Holman, R.A. (2011) Wave run-up during extreme storm conditions, J. Geophys. Res., 116, C07032, doi:10.1029/2010JC006819. Sheremet, A., Kaihatu, J.M, Su, S.F., Smith, E.R. & Smith, J.M. (2011) Modeling of nonlinear wave propagation over fringing reefs. Coastal Eng., 58 (12), 1125–1137. doi:10.1016/j.coastaleng.2011.06.007. Stewart, R.W. (1974) The air-sea momentum exchange. Boundary Layer Meteorology, 6, 151–167. Stockdon, H., Holman, R., Howd, P. & Sallenger, Jr., A. (2006) Empirical parameterization of setup, swash, and run-up. Coastal Engineering, 53, 573–588. Stokes, G.G. (1847) On the theory of oscillatory waves. Trans. Camb. Philos. Soc., 8, 441–473 Symonds, G., Huntley, D.A. & Bowen, A.J. (1982) Two dimensional surfbeat: Long wave generation by a time varying breakpoint. J. Geophys. Res., 87, 492–498. Swart, D.H. (1974) Offshore sediment transport and equilibrium beach profiles. Delft Hydraulics Lab Publication 131. Delft, the Netherlands. Takagaki, N., Komori, S., Suzuki, N., Iwano, K., Kuramoto, T., Shimada, S., et al. (2012) Strong correlation between the drag coefficient and the shape of the wind sea spectrum over a broad range of wind. Geophysical Research Letters, 39, L23604. Tanaka, S., Westerink, H., Cheung, K.F., Smith, T., Hamann, M., Minamide, M., et al. (2012) Tropical cyclone inundation potential on the Hawaiian Islands of Oahu and Kauai. Ocean Modelling, 52–53, 54–68. Thotagamuwage, D.T. & Pattiaratchi, C.B. (2014) Observations of infragravity period oscillations in a small marina. Ocean Eng., 88 (2014), 435–445. Tucker, M.J. (1950) Surf beats: Sea waves of 1 to 5 min. period. Proc. R. Soc. London, Ser. A, 202, 565– 573. Watson, G., Barnes, T. & Peregrine, D. (1994) The generation of low-frequency waves by a single wave group incident on a beach. In: Proceedings of 24th Conference on Coastal Engineering, Kobe, Japan. Vol. 1. Xie, L., Kiu, H., Liu, B. & Bao, S. (2011) A numerical study of the effect of hurricane wind asymmetry on storm surge and inundation. Ocean Model, 36, 71–79. Zhang, Y. & Baptista, A.M. (2008) SELFE: A semi-implicit Eulerian-Lagrangian finite-element model for cross-scale ocean circulation. Ocean Modelling, 21 (3–4), 71–96.

3 Sediment Transport Under Storm Conditions on Sandy Beaches Troels Aagaard and Aart Kroon Department of Geoscience and Natural Resources, University of Copenhagen, Copenhagen, Denmark

3.1

Introduction

Coastal erosion is defined as the movement of unconsolidated sediment or rock fragments from a subaerial to a subaqueous elevation on the coast causing a reduction in the volume of material contained in subaerial beaches, dunes and cliffs. Coastal erosion involves the movement of material and is distinct from passive coastal inundation that occurs, for example, as a result of sea level rise. On sandy beaches, coastal erosion is mainly driven by the action of waves and currents and it typically occurs under storm conditions. Morphologic change, including coastal erosion, due to wave and current processes is the result of gradients in the transport of sediment and can be calculated from the sediment volume continuity equation: dh =− dt

(

dqx dqy + dx dy

) (3.1)

where h is seabed elevation, qx , qy are the sediment transport rates in the cross-shore (x) and longshore (y) directions, and t is time. In order to understand the sediment transport processes that lead to coastal erosion on sandy beaches, it is convenient to distinguish between chronic erosion and punctuated erosion. Chronic erosion occurs when a coastal sector experiences persistent erosional loss on annual or decadal time scales. It is most commonly associated with systematic long-term gradients in the alongshore transport of sediment along the coast and not necessarily associated with storm wave activity. The longshore sediment transport is driven by waves that are obliquely incident to the coastline, and beaches erode if the net annual transport rate increases

Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

46

COASTAL STORMS: PROCESSES AND IMPACTS

in the downdrift direction (which causes a positive gradient in the longshore sediment transport). A positive transport gradient is related to increasing wave incidence angle and/or increasing wave height in a downdrift direction. Punctuated erosion is distinct from chronic erosion in the way that it is realised during a brief period of time, for example during a single (or a succession of) storm event(s) and, contrary to chronic erosion, it is mainly driven by cross-shore sediment transport processes. Gradients in the cross-shore sediment transport are in most cases much larger than longshore transport gradients; cross-shore transport directions and magnitudes are highly variable, changing rapidly under different wave conditions since several transport mechanisms interact, as will be detailed below. This chapter is mainly concerned with event-driven sediment transport during storms that lead to punctuated erosion.

3.2 Morphological consequences of coastal storms Prior to a discussion of storm-related coastal sediment transport processes, it is useful to consider briefly the morphological changes that typically occur during storms. An example coastal profile for a non-tidal beach is shown in Figure 3.1. Different wave and current regimes dominate across the profile; the lower shoreface is affected by shoaling wave processes that tend to move sand onshore, while wind- and tide-generated currents are additional significant contributors to the transport of sediment, particularly in the alongshore dimension, but also as up/downwelling currents during strong winds. The morphologically active upper shoreface that often comprises one or more nearshore bars is dominated by breaking wave processes in the surf zone. In addition to wind waves and swell incident from offshore, long-period infragravity waves and wave-generated mean currents are important in the movement of sediment here. The interaction between large horizontal gradients in suspended sediment load associated with zones of wave breaking and wave reformation across the upper shoreface, and the several types of nonlinearly interacting wave and current processes impose temporally and spatially variable sediment transport directions and rates, which may result in

16

upper shoreface lower shoreface surf zone shoaling wave zone

dune

elevation, m

12 8 4 0 –4 –8 –12

nearshore bars

MWL

backshore beach face swash zone 0

1000 2000 Distance from shoreline, m

3000

Figure 3.1 Example cross-shore profile for a non-tidal coast showing the morphological and hydrodynamic zones (italics) considered in this chapter. MWL is mean water level. The beach comprises the beach face and the backshore.

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

47

rapidly changing morphology. Note that Figure 3.1 shows the surf zone covering the full extent of the upper shoreface; this will only be the case during high-energy conditions. During fair-weather conditions, the surf zone contracts and covers only the inner part of the upper shoreface. Finally, the beach face is affected by the high-velocity uprush and backwash flows of the swash zone, while the backshore is dry under most conditions and sediment transport is in most cases accomplished by aeolian processes. During storms, however, the backshore is often inundated due to storm surge, large wave setup and infragravity motions (see Chapter 2), such that a swash zone is absent and the surf zone may extend to the base of the dunes where waves may directly attack the dune wall. For a beach exposed to a significant tidal range, an intertidal zone is located in between the beach face and the upper shoreface. The intertidal zone may comprise a number of intertidal bars and, depending upon the wave energy level and the tidal range, it may be subjected to the full range of swash, surf and shoaling wave processes over a tidal cycle. Sandy beaches may be classified into a sequence of morphodynamic states, ranging from dissipative beaches to reflective beaches and encompassing a finite number of intermediate states (Wright & Short, 1984; Scott et al., 2011). The dissipative end state exhibits a wide surf zone that may comprise up to ten surf bores, or more. The upper shoreface is gently sloping, with one or more subdued nearshore bars, waves typically break through spilling and longshore variation in morphology is small. In this state, the largest subaqueous sediment storage occurs. The opposite fair-weather end member of the sequence, the reflective beach is characterized by maximum subaerial sediment storage. Nearshore bars are absent in this state, unless they are dormant and inactive, and sediment is largely stored on the beach and in a berm. A surf zone does not exist and waves break at the beach face through plunging or surging. In between these two end states are a number of intermediate states that feature more pronounced bar topography and the bars possess varying degrees of alongshore three-dimensionality with alternating crescentic or transverse bar horns and bays. When the morphology is active (because significant sediment transport occurs) and in equilibrium with the hydrodynamic forcing, these states may be distinguished by the non-dimensional sediment fall velocity parameter (Gourlay, 1968): Ω=

Hb ws T

(3.2)

Where Hb is breaker height, ws is the sediment fall velocity and T is the wave period. For Ω > 5.5, the beach will be dissipative, whereas the reflective state occurs for Ω < 1.5 and the intermediate states occupy the range in between (Short, 1999). Breaker wave heights increase during storms and at the same time wave periods often decrease. This forces the beach towards the dissipative end state, but because of hysteresis it does not necessarily fully reach that state. If a storm arrives at a time when the beach is in an accreted reflective state, the beach face and berm erode (Masselink et al., 2008), sand is transferred to the upper shoreface and if a storm lasts long enough, a nearshore bar forms. If the beach is in an intermediate state at the onset of storm conditions and bar(s) are already present, the bar(s) typically migrate offshore (e.g. Thornton et al., 1996; Komar, 1998; Marino-Tapia et al., 2007; Ruessink et al.,

48

COASTAL STORMS: PROCESSES AND IMPACTS

2009). Fair-weather wave conditions after storms drive the beach back towards the reflective end state again and cause bars to move onshore. However, this classic profile evolution model is very simple and beach response to storms strongly depends on, for example, antecedent morphology. If the subaerial beach (and intertidal zone) is already flat and gently sloping, beach response to even intense storms may be surprisingly subdued (Aagaard et al., 2005; Kroon et al., 2007). Moreover, bar migration is not always offshore directed during dissipative storm conditions. On some beaches, typically possessing a gently sloping upper shoreface and exposed to significant changes in mean water level due to tides and/or storm surge activity, storm-driven bar migration is often onshore directed (Aagaard et al., 2004; Lindhorst et al., 2008; Bruneau et al., 2009; Anthony, 2013).

3.3 Sediment transport processes during storms The numerous field measurements of coastal sediment transport that have been conducted on the upper shoreface over the past approximately three decades have shown that net cross-shore transport at a given location on the cross-shore profile is complex. It is a result of several individual sediment transport components driven by a range of hydrodynamic processes and mediated by the existing morphology. Nevertheless, cross-shore sediment transport is now fairly well understood in a qualitative sense. Sediment transport can occur both in suspension and as bedload, but since bed shear stresses are large during storms, it is highly likely that suspended transport is dominant on the shoreface, which is consistent with the limited field experimental evidence that is available (Masselink et al., 2007; see also Aagaard & Hughes, 2013) Sediment is suspended from the seabed by bed shear stresses mainly within oscillatory wave boundary layers, although recent research has called attention to the fact that under strongly breaking (particularly plunging) waves, surface-injected turbulence may be a sometimes highly significant additional source of bed shear stress (Grasso et al., 2012) and sediment suspension (Scott et al., 2009). The mobilized sand can then be transported by a range of hydrodynamic processes and net cross-shore (suspended) sediment transport on the upper (and lower) shoreface is controlled by the balance between transport components driven by oscillatory wave motions at different frequencies and the component driven by wave-generated steady currents (Figure 3.2). Incoming shoaling waves (wind waves and swell) develop asymmetries both in their surface shape and in the associated velocity field (Freilich & Guza, 1984). The asymmetry initially develops about the horizontal axis (termed wave skewness, which causes orbital velocity skewness) such that onshore-directed orbital velocity beneath wave crests increases in magnitude but decreases in duration. On the other hand, wave troughs are shallow and longer, which cause weaker, but more prolonged offshore-directed orbital velocities. Maximum skewness typically occurs close to the wave breakpoint (Ruessink et al., 2012) and since sediment transport is a nonlinear function of fluid velocity, velocity skewness causes a net onshore-directed, wave-driven transport of sediment when integrated over a wave cycle. However, at least two exceptions to this simple concept exist. First, steep ripples on the seabed cause phase shifts between orbital velocity and sediment concentration, and can ultimately cause a reversal of the wave-driven suspended sediment transport component (O’Hara

distance above the seabed, m

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

0.25

0.25

0.2

0.2

0.15

0.15

0.1

0.1

0.05

0.05

0 –2

–1

0

1

sediment flux, kgm–2s–1 (a)

2

0 –3

–2

–1

49

0

1

2

sediment flux, kgm–2s–1 (b)

Figure 3.2 Example vertical profiles of cross-shore suspended sediment transport components during storms. Positive transport is onshore. Circles indicate oscillatory transport at incoming wave frequencies, dots are oscillatory transports at infragravity (IG) wave frequencies and crosses are Eulerian mean transports (due to undertow). The sum of these three components represents the net transport, which is indicated by squares. The data was collected using optical and fibre-optical backscatter sensors in the inner surf zone (h̄ ≈ 1 m) at Egmond Beach (the Netherlands) during a storm with offshore significant wave heights of ≈ 4 m. The IG transport component was oriented oppositely in the two cases, which strongly affected net transport magnitude.

Murray et al., 2011), although ripples are less likely to occur during storms in shallow water because bed shear stresses are large. Second, wave-driven transport may be offshore directed if bed shear stresses due to seaward-directed mean currents are sufficiently large to cause significantly increased sediment loads on the offshore wave stroke compared to the onshore stroke, and wave-driven transport may then become offshore directed. Landward of the wave breakpoint, wave skewness decreases and is gradually replaced by wave asymmetry associated with sawtooth-shaped surf bores. Wave asymmetry causes orbital acceleration skewness which again causes: (1) a thinner wave bottom boundary layer and thus larger bed shear stresses under the front face of the waves compared to the back slope of the wave; and (2) horizontal pressure gradients – larger gradients occur with the strongly accelerating flows under the wave front. These processes both increase sediment stirring and mobilization on the onshore wave phase relative to the offshore phase (Drake & Calantoni, 2001; Hsu & Hanes, 2004) and drive an onshore transport of sediment. Sediment flux predictions using a Meyer-Peter/Müller type of model (Austin et al., 2009) were more consistent with observed fluxes when fluid acceleration was included in the model, and large sediment suspension events tend to correlate well with large fluid accelerations (Puleo et al., 2003; Houser & Greenwood, 2007). Net oscillatory cross-shore transport of sediment caused by skewness and/or asymmetry of the incoming wind and swell waves is either augmented or opposed by transport due to (quasi-) steady Lagrangian and Eulerian currents. The former are related directly to the wave motions. Since wave orbital diameter depends on

50

COASTAL STORMS: PROCESSES AND IMPACTS

distance above the seabed, the diameter is larger under a wave crest than under a wave trough, which results in net onshore fluid mass transport over a wave cycle. The mass transport (Stokes drift; Xu & Bowen, 1994) is maximum at the free surface and decreases towards the seabed, and it cannot be measured using fixed instruments. The Stokes drift feeds a near-bed, offshore-directed (Eulerian) return current, the undertow, which ensures cross-shore mass continuity and which has its maximum near the seabed. Seaward of the wave breakpoint, Stokes drift and undertow are largely balanced in the vertical (Lentz & Fewings, 2012). Within the surf zone, however, undertow is enhanced significantly because of the added contribution of onshore mass transport in the upper part of the water column carried by surf bores/rollers (Svendsen, 1984). Undertow, which is more fully described in Chapter 2, often drives a large offshore directed sediment transport in the surf zone, especially during storm conditions, when wave dissipation is large (Figure 3.2). The speed of the undertow depends on the negative cross-shore gradient of wave radiation stress and this gradient again depends partly on beach slope; for given wave conditions, undertow is stronger on steeply sloping beaches than on gently sloping beaches (Longuet-Higgins, 1983; Aagaard et al., 2002). Hence, there are strong feedbacks between morphology and sediment transport. Within the surf zone, offshore-directed sediment transport due to undertow generally exceeds, and often significantly, the onshore-directed transport due to wind waves and swell, such that the net suspended sediment transport is offshore-directed (Figure 3.2), and this is particularly the case during storms when the beach is dissipative (Thornton et al., 1996; Conley & Beach, 2003; Aagaard et al., 2013). Hence, undertow is the main mechanism for net offshore sediment transfer and flattening of the beach (and intertidal zone) during storms. These patterns apply mostly in cases when the morphology is quasi-linear alongshore and the nearshore bars are parallel to the shoreline, such as on a dissipative beach. If the bar morphology is in an intermediate state with significant alongshore three-dimensionality at the onset of a storm, undertow is often replaced by rip circulation where rip currents flow offshore in gaps, or channels between bars; such rip channels typically have spacings between 100–500 m and rip current speeds may exceed 1 ms-1 (Brander & Short, 2000). Rip currents can carry large amounts of suspended sediment seaward (Aagaard et al., 1997; Greenwood et al., 2009; Thorpe et al., 2013) and this sediment is eroded from embayments in the lee of rip channels where wave heights are larger and mobilize more sediment compared to wave heights in the lee of bars (Komar, 1983; Thornton et al., 2007). The alongshore non-uniformity, consisting of alternating shoreline salients and embayments that are coupled to nearshore bar topography, may create strongly alongshore-varying erosion rates at the onset of a storm (Castelle et al., 2015). Erosion rates may be very large on embayed beaches where topographical constraints cause alongshore varying wave heights and water levels that favour the development of strong rip currents. One, or a few megarips may drain an entire embayment, and extreme storms can result in significant erosion and large loss of sediment far offshore through the action of megarips (Loureiro et al., 2012) but, for obvious reasons, field measurements of flow and sediment transport in megarips do not exist. In addition to the sediment transport components associated with incoming wind/swell waves and mean currents, several studies have shown that in particular

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

51

during storms and/or dissipative wave conditions, an important third component is associated with oscillatory flows due to oscillatory infragravity (IG) wave motions (Figure 3.2), see also Chapter 2. In some cases, IG-driven transport may in fact become dominant in the inner surf zone (Beach & Sternberg, 1988, 1991; Russell, 1993). Further offshore in the mid- and outer parts of the surf zone, IG-driven transport rates range from negligible (Conley & Beach, 2003) to large (Aagaard & Greenwood, 1995) and at a given point on the profile, IG-transport may be either onshore or offshore-directed (Houser & Greenwood, 2007; Aagaard et al., 2013). Consequently, the magnitude and direction of the IG-driven sediment transport on the upper shoreface is not well understood and, in fact, the reason why infragravity waves may impose a net cross-shore transport within the surf zone is unclear. Four hypotheses exist for net IG-driven sediment transport at a given location (Figure 3.3): (1) Large IG waves can suspend sand because of the large orbital velocities and associated bed shear stresses (Beach & Sternberg, 1988) and any IG wave skewness will impose a net transport direction (Figure 3.3a). Similarly, in cases when undertow is strong, more sand may be suspended on the offshore IG-phase leading to offshore IG transport. This mechanism may be especially relevant in the inner surf zone where the water depth is shallow and IG velocities may be large and exceed those of the wind and swell waves (Russell, 1993). (2) When the incoming wind/swell waves are unsaturated, the smaller water depths under IG wave troughs will result in increased sediment stirring by wind/swell waves because near-bed velocities increase with decreasing depth; transport at IG frequencies will be offshore directed (Smith & Mocke, 2002) unless the infragravity waves are cross-shore standing, since surface elevation and cross-shore velocity are then shifted by 90 degrees. This mechanism is analogous to the bound long wave case operating seaward of the surf zone (Figure 3.3b) where IG wave troughs are associated with large wave groups (Shi & Larsen, 1984; Ruessink et al., 1998), except that in the bound long wave case it is the increased wave heights rather than decreased water depth at long wave troughs that cause increased sediment stirring and offshore IG transport. (3) When the incoming wind/swell waves are depth-saturated in the surf zone, the waves are larger and may therefore suspend more sediment under IG wave crests (Figure 3.3c) resulting in onshore IG transport (Houser & Greenwood, 2007) unless the infragravity waves are cross-shore standing. (4) If cross-shore gradients exist in the amount of sediment stirred by wind/swell waves, such as, for example, caused by wave breaking over bars and wave reformation in troughs (Figure 3.3d), cross-shore standing infragravity waves may advect sand onshore from positions landward of the stirring maximum and offshore from positions seaward of the stirring maximum (Aagaard & Greenwood, 2008) because of sediment settling and the consequent unequal amounts of sand picked up at onshore and offshore long wave strokes. In all but the first hypothesis listed here, the IG waves are perceived as passive carriers of sediment that has been suspended by the (breaking) wind/swell waves, analogous to the role of mean currents. Referring back to Figure 3.2, which shows typical vertical sediment transport profiles from the inner surf zone during a storm, net cross-shore suspended sediment transport at any given surf zone location consists of the sum of the three transport components discussed above: (1) Net oscillatory transport due to skewed (and/or asymmetric) incoming wind waves and swell. The transport direction is onshore, but may be reduced or even reversed if steep ripples exist on the seabed, which is usually not the case under

52

COASTAL STORMS: PROCESSES AND IMPACTS

u+

(a) t

uq+ t q-

(b) x

(c) x

(d) x

Figure 3.3 Models for net IG-driven sediment transport: (a) shows schematic time series of an offshore-skewed IG velocity signal and the sediment transport rate associated with it, (b), (c) and (d) illustrate IG transport in the spatial domain (shoreline to the right) associated with (b) smaller water depths and/or larger incoming wave heights at IG wave troughs cause increased sediment suspension and offshore net transport, (c) larger incoming wave heights at IG wave crests cause increased suspension and onshore net transport, and (d) when IG waves are cross-shore standing, local sediment suspension maxima, for example at the wave breakpoint, may result in net IG transport away from the suspension maxima. The arrows indicate IG transport rates and directions. In panels (b), (c) and (d) the solid lines represent the incoming wave forms, the dotted lines are the infragravity wave shapes (envelopes) and the horizontal line is mean water level. See text for further explanation.

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

53

surf zone conditions, or if mean flows are sufficiently strong to suspend more sand on the offshore wave stroke. (2) Offshore-directed transport due to undertow or rip currents. (3) Net oscillatory transport due to IG waves that may be directed either way within the surf zone (Figure 3.2), but is often seaward directed near the beach face where eroded sediment is in transport. Under storm conditions, and in the surf zone, the mean current-driven component tends to be dominant in most cases, such that sum of these three components, which is the net transport, is offshore directed (Figure 3.2). When storms abate, wave heights decrease, the surf zone contracts and undertow becomes weaker. This typically results in onshore sediment transport and beach recovery; the beach recovery phase is significantly longer in general than the erosional phase because wave energy levels and sediment transport rates are smaller (Quartel et al., 2007). While individual storms tend to erode and flatten the beach on a time scale of hours, beach recovery is characterized by onshore migration and merging of intertidal bars near the high tide line, which occurs on a time scale of days to weeks. Onshore transport during beach recovery has been identified as resulting in general from two mechanisms: (1) When undertow weakens, net oscillatory transport due to wind waves and swell assume dominance such that sand is driven onshore. (2) When mean water levels decrease after a storm, the water depth across bars may be so shallow that undertow is inhibited. The surf zone is ‘choked’ with sand that has been eroded from the beach and the circulation of water becomes three-dimensional. Rip currents develop in gaps or low points between bars, and onshore mean flows evolve over the bars (Aagaard et al., 1998; MacMahan et al., 2005); onshore transport of sand carried by the mean flow drives the bar onshore until the rip channels are infilled. This net onshore movement of sediments is often further enhanced by the shifting water levels over tidal cycles: at lower stages of the tide and depending on the vertical position of the intertidal bar(s) on the profile, a bar may be exposed during part of the wave cycle and water is carried almost uni-directionally onshore by shallow surf bores (Aagaard et al., 2006). Swash processes finally emplace sediment back onto the eroded beach (see also Masselink et al., 2006).

3.4

Observations of sediment transport on the upper shoreface during storm events

Despite the many studies that describe morphological change of beach profiles and shoreline configurations during storm events, measurements of sediment transport in the surf zone during extreme events associated with coastal erosion are rare. First, there is an element of serendipity associated with having instruments deployed during such rare events and second, large difficulties are associated with maintaining sensors under adverse conditions. However, based on available evidence, patterns of cross-shore sediment transport on the shoreface are not vastly different from transport during more moderate energy conditions since the incoming waves are often saturated in the surf zone. Infragravity waves may be more energetic and hence potentially transport larger amounts of sand during extreme events, but even that is not necessarily the case, since energy at the higher IG frequencies may also be dissipated through wave breaking (De Bakker et al., 2014). However, during extreme events, the beach and the intertidal zone are typically inundated due to storm surge, and in cases when the beach is in an

54

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 3.4 Orthorectified image of Skallingen. Graadyb inlet is seen at the bottom of the image whilst the cuspate foreland of Blaavands Huk is at the upper left corner.

accretive state and the local slope is steeper than the upper shoreface slope, radiation stress and energy dissipation gradients may intensify, which causes large turbulence levels and strong undertows. Hence, offshore-directed suspended sediment transport rates can become large across the inundated beach/intertidal zone, which is then denuded. Cross-shore and longshore sediment transport has been the object of field experiments at Skallingen, which is an approximately 11 km long barrier spit facing the North Sea (Figure 3.4) and it is the northernmost barrier in the chain of barrier islands and spits stretching from Blaavands Huk in Denmark to Texel in the Netherlands. Whilst the spit accreted and rapid foredune aggradation occurred in the first half of the 20th century, the barrier has suffered chronic erosion since the mid-1970s and it receded at an average rate of 4.2 +/− 1.1 m/yr over the period 1981–2012. The reason for the onset of chronic erosion was increased net annual longshore sediment transport due to changes in wind/wave climate, combined with dredging activities in the Graadyb Inlet, which over the years have degraded the ebb delta shield and caused an anticlockwise rotation of the end of the spit (Aagaard & Sørensen, 2013). Since there is no sediment supply from updrift sources around the cuspate foreland of Blaavands Huk, the downdrift loss of sediment, which has increased to about 0.6 x 106 m3 a-1 , is all supplied by erosion of the barrier body at Skallingen. Figure 3.5 shows a series of cross-shore profiles at P6420 from 1981 to 2012. In the 30-year period, the dunes receded by about 150 m and they have now almost disappeared, although a slight recovery phase set in over the period 2008–2012. Whilst the spit suffers chronic erosion due to longshore drift divergence, the actual dune recession is caused by punctuated erosion that occurs during infrequent intense storms associated with significant storm surges. Figure 3.6 shows a ten-year time series (1999–2008) of mean water levels at Esbjerg Harbour. Storm surges, locally defined

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

55

14 1981

12 10

1988

elevation, m OD

8 6 4

2012 2008

1990

2 0 –2 –4 –150

–100

–50

0

50 100 150 200 distance from baseline, m

250

300

350

Figure 3.5 Cross-shore profiles at P6420 from the period 1981–2012. The dunes eroded consistently in the period 1981–2008 but have recovered somewhat since then.

as times when mean water levels exceed +3 m O.D. (Ordnance Datum) occur once or twice per year; in some years (e.g. 1999) water levels may be up to +4 m O.D., while in other years, no storm surges occur. Two large clustered surges in January 1990 (reaching +4.13 m O.D.) caused up to 44 m recession of the dunes. Field measurements of waves, currents and sediment transport were collected during a surge in October 2000, the timing of which is marked on the plot in Figure 3.6. The storm event lasted about 36 hours; wind speeds measured at Blaavand Lighthouse exceeded 25 ms-1 and offshore wave heights were up to Hs = 4.1 m with peak spectral periods of T = 9.5–12.8 s. Surge levels during this storm (+3.03 m DOD at Esbjerg Harbour) were exceeded only 0.12% of the time during the ten-year record shown in Figure 3.6. In comparison to mean water levels within the Wadden Sea at Esbjerg, the levels reached on the open coast were somewhat smaller, reaching only +2.60 m O.D. but they were still sufficiently high for waves to attack and erode the front of the dunes; slumping of the dune wall was observed near high tide and the dune toe had retreated by 5 m after the storm (Figure 3.7). The high intertidal bar that was located approximately 50 m from the baseline prior to the storm was completely eroded and the sand was deposited about 75 m further offshore to form a new intertidal bar at a lower level. The low intertidal bar existing prior to the storm (at x = 150 m) was also eroded, while the subtidal bar at x = 250 m from the baseline migrated onshore, which is a typical occurrence during storm surges at Skallingen (Aagaard et al., 2002, 2004). Overall, the profile within the surveyed section lost 22 m3 of sediment per metre of shoreline, most likely mainly downdrift (towards the south) and driven by the longshore current; the maximum recorded longshore current speed was V = 0.73 ms-1 .

56

COASTAL STORMS: PROCESSES AND IMPACTS

5 4

elevation, m OD

3 2 1 0

–1 –2 –3

0

500

1000

1500 2000 2500 days from Jan. 1, 1999

3000

3500

Figure 3.6 Mean water levels recorded at Esbjerg Harbour 1 January 1999–31 December 2008. The circle indicates the storm event in October 2000 during which sediment transport measurements were made.

8

elevation, m OD

6 4 2 0 –2 –4 –100

0

100 200 distance from baseline, m

300

Figure 3.7 Cross-shore profile change during the storm surge of 30–31 October 2000, at Skallingen. The grey dashed line is the profile prior to the storm, the thick solid line is the survey immediately after the storm on 1 November and the thin solid line is the profile on 6 November. Circles indicate positions where waves and currents were recorded and crosses signify positions of sediment transport measurements. The horizontal line indicates the maximum water level reached during data collection.

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

57

0 U (m/s)

Hs, h (m)

2

1

–0.2 –0.4

0 68

70

72

74

76

78

68

70

(a)

76

78

76

78

0.5 qx (kg/(m2s))

qx (kg/(m2s))

74 (b)

0.5 0 –0.5 –1 –1.5 68

72

70

72 74 run number (c)

76

78

0 –0.5 –1 68

70

72 74 run number (d)

Figure 3.8 Measurements of (a) mean water level (circles) and significant wave height (crosses) at the offshore subtidal station, (b) mean undertow speeds at the four instrument stations; the subtidal station is shown by circles and the most landward station is indicated by black dots, (c) cross-shore suspended sediment fluxes at the landward sediment transport station due to: incoming wind waves and swell (circles), IG waves (black dots), undertow (crosses) and the resulting net flux (squares), (d) similar data for the seaward sediment transport station. Negative current speeds and sediment fluxes are offshore directed. Instrument runs were spaced one hour apart.

Data was collected in the inner surf zone over a single tidal cycle during the storm peak on 30 October at the four positions indicated in Figure 3.7, and Figure 3.8 shows the wave and cross-shore suspended sediment transport measurements. The four instrument stations were equipped with pressure sensors and electromagnetic current meters, and two of those stations (the first and third from the landward end of the array and both located within the fair-weather intertidal zone) also held optical backscatter sensors that functioned during the event. Wave heights at the offshore subtidal station (which was well inside the wide surf zone during the storm event) peaked at Hs = 1.64 m (Figure 3.8a). Mean cross-shore current (undertow) speeds (Figure 3.8b) were not extreme at any of the stations, U ≈ −0.2 − 0.3 ms-1 except at the seaward subtidal station, where speeds were up to U = −0.42 ms-1 at the beginning of the event but then decreased, possibly as a result of the subtidal bar moving onshore (Figure 3.7). Despite the relatively modest cross-shore current speeds, net cross-shore suspended sediment transport in the intertidal zone was mainly driven by the undertow (Figures 3.8c and d) partly because oscillatory transports due to incoming and IG waves more or less counterbalanced each other, and partly because the IG component was quite variable with respect to both magnitude and direction. Both oscillatory components were, however, secondary in comparison with the transport driven by the undertow. On the other hand,

58

COASTAL STORMS: PROCESSES AND IMPACTS

during the subsequent beach recovery phase, IG waves were instrumental in transporting sand onshore, leading to landward migration of the intertidal bar (Houser & Greenwood, 2007). These measurements are generally consistent with the few other available datasets that have been published from storm and/or high energy events. Undertow typically dominates net cross-shore transport in high-energy surf zones (Conley & Beach, 2003), but IG motions may be instrumental in setting up significant cross-shore transport gradients that lead to bar erosion and migration (Aagaard & Greenwood, 1995). At times IG transport may exceed undertow-driven transport; during one of the pioneering field experiments of suspended sediment transport, Beach & Sternberg (1991) monitored an extremely dissipative beach on the Oregon coast with offshore Hs = 3–5 m and a 500 m wide surf zone. At their measurement location (h = 1.1–1.3 m), undertow speed was fairly moderate (U = −0.26 ms-1 ) and net transport was onshore directed and driven by IG waves. On the other hand, the early work by Russell (1993) documented equally large undertow- and IG-driven offshore-directed transport components during a storm in the intertidal zone of a macrotidal dissipative beach. Partly because of these inconsistent IG transport patterns, cross-shore sediment transport leading to beach erosion during storms is difficult to predict quantitatively.

3.5 Observations of sediment transport on the lower shoreface during storm events Whilst not of immediate importance to beach erosion/accretion during brief storm events, sediment transport on the lower shoreface is nevertheless of interest because it affects sediment supply/loss to and from the upper shoreface and thus exerts a strong impact on the evolution of a coast on long time scales. However, except for early observations by Wright et al. (1991), very few sediment transport measurements on the lower shoreface have been reported in the literature. Recently, a data set was collected seaward of the multi-barred upper shoreface off the barrier island of Fanø (the barrier island immediately south of Skallingen) at a mean water depth of h = 3.9 m. A Pulse-Coherent Acoustic Doppler Velocimeter was used to measure suspended sediment transport in the lower 30 cm of the water column (Aagaard, 2014) and Figure 3.9 shows significant wave height and suspended sediment transport components during the four-week period of deployment, where wave energy conditions ranged from low to high. Local wave heights were up to Hs = 2.5 m (Hs = 4.1 m at h = 16 m) and during peak wave conditions, waves were breaking at the sensor location. Cross-shore suspended sediment transport was separated into components due to oscillatory wave motion, and Eulerian (undertow; up/downwelling) and Lagrangian (Stokes drift) mean currents. Eulerian currents were measured, while Stokes drift was calculated from second-order wave theory. Figure 3.9c shows that over the four-week period, cumulated net transport was onshore directed and mainly caused by oscillatory wave motions (at incoming wave frequencies). The Eulerian and Lagrangian mean transports nearly balanced, but with a small onshore bias. The transport components are further resolved in Figure 3.9b, which shows the four-week time series of sediment

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

59

Hs, m

3 2 1 0

100

200

100

200

q, m3 m–1 s–1

× 10–4

500

600

400

500

600

400

500

600

0.5 0 –0.5 –1

Q (cumulated), m3 m–1

300 400 experimental hour (a)

300 (b)

10 8 6 4 2 0 –2

100

200

300 (c)

Figure 3.9 Cross-shore sediment transport measurements at the upper/lower shoreface transition: (a) shows local significant wave height during the measurement campaign; (b) illustrates cross-shore suspended sediment transport rates due to oscillatory wave motions (dotted grey line), Eulerian currents (thin solid) and Lagrangian currents (thick solid line), and (c) is cumulated wave (dotted grey), current (Eulerian plus Lagrangian; thin solid line) and resulting net (thick solid line) transport.

transport rates due to waves, Eulerian and Lagrangian currents. Eulerian mean transport was predominantly offshore directed, as expected, since it was presumably mainly due to undertow. Occasionally, however, it was onshore-directed due to either upwelling or bottom-stress induced onshore mean currents (Ozkan-Haller, 2014). Transport due to Stokes drift was (by necessity) consistently onshore directed. The resultant mean sediment transport vector was offshore directed during storm events and onshore directed during moderate and low energy conditions (Figure 3.9c). From these observations, we may infer that on a gently sloping beach/shoreface, such as at Fanø, a sediment transport convergence is located at the boundary between the surf and shoaling wave zones, where sand may be stored and from whence it can be brought onshore by shoaling waves as the surf zone contracts during low and moderate energy conditions. In addition, since the cross-shore sediment budget was positive over the experimental period, there was an onshore net supply of sand from the lower to the upper shoreface.

60

COASTAL STORMS: PROCESSES AND IMPACTS

3.6 Conclusions During punctuated storm events, beaches and dunes are eroded and sand is transported offshore. Net cross-shore transport is complex and determined by the balance of transport components due to a range of hydrodynamic processes (comprising incoming wind/swell waves and IG waves plus mean currents) that may interact nonlinearly and are affected by the antecedent beach and shoreface topography. While we have a reasonably good qualitative understanding of cross-shore sediment transport, also during extreme storm events, quantitative prediction is still some way in the future. Part of the reason is the present lack of understanding of the role of surface-injected turbulence in causing increased bed shear stress and sediment suspension. Another reason, which may be particularly relevant during storms, is the inconsistent transport patterns displayed by IG waves (Figures 3.2 and 3.8). IG waves are typically large in the inner surf zone during storms and may sometimes transport large amounts of sand either offshore or onshore, but sometimes they contribute very little to the net transport. At least four models exist for net IG transport on the shoreface (Figure 3.3) and each of these may be of relevance under different conditions and for different sectors of the coastal profile. Notwithstanding the cross-shore transport gradients that exist across the upper shoreface and that are the cause of morphological change and coastal erosion during storms, field measurements also indicate that a cross-shore transport convergence may exist at the boundary between the surf and shoaling wave zones (Figure 3.9) and that, over longer time scales, there may be an onshore directed supply of sand from the lower to the upper shoreface which is mainly realised during low and moderate wave conditions.

Acknowledgements This work was partly funded by the Danish Natural Sciences Research Council (Sediment Supply to Beaches, contract no. 09-070628). Kent Pørksen produced the line drawings.

References Aagaard, T. (2014) Sediment supply to beaches: Cross-shore sand transport on the lower shoreface. Journal of Geophysical Research – Earth Surface, doi: 10.1002/2013JF003041. Aagaard, T. & Greenwood, B. (1995) Suspended sediment transport and morphological response on a dissipative beach. Continental Shelf Research, 15, 1061–1086. Aagaard, T. & Greenwood, B. (2008) Oscillatory infragravity wave contribution to surf zone sediment transport – the role of advection. Marine Geology, 251, 1–14. Aagaard, T. & Hughes, M. (2013) Sediment Transport. In: J.F. Shroder (Ed.) Treatise on Geomorphology, Vol.10, pp.74–105. Academic Press, San Diego. Aagaard, T. & Sørensen, P. (2013) Sea level rise and the sediment budget of an eroding barrier on the Danish North Sea coast. Journal of Coastal Research, SI65, 434–439. Aagaard, T., Black, K.P. & Greenwood, B. (2002) Cross-shore suspended sediment transport in the surf zone: A field-based parameterization. Marine Geology, 185, 283–302.

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

61

Aagaard, T., Davidson-Arnott, R.G.D., Greenwood, B. & Nielsen, J. (2004) Sediment supply from shoreface to dunes: Linking sediment transport measurements and long term morphological evolution. Geomorphology, 60, 205–224. Aagaard, T., Greenwood, B. & Hughes, M.G. (2013) Sediment transport on dissipative, intermediate and reflective beaches. Earth Science Reviews, 124, 32–50. Aagaard, T., Greenwood, B. & Nielsen, J. (1997) Mean currents and sediment transport in a rip channel. Marine Geology, 140, 25–45. Aagaard, T., Hughes, M.G., Sørensen, R.M. & Andersen, S. (2006) Hydrodynamics and sediment fluxes across an onshore migrating intertidal bar. Journal of Coastal Research, 22, 247–259. Aagaard, T., Kroon, A., Andersen, S., Sørensen, R.M., Quartel, S. & Vinther, N. (2005) Intertidal beach change during storm conditions; Egmond, the Netherlands. Marine Geology, 218, 65–80. Aagaard, T., Nielsen, J. & Greenwood, B. (1998) Suspended sediment transport and nearshore bar formation on a shallow intermediate-state beach. Marine Geology, 148, 203–225. Anthony, E.J. (2013) Storms, shoreface morphodynamics, sand supply and the accretion and erosion of coastal dune barriers in the southern North Sea. Geomorphology, 199, 8–21. Austin, M.J., Masselink, G., O’Hare, T.J. & Russell, P.E. (2009) Onshore sediment transport on a sandy beach under varied wave conditions: Flow velocity skewness, wave asymmetry or bed ventilation? Marine Geology, 259, 86–101. Beach, R.A. & Sternberg, R.W. (1988) Suspended sediment transport in the surf zone: Response to cross-shore infragravity motion. Marine Geology, 80, 61–79. Beach, R.A. & Sternberg, R.W. (1991) Infragravity-driven suspended sediment transport in the swash, inner and outer surf zone. Proceedings Coastal Sediments 1991, ASCE, 114–128. Brander, R.W. & Short, A.D. (2000) Morphodynamics of a large-scale rip current system at Muriwai Beach, New Zealand. Marine Geology, 165, 27–39. Bruneau, N., Castelle, B., Bonneton, P., Pedreros, R., Almar, R., Bonneton, N., et al. (2009) Field observations of an evolving rip current on a meso-macrotidal well-developed inner bar and rip morphology. Continental Shelf Research, 29, 1650–1662. Castelle, B., Marieu, V., Bujan, S., Splinter, K.D., Robinet, A., Senechal, N., et al. (2015) Impact of the winter 2013–2014 series of severe Western Europe storms on a double-barred sandy coast: Beach and dune erosion and megacusp embayments. Geomorphology, 238, 135-148. Conley, D.C. & Beach, R.A. (2003) Cross-shore sediment transport partitioning in the nearshore during a storm event. Journal of Geophysical Research, 108, C3, 3065. De Bakker, A.T.M., Tissier, M.F.S. & Ruessink, B.G. (2014) Shoreline dissipation of infragravity waves. Continental Shelf Research, 72, 73–82. Drake, T.G. & Calantoni, J. (2001) Discrete particle model for sheet flow transport in the nearshore. Journal of Geophysical Research, 106, 19859–19868. Freilich, M.H. & Guza, R.T. (1984) Nonlinear effects on shoaling surface gravity waves. Philosophical Transactions Royal Society London, A311, 1–41. Gourlay, M.R. (1968) Beach and Dune Erosion Tests. Delft Hydraulics Lab., Rep. No M935/936. Grasso, F., Castelle, B. & Ruessink, B.G. (2012) Turbulence dissipation under breaking waves and bores in a natural surf zone. Continental Shelf Research, 43, 133–141. Greenwood, B., Brander, R.W., Joseph, E., Hughes, M.G., Baldock, T.E. & Aagaard, T. (2009) Sediment flux in a rip channel on a barred intermediate beach under low wave energy. In: C.A. Brebbia, G. Benassai, & G.R. Rodrigue, (Eds) Coastal Processes 2009, WIT Press, pp. 197–209. Houser, C. & Greenwood, B. (2005) Hydrodynamics and sediment transport within the inner surf zone of a lacustrine multiple-barred nearshore. Marine Geology, 218, 37–63.

62

COASTAL STORMS: PROCESSES AND IMPACTS

Houser, C. & Greenwood, B. (2007) Onshore migration of a swash bar during a storm. Journal of Coastal Research, 23, 1–14. Hsu, T-J. & Hanes, D.M. (2004) Effects of wave shape on sheet flow sediment transport. Journal of Geophysical Research, 109, C05025. Komar, P.D. (1983) The erosion of Siletz Spit, Oregon. In: P.D. Komar (Ed.) CRC Handbook of Coastal Processes and Erosion. CRC Press, pp. 65–76. Komar, P.D. (1998) Beach Processes and Sedimentation. Prentice-Hall, New Jersey, second edition, p. 544. Kroon, A., Quartel, S. & Aagaard, T. (2007) Impact of a major storm on sediment exchanges between the dunes, beach and nearshore. Proceedings Coastal Sediments 2007, 951–962, ASCE, New York. Lentz, S.J. & Fewings, M. (2012) The wind- and wave-driven inner-shelf circulation. Annual Reviews Marine Science, 4, 317–341. Lindhorst, S., Betzler, C. & Hess, H.C. (2008) The sedimentary architecture of a Holocene barrier spit (Sylt, German Bight): Swash bar accretion and storm erosion. Sedimentary Geology, 206, 1–16. Longuet-Higgins, M.S. (1983) Wave set-up, percolation and undertow in the surf zone. Proceedings Royal Society of London, A390, 283–291. Loureiro, C., Ferreira, O. & Cooper, J.A.G. (2012) Extreme erosion on high-energy embayed beaches: Influence of megarips and storm grouping. Geomorphology, 139–140, 155–171. MacMahan, J.H., Thornton, E.B., Stanton, T.P. & Reniers, A.J.H.M. (2005) RIPEX: Observations of a rip current system. Marine Geology, 218, 113–134. Marino-Tapia, I.J., Russell, P.E., O’Hare, T.J., Davidson, M.A. & Huntley, D.A. (2007) Cross-shore sediment transport on natural beaches and its relation to sandbar migration patterns: 1. Field observations and derivation of a transport parameterization. Journal of Geophysical Research, 112, CO3001. Masselink, G., Austin, M.J., O’Hare, T.J. & Russell, P. (2007) Geometry and dynamics of wave ripples in the nearshore zone of a coarse sandy beach. Journal of Geophysical Research, 112, C10022. Masselink, G., Austin, M., Tinker, J., O’Hare, T. & Russell, P. (2008) Cross-shore sediment transport and morphological response on a macrotidal beach with intertidal bar morphology, Truc Vert, France. Marine Geology, 251, 141–155. Masselink, G., Kroon, A. & Davidson-Arnott, R.G.D. (2006) Morphodynamics of intertidal bars in wave-dominated coastal settings – A review. Geomorphology, 73, 33–49. O’Hara Murray, R.B., Thorne, P.D. & Hodgson, D.M. (2011) Intrawave observations of sediment entrainment processes above sand ripples under irregular waves. Journal of Geophysical Research, 116, C01001. Ozkan-Haller, H.T. (2014) Vertical variability of undertow and longshore currents outside the surf zone. Journal of Waterway, Port, Coastal and Ocean Engineering, 140, 4–13. Puleo, J.A., Holland, K.T., Plant, N.G., Slinn, D.N. & Hanes, D.M. (2003) Fluid acceleration effects on suspended sediment transport in the swash zone. Journal of Geophysical Research, 108, C11, 14.1–14.12. Quartel, S., Ruessink, B.G. & Kroon, A. (2007) Weekly to seasonal cross-shore behavior of intertidal beach morphology. Earth Surface Processes and Landforms, 32, 1293–1307. Ruessink, B.G., Houwman, K.T. & Hoekstra, P. (1998) The systematic contribution of transporting mechanisms to the cross-shore sediment transport in water depths of 3 to 9 m. Marine Geology, 152, 295–324.

CH3 SEDIMENT TRANSPORT ON SANDY BEACHES

63

Ruessink, B.G., Pape, L. & Turner, I.L. (2009) Daily to interannual cross-shore sandbar migration: Observations from a multiple sandbar system. Continental Shelf Research, 29, 1663–1677. Ruessink, B.G., Ramaekers, G. & van Rijn, L.C. (2012) On the parameterization of the free-stream non-linear wave orbital motion in nearshore morphodynamic models. Coastal Engineering, 65, 56–63. Russell, P.E. (1993) Mechanisms for beach erosion during storms. Continental Shelf Research, 13, 1243–1265. Scott, N.V., Hsu, T.-J. & Cox, D. (2009) Steep wave, turbulence, and sediment concentration statistics beneath a breaking wave field and their implications for sediment transport. Continental Shelf Research, 29, 2303–2317. Scott, T., Masselink, G. & Russell, P. (2011) Morphodynamic characteristics and classification of beaches in England and Wales. Marine Geology, 286, 1–20. Shi, N.C. & Larsen, L.H. (1984) Reverse sediment transport induced by amplitude-modulated waves. Marine Geology, 54, 181–200. Short, A.D. (1999) Wave-dominated beaches. In: A.D. Short (Ed.) Handbook of Beach and Shoreface Morphodynamics. Wiley Science, pp. 173–203. Smith, G.G. & Mocke, G.P. (2002) Interaction between breaking/broken waves and infragravity-scale phenomena to control sediment suspension transport in the surf zone. Marine Geology, 187, 329–345. Svendsen, I.A. (1984) Mass flux and undertow in a surf zone. Coastal Engineering, 8, 347–365. Thornton, E.B., Humiston, R.T. & Birkemeier, W. (1996) Bar/trough generation on a natural beach. Journal of Geophysical Research, 101, 12097–12110. Thornton, E.B., MacMahan, J.H & Sallenger, A.H. (2007) Rip currents, megacusps and eroding dunes. Marine Geology, 240, 151–168. Thorpe, A., Miles, J., Masselink, G., Russell, P., Scott, T. & Austin, M. (2013) Suspended sediment transport in rip currents on a macrotidal beach. Journal of Coastal Research, SI65, 1880–1885. Wright, L.D. & Short, A.D. (1984) Morphodynamic variability of surf zones and beaches: A synthesis. Marine Geology, 56, 93–118. Wright, L.D., Boon, J.D., Kim, S.C. & List, J.H. (1991) Modes of cross-shore sediment transport on the shoreface of the Middle Atlantic Bight. Marine Geology, 96, 19–51. Xu, Z. & Bowen, A.J. (1984) Wave- and wind-driven flow in water of finite depth. Journal of Physical Oceanography, 24, 1850–1866.

4 Examples of Storm Impacts on Barrier Islands Nathaniel Plant, Kara Doran and Hilary Stockdon US Geological Survey, Saint Petersburg, Florida, USA

4.1

Introduction

Barrier islands are a commonly occurring geomorphic feature of many coastal regions. The many potential formation mechanisms and geologic settings have recently been described by Otvos (2012), and include prograding spits driven by alongshore sediment transport (Figure 4.1), as well as remnants of delta shorelines reworked by waves (Figure 4.2). The evolution of barrier spits can result in complicated behaviours and patterns (Ashton et al., 2001). Likewise, deltaic reworking can be controlled by complicated geologic histories (Penland et al., 1988). The two barrier types can be mixed, adding to the complexity of barrier island evolution as they respond to storms, changes in sea level, and sediment redistributions. Human development, coastal protection, and restoration are other activities that can alter the physical form or even the physical processes that are important to barrier island evolution (Hapke et al., 2013; Plant et al., 2014). As their name implies, barrier islands protect the mainland coast and estuary from the impacts of storms by forming a topographic barrier that can partially or completely block storm waves and storm-driven water levels from propagating past the barrier. Given that there can be enormous variability in the geomorphic form of barrier islands (e.g. Otvos, 2012) and there can be a wide range of storm characteristics, barrier islands can undergo a wide variety of responses to storms (Sallenger, 2000; Stockdon et al., 2007; Long et al., 2014). Here, we focus on the morphologic variability of barrier islands and on the differences in storm response. We describe different types of barrier island response to individual storms, as well as the integrated response of barrier islands to many storms. Our case study site is the Chandeleur Island chain (Figure 4.2), where a decadal time series of island elevation measurements have documented a wide range of barrier island responses to storms and long-term processes that are representative of barrier island behaviour at many other locations. These islands are low elevation, extremely vulnerable to storms (Stockdon et al., 2012) and exhibit a diversity of storm responses. Additionally, this location experiences a moderately high rate of relative sea-level rise, Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

66

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 4.1 Assateague Island (Virginia/Maryland, USA) showing 1 km of shoreline retreat (left) and 5 km of spit elongation (right). The historic shorelines (Himmelstoss et al., 2010) from the 1880s (red), 1940s (yellow), and 2000s (blue) are overlain on the imagery showing the long-term evolution.

increasing its vulnerability to the combined impacts of storms and long-term erosional processes (Gutierrez et al., 2014) (Figure 4.3). Finally, this location was subject to restoration activities that had short-term impacts and, perhaps, longer-term alteration of the island evolution (Lavoie et al., 2010). Understanding how natural processes, including storm impacts and intervening recovery periods interact with man-made restoration processes is also broadly relevant to understand the natural and human response to future storms.

4.2 Barrier island response to storms Over the long-term, the Chandeleur Islands have changed position and shape due to erosion along the seaward shoreline (Fearnley et al., 2009), overwash (Sallenger et al., 2009; overwash is also described in more detail in Chapter 9), breaching (Sherwood et al., 2014), and, more recently, restoration efforts (Plant et al., 2014). All of these processes occur in the context of the sources, transport, and deposition of a variety

CH4 STORM IMPACTS ON BARRIER ISLANDS

67

Figure 4.2 Barrier island migration illustrated at the Northern Chandeleur Islands (Louisiana, USA). The background images shows the island in 2015 and are overlain with shorelines (Miller et al., 2004) from the 1930s (left) and in 2001 (right).

68

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 4.3 Evolution of the Northern Chandeleur Islands. Multiple hurricanes and other storms caused significant land loss over a decade-long time span. A sand berm was constructed in 2010 and appears as a thin ribbon of sand at the northern tip of the island. The water level at the time of Landsat imagery acquisition and date of each image are labelled at the bottom and the most recent morphologic event is labelled at the top.

of sediment types that owe their initial characteristics to the geologic-scale processes (Twichell et al., 2009). These processes are common to most barrier islands (Figure 4.1). Elevation change is of particular interest because the island elevation, typically characterized by dune height, relative to storm-driven water levels, controls the regimes of island response (Sallenger, 2000). Storm-driven water levels include the contributions of storm surge, wave setup, and wave runup. The type and magnitude of barrier island response is divided into four regimes (Figure 4.4): beach erosion occurs when wave runup does not reach the elevation of the dune toe (the swash regime); dune erosion results when elevated water levels allow wave runup to reach the toe of

CH4 STORM IMPACTS ON BARRIER ISLANDS

69

Inundation Dune-crest elevation

Overwash

Dune-toe elevation

Collision

Swash Mean water level

Figure 4.4 Depiction of the Sallenger (2000) storm-response model showing storm-induced water levels corresponding to swash, collision, overwash and inundation regimes. Dashed lines show the storm-induced mean water level and solid lines indicate the wave runup level under each regime.

the dune (collision regime); erosion and formation of overwash deposits occurs when wave runup exceeds the dune-crest elevation (the overwash regime; see Chapter 9 for detailed description of overwash processes); and breaching and complete levelling of dunes occurs when storm-drive mean water levels (i.e. surge and wave setup) exceed the dune-crest elevation (inundation regime) (Sallenger, 2000). The Sallenger (2000) storm response regimes describe what sort of morphologic changes are expected under each regime and evaluation of these storm-response regimes has been shown to predict actual changes to beaches and dunes during storms (Stockdon et al., 2007; Plant & Stockdon, 2012). For instance, if overwash is minor, there may be simple landward translation of a barrier berm or dune (e.g. Figure 4.5) with minimal changes in the berm or dune elevation. If overwash is extreme, there can be dramatic changes to the island, ultimately including breach formation (Figures 4.2 and 4.3) where inundation occurs (Long et al., 2014). Along-island variation in the geomorphology, including dune-height variability, also causes variability in barrier island response. Over the course of a storm, the nature of barrier island response can evolve as storm surge and waves first erode the beach, then cause overwash or even inundation (Long et al., 2014; Sherwood et al., 2014). The rate at which the barrier island evolves also depends on other factors, such as beach or dune width, vegetation, and sediment characteristics. These other factors can control how quickly the dune

70

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 4.5 Elevation-change differences along the Chandeleur Islands after Hurricane Lili.

crest is lowered, thereby triggering a change from dune erosion to overwash (Plant & Stockdon, 2012; Long et al., 2014).

4.3 Quantifying the changes due to specific storms Here, we focus on the impact of a number of hurricanes using two metrics of storm response. The change to the shoreline position describes the response due to beach erosion or barrier-island migration associated with overwash. The change to the barrier elevation characterizes the response to dune erosion, overwash and inundation, and describes the change in vulnerability to future storms. Because the Chandeleur Islands are exceedingly low, dune or berm features typically used to represent barrier island vulnerability (Stockdon et al., 2009a) could not be identified and the maximum barrier elevation was used. Both the shoreline and elevation measures (among many others) were extracted from airborne lidar surveys of coastal topography, which is the most accurate and detailed approach to quantify coastal changes due to storms (Stockdon et al., 2007, 2009b). The specific impacts from Hurricane Lili (in 2002), Katrina (2005) and Gustav (2008) were documented by the US Geological Survey (Doran et al., 2009; Sallenger et al., 2009). We use these storm events to illustrate how barrier elevation and shoreline position are affected by different storm regimes. The response to Hurricane Lili can be characterized as primarily overwash because estimated storm-induced water levels exceeded the island elevation at many locations (Figure 4.6). For this storm, surge elevation was estimated from the maximum water level measured at a nearby tide gauge, and R2% (the two percent exceedance level due to setup and wave runup) was estimated using a parameterized model (Stockdon et al., 2006). Model inputs included measured beach slopes from topographic surveys and wave height and period measured at nearby buoys. In this case, the dominant overwash process caused an average shoreline retreat of 50 m and an average dune elevation change of a few centimetres. The storm-induced mean water level (labelled surge + setup in Figure 4.6) exceeded the island only at the lowest elevations, resulting in some inundation. The variability of the elevation changes ranged from 1.5 meters of elevation loss to somewhat less than a meter of elevation gain.

CH4 STORM IMPACTS ON BARRIER ISLANDS

71

Pre-Lili survey September, 2001 5 4.5 4

Elevation (m NAVD88)

3.5 3 2.5

Surge + R2%

2 Surge + Setup 1.5 Surge

1 0.5 0 3290

3295

3300

3305 3310 3315 UTM Northing (km)

3320

3325

3330

0.4 0.3

0.6

μ = –0.03 m Probability

Probability

0.35 0.25 0.2 0.15

0.4 0.3

0.1

0.2

0.05

0.1

0

–1.5

0 0.5 1 –1 –0.5 Lili maximum elevation change (m)

μ = –50 m

0.5

0

–300

–200 –100 0 100 Lili shoreline change (m)

200

Figure 4.6 Comparison of (top) pre-storm island elevation to the storm surge, wave setup, and R2% (which includes wave runup) estimated for Hurricane Lili. Resulting island elevation changes (bottom left) and shoreline changes (bottom right) measured with lidar. The mean values, 𝜇, are labelled with a black vertical line on each histogram. Zero change is denoted by the red vertical line.

In 2005, Hurricane Katrina produced conditions that inundated much of the Chandeleur Islands (Figure 4.7). Storm surge levels exceeded 3 m (Interagency Performance Evaluation Task Force (Ipet), 2007; Lindemer et al., 2010). Water levels due to contribution of surge and R2% (computed as described in the Hurricane Lili example) exceeded the island elevations everywhere. The resulting island response differed in magnitude from that during Hurricane Lili. Elevations decreased everywhere, by 1.4 m on average, and decreases were greater than 2.5 m at some locations. Shoreline loss was 250 m on average, exceeding 500 m in some locations, and there were some locations where there was no remnant of the island (Figure 4.3).

72

COASTAL STORMS: PROCESSES AND IMPACTS

7 Pre-Katrina max elevation Post-Katrina max elevation

Elevation (m, NAVD88, Geoid96)

6

Surge + R2%

5

4

3

Surge + Setup Surge

2

1

0.35 0.3 0.25 0.2 0.15 0.1 0.05 0

3295

3300

3305 3310 3315 UTM Northing (km)

μ = –1.35 m

0.6 Probability

Probability

0 3290

3320

3325

3330

μ = –250 m

0.5 0.4 0.3 0.2 0.1

–2.5

–2

–1.5

–1

–0.5

0

Katrina maximum elevation change (m)

0.5

0

–500 –400 –300 –200 –100 0 Katrina shoreline change (m)

100

Figure 4.7 Comparison of (top) pre- and post-storm island elevation to the storm surge, wave setup, and R2% estimated for Hurricane Katrina (asterisks mark alongshore locations of profiles shown in Figure 4.8). Histograms show variability of island elevation change (bottom left) and shoreline change (bottom right).

The differences in the barrier island response to Lili and Katrina are striking when topographic profiles are compared (Figure 4.8). In nearly all cases, Hurricane Lili resulted in erosion of the region between the shore and the dune crest, with deposition behind the pre-storm dune crest, which adds elevation to the barrier. If this additional elevation contributes to growth of new dunes then the overwash offers resilience to future storms even as the barrier island retreats. Hurricane Katrina, on the other hand, levelled the dunes with obvious net loss in sediment and elevation, leaving the island even more vulnerable to subsequent storms. After Hurricane Katrina, the island elevations increased somewhat due to natural berm and dune building processes that included swash during minor storms and aeolian transport, as well as development of dune vegetation. The average maximum elevation change was 0.26 m, though in some places elevation change exceeded 1 m. In 2008, the

CH4 STORM IMPACTS ON BARRIER ISLANDS

73

Figure 4.8 Profiles across the width of the Chandeleur Islands at three locations, showing classic overwash evolution associated with Hurricane Lili and island levelling due to the inundation associated with Hurricane Katrina. Profile locations are marked with an asterisk on Figure 4.7, with profile A at the northern-most location and profile C at the southern-most location.

74

COASTAL STORMS: PROCESSES AND IMPACTS

3 June 2008 Max elevation

Elevation (m, NAVD)

2.5

2

Approximate surge

1.5

1

0.5

0 3295

3300

3305

3310 UTM Northing (km)

3315

3320

3325

0.4 Probability

μ = –0.24 m

μ = –81 m

0.2

0

0 0.5 –2.5 –2 –1.5 –1 –0.5 Gustav maximum elevation change (m)

1 –500 –400 –300 –200 –100 0 100 Gustav shoreline change (m)

Figure 4.9 Comparison of pre-storm island elevation to the storm surge estimated for Hurricane Gustav, 2008 (top). The storm surge level was uncertain, so a range is indicated with two dashed lines. Resulting island elevation changes (bottom left) and shoreline changes (bottom right). Black lines are mean values; red lines are zero-change.

island was hit again by a storm, Hurricane Gustav (Figure 4.9). Surge levels were about 2 m, between those of Lili (about 1.5 m) and Katrina (3 m). The resulting elevation response was a mix of the two responses observed after previous storms. Similar to the response during Lili, island elevation changes ranged from 1.5 m of loss to 0.75 m of accretion, consistent with overwash processes. However, as during Katrina, the shoreline experienced extreme erosion, over 200 m in some places, and resulted in additional island loss (Figure 4.3). Comparing the shoreline-change histograms for all three storms (Lili, Katrina and Gustav) shows the difference between storms is most visible in the tails in the distribution, indicating the extreme retreat rates (Figure 4.10).

CH4 STORM IMPACTS ON BARRIER ISLANDS

75

Probability

Lili

μ = –50 m

0.6 0.4 0.2

Probability

0 –600

–500

–400

–300

–200 Katrina

–100

0

100

200

–200

–100

0

100

200

0

100

200

μ = –250 m

0.6 0.4 0.2 0 –600

–500

–400

–300

Probability

Gustav

μ = –81 m

0.6 0.4 0.2 0 –600

–500

–400

–300

–200 –100 Shoreline change (m)

Figure 4.10 Comparison of shoreline change histograms from three storms.

4.4

Resilience

The storm impacts in the examples that we have presented paint a picture of nearly continuous reduction in elevations and landward migration, and land loss from most storms. In fact, this level of vulnerability is predicted throughout most of the Gulf of Mexico because of low island elevations (Stockdon et al., 2012). However, as seen in the recovery after Hurricane Gustav, the longer-term evolution of barriers shows that they are capable of slow recovery between storms. This recovery may be due to the landward translation that is associated with overwash and deposition of material inland in shallower water that allows the island to keep up with sea-level rise (i.e. barrier island rollover). Hence, in the long term, barrier islands that are allowed to migrate can be resilient to major storms. Returning to the large-scale view of the barrier island (Figure 4.3, Figure 4.9), it is clear that the impacts of Katrina and Gustav reduced the island elevation, making it remain vulnerable to storms. However, by 2010, the island appeared to have recovered substantially. The island healed itself through continued overwash and shoreline retreat, building up elevation and filling the many breaches that dissected the island during Katrina, Gustav and other storms. The recovery pre-dates the placement of a man-made sand berm associated with the Deepwater Horizon oil spill. (The spill began in April 2010 and was not sealed for several months.) The man-made berm was destroyed relatively quickly (Figure 4.11) due to hurricanes and winter storms (Plant et al., 2014; Plant & Guy, 2014). However, it was estimated that this berm added sediment equal to at least four years’ worth of alongshore sediment transport. Because the island was already recovering naturally, it is quite likely that this additional sediment changed the

76

COASTAL STORMS: PROCESSES AND IMPACTS

323000

324000

323000

324000

3315000

3316000 NAD83/NAVD88 GEOID09 (M) High: 1.5 Less: –1

3313000

3314000

3315000 3314000 3313000

324000

3316000

323000

Figure 4.11 Elevation time series collected prior to man-made berm construction (March 2010), during construction (February 2011), and after winter storms and the tropical storm season (February 2012).

overall resilience of the island, perhaps leading to reduction of adverse elevation losses that could be caused by future storms.

4.5 Summary It is possible to predict the general nature, and even the details of barrier island response to storms given estimates of island elevation and storm-driven water levels (Roelvink et al., 2009; Stockdon et al., 2009b; Lindemer et al., 2010; Mccall et al., 2010; Plant & Stockdon, 2012; Stockdon et al., 2012; Long et al., 2014; Sherwood et al., 2014). Updated measurements of barrier island elevations can be supplied to high-resolution topography for numerical simulations (e.g. Mccall et al., 2010) utilizing ocean wave and surge models. The nature of barrier response to different storms leads to variability in the long-term vulnerability to future storms. For instance, Hurricane Lili resulted in a relatively narrow range of dune-height and shoreline position changes that had only a small effect on the average vulnerability to future storms. However, Hurricane Katrina caused severe reductions in elevation, which was associated with extreme shoreline retreat. This increased level of vulnerability allowed future storms, for example Gustav,

CH4 STORM IMPACTS ON BARRIER ISLANDS

77

to continue to cause severe impacts, particularly with respect to shoreline retreat and barrier loss. But, it is also clear that recovery of island elevation and shoreline position, which is not predicted by the current models, can be significant. The relative contributions of natural and man-made processes to barrier island resilience are not well understood and are topics of ongoing research that require both improved observations and models (Donnelly et al., 2006). Recent storm events, such as Hurricane Sandy that impacted the US Atlantic coast in 2012 (Sopkin, 2014) indicate an ongoing need to understand barrier island response to storms.

Acknowledgements Research on the impacts of storms on the Chandeleur Islands was initiated by Abby Sallenger, and this chapter has utilized some of his analyses. We thank Kristy Guy for drafting figures and for reviews by Joe Long and Jenna Brown.

References Ashton, A., Muray, A.B. & Arnault, O. (2001) Formation of coastline features by large-scale instabilities induced by high-angle waves. Nature, 414, 296–300. Donnelly, C., Kraus, N. & Larson, M. (2006) State of knowledge on measurement and modeling of coastal overwash. J. Coast. Res., 22 (4 ), 965–991 Doran, K.S., Stockdon, H.F., Plant, N.G., Sallenger, A.H., Guy, K.K. & Serafin, K.A. (2009) Hurricane gustav: Observations and analysis of coastal change, US Geological Survey Open-File Report OFR-2009-1279, p. 28. Fearnley, S., Miner, M., Kulp, M., Bohling, C., Martinez, L. & Penland, S. (2009) Hurricane impact and recovery shoreline change analysis and historical island configuration – 1700s to 2005. In: D. Lavoie (Ed.) Sand resources, regional geology, and coastal processes of the chandeleur islands coastal system – An evaluation of the Breton national wildlife refuge, US Geological Survey Scientific Investigations Report 2009, 5252. Gutierrez, B.T., Plant, N.G., Pendleton, E.A. & Thieler, E.R. (2014) Using a Bayesian network to predict shore-line change vulnerability to sea-level rise for the coasts of the united states, US Geological Survey Open-File Report 2014,1083, 26. Hapke, C.J., Kratzmann, M.G. & Himmelstoss, E.A. (2013) Geomorphic and human influence on large-scale coastal change. Geomorphology, 199, 160–170. Himmelstoss, E.A., Kratzmann, M., Hapke, C., Thieler, E.R. & List, J.H. (2010) The national assessment of shoreline change: A GIS compilation of vector shorelines and associated shoreline change data for the New England and Mid-Atlantic coasts. US Geological Survey Open-File Report 2010,1119. Interagency Performance Evaluation Task Force (IPET) (2007) Performance evaluation of the New Orleans and southeast Louisiana hurricane protection system. In: Final report of the interagency performance evaluation task force, edited, US Army Corps of Engineers, Washington, DC. Lavoie, D., Flocks, J.G., Kindinger, J.L., Sallenger, Jr., A.H. & Twichell, D.C. (2010) Effects of building a sand barrier berm to mitigate the effects of the deepwater horizon oil spill on Louisiana marshes. US Geological Survey, Open-File Report 2010, 1108, 7.

78

COASTAL STORMS: PROCESSES AND IMPACTS

Lindemer, C.A., Plant, N.G., Puleo, J.A., Thompson, D.M. & Wamsley, T.V. (2010) Numerical simulation of a low-lying barrier island’s morphological response to Hurricane Katrina. Coastal Engineering, 57 (11–12), 985–995. Long, J.W., d. Bakker, A.T.M. & Plant, N.G. (2014) Scaling coastal dune elevation changes across storm-impact regimes, Geophys. Res. Lett., 41 (doi:10.1002/2014GL059616), 8. McCall, R.T., Van Thiel de Vries, J.S.M., Plant, N.G., Van Dongeren, A.R. Roelvink, J.A. Thompson, D.M., et al. (2010) Two-dimensional time dependent hurricane overwash and erosion modeling at Santa Rosa Island. Coastal Engineering, 57 (7), 668–683. Miller, T.L., Morton, R.A., Sallenger, A.H. & Moore, L.J. (2004) The national assessment of shoreline change: A GIS compilation of vector shorelines and associated shoreline change data for the US Gulf of Mexico. USGS Open File Report 2004, 1089. Otvos, E.G. (2012) Coastal barriers – nomenclature, processes, and classification issues. Geomorphology, 139–140, 39–52. Penland, S., Boyd, R. & Suter, J. (1988) Transgressive depositional systems of the Mississippi delta plan: Model for barrier shoreline and shelf sand development, Journal of Sedimentary Petrology, 58, 932–949. Plant, N.G. & Guy, K.K. (2014) Change in the length of the southern section of the Chandeleur Islands oil berm, January 13, 2011, through September 3, 2012. US Geological Survey Open-File Report 2013, 1303, 8. Plant, N.G. & Stockdon, H.F. (2012) Probabilistic prediction of barrier-island response to hurricanes. JGR Earth Surface, 117 (F03015), 17. Plant, N.G., Flocks, J., Stockdon, H.F., Long, J.W., Guy, K., Thompson, D.M., et al. (2014) Predictions of barrier island berm evolution in a time-varying storm climatology. Journal of Geophysical Research Earth Surface, 119, 300–316. Roelvink, J.A., Reniers, A., Dogeren, A.V., d. Vries, J.V.T., McCall, R. & Lescinski, J. (2009) Modeling storm impacts on beaches, dunes and barrier islands, Coastal Engineering, 56, 1133–1152. Sallenger, A.H., Jr. (2000) Storm impact scale for barrier islands. J. Coast. Res., 16 (3), 890–895. Sallenger, A.H., Jr., Wright, C.W., Howd, P., Doran, K. & Guy, K. (2009) Extreme coastal changes on the Chandeleur Islands, Louisiana, during and after Hurricane Katrina. In: D. Lavoie (Ed.) Sand resources, regional geology, and coastal processes of the Chandeleur Islands coastal system – An evaluation of the Breton national wildlife refuge, pp. 27–36, US Geological Survey Scientific Investigations Report 2009, 5252. Sherwood, C.R., Long, J.W., Dickhudt, P.J., Dalyander, P.S., Thompson, D.M. & Plant, N.G. (2014) Inundation of a barrier island (Chandeleur Islands, Louisiana, USA) during a hurricane: Observed water-level gradients and modeled seaward sand transport. J. Geophys. Res. Earth Surf ., 119 (doi:10.1002/2013JF003069), 18. Sopkin, K.L., Stockdon, H.F., Doran, K.S., Plant, N.G., Morgan, K.L.M., Guy, K.K., et al. (2014) Hurricane Sandy – observations and analysis of coastal change. US Geological Survey Open-File Report, 2014,1088, 54. Stockdon, H.F., Holman, R.A., Howd P. & Sallenger A.H., Jr. (2006) Empirical parameterization of setup, swash, and runup. Coastal Engineering, 53, 573–588. Stockdon, H.F., Sallenger, A.H., Jr., Holman, R.A. & Howd, P. (2007) A simple model for the spatially-variable coastal response to hurricanes. Mar. Geol., 238, 1–20. Stockdon, H.F., Doran, K.S. & Sallenger, A.H., Jr. (2009a) Extraction of lidar-based dune-crest elevations for use in examining the vulnerability of beaches to inundation during hurricanes. J. Coast. Res., Special Issue (53), 59–65.

CH4 STORM IMPACTS ON BARRIER ISLANDS

79

Stockdon, H.F., Plant, N.G. & Sallenger, A.H., Jr. (2009b). National assessment of hurricane-induced coastal change vulnerability. Shore & Beach, 77 (93). Stockdon, H.F., Doran, K.J., Thompson, D.M., Sopkin, K.L., Plant, N.G. & Sallenger, A.H., Jr. (2012) National assessment of hurricane-induced coastal erosion hazards: Gulf of Mexico. US Geological Survey Open-File Report 2012, 1084, 51. Twichell, D., Pendleton, E., Baldwin, W. & Flocks, J. (2009) Subsurface control on seafloor erosional processes offshore of the Chandeleur Islands, Louisiana. Geo-Marine Letters, 29 (6), 349–358.

5 Storm Impacts on the Morphology and Sedimentology of Open-coast Tidal Flats Ping Wang and Jun Cheng School of Geosciences, University of South Florida, Tampa, USA

5.1

Introduction

A tidal flat is generally defined as an extensive, nearly horizontal, marshy or barren tract of land that is alternately covered and uncovered by water-level fluctuations associated with tides, and consisting of unconsolidated sediment, mostly mud and sand (Bates & Jackson, 1980). Tidal flat is also often referred to as an intertidal zone, although in some general discussions, it may include subtidal and supratidal zones. In this chapter, the term tidal flat strictly refers to the intertidal zone, lying between spring low tide and spring high tide levels, as defined in the Glossary of Geology (Bates & Jackson, 1980). The zonation of a tidal flat is generally divided based on the duration of submergence, which links directly to the differences in sedimentary characteristics. Based on distinctive characteristics of sediment, sedimentary structures and general trend of lamina thickness, a tidal flat is often divided into upper, middle and lower intertidal zones (Klein, 1976; Reineck & Singh, 1980). The transitions between the zones are gradual. Extensive coastal marsh typically distributes in the upper intertidal zone and further landward. Coastal wetlands function to stabilize shorelines and protect coastal communities during both normal and storm conditions (Gedan et al., 2011). However, the wetlands’ role in reducing shoreline erosion during extreme storm conditions has been challenged by Feagin (2008) and Feagin et al. (2009). This chapter mainly focuses on the sedimentology and morphology of the mostly barren intertidal zone. The role of vegetation in the morphodynamics of the upper intertidal zone and further landward are beyond the scope of this chapter. Generally, sediment transport and deposition on tidal flats are considered to be controlled by the regular fluctuations of tidal currents, whereas the comparatively irregular Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

82

COASTAL STORMS: PROCESSES AND IMPACTS

110E

120E 40 N

Korea Peninsula

Tianjin o R

Bohai Sea

he

ng

a Hu

Changjiang

Qingdao Yellow Sea

R

Shanghai 30 N

Guangzhou

Taiwan

East China Sea

20 N

South China Sea

0

200

400

600 km

open coast tidal flat

Figure 5.1 Distribution of open-coast tidal flats along the Chinese coast (modified from Li et al., 2004).

wave forcing, including storm impacts, is often regarded to be minimal and neglected. Eisma (1998) estimated that over 70% of tidal flats occur in wave-sheltered areas, such as bays, estuaries, lagoons and behind spits or barriers, while the remainder occurs along open coasts, the majority of which are characterized by low wave conditions. Fan (2012) argued that the Eisma (1998) estimate may have been influenced by the fact that publications on open-coast tidal flat were rather limited until after the year 2000. Fan (2012) suggests that due to their close association with large river deltas, for example Changjiang (Chen, 1998) and Huanghe deltas in China (Figure 5.1), open-coast tidal flats are more widely distributed than semi-enclosed ones. In comparison with extensively studied wave-sheltered tidal flats, open-coast tidal flats are characterized by: (1) facing an open ocean or sea with no morphologic barriers to block incident ocean waves; (2) flooding and ebbing tidal currents that are not extensively confined and regulated by tidal channels; and 3) receive large sediment input, particularly mud-sized sediments, from nearby rivers. Along open-coast tidal

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

83

flats, major tidal channels are generally absent and, therefore, do not significantly regulate flood and ebb currents. Therefore, sedimentary features, such as mega-ripples and dunes, associated with tidal channels are rare to non-existent. Waves propagate over the tidal flat directly from the open sea, although their energy is significantly dissipated over the wide and gentle muddy flats during low-wave conditions. Wave conditions are typically modulated by tidal water-level fluctuations. Intertidal and sub-tidal fluid mud may also induce significant damping to incident wave. However, during storms, due to the lack of wave barrier, open-coast tidal flats are vulnerable to the impact of high storm waves, both swells propagated from the open ocean and seas generated by local wind. The strong wave forcing may induce substantial reworking and re-deposition of tidal deposits. Sediment dynamics and morphodynamics along open-coast tidal flats carry strong regional characteristics controlled by both input sediment properties and regional oceanographic conditions. In this chapter, we discuss the morphodynamics and sedimentologic characteristics of open-coast tidal flats using those along the Changjiang River delta in China as examples (Figure 5.1). The open-coast tidal flats fringing the Changjiang River delta are significantly influenced by the tremendous mud-sized sediment supplies from the large river. This results in a regional accretionary trend of the extensive tidal flats. Mean sediment grain size along the Changjiang River delta open-coast tidal flat typically ranges from 4 to 8 phi (0.063–0.004 mm) (Li et al., 2000). The width of the tidal flats ranges from 3 to 4 km, with a maximum of nearly 10 km. The average slope of the tidal flat is typically 1:1000, with a maximum of 1:200 and a minimum of 1:5000 (Fan, 2012).

5.2

Sedimentologic characteristics

Sedimentologic characteristics vary substantially from one region to another, controlled by the regional sediment supply. Therefore, the examples here from the Changjiang River delta may not be directly applicable to other locations. However, the trends of temporal and spatial variations as controlled by tide and wave forcing, as well as the relative dominance of tide or wave forcing, are applicable in understanding morphodynamics of open-coast tidal flat in general. The upper intertidal zone, generally located between mean neap high tide and mean spring high tide, is dominated by clay-sized sediment with vegetation coverage, that is, coastal marsh. The sediment grain size increases seaward to mostly coarse silt in the barren lower intertidal zone between mean neap low tide and mean spring low tide. The surface of the barren middle to lower tidal flat is typically covered by wave ripples. Tidal flats are characteristic of a unique textbook set of sedimentary structures. The sequence of sedimentary structure variation from upper to lower intertidal zones as described by Reineck and Singh (1980) based on semi-enclosed estuarine tidal flats was also observed on open-coast tidal flats. The upper intertidal zone is characterized by finer sediment with thicker muddy laminae. Sandy lenticular bedding is common in the upper intertidal zone. The lower intertidal zone is characteristic of coarser sediment and thicker sandy laminae. Muddy flaser bedding is common in the lower intertidal zone. Wavy bedding is common in the middle intertidal zone.

84

COASTAL STORMS: PROCESSES AND IMPACTS

Ebb

Flood

Mud Sand Diastem

Erosion Deposition Ucrs Ucrm 0

Ucrm Ucrs

Figure 5.2 Schematic model showing the deposition and erosion of sandy and muddy laminae couplets. Ucrs is the threshold velocity for sand transport and Ucrm is the velocity below which mud deposition occurs (modified from Fan et al., 2004b).

It is generally accepted that four laminae may theoretically be deposited during one tidal cycle (Allen, 1985). Two sandy laminae may be formed during flood and ebb phases, respectively, and two muddy laminae deposited during high and low tide slack water (Figure 5.2). It was also found that thicker sandy laminae may correspond to relatively higher-energy events during spring tides, while thinner sand laminae correspond to relatively low energy events during neap tides (Boersma & Terwindt, 1981; Allen, 1985). Therefore, the quite regular thickness variations of sandy and muddy laminae are often interpreted as the results of spring-neap tidal cycles. This provides a valuable tool for the study of ancient tidal characteristics from rock records. Time-series analysis of laminae number and thickness has been applied to quantify the tide periodicities and sedimentation rates in both modern and ancient tidal deposits (Yang & Nio, 1985; Kvale et al., 1989; Tessier & Gigot, 1989; Kuecher et al., 1990; Kvale & Archer, 1990; Tessier, 1993; Miller & Eriksson, 1997). An approximate 14-day periodicity related to neap-spring cycles has been identified in most of the above studies. Deposition and erosion induced by waves have largely been neglected in the numerous tidal periodicity studies. Along open-coast tidal flats, the four laminae as illustrated in Figure 5.2 were rarely observed to have been deposited and preserved during one tidal cycle, due to their poor preservation potential, as discussed in the following (Li et al., 2000).

SDL

MDL

MDL

SDL

MDL

SDL

85

SDL

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

SDL

SDL

MDL

Figure 5.3 Alternation of sand-dominated layer (SDL) and mud-dominated layer (MDL). Left panel: SDL and MDL from modern Changjiang River delta tidal flat. Right panel: SDL and MDL from Upper Ordovician rock record in east-central China (right panel modified from Fan et al., 2004a).

Two different grouping patterns of sandy and muddy laminae were distinguished on open-coast tidal flat deposits (Figure 5.3). Groups with generally thicker sandy laminae than adjacent groups, are termed sand-dominated layers (SDL), while groups with generally thinner sandy laminae than adjacent groups are referred to as mud-dominated layers (MDL). Although determination of the exact boundaries between sand- and mud-dominated layers was somewhat subjective, the overall differences between adjacent sand- and mud-dominated layers were apparent (Figure 5.3). The thickness and number of sandy and muddy laminae in each sand- or mud-dominated layer were not necessarily identical. The alternating SDL and MDL (Figure 5.3 right panel) is also observed in an Upper Ordovician rock record in east-central China (Fan et al., 2004a). As an alternative to the spring-neap tidal cycle interpretation of the thickness variation of sand and mud laminae, Li et al. (2000) and Fan et al. (2004a) proposed that the MDL represent tidal depositions during calm weather, while SDL represents storm depositions. Li et al. (2000) argued that spring-neap tidal cycle interpretation would require a very high sedimentation rate, which is unrealistic to sustain for extensive open-coast tidal flats as those shown in Figure 5.1. Li et al. (2000) conducted a field study of an open-coast tidal flat over a four-month period, with two months at the end of a calm-weather season and two months at the beginning of a typhoon season. Sedimentation and/or erosion were measured at 35 locations relative to a series of scaled rods. A total of 22 time-series measurements were conducted during the four-month period. Overall, the intertidal zone was accreting during the calm-weather season (Figure 5.4), as indicated by the increasing of the average elevation at all the rods. During the storm season, net erosion (elevation decrease) was measured and the flat was covered by a sandy lamina that was much thicker than those deposited during the calm-weather season. Sharp elevation decrease was usually

COASTAL STORMS: PROCESSES AND IMPACTS

Change of average elevation of the profile

86

(m) 25 Change of average elevation Tidal range Wave height

20 15 10 5 0

4 (m)

Wave height

(m) 4

2

3 1

2 1

5.16

Tidal range

3

Typhoon season

Calm weather season 5.31

6.15

6.30

7.15 7.30 Date

8.14

8.26

9.13

Figure 5.4 Average tidal flat elevation change at Donghai Farm on the southern flank of the Changjiang (aka Yangtze) River delta, spring-neap tidal cycles and wave heights observed at the Ship Observation Station during a four-month study in 1992. Wave heights lower than 1 m were not plotted (modified from Li et al., 2000).

measured directly after the storms, indicating the erosion caused by storm waves. For the convenience of discussion, the calm- and storm-weather season was divided somewhat subjectively by the first significant typhoon impact in the study area. The frequent observations throughout the four-month period indicated that the tidal flat was generally muddier in the calm weather season than during the stormy season, suggesting that the deposition and preservation of mud was hindered by the energetic storm waves. The changes of average elevation were not related to the neap-spring tidal cycles, but were related to high wave events associated with storms (Figure 5.4). Deposition of a relatively thick sandy lamina was directly related to the high-energy wave events induced by the passage of a typhoon, instead of during spring-tide conditions. Fan et al. (2006)

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

87

conducted a detailed study documenting spatial variations of the deposition and erosion patterns across a tidal flat during calm and storm weather. Li et al. (2000) also examined sedimentation rate and preservation potential of individual lamina. The short-term (one spring-neap cycle) sedimentation was measured based on a series of plates inserted in the sediment. Deposition thickness and number of laminae were measured daily. The longer-term sedimentation was measured by counting the number of laminae and mud- and sand-dominated layers in core sections that were deposited over a hundred years (determined based on anthropogenic markers). The centennial sedimentation rate was found to be of the order of 4 cm per year. This high sedimentation rate is related to the tremendous sediment supply from the large Changjiang River. Such a high sedimentation rate should not be expected at locations without enormous sediment supply. Over the hundred-year period, the preservation potential of individual lamina, including both calm-weather and storm deposits, was found to be in the order of 0.2%, which was 45 times smaller than the 9% estimated during a short term of a neap-spring cycle. It is expected that the preservation potential decreases as the temporal interval increases. The hundred-year preservation potential of storm-induced sand-dominated layers was estimated to be of the order of 10%, 50 times higher than the overall lamina preservation of 0.2%. This suggests that the more energetic and thicker storm deposits tend to have a higher preservation potential. The assumption of 100% preservation implied in the spring-neap tidal cycle interpretation of the lamina thickness variation can only be true under special conditions. Fan et al. (2004a) conducted a similar study to that of Li et al. (2000), but in an ancient system of Upper Ordovician Tonglu rhythmites. The Tonglu rhythmites (Figure 5.3 right panel) illustrate apparent similarities with the modern open-coast tidal flat and are composed of sand- (lighter color) and mud-lamina (darker color) couplets. Grouping of sand- (SDL) and mud-dominated layers (MDL) is apparent. The SDLs are characterized by the presence of relatively thick sand laminae, erosional surfaces, mud pebbles and oscillation ripples. Comparison of these SDL-MDL cycles to modern tidal-flat facies suggests that the variation of sand-lamina thickness in the 6.2 m-thick cyclic rock outcrop should be related to wave energy variations associated with storm impacts rather than due to individual tides. Time-series analysis of sandy-lamina thickness variation, conducted by Fan et al. (2004a), revealed a peak period of approximately 14 laminae, with minor peaks occurring at 26 and 64 laminae (Figure 5.5). Although the 14-layer peak coincides with the neap-spring tidal cycles identified in other rhythmites, neap-spring variation was not accepted because of the requirement of an unrealistically high sedimentation rate of 3.43 m/yr. An alternative interpretation was proposed in which the thick sand laminae are related to storm deposition on the basis of erosional surfaces, mud pebbles and oscillatory wave ripples. The SDLs were interpreted as deposits from storm seasons, while MDLs corresponds to calm weather seasons. Applying the preservation potential of SDL obtained by Li et al. (2000) and Fan et al. (2002) from the Changjiang River delta open-coast tidal flat, a reasonable sedimentation rate of 3 cm/yr was obtained for the Tonglu rhythmites.

COASTAL STORMS: PROCESSES AND IMPACTS

Lamina thickness (mm)

88

A: A three-point averaged sand-lamina thickness

35

(a)

30 25 20 15 10 5 0

0 4.0 Power spectra

100

50

(b)

150

200

300 250 Lamina number

350

400

450

500

550

13.5 34.1

3.0 64 2.0

17.7

11.1

8.8

25.6 34.1

1.0 0.0

0.0

0.1

0.2

Frequency

0.3

0.4

0.5

Figure 5.5 Lamina-thickness variations in a section of the Upper Ordovician Tonglu rhythmite. (A) A three-point averaged sand-lamina thickness; (B) Frequency distribution of the thickness variation (modified from Fan et al., 2004a).

5.3 Erosion-deposition processes and morphodynamics of open-coast tidal flat The tidal wave becomes asymmetrical as it propagates over the wide tidal flat, with half of the tidal cycle (flooding phase) shorter but with faster flow, while the other half (ebbing phase) lasts longer with slower flow. Because transport rate is generally proportional to the velocity cubed, much greater rate of transport occurs during period of greater velocity (Figure 5.6). Therefore, this well-known time-velocity asymmetry results in a net transport in the direction of the faster (flood) flow. Time-velocity asymmetry has significant influence on the sedimentation and morphology of a tidal flat. The net onshore directed sediment transport due to the time-velocity asymmetry is attributable to the overall accretionary trend of a tidal flat. For open-coast tidal flat, this accretionary trend is further supplemented by abundant sediment supply. Erosion of cohesive sediment from the bed occurs when the bed shear stress from the fluid (current, wave, or combined) exceeds a critical value for erosion. The critical value for erosion is greater than the critical value for full deposition (Mehta, 1986). In other words, considerably more fluid power is needed to erode the cohesive sediments than to deposit them. This difference in the shear stresses for erosion and deposition, in addition to the time needed for the fine grains to settle, is responsible for the so-called settling lag and scouring lag, which is an important sedimentary process on many mud flats (Postma, 1961; Dyer, 1986; Bartholdy, 2000). The schematics of scour lag and settling lag are well illustrated (Figure 5.7) and explained by Dyer (1994). Theoretically, the scour and settling lag should result in a net deposition of fine grain sediment landward of the area, where maximum velocity during the tidal cycle equals the grain’s threshold

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

89

net transport

Bedload transoprt rate velocity

bedload transport rate

Time

threshold velocity

current velocity

Figure 5.6 Schematic illustration of time-velocity asymmetry. Because transport rate is generally proportional to velocity cubed, much more sediment is transported in the direction of the greater velocity, which results in a net transport toward that direction.

velocity. At many tidal flats, both semi-enclosed and open coast, the net sedimentation on the upper intertidal zone due to settling lag during calm season is often eroded by storm waves during storm season, resulting in a seasonal cycle (Allen & Duffy, 1998; Dyer et al., 2000; Yang et al., 2003; Fan et al., 2006; Talke & Stacey 2003, 2008). At a fine scale, deposition of cohesive sediment is a complicated process due to the concentration- and depth-dependent flocculation. Based on a series of laboratory experiments, Krone (1962) found that deposition occurs when the bed shear stress falls below a critical value for deposition, for example during slack tide. Adopting and expanding the concept of critical shear stress for deposition, Mehta and Partheniades (1975) conducted an extensive laboratory investigation on the deposition of cohesive sediment. They concluded that a given flow can maintain a constant fraction of sediment in suspension regardless of the absolute value of the concentration. Three deposition regimes were distinguished based on the bed shear stress, or the ratio of the equilibrium concentration and the initial concentration. They are full deposition, hindered or partial deposition, and no deposition. It is worth noting that wave motion, which generates additional shear stress, may prevent full deposition during slack tides. This can be quite relevant for open-coast tidal flat due to active wave forcing. The morphodynamics of tidal flats, especially the wide open-coast ones, are not well understood because of: (1) the complicated mixture of cohesive mud and non-cohesive coarse silt and sand (Wang, 2012); (2) further complications by vegetation coverage in the upper intertidal zone and intense bioturbation across the entire flat; and (3) extensive and flat topography and large tidal water-level fluctuation. These make the field measurements of sediment transport and morphology change very difficult. Mehta et al. (1996) conducted an extensive review on sediment transport and morphodynamics of

90

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

(b) Velocity

Velocity

2

1

7

1

Threshold

4

Threshold 5

2

4

3

5

6

3

Distance offshore

Distance offshore

Water trajectory

2 Time

3

Time

2 3 4

Water trajectory 4

Particle trajectory 5 6

Particle trajectory 7

5

Figure 5.7 Schematics of scour lag (A) and settling lag (B) for fine-grain sediments. A: Scour lag: a particle on the bed is suspended into the water column when the threshold velocity is exceeded at point 1. However, it does not achieve the depth-averaged velocity until point 2, a relatively seaward position. It then travels with the water trajectory to point 3, where we assume it is instantaneously deposited. On the following ebb tide, the particle is suspended, but again lags the flow until point 4 is reached. It is eventually re-deposited at point 5. A net landward movement has occurred during the tidal cycle because of the scour lag. B: Settling lag: at position 1, the particle is entrained from the bed and travels with water until point 2, where it starts to settle. Because of the settling lag, it reaches the bed at point 3. On the following ebb tide, it is not entrained until later in the tide cycle when the threshold velocity (greater than the velocity for settling) is reached. The deposition at low water is at position 6. Consequently, the particle has a net shoreward movement due the settling lag (modified from Dyer, 1994).

mud flats and developed a set of empirical formulas describing equilibrium shapes of mud flat. For accretion-dominated mud flat, the profile tends to be high and convex upward. Mehta et al. (1996) proposed the following formula: (𝜋 + 1)−1 L∗ = L 2

(5.1)

where L∗ is the length of the lower part of the profile, L is the distance from the low to high water mark. For erosion-dominated mud flat, the profile tends to be low and concave upward, the following formula is proposed by Metha et al. (1996): ) h (x) ( x 3 = 1− ho L 2

(5.2)

where h = h (x) is the depth of the tidal flat profile, ho is the high water depth at x = 0 and is equal to the tidal range, L is the distance from low to high water mark, and x is the horizontal distance perpendicular to shore.

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

91

Accretionary profile E I e v a t i o n

Erosive profile

Distance across shore

Figure 5.8 A conceptual model of mud shore equilibrium shape (the Mehby Rule). The two end members, accretion-dominated (modeled with Equation 5.1) and erosion-dominated (Equation 5.2) profiles, are illustrated (modified from Kirby, 2000).

Based on the equilibrium shapes of erosive and accretionary profiles developed by Mehta et al. (1996), Kirby (2000) proposed a conceptual model, referred to as the Mehby Rule (adopting the concept of Bruun Rule for sandy shores), describing the profile equilibration and evolution of mud flats (Figure 5.8). The model consists of two end members, including a dynamically stable, high and convex, accretion-dominated profile modeled with Equation 5.1 at one extreme, to a dynamically stable, low and concave, erosion-dominated profile modeled with Equation 5.2 at the other end. Mehta et al. (1996) and Kirby (2000) further proposed a stability number to evaluate the morphological trend of mud flats. The Kirby (2000) model (or the Mehby Rule) can be used to examine long-term morphology trend of tidal flat as controlled by sediment supply and short-term trend due to storms. The Macro-tidal mudflats flanking the Changjiang delta, which face the East China Sea without any morphologic barriers (Figure 5.1), are prone to substantial impact of waves (Figure 5.9). Variations in bed level of the open-coast tidal flat illustrate a strong relationship with the shelf waves (Yang et al., 2003; Fan et al., 2006). Erosion was measured under stormy conditions, including both proximal and distal storms. Subsequent accretion occurred immediately after the dissipation of the storms. In addition to the annual average of two typhoons making landfall on the greater Changjiang delta, several distant typhoons generated high offshore waves during the Fan et al., (2006) study, inducing substantial morphology changes on the tidal flats. The two high wave events shown in Figure 5.9 are both related to distal passage of storms. The cross-shore variations in bed-level changes are controlled by local water depth, tidal current velocity, wave energy, sediment properties, vegetation, and exposure time per tidal cycle. Seven sub-zones are divided in the Fan et al. (2006) study based on the above factors (Figure 5.10), including the mature marsh (M1 and M2), pioneer marsh (M3), upper and lower sections of the bare middle flats (B1 and B2), and upper and lower sections of the lower flats (B3 and B4). These sub-zones illustrated different magnitude and pattern of erosion and deposition under calm and stormy conditions. The Activity (A) and preservation coefficient (K), developed by Allen and Duffy (1998)

92

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

Wind speed Onshore wind Offshore wind

N NW

10

W

8

SW

6

S

4

SE

2

E

0

NE

Heights of high and low waters

−2 5 (m) 4

(b)

Spring

Neap

Tidal level Wave height

Spring

Neap

3

6

2

4

1

2

0

Wind direction

Wind speed (m/s)

12

23 24 25 26 27 28 29 30 31

1

2 3 4 5 6 7 Date (July–August, 1999)

8

9

Offshore wave heights (m)

14

0 10 11 12 13 14

Figure 5.9 (A) Time series of measured wind speed and direction, and (B) Time series of predicted astronomic tides and daily forecasted wave heights (H1∕10 ) at the Changjiang River mouth area from 23 July to 13 August 13, 1999. Two distal storms, with offshore waves exceeding 3 m, occurred during the study period (modified from Fan et al., 2006).

were used to evaluate the overall temporal trend of deposition and erosion. The Activity coefficient A is calculated as: ∑ A= (5.3) |D| where D is the bed level change, with deposition being positive and erosion being negative. A higher A value corresponds to a large overall change of bed level, and therefore indicates active deposition and erosion. The preservation coefficient K is calculated as: | ∑ D+ | | | (5.4) K = |∑ − | | D | | | where D+ denotes deposition and D− denotes erosion. A K value of greater than one indicates net deposition, that is, some of the deposited sediment is preserved. A K value of one suggests no net deposition or erosion. A K value of less than one indicates net erosion, that is, no deposition is preserved. Fan et al. (2006) found that the A and K values varied significantly among the seven sub-zones over their three-month field period. The K values are generally greater than one at the upper tidal flat zones (above the mean sea level) indicating an overall accretionary trend. The A values are greater at the lower tidal flat zones (below the mean sea level) suggesting more active deposition and erosion. The K values at the lower tidal flat zone are less than one during the storm season, indicating an overall erosive trend. Figure 5.11 illustrates erosion and deposition measured at the various intertidal

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

5

Seawall

MHWS M1

13 00 0.

Elevation (m WD)

4

3

93

M2

Upper M3

MHWN B1

Mean sea level

B2

2

0.0

01

Middle

Intertidal zone

MLWN

B3

Lower

1

B4

0.0

MLWS

00

0

7

Subtidal zone

0.0

005

−1 0

1000

2000

3000

4000

5000

6000

Distance seaward (m)

Figure 5.10 Tidal flat zonations based on vegetation cover, tidal level and morphodynamics, including: the Spartina alterniflora dominated marsh (M1), the Scirpus marsh (M2), the pioneer marsh (M3), the upper and lower sections of the bare middle mudflat (B1, B2), and the upper and lower sections of the lower mudflat (B3, B4). Elevation is referred to Wusong Datum (WD) (modified from Fan et al., 2006).

zones during a stormy season. Although the general trend can be summarized as above, substantial temporal and spatial variations were measured (Figure 5.11). Wave heights, particularly the heights of storm waves, are largely modulated by tidal water-level variations over the tidal flat. The breaking wave height, which is particularly active in transporting sediment, is essentially depth limited and is therefore controlled by the local water depth regardless of incident offshore wave heights. Therefore, erosion induced by storm waves is influenced by both the incident wave height and the tidal stage. A weaker storm occurring during spring tides can cause more erosion over a wider area than a stronger storm arriving during neap tides. The redistribution of breaking-wave-induced suspended sediments out of the wave-breaking zone is mainly controlled by tidal flows, which can be augmented by wind- and wave-driven currents. Transport capacity of onshore currents differs significantly from neap to spring tides, accounting for different erosion and deposition patterns on the intertidal zones. Weaker neap tidal currents tend to form accretion zone in the middle tidal flat zones (Figure 5.11) due to the settling of suspended sediment near the wave-breaking zone, while the upper tidal flat zones may suffer erosion from reformed waves. Stronger spring tidal currents can maintain the wave-breaking-induced suspended sediment in the water column for a longer time, and transport and deposit the sediment far from the erosion zone, resulting in high magnitude of erosion within the wave-breaking zone and accretion out of it. Yang et al. (2003) measured up to 4 cm of overwash deposition associated with a strong typhoon in the marsh that is above mean spring high tide, suggesting that energetic storms can be beneficial to high marsh.

94

COASTAL STORMS: PROCESSES AND IMPACTS

M3

B1

B2

B3

+1.1

+2.7

+2.4

+1.6

+4.3

9−13. Aug

−0.5

+1.1

−0.8

+1.3

7−9. Aug

−0.2

−0.7

−2.2

+0.5

−1.7

5−7. Aug

(f) −0.3

−2.0

−0.1

+1.4

+0.9

−3.5

(i)

4 2 0 −2 −4

(h)

2 0 −2 −4

Variations in bed level (cm)

M2 6 4 2 0 −2

12 8 4 0 −4

−0.1

During each elevation -monitoring period Mean tidal Mean wave range (m) height (m)

B4

3.51

2.0

2.39

2.0

1.99

3.5

2.64

2.3

3.26

2.8

3.50

2.5

3.36

2.5

2.48

3.8

1.83

1.8

(g)

3−5. Aug

8 4 0 −4 −8 −12 −16

(e)

8 4 0 −4 −8 −12

(d)

4 2 0 −2 −4 −6 −8

(c)

6 4 2 0 −2 −4 −6

(b) +0.5

4 2 0 −2 −4

(a) -0.3

+1.3

+0.5

+1.8

−0.7

−4.4

−2.2

+2.5 1−3. Aug

−0.6

+1.5

−0.7

+2.6

+0.7

30. Jul 25. Aug

−0.5

+1.4

+2.3

−0.5

−2.5 27−30. Jul

−0.4

−0.9

0.0

−1.7

25−27. Jul +1.1

+0.9

−1.5

+0.9

23−25. Jul 1.0

2.5 1.5 2.0 Distance from the seawall (km)

3.0

Figure 5.11 Short-term (2–3 days) changes in bed level across the tidal flat during a stormy period from 23 July to 13 August 1999, with mean tidal ranges and wave heights during each monitoring period listed. The six morphological zones responded differently to the typhoon-induced swells. Mean changes in bed level over each zone were calculated and indicated in the figures (modified from Fan et al., 2006).

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

95

Typhoon season along the Changjiang River delta coincides with the river flood season and pioneer plants growing season. This has significant influence on the erosional and depositional pattern on the tidal flat. The eroded material by high storm waves can be replenished rather quickly by huge riverine sediment input, and the tidal flat recovers rapidly to its ‘quasi-equilibrium’ profile several tidal cycles after the storms (Fan et al., 2006). Yang et al. (2003) also documented that the severe erosion of up to 10 cm in elevation caused by Typhoon Paibian in the upper intertidal zone recovered rapidly within a few days after the storm. Fan et al. (2006) found that the zone with maximum accretion is often located just seaward of the pioneer vegetation zone during their monitoring period. The accretionary bare flat favors seaward colonization of pioneer plants. Fan et al. (2006) developed a conceptual model depicting the morphodynamics of sediment-rich open-coast tidal flats (Figure 5.12). The model

wave breaking

wave damping

(a) Accretion Erosion Net sediment transport M1 + M2

M3 B1

B2

B3

B4 Subtidal

Intertidal reduced wave breaking reforming

(b)

wave breaking

wave damping

old profile new profile

M1 + M 2

M3 B1

B2

B3

B4 Subtidal

Intertidal reduced wave breaking reforming

(c) inner accretion zone

M1 + M2

wave breaking

middle accretion zone M3 B1

B2

Intertidal

B3

B4

wave damping

outer accretion zone Subtidal

Figure 5.12 A conceptual model depicting intertidal morphodymics and sediment transport in response to storm wave processes at different tidal regimes: (a) spring tide, (b) intermediate tide, and (c) neap tide (modified from Fan et al., 2006).

96

COASTAL STORMS: PROCESSES AND IMPACTS

incorporates forcing variations during spring and neap tides, as well as during stormy and calm conditions. The Fan et al. (2006) model shows considerable similarities with the earlier Yang et al. (2003) model, but included more factors controlling the deposition and erosion at an open-coast tidal flat, and with improved spatial resolution.

5.4 Conclusions Open-coast tidal flats distribute broadly worldwide along meso- to macro-tidal coasts with abundant fine-grain sediment input, for example from large rivers. Due to the lack of morphologic barrier blocking incident ocean waves, open-coast tidal flats are susceptible to impacts from high waves generated by both proximal and distal storms. Under calm weather conditions, due to time-velocity asymmetry and scour- and settling-lags associated with flooding and ebbing tidal flows, in addition to the typical abundant sediment supplies, open-coast tidal flats tend to be accretionary. High-wave energy associated with proximal and distal passages of storms provides the main mechanism for tidal flat erosion. Accretionary tidal flats tend to be convex upward, while erosive tidal flats tend to be concave upward. Spatial and temporal patterns of erosion and deposition associated with storm impact vary considerably, controlled by tide modulated wave breaking, and relatively strong flow during spring tides versus weak flow during neap tides. Open-coast tidal flats illustrate a characteristic sequence of sedimentary structures composed of muddy and sandy laminae. A distinctive group with thicker sandy laminae is referred to as Sand-Dominated Layer (SDL), while a group with thicker muddy laminae is referred to as Mud-Dominated Layer (MDL). Alternation of SDL and MDL is identified in both modern and ancient tidal flat environments. SDL is deposited during storm season, while MDL layer corresponds to calm weather season. The alternation of SDL and MDL therefore represents the variation of storm and calm weather season. This understanding is in contrast to the commonly used spring-neap tidal cycle interpretation.

References Allen, J.R.L. (1985) Principles of Physical Sedimentology. London-Boston-Sydney: George Allen & Unwin. Allen, J.R.L. & Duffy, M.J. (1998) Temporal and spatial depositional patterns in the Severn Estuary, southwestern Britain: Intertidal studies at spring-neap and seasonal scales, 1991–1993. Marine Geology, 146, 147–171. Bartholdy, J. (2000) Process controlling import of fine-grained sediment to tidal areas: A simulation model. In: K. Pye & J.R.L. Allen (Eds) Coastal and Estuarine Environments: Sedimentology, geomorphology, and geoarchaeology. Geological Society, London, Special Publications 175, pp.13–29. Bates, R.L. & Jackson, J.A. (Eds) (1980) Glossary of Geology. Alexandria, VA: American Geological Institute. Boersma, J.R. & Terwindt, J.H.J. (1981) Neap-spring tide sequences of intertidal shoal deposits in a mesotidal estuary. Sedimentology, 28, 151–170. Chen, X. (1998) Changjiang (Yangtze) river delta, China. Journal of Coastal Research, 14, 838–858.

CH5 STORM IMPACTS ON OPEN-COAST TIDAL FLATS

97

Dyer, K.R. (1986) Coastal and Estuarine Sediment Dynamics. Wiley, Chichester, p. 342. Dyer, K.R. (1994) Estuarine sediment transport and deposition. In: K. Pye (Ed.) Sediment Transport and Depositional Processes, Blackwell Scientific Publications, Oxford, pp. 193–216. Dyer, K.R., Christie, M.C. & Wright, E.W. (2000) The classification of intertidal mudflats. Continental Shelf Research, 20, 1039–1060. Eisma, D. (1998) Intertidal Deposits – River Mouths, Tidal Flats, and Coastal Lagoons. Boca Raton, FL: CRC Press. Fan, D.D. (2012) Open coast tidal flats. In: R.A. Davis & R.A. Dalrymple (Eds), Principles of Tidal Sedimentology, Springer, pp.187–230. Fan, D., Li, C.X., Archer, A.W. & Wang, P. (2002) Temporal distribution of diastems in deposits of an open-coast intertidal flat with high suspended sediment concentrations. Sedimentary Geology, 186, 211–228. Fan, D., Li, C. & Wang, P. (2004a) Influences of storm erosion/deposition on rhythmites of the late-Ordovician upper Wenchang formation around Tonglu, Zhejiang province, China. Journal of Sedimentary Research, 74, 527–536. Fan, D., Li, C., Wang, D., Wang, P., Archer, A.W. & Greb, S.F. (2004b) Morphology and sedimentation on open-coast intertidal flats of the Changjian Delta, China. Journal of Coastal Research, SI 43, 23–35. Fan, D., Guo, Y., Wang, P. & Shi, Z. (2006) Cross-shore variations in morphodynamic processes of an open-coast mudflat in the Changjiang Delta, China: With an emphasis on storm impacts. Continental Shelf Research, 26, 517–538. Feagin, R.A. (2008) Vegetation’s role in coastal protection. Science, 320, 176–177. Feagin, R.A., Lozada-Bernard, S.M., Ravens, T.M., Moller, I., Yeager, K.M. & Baird, A.H. (2009) Does vegetation prevent wave erosion of salt marsh edge? PNAS, 106, 10109–10113. Gedan, K.B., Kirwan, M.L., Wolanski, E., Barbier, E.B. & Silliman, B.R. (2011) The present and future role of coastal wetland vegetation in protecting shoreline: Answering recent challenges to the paradigm. Climate Change, 106, 7–29. Kirby, R. (2000) Practical implications of tidal flat shape. Continental Shelf Research, 20, 1061–1077. Klein, G. deV. (1976) Holocene Tidal Sedimentation. Stroudesburg: Dowden, Hutchinson & Ross. Krone, R.B. (1962) Flume studies of the transport of sediment in estuarial shoaling processes. Final Report, Hydraulic Engineering and Sanitary Engineering Research Laboratory, University of California at Berkeley. Kuecher, G.J., Woodland, B.G. & Broadhurst, F.M. (1990) Evidence of deposition from individual tides and of tidal cycles from the Francies Creek Shale. Sedimentary Geology, 68, 211–221. Kvale, E.P. & Archer, A.W. (1990) Tidal deposits associated with low sulfur coals, Brazil FM (Lower Pennsylvanian), Indiana. Journal of Sedimentary Petrology, 60, 563–574. Kvale, E.P., Archer, A.W. & Johnson, H.R. (1989) Daily, monthly, and yearly tidal cycles within laminated siltstones of the Mansfield Formation of Indiana. Geology, 17, 365–368. Li, C., Wang, P., Fan, D., Dang, B. & Li, T. (2000) Open-coast intertidal deposits and the preservation potential of individual lamina: A case study from east-central China. Sedimentology, 47, 1039–1051. Li, C., Wang, P. & Fan, D. (2004) Open coast tidal flat deposits. In: M. Schwartz (Ed.) Encyclopedia of Coastal Science, pp. 1206–1209. Mehta, A.J. (1986) Characterization of cohesive sediment properties and transport processes in estuaries. In: A.J. Mehta (Ed.) Estuarine Cohesive Sediment Dynamics, Springer, New York, pp. 290–325.

98

COASTAL STORMS: PROCESSES AND IMPACTS

Mehta, A.J. & Partheniades, E. (1975) An investigation of the deposition properties of flocculated fine sediments. Journal of Hydraulic Research, 12, 361–381. Mehta, A.J., Kirby, R. & Lee, S.-C. (1996) Some observations on mudshore dynamics and stability. Report to US Army Corps of Engineers, UFL/COEL/MP, 96/1, pp. 65. Miller, D.J. & Eriksson, K.A. (1997) Late Mississippian prodeltaic rhythmites in the Appalachian Basin: A hierarchical record of tidal and climatic periodicities. Journal of Sedimentary Research, 67, 653–660. Postma, H. (1961) Transport and accumulation of suspended matter in the Dutch Wadden Sea. Netherland Journal of Sea Research, 1, 148–190. Reineck, H.E. & Singh, I.B. (1980) Depositional Sedimentary Environments. New York: Springer-Verlag. Talke, S.A. & Stacey, M.T. (2003) The influence of oceanic swell on flows over an estuarine intertidal mudflat in San Francisco Bay. Estuarine Coastal and Shelf Science, 58, 541–554. Talke, S.A. & Stacey, M.T. (2008) Suspended sediment fluxes at an intertidal flat: The shifting influence of wave, wind, tidal, and freshwater forcing. Continental Shelf Research, 28, 710–725. Tessier, B. (1993) Upper intertidal rhythmites in the Mont-Saint-Michel Bay (NW France): Perspectives for paleo-reconstruction. Marine Geology, 110, 355–367. Tessier, B. & Gigot, P. (1989) A vertical record of different tidal cyclicities: An example from the Miocene Marine Molasse of Digne. Sedimentology, 36, 767–776. Wang, P. (2012) Principles of sediment transport applicable to tidal environments. In: R.A. Davis, & R.A. Dalrymple (Eds), Principles of Tidal Sedimentology, Springer, pp. 19–34. Yang, C. & Nio, S. (1985) The estimation of palaeohydrodynamic processes from subtidal deposits using time-series analysis methods. Sedimentology, 32, 41–57. Yang, S.-L., Friedrichs, C.T., Shi, Z., Ding, P.-X., Zhu, J. & Zhao, Q.-Y. (2003) Morphological response of tidal marshes, flats and channels of the Outer Yangtze River Mouth to a major storm. Estuaries, 26, 1416–1425.

6 Storm Impacts on Cliffed Coastlines Sue Brooks1 and Tom Spencer2 1 Department of Geography, Environment and Development Studies Birkbeck, University of London, UK 2 Cambridge Coastal Research Unit, University of Cambridge, UK

6.1

Introduction

Coastal cliffs provide some of the most spectacular landscape scenery around the world. Comprised of over-steepened slopes of varying elevation, gradient and geological composition, they demonstrate a diversity of process mechanisms, retreat rates and shoreline responses. Although it has not been substantiated, Emery and Kuhn’s (1982) assertion that 80% of the world’s coastline is cliffed is frequently quoted. The implication is that cliffs are ubiquitous worldwide, so understanding cliff responses to storm impacts is vital for populations, infrastructure, societal functioning and economic viability on a global scale. Establishing the timing, spatial distribution and scale of cliff-top retreat is, therefore, critical. One component of this more general concern is how cliff systems respond to storm impacts. In this chapter we focus on cliffs from around the United Kingdom, developed in a variety of geological formations, to exemplify the complex and diverse impacts storms can have on cliff responses. The locations of the cliffs we consider are shown in Figure 6.1. This is a difficult task for many reasons, involving at its core the fundamental interaction between: (1) site-specific, often complex, geological lithologies and stratigraphies and (2) multiple marine and subaerial erosive processes (Trenhaile, 1987; Sunamura, 1992). As we show below, and as has been exemplified by recent probabilistic modelling approaches to cliff retreat (e.g. Hall et al., 2002; Walkden & Hall, 2005; Young et al., 2011), these interactions have magnitude and frequency characteristics that are highly scale dependent (see also Cambers, 1976; Lee et al., 2001; Quinn et al., 2010). This scale dependency encompasses notions of, on the one hand, high levels of small-scale temporal and spatial variability in cliff morphodynamics and, on the other hand, episodicity in the infrequent (spatial and temporal) incidence of major cliff retreat events separated by periods of cliffline stasis. The presence of ‘hotspots’ illustrates both patterns, as captured in Kennedy’s description of cliff retreat as ‘regionally negligible and locally high’ at Sunset Cliffs, San Diego (in Young et al., 2011). And this linkage of space and time Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

100

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 6.1 Locations of cliffs from around the UK that are discussed in this chapter.

scales has been formalised in Drake and Phipps’ (2006) model for the chalk cliffs of Hunstanton, UK, which have long-term (1885–2004) retreat rates ranging from 0.02 to 0.25 m a−1 . Their model used successive field surveys along with ergodic substitution to develop a five-stage model of cliff retreat and thereby account for alongshore spatial variability in cliff form (Figure 6.2). The results emphasised the importance of geology in determining cliff response to storms, with the weaker basal layers of Red Chalk and Carstone being readily undercut leaving unstable blocks of White Chalk that fail along joints and bedding planes. Storms have a number of components that make them particularly effective agents of cliff retreat. They can simultaneously deliver large quantities of rainfall to the cliff, raise the still water level (sometimes by several metres) above astronomically predicted tidal

CH6 STORM IMPACTS ON CLIFFED COASTLINES

101

Figure 6.2 Five-stage model of cliff retreat for chalk cliffs, Hunstanton, UK (A) Early stage with few blocks at cliff base; (B) Undercutting of cliff and blocky collapse; (C) Overhanging large blocks fail through cliff base unloading; (D) Further undercutting and development of significant cliff overhang; (E) Cliff collapse and development of talus ramp at the cliff base (redrawn from Drake & Phipps, 2006).

102

COASTAL STORMS: PROCESSES AND IMPACTS

levels and drive the formation and propagation of waves that cross energy thresholds for lowering the fronting beach. With enhanced wave runup, and abrasion from gravel-sized beach particles, storms can lead to basal undercutting of cliffs. In this chapter we look at both the impact of individual storms but also stress the importance of sequences of storms and longer phases of storminess. We note that the Accumulated Excess Energy (AEE) concept, developed by Hackney et al. (2013) in a model of cliff response to elevated water levels and waves under sequences of storms, emphasises the importance of phases of storminess, rather than individual storms, in explaining variations in rates of cliffed shoreline retreat. In this model the effect of total energy (AEE) above a threshold is shown to be important rather than simply the energy associated with particular large individual storms. However, of primary significance in any meaningful discussion of storm impacts on cliffs is the matching of process environment (phases of storminess or individual storm impacts) with morphological response (storminess-related variations in cliff retreat rate or the imprint of individual storms). Seeing a storm impact signal is strongly related to the rate of cliff retreat. These rates vary greatly (French, 2001), with the primary control being lithology. It is widely accepted that ‘soft’ rock cliffs retreat at rates > 1 m a−1 (Collins & Sitar, 2008; Young et al., 2009; Brooks & Spencer, 2010) while ‘hard’ rock cliffs may retreat at just a few millimetres per year (Drake & Phipps, 2006; Rosser et al., 2007), although there is no precise definition of how these cliff types can be distinguished (Naylor et al., 2010). Here we suggest that cliffs lie on a spectrum from the most resistant consolidated geological formations to totally unconsolidated over-steepened sand dunes that are highly unstable and maintain their gradient through sustained negative pore water pressures (Hutchinson, 1970; Lahousse & Pierre, 2003; Brooks et al., 2012; Armaroli et al., 2013). This is illustrated in Plate 6.1. Analytical approaches across this spectrum need to vary: it is difficult to extract the storm impact signal in very resistant materials, whereas rapid retreat rates in soft rock cliffs mean that the record is only seen from historical information on where the cliff was in the past, not from its contemporary position and condition. It is only in moderately retreating cliff systems that both the process by which retreat is taking place and the record of that retreat process can be seen and quantified, such as for the Dorset cliffs of the UK (Brunsden & Jones, 1976; Brunsden & Chandler, 1996). Wave action (Carter & Stone, 1989; Adams et al., 2005; Rosser et al., 2007; Young et al., 2009; Castedo et al., 2012; Earlie et al., 2015) can undercut, oversteepen and initiate stress vibration within the cliff, while shear strength of cliff materials is highly dependent on prevailing positive and negative pore water pressure distributions (Hutchinson, 1970; Brooks et al., 2012). The prevailing conceptual model of cliff retreat involves a three-stage process (Trenhaile, 1987; Sunamura, 1992). Stage 1 involves the erosion of the cliff base through wave action thus increasing the slope gradient, undermining overlying structures and thereby lowering the stability of the cliff. Stage 2 involves cliff collapse largely brought about through subaerial terrestrial processes within the cliff. Cliff failure might involve the sudden collapse of the entire cliff face or take place as a series of smaller failures that propagate from the bottom up through the cliff face. This generates cliff foot deposits in the form of talus that are then removed by wave action during stage 3. It is difficult to place a timescale around these stages but Young et al. (2009) suggest stage 1 takes place over years, while stage 2 is often abrupt and linked to single events. Stage 3 can take weeks to years depending on the amount of material to be removed and the frequency, duration and magnitude of wave attack.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

(a)

(b)

(c)

(d)

(e)

(f)

103

Plate 6.1 Cliff spectrum: (A) 500–800 m high volcanic cliffs at Los Gigantes, Tenerife, Canary Islands (photo: T. Spencer); (B) White Chalk, Red Chalk and Carstone cliffs of Hunstanton, UK (photo: S. Brooks); (C) Liassic Clay cliffs of Charmouth, Dorset, UK (photo: S. Brooks); (D) Tertiary (Eocene) siltstones, sandstones and volcanic ash with overlying basalts on the cliffs of Oregon, USA; (E) Bluff erosion in soft sediments of Pacifica, California, USA; (F) Soft rock cliffs in glacial and pre-glacial sands and silts at Covehithe, Suffolk, UK (photo: S. Brooks).

104

COASTAL STORMS: PROCESSES AND IMPACTS

6.2 Methodologies and their application In theory at least, it is comparatively straightforward to quantify shoreline change for eroding cliffs. This is because change is unidirectional, generally progressing inland (although some localised cliff base sections might be accumulating material and exhibit short-term phases of advance). Much discussion on quantifying shoreline change has revolved around defining where the shoreline is located (Harley, 1972; Moore, 2000). Map sequences do not always use the same shoreline marker over time. For example, the Ordnance Survey of England and Wales have used Mean Level of Ordinary Tides (MLOT) on the earliest maps but Mean High Water Springs (MHWS) for later editions. Problems are compounded when switching to more recent aerial photographs, where traditional shoreline markers are not visible, and alternatives such as the vegetation line or beach morphological features need to be used. The position of many shoreline markers will vary depending on the time of day (and hence both tidal level and angle of illumination) and season in which the aerial photograph was captured. However, cliffs commonly present sharp and clear markers of the shoreline on both aerial photographs and historic maps (Brooks & Spencer, 2010). Hence we tend to see a consistent, clear and readily digitisable shoreline marker that enables maps and aerial photographs to be used together to quantify retreat rates on different timescales. The preferred method to quantify cliff retreat depends upon the overall rate of retreat, as error terms in the methodology need to be clearly separable from the actual rate at which the process is operating. For rapidly retreating cliffs (> 1 m a−1 ) the traditional approach has been the use of historic maps, aerial photographs (e.g. Moore, 2000) and direct ground survey (e.g. Sallenger et al., 2002; Dornbush et al., 2011). For slower rates, we might employ erosion pins (Greenwood & Orford, 2008) or marker-posts, or for much longer-term studies interrogate historical photographs, as in the classic studies of the sea cliffs of southern California (Shepard & Grant IV, 1947; Kuhn & Shepard, 1984). In recent years, however, the application of new technologies to monitor slow or single-event retreat has been considerable and has revealed significant new knowledge of the spatial patterning of cliff retreat. Thus, for example, at the relatively large scale, airborne LiDAR surveys are now being used routinely to repeat map basic two-dimensional (planimetric) patterns of cliff edge failure over kilometres of cliffed margins in many parts of the developed world (e.g. Hapke et al., 2009; Young et al., 2011), and supplemented by more detailed ground-based LiDAR in some locations (e.g. Young et al., 2009). These methods can provide detailed 3D images of cliff face change, including evacuation and accumulation of sediment at very high vertical and lateral resolution. Even finer scale mapping has recently been achieved using Terrestrial Laser Scanning (TLS) to quantify volume changes in the cliff face over short time periods (Young & Ashford, 2006; Rosser et al., 2007; Collins & Sitar, 2008; Rosser et al., 2013). Lim et al. (2011), for example, compiled a database of over 100,000 coastal rockfalls, down to volumes of 1.25 × 10−4 m3 , using monthly terrestrial laser scanning over a 20-month period on a coastal cliff rock face at Staithes, North Yorkshire, UK. These data-rich studies have highlighted the complexity and high degree of uncertainty over the precise roles of marine and terrestrial forcing of cliff retreat. This is mainly because of the spatial variability in cliff composition and structure that generates lags in response timing. Hence cliff collapse can take place weeks or months after the occurrence of a

CH6 STORM IMPACTS ON CLIFFED COASTLINES

105

Retreat rate (m a–1)

large storm as failure propagates upwards through the cliff face (Rosser et al., 2013) or because positive pore water pressure or reduced suction take time to develop within the cliff following rainfall (Hutchinson, 1970; Brooks et al., 2012). The latter is strongly linked to cliff geological structure, which can exhibit considerable alongshore variation over relatively short stretches of shoreline. Furthermore, contemporary field surveys of cliff-top position using the latest Real Time Kinematic (RTK) instrumentation can now be used alongside maps and remotely sensed imagery to gain a more detailed picture of the alongshore variations in cliff retreat rates. However, such technologies only cover the most recent time periods, so recourse to indirect methods remains necessary to answer wider questions. Indirect methods, such as use of computer models and historic maps or aerial photograph digitisation have the advantage of covering longer periods of time, enabling successions of storms to be assessed. They also enable analysis of periods when direct observation is not possible, such as historic change, but also projections of cliffline retreat into the future. The indirect methods have been greatly facilitated by the development of techniques such as the Digital Shoreline Analysis System (DSAS) of the USGS (Thieler et al., 2009). This method, in which cross-shore transects can be cast within a GIS framework and retreat calculated between different time intervals, enables a very high level of spatial densification in alongshore retreat rates to be achieved. Digitised shorelines from maps, aerial photographs and RTK field data have been used along with DSAS for both Holderness (Castedo et al., 2015) and the East Anglian (Brooks & Spencer, 2010, 2012, 2014) soft rock cliffs of the UK. This methodology provides a quick and accurate assessment of cliff-top retreat that takes account of the

9 8 7 6 5 4 3 2 1 0

EPR 1883–2008 1992–2008 1992–2001 2001–2008

0

250

500 750 Distance (m)

1000

0– 2 2– 4 4– 6 6– 8 8– 10 10 –1 2 >1 2

Net Shoreline Movement (m) NSM

Figure 6.3 Shoreline retreat in soft rock cliffs over different timescales: (A) Long-term and medium-term rates of change; (B) Event-driven short-term retreat.

106

COASTAL STORMS: PROCESSES AND IMPACTS

whole shoreline and not just point measurements. When allied to the more recent RTK field surveys, a detailed picture can be developed around the changes in cliff-top position on different timescales. For the rapidly retreating cliffs of the Suffolk coast, Figure 6.3 shows shoreline position over the long term (hundred-year timescale), medium term (decadal timescale) and following the 5–6 December 2013 storm surge. Long-term retreat has been steady at between 3 and 4 m a−1 with a smooth alongshore trend. For the past two decades there is a strong contrast in retreat around the long-term mean, with 1992–2000 having rates between 4 and 7 m a−1 , while 2000–2008 has rates ranging from 1 to 5 m a−1 . There is greater alongshore variability than for the longer timescale, reflecting the role of shorter phases of storminess or occurrence of individual storms. The 1990s is known to have had more storms than the 2000s (Brooks et al., 2012) and this signal is region-wide. The alongshore change in geology between till capped cliffs and cliffs having just interbedded soft sediments is also shown. The decadal signal is only evident for the till capped cliffs, further suggesting a different response to storms, with individual large events generating most of the retreat. In the second system there is a more steady retreat through time, further suggesting that different retreat mechanisms apply. On the shortest timescale, that of the individual storm surge event, retreat can be up to 12 m (Spencer et al., 2015) at individual locations along the cliffline.

6.3 Storminess and the cliff record The East Anglian example above shows the difficulty in trying to map synoptic scale storm events, and even periods of storminess, onto a climatological index derived from atmospheric conditions over a number of months and used to characterise a particular climatological year/years (Burningham & French, 2012). However, when the climatological signal is clear, there are possibilities for linking changes in storminess to cliff responses. Here we give two examples. In the first, the cliff system, in very resistant materials, is largely passive, but provides a pathway whereby storm impacts on the cliff face are recorded in cliff-top locations. In the second, cliff recession rates are related to clear, climatologically-driven variations in still water level and wave heights. Along the western and northern cliffed margins of the British Isles – Shetland, Orkney, Caithness and the Outer Hebrides in Scotland and the Aran Islands, Galway Bay in Ireland – scoured cliff-top surfaces are often covered with a boulder ridge up to 20 m above sea level, individual boulders and imbricated boulder clusters to + 35 m, and air-thrown material at even greater elevations, known collectively as Cliff-top Storm Deposits (CTSDs) (Hall et al., 2006). Severe wave conditions occur during storms in the North Atlantic and it is not uncommon for deep water waves in the sea areas to the west of the Shetlands to achieve maximum heights of 20 m or more. Often with deep water immediately offshore, these waves lose little of their considerable energy when encountering steep sea cliffs. Rather than concentrating wave attack at the cliff base, these giant waves are capable of generating cliff-top forces sufficient to fracture bedrock and to pluck, rotate and lift boulders as large as 277 m3 . Furthermore, the cliff-top platforms and ramps can be overtopped by such waves and inundated by fast-moving, ‘green water’ bores capable of transporting boulders of up to 40 m3 over tens of metres inland (Hansom et al., 2008). Whilst many of these

CH6 STORM IMPACTS ON CLIFFED COASTLINES

107

deposits are ‘fresh’ and can be traced back to major storms in the 1990s, it is clear that they often veneer much older CTSDs. From the cliffs of Shetland, radiocarbon dating of peats beneath storm-deposited boulders and of marine shells buried inside CTSD boulder ridges, and optically stimulated luminescence (OSL) ages on sands interbedded with storm-derived boulders, suggest that major phases of storm activity occurred during the periods AD 400–550, 700–1050 and 1300–1900, as well as between 1950 and 2005. This record correlates well with the Greenland sea salt (Na+) record in the GISP2 ice core, which itself reflects patterns of general storminess in the North Atlantic Ocean (Figure 6.4; Hansom & Hall, 2009). Expansion of Icelandic sea ice creates stronger thermal gradients and greater capacity for cyclogenesis to generate strong extra-tropical storms with dominantly south-westerly winds that then cross northern Europe. Conversely, ice retreat reduces the capacity for cyclogenesis and northern Europe experiences less stormy phases. Dawson et al. (2004) refer to this as the ‘see-saw’ effect, reflected in the strength of the Icelandic Low. The Low forms one half of the North Atlantic Oscillation (NAO) index, a winter (October to March) measure of the difference in normalised sea-level pressures across the North Atlantic Ocean between a sub-tropical high and a polar low (Hurrell, 1995). Interestingly, the third cluster of CTSD dates correspond to the onset of the Little Ice Age (ca. AD 1400) which was associated with an increase in sea ice cover (Hansom & Hall, 2009). Considerable proportions of the coast of the US States of California, Oregon and Washington are cliffed. Typically, embayments backed by uplifted and cliffed Pleistocene terraces lie between steep headlands in more resistant sandstones, mudstones and conglomerates, and in places highly resistant crystalline rocks, which plunge into deep water immediately offshore (Hapke et al., 2009; Committee on Sea-level rise in California, Oregon and Washington, 2012). In California, Moore et al. (1999) reported lithology – dependent cliff and bluff erosion rates of 0.02–0.2 m a−1 (1932–1994) in San Diego County, and 0.06–14 m a−1 (1953–1994) in Santa Cruz County, California. Griggs and Patsch (2004) state that Californian cliffs and bluffs in sedimentary rocks typically erode at rates of 0.15–0.30 m a−1 . On the Oregon coast, Priest (1999) reported cliff and bluff retreat at rates < 0.19 m a−1 (1939–1991), rising to up to 0.5 m a−1 in landslide-prone bluffs. These background rates, however, are orders of magnitude lower than the rates of retreat that result from the episodic combination of high water levels and storm waves on this high energy coastline. Water level records along the west coast of the Americas show considerable inter-annual variability. In recent times, exceptionally high annual sea levels occurred during 1982–1983 and 1997–1998. Both of these periods included major El Niño events, the warm phase of the large-scale atmosphere – ocean oscillation, known as ENSO (El Niño – Southern Oscillation), that characterises the equatorial Pacific Ocean. The wind stress imparted by the SE Trade Winds over the great distances of the Pacific, results in a Western Pacific ‘warm pool’ which typically sits at a sea level ca. 45 cm higher than the sea surface in the Eastern Pacific. When in El Niño events the Trade Winds weaken, and are replaced by westerly wind bursts, this warm pool ‘sloshes back’ along the equator, in the form of unusually long (ca. 10,000 km), relatively fast-moving (8 km hr−1 ) ‘Kelvin waves’. When a Kelvin wave hits South America, it splits and moves both northward and southward along the coast; Enfield and Allen (1980) have identified coherent sea-level highs related to El Niño as far north as Alaska and as far south as Valparaiso. With the addition of a warm water effect and stronger northward

108

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 6.4 Historic storminess identified from cliff-top storm deposits: (A) Icelandic low proxy record; (B) Greenland sea salt observed record; (C) Shetland cliff-top storm record (redrawn from Hansom & Hall, 2009).

CH6 STORM IMPACTS ON CLIFFED COASTLINES

109

currents, mean monthly average water levels were 26 cm (1982–1983 El Niño) and 33 cm (1997–1998) higher than the long-term averages (1967–1999) (Komar & Allen, 2002). However, the impacts of the 1982–1983 storms were greater because, unlike 1997–1998, they coincided with unusually high predicted tidal levels (Storlazzi et al., 2000). Komar (2004) calculates that a 50 cm increase in water level on a beach with a slope of 1:25 (0.04), typical of the Oregon coast, will shift the mean shoreline landward by 12.5 m, increasing the probability that wave run-up will reach the cliffs backing the beaches, resulting in their erosion. This effect is magnified because El Niño phases also result in the reorganisation of atmospheric circulation patterns over the Pacific Ocean, resulting in an increased frequency and severity of winter storms. Storm-generated wave heights can be very high on this coast; the 1997–1998 El Niño storms were characterised by 10.5 m waves, with breaker heights, reaching 11.7 m; the ‘hundred-year wave’ is calculated to be 15–16 m (Allan & Komar, 2006). If even a 1-m-high storm surge is superimposed on an already elevated monthly water level, the total enhanced water level of 1.5 m will shift the shoreline landward by 38 m, a substantial portion of the width of most Oregon beaches (Komar, 2004). Furthermore, in El Niño years, storms often approach continental margins along a south-westerly track. Within the embayments of the Pacific NW coast, these storms drive alongshore movements of beach sediments, creating rip current embayments (Shih & Komar, 1994; Sallenger et al., 2002) and cause the migration of tidal inlets and river mouths, both creating local erosional ’hotspots’. It is not surprising, therefore, that El Niño climatologies generate typical accelerated cliff line retreat on the Oregon coast of 10–45 m (Allan, 2006). In 1997–1998, 130 m of retreat was recorded at Willapa Bay, Oregon (USGS, 2013). On the coast of central California, 76% of cliff erosion appears due to El Niño-related storms (Storlazzi & Griggs, 2000). Between October 1997 and April 1998, following a series of El Niño-related winter storms, LiDAR profiling of the cliffs at La Pacifica, south of San Francisco Bay showed a retreat of 10–13 m. The long-term cliff retreat rate here is 0.2 m a−1 , suggesting that these storms accounted for 50 years of ‘normal’ retreat over a single winter season (USGS, 2013). The coastal damage in southern and central California during the 1982–1983 El Niño was exceptionally severe, in part because of the coincidence of a high storm surge and the highest tides in four years (Flick, 1998). Seven large wave or storm events occurred during the first three months of 1983, when most coastal erosion took place, and the arrivals of these large waves coincided with times of very high tides, thereby concentrating more wave action directly on the shoreline and cliffs. This El Niño-determined chronology of cliffline retreat is, however, complicated by two additional factors. First, there are large-scale variations in sea-level change along the Pacific coast as a result of the tectonics of the Pacific-North America plate collision zone. Thus, sea levels are falling in northern Washington and southern Oregon but rising in northern Oregon and California (Komar & Shih, 1993; Komar et al., 2011). Thus, enhanced wave runup in El Niño phases is in part offset in those regions of tectonic uplift, but augmented on the submerging sections of this coast. Second, these El Niño effects need to be seen in the context of decadal-scale changes in wave climate. There was a 65% increase in the frequency of cyclonic disturbances between 1948 and 1997 (Graham & Diaz, 2001). Analyses of wave buoy measurements collected during the past 25–30 years in the eastern North Pacific have shown that deep-water wave heights and periods have increased, particularly off the coast of Washington, where the mean

110

COASTAL STORMS: PROCESSES AND IMPACTS

winter deep-water significant wave height has increased by 0.8 m in 25 years, with the maximum winter wave heights rising by well over 2 m (Allan & Komar, 2006; Ruggiero, 2013). Again, these differences have been regionally variable – with no decadal change in southern California – as a result of the changes in regional climatology, which have sent major storm paths to the north. Indeed, the most severe storms, with wave heights in excess of 14 m, have not been related to El Niño dynamics. What this example demonstrates is that, on the one hand, cliff retreat can be strongly determined by ocean-scale fluctuations in atmosphere-ocean climatology, but also, on the other hand, that such storminess controls are further complicated by other climatic and non-climatic controls. This makes the future prediction of the magnitude and location of accelerated cliff retreat in storms on the Pacific coast and elsewhere a challenging field of research.

6.4 Case study: Soft rock cliff geology and responses to storms The geological structure of soft-rock cliffs has been closely associated with the nature of mass movement (e.g. Brunsden & Jones, 1976; Brunsden & Chandler, 1996; Brunsden & Lee, 2004; Gray, 1988). In particular, layered cliff stratigraphies have the potential to generate complex hydrological and geotechnical responses (Rulon & Freeze, 1985; Rulon et al., 1985). Hence a link between high spatio-temporal variability in cliff retreat rates and the considerable variability in the cliff structures that dictate the nature, rate and occurrence of mass movement might be expected. Thus, mass movement in cliffs may or may not be closely coupled with basal removal and, in places may act as the independent supplier of material to the beach below. To assess this in greater detail indirect methods are used as these can couple alongshore variation in retreat (from maps and aerial photographs) with cliff responses for different geological formations. For terrestrial hillslopes, mass movement involving dynamic hydrological responses has been associated with periods of high rainfall and suction loss within the unsaturated zone (e.g. Campbell, 1975; Brooks & Anderson, 1995; Wilkinson et al., 2002; Brooks et al., 2004; Gofar & Lee, 2008; Lee et al., 2009). Suction loss has been shown to be an important failure mechanism under high permeability rates in relatively coarse-grained materials, especially where layering (strata) is present. It is also a significant control on slope stability where materials are very deep. Research has been carried out in both tropical (Anderson et al., 1994; Anderson et al., 1996; Lu & Griffiths, 2004; Lee et al., 2009) and temperate (Brooks & Anderson, 1995; Brooks et al., 2002; Brooks et al., 2004) settings. This potentially significant control on mass failure has recently been highlighted for coastal cliffs, where it has been demonstrated to be significant in moderately-cemented sands forming coastal bluffs in northern California (Hampton & Dingler, 1998; Collins & Sitar, 2008). This phenomenon can be readily exemplified from application of a physically-based, two-dimensional model that can simulate hydrological responses to rainfall in soft rock cliffs. Here we use as an example the soft rock cliffs of Suffolk, which have long-term retreat rates of between 3 and 4 m a−1 (Brooks & Spencer, 2010). Their geological composition and location is illustrated in Figure 6.5. The Suffolk cliffline is composed of Pliocene and early-mid Pleistocene marine deposits, generally termed Crag, overlying the Palaeogene and Cretaceous basement (Hamblin et al., 1997; Gibbard et al., 1998).

CH6 STORM IMPACTS ON CLIFFED COASTLINES

111

Figure 6.5 (A) Location and; (B) geological composition of soft rock cliff, Covehithe, Suffolk, UK.

The Crag deposits contain clays and silt-clays from the Pre-Pastonian/Baventian era, overlain by marine Pleistocene sand and gravels with thin layers of laminated silts (West, 1980) and thicker gravel lenses of the Westleton Beds (Hey, 1967). In places along the cliffline there is a pronounced and significant exposure of loamy diamicton which we interpret as decalcified Lowestoft Formation till (Anglian Stage).

112

COASTAL STORMS: PROCESSES AND IMPACTS

System 1 30

15

Slow percolation 3 Till

3 6 9 12

Depth (m)

10

18

Suction loss

15

Soft medium send

21 27 24 30

High suction maintained Weakly cemented medium sand Baventian fine sand with silt laminae Water-tabke soft medium sand Baventia (silty-clay)

5 Limited effect

0

–30

–25

–20

0 5 –15 –10 –5 Pore-water pressure (kPa)

10

15

System 2 15

Rapid percolation Depth (m)

10 3 6

9 12 15 18 21

5

Suction loss

Soft medium sand Basal debris (largely unconsolidated sand)

Limited effect

Water-table Baventian (silty clay) 0

–30

–25

–20

–15 –10 –5 0 5 Pore-water pressure (kPa)

10

15

System 3 15

Depth (m)

10 Rapid percolation

5

21 1512 9 6 18

3 Suction loss Soft medium sand Limited efffect Water-table Baventian (silty clay)

0 –30

–25

–20

–15 –10 –5 0 5 Pore-water pressure (kPa)

10

15

Figure 6.6 Modelled outputs showing changing negative pore water pressure (suction) responses in three cliff types of contrasting geological composition. Numbers refer to days since start of simulated rainfall.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

113

Within the geology of the Suffolk cliffs, the main elements of importance are the layered structure of the cliffs, the existence of a S-N dipping clay base, which forms the shore platform, and the occurrence of Lowestoft Till at some cliff-top locations. GeoSlope Office was used to look at internal hydrological responses (for more detail on parameterisation and validation see Brooks et al., 2012). Modelled suction responses are shown in Figure 6.6 and show that the same rainstorm can deliver dramatically different results in cliff response depending on the geological composition and structure. Thus, with a capping of denser slowly-percolating till (System 1), there is gradual percolation and loss of suction that reaches a depth of 7 m below the cliff-top after 30 days, and development of a large zone of reduced suction providing greater potential for cliff-top failure. Hence there is likely to be a cumulative effect from successions of storms that deliver large (>30 mm) quantities of rainfall. In the absence of this stratum, more rapid infiltration and percolation through the soft sediments also facilitates suction loss (Systems 2 and 3) but not to the same extent or for the same length of time. Balancing these terrestrial dynamics with associated marine influences is important when considering spatial variability in the effect of storms on cliff retreat. Large storms readily remove unconsolidated beach material to expose the shore platform, in this example cut in the Baventian Clay, and notch the cliff base. The water level reached by the sea (still water level plus waves) needs to be assessed relative to cliff base elevation as that determines the marine role. Figure 6.7 shows water levels recorded in the Lowestoft tide gauge and wave heights recorded in the Southwold approaches wave buoy for an exceptionally stormy period from 7–11 November, 2010, when cliff retreat was highly variable alongshore. During this period high spring tides coincided with strong onshore winds that were sustained over several high tides. Waves of up to 3.5 m resulted. There was over 40 mm of rain in the period, generating suction loss in the cliffs. In the highest till-capped cliffs (System 1) there was complete removal of the beach, exposure of the shore platform and notched cliff bases developed, with associated undermining and collapse of the overlying layers by as much as 12 m inland. In the softer cliffs at lower elevations (System 2), the responses were less dramatic, with evidence of sediment redistribution on the beach itself and moderate retreat at the cliff-top. A model for the alongshore variation in response to storms for these soft sediment cliffs is suggested in Figure 6.8, in which the varying geological composition of the cliffs is thought to play a major role in dictating the mode and timing of cliff retreat.

Water level (m OD)

Water height (m)

1 0.5 0 –0.5 –1 –1.5 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00

3.5 3 2.5 2 1.5 1 0.5 0

Wave height (m)

Water level (m CD)

1.5

Figure 6.7 Waves recorded at Southwold approaches (http://www.cefas.defra.gov.uk/our-science/ observing-and-modelling/monitoring-programmes/wavenet.aspx) plotted alongside tidal variations at Lowestoft (http://www.ntslf.org/data/uk-network-real-time) to show coincident high still water levels and large waves.

114

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 6.8 Suggested retreat mechanisms for soft rock cliffs of varying geological composition, based on modelled and observed responses at Covehithe at different points alongshore.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

115

Spatial variation in cliff retreat is frequently observed along this cliffed coast and recent studies are providing growing evidence for the complex nature of cliff response to storms (Brooks et al., 2012; Brooks & Spencer, 2014). The example provided above illustrates this alongshore variation and suggests that, even for the largest storms, cliff response can be strongly controlled by geological composition and the way it affects the balance between marine and terrestrial processes. Both domains play an important part in determining spatial variation in cliff retreat and combine in different ways. Accompanying cliff retreat is an alteration in the stratigraphy of the glacial sediments, especially where retreat rates are high, which introduces greater complexity. A further important issue is the interplay between shoreline retreat (specifically erosion) and deposition of the released sediment. Soft sediment cliffs that reach high (>10 m) elevations and extend alongshore for distances of kilometres release considerable volumes of sediment (Cambers, 1976; Brooks & Spencer, 2010; Montreuil & Bullard, 2012), which can be used to build beaches (Young et al., 2009) and develop nearshore and offshore structures (Zhou et al., 2014). These in turn offer shoreline protection either directly, in the case of beaches at the cliff base or indirectly, in the case of bathymetric shallowing leading to a reduction in wave height (Stansby et al., 2006). ‘Switching on’ and ‘switching off‘ behaviour has been observed where shorelines retreat rapidly into ground of changing elevation. Retreat across ground of increasing elevation enables the sediment supply system to accelerate to higher volumes released, while retreat into areas of decreasing land elevation slows down the sediment release; indeed, the supply can ‘switch off’ completely. This can drive episodicity in cliff retreat depending on the level of development of protective structures from the sediment. Hansen and Barnard (2010) link shoreline change to various wave and topographic/bathymetric parameters to demonstrate that while storm impacts are important in the short term, over historic periods it is changes in the bathymetry that largely govern shoreline change. The role of bathymetric change in governing cliff response to storms is deserving of far greater attention. Pye and Blott (2006) have cited the development of the Dunwich-Sizewell Bank, Suffolk, UK as a major factor in the reduction of storm impacts and cliff retreat for the Dunwich cliffs, which lie to landward of the bank. Storms therefore play a direct role in ensuring temporal variability in cliff retreat as they cross erosional thresholds and provide the power to drive the retreat process, but also have an indirect effect by influencing sediment supply, sediment redistribution and the temporal variability in the size and location of natural protective structures.

6.5

Modelling shoreline retreat for cliffed coasts and the incorporation of storminess

Over the past decade there has been a proliferation in the development and application of modelling approaches to shoreline change, spurred on by a growing awareness of sea-level rise (Woodworth et al., 2009; Wahl et al., 2013) and the need to plan for the future beyond 2100. Modelling cliff retreat presents specific challenges because, as shown in the previous section, it is often driven by high magnitude, low frequency storms and these provide few and unpredictable opportunities for data collection for model parameterisation, calibration, verification and validation. Cliff retreat is

116

COASTAL STORMS: PROCESSES AND IMPACTS

also a highly complex process involving interfacing of eroding geological structures with highly dynamic beach and offshore depositional structures. One of the earliest approaches to modelling shoreline retreat was the development of the Bruun Rule, formulated under the assumption that all sediment remains within the active profile and is therefore only applicable to low-lying shores (Bruun, 1988). Criticisms of the model are based around the observation that it frequently underestimates retreat, sometimes by more than an order of magnitude (e.g. Cooper & Pilkey, 2004; Ranasinghe & Stive, 2009). However this approach has been modified for cliffs with the introduction of terms for cliff elevation (B) and the proportion of sediment remaining within the active profile (P). This overcomes some of the earlier restrictions (Weggel, 1979; Hands, 1983). For cliffs of significant elevation, Dean (1991) has produced a modified Bruun Rule as follows: ( ) S2 − S1 L∗ R2 = R1 + ( ) P B + h∗ where: R2 = future retreat rate (m a−1 ) R1 = historic retreat rate (m a−1 ) S1 = historic sea-level rise (mm a−1 ) S2 = future sea-level rise (mm a−1 ) L∗ = length of active cross-shore profile (m) P = proportion of sediment remaining within the active profile (% sand and gravel) B = cliff elevation (m) h∗ = depth of closure (m) Known sea-level rise in past periods, along with the associated retreat, can be included as the historic parameters, while projections of sea-level rise in the future are included to quantify future retreat rates. Bray and Hooke (1997) describe the model of Dean as being ‘the most easily applied and realistic adaptation of the Bruun Rule for eroding cliffs’ but there are some remaining restrictions. One in particular involves the depth of closure (h∗ ), which is the water depth offshore of which there is no temporal change (daily, seasonal or over longer periods) in geometry. Thus, this depth represents the point of transition from the mobile beach to the stable offshore zone (Hallermeier, 1981). Nicholls et al. (1998) argue that this term is not independent of timescale because when more extreme events (and thus greater wave heights) are incorporated over long timescales a large range of closure depths is likely to result. The Sunamura model (Sunamura, 1988) offers a somewhat different approach for modelling retreat of cliffed shorelines with no dissipative beach or shoreface sediment layer. The model was derived for cliffs retreating under high wave power and material strength (as might be termed ‘hard rock’ cliffs (Naylor et al., 2010)). In this model the second use of R1 includes the long-term joint effect of both wave power and material strength as follows: ) ( S2 − S1 R2 = R1 + h ∗∕(R +L ) 1 ∗ The calibrated value of R1 reflects the long-term combined influence of wave power and material strength on shoreline retreat and simplifies the data required to parameterise

CH6 STORM IMPACTS ON CLIFFED COASTLINES

117

the model. However, this necessitates the assumption that there is no change in wave power or material strength through time (Bray & Hooke, 1997), which is not always the case. Since these early cliff response models there has been the development of probabilistic approaches (allowing parameter and output uncertainty to be evaluated) that include a much greater range of forcing factors (Hall et al., 2002). SCAPE (Soft Cliff And Platform Erosion) is widely used for soft rock cliffs where retreat is highly episodic, driven by cliff base erosion during storms and mass movements in the cliff (Walkden & Hall, 2005). This model has at its heart the shore platform as the regulator of retreat. Originally formulated and tested for soft rock cliffs (glacial tills, clays and mixed sands and gravels) that occur widely in eastern England, the model has been simplified for cliffs with low beach volumes (< 30 m3 m−1 ) through identification of parameter redundancy (Walkden & Dickson, 2008). The simple form of the model is: √ R2 = R1

S2 S1

This is very similar to the historic extrapolation of Leatherman (1990) in which shoreline retreat takes place in direct proportion to the acceleration in sea-level rise. This formulation of the SCAPE model was found to provide the closest approximation to cliff retreat along rapidly retreating soft rock cliffs in Suffolk, UK (Brooks et al., 2012) and was therefore used to predict the position of future clifflines under rising sea level. A similar approach involving SCAPE and other model formulations was adopted for the Holderness cliffs of the UK (Castedo et al., 2015). Importantly a methodology has been developed to estimate alongshore extent and elevation of emerging clifflines in order to provide predictions of sediment output under future shoreline retreat. Future sediment output is likely to be orders of magnitude higher than previous estimates have suggested, which is important in addressing the issue raised earlier on the link between the sediment supply from cliff retreat and the development of offshore and nearshore bathymetry.

6.6

Future storm impacts on clifflines under accelerated sea-level rise and changing storminess

The modelling approaches described in the previous section are not just concerned with how shorelines and cliffs respond to storms and sea-level rise, but are fundamentally aimed at predicting the changing future rate of retreat as these drivers change. Cliff retreat during high magnitude storms is dependent on the combination of rainfall, plus still water level and wave height, the period of time for which these components remain high as well as their interaction (Carter & Stone, 1989; Brooks et al., 2012). We know that these components combine in different ways to drive cliff retreat. Large waves coinciding with high spring tides have significant impacts on cliffs, as exemplified by the 7–11 November 2010 storms on the Suffolk coast, UK. But equally, significantly elevated still water levels during storm surges can be just as effective even when lower

118

COASTAL STORMS: PROCESSES AND IMPACTS

wave heights are involved, as exemplified by the 5–6 December, 2013 storm surge on the East Anglian coast (Spencer et al., 2015). Lessons can be learnt from contemporary occurrence of high magnitude storms. We are in an unprecedented situation in being able to gather detailed and accurate data to capture the combination of waves, still water levels and rainfall and link this to cliff response. This is well exemplified by the storms during the winter 2013–2014 that affected large areas of the UK, including both hard rock cliffs in the west and soft rock cliffs in the east. The period was one of the most energetic since the 1950s. During that winter a powerful jet stream was responsible for a long succession of low pressure systems across the Atlantic Ocean (Wallace et al., 2014). The first storm of the winter resulted in an intense easterly-tracking cyclone, with Beaufort Force 9 (strong gale) to 11 (violent storm) winds, on 5 December. Around the shores of the southern North Sea intense low pressure and strong winds produced a storm with surge residuals reaching more than 2 m. Still water levels were able to reach over 5 m ODN in places because the maximum surge residual coincided with the time of high water. Waves of 2 m (not significantly high) acted on the still water level and generated 10 m of cliff retreat in soft sediment cliffs, with cliff retreat up to four times the long-term annual average. Similarly, extreme waves were experienced in the Atlantic during that winter, driven onshore by westerly propagation of storms and affecting hard rock cliffs in the west of the UK. The period 31 January–6 February 2014 enabled the impacts of storms on hard rock cliffs to be uniquely captured in situ and using sophisticated remote instrumentation (Earlie et al., 2015). Again, for hard rock cliffs, the impacts were one or two orders of magnitude greater than the long-term mean, either expressed as retreat, vertical ground movement or cliff face loss. Hence the most recent research highlights importance of storms under extreme conditions in producing large responses in cliffs that lie at opposing ends of the cliff strength spectrum. Using historical information to calibrate a shoreline response model for future prediction is one approach to shoreline modelling that is largely tied to sea-level rise. Indeed most modelling approaches are focussed on sea-level rise even though there are numerous examples of the significant role of storms in driving cliff retreat. Changing storminess will accompany future sea-level rise, although the precise future role of enhanced storminess is not thought to be as significant as future sea-level rise. Nonetheless modelling approaches need to include both drivers of change as their combined influence dictates shoreline retreat. Recently, storm impacts have been modelled using the concept of accumulated excess energy (Hackney et al., 2013). This approach arises from the observation that the key factor controlling both the process as well as the rate of cliff retreat is the action of waves (Sunamura, 1992; Trenhaile, 2009). In this approach, the accumulated excess energy enables the effect of multiple storms to be assessed (periods of storminess) developing from earlier attempts to model storm impacts based around peak wave energy (Quinn et al., 2010), which consider individual events and only provide at most 62% predictive power (Robinson, 1977; Amin & Davidson-Arnott, 1997). The AEE model focuses on event duration above a given erosion threshold, with longer duration resulting in greater coastal erosion for a given magnitude (Ciavola et al., 2007; Darby et al., 2010; Armaroli et al., 2012). The AEE model has a predictive power of between 46% and 89% for retreat in the Greensand and Gault Clay cliffs of the Isle of Wight and for the soft cliffs of the Suffolk coast, respectively. The model has been used to provide future predictions of cliff retreat

CH6 STORM IMPACTS ON CLIFFED COASTLINES

119

using output from a downscaled HadCM3 Global Climate model. Results suggest that over the next century changes in sea level and wave height will generate increases in cliff retreat of 0.5 m a−1 for the Atlantic cliffs of the Isle of Wight and 0.3 m a−1 for the North Sea cliffs of Suffolk. This is a useful approach as in future the combination of accelerated sea-level rise and changing storminess will drive cliff retreat. Robust modelling and accurate data to drive the models will strengthen our ability to plan for change.

6.7

Conclusions

Cliffs lie on a spectrum from the most resistant, massive rocky structures to the highly erodible soft sediments and oversteepened dunes made of unconsolidated non-cohesive sediment. Individual extreme storms can be responsible for long-term cliff change. Future storm impacts will depend on the rate of accelerated sea-level rise as this sets the base level to ever-increasing elevations, enabling storms to become more damaging. While sea-level rise estimates have been recently revised in the IPCC Fifth Assessment Report (Church et al., 2013) and suggest that global mean sea level will rise between 26 to 55 cm (for the low emissions scenario RCP2.6), 32 to 63 cm (for medium emissions RCP4.5) and 33 to 63 cm (for high emissions RCP6) for 2081−2100 relative to 1986–2005, data concerning future storminess remains very inconclusive. No unequivocal evidence for systematic long-term changes in storminess is reported in either the IPCC Special Report on extremes (Field et al., 2012) or the IPCC Fifth Assessment Report (Kovats & Valentini, 2014). Furthermore for the period 1950–2000, UK storm frequency was dominated by natural variability rather than any kind of systematic change (Allan et al., 2009). Nevertheless the main focus for coastal management in future needs to be on linking storm impacts on rapidly retreating soft rock cliffs with sediment delivery and the concurrent growth in natural structures that can offer future shoreline protection. Sediment is vital if such structures are to keep pace with sea-level rise and remain resilient against storms that will increasingly act at ever-higher elevations. Estuarine and coastal ecosystem (ECE) protection (sand bars, beaches, seagrass beds, salt marshes, mangroves and sand dunes) provide a vital role in wave attenuation, and shoreline management needs to develop a holistic framework as sea-level rises over the coming century. Understanding cliff response to storm impacts is a vitally important component of a much wider picture.

Acknowledgements The writing of this paper was facilitated by a Leverhulme Research Fellowship awarded to the lead author in 2015. This paper is a contribution to EU FP7 Collaborative Project Resilience-Increasing Strategies for Coasts – toolkit (RISC_KIT).

References Adams, P.N., Storlazzi, C.D. & Anderson, R.S. (2005) Nearshore wave-induced cyclical flexing in sea cliffs. Journal of Geophysical Research, 110, F02002, doi:10.1029/2004JF000217.

120

COASTAL STORMS: PROCESSES AND IMPACTS

Allan, J. (2006) Extreme storms, El Niños, and sea-level rise due to earth’s changing climate: The changing face of the Oregon coast. Available at: http://www.oregongeology.org/sub/projects/ ccig/OR_ccig_mtg_072706.pdf Allan, J.C. & Komar P.D. (2002) Extreme storms on the Pacific Northwest Coast during the 1997–98 El Niño and 1998–99 La Niña. Journal of Coastal Research, 18, 175–193. Allan, J.C. & Komar, P.D. (2006) Climate controls on US West Coast erosion. Journal of Coastal Research, 22 (3), 511–529. Allan, R., Tett S. & Alexander, L. (2009) Fluctuations in autumn – winter severe storms over the British Isles: 1920 to present. International Journal of Climatology, 29, 357–371. Amin, S.M.N. & Davidson-Arnott, R.G. (1997) A statistical analysis of the controls on shoreline erosion rates, Lake Ontario. Journal of Coastal Research, 13, 1093–1101. Anderson, M.G., Lloyd, D.M. & Othman, A. (1994) Using a combined slope hydrology/slope stability model for cut slope design in the Tropics. Malaysian Journal of Tropical Geography, 25, 1–10. Anderson, M.G., Collison, A.J.C., Hartshorne, J., Lloyd, D.M. & Park, A. (1996) Developments in slope hydrology – stability modelling for tropical slopes. In: M.G. Anderson & S.M. Brooks (Eds) Advances in Hillslope Processes. John Wiley and Sons Ltd., Chichester, pp. 799–821. Armaroli, C., Ciavola, P., Perini, L., Calabrese, L., Lorito, S., Valentini, A., et al. (2012) Critical storm thresholds for significant morphological changes and damage along the Emilia-Romagna coastline, Italy. Geomorphology, 143–144, 34–51. Armaroli, C., Grottoli, E., Harley, M.D. & Ciavola, P. (2013) Beach morphodynamics and types of foredune erosion generated by storms along the Emilia-Romagna coastline, Italy. Geomorphology, 199, 22–35. Bray, M.J. & Hooke, J.M. (1997) Prediction of soft-cliff retreat with accelerating sea-level rise. Journal of Coastal Research, 13 (2), 453–467. Brooks, S.M. & Anderson, M.G. (1995) The determination of suction controlled slope stability in humid-temperature environments. Geografiska Annaler, 77A, 11–22. Brooks, S.M. & Spencer, T. (2010) Temporal and spatial variations in recession rates and sediment release from soft rock cliffs, Suffolk coast, UK. Geomorphology, 124 (1–2), 26–41. Brooks, S.M. & Spencer, T. (2012) Shoreline retreat and sediment release in response to accelerating sea-level rise: Measuring and modelling cliffline dynamics on the Suffolk Coast, UK. Global and Planetary Change, 80–81, 165–179. Brooks, S.M. & Spencer, T. (2014) Importance of decadal scale variability in shoreline response: Examples from soft rock cliffs, East Anglian coast, UK. Journal of Coastal Conservation: Policy and Management, 18, 581–593. Brooks, S.M., Crozier, M.J., Preston, N.J. & Anderson, M.G. (2002) Regolith stripping and the control of shallow translational hillslope failure: Application of a two-dimensional coupled soil hydrology-slope stability model, Hawke’s Bay, New Zealand. Geomorphology, 45, 165–179. Brooks, S.M., Crozier, M.J., Glade, T.W. & Anderson, M.G. (2004) Towards establishing climatic thresholds for slope stability. Pure and Applied Geophysics, 161, 881–905. Brooks, S.M., Spencer, T. & Boreham, S. (2012) Mechanisms for cliff retreat in rapidly receding soft-rock cliffs: Marine and terrestrial influences, Suffolk coast, UK. Geomorphology, 153–154, 48–60. Brunsden, D. & Chandler J.M. (1996) Development of an episodic landform change model based upon the Black Ven Mudslide, 1946–1995. In: M.G. Anderson & S.M. Brooks, S.M. (Eds) Advances in Hillslope Processes. John Wiley and Sons Ltd, Chichester.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

121

Brunsden, D. & Jones, D.K.C. (1976) The evolution of landslide slopes in Dorset. Philosophical Transactions of the Royal Society of London, A283, 605–631. Brunsden, D. & Lee, E.M. (2004) Behaviour of coastal landslide systems: An interdisciplinary view. Zeitschrift für Geomorphologie, 134, 1–112. Bruun, P. (1988) The Bruun Rule of erosion by sea-level rise: A discussion on large scale twoand three-dimensional usages. Journal of Coastal Research, 4 (4), 627–648. Burningham, H. & French, J. (2012) Is the NAO winter index a reliable proxy for wind climate and storminess in northwest Europe? International Journal of Climatology, 33, 2036–2049. Cambers, G. (1976) Temporal scales in coastal erosion systems. Transactions of the Institute of British Geographers, 1 (2), 246–256. Campbell, R.H. (1975) Soil slips, debris flows and rainstorms in the Santa Monica Mountains and vicinity, Southern California. United States Geological Survey Professional Paper, 851. Carter, R.W.G. & Stone, G. (1989) Mechanisms associated with the erosion of sand dune cliffs, Magilligan, Northern Ireland. Earth Surface Processes and Landforms, 14, 1–10. Castedo, R., Murphy, W., Lawrence, J. & Paredes, C. (2012) A new process-response coastal recession model of soft rock cliffs. Geomorphology, 177, 128–143. Castedo, R., de la Vega-Panizo, R., Fernández-Hernández, M. & Paredes, C. (2015) Measurement of historical cliff-top changes and estimation of future trends using GIS data between Bridlington and Hornsea – Holderness Coast (UK). Geomorphology, 230, 146–160. Church, J.A., Clark, P.U., Cazenave, A. Gregory, J.M., Jevrejeva, S., Levermann, A., et al. (2013) Sea Level Change. In: T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, et al. (Eds) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Ciavola, P., Armaroli, C., Chiggiato, J., Valentini, A., Deserti, M., Perini, L., et al. (2007) Impact of storms along the coastline of Emilia-Romagna: The morphological signature of the Ravenna coastline (Italy). Journal of Coastal Research, 50, 540–544. Collins, B.D. & Sitar, N. (2008) Processes of coastal bluff erosion in weakly lithified sands, Pacifica, California, USA. Geomorphology, 97 (3–4), 483–501. Committee on Sea-level rise in California, Oregon and Washington (2012) Sea-level rise for the coasts of California, Oregon, and Washington: Past, present, and future. The National Academies Press, Washington, D.C. Cooper, J.A.G. & Pilkey, O.H. (2004) Sea-level rise and shoreline retreat: Time to abandon the Bruun Rule. Global and Planetary Change, 43 (3–4), 157–171. Darby S.E., Trieu H.Q., Carling, P.A., Sarkkula, J., Koponen J., Kummu, M., et al. (2010) A physically-based model to predict hydraulic erosion of fine-grained river banks: The role of form roughness in limiting erosion. Journal of Geophysical Research, 115, F04003, doi:10.1029/2010JF001708. Dawson, S., Smith, D.E., Jordan, J. & Dawson, A.E. (2004) Late Holocene coastal sand movements in the Outer Hebrides, N.W. Scotland. Marine Geology, 210, 281–306. Dean, R.G. (1991) Equilibrium beach profiles: Characteristics and applications. Journal of Coastal Research, 7 (1), 53–84. Dornbush, U., Robinson, D.A., Moses, C.A. & Williams, R.B.G. (2011) Temporal and spatial variations of chalk cliff retreat in East Sussex, 1873 to 2001. Marine Geology, 249, 271–282.

122

COASTAL STORMS: PROCESSES AND IMPACTS

Drake, A.R. & Phipps, P.J. (2006) Cliff recession and behaviour studies, Hunstanton, UK. Proceedings of the Institution of Civil Engineers Maritime Engineering, 160, 3–17. Earlie, C.S., Young, A.P., Masselink, G. & Russell, P.E. (2015) Coastal cliff ground motions and response to extreme storm waves. Geophysical Research Letters, 42, 847–854. Emery, K.O. & Kuhn, G.G. (1982) Sea cliffs: Their processes, profiles and classification. Geological Society of America Bulletin, 93, 644–654. Enfield, D.B. & Allen, J.S. (1980) On the structure and dynamics of monthly mean sea level anomalies along the Pacific Coast of North and South America. Journal of Physical Oceanography, 10, 557–578. Field, C.B., Barros, V., Stocker, T.F., Dahe, Q., Dokken, D.J., Ebi, K.L., et al. (2012) Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Flick, R.E. (1998) Comparison of tides, storm surges, and mean sea level during the El Niño winters of 1982–83 and 1997–98. Shore & Beach, 66 (3), 7–17. French, P. (2001) Coastal Defences: Processes, Problems and Solutions. Routledge, London. Gibbard, P.L., Zalasiewicz, J.A. & Mathers, S.J. (1998) Stratigraphy of the marine Plio-Pleistocene crag deposits of East Anglia. Mededelingen Nederlands Instituut voor Toegepaste Geowetenschappen, TNO 60, 239–262. Gofar, N. & Lee, L.M. (2008) Extreme rainfall characteristics for surface slope stability in the Malaysian Peninsular. Georisk Assessment and Management of Risk for Engineered Systems and Geohazards, 2, 65–78. Graham, N.E. & Diaz, H.F. (2001) Evidence for intensification of North Pacific wintercyclones since 1948. Bulletin of the American Meteorological Society, 82, 1869–1893. Gray, J.M. (1988) Coastal cliff retreat at the Naze, Essex since 1874: Patterns, rates and processes. Proceedings of the Geologists’ Association, 99, 335–338. Griggs, G.B. & Patsch, K.B. (2004) California’s coastal cliffs and bluffs. In: M.A. Hampton & G.B. Griggs (Eds) Formation, Evolution and Stability of Coastal Cliffs – Status and Trends. USGS Professional Paper 1693, pp. 53–64. Hackney, C., Darby, S.E. & Leyland, J. (2013) Modelling the response of soft cliffs to climate change: A statistical, process-response model using accumulated excess energy. Geomorphology, 187, 108–121. Hall, J.W., Meadowcroft, I.C., Lee, E.M. & van Gelder, P. (2002) Stochastic simulation of episodic soft coastal cliff recession. Coastal Engineering, 46 (3), 159–174. Hall, A.M., Hansom, J.D., Williams, D.M. & Jarvis, J. (2006) Distribution, geomorphology and lithofacies of cliff-top storm deposits: Examples from the high-energy coasts of Scotland and Ireland. Marine Geology, 232, 131–155. Hallermeier, R.J. (1981) A profile zonation for seasonal sand beaches from wave climate. Coastal Engineering, 4 (3), 253–277. Hamblin, R.J.0., Moorlock, B.S.P., Booth, S.J., Jeffrey, D.H. & Morigi, A.N. (1997) The Red Crag and Norwich Crag formations in eastern Suffolk. Proceedings of the Geologists’ Association, 108, 11–23. Hampton, M.A. & Dingler, J. (1998) Short-term evolution evolution of three coastal cliffs in San Mateo County, California. Shore and Beach, 66, 24–30. Hands, E.B. (1983) The Great Lakes as a test model for profile responses to sea-level changes. In: P.D. Komar (Ed.) Handbook of Coastal Processes and Erosion. Boca Raton, Florida: CRC Press, pp. 176–189.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

123

Hansen, J.E. & Barnard, P.L. (2010) Sub-weekly to interannual variability of a high-energy shoreline. Coastal Engineering, 57 (11–12), 959–972. Hansom, J.D. & Hall, A.M. (2009) Magnitude and frequency of extra-tropical North Atlantic cyclones: A chronology from cliff-top storm deposits. Quaternary International, 195, 42–52. Hansom, J.D., Barltrop, N.D.P. & Hall, A.M. (2008) Modelling the processes of cliff-top erosion and deposition under extreme storm waves. Marine Geology, 253, 36–50. Hapke, C.J., Reid, D. & Richmond, B. (2009) Rates and trends of coastal change in California and the regional behaviour of the beach and cliff system. Journal of Coastal Research, 25 (3), 603–615. Harley, J.B. (1972) Maps for the local historian: A guide to the British sources. National Council of Social Service for the Standing Conference for Local History. Hey, R.W. (1967) The Westleton Beds reconsidered. Proceedings of the Geologists’ Association, 87, 69–82. Hurrell, J.V. (1995) Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science, 269, 676–679. Hutchinson, J.N. (1970) A coastal mudflow on the London Clay cliffs at Beltinge, North Kent. Geotechnique, 20, 412–438. Komar, P.D. (2004) Oregon’s coastal cliffs: Processes and erosion impacts. In: M.A. Hampton & G.B. Griggs (Eds) Formation, Evolution and Stability of Coastal Cliffs – Status and Trends. USGS Professional Paper 1693, pp. 65–80. Komar, P.D. & Allan, J.C. (2002) Nearshore-process climates related to their potential for causing beach and property erosion. Shore & Beach, 70 (3), 31–40. Komar, P.D. & Shih, S-M. (1993) Cliff erosion along the Oregon coast: A tectonic-sea level imprint plus local controls by beach processes. Journal of Coastal Research, 9, 747–765. Komar, P.D., Allan, J.C. & Ruggiero, P. (2011) Sea level variations along the US Pacific Northwest coast: Tectonic and climate controls. Journal of Coastal Research, 27 (5), 808–823. Kovats, S. & Valentini, R. (2014) Chapter 23. Europe. In: C.B. Field, V. Barros, T.F. Stocker, Q. Dahe, D.J. Dokken, K.L. Ebi, et al. (Eds) Climate Change 2014: Impacts, Adaptation and Vulnerability Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. Kuhn, G.G. & Shepard, F.P. (1984) Sea cliffs, beaches, and coastal valleys of San Diego County: Berkeley, California. University of California Press. Lahouse, P. & Pierre, G. (2003) The retreat of chalk cliffs at Cape Blanc-Nez (France): Autopsy of an erosional crisis. Journal of Coastal Research, 19 (2), 431–440. Leatherman, S.P. (1990) Modeling shore response to sea-level rise on sedimentary coasts. Progress in Physical Geography, 14, 447–464. Lee, E.M. (2008) Coastal cliff behaviour: Observations on relationship between beach levels and recession rates. Geomorphology, 101, 558–571. Lee, E.M., Hall, J.W. & Meadowcroft, I.C. (2001) Coastal cliff recession: The use of probabilistic prediction methods. Geomorphology, 40, 253–269. Lee, M.L., Gofar, N. & Rahardjo, H. (2009) A simple model for preliminary evaluation of rainfall-induced slope instability. Engineering Geology, 108 (3–4), 272–285. Lim, M., Rosser, N.J., Petley, D.N. & Keen, M. (2011) Quantifying the controls and influence of tide and wave impacts on coastal rock cliff erosion. Journal of Coastal Research, 27, 46–56. Lu, N. & Griffiths, D.V. (2004) Profiles of steady-state suction stress in unsaturated soils. Journal of Geotechnical and Geoenvironmental Engineering, 130, 1063–1076.

124

COASTAL STORMS: PROCESSES AND IMPACTS

Montreuil, A-L. & Bullard, J.E. (2012) A 150-year record of coastline dynamics within a sediment cell: Eastern England. Geomorphology, 179, 168–185. Moore, L.J. (2000) Shoreline mapping techniques. Journal of Coastal Research, 16, 111–124. Moore, L.J., Benumof, B.T. & Griggs, G.B. (1999) Coastal erosion hazards in Santa Cruz and San Diego Counties, California. Journal of Coastal Research Special Issue, 28, 121–139. Naylor, L.A., Stephenson, W.J. & Trenhaile, A.S. (2010) Rock coast geomorphology: Recent advances and future research directions. Geomorphology, 114 (1–2), 3–11. Nicholls, R.J., Birkemeier, W.A. & Lee, G.-H. (1998) Evaluation of depth of closure using data from Duck, NC, USA. Marine Geology, 148, 179–201. Priest, G.R. (1999) Coastal shoreline change study northern and central Lincoln County, Oregon. Journal of Coastal Research, Special Issue, 28, 140–157. Pye, K. & Blott, S.J. (2006) Coastal processes and morphological change in the Dunwich-Sizewell area, Suffolk, UK. Journal of Coastal Research, 22 (3), 453–473. Quinn, J.D., Rosser, N.J., Murphy, W. & Lawrence J.A. (2010) Identifying the behavioural characteristics of clay cliffs using intensive monitoring and geotechnical numerical modelling. Geomorphology, 120, 107–122. Ranasinghe, R. & Stive, M.J.F. (2009) Rising seas and retreating coastlines. Climatic Change, 97, 465–468. Robinson, L.A. (1977) Marine erosive processes at the cliff foot. Marine Geology, 23, 257–271. Rosser, N.J., Brain, M.J., Petley, D.N., Lim, M. & Norman, E.C. (2013) Coastline retreat via progressive failure of rocky coastal cliffs. Geology, 41, 939–942. Rosser, N.J., Lim, M., Petley, D.N., Dunning, S. & Allison, R.J. (2007) Patterns of precursory rockfall prior to slope failure. Journal of Geophysical Research, 112, F04014. doi: 10.1029/2006JF000642. Ruggiero, P. (2013) Is the intensifying wave climate of the US Pacific Northwest increasing flooding and erosion risk faster than sea-level rise? Journal of Waterway, Port, Coastal, Ocean Engineering, 139 (2), 88–97. Rulon, J.J. & Freeze, R.A. (1985) Multiple seepage faces on layered slopes and their implications for slope stability analysis. Canadian Geotechnical Journal, 22, 347–356. Rulon, J.J., Rodway, R. & Freeze, R.A. (1985) The development of multiple seepage faces on layered slopes. Water Resources Research, 21, 1625–1636. Sallenger, A.H., Krabill, W., Brock, J., Swift, R., Manizade, S. & Stockdon, H. (2002) Sea-cliff erosion as a function of beach changes and extreme wave run-up during the 1997–1998 El Nino. Marine Geology, 187 (3), 279–297. Shepard, F.P. & Grant, U.S. IV,. (1947) Wave erosion along the southern California coast. Bulletin of the Geological Society of America, 58, 919–926. Shih, S-M. & Komar, P.D. (1994) Sediments, beach morphology and sea cliff erosion within an Oregon Coast littoral cell. Journal of Coastal Research, 10, 144–157. Spencer, T., Brooks, S.M., Möller, I., Evans, B. & Tempest, J. (2015) Improving understanding of the geomorphic impacts of major storm surges: Southern North Sea Event of 5 December 2013. Earth Science Reviews, 146, 120–145. Stansby, P., Kuang C.-P., Laurence, D. & Launder, B. (2006) Sandbanks for coastal protection: Implications of sea-level rise. Part 1: Application to East Anglia. Tyndall Centre for Climate Change Research, Working Paper 86. Storlazzi, C. & Griggs, G.B. (2000) The influence of El Niño – Southern Oscillation (ENSO) events on the evolution of central California’s shoreline. Geological Society of America Bulletin, 112 (2), 236–249.

CH6 STORM IMPACTS ON CLIFFED COASTLINES

125

Storlazzi, C., Willis, C.M. & Griggs, G.B. (2000) Comparative impacts of the 1982–82 and 1997–98 El Niño winters on the central California coast. Journal of Coastal Research, 16, 1022–1036. Sunamura, T. (1988) Projection of future coastal cliff recession under sea-level rise induced by the greenhouse effect: Nii-jima Island, Japan. Transactions of the Japan. Geomorphological Union, 9 (1), 17–33. Sunamura, T. (1992) Geomorphology of Rocky Coasts. John Wiley and Sons, Chichester. Thieler, E.R., Himmelstoss, E.A., Zichichi, J.L. & Ergul, A. (2009) Digital Shoreline Analysis System (DSAS) version 4.0 – An ArcGIS extension for calculating shoreline change. United States Geological Survey Open-File Report 2008, 1278. Trenhaile, A.S. (1987) The Geomorphology of Rock Coasts. Oxford University Press, Oxford. Trenhaile, A.S. (2009) Modelling the erosion of cohesive clay coasts. Coastal Engineering, 56, 59–72. USGS (2013) Coastal erosion along the US West Coast during the 1997 98 El Niño: Expectations and observations. At: http://coastal.er.usgs.gov/lidar/AGU_fall98/index.html Wahl, T., Haigh, I.D., Woodworth, P.L., Albrecht, F., Dillingh, D., Jensen, J., et al. (2013) Observed mean sea level changes around the North Sea coastline from 1800 to the present. Earth-Science Reviews, 124, 51–67. Walkden, M. & Dickson, M. (2008) Equilibrium erosion of soft rock shores with a shallow or absent beach under increased sea-level rise. Marine Geology, 251 (1–2), 75–84. Walkden, M.J.A. & Hall, J.W. (2005) A predictive mesoscale model of the erosion and profile development of soft rock shores. Coastal Engineering, 52 (6), 535–563. Wallace, J.M., Held, I.M., Thompson, D.W.J., Trenberth, K.E. & Walsh, J.E. (2014) Global warming and winter weather, Science, 343, 729–730. Weggel, R.J. (1979) A method for estimating long-term erosion rates from a long-term rise in water level. Coastal Engineering Technical Aid, 79–2, CERC. West, R.G. (1980) The Pre-glacial Pleistocene of the Norfolk and Suffolk Coasts. Cambridge University Press, Cambridge. Wilkinson, P.L., Anderson, M.G. & Lloyd, D.M. (2002) An integrated hydrological model for slope stability. Earth Surface Processes and Landforms, 27, 1267–1283. Woodworth, P.L., Terferle, F.N., Bingley, R.M., Shennan, I. & Williams, S.D.P. (2009) Evidence for the accelerations of sea level on multi-decade and century timescales. International Journal of Climatology, 29 (6), 777–789. Young, A.P. & Ashford, S.A. (2006) Application of Airborne LIDAR for Seacliff Volumetric Change and Beach-Sediment Budget Contributions. Journal of Coastal Research, 22, 307–318. Young, A.P., Guza, R.T., Flick, R.E., O’Reilly, W.C. & Gutierrez, R. (2009) Rain, waves and short-term evolution of composite seacliffs in southern California. Marine Geology, 267, 1–7. Young, A.P., Guza, R.T., O’Reilly, W.C., Flick, R.E. & Gutierrez, R. (2011) Short-term retreat statistics of a slowly eroding coastal cliff. Natural Hazards Earth Systems Science, 11, 205–217. Zhou, L., Liu, J., Saito, Y., Zhang, Z., PlaChu, H. & Hu, G. (2014) Coastal erosion as a major sediment supplier to continental shelves: Example from the abandoned Old Huanghe (Yellow River) delta. Continental Shelf Research, 82, 43–59.

7 Storms in Coral Reefs Ana Vila-Concejo1 and Paul Kench2 1 Geocoastal 2 School

7.1

Research Group, School of Geosciences, The University of Sydney, NSW, Australia of Environment, University of Auckland, New Zealand

Introduction

Coral reefs are among the most iconic coastal-marine landscapes on Earth. Distributed throughout the tropical oceans, coral reefs occupy approximately 300,000 km2 of the Earth’s surface (Spalding et al., 2001). Despite their modest spatial extent, compared with other coastal systems, coral reefs sustain and protect millions of people in reef-associated coastal communities. Indeed, coral reefs are highly valued ecosystems as a consequence of their high biological diversity and the ecosystem services they provide (Best & Bornbusch, 2005). A critical element of these ecosystem services is the geomorphic value reefs afford. In particular, the physical structure of reefs provides habitat for coral communities, provides the physical foundation of habitable land in mid-ocean atolls and fringing reef systems, regulates oceanographic processes and the wave energy impacting coastlines, and may act as a buffer to coastal erosion. Cyclones (hurricanes and typhoons) and severe weather events are one of the major natural disturbance events on coral reefs and consequently have been a topic of global interest for many decades (Stoddart, 1971; Harmelin-Vivien, 1994). The direct physical effects of storms on reefs and their associated communities commonly record catastrophic impacts. Perhaps less well recognised and reported are the more passive reef responses, and the constructive effects of storms. As the largest biological constructions on Earth, coral reefs are dynamic geomorphological systems, that demonstrate a complex interplay between physical and biological processes (Woodroffe, 2003). As geomorphic features, coral reefs are three-dimensional wave-resistant structures (Done, 2011a) that consist of a veneer of living coral communities overlying vast sequences of previously deposited calcium carbonate. Across millennial timescales these structures produce a number of characteristic landform types, including atolls, barrier reefs, fringing reefs and reef platforms (Kench, 2013). These reef types vary in extent from less than 1 km2 , in the case of smaller patch reefs, to more than 100 km2 , with some reef networks forming barrier complexes, such as the Great Barrier Reef, which is more than 2400 km in length. In addition to gross structure, reef platforms also support a number of conspicuous sedimentary landforms constructed by the transport and accumulation of detrital sediment deposited by wave and current Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

128

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

(b)

Figure 7.1 Main shallow morphological elements in coral reefs: (a) Schematic coral reef cross-section; (b) Satellite image of the SE corner of One Tree Reef in the southern Great Barrier Reef showing the coral zones.

processes on, or adjacent to, coral reefs (Kench, 2013; Figure 7.1). Autochthonous sediments are mostly produced on the forereef and reef through coral growth and other benthic organisms such as foraminifera and calcareous algae. This sediment is then eroded and transported to the backreef through dynamic processes contributing to the sedimentary landforms (Davies & Kinsey, 1977; Davies & Marshall, 1980; Hopley et al., 2007). Such deposits include subtidal sand aprons, which contribute to lagoon filling and that cover on average 20% of the world’s reefs (Rankey & Garza-Pérez, 2012), and subaerial landforms such as islands, coastal plains and beaches (Figure 7.1). Of interest to the analysis of storm effects is the development and ongoing morphological adjustment of these deposits, which are geomorphically important at the human timescale as they form the foundation for coastal communities and provide the only habitable land in a number of mid-ocean atoll nations, such as Tuvalu, Kiribati and the Maldives. This chapter examines the effects of storm events on coral reefs and reef-associated landforms. It begins with a brief examination of the geomorphic units of reef systems,

CH7 STORMS IN CORAL REEFS

129

places the physical dynamics of reefs within an eco-morphodynamic framework and highlights the unique aspects of the interaction of waves and storm waves with reef systems that force ecological and geomorphic change. The focus of the chapter is on the event-scale effects of storm events. Storm effects are considered with respect to both the structure of coral reefs, which is further considered on the different morphological components, and sedimentary landforms. Storm driven effects are placed within the eco-morphodynamic framework that reflects the interaction of biological and physical processes. The chapter also highlights contemporary research questions in understanding the influence of storms on the current dynamics and future trajectories of coral reef structure and associated sedimentary landforms.

7.2

Geomorphic units of reefs

Reefs have three main morphological elements: forereef, reef and backreef, each with its own biota and sediment types (Collins, 2011) and following are the characteristics for each morphological region: •

•

Forereef (Figure 7.1): slopes seaward at steep gradient and is composed of reefbuilding organisms and reef debris. It can be further divided into three parts (Blanchon, 2011): ∘ A shallow gently sloping upper section exposed to incoming waves and that may show spurs and grooves which are finger-like formations consisting of a series of ridges (spurs) and channels (grooves) mostly perpendicular to the reef crest. ∘ A steeper section that can sustain coral growth as deep as 100 m. ∘ A forereef tallus apron where sediment and coral fragments accumulate at the toe of the reef. The reef crest is narrow-crested or flat-topped zone where calcareous reef framework builders are active, often forming an algal rim, but with little coral development (Blanchon, 2011). Reef flats (Figure 7.1) are the most recent expression of coral reef growth at sea level. Thornborough and Davies (2011) differentiate two types of reef flats: ∘ Coral-dominated, not linked to storms. ∘ Rubble-dominated, which are common features of exposed high-energy reefs. Reef width is considered one of the limiting controls on wave attenuation/dissipation, the wider the reef, the larger the wave attenuation. ∘ Coral-dominated flats (Figure 7.1b) have a clear zonation that includes a mostly bare zone (algal rim and rubble band); a live-coral zone (coral windrows and coral patches); and a sand apron which is a depositional zone (Thornborough & Davies, 2011). Formation and development of coral-dominated flats is not linked to storm events. ∘ Rubble-dominated flats (Figure 7.1b) are linked to high-energy incident waves. Their zonation is not as clear as for coral-dominated flats, but in general there is an algal rim in the ocean edge which is followed by a succession of rubble-covered areas where rubble size is maximum near the ocean and minimum near the lagoon (Thornborough & Davies, 2011). Rubble-dominated flats show a seaward gentle slope (Thornborough & Davies, 2011), where storm

130

COASTAL STORMS: PROCESSES AND IMPACTS

ridges (ramparts) can migrate lagoonwards and coalesce to form (windward) islands (Scoffin, 1993). • The backreef environment consists of coral communities and reef debris swept landward behind the reef, including sand aprons (Figure 7.1) to mud deposits in lagoonal environments. It is characterised by mostly low-energy conditions as waves dissipate their energy over the forereef and reef and is therefore a depositional environment.

7.2.1 Reefs as ecomorphodynamic structures The distribution of coral reefs is influenced by environmental factors such as light, sea-surface temperature and carbonate saturation state, being mostly limited to waters with sea-surface temperatures between 17–18∘ C and 33–34∘ C (Woodroffe, 2003). Coral reefs are unique coastal environments as their structure and contemporary morphology reflects the delicate balance between constructive ecological processes (organisms that produce calcium carbonate (CaCO3 ) sediments from their skeletal remains (e.g. corals, calcareous algae molluscs and foraminifera), and destructive ecological and physical processes that act to break down and redistribute CaCO3 sediment within reef systems (Figure 7.2, Perry et al., 2012; Kench, 2013). This interplay between ecological and physical processes, termed ecomorphodynamics (Kench, 2011), is responsible for the structural development of reefs, but also the formation and dynamics of secondary geomorphic features, including sedimentary landforms such as islands, coastal plains and beaches. Ecomorphodynamics provides a useful framework to examine how storm events can perturb the reef system leading to fundamental shifts in ecological and geomorphic state (Figure 7.2). In particular, the framework highlights a number of key aspects of

Environmental controls

Physical processes

Oceanographic climate

Coastal processes

Extreme ‘pulse’ events • Cyclones • Long period swell • Tsunami

Long-term change • Sea level • Wave climate • Water chemistry • Water temperature

Tectonic processes • Earthquakes

Water level Waves Currents

Coral reef system

Biological processes Reef ecology community structure

Reef platform development (Reef growth)

Gross carbonate production

Carbonate cycling Sediment generation

Sediment transport Geomorphic change • Islands • Lagoon infill

• Coastal plains • Reef structure

Figure 7.2 Conceptual model showing ecomorphodynamics of coral reefs. Shaded components represent storm effects.

CH7 STORMS IN CORAL REEFS

131

reef dynamics that are critical to understanding the influence of storms. First, storms are one of a set of boundary controls that can force changes in reef health, carbonate cycling and consequently geomorphic state. Second, the influence of storms on short-term carbonate cycling and geomorphic change can be evaluated at the ‘event scale’. However, aggregated over geological-timescales multiple storms can also influence long-term reef development. Third, the system contains feedbacks that in the case of storms are temporally specific; short-term changes in incident energy alter the structure of ecological communities (Chappell, 1980), reef morphology, sedimentation processes and short-term geomorphic change of beach and island shorelines (Sheppard, 2005). Fourth, feedbacks may be non-linear and can be associated with significant time lags. For instance, short-term stresses such as storms may cause changes in the reef that may propagate through the system, altering the carbonate budget and ultimately the ecology of the reef in the decades to come. Development of a robust understanding of the effects of storms on coral reefs must recognise geographic variations in the frequency and magnitude of storms: that storms affect reef structure and functioning at a range of spatial and temporal scales, and that the net effect of storms is also spatially and temporally variable. For example, at the macro-geographic scale (Figure 7.3) approximately 70% of tropical coral reefs can be found within the major zones of cyclogenesis corresponding to 7–25∘ north and south of the equator (Done, 2011b). However, in the equatorial region reefs are rarely, if ever, exposed to extreme storm events. Such variations in energetic regime have imparted discernible differences in the structure and susceptibility of reefs to extreme events. At the meso-scale (atoll or individual reef platform) within reef, variations of impacts occur as a result of the localised direction of storm approach. Coupled with reef location, with respect to zones of cyclogenesis, is the frequency of storms, the ability of reefs to withstand storm impacts and relaxation period for reef recovery/adjustment. In the storm

N

Historical tropical storm tracks (1842–2014) Cat 1 Cat 2 Cat 3 Cat 4 Cat 5

Figure 7.3 World map showing coral reefs of the world (database downloaded from http://www.wri .org/publication/reefs-risk-revisited) superimposed on historical tropical storm tracks between 1842 and 2014 (database downloaded from NOAA; Knapp et al. (2010, updated in 2014). Please note that bright red large points correspond to coral reef location.

132

COASTAL STORMS: PROCESSES AND IMPACTS

belt, reefs have been subject to thousands of cyclones throughout the past 6000 years and arguably the structure and function of reefs systems is a reflection of the frequency of these high-energy events. In contrast, reefs in equatorial regions are rarely influenced by extreme events. Coral reefs also receive the influence of long-period swells generated by distant storms; the effect of these swells may be dramatic in those reefs located near the equator, outside the cyclone belt. It is therefore important to note that coral reef structure is a consequence of a suite of environmental factors, including frequency and intensity of storm events. This in turn conditions the effects and timescales of recovery of a given reef and ultimately means that storm-associated morphological changes may vary significantly between localities.

7.2.2 Unique interactions of storm waves with coral reefs Coral reefs represent the interface between deep ocean waves and reef landforms, and the physical structure creates a hydrodynamic environment markedly different from continental beach systems. The interaction of incident ocean swell with coral reefs is widely known to modulate oceanographic, ecological and geological processes in coral reef systems (Roberts et al., 1992). In particular, waves and currents are the driving forces for sediment entrainment and transport, and ecological and geomorphic change (Macintyre et al., 1987; Smith & Buddemeier, 1992; Kench, 1998; Kench & Brander, 2006). Maintenance of healthy ecological reef systems is also heavily dependent on waves and currents that deliver nutrient supplies and renew water and oxygen for uptake by corals (Jokiel, 1978; Hearn et al., 2001); removal of wastes (Frith, 1983; Kraines et al., 1998); and the dispersal and recruitment of larvae (Hamner & Wolanski, 1988). There are a number of key hydrodynamic features that are unique to reef systems. First, the coral reef hydrodynamic environment is fundamentally determined by the interaction of waves across a steep transition from relatively deep to shallow water. This sharp transition promotes wave dissipation through breaking at the reef crest. On steep-faced reefs, such as those in atolls, wave breaking is the main dissipation mechanism; waves primarily break on the forereef or reef crest depending on the water depth available for wave propagation (Gourlay, 1994). For reefs with a more gentle gradient, such as those in the Caribbean, friction over the forereef plays a substantial role in energy dissipation, reducing wave heights by 20% before wave breaking occurs (Roberts et al., 1975). Ferrario et al. (2014) analysed 255 studies on coral reef and wave attenuation, revealing that coral reefs dissipate 97% of the incoming wave energy, with the reef crests alone dissipating most of it (86%). Second, wave breaking at the reef edge generates a suite of secondary wave motions and setup gradients that promote cross-reef currents and shoreline setup processes (Symonds et al., 1995). Third, not all energy is dissipated at the reef edge and at higher water levels considerable energy propagates onto and across reef surfaces toward lagoons and island shorelines (Brander et al., 2004). Fourth, wave processes at the reef edge and across reef flats are modulated by relative water depth across reefs, with studies noting across reef energy decay and a parallel reduction in sediment transport potential under mean energy conditions (Kench & McLean, 2004; Kench & Brander, 2006; Vila-Concejo et al., 2014; Harris et al., 2015a). Furthermore, the depth constraints on hydrodynamics imply that geomorphic activity is constrained to higher tidal stages when wave and current energy

CH7 STORMS IN CORAL REEFS

133

can penetrate across the reef surface (Kench & Brander, 2006; Storlazzi et al., 2011, Harris et al., 2014a). Based on observations that wave breaking occurs at the reef edge, reefs are considered to act as protective buffers to incident ocean swell (Ferrario et al., 2014). However, such assertions are based on analysis of wave studies undertaken under mean energy conditions. Of significance to this chapter is the fundamental shift in the hydrodynamic regime of reefs as a consequence of storms. Notably, intense storms can elevate incident wave heights impacting reefs by an order of magnitude, increasing the energy impacting the forereef and reef edge by two orders of magnitude. Storm surge also super-elevates water levels across reefs, which fundamentally transforms the reef flat wave environment, allowing much greater swell energy to propagate across reefs and across island surfaces (Maragos et al., 1973). In such instances the temporal window for geomorphic work is extended across entire tidal cycles and often across the duration of a storm event. Few studies have quantitatively recorded such events on reefs. However, numerous post-event observations have been undertaken on the transformations in ecological cover and geomorphology (e.g. Table 7.1). Numerous studies have noted the catastrophic effects of storms on living coral communities (Maragos et al., 1973; Harmelin-Vivien, 1994; Gardner et al., 2005) and island geomorphology (e.g. Stoddart, 1962; Maragos et al., 1973). Few studies have documented the changes in the process regime. Storms bring larger waves and stronger winds, generating strong currents and high shear stresses; however, the setup-driven increased water depth over the reef flat is Table 7.1 Summary of destructive and constructive effects on the zones of coral reefs as described in this chapter. Reef zone

Destructive

Constructive

Forereef

Loss of coral. Disappearance of spurs and grooves.

Sediment (rubble) production.

Reef flat

Loss of coral. Peel away parts of the reef.

Coral rubble generation. Rubble sedimentary features: 1. Rubble ridges or ramparts 2. Rubble tongues or spits 3. Exceptionally large boulders

Backreef: Sand Aprons

No clear effects either constructive or destructive.

No clear effects either constructive or destructive.

Reef Islands

Island erosion (mostly sandy and mangrove cays): 1. Shoreline displacement 2. Island disappearance 3. Damage to vegetation and human infrastructure

Formation and lateral growth of rubble islands through rubble transport from the reef flats. Vertical accretion of sandy and rubble islands through overwash processes.

134

COASTAL STORMS: PROCESSES AND IMPACTS

key for increasing sediment transport (Storlazzi et al., 2011). Kench and McLean (2004) quantified increases in cross reef flows and sediment transport rates, by two orders of magnitude, in response to TC Graham (December 1991) impacting the Cocos (Keeling) Islands. In a further study, results from numerical modelling in Hawaii showed that storm conditions are the dominant contributor to annual sediment flux, representing 63% of it (Storlazzi et al., 2011). The destructive and constructive effects of storms on reef systems (Table 7.1) are explored in greater depth in the following sections.

7.3 Storms on the forereef: Role of spurs and grooves The forereef is characterised by the inclination of the slope, which can be marked by breaks or platforms and the occurrence of spur and groove systems in the upper parts (Cabioch, 2011). Spurs and grooves are one of the most biodiverse and productive zones of modern reefs (Perry et al., 2012) and yet, their formation, processes and evolution remain a mystery for science. Spurs and grooves (Figure 7.1 and Figure 7.4a) are three-dimensional finger-like formations (ridges and channels) that are mostly ubiquitous in coral reefs’ fronts. They act

(a)

(c)

(b)

(d)

(e)

Figure 7.4 One Tree Reef, Southern Great Barrier Reef, Australia (see Figure 7.1): (a) A diver on the forereef, a spur can be seen to the left of the photo and a groove to the right, the spur is covered with laminar corals (plates); (b) Laminar corals on the forereef; (c) A rubble dominated reef flat showing old (dark colours) and new (white) coral rubble; (d) Close-up photo of rubble dominated reef flat, presence of laminar corals (plates); (e) Another view of the rubble dominated reef flat, SE corner of One Tree Reef.

CH7 STORMS IN CORAL REEFS

135

as natural breakwaters and play an important role in dissipating wave energy (Munk & Sargent, 1954; Roberts et al., 1977); they are therefore critical to the reef’s stability to resist erosion (Sheppard, 1981). Spurs and grooves have been documented in every coral reef region of the world, including the Great Barrier Reef, the Red Sea, the Caribbean region and the Indian, Pacific and Atlantic oceans. They have also been found in fossil reef structures (e.g. Wood & Oppenheimer, 2000). The size, morphology, depth, alignment, spacing and species composition of spur and groove systems are related to dominant wave conditions and vary markedly from reef to reef (Storlazzi et al., 2003; Gischler, 2010). The distance between consecutive grooves (wavelength) ranges from 5 m (Sheppard, 1981) to 150 m (Storlazzi et al., 2003). Spur heights, the difference in elevation between consecutive spurs and grooves, vary between 0.6 m (Cloud, 1959) and 7 m (Newell, 1954). According to Duce et al. (2014), spurs and grooves extend from 1 m to approximately 35 m below mean sea level, showing lengths that vary from tens to hundreds of metres. There is some consensus in these variations in wavelength; height and length being related to the wave climate at each reef, with less wave energy correlating with less defined spurs and grooves (Roberts et al., 1977). The alignment of spurs and grooves is typically related to the direction of the dominant wave energy (e.g. Shinn, 2011). This relationship was clarified by Duce et al., (2016) who established a morphometric classification for spurs and grooves. The role that spurs and grooves play on dissipating wave energy is unclear, although there is a widespread consensus that their 3D morphology contributes to enhancing wave attenuation. Foley et al. (2014) suggest that the morphology of spurs and grooves is important not only for wave energy dissipation but also in providing a shelter for coral, allowing post-storm recovery by promoting coral recruitment and growth.

7.3.1 Destructive effects of storms in the forereef and spur and groove Storms on forereefs are mostly destructive processes (Table 7.1). Storms (including cyclones, hurricanes, typhoons and strong wind events) have catastrophic impacts on coral reefs worldwide. Harmelin-Vivien and Laboute (1986) report that more than 50% of the corals on the reef slopes of French Polynesia were lost after a series of hurricanes in the early 1980s. Woolsey et al. (2012) report on the effects of a category 4 cyclone passing near One Tree Reef (OTR) on the southern Great Barrier Reef. They found a significant decrease on live corals in general and complete disappearance (even two years after cyclone impact) of laminar corals on the exposed flanks of OTR (Figure 7.4a and b). Scoffin (1993) explained that a cyclone/hurricane can break out corals in the forereef up to a depth of 20 m; material broken is then transported to the forereef apron, stays in place or is transported up to the rubble flat. Thornborough and Davies (2011) explained that the broken coral accumulates in situ and is then picked up and transported up to the reef flats. The complete disappearance of spurs and grooves resulting in the flattening of the forereef after a severe tropical storm has also been reported in the literature (Stoddart, 1962). Destruction caused by storms is variable and it mostly depends on the forereef slope and angle of approach of the storm. The forereef slope and angle of approach of storm waves can control the degree of damage, with quasi-vertical forereefs showing

136

COASTAL STORMS: PROCESSES AND IMPACTS

less damage than gently sloping forereefs; and forereefs facing the waves showing more damage than forereefs parallel to the direction of wave approach (Etienne, 2012). Coral coverage can be affected by physical, chemical and ecological factors like storms, water conditions and interactions between species. Subsequent recovery varies depending on the reef, for example Woolsey et al. (2012) estimated that OTR would recover from cyclone Hamish in less than ten years.

7.3.2 Constructive effects of storms in the forereef Coral loss in the forereef results in sediment rubble, which may form sedimentary deposits typical of high-energy coral reefs. Rubble deposits are essential for the long-term resilience of coral reefs. Rubble formed during a storm may be stored in the forereef; the same or subsequent storms may then entrain and transport this sediment down the forereef or up to the reef crest and flats (see section 7.4). The percentage of coral reef fragments that are transported from the forereef into the reef flats during subsequent high-energy events has not been quantified. There are, however, reports of large quantities of coral rubble being transported to the reef flats after a storm (e.g. Woolsey et al., 2012). As evidenced in other chapters of this book (for example Chapters 6, 9 and 10), it is worth noticing that small differences in storm characteristics can alter its overall ‘impact’ and that ‘storms’, so-defined because of their wave height, can be accretionary by producing sediments that then accumulate elsewhere.

7.4 Storms on the reef flats: Development of rubble flats and rubble spits 7.4.1 Waves on the reef flats Reef flats are near-horizontal hard surfaces of coral framework that may (coraldominated) or may not (rubble dominated) have coral growing on them. Sheltered coral reef flats are largely coral-dominated, while exposed reef flats are typically rubble dominated. Most of the wave energy is dissipated on the forereef, which means that, even under storms, the waves on the reef flats remain small (Harris et al., 2014b). Therefore, only an extremely high sea level caused by a combination of high astronomical tides and storm surge, would limit wave dissipation over the forereef and cause large waves to propagate over the reef flats. Such conditions may occur under cyclone conditions.

7.4.2 Destructive effects of storms on reef flats Given their largely featureless morphology, reef flats are typically resilient to storms (Table 7.1). On coral-dominated reef flats storms and cyclones dislodge coral colonies and cause coral mortality and loss of coral cover on coral-dominated reef flats, especially on the reef crests (Massel & Done, 1993; Madin & Connolly, 2006). Very large severe tropical storms can even peel away parts of the reef flat (e.g. Scoffin, 1993)

CH7 STORMS IN CORAL REEFS

137

7.4.3 Constructive effects of storms on reef flats Storms are mostly constructive processes on rubble-dominated reef flats (Table 7.1) as they bring coral rubble that can accumulate on the reef flats. High-energy waves entrain and transport the coral rubble generated in the forereef and part of it can be transported into the reef flats (Figure 7.4c, d, and e). This coral rubble can then be broken down into smaller particles that can be transported into the backreef. High-energy waves generated by storms and cyclones create rubble depositional features that are typical of exposed reefs. The most common are: (1) storm ridge or ramparts; (2) rubble tongues or spits; and, (3) exceptionally large boulders or blocks of coral framework. A storm ridge or rampart (Figure 7.4c and 7.5e) is a shore-parallel accumulation of coral rubble (shingle) near the seaward margin of the reef flat (Scoffin, 1993). Storm ridge formation is the result of stronger than average wave energy and can be the product of swash or drift alignments (Scheffers et al., 2012). Swash-aligned storm ridges run strongly parallel to each other, while drift-aligned storm ridges increase their spacing downdrift (Scheffers et al., 2012). Ridges may extend round the reef but usually occur just on the windward side (Scoffin, 1993). Storm ridges are then reworked by subsequent storms and, as result, may migrate towards the lagoon and even coalesce to form or enlarge reef islands (Figure 7.4e and 7.5f). The migration is often encompassed by a change in the seaward slope of the storm ridge, which changes from convex to concave (Baines & McLean, 1976). Sometimes, storm ridges are reworked by storms that cause the spreading of the coral rubble creating a sheet of rubble (Scoffin, 1993), which could end up covering the entire reef flat and becoming a rubble-dominated flat (Figure 7.4e, 7.5c and d). Storm ridges and rubble flats may show across-shore rubble accumulation deposits called rubble tongues, spits or spurs (Figure 7.4e), that can be associated with rubble ridges (Scoffin, 1993) or appear as an independent feature (Etienne & Terry, 2012). Shannon et al. (2013) presented the evolution of a rubble flat and associated rubble spits on the eastern (windward) reef flat of OTR over 45 years. Their results show that both rubble flats and rubble spits prograded towards the lagoon between 1964 and 2009; rubble flat progradation (average 0.5 m/yr) was significantly slower than the progradation of the rubble spits (average 2 m/yr). The availability of rubble, the underlying substrate, the energy regime and storm frequency controlled the rate of progradation, and, interestingly, they found that while rubble flats only existed on the high-energy (windward) margin, rubble spits could be found reef-wide. Their results showed that rubble spits on the low-energy flanks prograded faster than those on the high-energy flanks, and that was due to low-energy spits prograding over shallower areas and thus needing less rubble to advance. They also identified that rubble spits seemed to occupy preferential positions along the reef flats, with new rubble spits developing over the top of existing rubble spits. They hypothesised that their preferred location was related to the forereef morphology, identifying larger grooves through which storm waves would focus their energy, causing a preferential pathway for rubble transport. A similar process was identified by Hamylton and Spencer (2011) in the Seychelles, where they observed a distinctive stripped pattern in sediment deposition on the reef flat that they associated to the forereef morphology of spur and groove.

138

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

(b)

(c)

(d)

(e)

(f)

Figure 7.5 Storm depositional products on coral reefs. Reef flat storm blocks in Bonnaire (a) and Funafuti atoll, Tuvalu (b), intertidal rubble sheet increasing reef elevation at Funafuti atoll (c), shingle ridge at Bewick Island, Great Barrier Reef (d), hurricane Bebe storm rampart (e) and subsequent deposition of gravel against island shoreline, Funafuti atoll, Tuvalu (f).

Another storm/severe tropical storm effect is the appearance of exceptionally large boulders of coral or blocks of reef framework (Figure 7.5a and b) near the windward margin of some reefs (Scoffin, 1993). Etienne (2012) examined fresh coral boulders tossed on the reef flat by tropical cyclone Oly in French Polynesia; he used his measurements of boulder dimensions to estimate flow velocity following the equations of Nott (2003) and Nandasena et al. (2011). Etienne (2012) obtained flow velocities exceeding 3 m/s in the upper part of the forereef (up to 10 m depth) and estimated flow velocities

CH7 STORMS IN CORAL REEFS

139

on the reef flat between 3 and 5 m/s. Another study using similar methods, but this time in Fiji after cyclone Tomas, yielded flow velocities between 2 and 4 m/s (Etienne & Terry, 2012).

7.5

Storms on the backreef: Sand aprons, reef islands and beaches

7.5.1 Sand aprons Sand aprons (also known as sand flats or sand sheets) are sandy deposits often found in the backreef environment. Sand aprons can be found in the backreef of both rubble-dominated and coral-dominated reef flats. In the case of coral-dominated reef flats, the sand apron provides a substrate of suitable depth on which corals can colonise, allowing coral growth towards the lagoon (Thornborough & Davies, 2011). The sand in the sand aprons is typically composed of bioclasts formed by physical processes and bioerosion in the nearby reef flats; large benthic foraminifera (LBF) are also abundant and may represent up to 70% of the composition of sand in Indo-Pacific reef platforms (Yamano et al., 2000). Hydrodynamics exert controls on the morphology and sedimentology of sand aprons, with waves and tides being the primary controls. On meso- and macrotidal environments, waves can only propagate over the reef flats during high tide (Harris et al., 2014a, b, 2015a; Vila-Concejo et al., 2014). Grain size results show coarser sediment next to the reef flat and finer sediment towards the lagoon (Harris et al., 2011; Wasserman & Rankey, 2014). The effects of storms on sand aprons are not well understood. Vila-Concejo et al. (2013) showed the southern (exposed) sand apron at One Tree Reef before and after the impact of Cyclone Hamish (described in Woolsey et al., 2012) in 2009. Their findings show that the sand apron did not undergo any major changes (Figure 7.6) despite the significant changes observed on the rubble flats. Further research on the sediment transport and hydrodynamics on the sand apron, showed that under modal conditions there is enough energy to transport sediment, but that this transport is small and does not contribute to sand apron progradation (Harris et al., 2014a; Vila-Concejo et al., 2014). Holocene studies looking into core samples of the sand apron, microatolls, and other proxies established that the sand apron at OTR is a relict feature that was formed between 6000 and 2000 calibrated years before present (Harris et al., 2015b) and that currently there is no evidence of sediment transport conducive to sand apron progradation and lagoon infilling (Vila-Concejo et al., 2015). The sand apron is no longer prograding and it appears to be sheltered from present day conditions.

7.5.2 Reef islands 7.5.2.1 Destructive processes on reef islands Island erosion has been highlighted as a conspicuous outcome of extreme storm waves interacting with unconsolidated sedimentary shorelines in reef environments. In particular, shoreline sediment can be remobilised under higher energy episodes and transported: alongshore, through reef passages to contribute to sand apron progradation and lagoon filling; off-reef into barrier lagoons or down forereef slopes, where it is lost to the sediment reservoir. While

140

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

(b)

Figure 7.6 Long-term evolution of sedimentary deposits at One Tree Reef, Great Barrier Reef, Australia: (a) An example of rubble spit formation and evolution showing its recurrent position on the northern margin. (b) Sand apron evolution on the southern margin (modified from Vila-Concejo et al. (2013)). In both cases, the background image is WorldView 2 from December 2009. GIS analyses by Amelia Shannon.

there are multiple forms of erosional signature on islands (see Stoddart, 1971) the net effect is generally a localised loss of sediment volume promoting lateral shoreline displacement. For example, Hurricane Hattie in October 1961 caused major erosional impacts on reefs and sand cays in the Belize Barrier reef (Stoddart, 1963). Five cays were completely devegetated, with sediment dispersed across the reef platforms, resulting in

CH7 STORMS IN CORAL REEFS

141

Figure 7.7 Erosion impacts on islands in the Belize barrier reef following Hurricane Hattie 1961. Grey shading is the island area prior to the hurricane. Black area is the post hurricane island area. Based on Stoddart (1969).

the entire loss of islands. A number of other islands also experienced more than 50% loss in vegetated area (Figure 7.7). Similarly, in the Marshall Islands major typhoons have been associated with significant island erosion (Blumenstock, 1958). In particular, an event in 1905 impacted the atoll of Nadikdik, promoting erosion and disaggregation of islands (Ford & Kench, 2014) while in January 1958 a very intense typhoon ‘Ophelia’ impacted Jaluit atoll, also resulting in erosion signatures on islands. In addition to shoreline and island erosion, these storms also cause significant damage to vegetation complexes and human infrastructure. While the number of direct observations of storm damage on reef islands is still limited, the erosional impacts are commonly associated with sand islands and mangrove cays.

7.5.2.2 Constructive processes on reef islands There are a number of constructive depositional products that are also associated with extreme events. As noted earlier, cyclonic waves can decimate the living coral veneer of reefs to wave base water depths, transporting this material onto the reef platform surface. While this action resets the physical substrate of the forereef for recolonisation, the process can result in a pulse of sediment added to the detrital sediment reservoir of the reef platform. These deposits form a range of conspicuous features including: isolated large coral or reef framework boulders that perch on reef surfaces; subtidal and intertidal rubble sheets that can contribute to vertical reef development; subaerial rubble tracts located on the reef surface, and islands (Figure 7.5). The delivery of rubble to the reef surface can directly and indirectly contribute to island accretion. For example Lady Elliot Island, Great Barrier Reef, formed through the sequential deposition of storm rubble banks (Chivas et al., 1986). McLean and Hosking (1991) also attribute the formation of rubble islands on the eastern rim of Funafuti atoll directly to extreme wave events. Storm deposits on the reef surface can also provide a reservoir to contribute to island accretion through secondary reworking that may occur for years to decades following an event. For example, in October 1972 Tropical Cyclone Bebe directly affected the

142

COASTAL STORMS: PROCESSES AND IMPACTS

atoll of Funafuti, Tuvalu. The event ripped up the living coral, and extant carbonate sediment from the fore-reef, depositing a boulder rampart on the reef flat. The rampart contained approximately 1.4 x 106 m3 of storm rubble, had a mean width and elevation of 37 m and 2.5 m, and extended 18 kilometres along the east to south-eastern rim of the atoll (Maragos et al., 1973) (Figure 7.5e). The rapid appearance of this deposit had immediate effects on morphology and processes on the atoll rim, including blocking of passages between islands impacting lagoon flushing, and shutting down the process window at island shorelines. Subsequent monitoring of this rubble bank has shown that successive storms have reworked the rubble through overwash processes, forcing the bank to migrate landward and fuse to islands, with rubble being transferred alongshore to extend islands (Baines & McLean, 1976). Collectively this reworking of storm material has increased island area over the past 40 years (Kench et al., 2015) (Figure 7.8). An additional constructional effect of storms on islands is washover sedimentation as a result of extreme waves overtopping shoreline ridges. Waves entrain nearshore sediments, which are transported and deposited onto and across the island surface (a)

(b)

(c)

Figure 7.8 The modification of the Hurricane Bebe storm rampart on Funafuti Atoll, Tuvalu: (a) Sequential surveys showing the migration of rampart rubble toward the shoreline. (b) The rampart deposited on the reef flat. (c) Expansion of Funamanu Island, Funafuti 1971–2013 as a result of delivery of rampart rubble to the shoreline.

CH7 STORMS IN CORAL REEFS

143

in the overwash flow (see also Chapter 4). A number of early studies qualitatively described such washover deposits that range up to almost 0.80 m in depth (Blumenstock, 1958; Stoddart, 1971). More recently, studies have quantitatively assessed the role of washover on island change (Kench et al., 2006; Smithers & Hoeke, 2014). Washover deposits typically have their maximum depth near the island ridge and taper landward. Consequently, the spatial extent of such deposits is determined by the extent that waves and broken bores can propagate across the island surface. Commonly, washover sheets are comprised of sand-size material (Kench et al., 2006; Smithers & Hoeke, 2014). However, large coral boulders can also be tossed onto island surfaces under extreme conditions. Such washover deposition can be an active mechanism of vertical island building, particularly where islands experience multiple wave inundation events. For example, Kench et al. (2005) found conspicuous washover deposition on Maldivian reef islands following the Indian Ocean tsunami in December 2004. Subsequent surveys have shown that additional washover deposition occurred in 2007 as a result of an episode of high waves with long period, generated by high latitude storms (Figure 7.9). A similar high latitude storm event occurred in the Pacific in December 2008 and delivered long-wave length wind-waves to the tropical Pacific. Waves were reported to overwash islands for several days, at higher tidal stages (Smithers & Hoeke, 2014). On Nukutoa, Takuu atoll in Papua New Guinea, this event

(a)

(c)

(b)

Figure 7.9 Storm and high wave washover deposits in the Maldives: (a) Limits of a washover sand sheet. (b) Depth of the sand sheet as a consequence of the 2004 Indian Ocean tsunami. (c) Multiple washover layers resulting from the 2004 tsunami and 2007 long-period swell event.

144

COASTAL STORMS: PROCESSES AND IMPACTS

deposited a washover sand sheet that covered 13% of the island with a maximum depth of 0.22m. Bayliss-Smith (1988) following analysis of the post-storm impacts of TC Annie on Ontong Java, Solomon Islands, proposed one of the earliest morphodynamic models for reef islands, which synthesised these differences in morphological response, dependent upon variations in the interplay between the calibre of island sediments (gravel or sand) and the frequency and intensity of storms (Figure 7.10). The model shows the contrasting impact trajectories of reef landforms and differing lag period responses. For example, while a sand island may experience erosion during a large event, over the ensuing years sand accumulation may rebuild the landform similar to a continental sandy shoreline. However, in storm adjusted locations storms may deliver pulses of sediment to islands, which increases the sediment volume. Over the ensuing decades, without continual input of new material, the storm deposit is reworked and material is gradually removed from the system towards depths where it cannot be reworked by ordinary waves, reducing island volume.

(a)

(b)

Figure 7.10 Conceptual model of reef island shoreline dynamics in response to storm events: (a) Storm frequency and intensity. (b) Morphological response of sand and gravel islands as reflected in sediment volume. After Bayliss-Smith (1988).

CH7 STORMS IN CORAL REEFS

7.6

145

Conclusion

It is important to recognise that the presence of constructional and destructional signatures resulting from individual storm events are not mutually exclusive. Both processes are commonly found on a single island or reef, or on adjacent islands or reefs, depending on localised exposure to extreme waves, and reflect the aggregate morphological response of these systems, mediated through adjustments in the sediment reservoir in response to extreme changes in incident energy. For example, in the extreme wave event that impacted islands of the Pacific in 2008, Smithers and Hoeke (2014) documented that 60% of the Takuu shoreline showed evidence of erosion, while washover deposition also occurred. Kench et al. (2006) also found island erosion (maximum of 9%) and washover sedimentation (up to 17% of island area) on Maldivian reef islands following wave inundation from tsunami. We have shown that storms and cyclones may cause destructive and constructive processes in different zones of coral reefs. Whether storms are constructive or destructive (and to which degree) depends on the modal conditions upon which a given coral reef exists. Coral reefs located within the cyclone belt, are prepared to receive storm energy and may thrive under it. Those coral reefs that do not often receive high-energy storms are the ones that will show the most dramatic effects under storm conditions. Numerous studies shown in this chapter demonstrate that coral reefs can recover and even thrive on storm impacts. Coral reefs are iconic. They are some of the most biodiverse ecosystems on Earth and store large quantities of carbon. This chapter demonstrates that we still need to investigate further to understand the processes that shape coral reefs, with special attention to storm effects. We live in a world where climate (including storm regimes) is changing and sea levels are rising. Global warming associated with climate change might cause the tropicalisation of some temperate zones; climate change will modify the intensity and frequency of storms such that coral reefs that are currently outside the storm belt will start to be regularly impacted by high-energy storms. Some coral reefs may thrive and some may erode beyond recovery, coral reefs may grow in new areas. Field-based studies to understand current processes, complemented by remote sensing and numerical modelling to forecast other scenarios will see our knowledge increase, we hope, to critical levels where we can prepare for what the future may bring.

Acknowledgements AVC acknowledges funding from Australian Research Council Future Fellowship FT100100216. Underwater photos in the Spur and Groove were provided by Belinda Dechnik and Stephanie Duce.

References Baines, G.B.K. & MClean, R.F. (1976) Sequential studies of hurricane deposit evolution at Funafuti atoll. Marine Geology, 21, M1–M8. Bayliss-Smith, T.P. (1988) The role of hurricanes in the development of reef islands, Ontong Java Atoll, Solomon Islands. Geographical Journal, 377–391.

146

COASTAL STORMS: PROCESSES AND IMPACTS

Best, B. & Bornbusch, A. (2005) Global Trade and Consumer Choices: Coral Reefs in Crisis. Washington, DC, American Association for the Advancement of Science. Blanchon, P. (2011) Geomorphic zonation. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Blumenstock, D.I. (1958) Typhoon effects at Jaluit Atoll in the Marshall Islands. Nature, 182, 1267–1269. Brander, R.W., Kench, P.S. & Hart, D. (2004) Spatial and temporal variations in wave characteristics across a reef platform, Warraber Island, Torres Strait, Australia. Marine Geology, 207, 169–184. Cabioch, G. (2011) Forereed/Reef front. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. Springer. Chappell, J. (1980) Coral morphology, diversity and reef growth. Nature, 286, 249–252. Chivas, A., Chappell, J., Polach, H., Pillans, B. & Flood, P. (1986) Radiocarbon evidence for the timing and rate of Island development, beach-rock formation and phosphatization at Lady Elliot Island, Queensland, Australia. Marine Geology, 69, 273–287. Cloud, P.E.J. (1959) Geology of Saipan, Mariana Islands. US Geological Survey Professional Paper, 280K, 361–445. Collins, L. (2011) Reef structure. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. Springer. Davies, P.J. & Kinsey, D.W. (1977) Holocene reef growth – One Tree Island, Great Barrier Reef. Marine Geology, 24, M1–M11. Davies, P.J. & Marshall, J.F. (1980) A model of epicontinental reef growth. Nature, 287, 37–38. Done, T. (2011a) Coral Reef, Definition. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Done, T. (2011b) Tropical Cyclone/Hurricane. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Duce, S., Vila-Concejo, A., Hamylton, S.M., Bruce, E. & Webster, J.M. (2014) Spur and groove distribution, morphology and relationship to relative wave exposure, Southern Great Barrier Reef, Australia. Journal of Coastal Research, SI(70), 115–120. Duce, S., Vila-Concejo, A., Hamylton, S.M., Webster, J.M., Bruce, E. & Beaman, R.J. (2016) A morphometric assessment and classification of coral reef spur and groove morphology. Geomorphology, 265, 68–83. Etienne, S. (2012) Marine inundation hazards in French Polynesia: Geomorphic impacts of Tropical Cyclone Oli in February 2010. Geological Society, London, Special Publications, 361, 21–39. Etienne, S. & Terry, J.P. (2012) Coral boulders, gravel tongues and sand sheets: Features of coastal accretion and sediment nourishment by Cyclone Tomas (March 2010) on Taveuni Island, Fiji. Geomorphology, 175–176, 54–65. Ferrario, F., Beck, M.W., Storlazzi, C., Micheli, F., Shepard, C.C. & Airoldi, L. (2014) The effectiveness of coral reefs for coastal hazard risk reduction and adaptation. Nature Communications, 5. Foley, M., Stender, Y., Singh, A., Jokiel, P. & Rodgers, K. (2014) Ecological engineering considerations for coral reefs in the design of multifunctional coastal structures. International Conference Coastal Engineering, 2014 South Korea. Coastal Engineering Research Council. Ford, M.R. & Kench, P.S. (2014) Formation and adjustment of typhoon-impacted reef islands interpreted from remote imagery: Nadikdik Atoll, Marshall Islands. Geomorphology, 214, 216–222. Frith, C.A. (1983) Some aspects of lagoon sedimentation and circulation at One Tree Reef, Southern Great Barrier Reef. BMR Journal of Australian Geology and Geophysics, 8, 211–221.

CH7 STORMS IN CORAL REEFS

147

Gardner, T.A., Côté, I.M., Gill, J.A., Grant, A. & Watkinson, A.R. (2005) Hurricanes and Caribbean coral reefs: Impacts, recovery patterns and role in long-term decline. Ecology, 86, 174–184. Gischler, E. (2010) Indo-Pacific and Atlantic spurs and grooves revisited: The possible effects of different Holocene sea-level history, exposure, and reef accretion rate in the shallow fore reef. Facies, 56, 173–177. Gourlay, M.R. (1994) Wave transformation on a coral reef. Coastal Engineering, 23, 17–42. Hamner, W.H. & Wolanski, E. (1988) Hydrodynamic forcing functions and biological processes on coral reefs: A status review. International Coral Reef Symposium, 1988 Townsville, Australia. 8–12. Hamylton, S. & Spencer, T. (2011) Geomorphological modelling of tropical marine landscapes: Optical remote sensing, patches and spatial statistics. Continental Shelf Research, 31, S151–S161. Harmelin-Vivien, M.L. (1994). The effects of storms and cyclones on coral reefs: A review. Journal of Coastal Research, 211–231. Harmelin-Vivien, M.L. & Laboute, P. (1986) Catastrophic impact of hurricanes on atoll outer reef slopes in the Tuamotu (French Polynesia). Coral Reefs, 5, 55–62. Harris, D.L., Webster, J.M., De Carli, E.V. & Vila-Concejo, A. (2011) Geomorphology and morphodynamics of a sand apron, One Tree Reef, Southern Great Barrier Reef. Journal of Coastal Research, SI, 760–764. Harris, D.L., Vila-Concejo, A. & Webster, J.M. (2014a) Geomorphology and sediment transport on a submerged back-reef sand apron: One Tree Reef, Great Barrier Reef. Geomorphology, 222, 132–142. Harris, D.L., Vila-Concejo, A., Power, H.E. & Webster, J.M. (2014b) Wave processes on coral reef flats during modal and storm conditions. In: AGU, Ed. AGU Fall Meeting, 2014, San Francisco. Harris, D.L., Vila-Concejo, A., Webster, J.M. & Power, H.E. (2015a) Spatial variations in wave transformation and sediment entrainment on a coral reef sand apron. Marine Geology, 363, 220–229. Harris, D.L., Webster, J.M., Vila-Concejo, A., Hua, Q., Yokoyama, Y. & Reimer, P.J. (2015b) Late Holocene sea-level fall and turn-off of reef flat carbonate production: Rethinking bucket fill and coral reef growth models. Geology, 43, 175–178. Hearn, C.J., Atkinson, M.J. & Falter, J.L. (2001) A physical derivation of nutrient uptake rates in coral reefs: Effects of roughness and waves. Coral Reefs, 20, 347–356. Hopley, D., Smithers, S.G. & Parnell, K.E. (2007) The Geomorphology of the Great Barrier Reef. Development, Diversity and Change. Cambridge, UK, Cambridge University Press. Jokiel, P.L. (1978) Effects of water motion on reef corals. Journal of Experimental Marine Biology and Ecology, 35, 87–97. Kench, P.S. (1998) Physical controls on development of lagoon sand deposits and lagoon infilling in an Indian Ocean atoll. Journal of Coastal Research, 14, 1014–1024. Kench, P.S. (2011) Eco-morphodynamics. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Kench, P.S. (2013) Coral systems. In: J. Shroder & D.J. Sherman (Eds) Treatise on Geomorphology, Vol. 10, Coastal Geomorphology. San Diego: Academic Press. Kench, P.S. & MClean, R.F. (2004) Hydrodynamics and sediment flux of hoa in an Indian Ocean atoll. Earth Surface Processes and Landforms, 29, 933–953. Kench, P.S. & Brander, R.W. (2006) Wave processes on coral reef flats: Implications for reef geomorphology using Australian case studies. Journal of Coastal Research, 22, 209–223. Kench, P.S., MClean, R. & Nichol, S. (2005) New model of reef-island evolution: Maldives, Indian Ocean. Geology, 33, 145.

148

COASTAL STORMS: PROCESSES AND IMPACTS

Kench, P.S., MClean, R.F., Brander, R.W., Nichol, S.L., Smithers, S.G., Ford, M.R., et al. (2006) Geological effects of tsunami on mid-ocean atoll islands: The Maldives before and after the Sumatran tsunami. Geology, 34, 177–180. Kench, P.S., Thompson, D., Ford, M., Ogawa, H. & MClean, R. (2015) Coral islands defy sea-level rise over the past century: Records from a central Pacific atoll. Geology, 43, 515–518. Knapp, K.R., Kruk, M.C., Levinson, D.H., Diamond, H.J. & Neumann, C.J. (2010, updated in 2014) The International Best Track Archive for Climate Stewardship (IBTrACS): Unifying tropical cyclone best track data. Bulletin of the American Meteorological Society, 91, 363–376. Kraines, B.S., Yanagi, T., Isobe, M. & Komiyama, H. (1998) Wind-wave driven circulation on the coral reef at Bora Bay, Miyako Island. Coral Reefs, 17, 133–143. Macintyre, I.G., Graus, R.R., Reinthal, P.N. & Litter, D.S. (1987) The barrier reef sediment apron: Tobacco Reef, Belize. Coral Reefs, 6, 1–12. Madin, J.S. & Connolly, S.R. (2006) Ecological consequences of major hydrodynamic disturbances on coral reefs. Nature, 444, 477–480. Maragos, J.E., Baines, G.B. & Beveridge, P.J. (1973) Tropical cyclone bebe creates a new land formation on funafuti atoll. Science (New Yo), 181, 1161–1164. Massel, S.R. & Done, T.J. (1993) Effects of cyclone waves on massive coral assemblages on the Great Barrier Reef: Meteorology, hydrodynamics and demography. Coral Reefs, 12, 153–166. MClean, R. & Hosking, P. (1991) Geomorphology of reef islands and atoll motu in Tuvalu. South Pacific Journal of Natural Science, 11, 167–189. Munk, W.H. & Sargent, M.C. (1954) Adjustment of Bikini Atoll to Ocean Waves. US Geological Survey Professional Paper, 260 C, 275–280. Nandasena, N.A.K., Paris, R. & Tanaka, N. (2011) Reassessment of hydrodynamic equations: Minimum flow velocity to initiate boulder transport by high energy events (storms, tsunamis). Marine Geology, 281, 70–84. Newell, N.D. (1954) Reefs and sedimentary processes of Raroia. Atoll Research Bulletin. Nott, J. 2003. Waves, coastal boulder deposits and the importance of the pre-transport setting. Earth and Planetary Science Letters, 210, 269–276. Perry, C.T., Edinger, E.N., Kench, P.S., Murphy, G.N., Smithers, S.G., Steneck, R.S. et al. (2012) Estimating rates of biologically driven coral reef framework production and erosion: A new census-based carbonate budget methodology and applications to the reefs of Bonaire. Coral Reefs, 31, 853–868. Rankey, E.C. & Garza-Pérez, J.R. (2012) Seascape Metrics of shelf-margin reefs and reef sand aprons of holocene carbonate platforms. Journal of Sedimentary Research, 82, 53–71. Roberts, H.H., Murray, S.P. & Suhayda, J.N. (1975) Physical processes in fringing reef systems. Journal of Marine Research, 33, 233–260. Roberts, H.H., Murray, S.P. & Suhayda, J.N. (1977) Physical Processes in a Fore-Reef Shelf Environment. Third International Coral Reef Symposium, 1977 Miami, USA. Roberts, H.H., Wilson, P.A. & Lugofernandez, A. (1992) Biologic and geologic responses to physical processes – examples from modern reef systems of the Caribbean-Atlantic region. Continental Shelf Research, 12, 809–834. Scheffers, A.M., Scheffers, S.R., Kelletat, D.H., Squire, P., Collins, L., Feng, Y., et al. (2012) Coarse clast ridge sequences as suitable archives for past storm events? Case study on the Houtman Abrolhos, Western Australia. Journal of Quaternary Science, 27, 713–724. Scoffin, T.P. (1993) The geological effects of hurricanes on coral reefs and the interpretation of storm deposits. Coral Reefs, 12, 203–221. Shannon, A., Power, H., Webster, J. & Vila-Concejo, A. (2013) Evolution of coral rubble deposits on a reef platform as detected by remote sensing. Remote Sensing, 5, 1–18. Sheppard, C.R.C. (1981) The groove and spur structures of Chagos Atolls and their coral zonation. Estuarine Coastal and Shelf Science, 12, 549.

CH7 STORMS IN CORAL REEFS

149

Sheppard, S.R.J. (2005) Landscape visualisation and climate change: The potential for influencing perceptions and behaviour. Environmental Science & Policy, 8, 637–654. Shinn, E.A. (2011) Spurs and grooves. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Smith, S.V. & Buddemeier, R.W. (1992) Global change and coral reef ecosystems. Annual Review of Ecology and Systematics, 23, 89–118. Smithers, S. & Hoeke, R. (2014) Geomorphological impacts of high-latitude storm waves on low-latitude reef islands – Observations of the December 2008 event on Nukutoa, Takuu, Papua New Guinea. Geomorphology, 222, 106–121. Spalding, M., Ravilious, C. & Green, E.P. (2001) World Atlas of Coral Reefs. Univ. of California Press. Stoddart, D.R. (1962) Catastrophic storm effects on the British Honduras reefs and cays. Nature, 196, 512–515. Stoddart, D.R. (1963) Effects of Hurricane Hattie on the British Honduras reefs and cays, October 30–31, 1961. National Academy of Sciences, National Research Council, Washington DC, Pacific Science Board. Stoddart, D.R. (1969) Ecology and morphology of recent coral reefs. Biological Reviews, 44, 433–498. Stoddart, D.R. (1971) Coral reef and islands and catastrophic storms. Applied Coastal Geomorphology. London: Macmillan. Storlazzi, C.D., Logan, J.B. & Field, M.E. (2003) Quantitative morphology of a fringing reef tract from high-resolution laser bathymetry: Southern Molokai, Hawaii. Geological Society of America Bulletin, 115, 1344–1355. Storlazzi, C.D., Elias, E., Field, M. & Presto, M. (2011) Numerical modeling of the impact of sea-level rise on fringing coral reef hydrodynamics and sediment transport. Coral Reefs, 30, 83–96. Symonds, G., Black, K.P. & Young, I.R. (1995) Wave-driven flow over shallow reefs. Journal of Geophysical Research: Oceans, 100, 2639–2648. Thornborough, K.J. & Davies, P.J. (2011) Reef flats. In: D. Hopley (Ed.) Encyclopedia of Modern Coral Reefs. The Netherlands: Springer. Vila-Concejo, A., Harris, D.L. & Power, H.E. (2015) Sand transport in coral reefs: Are lagoons infilling? Coastal Sediments, 2015 San Diego. ASCE. Vila-Concejo, A., Harris, D.L., Shannon, A.M., Webster, J.M. & Power, H.E. (2013) Coral reef sediment dynamics: Evidence of sand-apron evolution on a daily and decadal scale. Journal of Coastal Research, SI, 606–611. Vila-Concejo, A., Harris, D.L., Power, H.E., Shannon, A.M. & Webster, J.M. (2014) Sediment transport and mixing depth on a coral reef sand apron. Geomorphology, 222, 143–150. Wasserman, H.N. & Rankey, E.C. (2014) Physical oceanographic influences on sedimentology of reef sand aprons: Holocene of Aranuka Atoll (Kiribati), Equatorial Pacific. Journal of Sedimentary Research, 84, 586–604. Wood, R. & Oppenheimer, C. (2000) Spur and groove morphology from a Late Devonian reef. Sedimentary Geology, 133, 185–193. Woodroffe, C.D. (2003) Coasts. Cambridge University Press. Woolsey, E., Bainbridge, S., Kingsford, M. & Byrne, M. (2012) Impacts of cyclone Hamish at One Tree Reef: Integrating environmental and benthic habitat data. Marine Biology, 159, 793–803. Yamano, H., Miyajima, T. & Koike, I. (2000) Importance of foraminifera for the formation and maintenance of a coral sand cay. Coral Reefs, 19, 51–58.

8 Storm Clustering and Beach Response Nadia Sénéchal1,2 , Bruno Castelle1,2 and Karin R. Bryan3 1 Univ.

Bordeaux, UMR EPOC, Pessac, France UMR EPOC, Pessac, France 3 School of Science, University of Waikato, Hamilton, New Zealand 2 CNRS,

8.1

Introduction

Over the past decade, there has been an increasing interest in the impact of coastal storm groups on shoreline and beach dynamics (e.g. Ferreira, 2005, 2006; Vousdoukas et al., 2012; Loureiro et al., 2012; Splinter et al., 2014a; Coco et al., 2014; Karunarathna et al., 2014). One of the central questions is to determine whether or not there is a cumulative impact in the presence of coastal storm grouping, or in other words, if the combined erosion of a series of storms is larger than the sum of the average erosion of each individual storm (e.g. Morton et al., 1994; Lee et al., 1998; Ferreira, 2005, 2006) or not (e.g. Yates et al., 2009; Coco et al., 2014). This also raises the question of whether multiple sequential minor events with low return periods may be aggregated to cause the same impact as a single-event with a far higher return period (e.g. Cox & Pirrello, 2001; Ferreira, 2005, 2006; Callaghan et al., 2008; Karunarathna et al., 2014, Splinter et al., 2014a, b). For example, Figure 8.1 (from Karunarathna et al., 2014) displays the logistic exceedance return period contour for Long Reef Point storm data following the statistical procedure used by Callaghan et al. (2008). The lines represent exceedance return levels derived from the model and crosses represent measured return periods for single storms. The blue and red coloured circles represent the average and maximum significant wave heights, Hsmax , of each single storm contained in groups of sequential storms. Note that these grouped values are not exceptionally large and they have return periods of less than one year in the case of groups of two storms and five years in the case of groups of three storms. The authors then calibrated the XBeach model (see Chapter 10 for a description of the numerical model) with the characteristics of both single storms and sequences of storms representing each of the storm groups presented in Figure 8.1 to compare beach erosion differences (Figure 8.2 by Karunarathna et al., 2014). The erosion volume induced by groups of two and three storms, where the characteristics of each storm was Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

152

COASTAL STORMS: PROCESSES AND IMPACTS

200 1 yr

175

2 yrs

D (hrs)

150

5 yrs 10 yrs

125

25 yrs 50 yrs

100 75 50 25 4.5

6.5

8.5

10.5

12.5

14.5

Hsmax (m)

Figure 8.1 Logistic exceedance return period contour for Long Reef Point storm data. Crosses indicate measured storms; solid red circle – average of a group of two storms; hollow red circle – maximum of a group of two storms; solid blue circle – average of a group of three storms; hollow blue circle – maximum of a group of three storms. Black curves correspond to a range of storm return periods from one to fifty years (figure by Karunarathna et al., 2014).

Average ΔV (m3/m)

100 80 60 40 20 0 0

10

20 30 Storm return period (years)

40

50

Figure 8.2 Beach volume change against storm return period (black line with dots – single storms; blue solid line – average of a group of two storms; blue dotted line – maximum of a group of two storms; red solid line – average of a group of three storms; red broken line – maximum of a group of three storms) (figure by Karunarathna et al., 2014).

equal the average of the storms that comprised the group, corresponded to erosion volume induced by a single storm with a return period of two years or more (even though the average storm return period for these cases was less than a year, Figure 8.1, hollow circles). Erosion volumes induced by groups of two and three storms where the characteristics were equal to the maximum of the individual storms comprising the cluster (with the storm return period of less than one and five years, respectively, Figure 8.1, solid circles) correspond to erosion volume induced by single storms with return periods of ten or more years.

CH8 STORM CLUSTERING AND BEACH RESPONSE

153

These findings, including those of Ferreira (2005, 2006), Callaghan et al. (2008) and Splinter et al. (2014a) show that the erosion caused by a group of storms associated with short return periods can exceed the erosion of an single event with a longer return period. In the literature, there is a good agreement that the response to multiple events is not equal to the linear sum of the effects of each storm, but that it rather results from a function of exceeded thresholds and triggered feedback mechanisms (e.g. Masselink & van Heteren, 2014), resulting in non-linear behaviour. This can result in seemingly disproportionally-large impacts of moderate storms (e.g. Lee et al., 1998; Ferreira, 2005, 2006; Castelle et al., 2007; Furmanczyk et al., 2012) or the absence of the expected cumulative storm effect (e.g. Coco et al., 2014) making predictions a difficult challenge. Thus it is not possible to scale-up single-storm erosion studies into predictions of erosion caused by groups of storms. In this chapter, storm cluster genesis and definitions will be addressed in the first section; storm clusters can be defined in many different ways but the term generally refers to a sequence of coastal storm events separated by a short time interval. Then in the second section a short summary of the approaches and methodologies used to address storm clustering will be provided. In the last section, a summary and discussion will be provided on the key processes that should be included to more specifically address storm clustering impact on shoreline and beach morphology.

8.2

Storm clustering: Genesis and definitions

8.2.1 Genesis The wave climate on the world’s coasts is highly variable both spatially and temporally (e.g. Semedo et al., 2011) and so are also the extreme significant wave heights in the ocean; both of these have received increasing attention in the literature over recent decades (e.g. Betts et al., 2004; Keim et al., 2004; Izaguirre et al., 2010, 2011; Ruggiero et al., 2010; Reguero et al., 2013). In particular, increasing attention has focused on quantifying increasing storminess in a context of global climate change and determining the link with variations in the atmosphere and ocean systems (e.g. Allan & Komar, 2002; Donnelly & Woodruff, 2007; Arpe & Leroy, 2009; Ruggiero et al., 2010; Izaguirre et al., 2011). Because wind waves (including wind sea and swells) are primarily driven by wind friction on the ocean surface, many regions around the world often exhibit a strong seasonality in the degree of storminess and, in turn, in wave conditions (e.g. Butel et al., 2002; Méndez et al., 2008; Jonathan & Ewans, 2011). This is particularly evident in the case of the mid-latitude areas where storms form in the zone of major eddies in the atmosphere. A remarkable feature of the most extreme storms, such as those hitting Europe in 1990, 1999 and recently in 2014, is their tendency to occur in groups (or ‘clusters’), with storm peaks occurring in rapid succession separated by 2–3 days (e.g. Kvamsto et al., 2008; Vitolo et al., 2009). The various factors leading to the occurrence of storm clusters include (e.g. Pinto et al., 2014): an extended and intensified upper-level jet stream; Rossby wave breaking (an upper tropospheric wave which amplifies and overturns, e.g. Hanley & Caballero, 2012); steering from large-scale variability modes (like the North Atlantic oscillation, the ENSO); and secondary cyclogenesis over the Eastern

154

COASTAL STORMS: PROCESSES AND IMPACTS

North Atlantic (e.g. Mailier et al., 2006). Extra tropical cyclones may also be clustered in space and time because they form part of a coherent entity such as wavepackets (known as the Poisson cluster process). Over the last decade, studies have substantiated the concept of downstream development, on the basis of the observational evidence that synoptic-scale baroclinic eddies do exhibit group-velocity propagation along storm tracks in the form of wavepackets (e.g. Lee & Held, 1993; Swanson & Pierrehumbert, 1994; Rao et al., 2002). In the mid-Atlantic, variations in storm activity can also be influenced by southward intrusion and strengthening of the jet stream (Betts et al., 2004). However, such patterns may be complicated, for example, by the existence of a climate ‘see-saw’ like the one observed between western Greenland and northern Europe (e.g. Dawson et al., 2004). Thompson & Barnes (2014) recently provided evidence that southern hemisphere mean atmospheric wave activity exhibits periodic variability over ∼20–30 days time scales. These are linked to the observed 20–30 days periodicity of the large-scale southern hemisphere atmospheric circulation consistent with two-way feedbacks between the baroclinicity and the eddy fluxes of heat. This variability in the climate system thus induces variations in both the generation and amplitude of wave activity throughout much of the southern hemisphere.

8.2.2 Definitions While storm and serial clustering are generally well-defined from a meteorological point a view, there is no broadly-accepted definition in coastal erosion studies and coastal storm clusters can be defined in many different ways (e.g. Lee et al., 1998; Ferreira, 2005; Callaghan et al., 2008; Vousdoukas et al., 2012; Coco et al., 2014). Although large-scale meteorological patterns may affect the offshore wave climate, the propagation of waves towards the shore and associated transformations are a key process in assessing the impact of the storm on the shore (e.g. Cooper et al., 2004; Regnauld et al., 2004; Callaghan & Wainwright, 2013). Because of the lack of clear definition, even the term ‘cluster’ is still nowadays not commonly used in the nearshore community. Authors might sometimes use the term storm groups (e.g. Lee et al., 1998; Ferreira, 2005, 2006; Loureiro et al., 2012), consecutive or sequence of storms (e.g. Vousdoukas et al., 2012; Coco et al., 2014; Castelle et al., 2015), with only a few publications clearly mentioning the term storm clustering (e.g. Karunarathna et al., 2014; Splinter et al., 2014b; Senechal et al., 2015). In principle, coastal storm clusters (or groups) refer to a sequence of coastal storm events separated by a short time interval. Clearly, the definition depends on the explanation of a coastal storm (which is usually defined as an event in which the wave height exceeds a threshold) and the definition of a ‘short’ time interval. For example, Callaghan et al. (2008) choose a threshold of 3 m for analysis of the wave climate of Australia (also used in Lord & Kulmar, 2000; Kulmar et al., 2005; Karunarathna et al., 2014), and Short & Trenaman (1992) use 2.5 m. Ferreira (2005), on the Portuguese west coast, used 6 m to insure that they only considered storms that were responsible for significant erosion as single events. To remove subjectivity, the threshold might have been determined using the probability distribution of the wave height, for example the 99.5% exceedance level (Luceno et al., 2006) or the 95% exceedance level (Masselink et al., 2014). In terms of separation interval (also called inter-exceedance, Fawcett & Walshaw (2008) or interarrival time, Salvadori (2014)), anything shorter than three days is considered to be

CH8 STORM CLUSTERING AND BEACH RESPONSE

155

part of the same storm (Luceno et al., 2006), or six hours in the case of Li et al. (2014). Also arbitrary is the maximum time for which two storms are considered part of the same cluster, for example nine days in Karunarathna et al. (2014), 14 and 21 days in Ferreira (2005) and up to 39 days in Lee et al. (1998). This large variability in definitions is essentially driven by the ‘morphological’ definition of storm cluster that states that storm clusters occur when the interval between two consecutive storms is shorter than the beach recovery period for individual storms (e.g. Morton et al., 1995). This definition is still widely used in coastal erosion studies (e.g. Loureiro et al., 2012; Senechal et al., 2015). In this approach, clusters are defined with respect to the erosional response rather than the hydrodynamic forcing, in which case the time interval between clustered storms is shorter than the recovery time of the beach (which depends on parameters such as grain size, sediment supply and storage, as well as beach history and storm intensity). This is also called ‘event merging’ (Callaghan et al., 2008). This recovery period is thus highly site-specific and yet not clearly addressed in the literature. A definition of clusters based on beach recovery time supposes that beach recovery occurs and yet this might not be the case, especially if wave conditions following the energetic events do not allow onshore sediment transport (e.g. because of wave incidence angle or waves not being energetic enough, Ruessink et al., 2007). Figure 8.3 and Table 8.1 provide an overview of the sensitivity of coastal storm serial clustering classification due to differences in definition. Data clearly indicate that both the storm definition used and the interval applied may strongly modify the number and duration of storm clusters as well as the number of storms within a cluster. Some authors use a much more strict definition of cluster (following the meteorological approach), in which the storms not only must be close together, but also must be correlated (in which case the probability of a storm occurring after another storm is not random). In this case, a cluster is defined using the probability of inter-exceedance times (Ferro & Segers, 2003). Such statistical dependence of clustered events only occurs when the interval time is very short (Fawcett & Walshaw, 2008). There is a large body of literature on the statistics of extreme events, which uses methods such as the ‘blocks’ and ‘runs’ method to investigate sequences of extremes (for more on this refer to, e.g.

Figure 8.3 Time series of significant wave height (black) and peak wave period (blue) during the winter of 2013/2014 measured in about 50 m depth offshore of Truc Vert beach (SW France). The horizontal colour lines indicate the occurrence of storm and storm clusters using six different definitions.

156

COASTAL STORMS: PROCESSES AND IMPACTS

Table 8.1 Storm and storm cluster characteristics computed from the storm sequence shown in Figure 8.3 using six storm cluster definitions. Threshold in Hs

dts (days)

Nb of storms

Nb of storm clusters

Avg cluster duration (days)

Maximum storm cluster duration (days)

Avg nb of storms within a cluster

Hs,99.5% Hs,99.5% Hs,99.5% Hs,99.5% Hs,99.5% Hs,99.5%

2 days 3 days 9 days 2 days 3 days 9 days

7 3 1 5 2 1

1 3 3 5 5 3

3.25 3.8 7.62 3.78 5.56 13.97

3.25 4.33 15.79 5.75 9.92 18.92

2 2 2.66 2.8 3.4 6.33

= 6.22 m = 6.22 m = 6.22 m = 5.54 m = 5.54 m = 5.54 m

Smith & Weissman, 1994), which may be adopted by coastal scientists in the future. However, this approach may not be useful in areas that are exposed to different sources of low pressure systems that will consequently not necessarily be correlated. From the authors’ point of view, definitions of storm clusters based on recovery time, even if they make sense, are difficult to address, essentially because of the lack of comprehensive datasets assessing beach recovery concepts and processes, and because the rate of recovery depends on the beach history (i.e. the erosional state prior to the occurrence of the storm cluster). This approach is also hard because of the triggering effect of temporal scales: storms are generally occurring over short-term periods and some areas are experiencing long-term recession (suggesting no recovery) not only associated with storms. Wave and storm statistics seems to be the most relevant method to develop a robust storm cluster definition despite the lack of universal definition of coastal storms.

8.3 Approaches used to assess storm clustering impact on coasts Assessing storm clustering impacts on coasts implies that we are able to study the impact of each storm event within the cluster and not just consider the cluster as a ‘single’ large event. As highlighted in the previous section, the definition of a cluster is imprecise. However, in most cases, the cluster designates a rapid succession of energetic events that make the usual approaches used to assessing ‘single’ storms inappropriate.

8.3.1 Data collection Field data collections are an important source of information to assess storm clustering and coastal impact. One can generally define three kinds of field data collections: (1) sedimentary archives (e.g. Donnelly et al., 2001, 2004); (2) observations collected before and after the energetic events consisting essentially of visual observations and topographic surveys (e.g. Lee et al., 1998; Birkemeier et al., 1999; Kish & Donoghue, 2013; Smithers & Hoeke, 2014; Castelle et al., 2015); and (3) intensive field data collected during the energetic events consisting of high frequency topographic surveys

CH8 STORM CLUSTERING AND BEACH RESPONSE

157

using DGPS methodology coupled to either Eulerian and/or Lagrangian hydrodynamic sensors and remote sensing (Senechal et al., 2011; Coco et al., 2014; Almeida et al., 2015; Earlie et al., 2015). While sedimentary archives allow a longer time period to be considered, intensive field data (which are difficult to collect) provide the only way to study hydrodynamic and sediment transport processes in the presence of natural conditions, including the processes of between-storm recovery. Conversely, during field experiments, field data are generally collected over a limited area (generally < 1 km alongshore) and do not allow large-scale alongshore variability to be accounted for, especially in the presence of non-uniform coasts. On the whole, there is still a lack of comprehensive field data to assess storm cluster impact. Most of the intensive field experiments (e.g. Thornton et al., 1996; Gallagher et al., 1998; Ruessink et al., 1998, 2001; Aagaard et al., 2005; Masselink et al., 2008; Bruneau et al., 2009; Almeida et al., 2015; Ogawa et al., 2015) were conducted under offshore significant wave height not exceeding 5 m. Only a few datasets were collected during a rapid succession of high to extremely high-energy waves (e.g. Senechal et al., 2011). Remote sensing of shoreline erosion is commonly used, and if the storm clusters are not too close together, accurate measurements of both the alongshore and cross-shore variations of shoreline response can be made (e.g. Jimenez et al., 1997; Ciavola et al., 2007; Almar et al., 2010; Kish & Donoghue, 2013; van der Lageweg, 2013). However this approach alone, as well as the more recent lidar (e.g. Revelle et al., 2002; Stockdon et al., 2002; Sallenger et al., 2003; List et al., 2006) and optical satellite imagery (e.g. Castelle et al., 2015) approaches usually do not allow for storm cluster assessments because they generally address a storm cluster as a ‘single’ large event. Recently Unmanned Aerial Vehicles (UAV) have been developped and used, for example, to study wave runup by identifying the maximum wave runup (e.g. Casella et al., 2014). The use of UAV, even if restricted to a smaller spatial coverage, allows rapid deployment and thus data collection between two consecutive storms occuring within the same storm cluster. In recent decades, low-cost, long-term optical video measurements of the nearshore have been increasingly popular. Time exposure images have been used extensively for detecting morphological patterns and shoreline proxies, even at the cluster ‘time scale’ (e.g. Almar et al., 2010; Vousdoukas et al., 2012; Masselink et al., 2014; Senechal et al. 2015, Figure 8.4). In this figure, time exposure images collected under a sequence of energetic winter storm events occuring over a six-week window showed the straightening of the outer sandbar (panels a to d) during the most energetic event, followed by the reconstruction of the outer bar cresecentic patterns under less severe conditions (panels e and f).

8.3.2 Numerical models The detailed morphological evolution of beaches during storms involves many nonlinear processes and complex hydrodynamic-morphodynamic feedbacks, making the simulation of storm-driven beach erosion, and even more storm-cluster-driven beach erosion, a challenging task. On most of open coasts, storm-driven beach and dune erosion is mostly a process of cross-shore sediment transport, with longshore processes typically acting on longer timescales (e.g. Hansen & Barnard, 2010), except in the vicinity of hard structures such as groins and headlands (Cooper et al., 2004).

158

COASTAL STORMS: PROCESSES AND IMPACTS

bays

1000

horns

800 X (m)

600 400

(a)

200 1000

800 X (m) 600 400 (b)

200 1000

800 X (m) 600 400 (c)

200 1000

800 X (m) 600

transverse intermediate bar

400 (d)

200 horns

1000 800 X (m) 600 400 (e)

rips

200 horns

1000 800 X (m) 600 400

rips

200 (f)

1000

500

–500

0

–1000

Y (m)

Figure 8.4 Timex collected under consecutive storm conditions showing (from (a) to (f)) the straigthening and the reconstruction of the outer bar crescentic patterns (from Almar et al., 2010).

To simulate storm-cluster-driven erosion, one can resort to two types of numerical models: (1) process-based and fully coupled (wave-current-sediment transport-bottom evolution) models that rely on a description of the underlying physical processes (see Chapter 10), and allow investigation of the erosive effect of storm clusters (e.g.

CH8 STORM CLUSTERING AND BEACH RESPONSE

159

Figure 8.2 from Karunarathna et al., 2014); or (2) data-driven models. It is noteworthy that all these models need, to different degrees, field data for calibration/training. Deterministic or probabilistic data-driven models combine statistical techniques and previous measurements at a given site to decipher the overall behaviour of the morphological system and provide beach change projections given a forecast of wave conditions. Statistical techniques include, non-exhaustively, neural networks, empirical mode decomposition methods and methods based on signal covariance structure such as empirical orthogonal functions and Canonical correlation analysis (amongst others, ̇ Larson et al., 2000; Rózyoski et al., 2001). It is noteworthy that the data-driven model skill depends heavily on the quantity and quality of the data available at a given site. One can therefore apply a data-driven approach to address storm-driven erosion only if long-term, high-quality and high-frequency observations on the timescale of years to decades are available, including a large and representative range of storm and storm-cluster wave conditions, from which projections can be made. Data-driven models applied to storm cluster impacts can be divided into two main categories: (1) hybrid and (2) probabilistic models. Hybrid models rely on a number of simplifying assumptions in the governing equations and typically only retain key processes governing the morphological evolution of beaches. These models have been commonly shown to provide more accurate large-scale (in both time and space) coastal evolution projections than complex process-based models, in which mis-specifications of the physics and boundary conditions typically cascade up through the scales, resulting in an inescapable build-up of errors and unreliable large-scale simulations. Within the hybrid model family, equilibrium-based semi-empirical shoreline models (e.g. Miller & Dean, 2004; Yates et al., 2009; Davidson et al., 2010, 2013; Long & Plant, 2012; Castelle et al., 2014; Splinter et al., 2014b) have been increasingly used to address shoreline response on the timescale of a few years on cross-shore transport dominated beaches, comprising the short-time shoreline response during storms and storm clusters. In essence, these models consider that shoreline response is driven by both the evolving disequilibrium state of a beach through time and the rapidly-varying wave forcing by prevailing wave conditions. Contrary to the above deterministic models, probabilistic models usually embed at least one deterministic reduced-complexity or process-based model (e.g. Callaghan et al., 2013, for an application with a process-based model), which is run many times to build a Monte Carlo simulation or a Bayesian network that can further provide probabilistic predictions of shoreline change in response to different wave condition scenarios. This also means that the skill of a probabilistic beach erosion model depends on an accurate calibration/training of the determinist model. Such models can include the joint probability of consecutive storm events, and so can simulate the impact of clusters.

8.4

Beach response to storm cluster

8.4.1 Bar dynamics under storm clustering Lee et al. (1998) analysed ten years of nearly bi-weekly topographic surveys to study the role of storm groups in controlling profile evolution at Duck (USA East Coast). In particular they showed that a storm group with a similar or even smaller wave power

160

COASTAL STORMS: PROCESSES AND IMPACTS

than an individual storm caused the offshore migration of a pre-existing transitional or outer bar to a more seaward position, combined with its substantial growth. In contrast, their data set indicated that no individual storm, even though they had the same wave average power, caused those kinds of changes. One explanation suggested by the authors for why the storm groups have such a significant impact is that the first storm destabilizes the profile by resuspending and transporting sediment across the profile. These newly-deposited sediments along the beach profile are loosely consolidated and can thus be easily eroded. With the second storm arriving in quick succession, the profile is easily changed compared to the case where sand had time to compact. Similarly to the observations of Lee et al. (1998), Van Enckevort & Ruessink (2003) and Vousdoukas et al. (2012) reported that the temporal scales of nearshore bar position fluctuations were less related to individual events than to storm series. The dynamic of the outer bar is a key process in beach response to storm clusters because of possible morphological feedback: the outer bar generally acts as a buffer of sediment or prevents the beachface from experiencing intense wave breaking. Many studies have shown that the cumulative impact of storm clusters can outweigh that of a single storm, with a return period of tens of years (e.g. Birkemeier et al., 1999; Ferreira, 2005; Castelle et al., 2007; Karunarathna et al., 2014), which may be caused by the dynamics of the nearshore bars. Past definitions of storm cluster have highlighted the importance of recovery time, and the initial state of the beach. In particular, the way how the initial state of the beach between the first storm of a cluster and subsequent storms might impact the erosional response, shoud be further investigated.

8.4.2 Morphological feedback Splinter et al. (2014a) used the XBeach model (see Chapter 10) and showed that storm sequencing does not significantly affect the total beach erosion volume, but that the volume eroded during individual storms is influenced by the antecedent state of the beach (i.e. prior cumulative erosion). Castelle et al. (2007) analysed observations from the Gold Coast (Australia) and showed that the increased beach erosion rate observed during the two last, much less energetic, storms of a sequence of closely-spaced storms could be explained by the outer bar decay driven by the first, severe, storm of the sequence. The authors hypothesized that the decayed outer bar did not provide any significant protection during the subsequent two storm wave events, which resulted in intense erosion. This outer-bar decay was subsequently shown to be part of an intrinsic Net Offshore Migration (NOM) cycle at this site (Ruessink et al., 2009), corroborating earlier work (Shand et al., 2004) on the link between rapid, but unrelated to wave conditions, beach erosion and NOM cycles. Senechal et al. (2015) used analysis of the nearly daily dynamics of the shoreline and the inner bar at Biscarrosse Beach, an open sandy intermediate beach on the south-west part of the French Atlantic coast, to report another form of impact of surf zone bars on shoreline evolution. During the first storm associated with significant wave height > 6 m, the shoreline remained relatively stable (or experienced rapid post-storm recovery – due to wave conditions, shoreline extraction was not possible during the apex of the storm), while the inner bar experienced rapid up-state transition. A few days later, the beach experienced energetic conditions but weaker than the ones experienced a few days earlier (wave height ∼3–4 m) and the

CH8 STORM CLUSTERING AND BEACH RESPONSE

Surf-zone

161

Sub-aerial profile Rapid erosion/recovery

No bar or degenerated bar

Rapid erosion

Beach-face erosion slows down

Surf-zone sediment redistribution

Storms

Consecutive storms Increased storm intensity,

Nearshore bar growing

Response slows down Storms

Further erosion

Well defined nearshore bar Enhanced wave breaking/attenuation

Recovery

(c) PHASE 3: Equilibrium state

No storm

Morphologically active beach-face

Decayed nearshore bar

Gradual surf-zone sediment redistribution

(a) PHASE 2: Transition to dissipative state

No storm

Recovery

(a) PHASE 1: Rapid profile pivoting

Enhanced wave dissipation

Beach-face in equilibrium

Equilibrium

Well defined nearshore bar, surf-zone in equilibrium

No storms

Long term recovery

surge, or tidal level

Figure 8.5 Conceptual model describing beach morphological evolution phases of steep-sloping beaches during consecutive storms: (a) beach-face shows rapid erosion/recovery; (b) the morphological change at the (eroded) sub-aerial beach decelerates and adaptation of the surf-zone bathymetry becomes more important; (c) the beach reaches equilibrium, becomes more resilient to wave forcing and is vulnerable mostly to increased mean water levels (figure by Vousdoukas et al., 2012).

shoreline experienced rapid erosion. Image analysis indicates that during this energetic period, waves were not breaking on the inner bar around mid-tide to high tide. Vousdoukas et al. (2012) proposed a conceptual model of the morphological response of steep sloping beaches under a series of storm events (Figure 8.5). Their model shows that if no bar or a degenerated bar is present in the surf zone then the sub-aerial profile undergoes rapid erosion and the beach profile evolves to a more dissipative state, including nearshore bar growth. Then, morphological change is restricted to the lower (submerged) profile, where most wave attenuation takes place and the sub-aerial profile response slows down. They concluded that under a series of storm events, the antecedent morphological state may initially be the dominant controlling factor of beach response. Some models also suggest a similar negative feedback mechanism, with the sand eroded from the beachface transferred to the inner surf zone and therefore increasing wave dissipation by the time waves reach the swash zone. This makes the storms within a storm cluster less and less effective in eroding the beach over time. Similarly, in a study establishing storm thresholds for the Spanish Gulf of Cadiz coast, Del Rio et al. (2012) indicated that no cumulative effect was generally observed in this area because during the first storms the beaches increased their dissipativeness, thus facilitating ‘profile self-protection’. Nearshore sandbar dynamics is not the only phenomom affecting shoreline erosion during storms and storm clusters, as the localized beach and dune erosion also play a role. Many authors (e.g. Dalon et al., 2007; Thornton et al., 2007) have shown that the morphology of the nearshore sandbar prior to the storm(s) can act as a morphological template for beach and dune erosion with the megacusp embayments, where maximum erosion occurs, which are aligned with rip channels. During high tides and storm waves, dune erosion occurs in the embayments where the beach is both the narrowest and

162

COASTAL STORMS: PROCESSES AND IMPACTS

lowest. More recently, Castelle et al. (2015) indicated that both the antecedent outer sandbar morphology and storm wave characteristics, including period and angle of incidence, govern patterns of beach and dune erosion along an open multiple-barred sandy coast during the sequence of severe storms shown in Figure 8.6. They showed that in early January 2014 the outstanding shore-normal incident storm swell Hercules hitting the western Atlantic coasts, with significant wave height and peak wave period reaching 9.6 m and 22 s, respectively, triggered the formation of localized megacusp embayments along the French soutwest linear sandy coasts that persisted throughout the storm clusters in January, February and March 2014 (Figure 8.6). The persistence and cumulative effect of megarips during storm series at three embayed beaches have also been reported by Loureiro et al. (2012). Their observations showed that extreme erosion occurred when megarips and feeder channels persisted during successive storms and thus promoted continued erosion and offshore sediment export. Their observations also showed that once initiated, megarip channels can persist for several months even under calm conditions, acting as conduits for seaward sediment export under non-storm conditions. The maintenance of such rip circulation systems, driven by morphodynamic feedback, reduces beach recovery ability until the rip-neck and feeder are infilled.

8.4.3 The dynamic equilibrium concept The concept of an equilibrium beach shoreline and profile (e.g. Wright & Short, 1984; Yates et al., 2009; Davidson et al., 2013) significantly changes the way in which we consider the erosion caused by a cluster of storms. For example in the conceptual model proposed by Vousdoukas et al. (2012) (Figure 8.5), the final stage is the equilibrium state. Additional erosion will probably occur only if wave energy and/or water levels exceed previous conditions. Thus, a small storm followed by a large storm can have quite a different effect from a large storm followed by a small storm, even when the mean energy delivered to the coastline over the cluster time period is the same. Yates et al. (2009) show an example where different sequencing of a wave record with the same mean energy causes erosion in one case and accretion in the other case. The reason for this is that the first storm leaves the beach in a different state depending on its energy level. For example, for a given wave condition a shoreline that is in an accreted state will erode more than a shoreline that is in a more eroded state. Therefore the second storm of the cluster will respond in a way that depends on the size of (and response to) the first storm. In temperate environments the first winter storms typically drive the most pronounced erosion events because the wave energy disequilibrium is large (e.g. Yates et al., 2009; Castelle et al., 2014). Given that during storm clusters the first rank storms become part of the recent history, these models therefore suggest that closely-spaced storms are increasingly ineffective in eroding the beach over time as the beach reaches a new equilibrium with the prevailing high-energy wave conditions. As such, equilibrium-based semi-empirical shoreline models could be successful in explaining the surprisingly small erosion driven by the storm clusters studied by Coco et al. (2014) that occurred at the end of the winter 2008 at Truc Vert beach. During storm wave conditions driving beach and dune erosion, a large amount of sediment is transferred from the beach face to the inner surf zone where a terrace forms rapidly. Storm

CH8 STORM CLUSTERING AND BEACH RESPONSE

163

Figure 8.6 (a) Aerial photograph of the Gironde coast (SW France) on 7 March 2014, showing megacusp embayments cutting the dune (dune foot indicated by the red dashed line) with, at this section of coast, a mean alongshore and cross-shore length scale of 500 m and 20 m, respectively (photo Julien Lestage). (b, c) LANDSAT satellite images of the Gironde coast (b) prior to the winter of 2013/2014 on 10 July 2013 for waves with Hs≈0.6 m and Tp≈8.8 s and (c) after the winter of 2013/2014 on 23 March 2014, for waves with Hs≈3.6 m and Tp≈12.9 s. In (b, c) the shoreline (dune foot proxy) measured with the ATV on 3–4 April 2014 is superimposed with colour bar indicating the deviation from mean shoreline in meters. The orange arrows in (c) show the location of rip current occurrence during storm waves clearly facing megacusp embayments. The grey line in (b, c) indicates the location of hard coastal structures in the coastal town of Lacanau.

164

COASTAL STORMS: PROCESSES AND IMPACTS

waves subsequently dissipate more energy through breaking across this terrace, with a reduction in erosion rate. As a result, the cumulative impact of the subsequent storm does not accelerate the erosion rates. Aagaard et al. (2005) analysed data collected in the swash and inner surf zone through successive tidal cycles and in presence of three large consecutive storm events. They reported that gently sloping shoreline salients were highly stable in position and displayed only minor slope adjustments through the storm periods. This is consistent with previous observations, which reported that at times the beach can be surprisingly resilient on an intra-storm time scale and the intertidal zone slope can remain almost constant over storm events (e.g. Aagaard et al., 1998; Coco et al., 2014). From a morphological perspective and in the course of storm clustering, this could be attributed to the equilibrium concept. However the data collected by Aagaard et al. (2005) indicate that the subdued net beach response results from offshore sediment export occurring under surf zone conditions at high tide being compensated by onshore sediment transport at low tide under dominant swash zone processes. Such temporally varying, tidally-modulated sediment transport rates and direction are suggested as a possible mechanism to maintain a quasi-equilibrium intertidal beach slope.

8.4.4 Water level The water level and in particular the tide effect has been addressed in the Storm Impact Scale model proposed by Sallenger (2000) (e.g. Stockdon et al., 2007; Plant & Stockdon, 2012; Masselink & van Heteren, 2014) that is often used to predict storm impacts on beach and dune systems. Storm tide, defined as the absolute water level associated with a storm, as measured relative to a fixed datum such as mean sea level, will certainly in turn govern the Storm Impact Scale. Monitoring frontal dune erosion and accretion on the Sefton coast in northwest England over the past 50 years by Pye and Blott (2008) has revealed that relatively high dune erosion rates at Formby Point 1958–1968 were associated with a relatively large number of storm tides. On the other hand slower erosion at Formby, and even relatively rapid accretion in areas to the north and south, were reported during the 1970s and 1980s, when the frequency of major storm tides decreased. In the presence of storm clusters, the positive storm surge may last for several days and the probability that it coincides with a high tide or even brackets several tidal cycles, including a spring high tide, is enhanced. For example, Vousdoukas et al. (2012) concluded that the observed morphological changes during consecutive storm events, suggest that hydrodynamic forcing, and especially the tide and the storm tide, control storm impact (Figure 8.5). Del Rio et al. (2012) indicated that an example of exceptionally severe effects of storm group for the Spanish Gulf of Cadiz coast was associated with the long duration of the event and the coincidence of some storm peaks with spring tide conditions leading to an unusual cumulative effect for the area. Coco et al. (2014) analysing a sequence of daily beach survey over one month in the presence of a cluster of storms concluded that the first largest measured erosive events on the upper beach face (∼25 m3 /m) was associated with very energetic conditions (S2 event, a storm with an estimated return period of ten years) occurring during spring tide conditions (Figure 8.7, S2 event). In particular, erosion of the embryo dunes

CH8 STORM CLUSTERING AND BEACH RESPONSE

S2

S3

165

S4

Alongshore average position (m) Hs (m)

10 5 0 high tide mean waterline position

180 0.0 160 0.5 140 1.0 1.5 120 2.0 2.5 3.0 100 3.5 4.0 4.5 80 5.0 28/02

07/03

14/03

21/03

28/03

04/04

11/04

Time (dd/mm) Figure 8.7 Alongshore variability beach response. Top panel: Offshore Hs. Bottom panel: Each black line represents the mean alongshore position for different contour lines (each contour line is labelled on the left). Vertical bars indicate the alongshore variability denoted by one standard deviation of the mean contour line position. The blue line indicates the high tide mean waterline position as measured from a pressure sensor on the inner sandbar. Offshore direction is upwards (modified from Coco et al., 2014).

(contour lines above 3 m) was reported during the S2 and was associated with the elevated water levels. The long duration and smaller magnitude of S4 resulted in erosion of the upper part of the beach, but with erosive events being correlated with the high tide mean waterline positions. Finally, the second largest erosive event observed on the upper beach face (5–7 April) was associated with very low waves (Hs < 1 m), but spring tide conditions and concerned contour levels up to 4 m. This data set led the authors to the conclusion that storms that can potentially cause significant erosion in terms of Hs had a limited effect on the morphology, because among other factors, the large wave height was coupled to neap tides. On the other hand, ‘recovery’ conditions in terms of Hs had an important effect on the morphology because of spring tide conditions and wave incidence.

8.4.5 Recovery periods Recovery periods have received much less attention in the literature, even being neglected in most studies evaluating the cumulative impact of storm clusters on beach erosion (e.g. Splinter et al., 2014a; Karunarathna et al., 2014). However, storm

166

COASTAL STORMS: PROCESSES AND IMPACTS

clustering depends both on the storm definition and the definition of a ‘short ‘interval’. And yet, the definitions used for a ‘short interval’ are either an interval too short to ensure beach recovery (e.g. Callaghan et al., 2008; Karunarathna et al., 2014; Senechal et al., 2015) or any interval of less than three days. Studies have shown that the initial recovery from storm-induced erosion can be extremely fast (e.g. List et al., 2006; Roberts et al., 2013), but complete recovery may also last for years if the foredunes have been eroded (Birkemeier, 1979; Wang et al., 2006; Suanez et al., 2012; Bramatao et al., 2012; Houser et al., 2015). Therefore, by definition, the so-called recovery time (generally defined as the calm period between two consecutive storms) will always be shorter than the time needed for complete recovery. Recovery from any storm is obviously a key component of preconditioning for the next storm, but might also in some cases influence the long-term response of the beach (e.g. Houser et al., 2009, 2015; Roberts et al., 2013). Hence Forbes et al. (2004) suggested that large longshore and interdecadal variance observed along barrier shores in the southern Gulf of St Lawrence might reflect recovery from an episode of widespread overwash prior to 1935, possibly initiated by intense storms or groups of storms in the latter half of the 19th century. Roberts et al. (2013) analyzed 18 almost monthly beach-profile surveys at 46 locations and found that the timescale of the beach cycle relates to the frequency and intensity of storm impact and duration of the inter-storm recovery instead of simple seasonality. However, the question arising here is not really if complete recovery will occur during this ‘recovery period’, but rather what can be considered as initial recovery and which duration should be considered. Morton et al. (1995) defined four categories of post-storm response: continued erosion, partial recovery, complete recovery and over-recovery. Here the key question is obviously to define the partial recovery, but also if short-term partial recovery can be considered as positive regarding the response of the beach to a cluster of storms. Indeed recovery is generally been considered as a positive process for the beach; however, some authors (e.g. Del Rio et al., 2012) indicate that the lack of beach recovery between two storms within the same storm group may also explain why the effects of several medium-energy storms are not higher than those of a single higher-energy event. Beach recovery generally takes place under falling wave conditions. However, recent field data has provided evidence of beach erosion even under calm periods following severe storm clustering conditions (e.g. Coco et al., 2014) or persisent shoreline positions and/or beach profile under energetic conditions (e.g. Aagaard et al., 2005; Quartel et al., 2008; Senechal et al., 2009; Coco et al., 2014). Almar et al. (2010), using video data, reported a shoreward propagation accretionary sand wave under severe storm conditions (Hs > 8 m). On the other hand, the maintenance of rip circulation systems under calm conditions generated during a series of storms, might reduce beach recovery ability even under non-storm conditions (e.g. Loureiro et al., 2012). The rate of beach recovery between consecutive storm events can also depend on tidal range. Morphological relaxation of sandbar systems under storm conditions has also been reported by Sedrati & Anthony (2007) during an intensive field experiment conducted on a macro-tidal beach, and more recently by Masselink et al. (2014) based on long-term observations (15 years). Both reported slow morphological responses which were explained by extended relaxation times attributed to the large tidal range at the study site where the processes might be only active for part of the tidal cycle.

CH8 STORM CLUSTERING AND BEACH RESPONSE

167

Indeed, the recovery periods within the same cluster will certainly primarly depend on the hydrodynamic conditions that should be considered relative to the previous conditions (including storms and moderate energy conditions). For example, waves and tide conditions may or may not favor onshore sediment transport; but recovery will also depend on the beach geomorphology and geologic settings (e.g. Splinter et al., 2011; Gallop et al., 2012; Anthony, 2013); the local availability of sediment that also depends on the Storm Impact Scale and where the eroded sediments have been deposited (Forbes et al., 2004; Masselink & van Heteren, 2014); the usual (modal) beach state will certainly also be relevant, low-energy coasts being more vulnerable than their higher-energy counterparts (e.g. Qi et al., 2010; Yu et al., 2013; Masselink & Van Heteren, 2014); and finally the sand transfer with the foredunes (e.g. Hesp, 1988; Suanez et al., 2012; Anthony, 2013) by aeolian sand transport and interactions with the vegetation (e.g. Pries et al., 2008; Priestas et al., 2010; Suanez et al., 2012; Seablom et al., 2013).

8.5

Conclusions

The erosional response to clusters of storms is generally much larger than the sum impact of the component storms that make up the cluster. The reasons why this occurs are several: (1) the equilibrium response of the beach depends on the antecedent morphological conditions; (2) the beach may not have adequately recovered between events within clusters; (3) the profile, although recovered, might not have compacted between storms; (4) the configuration of offshore sand bars may not be the same during storms, and so the protective shield against storms might be different; (5) the tidal conditions may be different between storms that comprise the cluster. A consequence of the sensitivity to storm clustering is that it is not possible to use weekly or monthly-averaged wave conditions to model shoreline change, even when the verification data (often profile data) may only be available at monthly or longer time intervals. Empirical or hybrid models that require training datasets clearly need a range of storm sequences for training, making it necessary for longer training datasets. Nowadays, there is a lack of observational studies focusing on beach recovery under storm clustering conditions, probably also because we still do not understand when recovery begins and finishes, even in presence of a single storm. It is critical that our future research is better able to quantify the widely-varying differences between storm-recovery rates that appear to be the key to understanding the impact of clustering.

References Aagaard, T. (1998) Rhythmic beach and nearshore topography: Examples from Denmark. Geografisk Tidsskrift, 88, 55–60. Aagaard, T., Kroon, A., Andersen, S., Sorenson, R.M., Quartel, S. & Vinther, N. (2005) Intertidal beach change during storm conditions; Egmond, The Netherlands. Marine Geology, 218, 65–80. Allan, J.C. & Komar, P.D. (2002) Extreme storms on the Pacific Northwest Coast during the 1997–98 El Niño and 1998–99 La Niña. Journal of Coastal Research, 18 (1), 175–193.

168

COASTAL STORMS: PROCESSES AND IMPACTS

Almar, R., Castelle, B., Ruessink, B.G., Sénéchal, N., Bonneton, P. & Marieu, V. (2010) Twoand three-dimensional double-sandbar system behaviour under intense wave forcing and a meso-macro tidal range. Continental Shelf Research, 30 (7), 781–792. Almeida, L.P., Vousdoukas, M.V., Ferreira, O., Rodrigues, B.A. & Matias, A. (2012) Thresholds for storm impacts on an exposed sandy coastal area in southern Portugal. Geomorphology, 143–144, 3–12. Almeida, L.P., Masselink, G., Russell, P.E. & Davidson, M.A. (2015) Observations of gravel beach dynamics during high-energy wave conditions using a laser scanner. Geomorphology, 228, 15–27. Anders, F.J. & Byrnes, M.R. (1991) Accuracy of shoreline change rates as determinated from maps and aerial photographs. Shore and Beach, 59, 17–26. Anthony, E.J. (2013) Storms, shoreface, morphodynamics, sand supply, and the accretion and erosion of coastal dune barriers in the southern North sea. Geomorphology, 199, 8–21. Arpe, K. & Leroy, S.A.G. (2009) Atlantic hurricanes – Testing impacts of local SSTs, ENSO, stratospheric QBO – Implications for global warming. Quaternary International, 195, 4–14. Betts, N.L., Orford, J.D., White, D. & Graham C.J. (2004) Storminess and surges in the south-western approaches of the eastern North Atlantic: The synoptic climatology of recent extreme coastal storms. Marine Geology, 210, 227–246. Birkemeier, W.A. (1979) The effects of the 19 December 1977 coastal storm on beaches in North Carolina and New Jersey. Shore and Beach, 47, 7–15. Birkemeier, W.A., Nicholls, R.J. & Lee, G. (1999) Storms, storm groups and nearshore morphologic change. In: Proc. Coastal Sediments ’99. ASCE, New York, pp. 1109–1122. Bramato, S., Ortega-Sãnchez, M., Mans, C. & Losada, M.A. (2012) Natural recovery of a mixed sand and gravel beach after a sequence of short duration storm and moderate sea states. Journal of Coastal Research, 28, 89–101. Bruneau, N., Castelle, B., Bonneton, P., Pederos, R., Almar, R., Bonneton, N., et al. (2009) Field observations of an evolving rip current on a meso-macrotidal inner bar and rip morphology. Continental Shelf Research, 29, 1650–1662. Butel, R., Dupuis, H. & Bonneton, P. (2002) Spatial variability of wave conditions on the French Atlantic Coast using in-situ data. Journal of Coastal Research, SI36, 96–108. Callaghan, D.P., Nielsen, P., Short, A. & Ranasinghe, R. (2008) Statistical simulation of wave climate and extreme beach erosion. Coastal Engineering, 55 (5), 375–390. Callaghan, D.P., Ranasinghe, R. & Roelvink, D.A. (2013) Probabilistic estimation of storm erosion using analytical, semi-empirical, and process based storm erosion models. Coastal Engineering, 82, 64–75. Callaghan, D.P. & Wainwright, D. (2013) The impact of various methods of wave transfers from deep water to nearshore when determining extreme beach erosion. Coastal Engineering, 74, 50–58. Casella, E., Rovere, A., Pedroncini, A., Mucerino, L., Casella, M., Cusati, L.A., et al. (2014) Study of wave runup using numerical models and low-altitude aerial photogrammetry: A tool for coastal management. Estuarine, Coastal and Shelf Science, 149, 160-167. Castelle, B., Marieu, V., Bujan, S., Ferreira, S., Parisot, J.P., Capo, S., et al. (2014) Equilibrium shoreline modelling of a high-energy meso-macrotidal multiple-barred beach. Marine Geology, 347, 85–94. Castelle, B., Marieu, V., Bujan, S., Splinter, K.D., Robinet, A., Sénéchal, N., et al. (2015) Impact of the winter 2013–2014 series of severe Western Europe storms on a double-barred sandy coast: Beach and dune erosion and megacusp embayments. Geomorphology, 238, 135–148.

CH8 STORM CLUSTERING AND BEACH RESPONSE

169

Castelle, B., Turner, I.L. & Ruessink, B.G. (2007) Impact of storms on beach erosion: Broadbeach (Gold Coast, Australia). Journal of Coastal Research, Special Issue, 50, 534–539. Ciavola, P., Armaroli, C., Chiggiato, J., Valentii, A., Deserti, M., Perini, L., et al. (2007) Impact of storms along the Coastline of Emilia-Romagna: The morphological signature of the Ravenna coastline (Italy). Journal of Coastal Research, SI50. Coco, G., Senechal, N., Rejas, A., Bryan, K.R., Capo, S., Parisot, J.P., et al. (2014) Beach response to a sequence of storms. Geomorphology, 204, 493–501. Cooper, J.A.G., Jackson, D.W.T., Navas, F., McKenna, J. & Malvarez, G. (2004) Identifying storm impacts on an embayed, high-energy coastline: Examples from western Ireland. Marine Geology, 210, 261–280. Cox, J.C. & Pirrello, M.A. (2001) Applying joint probabilities and cumulative effects to estimate storm-induced erosion and shoreline recession. Shore and Beach, 69 (2), 5–7. Dalon, M.M., Haller, M.C. & Allan, J. (2007) Morphological characteristics of rip current embayments on the Oregon Coast. In: Proc. Coastal Sediments ’07. ASCE, New York, pp. 2137–2150. Davidson, M.A., Lewis, R.P. & Turner, I.L. (2010) Forecasting seasonal to multi-year shoreline change. Coastal Engineering, 57, 620–629. Davidson, M.A., Splinter, K.D. & Turner, I.L. (2013) A simple equilibrium model for predicting shoreline change. Coastal Engineering, 73, 191–202. Dawson, A., Elliott, L., Noone, S., Hickey, K., Holt, T., Wadhams, P., et al. (2004) Historical storminess and climate ‘see-saws’ in the North Atlantic region. Marine Geology, 210, 247–259. Del Rio, L., Plomaritis, T.A., Benavente, J., Valladares, M. & Ridera, P. (2012) Establishing storm thresholds for the Spanish Gulf of Cadiz coast. Geomorphology, 143–144, 13–23. Donnelly, J.P., Butler, J., Roll, S., Wengren, M. & Webb III,, T. (2004) A backbarrier overwash record of intense storms from Brigantine, New Jersey. Marine Geology, 210, 107–121. Donnelly, J.P., Roll, S. Wengren, M., Butler, J., Lederer, R. & Webb III,, T. (2001) Sedimentary evidence of intense hurricane strikes from New Jersey. Geology, 29 (7), 615–618. Donnelly, J.P. & Woodruff, J.D. (2007) Intense Hurricane activity over the past 5000 years controlled by El Niño and the West African monsoon. Nature, 447, 465–468. Earlie, C.S., Young, A.P., Masselink, G. & Russell, P.E. (2015) Coastal cliff ground motions and response to extreme storm waves. Geophysical Research Letters, doi: 10.1002/2014GL062534. Fawcett, L. & Walshaw, D. (2008) Bayesian inference for clustered extremes. Extremes, 11, 217–233. Ferreira, O. (2005) Storm groups versus extreme single storms: Predicted erosion and management consequences. Journal of Coastal Research, SI 42, 221–227. Ferreira, O. (2006) The role of storm groups in the erosion of sandy coasts. Earth Surface Processes and Landforms, 31, 1058–1060. Ferro, C.A.T. & Segers, J. (2003) Inference for clusters of extreme values. J. of the Royal Statistical Society: series B, 65 (2), 545–556. Forbes, D.L., Parkes, G.S., Manson, G.K. & Ketch, L.A. (2004) Storms and shoreline retreat in the southern Gulf of St Lawrence. Marine Geology, 210, 169–204. Furmanczyk, K.K., Dudzinska-Nowak, J., Furmanczyk, K.A., Paplinska-Swerpel, B. & Brzezowska, N. (2012) Critical storm thresholds for the generation of significant dune erosion at Daziwnow Spit, Poland. Geomorphology, 143–144, 62–68. Gallagher, E.L., Elgar, S. & Guza, R.T. (1998) Observations of sand bar evolution on a natural beach. Journal of Geophysical Research, 103, 3203–3215. Gallop, S.L., Bosserelle, C., Eliot, I. & Pattiaratchi, C.B. (2012) The influence of linestone reefs on storm erosion and recovery of perched beach. Continental Shelf Research, 47, 16–27.

170

COASTAL STORMS: PROCESSES AND IMPACTS

Hanley, J. & Caballero, R. (2012) The role of large-scale atmospheric flow and Rossby wave breaking in the evolution of extreme windstorms over Europe. Geophysical Research Letters, 39, L21708, doi: 10.1029/2012GL053408. Hansen, J.E. & Barnard, P.L. (2010) Sub-weekly to interannual variability of a high-energy shoreline. Coastal Engineering, 57, 959–972. Harris, D.L. (1963) Characteristics of hurricane storm surge. Technical paper No. 48 US Department of Commerce, Weather Bureau, Washington DC. Hesp, P.A. (1988) Surf zone, beach and foredune interactions on the Australian south east coast. Journal of Coastal Research, SI3, 15–25. Houser, C. & Hamilton, S. (2009) Sensitivity of post-hurricane beach and dune recovery to event frequency. Earth, Surface Processes and Landforms, 34, 613–628. Houser, C., Wernette, P., Rentschlar, E., Jones, H., Hammond, B. & Trimble, S. (2015) Post-storm beach and dune recovery: Implications for barrier island resilience. Geomorphology, 234, 54–63. Izaguirre, C., Méndez, F.J., Menéndez, M., Lucenõ, A. & Losada, I.J. (2010) Extreme wave climate variability in southern Europe using satellite data. Journal of Geophysical Research, 115, C04009, doi: 10.1029/2009JC005802. Izaguirre, C., Méndez, F.J., Menéndez, M. & Losada, I.J. (2011) Global extreme wave height variability based on satellite data. Geophysical Research Letters, 38, L10607, doi: 10.1029/2011GL047302. Jimenez, J.A., Sanchez-Arcilla, A., Bou, J. & Ortiz, M.A. (1997) Analysing short-term shoreline changes along the Ebro Delta (Spain) using aerial photographs. Journal of Coastal Research, 13 (4), 1256–1266. Jonathan, P. & Ewans, K. (2011) Modeling the seasonality of extreme waves in the Gulf of Mexico. Journal of Offshore Mechanics and Arctic Engineering, 133, 021104-1:021104-9. Karunarathna, H., Pender, D., Ranasinghe, R., Short, A.D. & Reeve, D.E. (2014) The effect of storm clustering on beach profile variability. Marine Geology, 348, 103–112. Keim, B.D., Muller, R.A. & Stone, G.W. (2004) Spatial and temporal variability of coastal storms in the North Atlantic Basin. Marine Geology, 210, 7–15. Kish, S.A. & Donoghue, J.F. (2013) Coastal response to storms and sea-level rise: Santa Rosa island, Northwest Florida, USA. Journal of Coastal Research, SI63 – Understanding and Predicting change in the coastal ecosystems of the Northern Gulf of Mexico, 131–140. Kvamsto, N.G., Song, Y., Seierstad, I.A., Sorteberg, A. & Stephenson, B. (2008) Clustering of cyclones in the ARPEGE general circulation model. Tellus A, 60 (3), 547–556. Kulmar, M., Lord, D. & Sanderson, B. (2005) Future directions for wave data collection in New South Wales. Proceedings of Coasts and Ports: Coastal Living-Living Coast. Australian Conference. van der Lageweg, W.I., Bryan, K.R., Coco, G. & Ruessink, B.G. (2013) Observations of shoreline-sandbar coupling on an embayed beach, Marine Geology, 344, 101–114. Larson, M., Capobianco, M. & Hanson, H. (2000) Relationship between beach profiles and waves at Duck, North Carolina, determined by canonical correlation analysis. Marine Geology, 163, 275–288. Lee, S. & Held, I.M. (1993) Baroclinic wave packets in models and observations. Journal of Atmospheric Science, 60, 1490–1503. Lee, G., Nicholls, R.J. & Birkemeier, W.A. (1998) Storm-driven variability of the beach-nearshore profile at Duck, North Carolina, USA, 1981–1991. Marine Geology, 148, 163–177. Li, F., Van Gelder, P.H.A.J.M., Ranasinghe, R., Callaghan, D.P. & Jongejan, R.B. (2014) Probabilistic modelling of extreme storms along the Dutch coast, Coastal Engineering, 86, 1–13, http://dx.doi.org/10.1016/j.coastaleng.2013.12.009.

CH8 STORM CLUSTERING AND BEACH RESPONSE

171

List, H.J., Farris, A.S. & Sullivan, C. (2006) Reversing storm hotspots on sandy beaches: Spatial and temporal characteristics. Marine Geology, 226, 261–279. Long, J.W. & Plant, N.G. (2012) Extended Kalman Filter framework for forecasting shoreline evolution. Geophysical Research Letters, 39, L13603, doi: 10.1029/2012GL052180. Lord, D. & Kulmar, M. (2000) The 1974 storms revisited: 25 years’ experience in ocean wave measurement along the South-East Australian Coast. 17th International Conference on Coastal Engineering, Sydney. Loureiro, C., Ferreira, O. & Cooper, J.A. (2012) Extreme erosion on high-energy embayed beaches: Influence of megarips and storm grouping. Geomorphology, 139–140, 155–171. Lucenõ, A., Menendez, M. & Mendez, F.J. (2006) The effect of temporal dependence on the estimation of the frequency of extreme ocean climate events. Proc. R. Soc. A., 462, 1683–1697. doi: 10.1098/rspa.2005.1652. Mailier, P.J., Stephenson, D.B., Ferro, C.A.T. & Hodges, K.I. (2006) Serial clustering of extratropical cyclones. Monthly Weather Review, 134, 2224–2240. Masselink, G., Austin, M., Tinker, J., O’Hara, J. & Russell, P. (2008) Cross-shore sediment transport and morphological response on a macrotidal beach with intertidal bar morphology, Truc Vert, France. Marine Geology, 251, 141–155. Masselink, G., Austin, M., Scott, T., Poate, T. & Russell, P. (2014) Role of wave forcing, storms and NAO in outer bar dynamics on a high-energy macro-tidal beach. Geomorphology, 226, 76–93. Masselink, G. & van Heteren, S. (2014) Response of wave-dominated and mixed-energy barriers to storms. Marine Geology, 352, 321–347. Méndez, F.J., Menéndez, M., Luceño, A., Medina, R. & Graham, N.E. (2008) Seasonality and duration in extreme value distribution of significant wave height. Ocean Engineering, 35 (1), 131–138. Miller, J.K. & Dean, R.G. (2004) A simple new shoreline change model. Coastal Engineering, 51, 531–556. Morton, R.A., Paine, J.G. & Gibeaut, J.C. (1994) Stages and durations of post-strom beach recovery; southeastern texas coats, USA. Journal of Coastal Research, 10, 884–908. Morton, R.A., Gibeaut, J.C. & Paine, J.G. (1995) Mesoscale transfer of sand during and after storms: Implications for prediction of shoreline movement. Marine Geology, 126, 161–117. Ogawa, H., Dickson, M.E. & Kench, P.S. (2015) Hydrodynamic constraints and storm wave characteristics o a sub-horizontal shore platform. Earth Surface Processes and Landforms, 40 (1), 65–77. Pinto, J.G., Gomara, I., Masato, G., Dacre, H.F., Woollings, T. & Caballero, R. (2014) Large-scale dynamics associated with clustering of extratropical cyclones affecting Western Europe. Journal of Geophysical Research: Atmospheres, doi: 10.1002/2014JD022305. Plant, N.G. & Stockdon, H.F. (2012) Probabilistic prediction of barrier-island response to hurricanes. Journal of Geophysical Research: Earth Surface, 117 (F3), doi: 10.1029/2011JF002326. Pries, A.J., Miller, D.L. & Branch, L.C. (2008) Identification of structural and spatial features that influence storm-related dune erosion along a barrier-island ecosystem in the Gulf of Mexico. Journal of Coastal Research, 24, 168–175. Priestas, A.M. & Fagherazzi, S. (2010) Morphological barrier island changes and recovery of dunes after Hurricane Dennis, St George Island, Florida. Geomorphology, 114, 614–626. Pye, K. & Blott, S.J. (2008) Decadal-scale variation in dune erosion and accretion rates: An investigation of the significance of changing storm tide frequency and magnitude on the Sefton coast, UK. Geomorphology, 102, 652–666. Qi, H., Cai, F., Lei, G., Cao, H. & Shi, F. (2010) The response of three main beach types to tropical storms in South China. Marine Geology, 275, 244–254.

172

COASTAL STORMS: PROCESSES AND IMPACTS

Quartel, S., Kroon, A. & Ruessink, B.G. (2008) Seasonal accretion and erosion patterns of a microtidal sandy beach. Marine Geology, 250, 19–33. Rao, V.B., do Carmo, A.M.C. & Franchito, S.H. (2002) Seasonal variations in the southern hemisphere storm tracks and wave propagation. Journal of Atmospheric Science, 59, 1029–1040. Raubenheimer, B. & Guza, R.T. (1996) Observations and predictions of run-up. Journal of Geophysical Research, 101 (C10), 25, 575–587. Regnauld, H., Pirazzoli, P.A., Morvan, G. & Ruz, M. (2004) Impacts of storms and evolution of the coastline in western France. Marine Geology, 210, 325–337. Reguero, B.G., Méndez, F.J. & Losada, I.J., 2013. Variability of multivariate wave climate in Latin America and the Caribbean. Global and Planetary Change, 100, 70–84. Revell, D.L., Komar, P.D. & Sallenger Jr.,, A.H. (2002) An application of LIDAR to analyses of El Ninõ erosion in the Netarts Littoral Cell, Oregon. Journal of Coastal Research, 18 (4), 792–801. Roberts, T.M., Wang, P. & Puleo, J.A. (2013) Storm-driven cyclic beach morphodynamics of a mixed sand and gravel beach along the Mid-Atlantic Coast, USA. Marine Geology, 346, 403–421. Roelvink, J.A., Reniers, A.J.H.M., van Dongeren, A., de Vries, J.V., McCall, R. & Lescinski, J. (2009) Modelling storm impacts on beaches, dunes and barrier islands. Coastal Engineering, 56, 1133–1152. ̇ Rózyoski, G., Larson, M. & Pruszak, Z. (2001) Forced and self-organized shoreline response for a beach in the southern Baltic Sea determined through singular spectrum analysis. Coastal Engineering, 43, 41–58. Ruessink, B.G., Houwman, K.T. & Hoekstra, P. (1998) The systematic contribution of transporting mechanisms to the cross-shore sediment transport in water depths of 3 to 9 m. Marine Geology, 152, 295–324. Ruessink, B.G., Miles, J.R., Feddersen, F., Guza, R.T. & Elgar, S. (2001) Modeling the alongshore current on barred beaches. J. Geophysical Research, 106, 22, 451–22, 463. Ruessink, B.G., Kuriyama, Y., Reniers, A.J.H.M., Roelvink, J.A. & Walstra, D.J.R. (2007) Modeling cross-shore sandbar behavior on the time scale of weeks DOI:10.1029/2006JF000730#_blank. Journal of Geophysical Research. Earth Surface, 112 (F3), F03010. Ruessink, B.G., Pape, L. & Turner, I.L. (2009) Daily to interannual cross-shore sandbar migration from a multiple sandbar system. Continental Shelf Research, 29, 1663–1677. Ruggiero, P., Komar, P.D. & Allan, J.C. (2010) Increasing wave heights and extreme value projections: The wave climate of the US Pacific Northwest. Coastal Engineering, 57, 539–552. Sallenger Jr,, A.H. (2000) Storm impact scale for barrier islands. Journal of Coastal Research, 16, 890–895. Sallenger, A.H., Krabill, W., Swift, R., Brock, J., List J., Hansen M., et al. (2003) Evaluation of airborne scanning lidar for coastal change applications. Journal of Coastal Research, 19, 125–133. Seablom, E.W., Ruggiero, P., Hacker, S.D., Mull, J. & Zarnetske, P. (2013) Invasive grasses, climate change, and exposure to storm-wave overtopping in coastal dune ecosystems. Global Change Biology, 19, 824–832, doi: 10.1111/gcb.12078. Sedrati, M. & Anthony, E.J. (2007) Storm-generated morphological change and longshore sand transport in the intertidal zone of a multi-barred macrotidal beach. Marine Geology, 244, 209–229. Semedo, A., Suselj, K., Rutgersson, A. & Sterl, A. (2011) A global view on the wind sea and swell climate and variability from ERA-40. Journal of Climate, 24, 1461–1479.

CH8 STORM CLUSTERING AND BEACH RESPONSE

173

Sénéchal N., Gouriou T., Castelle B., Parisot J.P., Capo S., Bujan S., et al. (2009) Morphodynamic response of a meso- to macro-tidal intermediate beach based on a long-term data-set. Geomorphology, 107, 263–274. Sénéchal, N., Abadie, S., Gallagher, E., MacMahan, J.H.M., Masselink, G., Michallet, H., et al. (2011) The ECORS-Truc Vert ’08 nearshore field experiment: Presentation of a three-dimensional morphologic system in a macro-tidal environment during consecutive extreme storm conditions. Ocean Dynamics, doi: 10.1007/s10236-011-0472-x. Senechal, N., Coco, G., Castelle B. & Marieu, V. (2015) Storm impact on the seasonal shoreline dynamics of a meso- to macrotidal open sandy beach. Geomorphology, 228, 448–461. Shand, R., Hesp, P. & Shepherd, M. (2004) Beach cut in relation to net offshore bar migration, Journal of Coastal Research, Special Issue, 39, 334–340. Short, A.D. & Trenaman, N.L. (1992) Wave climate of the Sydney region, an energetic and highly variable ocean wave regime. Australian Journal of Marine and Freshwater Research, 43, 765–791. Smith, R.L. & Weissman, I. (1994) Estimating the extremal index. J. R. Statist. Soc. B., 56 (3), 515–528. Smithers, S.G. & Hoeke, R.K. (2014) Geomorphological impacts of high-latitude storms waves on low-latitude reef islands – Observations of the December 2008 event on Nukutoa, Takuu, Papua New Guinea. Geomorphology, 222, 106–121. Splinter, K.D. Strauss, D.R. & Tomlinson, R.B. (2011) Assessment of post-storm recovery of beaches using video imaging techniques: A case study at Gold coast Australia. IEEE Transactions on Geoscience and Remote Sensing, 49 (12), 4704–4716. Splinter, K.D., Carley, J.T., Golshani, A. & Tomlinson, R. (2014a) A relationship to describe the cumulative impact of storm clusters on beach erosion. Coastal Engineering, 83, 49–55. Splinter, K.D., Turner, I.L., Davidson, M.A., Barnard, P., Castelle, B. & Oltman-Shay, J. (2014b), A generalized equilibrium model for predicting daily to interannual shoreline response, J. Geophys. Res. Earth Surf., 119, 1936–1958, doi: 10.1002/2014JF003106. Stockdon, H.F., Sallenger, A.H., List, J.H. & Holma, R.A. (2002) Estimation of shoreline position and change from airborne topographic lidar data. Journal of Coastal Research, 18, 502–513. Stockdon, H.F., Sallenger Jr.,, A.H., Holman, R.A. & Howd, P.A. (2007) A simple model for the spatially-variable coastal response to hurricanes. Marine Geology, 238, 1–20. Suanez, S., Cariolet, J.-M., Cancouët, R., Ardhuin, F. & Delacourt, C. (2012) Dune recovery after storm erosion on a high-energy beach: Vougot Beach, Brittany (France). Geomorphology, 139–140, 16–33. Swanson, K.L. & Pierrehumbert, R.T. (1994) Nonlinear wave packet evolution on a baroclinically unstable jet. Journal of Atmospheric Science, 51, 384–396. Thompson, D.W.J. & Barnes, E.A. (2014) Periodic variability in the large-scale southern hemisphere atmospheric circulation. Science, 343, 641–645. Thornton, E.B., Humiston, R.T. & Birkemeier, W. (1996) Bar/trough generation on a natural beach. Journal of Geophysical Research, 101 (C5), 12,097–12,110. Thornton, E.B., MacMahan, J.H. & Sallenger Jr,, A.H. (2007) Rip currents, mega-cusps, and eroding dunes. Marine Geology, 240, 151–167. Van Enckevort, I.M.J. & Ruessink, B.G. (2003) Video observations of nearshore bar behaviour. Part 1: Alongshore uniform variability. Continental Shelf Research, 23 (5), 501–512. Vitolo, R., Stephenson, D., Cook, I. & Mitchell-Vallace, K. (2009) Serial clustering of intense European windstorms. Meteorologische Zeitschrift, 18, 411–424.

174

COASTAL STORMS: PROCESSES AND IMPACTS

Vousdoukas, M.I., Almeida, L.P.M. & Ferreira, O. (2012) Beach erosion and recovery during consecutive storms at a steep-sloping, meso-tidal beach. Earth Surface Processes and Landforms, 37, 583–593. Wang, P., Kirby, J.H., Haber, J.D., Horwitz, M.H., Knorr, P.O. & Krock, J.R. (2006) Morphological and sedimentological impacts of Hurricane Ivan and immediate poststorm beach recovery along the northwestern Florida barrier-island coast. Journal of Coastal Research, 22, 1382–1402. Wright, L.D. & Short, A.D. (1984) Morphodynamic variability of surf zones and beaches: A synthesis. Marine Geology, 56, 93–118. Yates, M.L., Guza, R.T. & O’Reilly, W.C. (2009) Equilibrium shoreline response: Observations and modeling. Journal of Geophysical Research, 114 (C09014). Yu, F., Switzer, A.D., Lau, A.Y.A., Yeug, H.Y.E., Chik, S.W., Chiu, H.C., et al. (2013) A comparison of the post-storm recovery of two sandy beaches on Hong Kong island, southern China. Quaternary International, 304, 163–175.

9 Overwash Processes: Lessons from Fieldwork and Laboratory Experiments Ana Matias1 and Gerhard Masselink2 1 Centro 2 School

9.1

de Investigação Marinha e Ambiental (CIMA), Universidade do Algarve, Portugal of Marine Science and Engineering,University of Plymouth,Plymouth, UK

Introduction

The overarching aim of this chapter is to provide an overview of the current knowledge of the hydrodynamics and morphodynamics associated with overwash processes. The subject is approached from a short-term perspective (seconds to days) and will focus on overwash hydraulics, coastal impacts, sediment transport/deposition and morphological evolution during overwash episodes. This chapter will not discuss the role of overwash in barrier dynamics long-term coastal evolution (years to millennia), nor will it deal with the numerical modelling of overwash processes (described in Chapter 10). We will first define overwash and overwash-induced morphologies; their occurrence worldwide is presented, as well as the importance of overwash for barrier morphodynamics. A synthesis of methods to study overwash is given in section 9.2, covering fieldwork data collection and experimental work. The current state of knowledge on hydrodynamic processes during overwash is detailed in section 9.3. Hydrodynamics are described from the oceanographic conditions during overwash to the kinematics of the overwash flow itself as quantified by variables such as overwash depth, velocity, discharge and intrusion. Detailed information on the small-scale morphodynamic processes during overwash episodes on sand and gravel barriers is provided in section 9.4 and will include characteristics of washover geometry, volume and sedimentology, as well as a discussion of overwash sediment dynamics.

9.1.1 Overwash definition The terminology used in the literature to describe overwash and its resulting sedimentation has varied widely. In this work, overwash is the discontinuous transport of seawater and sediment over the barrier crest generated by wave runup (Figure 9.1). When the Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

176

(a)

COASTAL STORMS: PROCESSES AND IMPACTS

(b)

(c)

Figure 9.1 Pictures of overwash. (a) Overview of overwash flow over the crest (left) towards the backbarrier (right) contouring dune remnants on Barreta Island, Ria Formosa, Portugal. (b) Overwash at the barrier crest confined at throat by dune scarps on Barreta Island, Ria Formosa, Portugal. (c) Overwash over Roi-Namur Island, Republic of the Marshall Islands, during 2 March 2014 (Swarzenski, 2014). Photographs (a) and (b) taken by Alexandra Cunha; and (c) taken by Peter Swarzenski.

mean water level (including the tide level and/or storm surge elevation) is higher than the barrier crest, inundation occurs. Therefore, overwash is intrinsically a wave-driven process, related to swash activity, which may be associated with storm surge or equinoctial spring tides, but not necessarily. An overwash episode can last from minutes to days and is composed of a number of overwash events that often last less than a minute. The number of overwash events per second is referred to as the overwash frequency and the average time between two consecutive overwash events is referred to as the return period. Overwash processes are part of a continuum that starts as a prolongation

CH9 OVERWASH PROCESSES

177

of swash processes over the barrier crest and ends just before the barrier experiences inundation, which can result in catastrophic lowering of the barrier crest, breaching and/or inlet opening. Orford and Carter (1982) introduced the term overtopping: a process in which swash excursions just carry over the crest, causing vertical accretion at the swash-limit in the presence of rapid crestal percolation (Figure 9.2a and 9.2c). Overtop and overwash processes may occur at the same place on the barrier crest under conditions of increasing or decreasing wave height, rising or falling tidal stage, and waxing or waning meteorological surge. Overtopping as defined by Orford and Carter (1982) should not be mistaken for overtopping of coastal engineering structures (Figure 9.2b). Overtopping of coastal structures or sea defences occurs because of waves running up the face of a structure (Pullen et al., 2007). Washover is the morphology generated by overwash flows (Figure 9.3). However, not all morphologies named washovers are generated by overwash; often they result from barrier inundation. Coastal landforms associated with overwash have been extensively described in the literature. Two main washover types can occur: (1) washover plains (Figure 9.1a) are wide low-lying denuded areas; and (2) washover lobes are localised features that cut the dune field in specific places (Figure 9.1b) and consist of a mouth, a channel and a fan. Other authors have given different names to washover morphologies, for example washover plains have been named washover ramps (Fisher & Simpson, 1979), washover sheets (e.g. Ritchie & Penland, 1988), washover flats (e.g. Holland et al., 1991), and washover terraces (e.g. Bray & Carter, 1992). Such proliferation of terms has not facilitated straightforward comparison between different overwash studies.

9.1.2 Occurrence of overwash Overwash is a natural process that occurs in a variety of environments: oceanic coasts (e.g. Morton & Sallenger, 2003), estuarine coasts (e.g. Jennings & Coventry, 1973), barrier reefs (e.g. Bayliss-Smith, 1988), sea-shores (e.g. Guillén et al., 1994) and lacustrine coasts (e.g. Davidson-Arnott & Fisher, 1992). Nevertheless, their occurrence on barrier islands is most commonly described in the literature. Most studies are not specifically dedicated to the study of the actual overwash process; rather, the focus has generally been on overwash representing a storm impact on the coast (e.g. Morton & Sallenger, 2003), a significant process in barrier dynamics (e.g. Orford et al., 1995), or an important element of coastal sedimentology (e.g. Schwartz, 1982). Published observations and measurements of overwash or overwash-induced morphologies include occurrences in Australia (e.g. Baldock et al., 2008), Brazil (Silva et al., 2014), Canada (e.g. Armon & McCann, 1979), Colombia (e.g. Morton et al., 2000), Denmark (e.g. Kroon et al., 2013), France (e.g. Stéphan et al., 2010), Germany (e.g. Hofstede, 1997), Ireland (e.g. Orford & Carter, 1982), Italy (e.g. Armaroli et al., 2012), Morocco (e.g. Raji et al., 2015), Mexico (Cooper et al., 2007), the Netherlands (e.g. Hoekstra et al., 2009), New Zealand (e.g. Tribe & Kennedy, 2010), People’s Republic of China (e.g. Qi et al., 2010), Portugal (e.g. Matias et al., 2010), the Republic of Marshall Islands (Swarzenski, 2014), Soloman Islands (Bayliss-Smith, 1988), Spain (e.g. Benavente et al., 2006), Thailand (Phantuwongraj et al., 2013), UK (e.g. Bradbury & Powell, 1992), USA. (e.g. Leatherman, 1976), and Vietnam (e.g. Tuan & Verhagen, 2008). Overwash has almost certainly occurred, and is occurring, in other countries for which publications are non-existent or not easily found. In spite of the increasing number of

178

(a)

COASTAL STORMS: PROCESSES AND IMPACTS

(b)

(c)

Figure 9.2 Illustration of both applications of the word ‘overtopping’. (a) Photograph of overtopping on Barreta Island, Ria Formosa, Portugal. (b) Photograph of overtopping of a groin in Albufeira, Portugal. (c) Scheme with definition of overtopping and overwash processes (modified from Orford & Carter, 1982).

CH9 OVERWASH PROCESSES

179

Figure 9.3 Picture of washovers in Salthouse, East England, UK, after North Sea storm surge in December 2014. Photo courtesy of Mike Page, www.mike-page.co.uk.

180

COASTAL STORMS: PROCESSES AND IMPACTS

studies about overwash processes throughout the world, the majority of coastal and geoscience investigations related to overwash have been conducted in the USA (more than 60% of bibliographic references until 2015).

9.1.3 Relevance of overwash Overwash associated with major storms can be catastrophic (such as the impact of Hurricane Katrina on the USA coastal areas); however, from a longer-term point of view (hundreds of years), overwash can be considered a constructive and natural process that contributes to the shaping, reshaping and maintenance of barrier systems. Repeated overwash processes are important for long-term natural evolution of transgressive barrier islands, whereby the net volume of sand contained in the barrier structure is often maintained whilst the environments translate landward (e.g. Dolan & Godfrey, 1973). Overwash is instrumental in the landward migration of barrier islands by the ‘rollover’ mechanism (e.g. Carter & Orford, 1993). The rollover mechanism involves onshore-directed sediment transport driven by storm waves through erosion of the front of the barrier, transfer across the barrier crest and deposition at the back of the barrier in the form of washover deposits (Figure 9.3). Overwash also contributes to other patterns of barrier evolution, such as breaching (Kraus et al., 2002), barrier breakdown (Pye & Blott, 2009), inlet formation (Vila-Concejo et al., 2006), dynamics of lagoon entrance (Baldock et al., 2008) and outlet closure (Orford et al., 1988). In some cases, overwash is an important process in salt-marsh development. A thorough description about overwash and other storm-related processes in barrier dynamics is given in Chapter 4. Overwash occurrence in developed areas often represents a hazard, with significant socio-economic impacts, such as: injury and loss of human life; burial of roads and paths; destruction of private property, beach facilities, electric equipment and infrastructures; touristic activities decline; pollution due to sewage destruction; intrusion of salt and sand into agriculture soils; and inhibition of channel navigation.

9.2 Methods to study overwash processes Overwash occurs both on sand and gravel barriers. Field studies of overwash in sandy environments are more common (e.g. Leatherman, 1976; Holland et al., 1991) than on gravel beaches. Nevertheless, important field studies on gravel barrier are reported by Lorang (2002) and Orford et al. (2003). Overwash mainly occurs during storms, when accurate in situ field measurements are hazardous and difficult or impossible to obtain with current equipment. Measurements of overwash are mainly undertaken in two ways: fieldwork measurements and laboratory experiments.

9.2.1 Fieldwork measurements Field observations of overwash processes are sometimes carried out during overwash processes, but more commonly are made before and after overwash occurrence. Analysis of ground photographs and vertical aerial photographs (e.g. Rodríguez et al., 1994) has been used to locate, describe and measure overwash. The evaluation of overwash-induced morphological changes, volume of sediment deposition and

CH9 OVERWASH PROCESSES

181

overwash water maximum intrusion across the barrier have been studied by the measurement of pre- and post-storm barrier morphology using topographic surveying techniques, such as total station surveying (e.g. Stone et al., 2004), GPS surveying (e.g. Matias et al., 2009) or LiDAR surveying (e.g. Stockdon et al., 2009). Most pre-2000 studies used cross-shore profiles to obtain overwash deposition in m3 m−1 inferred from specific locations in the washovers. With the advent of dGPS, and particularly LiDAR surveying, data on three-dimensional washover morphology has become easily available and more accurate total sedimentation volumes can be obtained. In other cases, overwash sediment deposition is evaluated by the combined technique of plan-view measurement of washover perimeter and elevation, with the thickness of overwash deposits obtained by coring and/or ground penetrating radar (e.g. Carruthers et al., 2013). There are only seven existing datasets collected during overwash: (1) Assateague Island, USA (Fisher et al., 1974; Leatherman, 1976; Fisher & Stauble, 1977); (2) Nauset Spit-Eastham, USA (Leatherman & Zaremba (1987); (3) Trinity Island, USA (Holland et al., 1991); (4) Trabucador Spit, Spain (Guillén et al., 1994); (5) Shoeldon’s Marsh barrier, USA (Bray & Carter, 1992); (6) Belongil Beach, Australia (Baldock et al., 2008); and (7) Barreta Island, Portugal (Matias et al., 2010). The field investigations all measured morphological changes during overwash and some measured overwash hydrodynamics. However, these datasets are mixed in quality and scope, ranging from single hydrodynamic measurements using rough methods, to more extensive measurements using more sophisticated methods. Topography was measured using a grid of wooden stakes superimposed on a transverse slice of the island, a grid of elevation stations that were measured with an automatic level and survey rod, total station and GPS. Overwash hydraulics was measured with bi-directional electromagnetic current meters, flow meters, by timing wooden floats between markers’ progression of the overwash bore front between rods, pressure transducers and capacitance-type wave staffs.

9.2.2 Laboratory experiments Laboratory experiments on overwash provide numerous advantages in relation to fieldwork, including the ability to control the hydrodynamic conditions to ensure overwash occurs and logistical benefits (e.g. more preparation time, capacity to recruit researchers, less stressful schedules, simplification of equipment installation and power supply, and provision of a safe environment for people and equipment). Experiments of overwash on mobile bed have been made in eight facilities: (1) at DuPont Hall, the University of Delaware (USA), described in Hancock and Kobayashi (1994), Tega and Kobayashi (1996, 1999), Figlus et al. (2010, 2011); (2) at HR Wallingford (UK), described in Obhrai et al. (2008); (3) at US Army Corps of Engineers Coastal Hydraulic Laboratory in Vicksburg (USA), described in Donnelly (2008); (4) at Texas A&M University (USA), described in Park and Edge (2010); (5) at ‘Air-Sea’ tank in the Coastal and Oceanographic Laboratory at the University of Florida (USA), described in Srinivas et al. (1992) and Pirrello (1992); (6) at the University of Queensland (Australia), described in Baldock et al. (2005); (7) at CIEM of the Polytechnic University of Catalonia (Spain), described in Alessandro et al. (2010); and (8) at the Delta flume (Figure 9.4), Deltares (the Netherlands), described in Matias et al. (2012, 2013). Experiments undertaken in facilities (1) to (7) are considered small-scale experiments, with wave heights in the range of 0.14–0.33 m. Large-scale experiments on

182

COASTAL STORMS: PROCESSES AND IMPACTS

(a)

(b)

(c)

(d)

Figure 9.4 Pictures of BARDEX and BARDEX II experiments. (a) View towards the ‘lagoon’, with scaffolding on top of gravel barrier of BARDEX experiment. (b) Time-series of overwash depth recorded during Test Series D34 (BARDEX II), with the peak depth of each overwash event marked with a circle. (c) View towards the paddle, with overwash on top of the gravel barrier of BARDEX experiment. (d) View towards the paddle, with overwash on top of the sandy barrier of BARDEX II experiment.

overwash have been undertaken during BARDEX (Barrier Dynamics Experiment) experiments, when significant wave heights reached 1.0 m. Most experiments used sand grain-size sediments to build the barriers inside the flumes; however, Obhrai et al. (2008), Bradbury and Powell (1992) and Matias et al. (2012) constructed barriers out of gravel-sized material. All experiments measured barrier profiles before and after overwash runs, thus enabling accurate quantification of barrier morphologic change and overwash sedimentation volumes. Morphological measurements were made using various

CH9 OVERWASH PROCESSES

183

devices (e.g. laser scanner, ultrasound profiler, mechanical profiler, video cameras for photogrammetric surveys) with measuring instruments located at fixed positions or mobile from overhead positions. During BARDEX experiments, the subaerial barrier morphologic evolution during overwash was measured at 4 Hz using 45 acoustic Bed-Level Sensors (BLS). BLSs provide a high-resolution measurement of the morphological evolution on each of the discrete points of measure, but only record the bed level when the bed is dry. Measurements of the main hydrodynamic variables of overwash (velocity, depth and discharge) are much more difficult to obtain; therefore, datasets are very limited. Overwash water depth and velocity were measured using a video camera, and using the BLSs by cross-correlating the time series of adjacent sensors. Some experiments inferred overwash discharge from measurements in backbarrier reservoirs, either with wave gauges inside the reservoir or by tracking the pumping rates necessary to maintain a certain water level.

9.3

Hydrodynamic processes during overwash

9.3.1 Oceanographic conditions Overwash events during storms are undoubtedly the most common situation described worldwide in the literature. Of particular attention has been the occurrence of overwash in USA barriers during hurricanes, as identified by Hayes (1967), and highly relevant in the context of the extensive flooding of human development during Hurricane Katrina in August 2005. Overwash has also occurred as a consequence of other types of storms. Storm wave heights that induce overwash can vary significantly. Existing studies relate storm-induced overwash on sandy barriers during storm events with offshore wave heights from less than 4 m (March 1975, Assateague Island, MD, USA; Leatherman, 1976), to more than 9 m (October 1991, Devereaux Beach, MA, USA; FitzGerald et al., 1994). Overwash on gravel barriers has been identified during storms with wave heights from 3.5 m (e.g. Hurst Spit, UK, Bradbury & Powell, 1992) to 6.5 m (Chesil Beach, UK, May & Hansom, 2003). Water levels during storms are critical to the occurrence of overwash. Storm-driven surge and runup are highest when storms coincide with perigean spring tides, especially for areas where surge amplitudes are relatively small when compared to the tidal range (Anthony, 2013). It should be stressed that overwash can also occur during non-storm conditions. Non-storm overwash may be caused by El Niño flooding (Morton et al., 2000), extreme tide levels in lakes (Schwartz, 1975), or due to severe lagoon floods (Nguyen et al., 2006). Nonetheless, the location and magnitude of overwash is not only dependent on the hydrodynamic forcing (wave conditions and water level), but is also controlled by the site-specific geomorphological context.

9.3.2 Hydraulics of overwash flows After wave runup reaches the barrier crest, either during storm or non-storm conditions, overwash flow starts to run over the barrier crest. Not much information exists about the properties of the generated flow, or for the morphological and sedimentological evolution of coastal barriers during overwash. Nevertheless, some data are available

184

COASTAL STORMS: PROCESSES AND IMPACTS

on the main hydrodynamic properties of overwash flows (velocity, depth, discharge, intrusion). Leatherman (1977) obtained a mean overwash flow velocity of 1.95 ms−1 in Assateague Island (USA); Leatherman and Zaremba (1987) measured 0.5–2.0 ms−1 overwash flow velocities at Nauset Spit (USA); the maximum overwash flow through the Trabucador Bar (Spain) was 1.5 ms−1 (Guillén et al., 1994); mean velocities of 2.0 ms−1 were obtained by Holland et al. (1991) at the Isles Dernieres (USA); Bray and Carter (1992) measured overwash flow velocities of 1–3 ms−1 at a barrier in Lake Erie (USA) and Matias et al. (2010) measured average velocities of 2.2–2.3 ms−1 for non-storm overwash on Barreta Island (Portugal). Figure 9.5 represents the average and standard deviation of overwash flow velocity and depth measured in the field. As for experimental studies in flumes, measurements of overwash velocities at the crest are also limited. Srinivas et al. (1992) measured 0.8–1.2 ms−1 overwash velocity over a sandy barrier; Schüttrumpf and Oumeraci (2005) measured up to 0.7 ms−1 overtopping velocity over an impermeable dike; Donnelly (2008) measured bore front velocities smaller than 1.5 ms−1 on a sandy barrier; Matias et al. (2014) measured average overwash velocities of 3.3 ms−1 on a gravel barrier; and Matias et al. (2016) measured average overwash velocities of 2.0 ms−1 on a sandy barrier. Fieldwork undertaken during overwash has shown a range of overwash peak flow depths: 0.7 m (Fisher & Stauble, 1977), 0.45 m (Fisher et al., 1974; Leatherman & Zaremba, 1987), 0.19 m (Leatherman, 1976), 0.15 m (Matias et al., 2010), 0.13 m (Holland et al., 1991) and 0.10 m (Bray & Carter, 1992; Figure 9.5). Typically, laboratory experiments exhibit very shallow flows, such as 0.04–0.09 m (Donnelly, 2008) due to the small scale used, with the exception of the BARDEX and BARDEX

4 F. & S., 1977

Overwash velocity (ms−1)

3.5

M. et al., 2010

3 L., 1976

2.5 L. & Z., 1977

B. & C., 1992

2

F. et al., 1974

1.5 H. & et al., 1991

1 0.5 0

0

0.1

0.2

0.3

0.4 0.5 Overwash depth (m)

0.6

0.7

0.8

Figure 9.5 Relationship between overwash depth and overwash velocity for available fieldwork data sets. Symbols represent average values and bars represent standard deviation of velocity. Data sets used in this plot are identified by the authors of studies as: Fisher et al., 1974; Leatherman, 1976; Fisher & Stauble, 1977; Leatherman & Zaremba, 1987; Holland et al., 1991; Bray & Carter, 1992; Matias et al., 2010.

CH9 OVERWASH PROCESSES

185

II experiments during which a maximum overwash depth of 0.77 m (Matias et al., 2014) and 0.46 m (Matias et al., 2016), respectively, was measured. Considering all available fieldwork datasets, summarised in Figure 9.5, most overwash events are super-critical flows with average Froude numbers of c. 1.5, demonstrating a shallow and fast, super-critical and turbulent flow. However, Leatherman and Zaremba (1987) recorded critical flows and Fisher et al. (1974) measured subcritical flows because flows were deeper. Direct measurements of overwash water discharge were only undertaken in laboratory experiments. Tega and Kobayashi (1996) obtained an overwash discharge of 0.4–8.7 lm−1 s−1 , while Figlus et al. (2010) recorded 17–20 lm−1 s−1 . Naturally, overwash flow parameters are highly variable between sites, between different episodes at the same site and between different events during the same episode. As mentioned previously, each overwash episode consists of a number of overwash events over the crest (Figure 9.4b as an example) and although a mean value is a succinct way to express and compare episodes, additional statistical properties (e.g. 2% exceedance value) may be more appropriate.

9.4

Morpho-sedimentary dynamics by overwash processes

9.4.1 Morphological changes induced by overwash The overwash flow passage over the barrier island induces sediment transport and deposition, and, hence, changes in the barrier morphology. Figure 9.6 represents examples of the morphological changes due to overwash measured in barriers located in Europe (France, Portugal and Italy) and the USA (Florida and Maryland). During overwash, most barriers experience landward retreat of the barrier crest on the order of tens of metres (e.g. 20 m in Sillon de Talbert, France; Stéphan et al., 2010); and crest elevation decrease (Figure 9.6a, 9.6c and 9.6d), stability (Figure 9.6b and 9.6f) or increase (Figure 9.6e). It is noticeable that overwash tends to smooth the barrier cross-shore profile; however, the shape and volume of the overwash deposit is variable. In plan view, an isolated overwash occurrence generates a conspicuous shape of lobate to elongate washover fans, whereas widespread overwash generates a semi-rectilinear seaward edge and a crenulated inland edge (Figure 9.7). In more dramatic cases, there is generalised erosion of almost the entire barrier island (Figure 9.7a), with accretion occurring on the middle to distal barrier margin (Figure 9.7b), or even at the lagoon/channel/sound (Figure 9.7a). Morton and Sallenger (2003) noted that washover volumes are related to the type of washover, increasing from confined fans, to terraces, to sheet overwash deposits. Existing studies have reported overwash depositional volumes that vary between tens and hundreds of m3 m−1 . For example, the smallest fieldwork measurement of overwash deposition found in the literature was 3 m3 m−1 (Leatherman, 1976); however, most studies report deposition volumes between 10 m3 m−1 (FitzGerald et al., 1994) and 150 m3 m−1 (Leatherman & Zaremba, 1987). Exceptionally high depositional volumes of 225 m3 m−1 were measured in Matagorda Peninsula (Texas, USA) by Morton and Sallenger (2003). For more details about the large (in space) and long (in time) scale response of barrier islands, the reader should refer to Chapter 4.

186

COASTAL STORMS: PROCESSES AND IMPACTS

(a) Santa Rosa Island, USA

(b) Sillon de Talbert, France

5

pre-overwash post-overwash 12 10

3

Elevation (m)

Elevation (m)

4

2 1

8 6 4

0 −1

2 0

50

100 150 200 Distance (m)

250

300

0

50

100 150 Distance (m)

200

250

(d) Barreta Island, Portugal 5

5

4

4

3

Elevation (m)

Elevation (m)

(c) St. George Island, USA

3 2

1 0

1

−1 −2

0 0

25

50

75 100 125 Distance (m)

175

150

200

0

(e) Assateague Island, USA

20

40

60 80 Distance (m)

100

120

(f) Lido di Dante, Italy

4

3 Elevation (m)

3 Elevation (m)

2

2 1 0 −1 −2

2 1 0 −1 −2

0

20

40

60 80 Distance (m)

100

120

140

0

20

40 60 Distance (m)

80

100

Figure 9.6 Example of pre- and post-overwash cross-shore barrier morphology. (a) Profiles located on Santa Rosa Island, Florida, USA, before and after Hurricane Opal in 1995; adapted from Stone et al. (2004). (b) Profiles located on Sillon de Talbert, Brittany, France, before and after March 2008 storm; adapted from Stéphan et al. (2010). (c) St George Island, Florida, USA, before and after Hurricane Dennis in 2005; adapted from Priestas & Fagherazzi (2010). (d) Profiles located on Barreta Island, Ria Formosa, Portugal, before and after equinoctial spring tides of September 2012; adapted from Matias et al. (2009). (e) Profiles located on Assateague Island, Maryland, USA, before and after northeastern storm of August 1976; adapted from Fisher & Stauble (1977). (f) Profiles located on Lido di Dante, Emilia-Romagna Region, Italy, before and after December 2008 storm, adapted from Armaroli et al. (2012). The sea is always to the left of the profiles. Note that the horizontal and vertical scales are different for the various examples.

CH9 OVERWASH PROCESSES

187

(a) Dauphin Island, USA

Mississippi Sound

Difference Erosion Accretion

Gulf of Mexico

(b) St. George Island, USA 2

Washover deposition

1

Distance (m)

150

0 300

Road Foredune removal

450

Fan

Throat

Apron

600

−1 −2 −3

Figure 9.7 (a) Elevation change map of Dauphin Island, Alabama, USA, after Hurricane Katrina in 2005, elevation gains in green and losses in red (Sallenger et al., 2007). (b) Elevation change map of St George Island, Florida, USA, after Hurricane Dennis in 2005, bar in meters of elevation (from Priestas & Fagherazzi, 2010).

Overwash episodes do not happen as a single overwash event, but a set of events (Figure 9.4b). These multiple events are recorded in washovers as laminations throughout the deposit (e.g. Leatherman et al., 1977; Schwartz, 1982), although these layers are not always seen in sediment cores (Leatherman et al., 1977). The initial events of overwash can erode the throat and pre-overwash surface in the backbarrier, resulting in a reactivation surface (Kochel & Dolan, 1986). A set of well-laminated, sub-horizontal to landward dipping beds is the main sedimentary structure can be found on overwash stratigraphic sequences (Hobday & Jackson, 1979; Nichol & Boyd, 1993; Davidson-Arnott & Reid, 1994). Also common is the presence of a basal gravel layer (e.g. Davidson-Arnott & Fisher, 1992; Anthony et al., 1996).

9.4.2 Morphodynamic processes during overwash Fieldwork and laboratory observations have revealed that an overwash episode is generally composed of three different phases: overtopping, minor overwash and overwash. The details and differences in barrier evolution during overwash on gravel and sandy barriers given here are solely based on measurements undertaken during large-scale BARDEX experiments. Interesting and useful insights into feedback

188

COASTAL STORMS: PROCESSES AND IMPACTS

processes arise from these observations, but it should be emphasised that these insights cannot automatically be transferred to the real world without complementary field observations. At the start of an overwash episode, the wave runup starts to overtop the barrier crest (Figure 9.2a) and the sand transported across the beach face is mostly deposited on the crest. Once the flow crosses the barrier top, usually just before the peak of the storm or high tide, the sedimentary processes are similar to those occurring during beach berm development, and defined by Orford and Carter (1982) as beach crest overtopping (definitions of overtopping and overwash can be found in the appropriate section). Sandy barrier may experience almost negligible crest elevation variations during overtopping or a thin layer of sand is deposited by the shallow overtopping flows. On gravel beaches flows are generally strong, but do not result in overwash at this earlier stage because of greater infiltration. Crest accretion on gravel barriers during overtopping provides a negative feedback on morphological evolution through raising the barrier crest, and inhibiting overwash. This would suggest that gravel barriers are more resilient than sandy barriers under conditions around the threshold between overtopping and overwash, because the gravel barrier crest build-up delays the onset of overwash. As the storm increases and/or the tide rises, minor overwash begins and evolves by means of an increase in overwash frequency and flow velocity/depth. Sediment is transported further into the barrier and eventually onto backbarrier slope (e.g. Figure 9.6d). When overwash intrusion is long, sedimentation can occur at the rear slope of the barrier (e.g. Figure 9.6b). These deposits create back-barrier slope instabilities, which periodically fail and avalanche down the submerged rear-side of the barrier forming a steep prograding surface approximately parallel to the original slope (Figure 9.8b).

(a) Series E, BARDEX II

(b) Series E10, BARDEX 2008

Figure 9.8 Barrier cross-shore profiles from test series with overwash. (a) Profiles from Test Series E of BARDEX II; and (b) Profiles from Test Series E10 of BARDEX (from Matias et al., 2014). Comparison of gravel and sand profiles and feedback processes of BARDEX 2008 and BARDEX II.

CH9 OVERWASH PROCESSES

189

The lagoon water level is important in controlling the geometry of the back-barrier deposit, particularly at the interception between the subaerial back-barrier deposits and those below lagoon water level. In situ field records during overwash on comparable sand and gravel barriers are inexistent. The comparison between gravel and sandy barrier behaviour during intense barrier overwash phase described here is solely based on measurements undertaken during BARDEX experiments. Figure 9.8 shows the difference in morphological evolution during overwash of a sandy barrier (Figure 9.8a, from the experiment BARDEX II) and a gravel barrier (Figure 9.8b, from the experiment BARDEX 2008). BARDEX II morphology mimics sand barriers with a backbarrier lagoon and no dune development, which can be found, for example, in the Ria Formosa barrier system, Portugal (Figure 9.1a and 9.2a). Similarly, BARDEX 2008 morphology mimics gravel barriers such as Slapton Sands, Devon, UK (Austin & Masselink, 2006). In the gravel barrier case (Figure 9.8b), the beach face started very steep and flattened during overwash. On the sandy barrier (Figure 9.8a), the beach face started equally steep, but remained steep during overwash. On the gravel barrier, sediments were transported onshore, whereas on the sandy barrier sediments were transported onshore to the backbarrier and offshore to the inner surf zone. This offshore sediment transport resulted in the development of a submerged bar (Figure 9.8a) that enhanced wave energy dissipation. The inner-surf build-up on the sandy barrier provides a negative feedback process that tends to hinder further overwash. Therefore, BARDEX overwash experiments provide examples both of positive and negative feedback processes during overwash.

9.5

Conclusion

Overwash is the discontinuous transport of seawater and sediment over the bar crest generated by wave runup, resulting in the development of washover plains and washover lobes. Overwash processes are part of a continuum of process regimes that starts with overtopping, when wave runup just reaches the barrier crest, and ends with inundation, when the barrier crest is continuously submerged. Overwash associated with major storms can be catastrophic; however, from a longer-term point of view (hundreds of years), overwash can be considered a constructive and natural process that contributes to the shaping, reshaping and maintenance of barrier systems. Overwash is generally associated with storm conditions with significant wave heights ranging from 3 to 9 m; however, overwash can also occur during non-storm conditions. Importantly, the occurrence and magnitude of overwash is not only dependent on the hydrodynamic forcing (wave conditions and water level), but is also controlled by the site-specific geomorphological context (see also Chapter 4). Due to the hazardous and challenging conditions during overwash, in situ field measurements of overwash processes are very limited. Overwash depth and velocity have been obtained in only a handful of field studies; these show depths of up to 0.7 m and velocities of 1–3.5 m s−1 . Most information on overwash dynamics has been obtained through comparison of barrier morphology before and after overwash occurrence, and existing studies have reported overwash depositional volumes that vary between tens and hundreds of m3 m−1 . Proto-type laboratory experiments have suggested two contrasting preservation strategies of sandy and gravel barriers during overwash conditions. On gravel barriers,

190

COASTAL STORMS: PROCESSES AND IMPACTS

crest deposition due to swash infiltration may delay the transition from overtopping to overwashing. On sandy barriers, offshore sediment transport, as well as onshore transport, within the overwashing regime, can lead to the development of a nearshore bar, and the enhanced wave breaking on the bar reduces the potential for overwash.

Acknowledgements This study was supported by RUSH project (PTDC/CTE-GIX/116814/2010) funded by FCT and the European Community’s 7th Framework Programme BARDEX II project (HYDRALAB IV, Contract no. 261520). Ana Matias was supported by Investigator Programme, funded by FCT.

References Alessandro, F., Fortes, C.J., Ilic, S., James, M., Sancho, F., Schüttrumpf, H. & Tomasicchio, G.R. (2010) Wave storm induced erosion and overwash in large-scale flume experiments. HYDRALAB III Joint User Meeting. Hanover, Germany. Anthony, E.J. (2013) Storms, shoreface morphodynamics, sand supply, and the accretion and erosion of coastal dune barriers in the North Sea. Geomorphology, 199, 8–21. Anthony, E.J., Lang, J. & Oyédé, L.M. (1996) Sedimentation in a tropical, microtidal, wave-dominated coastal-plain estuary. Sedimentology, 43, 665–675. Armaroli, C., Ciavola, P., Perini, L., Calabrese, L., Lorito, S., Valentini, A. et al. (2012) Critical storm threshold for significant morphological changes and damage along the Emilia-Romagna coastline, Italy. Geomorphology, 143–144, 34–51. Armon, J.W. & McCann, S.B. (1979) Morphology and landward sediment transfer in a transgressive barrier island system, Southern Gulf of St Lawrence, Canada. Marine Geology, 31, 333–344. Austin, M.J. & Masselink, G. (2006) Observations of morphological change and sediment transport on a steep gravel beach. Marine Geology, 229, 59–77. Baldock, T.E., Hughes, M.G., Day, K. & Louys, J. (2005) Swash overtopping and sediment overwash on a truncated beach. Coastal Engineering, 52, 633–645. Baldock, T.E., Weir, F. & Hughes, M.G. (2008) Morphodynamic evolution of a coastal lagoon entrance during swash overwash. Geomorphology, 95, 398–411. Bayliss-Smith, T.P. (1988) The role of hurricanes in the development of reef islands, Ontong Java Atoll, Solomon Islands. The Geographical Journal, 154, 377–391. Benavente, J., Del Río, L., Gracia, F.J. & Martínez-Del-Pozo, J.A. (2006) Coastal flooding hazard related to storms and coastal evolution in Valdelagrana spit (Cadiz Bay Natural Park, SW Spain). Continental Shelf Research, 26, 1061–1076. Bradbury, A.P. & Powell, K.A. (1992) The short term profile response of shingle spits to storm wave action. 23rd International Conference on Coastal Engineering. Venice, Italy. Bray, T.F. & Carter, C.H. (1992) Physical processes and sedimentary record of a modern, transgressive, lacustrine barrier island. Marine Geology, 105, 155–168. Carruthers, E.A., Lane, D.P., Evans, R.L., Donnelly, J.P. & Ashton, A.D. (2013) Quantifying overwash flux in barrier systems: An example from Martha’s Vineyard, Massachusetts, USA. Marine Geology, 343, 15–28. Carter, R.W.G. & Orford, J.D. (1993) The morphodynamics of coarse clastic beaches and barriers: A short term and long term perspective. Journal of Coastal Research, SI 15, 158–179.

CH9 OVERWASH PROCESSES

191

Cooper, J.A.G., Lewis, D.A. & Pilkey, O.H. (2007) Fetch-limited barrier islands: Overlooked coastal landforms. GSA Today, 17, 4–9. Davidson-Arnott, R.G.D. & Fisher, J.. (1992) Spatial and temporal controls on overwash occurrence on a Great Lakes barrier spit. Canadian Journal of Earth Sciences, 29, 102–117. Davidson-Arnott, R.G.D. & Reid, H.E.C. (1994) Sedimentary processes and the evolution of the distal bayside of Long Point, Lake Erie. Canadian Journal of Earth Sciences, 31, 1461–1473. Dolan, R. & Godfrey, P. (1973) Effects of Hurricane Ginger on the barrier islands of North Carolina. Geological Society of America Bulletin, 84, 1329–1334. Donnelly, C. (2008) Coastal overwash: Processes and modelling. PhD, Lund University, Sweden. Figlus, J., Kobayashi, N., Gralher, C. & Iranzo, V. (2010) Wave-induced overwash and destruction of sand dunes. 32nd International Conference on Coastal Engineering. Shanghai, China. Figlus, J., Kobayashi, N., Gralher, C. & Iranzo, V. (2011) Wave overtopping and overwash of dunes. Journal of Waterway, Port, Coastal, and Ocean Engineering, 137, 26–33. Fisher, J.J. & Simpson, E.J. (1979) Washover and tidal sedimentation rates as environmental factors in development of a transgressive barrier shoreline. In: S.P. Leatherman (Ed.) Barrier islands from the Gulf of St. Lawrence to the Gulf of Mexico. Academic Press, New York. Fisher, J.S., Leatherman, S.P. & Perry, F.C. (1974) Overwash processes on Assateague Island. 14th International Conference on Coastal Engineering, Copenhagen, Denmark. Fisher, J.S. & Stauble, D.K. (1977) Impact of Hurricane Belle on Assateague Island washover. Geology, 5, 765–768. Fitzgerald, D.M., Van Heteren, S. & Montello, T.M. (1994) Shoreline processes and damage resulting from the Halloween Eve Storm of 1991 along the North and South shores of Massachusetts, USA. Journal of Coastal Research, 10, 113–132. Guillén, J., Camp, J. & Palanques, A. (1994) Short-time evolution of a microtidal barrier – Lagoon system affected by storm and overwashing: The Trabucador Bar (Ebro Delta, NW Mediterranean). Zeitschrift fur Geomorphologie, 38, 267–281. Hancock, M.W. & Kobayashi, N. (1994) Wave overtopping and sediment transport over dunes. 24th Conference on Coastal Engineering. Kobe, Japan: ASCE. Hayes, M.O. (1967) Hurricanes as geological agents: Case studies of Hurricane Carla, 1961, and Cindy, 1963. Report of Investigation 61, Texas Bureau of Economic Geology. Hobday, D.K. & Jackson, M.P.A. (1979) Transgressive shore zone sedimentation and syndepositional deformation in the Pleistocene of Zululand, South Africa. Journal of Sedimentary Petrology, 49, 145–158. Hoekstra, P., Haaf, M., Buijs, P., Oost, A., Breteler, R.K., Van Der Giessen, K., et al. (2009) Washover development on mixed-energy, mesotidal barrier island system. Coastal Dynamics 2009. Tokyo, Japan. Hofstede, J.L.A. (1997) Process-response analysis for the North Frisian supratidal sands (Germany). Journal of Coastal Research, 13, 1–7. Holland, K.T., Holman, R.A. & Sallenger, A.H. (1991) Estimation of overwash bore velocities using video techniques. Coastal Sediments ’91. Seattle, Washington, USA: ASCE. Jennings, J.N. & Coventry, R.J. (1973) Structure and texture of a gravelly barrier island in the Fitzroy Estuary, Western Australia, and the role of mangroves in the shore dynamics. Marine Geology, 15, 145–167. Kochel, R.C. & Dolan, R. (1986) The role of overwash on a mid-Atlantic coast barrier island. Journal of Geology, 94, 902–906. Kraus, N.C., Militello, A. & Todoroff, G. (2002) Barrier breaching processes and barrier spit breach, Stone Lagoon, California. Shore and Beach, 70, 21–28.

192

COASTAL STORMS: PROCESSES AND IMPACTS

Kroon, A., Kabuth, A.K. & Westh, S. (2013) Morphologic evolution of a storm surge barrier system. Journal of Coastal Research, 529–534. Leatherman, S.P. (1976) Quantification of overwash processes. PhD, University of Virginia. Leatherman, S.P. (1977) Overwash hydraulics and sediment transport. Coastal Sediments ’77. Charleston, South Carolina, USA. Leatherman, S.P., Williams, A.T. & Fisher, J.S. (1977) Overwash sedimentation associated with a large-scale northeaster. Marine Geology, 24, 109–121. Leatherman, S.P. & Zaremba, R.E. (1987) Overwash and aeolian processes on a US Northeast coast barrier. Sedimentary Geology, 52, 183–206. Lorang, M.S. (2002) Predicting the crest height of a gravel beach. Geomorphology, 48, 87–101. Matias, A., Vila-Concejo, A., Ferreira, Ó., Morris, B. & Dias, J.A. (2009) Sediment dynamics of barriers with frequent overwash. Journal of Coastal Reseach, 25 (3), 768–780. Matias, A., Ferreira, Ó., Vila-Concejo, A., Morris, B. & Dias, J.A. (2010) Short-term morphodynamics of non-storm overwash. Marine Geology, 274, 69–84. Matias, A., Williams, J.J., Masselink, G. & Ferreira, Ó. (2012) Overwash threshold for gravel barriers. Coastal Engineering, 63, 48–61.Matias, A., Masselink, G., Kroon, A., Blenkinsopp, C.E. & Turner, I.L. (2013) Overwash experiment on a sandy barrier. Journal of Coastal Research, SI 65, 778–783. Matias, A., Blenkinsopp, C.E. & Masselink, G. (2014) Detailed investigation of overwash on a gravel barrier. Marine Geology, 350, 27–38. Matias, A., Masselink, G., Castelle, B., Blenkinsopp, C. E. & Kroon, A. (2016) Measurements of morphodynamic and hydrodynamic overwash processes in a large-scale wave flume. Coastal Engineering, 113, 33–46. May, V.J. & Hansom, J.D. (2003) Coastal Geomorphology of Great Britain. Geological Conservation Review Series. Peterborough. Morton, R.A., Gonzalez, J.L., Lopez, G.I. & Correa, I.D. (2000) Frequent non-storm washover of barrier islands, Pacific coast of Colombia. Journal of Coastal Research, 16, 82–87. Morton, R.A. & Sallenger, A.H. (2003) Morphological impacts of extreme storms on sandy beaches and barriers. Journal of Coastal Research, 19, 560–576. Nguyen, X.T., Donnelly, C. & Larson, M. (2006) A new empirical formula for coastal overwash volume. Vietnam-Japan Estuary Workshop 2006. Hanoi, Vietnam. Nichol, S.L. & Boyd, R. (1993) Morphostratigraphy and facies architecture of sandy barriers along the Eastern Shore of Nova Scotia. Marine Geology, 114, 59–80. Obhrai, C., Powell, K. & Bradbury, A. (2008) A laboratory study of overtopping and breaching of shingle barrier beaches. 31st International Conference on Coastal Engineering. Hamburg, Germany. Orford, J.D. & Carter, R.W.G. (1982) Crestal overtop and washover sedimentation on a fringing sandy gravel barrier coast, Carnsore Point, Southeast Ireland. Journal of Sedimentary Petrology, 52, 265–278. Orford, J.D., Carter, R.W.G., Forbes, D.L. & Taylor, R.B. (1988) Overwash occurrence consequent on morphodynamic changes following lagoon outlet closure on a coarse clastic barrier. Earth Surface Processes and Landforms, 13, 27–35. Orford, J.D., Carter, R.W.G., Jennings, S.C. & Hinton, A.C. (1995) Processes and timescales by which a coastal gravel-dominated barrier responds geomorphologically to sea-level rise: Story Head Barrier, Nova Scotia. Earth Surface Processes and Landforms, 20, 21–37. Orford, J.D., Jennings, S. & Pethick, J. (2003) Extreme storm effect on gravel-dominated barriers. Coastal Sediments ’03. Florida: ASCE.

CH9 OVERWASH PROCESSES

193

Park, Y.H. & Edge, B.L. (2010) An empirical model to estimate overwash. Journal of Coastal Reseach, 26, 1157–1167. Phantuwongraj, S., Choowong, M., Nanayama, F., Hisada, K., Charusiri, P., Chutakositkanon, V., et al. (2013) Coastal geomorphic conditions and styles of storm surge washover deposits from Southern Thailand. Geomorphology, 192, 43–58. Pirrello, M.A. (1992) The role of wave and current forcing in the process of barrier island overwash. MSc, University of Florida. Priestas, A.M. & Fagherazzi, S. (2010) Morphological barrier island changes and recovery of dunes after Hurricane Dennis, St George Island, Florida. Geomorphology, 114, 614–626. Pullen, T., Allsop, N.W.H., Bruce, T., Kortenhaus, A., Schuttrumpf, H. & Van Der Meer, J.W. (2007) Wave overtopping of sea defences and related structures: Assessment manual. EurOtop. Die Küste, 73. Pye, K. & Blott, S.J. (2009) Progressive breakdown of gravel-dominated coastal barrier, Dunwich-Walberswick, Suffolk, UK: Processes and implications. Journal of Coastal Reseach, 25, 589–602. Qi, H., Cai, F., Lei, G., Cao, H. & Shi, F. (2010) The response of three main beach types to tropical storms in South China. Marine Geology, 275, 244–254. Raji, O., Dezileau, L., Von Grafenstein, U., Niazi, S., Snoussi, M. & Martinez, P. (2015) Extreme sea events during the last millennium in the northeast of Morocco. Natural Hazards and Earth System Sciences, 15. Ritchie, W. & Penland, S. (1988) Rapid dune changes associated with overwash processes on the deltaic coast of South Louisiana. Marine Geology, 81, 97–122. Rodríguez, R.W., Webb, R.M.T. & Bush, D.M. (1994) Another look at the impact of Hurricane Hugo on the shelf and coastal resources of Puerto Rico, USA. Journal of Coastal Research, 10, 278–296. Sallenger, A., Wright, C.W. & Lillycrop, J. (2007) Coastal-change impacts during Hurricane Katrina: An overview. Coastal Sediments ’07. New Orleans: ASCE. Schüttrumpf, H. & Oumeraci, H. (2005) Layer thicknesses and velocities of wave overtopping flow at seadikes. Coastal Engineering, 52, 473–495. Schwartz, R.K. (1975) Nature and genesis of some storm washover deposits. Technical memorandum 61. Coastal Engineering Research Center. Schwartz, R.K. (1982) Bedform and stratification characteristics of some modern small-scale washover sand bodies. Sedimentology, 29, 835–849. Silva, A.L.C., Silva, M.A.M., Gambôa, L.A.P. & Rodrigues, A.R. (2014) Sedimentary architecture and depositional evolution of the Quaternary coastal plain of Maricá, Rio de Janeiro, Brazil. Brazilian Journal of Geology, 44 (2). Srinivas, R., Dean, R.G. & Parchure, T.M. (1992) Barrier island erosion and overwash study – Volume 1. Coastal and Ocean Engineering Department, University of Florida. Stéphan, P., Suanez, S. & Fichaut, B. (2010) Franchissement et migration des cordons de galets par rollover. Impact de la tempête du 10 mars 2008 dans l’évolution récente du Sillon de Talbert (Côtes-d’Armor, Bretagne). Norois, 215, 59–75. Stockdon, H., Doran, K.S. & Sallenger, A.H. (2009) Extraction of lidar-based dune-crest elevations for use in examining the vulnerability of beaches to inundation during hurricanes. Journal of Coastal Reseach, 59–65. Stone, G., Liu, B., Pepper, D.A. & Wang, P. (2004) The importance of extratropical and tropical cyclones on the short-term evolution of barrier islands along the northern Gulf of Mexico, USA. Marine Geology, 210, 63–78.

194

COASTAL STORMS: PROCESSES AND IMPACTS

Swarzenski, P. (2014) Assessing the vulnerability of Pacific atolls to climate change. Sound Waves, March/April 2014, 1–3. Tega, Y. & Kobayashi, N. (1996) Wave overwash of subaerial dunes. 25th International Conference on Coastal Engineering. Orlando, Florida, USA, pp. 4148–4160. Tega, Y. & Kobayashi, N. (1999) Numerical modeling of overwashed dune profiles. Coastal Sediments ’99. New York: ASCE. Tribe, H.M. & Kennedy, D.M. (2010) The geomorphology and evolution of a large barrier spit: Farewell Spit, New Zealand. Earth Surface Processes and Landforms, 35, 1751–1762. Tuan, T.Q. & Verhagen, H.J. (2008) Breach initiation by the response of coastal barriers during storm surges. 7th International Conference on Coastal and Port Engineering in Developing Countries (COPEDEC 2008). Dubai, United Arab Emirates: PIANC. Vila-Concejo, A., Matias, A., Pacheco, A., Ferreira, Ó. & Dias, J.A. (2006) Quantification of inlet-related hazards in barrier island systems. An example from the Ria Formosa (Portugal). Continental Shelf Research, 26, 1045–1060.

10 Modeling the Morphological Impacts of Coastal Storms Ap van Dongeren1 , Dano Roelvink1,2 , Robert McCall1 , Kees Nederhoff1 and Arnold van Rooijen1,3 1 Deltares,

Delft, The Netherlands Institute, Delft, The Netherlands 3 University of Western Australia, Crawley, WA, Australia 2 UNESCO-IHE

10.1

Introduction

This chapter discusses the modeling of morphodynamical storm impacts on coasts. We will give an overview of classes of models, the physical processes that each class resolves, and the model class applicability on the different coastal environments discussed in the previous chapters. We discuss recent advances in coastal storm impact modeling, with examples of applications, and an outlook of modeling challenges and opportunities ahead. Any model is a schematized representation of reality. The real world is usually too complex to represent with a model, and the challenge is thus to capture reality at a level such that a model is still useful (i.e. still resolves the essential processes) but is not too complex and cumbersome to use. This is true in general but also for coasts, which have many time and spatial scales and details. For every coastal environment we need to ask the questions: what are the relevant and dominant processes that control storm impacts? And: do we need to simulate these processes directly, or can we represent them in some way, or even can we neglect them altogether? Thus there is a triage of ‘direct simulation’, ‘representation’ or ‘elimination’. The approach taken in this chapter is somewhat biased towards sandy coasts. This is for good reason, as processes on these types of coasts have been studied most intensively in the past. The reason for this is that sandy coasts at the mid-latitudes protect high-investment coastal zones (Europe, USA mid-Atlantic) and have therefore received the most attention. However, as has been shown in the previous chapters, more and more attention is being paid to other types of coasts as economic development is increasing, climate-change effects are being felt and the eco-system services that these coastal types provide are being appreciated more. These coastal types are discussed in this chapter as well. Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

196

COASTAL STORMS: PROCESSES AND IMPACTS

With respect to coastal impact modeling, and specifically impacts on sandy coasts, two general types of model classes can be identified: empirical models and process-based models. We will find that there are ‘pure’ empirical models, but there are no practical ‘pure’ process-based models. To date, every process-based model includes some representation of processes by empirical closure. We note that besides these two processes, Roelvink and Brøker (1993) also name descriptive or conceptual models (such as Sallenger, 2000), but as this type does not provide quantitative response information, it is not included here.

10.1.1 Empirical models Empirical models typically use observations to relate the response of a coast to the forcing. Purely empirical models are thus ‘black box’ models that do not consider or incorporate the underlying physical processes. Examples of this type of model are Dean (1977), Van de Graaff (1977), Vellinga (1986) and Van Gent et al. (2008), which relate the post-storm profile shape to stationary wave properties and sediment properties, independent of the pre-storm state. This type of model relies on observations taken in the field or the laboratory. Field observations are typically sparse and may not include all forcing parameters, which necessitates a simplification of the response model and thus limits the predictive value. Laboratory data is usually more controlled, that is, more quantitative data of more forcing and response parameters are taken under known conditions. However, laboratory experiments are performed with explicit (and implicit) assumptions and limitations. Typically, laboratory data is taken in a flume, which assumes longshore uniformity in the wave and water level forcing, the initial state of the coast and its response. This means that the coastal response process is considered to be a 1D process. Moreover, empirical models are only valid for the range of data obtained; hence empirical models are strictly limited to the coastal profile, sediment characteristics and the range of forcing parameters (wave heights and water levels). Furthermore, the sediment transport does not scale according to the hydrodynamical Froude Scale, which results in so-called vertically-distorted scale models. For a discussion on this aspect we refer to Vellinga (1986) and Van Thiel de Vries (2009) The benefit of these models is that they are computationally fast, but the drawback is that their applicability strictly is limited. Another type of empirical model is the convolution model that describes the time evolution of the profile (Kriebel, 1982; Kriebel & Dean, 1993; Madsen & Plant, 2001). This model is based on simple analytical solutions to predict the longshore uniform, time-dependent beach and dune profile response to storm scale variations in water level and breaking wave height. The underlying assumption is that the coastal profile will exponentially tend to an equilibrium shape for a given wave condition and characteristic erosion time scale. Yates et al. (2009) presented a model to predict the shoreline position on the US Pacific coasts, where the predicted shoreline change is a function of the previous state (shoreline location) and the equilibrium energy state. They did not include the entire profile, however. Yates et al. (2011) extended the model to storm conditions, and showed that it was applicable to other sites as well. Davidson et al. (2013) concurrently developed a similar model for the Australian East Coast. These models are also discussed in Chapter 10.

CH10 MODELING THE MORPHOLOGICAL IMPACTS

197

10.1.2 Process-based models Process-based models rely on the principle that the coastal response is governed by known physical processes of wave dynamics and sediment transport dynamics. This type of model requires that the coastal system and its governing processes are known well enough such that the relevant and dominant processes can be included in the model. This also means that they are usually far more complex and computationally more intensive than empirical models. However, since they are based on universal physical principles, these are more generally applicable. For instance, various regimes in coastal forcing (e.g. Sallenger, 2000) can be then be simulated seamlessly: alongshore non-uniformity in the coastal shape, forcing and response can be taken into account; different coastal environments can be modeled; and aspects such as buildings in the coastal impact zone can be modeled. In reality and in practicality, ‘pure’ process-based models, that is, models that are fully-dependent on physical process formulation are not feasible. In our case of coastal storm impact, this would require Direct Numerical Simulations of the Navier-Stokes equations, including sediment transport in three dimensions over a range of time scales (from the time-scale of turbulence to the time-scale of the storm event). Thus, only part of the physical processes is simulated directly, whereas the rest is represented or neglected. The model developer’s trade is to analyze the processes and determine which are the essential processes to be modeled directly and which can or must be modeled by representation. The reason for the latter may be that either the processes are unknown or are too computationally expensive. Even with the most computationally-expensive processes parameterized, the simulations of even the essential processes may be demanding. In practice this has led to two versions of process-based models: profile (1D) models and area (2D) models. Similarly to empirical models, profile models implicitly assume alongshore uniformity in the coastal properties, forcing and response. However, they are more generally applicable and can include various Sallenger regimes, wave directional spreading, the presence of seawalls, cross-shore variability in coastal properties and shape, and time-dependent forcing and response. Two-dimensional models are required in the case of strong coastal curvature, alongshore variation in forcing by surge or water levels, transitions from one coastal type to another. Examples of 1D and 2D models are detailed in the following sections. A more thorough discussion can be found in Ciavola et al. (2015).

10.1.2.1 1D (profile) models Many 1DV semi-empirical models are used throughout the world. For the purpose of this discussion, we will look at models that include the calculation of the eroded volume in the submerged profile as well as in subaerial profile. Therefore, profile models such as UNIBEST-TC (Reniers et al., 1995), LITPROF (Brøker-Hedegaard et al., 1991) and COSMOS (Nairn & Southgate, 1993), which only simulate the submerged profile are of limited use in this respect. Here, we will discuss the internationally well-known models SBEACH, DUROSTA and CSHORE, which do include the subaerial part. SBEACH (Larson & Kraus (1989), Larson et al. (1990) and Larson et al. (2004a)) is a 1DV semi-empirical time-dependent dune erosion model which uses time-averaged process-based wave transformation and empirical-based sediment transports. It thus couples a stationary hydrodynamic approach to a non-stationary morphodynamic one,

198

COASTAL STORMS: PROCESSES AND IMPACTS

on the assumption that the morphodynamic time scale is slower that the hydrodynamic one so that the waves have ample time to adjust. The dry dune supplies sediment through an empirical relation between wave impact and the weight of the sand which is eroding (Overton & Fisher, 1988), with innovations by Nishi & Kraus (1996), Larson et al. (2004b) and Palmsten & Holman (2011). While it is technically a process-based model, it is for a large part based on empirical formulations. A different type of model in this class is represented by DUROSTA (Steetzel, 1993), also known as UNIBEST-DE. It solves the processes of wave transformation (refraction, wave breaking), cross-shore and alongshore wave-induced currents, sediment transport and morphological change for time-varying hydraulic conditions. While strictly a 1D model, it does allow for the parameterized effects of wave obliquity, alongshore current gradients and coastal curvature. Dune erosion is represented by extrapolating near dune sediment transports over the dry dune face using an estimate for the wave runup. Finally, a third type of model is CSHORE (Kobayashi et al., 2009) which was developed for use by the US Corps of Engineers for ocean and great lake coasts for cases with long straight coasts where gradients in longshore directed transport are negligible and waves constitute the principal generation mechanism for sediment suspension. In CSHORE waves, currents and sediment transport are computed simultaneously through an iterative landward-marching procedure. The model development has focussed on process-based sediment transport formulations for a nearshore breaking wave environment. The model has since been extended to two horizontal dimensions; see below. The above models assume along-shore uniform conditions, both in the hydrodynamic forcing and in the coastal response, and have been applied successfully along relatively undisturbed coasts. However, as stated above, there are a number of limiting conditions for its general applicability, which is where 2DH area models are fit for purpose.

10.1.2.2 2D (area) models In complex coastal environments, empirical models and process-based 1D profile models do not suffice. This complexity can take a number of forms. The coast may: • Have an alongshore-varying topography, for instance if the dune elevation displays a spatial variation in natural dune system • Have an alongshore-varying bathymetry, for instance it may be fronted by deep channels, or be associated with a nearby tidal inlet • Be strongly curved, for instance in the case of barrier island heads • Be forced by alongshore-varying wave and water level conditions • Show a transition between coastal types, for instance a rocky headland with a sheltered pocket beach • Be partly vegetated • Include a transition of hard (engineered) coastal protection to a sandy (softer) environment • Include discrete buildings

CH10 MODELING THE MORPHOLOGICAL IMPACTS

199

Figure 10.1 Areas (marked in red) along the Dutch coast where the assumption of alongshore uniformity is violated (courtesy of Dr M. Boers, Deltares).

An analysis for the (mostly sandy) Dutch coast revealed that an estimated 40% of its alongshore length violates the assumption of alongshore uniformity for a number of the above reasons, see Figure 10.1. These situations call for a 2D-process-based model, which includes hydrodynamics and morphodynamics on the storm time-scale. We will discuss two 2DH models of this type.

200

COASTAL STORMS: PROCESSES AND IMPACTS

One model is C2Shore Johnson & Grzegorzewski (2011); Sleath-Grzegorzewski et al., 2013). This model is an extension of the CShore model and is coupled with a spectral wave model STWAVE (Smith et al., 2001) and a circulation model (Westerink et al., 1994), which computes currents and (tidal and surge) water levels. Johnson & Grzegorzewski (2011) present the model formulations and an application to Ship Island, Mississippi, USA for the case of Hurricane Katrina. They find that beach erosion and shoreline retreat were well predicted, but the deposition on the lee of the island was severely overestimated. They note substantial uncertainty in hydraulic boundary conditions and a lack of conservation of sand in the observations. Sleath-Grzegorzewski et al. (2013) applied the model to study the effect of Ship Island restoration. Another, and more widely-used, model is XBeach (Roelvink et al., 2009), which was developed to simulate the seamless hydrodynamic and morphological impact of storms and hurricanes on complex coasts. The model has two modes: a hydrostatic or ‘surfbeat’ mode and a non-hydrostatic mode. In the hydrostatic model, the hydrodynamic processes are separated into motions at the time scale of the short waves and motions at longer time scales, such as currents and long (infragravity) waves. A principle sketch is given in Figure 10.2. The short-wave motion is solved using the wave action equation, which solves the variation of short-waves envelope (wave amplitude) on the scale of wave groups (dark blue line), rather than the time trace of the individual short waves themselves (black line). It employs a dissipation model for use with wave groups (Roelvink, 1993; Daly et al., 2012) and a roller model (Svendsen, 1984; Nairn et al., 1990; Stive & de Vriend, 1994) to represent momentum stored at the surface after breaking. These variations, through radiation stress gradients (Longuet-Higgins & Stewart 1962, 1964) exert a force on the water column and drive longer period waves (infragravity waves) and unsteady currents, which are solved by the nonlinear shallow water equations (e.g. Phillips, 1977; Svendsen, 2003). The infragravity wave motions typically consist of incoming waves that propagate with (and are bound to) the wave groups (light blue line), as well as free components, which typically propagate offshore (red line). The hydrodynamics drive sediment transports under wave and flow conditions, following Van Rijn et al. (2007a,b) and Van Thiel-Van Rijn (Van Thiel de Vries, 2009)

short waves envelope Cg short waves

gh

bound long wave leaky wave

Figure 10.2 A principle sketch of short wave motions (black), the short wave envelope (dark blue), the incoming bound long wave (light blue) and the reflected free long wave (red) (courtesy of Dr Ad Reniers).

CH10 MODELING THE MORPHOLOGICAL IMPACTS

201

transport equations. The sediment transport includes an empirical formulation for avalanching (slumping) of dune front. On the basis of transport gradients the bathymetric update is computed. We refer to Roelvink et al. (2009) for a full description of the model. The model is applicable on spatial scales of order 10 × 10 kilometers, which includes the wave shoaling and surfzone, barrier islands and the back-barrier lagoon system, if present. It allows for the modeling of ‘hard’ structures such as seawalls and buildings. The model is boundary curve fitting (curvilinear) and can be driven with measured or modeled boundary conditions, obtained from larger area models. The model has been validated with a series of analytical, laboratory and field test cases (Roelvink, et al., 2009; van Thiel de Vries, 2009; Van Dongeren et al., 2009), and applied in a number of coastal environments, which will be addressed below. In the non-hydrostatic mode, the depth-averaged flow due to waves and currents is computed using the non-linear shallow water equations, but includes a non-hydrostatic pressure term, so that the dispersive short wave motion (black line in Figure 10.2) is resolved. The depth-averaged normalized dynamic pressure is derived in a method similar to a one-layer version of the SWASH model (Zijlema et al., 2011). The depth averaged dynamic pressure is computed from the mean of the dynamic pressure at the surface and at the bed by assuming the dynamic pressure at the surface to be zero and a linear change over depth. The main advantages of the non-hydrostatic mode are that the incident-band (short wave) runup and associated overwash are included, which is especially important on steep slopes such as gravel beaches. Another advantage is that the wave asymmetry and skewness are resolved by the model and no approximate local model or empirical formulation is required for these terms. Finally, in cases where diffraction is a dominant process, wave-resolving modeling is needed as it is neglected in the short wave averaged mode. When using the non-hydrostatic model, a much higher spatial resolution is needed to resolve the short waves. In an explicit numerical scheme, this results in smaller time steps, making this mode much more computationally expensive.

10.1.3 Process-model applications In this section, we will discuss the applicability of process-based models (for which we took XBeach as the example) in various coastal environments: sandy, gravel, coral reef, vegetated and urbanized coasts. We will show not only the model’s merits but also its drawbacks and the need for further development of process-based models in general.

10.1.3.1 Sandy coasts XBeach has been extensively applied and tested on various sandy coasts. One of the first applications of XBeach was on Hurricane Ivan impact on Santa Rosa Island, Florida. This barrier island is part of the Gulf Coast National Seashore, which is sparsely vegetated and not urbanized. Hurricane Ivan made landfall on 16 September 2004 just to the west of the island, with maximum onshore winds, waves and surge occurring to the east of the hurricane eye. McCall et al. (2010) showed that the island was subject to a sequence of collision, overwash and inundation regimes, which caused dune erosion and deposition in the nearshore at first, followed

202

COASTAL STORMS: PROCESSES AND IMPACTS

Figure 10.3 Four stages of the morphodynamical impact of Hurricane Ivan on Santa Rosa Island, Florida. The Gulf of Mexico is in the front and the Santa Rosa Sound is at the back. Top left: collision regime, top right: overwash regime, bottom left: inundation regime, bottom right: topography after recession of the flood. Reprinted from McCall et al., 2010, Coastal Engineering, with permission from Elsevier.

by overwash erosion and deposits on and behind the island in typical overwash fans, see Figure 10.3. The little infrastructure on this section of the island, mostly consisting of roads and parking lots was destroyed. The model, driven by wave and surge time series based on field data and large-scale numerical model results, was capable of predicting the morphological changes. The skill of the model was high (66% of variance explained, maximum bias −0.21 m), albeit these results were obtained using a sheet flow sediment transport limiter, which maximizes sediment transport in the case of extreme high flows (Froude numbers) that are outside the calibration range of the Van Thiel-Van Rijn sediment transport formulation. To overcome this incomplete formulation, De Vet et al. (2015) removed the sheet flow limiter, and used a more physics-based approach of a better approximation of the bed roughness and wave skewness and asymmetry in the case of overwash and breaching on a Long Island barrier island under Hurricane Sandy conditions. Lindemer et al. (2010) applied the model to the Chandeleur Islands (Louisiana, USA), which is a detached (almost relic) barrier island off the coast of the Mississippi River birdfoot, and was completely inundated during the storm. The authors show that qualitatively the patterns of erosion and channel formation were predicted well, but the magnitude of the erosion was underpredicted. Specifically, XBeach correctly

CH10 MODELING THE MORPHOLOGICAL IMPACTS

203

predicted erosion of the sandy berm and regions that became subaqueous, but areas that remained subaerial were not well predicted. They cite as causes, uncertainties in pre-storm topography (derived from older data), as well as incomplete sediment transport formulations due to sediment diameter variations and the existence of vegetation. Uncertainties in the hydrodynamic forcing were found to have a small effect on the inundation regime. Splinter & Palmsten (2012) evaluated XBeach and two parametric models for the case of storm impact on the East coast of Australia. They found that XBeach could reproduce both the dune toe retreat and dry beach volume change, but only after careful calibration of its parameters. Without calibration, the empirical model proposed by Palmsten & Holman (2012) performed best on dune toe retreat, but underestimated the dry beach volume change. Van Dongeren et al. (2009) presented an overview of beach profile changes due to storms on eight European beaches, including a comparison of model results obtained with off-the-shelf models. The results showed that the XBeach has skill in predicting the coastal profile, albeit that in most cases the erosion around the mean water line is over predicted and the depositions at the lower beach face are over predicted. In follow-up papers on beaches included in that study, Vousdoukas et al. (2012) extensively calibrated the model to predict morphological response to storm events along a meso-tidal, steeply sloping beach near Faro (Portugal). They found that in the case of steeper sloped beaches, the default parameter set derived for dissipative beaches over predicts the morphological change, with resulting Brier Skill Scores (Van Rijn et al., 2003) from 0.2 to 0.72. Values below zero are labelled ‘bad’; in the range of 0–0.3, ‘poor’; 0.3–0.6, ‘reasonable’; 0.6–0.8, ‘good’; and 0.8–1.0, ‘excellent’. Thus, the computed values in this case range between poor and good. Armaroli et al. (2013) applied the model to a sandy Adriatic beach, which is protected with offshore breakwaters. They found that the erosion of the upper beach and dune toe was reasonably well predicted, but the model did not reproduce the slope of the dune, as it does not account for biotic factors (e.g. plant roots), which explains the steeper observed dune slopes. Dissanayake et al. (2014) applied the model to evaluate the storm impact on the Sefton coast in north-west UK. Nested with a larger area model, XBeach predicted the beach change quite accurately, with BSS scores of 0.8 and above. Callaghan et al. (2013) considered the use of XBeach in a probabilistic approach to estimate storm erosion volumes. Compared with alternative (and computationallycheaper) storm erosion models such as the convolution model by Kriebel & Dean (1993) and the semi-empirical SBEACH model, the XBeach model performed well for a case study in Australia, provided that the entire erosion volume data set is used to calibrate XBeach. The advantage that XBeach predicts physically more realistic behavior offsets the relatively high computational demand. Splinter et al. (2014) hindcasted the cumulative impact of a series of relatively small storms that impacted the Gold Coast of Australia. In this case, four clustered storms caused more erosion than a single normative (1/100 per year annual probability) event. XBeach could reproduce the observed dry beach erosion volume to about 20% and shoreline retreat by about 10%. They show that artificial changes in the sequence of storms did not affect the total erosion volume. Karunarathna et al. (2014) concurrently analysed the impact of sequences of storms on a beach in New South Wales, Australia. They used XBeach to estimate the post-storm profile, as profile observations are often

204

COASTAL STORMS: PROCESSES AND IMPACTS

taken too long after the occurrence of an event and thus include part of the beach recovery phase. Thus, the model was used to fill data gaps, which made the analysis more precise. They also found that erosion due to a sequence of storms is consistently higher than for a comparable single event (in terms of wave power), but also that the time between storms and the rate of recovery in the intermediate periods play a role. The above shows that the storm impact model performed adequately but that some physical processes need attention, such as the erosion in the presence of vegetation, wave-driven onshore transport of sediment, sediment transport under sheet flow conditions and the effect of topographic roughness.

10.1.3.2 Gravel coasts Despite their wide-ranging use as cost-effective and sustainable forms of coastal defence, relatively little research has been directed at understanding the morphodynamics of gravel beaches in comparison to their sandy counterparts (Mason & Coates, 2001), and in particular their morphodynamic response to energetic wave conditions (Poate et al., 2013). Due to this lack of understanding of fundamental processes, few process-based models have been developed that are able to simulate the morphodynamics of gravel beaches, and even fewer have been applied and validated for storm impacts. Gravel beaches differ from sandy beaches during storms in three important aspects. First, gravel beaches are generally steep (𝛽 = 0.05–0.20), reflective and have a very narrow surfzone, leading to dominant forcing at the incident wave band over the infragravity band. Second, gravel beaches are relatively permeable, leading to substantial infiltration losses in the swash and subsequent asymmetry in the uprush and backwash volume. Finally, sediment transport on gravel beaches is dominated by bed load and sheet flow transport in the swash, due to the relatively high fall velocity of gravel and the absence of a dissipative surfzone leading to highly energetic conditions in the swash (Buscombe & Masselink, 2006). Van Gent (1995a, 1995b, 1996) presented the first promising numerical process-based model for the morphodynamic simulation of storm impacts on gravel beaches. The model simulates intra-wave motions of shallow water waves and groundwater inside the porous beach, and uses a critical threshold of motion to displace particles on the bed in an upslope or downslope direction. The model was validated using data from physical model experiments and one berm breakwater in the USA, in conditions that ranged from low to high energy. Pedrozo-Acuña et al. (2006, 2007) applied a modification of an existing Boussinesq wave model (COULWAVE; Lynett et al., 2002) to gravel beaches under mildly energetic forcing conditions. Although the model did not include groundwater processes, the model was found to reproduce the berm-building conditions observed in physical model experiment relatively well if the sediment friction factor in the uprush was increased with respect to that of the backwash. Groundwater processes, as well as the effects of acceleration in the swash and sediment fluidization under plunging breakers were hypothesized to cause the apparent difference in the sediment friction factor. Williams et al. (2012) and Jamal et al. (2014) applied a modified version of XBeach to simulate the morphodynamic response of gravel beaches during overwash and berm-building conditions, respectively. Both studies found the permeability of the gravel beach to be important in the simulation of morphological change. While neither XBeach model used in the study explicitly computed the incident swash, Jamal et al.

CH10 MODELING THE MORPHOLOGICAL IMPACTS

205

(2014) found that an additional parameterization of the bed return flow was required to reduce the dominantly offshore-directed transport of the surf-beat type approach. The latest process-based model to be applied successfully to model storm impacts on gravel beaches is XBeach-G (McCall et al., 2014). The model is a derivative of XBeach that includes the non-hydrostatic computation of the incident and infragravity wave band (Smit et al., 2010) and a groundwater model to account for swash infiltration losses (McCall et al., 2012). Through comparison with data collected during physical model experiments, as well as data collected at six natural gravel beaches, the model has been shown to simulate storm hydrodynamics, including wave runup and overtopping well (Masselink et al., 2015). Furthermore, the model has been shown to have considerable skill in predicting the morphodynamic response of gravel barriers across a wide range of forcing conditions and barrier response types (Figure 10.4), from berm building to barrier rollover, with minimal calibration (McCall et al., 2015). Berm building

Beach erosion 10

Elevation (m)

4 3

5

2 1 0

0 –10

0

10

20

–30

30

–20

Crest overtopping 6

8

Elevation (m)

10

4 2 0 –40

–20

0

20

0

10

20

Crest lowering

8

–2 –60

–10

40

6 4 2 0

0

150 50 100 Cross-shore distance (m)

200

Barrier rollover 5 4 3 2 1 0 –20

–10 0 10 20 Cross-shore distance (m)

30

Figure 10.4 Example of measured (black) and modeled (red) post-storm beach profiles for five separate events and locations. The pre-storm profiles are shown in grey, whereas the maximum still water levels are represented by the blue lines. Modified from McCall et al. (2015) under Creative Commons Attribution License (CC BY).

206

COASTAL STORMS: PROCESSES AND IMPACTS

10.1.3.3 Coral and rocky platform coasts Even though a large proportion of the world’s coastlines, perhaps as high as 80% (Emery & Kuhn, 1982), contain a broad class of submerged reef structures, including tropical coral reefs, very little work (compared to sandy beaches) has addressed the range of nearshore hydrodynamic processes in reef environments. Field and modeling reef studies (see Van Dongeren et al., 2013 for a review) show that reefs exhibit similar processes as seen on sandy beaches, with two important differences: the slope of a coral reef is generally much steeper and rougher, followed by a relatively flat-sloped reef top and lagoon. Therefore, the surfzone is much further away from the shoreline than on sandy beaches, which allows for a clear distinction between zones of wave generation, propagation and decay. Also, due to the presence of the lagoon momentum, separation between wave-induced circulation currents and setup takes place. The XBeach model had to be extended by introducing a dissipation term due to bottom friction in the wave action equation, following Jonsson (1966). This introduces a free parameter fw which can be constrained using field data by Lowe et al. (2007). Storm impacts are important on reef-lined coasts. Especially on small (atoll) islands, (swell) wave-induced runup, overtopping and inundation causes not only flooding hazards, but also salinization of the aquifer. Damlamian et al. (2013) used XBeach to create inundation risk maps for five atolls in French Polynesia, after careful calibration of short wave and long wave propagation and dissipation over a reef flat and accompanied by an extensive sensitivity study. Quataert et al. (2015) modeled runup using XBeach on Roi Namur (Kwajalein, Republic of Marshall Islands). While the wave transformation and mean water levels (due to setup) were predicted correctly, the most extreme runup events were under predicted. A likely cause for this is the fact that in the hydrostatic surfbeat model short waves are not taken into account, which means that in these cases a non-hydrostatic model should be used. 10.1.3.4 Vegetated coasts Coastlines, especially those between the tropics, may be fronted by different types of vegetation such as kelp, sea grass and mangrove forests. This vegetation has an effect on the impact of waves, currents and water levels on the hinterland and may help to reduce hydraulic loads and thus flood risk. The mechanism by which vegetation reduces the wave height is well known (Dalrymple et al., 1984; Løvås, 2000; Løvås & Tørum, 2001; Mendez & Losada, 2004). However, the effect of vegetation on the mean water level (or wave induced setup) is less well known, with just a few theoretical (Dean & Bender, 2006) and experimental studies (Wu et al., 2011) showing that, under certain hydrodynamic conditions, the presence of vegetation results in a lower mean water level near the coast. The effect of mangrove vegetation during storms has been documented by Mazda et al. (2006), Quartel et al. (2007), Bao (2011) and Phan et al. (2015), and the effect of salt marsh vegetation by Moller et al. (1999, 2014), among other references. A reduction through vegetation thus reduces wave runup, overtopping and morphological impact. Van Rooijen et al. (2015) implemented a more complete vegetation dissipation formulation in XBeach. The model was tested using data from different physical experiments. Figure 10.5 shows the wave height transformation and the wave-induced setup for the conventional case of no vegetation (black lines represent the model results and circles the observations), where waves decay due to depth-induced breaking, which causes radiation stress gradients and a balancing wave-induced setup. In the case of

CH10 MODELING THE MORPHOLOGICAL IMPACTS

207

Hmean [m]

0.08 0.06 0.04 0.02

X Beach (no VEG) XB–VEG. XB–VEG. + Measured: no VEG. Measured: with VEG.

0 10 ηmean [mm]

8 6 4 2 0 –2

z [m]

0.2 0 –0.2 –0.4 2

4

6

8

10 x [m]

12

14

16

Figure 10.5 Example result of measured (squares and circles) and modeled wave height (top panel) and mean water level (middle panel). The effect on the mean water level is indicated by the difference between the red and blue line. Figure, courtesy of Arnold van Rooijen, Deltares.

emerged vegetation, the short-wave height decays not only through breaking but also through dissipation in the canopy (blue line and squares), which is quite accurately modeled using the formulation by Mendez & Losada (2004); this is widely used and implemented in numerical models. By itself, the change in short wave height transformation has an effect in the spatial distribution of the radiation stress gradients, and thus on the setup profile, as evidenced by the blue line in the middle panel. However, there is still a mismatch between this result and the observations (see blue line), which is due to the effect of wave skewness that causes an onshore-directed net forcing on the vegetation field, thus reducing the setup. Incorporating these effects in the momentum equations yields a prediction, which shows that vegetation can dramatically reduce or even eliminate setup (red line in the middle panel).

10.1.3.5 Urbanized/hard structure coasts In the previous sections, natural coastal systems were discussed. However, in urbanized environments, the coast may contain hard elements such as dune foot revetments, seawalls, groins and buildings. In the presence of a structure and for the specific case of sandy coasts, dune erosion and overwash are strongly affected both in cross-shore and in longshore direction, as visualized in Figure 10.6.

208

COASTAL STORMS: PROCESSES AND IMPACTS

Side view

Top view

Figure 10.6 The impact of hard elements – both in cross-shore (left) and in the longshore direction(right). The blue line indicates the response of a coast without the structure, while the red line indicates the response in the presence of a structure. Figure, courtesy of Kees Nederhoff.

In the case of a sandy coast, during storm conditions sand is eroded from the dune and deposited in the nearshore area, which helps to reduce the wave impact on and erosion of the remaining dune. When in the cross-shore direction the dune face is intersected with a hard element, part of the sediment supply from the dune to the nearshore is blocked, while the initial offshore transport capacity caused by the attacking waves in front of the structure remains. As a result, a scour hole can develop (WL Delft Hydraulics, 1987). The amount of erosion (scour) can vary considerably and depends on (amongst other factors) on whether the waves reflect, overtop or break at the structure, and on sediment characteristics (Sumer & Fredsoe, 2002). A positive effect of the cut-off of sediment is that on the whole less erosion by volume in the cross-shore direction will occur. Irish et al. (2013) showed, using a Bousssinesq-type model without morphological change, that during Hurricane Sandy a seawall near Bay Head, NJ, reduced the momentum flux at the longshore transect by at least 50%. XBeach is used in a subsequent numerical study, which does include morphological change (Smallegan et al., 2015). Besides the effect of hard structures in the cross-shore direction, an alongshore interaction is also expected. Hard structures can increase the erosion volume of the adjacent coasts (WL Delft Hydraulics, 1993). There are two drivers for this effect: 1.

An alongshore exchange of sediment from the ‘sandy’ towards the ‘hard’ cross-section that is driven by setup differences. Hard-structure cross-sections are less dissipative due to the cut-off of sediment supply and therefore waves break right in front of a structure rather than at a distance offshore in the case of a soft cross-section. This initiates an alongshore setup difference and thus alongshore sediment transport (Van Geer et al., 2012). 2. Locally higher waves will impact the soft cross-section that will result in more erosion. These waves are more energetic, due to the weaker soft cross-section (due to driver 1) and diffraction around the construction, which increases the offshore sediment transport (Nederhoff et al., 2015). The alongshore effect of constructions have both been reproduced in laboratory experiments and in hindcasts of Hurricane Sandy. Boers et al. (2011) showed in the

CH10 MODELING THE MORPHOLOGICAL IMPACTS

10

0

0 400

600

400

zb (m)

zb (m)

10

–10 800

209

200

0

–10 800

600 800 0 1000 Longshore (m)

200

0 400 600

400

200

200

600 800 0 1000 Longshore (m)

Cross-shore (m)

Cross-shore (m) zb (m) –5

–4

–3

–2

–1

0

1

2

3

4

5

Figure 10.7 Pre (left) and post-Sandy (right) in a three dimensional plot with both bed and water levels as simulated by XBeach (Courtesy Kees Nederhoff).

laboratory that a dune-dike transition resulted in 27% more erosion at the adjacent coast and that this percentage can increase up to 88% for a breach in a dike. Nederhoff et al. (2015) made a hindcast of the impact of Hurricane Sandy on a condominium building at Camp Osborne, Brick, NJ (USA), see Figure 10.7. The presence of the building resulted in an increase of the erosion volume of the adjacent coast with a maximum of +32% (52 m3 /m) over a length of 266 m. Remarkable is the fact that this pattern of increase in erosion only occurred at one side, which was found to be related to the obliqueness of the incoming waves.

10.1.4 Operational models While storm impact morphological simulations are still computationally expensive, the incorporation of such models in operational models comes into view. Haerens et al. (2012) showed results from the EU-funded MICORE project of the construction of a number of operational storm early warning systems in Europe. Vousdoukas et al. (2012) demonstrated the result of the Faro (Portugal) case study site from the same project where nested XBeach models were forced by an existing operational wave-forecast model to generate daily forecasts of storm impacts, whereas Harley et al. (2011) demonstrated a system for the Italian Emilia-Romagna Coast. In Van Dongeren et al. (2014) and Van Verseveld et al. (2015), this approach is furthered to include damage due to inundation, wave impact and morphological change, for a number of case study sites. On an even larger scale, Barnard et al. (2014) incorporated hundreds of XBeach transect models in an operational morphodynamic forecast system for the southern California coast, with the objective to predict cliff failure. The model system identified coastal sections that are vulnerable to a range of current and future oceanographic coastal hazards.

10.2

Outlook

This chapter gives an overview of model classes with the physical processes that each class resolves, and the model class applicability on the different coastal environments, such as sandy, gravel and coral/rock coasts, as well as coasts with hard structures and vegetation. It is clear that storm impact processes on each of these coasts are complex

210

COASTAL STORMS: PROCESSES AND IMPACTS

and are composed of many sub-processes, such as wave dissipation, dune avalanching, offshore/alongshore and onshore transports, to name just a few. This implies careful understanding of each process in isolation, which is best done in controlled laboratory environments. With this data and understanding, process formulations can be calibrated. Many experimental results have already been collected but more are needed in order to further test coastal impact models. Field data of storm events, with well-documented pre-existing conditions, hydrodynamic boundary conditions of waves, wind and surge, and the storm impact measured directly after the storm, are needed to validate models on the prototype scale. We foresee that in the future many more physical processes which act at the storm time scale will be implemented in models. One can think of the effect of vegetation on morphological change, the effect of buildings on coastal change, sediment transports on beaches composed of gravel and sand, but also the inundation of the hinterland, the infiltration of seawater in aquifers and the damage of infrastructure. In addition, the recovery processes after a storm will also become important, to answer the question what the long-term behavior of the coast, which is hit successively by storms, will be. More complex models need good calibration of subprocesses and validation on field data, but also need to be manageable in terms of computational expense. Here, besides an expected increase in raw computational power, smart techniques such as multi-processor implementation, code optimization and cloud computing will bring the simulation of coastal storm impacts closer to engineering and forecasting practice.

Acknowledgements This chapter is a contribution to the RISC-KIT project (EU contract 603458, www .risckit.eu) and is partially funded by the Deltares Research program on ‘Hydro- and Morphodynamics during Extreme Events’ (project number 1220002).

References Armaroli, C., Grottoli, E., Harley, M.D. & Ciavola, P. (2013) Beach morphodynamics and types of foredune erosion generated by storms along the Emilia-Romagna coastline, Italy. Geomorphology, 199, 22–35. DOI: 10.1016/j.geomorph.2013.04.034. Bao, T.Q. (2011) Effect of mangrove forest structures on wave attenuation in coastal Vietnam. Oceanologia, 53 (3), 807–818. Barnard, P.L., van Ormondt, M., Erikson, L.H., Eshleman, J., Hapke, C., Ruggiero, P., et al. (2014) Development of the Coastal Storm Modeling System (CoSMoS) for predicting the impact of storms on high-energy, active-margin coasts. Natural Hazards, DOI: 10.1007/s11069-014-1236-y. Boers, M., Van Geer, P & Van Gent, M. (2011) Dike and dune revetment impact on dune erosion. Procedings Coastal Sediments 2011, Miami, FL. Brøker-Hedegaard, I., Deigaard, R. & Fredsøe, J. (1991) Onshore/offshore sediment transport and morphological modeling of coastal profiles. Coastal Sediments ’91, 643–657. Buscombe, D. & Masselink, G. (2006) Concepts in gravel beach dynamics. Earth-Science Reviews, 79, 33–52.

CH10 MODELING THE MORPHOLOGICAL IMPACTS

211

Callaghan, D.P., Ranasinghe, R. & Roelvink, D. (2013) Probabilistic estimation of storm erosion using analytical, semi-empirical, and process based storm erosion models. Coastal Engineering, 82, 64–75. DOI: 10.1016/j.coastaleng.2013.08.007 Ciavola, P., Ferreira, O., Van Dongeren, A., Van Thiel de Vries, J., Armaroli, C. & Harley, M. (2015) Prediction of storm impacts on beach and dune systems. In: P. Quevauviller (Ed.) Hydrometeorological Hazards: Interfacing Science and Policy, first edition. John Wiley & Sons, Ltd. Dalrymple, R.A., Kirby, J.T. & Hwang, P.A. (1984) Wave diffraction due to areas of energy dissipation. J. Waterw. Port Coast. Ocean Eng., 110 (1), 67–79. Daly, C., Roelvink, J.A., van Dongeren, A.R., van Thiel de Vries, J.S.M. & McCall, R.T. (2012) Validation of an advective-deterministic approach to short wave breaking in a surf-beat model, Coastal Engineering, 60, (Feb.), 69–83, ISSN 0378-3839, 10.1016/j.coastaleng.2011.08.001. Damlamian, H., Kruger, J., Turagabeci, M. & Kumar, S. (2013) Cyclone Wave Inundation Models for Apataki, Arutua, Kauehi, Manihi and Rangiroa Atolls, French Polynesia. SPC SOPAC Technical Report (PR176), September. Davidson, M.A., Splinter, K.D. & Turner, I.L. (2013) A simple equilibrium model for predicting shoreline change. Coastal Engineering, 73, 191–202. Dean, R.G. (1977) Equilibrium beach profiles: US Atlantic and the Gulf Coasts. Department of Civil Engineering, Ocean Engineering Report No. 12, University of Delaware, Newark, DE. Dean, R.G. & Bender, C.J. (2006) Static wave setup with emphasis on damping effects by vegetation and bottom friction. Coastal Engineering, 53 (2), 149–156. De Vet, P.L.M., McCall, R.T., Den Bieman, J.P., Stive, M.J.F. & Van Ormondt, M. (2015) Modeling dune erosion, overwash and breaching at Fire Island (NY) during Hurricane Sandy. Proc. Coastal Sediments 2015. Dissanayake, P., Brown, J. & Karunarathna, H. (2014) Modeling storm-induced beach/dune evolution: Sefton coast, Liverpool Bay, UK. Marine Geology, 357, 225–242. DOI: 10.1016/j.margeo.2014.07.013 Dolan, R. & Davis, R.E. (1992) An intensity scale for Atlantic coast northeast storms. Journal of Coastal Research, 8 (4), 840–853. Emery, K.O. & Kuhn, G.G. (1982) Sea cliffs: Their processes, profiles, and classification. Geological Society of America Bulletin, 93 (7), 644–654. Haerens, P., Ciavola, P., Ferreira, O., van Dongeren, A., van Koningsveld, M. & Bolle, A. (2012) Online operational early warning system prototypes to forecast coastal storm impacts (CEWS). Proceedings of Coastal Engineering 2012, in CD-ROM. Harley, M., Armaroli, C. & Ciavola, P. (2011) Evaluation of XBeach predictions for a real-time warning system in Emilia-Romagna, Northern Italy. Journal of Coastal Research, SI 64, 1861–1865. Irish, J.L., Lynett, P.J., Weiss, R, Smallegan, S.M. & Cheng, W. (2013) Buried relic seawall mitigates Hurricane Sandy’s impacts. Coast. Eng., 80, 79–82, DOI: dx.doi.org/10.1016/ j.coastaleng.2013.06.001, 2013. Jamal, M.H., Simmonds, D.J. & Magar, V. (2014) Modeling gravel beach dynamics with XBeach. Coastal Engineering, 89, 20–29. DOI: 10.1016/j.coastaleng.2014.03.006. Johnson, B.D. & Sleath-Grzegorzewski, A. (2011) Modeling nearshore morphological evolution of Ship Island during Hurricane Katrina. Proceedings Coastal Sediments 2011. Jonsson, I.G. (1966) Wave boundary layers and friction factors, Paper Proc. Tenth Conference on Coastal Engineering. ASCE, Tokyo, Japan.

212

COASTAL STORMS: PROCESSES AND IMPACTS

Karunarathna, H., Pender, D., Ranasinghe, R., Short, A.D. & Reeve, D.E. (2014) The effects of storm clustering on beach profile variability. Marine Geology, 348, 103–112. DOI: 10.1016/j.margeo.2013.12.007. Kobayashi, N., Payo, A. & Johnson, B.D. (2009) Suspended sand and bedload transport on beaches. In: Handbook of Coastal and Ocean Engineering. World Scientific. Chapter 28, pp. 807–823. Kriebel, D.L. (1982) Beach and dune response to hurricanes. Unpublished MS thesis, Department of Civil Engineering, University of Delaware, Newark, NJ, USA. Kriebel, D.L. & Dean, R.G. (1993) Convolution method for time-dependent beach-profile response. Journal of Waterway, Port, Coastal, and Ocean Engineering, 119 (2), 204–226. Larson, M. & Kraus, C.N. (1989) SBEACH: Numerical model for simulating storm-induced beach change; Report 1, Empirical foundation and model development. Technical Report CERC-89-9, Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Larson, M., Kraus, C.N. & Byrnes, M.R. (1990) SBEACH: Numerical model for simulating storm-induced beach change; Report 2, Numerical formulation and model tests. Technical Report CERC-89-9, Coastal Engineering Research Center, US Army Engineer Waterways Experiment Station, Vicksburg, Mississippi. Larson, M., Wise, R.A. & Kraus, N.C. (2004) Modeling dune response due to overwash transport. Proc. Coast. Eng., Lisbon, Portugal, 2133–2145. Larson M., Erikson L. & Hanson H. (2004) An analytical model to predict dune erosion due to wave impact. Coastal Engineering, 51 (8–9), 675–696. Lindemer, C.A., Plant, N.G., Puleo, J.A., Thompson, D.M. & Wamsley, T.V. (2010) Numerical simulation of a low-lying barrier island’s morphological response to Hurricane Katrina. Coastal Engineering, 57 (11–12), 985–995. DOI: 10.1016/j.coastaleng.2010.06.004. Longuet-Higgins, M.S. & Stewart, R.W. (1962) Radiation stress and mass transport in gravity waves with application to ‘surf-beats’. J. Fluid Mech., 8, 565–583. Longuet-Higgins, M.S. & Stewart, R.W. (1964) Radiation stress in water waves, a physical discussion with applications. Deep Sea Res., 11, 529–563. Løvås, S.M. (2000) Hydro-physical conditions in kelp forests and the effect on wave damping and dune erosion: A case study on Laminaria hyperborea. PhD thesis, University of Trondheim, Norwegian Institute of Technology, Trondheim, Norway. Løvås, S.M. & Tørum, A. (2001) Effect of the kelp Laminaria hyperborea upon sand dune erosion and water particle velocities. Coastal Engineering, 44 (1), 37–63. Lowe, R.J., Falter, J.L., Koseff, J.R., Monismith, S.G. & Atkinson, M.J. (2007) Spectral wave flow attenuation within submerged canopies: Implications for wave energy dissipation. Journal of Geophysical Research, 112, C05018 10.1029/2006JC003605. Lynett, P.J., Wu, T.-R. & Liu, P.L.-F. (2002) Modeling wave runup with depth-integrated equations. Coastal Engineering, 46 (2), 89–107. Madsen, A.J. & Plant, N.G. (2001) Intertidal beach slope predictions compared to field data. Marine Geology, 173, 121–139. Mason, T. & Coates, T.T. (2001) Sediment transport processes on mixed beaches: A review for shoreline management. Journal of Coastal Research, 17 (3), 645–657. Masselink, G., McCall, R., Poate, T. & Van Geer, P. (2015) Modeling storm response on gravel beaches using XBeach-G. Maritime Engineering, 167 (4), 173–191. Mazda, Y., Magi, M., Ikeda, Y., Kurokawa, T. & Asano, T. (2006) Wave reduction in a mangrove forest dominated by Sonneratia sp. Wetlands Ecology and Management, 14 (4), 365–378.

CH10 MODELING THE MORPHOLOGICAL IMPACTS

213

McCall, R.T., Van Thiel de Vries, J.S.M., Plant, N.G., Van Dongeren, A.R., Roelvink, J.A., Thompson, D.M., et al. (2010) Two-dimensional time dependent hurricane overwash and erosion modeling at Santa Rosa Island. Coastal Engineering, 57 (7), 668–683. McCall, R., Masselink, G., Roelvink, J., Russell, P., Davidson, M. & Poate, T. (2012) Modeling overwash and infiltration on gravel barriers. In: Proceedings of the 33rd International Conference on Coastal Engineering. McCall, R.T., Masselink, G., Poate, T.G., Roelvink, J.A., Almeida, L.P., Davidson, M., et al. (2014) Modeling storm hydrodynamics on gravel beaches with XBeach-G. Coastal Engineering, 91, 231–250. DOI: 10.1016/j.coastaleng.2014.06.007. McCall, R., Masselink, G., Poate, T. & Roelvink, J. (2015) Modeling storm morphodynamics on gravel beaches with XBeach-G. Coastal Engineering, 103, 52–66. Mendez, F.J. & Losada, I.J. (2004) An empirical model to estimate the propagation of random breaking and nonbreaking waves over vegetation fields. Coastal Engineering, 51 (2), 103–118. Möller, I., Spencer, T., French, J.R., Leggett, D.J. & Dixon, M. (1999) Wave transformation over salt marshes: A field and numerical modeling study from north Norfolk, England. Estuarine, Coastal and Shelf Science, 49 (3), 411–426. Möller, I., Kudella, M., Rupprecht, F., Spencer, T., Paul, M., van Wesenbeeck, B., et al. (2014) Wave attenuation over coastal salt marshes under storm surge conditions. Nature Geoscience, 7 (Oct.). DOI: 10.1038/NGEO2251. Nairn, R.B., Roelvink, J.A. &. Southgate, H.N. (1990) Transition zone width and implications for modeling surfzone hydrodynamics, Proc. 22nd International Conference on Coastal Engineering, Am. Soc. Of Civ. Eng., New York. Nairn, R.B. & Southgate, H.N. (1993) Deterministic profile modeling of nearshore processes. Part 2. Sediment transport and beach profile development. Coastal Engineering, 19 (1–2), 57–96. Nederhoff, C.M., Lodder, Q.J., Boers, M., den Bieman, J.P. & Miller, J.K. (2015) Modeling the effects of hard structures on dune erosion and overwash: A case study of the impact of Hurricane Sandy on the New Jersey coast. Proceedings Coastal Sediments 2015, San Diego, CA. Nishi, R. & Kraus, N. (2000) Mechanism and calculation of sand dune erosion by storms. Proc. Int. Conf. on Coastal Engineering, 1 (25). DOI: 10.9753/icce.v25. Overton, M.F. & Fisher, J.S. (1988) Laboratory investigation of dune erosion. J. Waterway, Port, Coastal and Ocean Engineering, 114 (3), 367–373. Palmsten M. & Holman R.A. (2011) Infiltration and instability in dune erosion. Journal of Geophysical Research, 116, C10. Palmsten, M.L. & Holman, R.A. (2012) Laboratory investigation of dune erosion using stereo video. Coastal Engineering, 60 (1), 123–135. DOI: 10.1016/ j.coastaleng.2011.09.003. Pedrozo-Acuña, A., Simmonds, D., Otta, A. & Chadwick, A. (2006) On the cross-shore profile change of gravel beaches. Coastal Engineering, 53 (4), 335–347. Pedrozo-Acuña, A., Simmonds, D.J., Chadwick, A.J. & Silva, R. (2007) A numerical-empirical approach for evaluating morphodynamic processes on gravel and mixed sand-gravel beaches. Marine Geology, 241, 1–18. Phan, L.K., Van Thiel De Vries, J.S.M. & Stive, M.J.F. (2015) Coastal mangrove squeeze in the Mekong Delta. Journal of Coastal Research, 31 (2), 233–243. Phillips, O.M. (1977) The Dynamics of the Upper Ocean. Cambridge Univ. Press, New York, Poate, T., Masselink, G., Davidson, M., McCall, R., Russell, P. & Turner, I. (2013) High frequency in-situ field measurements of morphological response on a fine gravel beach during energetic wave conditions. Marine Geology, 342, 1–13.

214

COASTAL STORMS: PROCESSES AND IMPACTS

Quartel, S., Kroon, A., Augustinus, P.G.E.F., Van Santen, P. & Tri, N.H. (2007) Wave attenuation in coastal mangroves in the Red River Delta, Vietnam. Journal of Asian Earth Sciences, 29 (4), 576–584. Quataert E., Storlazzi, C., Van Rooijen, A., Van Dongeren, A.R. & Cheriton, O. (2015) The influence of coral reefs and climate change on wave-driven flooding of tropical coastlines. Geoph. Res. Letters, 42, 6407–6415, Reniers, A.J.H.M., Roelvink, J.A. & Walstra, D.J.R. (1995) Validation Study of Unibest-TC Model, Report H2130, Delft Hydraulics, Delft, The Netherlands. Roelvink, D., Reniers, A., van Dongeren, A., van Thiel de Vries, J., McCall, R. & Lescinski, J. (2009) Modeling storm impacts on beaches, dunes and barrier islands. Coastal Engineering, 56, 1133–1152. Roelvink, J.A. (1993) Surf beat and its effect on cross-shore profiles. PhD thesis. Delft Univ. of Technology, Delft, Netherlands. Roelvink, J.A. & Brøker, I. (1993) Cross-shore profile models. Coastal Engineering, 21, 163–191. Sallenger, A.H. (2000) Storm Impact Scale for Barrier Islands. Journal of Coastal Research, 16 (3), 890–895. Sleath-Grzegorzewski A.S., Johnson, B.D., Wamsley, T.V. & Dean Rosati, J. (2013) Sediment Transport and morphology modeling of Ship Island, Mississippi. USA, during storm evens. Proc. Coastal Dynamics 2013. Smallegan, S.M., Irish, J.L., den Bieman, J.P. & van Dongeren, A.R. (2015) Numerical investigation of developed and undeveloped barrier island response to Hurricane Sandy. Proc. Coastal Structures & Solutions to Coastal Disasters. Smit, P., Stelling, G., Roelvink, J., Van Thiel de Vries, J., McCall, R., Van Dongeren, A., et al. (2010) XBeach: Non-hydrostatic model: Validation, verification and model description. Tech. rep., Delft University of Technology. Smith, J.M., Sherlock, A.R. & Resio, D.T. (2001) STWAVE: Steady-state spectral wave model user’s guide for STWAVE Version 4. Special Report ERDC/CHL-01-01 US Army Engineer Research and Development Center, Vicksburg, MS. Splinter, K.D. & Palmsten, M.L. (2012) Modeling dune response to an East Coast Low. Marine Geology, 329–331, 46–57. DOI: 10.1016/j.margeo.2012.09.005. Splinter, K.D., Carley, J.T., Golshani, A. & Tomlinson, R. (2014) A relationship to describe the cumulative impact of storm clusters on beach erosion. Coastal Engineering, 83, 49–55. DOI: 10.1016/j.coastaleng.2013.10.001. Steetzel, H.J. (1993) Cross-shore Transport during Storm Surges. PhD thesis. Delft University of Technology. Stive, M.J.F. & de Vriend, H.J. (1994) Shear stresses and mean flow in shoaling and breaking waves, Proc. 24th International Conference on Coastal Engineering, Am. Soc. of Civ. Eng., New York, 1994. Sumer, B. & Fredsoe, J. (2002) The mechanics of scour in the marine environment. In: Advanced Series on Ocean Engineering, World Scientific, 17th edition. Svendsen, I.A. (1984) Mass flux and undertow in the surfzone. Coastal Engineering, 8, 347–365. Svendsen, I.A. (2003) Introduction to nearshore hydrodynamics. Advanced Series on Ocean Engineering: Volume 24, World Scientific, Singapore. Symonds, G., Huntley, D.A. & Bowen, A.J. (1982) Two dimensional surf-beat: Long wave generation by a time-varying break point. Journal of Geophysical Research, 87 (C1), 492–498. van Thiel de Vries, J.S.M. (2009) Dune erosion during storm surges. PhD thesis, Delft University of Technology, Delft. Van de Graaff, J. (1977) Dune erosion during a storm surge. Coastal Engineering, 1, 99–134.

CH10 MODELING THE MORPHOLOGICAL IMPACTS

215

Van Dongeren, A., Bolle, A., Vousdoukas, M., Plomaritis, T., Eftimova, P., Williams, J., et al. (2009) MICORE: Dune erosion and overwash model validation with data from nine European field sites, Proc. Coastal Dynamics. Van Dongeren, A.R., Lowe, R., Pomeroy, A., Duong, M.T., Roelvink, J.A., Symonds, G., et al. (2012) Numerical modeling of low-frequency wave dynamics over a fringing coral reef. Coastal Engineering. DOI: 10.1016/j.coastaleng.2012.11.004. Van Dongeren, A.R., Ciavola, P., Viavattene, C., de Kleermaeker, S., Martinez, G., Ferreira, O., et al. (2014) RISC-KIT: Resilience-increasing strategies for coasts – toolkit. In: A.N. Green & J.A.G. Cooper (Eds) Proceedings 13th International Coastal Symposium (Durban, South Africa),Journal of Coastal Research, Special Issue, 66 Van Geer, P., De Vries, B., Van Dongeren, A. & Van Thiel de Vries, J.S.M. (2012) Dune erosion near sea walls: Model-data comparison. Proceedings International Conference on Coastal Engineering, 1–9. Van Gent, M. (1995a) Wave interaction with berm breakwaters. Journal of Waterway, Port, Coastal, and Ocean Engineering, 121 (5), 229–238. Van Gent, M. (1995b) Wave interaction with permeable coastal structures. PhD thesis, Delft University of Technology. Van Gent, M. (1996) Numerical modeling of wave interaction with dynamically stable structures. In: Proceedings of 25th Conference on Coastal Engineering, Orlando, Florida. pp. 1930–1943. Van Gent, M.R.A., van Thiel de Vries, J.S.M., Coeveld, E.M., de Vroeg, J.H. & van de Graaf, J. (2008) Large-scale dune erosion tests to study the influence of wave periods. Coastal Engineering, 55, 1041–1051. Van Rijn, L.C., Wasltra, D.J.R., Grasmeijer, B., Sutherland, J., Pan, D. & Sierra, J.P. (2003) The predictability of cross-shore bed evolution of sandy beaches at the time scale of storms and seasons using process-based profile models. Coastal Engineering, 47 (3), 295–327. DOI: 10.1016/S0378-3839(02)00120-5. Van Rijn, L. (2007a) Unified view of sediment transport by currents and waves. Initiation of motion, bed roughness, and bed-load transport. J. Hydraul. Eng., 133 (6), 649–667. Van Rijn, L. (2007b) Unified view of sediment transport by currents and waves. II: Suspended transport. J. Hydraul. Eng., 133 (6), 668–689. Van Rooijen, A.A., McCall, R.T., Van Thiel de Vries, J.S.M., Van Dongeren, A.R., Reniers, A.J.H.M. & Roelvink, J.A (2016) Modeling the effect of wave-vegetation interaction on wave setup. J. Geophys. Res. Oceans, 121. DOI: 10.1002/2015JC011392. Van Verseveld, H.C.W., van Dongeren, A.R., Plant, N.G., Jaeger, W.S. & den Heijer, C. (2015) Modeling Multi-hazard hurricane damages on an urbanized coast with a Bayesian network approach. Coastal Engineering, 103, 21–35. Vellinga, P. (1986) Beach and dune erosion during storm surges. PhD thesis. Delft University of Technology. Delft Hydraulics Communication No. 372. Vousdoukas, M., Ferreira, Ó., Almeida, L. & Pacheco, A. (2012) Toward reliable storm hazard forecasts: XBeach calibration and its potential application in an operational early-warning system. Ocean Dynamics, 62, 1001–1015. Westerink, J.J., Blain, C.A., Luettich, R.A. & Scheffner, N.W. (1994) Adcirc: An advanced three-dimensional circulation model for shelves, coasts, and estuaries. Technical Report DRP-92-6 US Army Williams, J., de Alegría-Arzaburu, A.R., McCall, R.T., van Dongeren, A. (2012) Modeling gravel barrier profile response to combined waves and tides using XBeach: Laboratory and field results. Coastal Engineering, 63, 62–80, BARDEX: A large-scale laboratory study of gravel barrier dynamics. 10.1016/j.coastaleng.2011.12.010.

216

COASTAL STORMS: PROCESSES AND IMPACTS

WL Delft Hydraulics (1987) Systematic research into the effect of dune base defenses (Dutch: Systematisch onderzoek naar de werking van duinvoetverdedigingen). Technical report, M2051 -part 2. WL Delft Hydraulics (1993) Effect of objects on dune erosion (Dutch: Effect bebouwing duinafslag).Technical report, H1696, Delft. Wu, W., Ozeren, Y., Wren, D., Chen, Q., Zhang, G., Holland, M., et al. (2011) Phase I Report for SERRI Project No. 80037: Investigation of surge and wave reduction by vegetation. Laboratory Publication, 1, 315. Yates, M.L., Guza, R.T. & O’Reilly, W.C. (2009) Equilibrium shoreline response: Observations and modeling. Journal of Geophysical Research, 114, C09014. 10.1029/2009JC005359. Yates, M.L., Guza, R.T., O’Reilly, W.C., Hansen, J.E. & Barnard, P.L. (2011) Equilibrium shoreline response of a high wave energy beach. J. Geophys. Res., 116, C04014. DOI: 10.1029/2010JC006681. Zijlema, M., Stelling, G. & Smit, P. (2011) SWASH: An operational public domain code for simulating wave fields and rapidly varied flows in coastal waters. Coastal Engineering, 58 (10), 992–1012. DOI: 10.1016/j.coastaleng.2011.05.015.

11 Preparing for the Impact of Coastal Storms: A Coastal Manager-oriented Approach José Jiménez1 , Clara Armaroli2 and Eva Bosom3 1 Laboratori

d’Enginyeria Marítima, Universitat Politécnica de Catalunya Barcelona Tech, Barcelona, Spain 2 Department of Physics and Earth Sciences, University of Ferrara, Ferrara, Italy 3 Laboratori d’Enginyeria Marítima, Universitat Politécnica de Catalunya Barcelona Tech, Barcelona, Spain

11.1

Introduction

The impact of coastal storms induces a series of potentially harmful hazards, such as coastal erosion and inundation. At the same time, the continuous development of coastal zones during recent decades has produced an increase in the number of people, activities and economic interests exposed to these and other natural hazards. The combination of these two factors has resulted in a scenario where natural hazard-induced problems in coastal areas are becoming more frequent and more harmful (e.g. Zhang et al., 2000; Jiménez et al., 2012). However, as Kron (2012) states, disasters are the net result of the negative effects of extreme natural events and the positive responses to them (including the prepared responses). In this sense, one of most common (and important) actual needs of coastal managers is to properly know the source of existing potential hazards, to evaluate the expected magnitude of processes inducing damages along the coast and to assess the associated coastal vulnerability and risk. During recent decades, coastal vulnerability has emerged as an important concept in understanding and managing coastal risks, due to the growing concern about their intensity and consequences (e.g. Alcántara-Araya, 2002; Gaddis et al., 2007). Furthermore, during recent years, many authors have highlighted the relevance of developing tools to integrate vulnerability into coastal management frameworks (e.g. McFadden et al., 2007; Meur-Férec et al., 2008). Different conceptual and methodological approaches to characterize coastal storm-induced vulnerability at different spatial scales have been developed in recent decades. Among them, the approach developed by the Unites States Geological Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

218

COASTAL STORMS: PROCESSES AND IMPACTS

Survey (USGS) for estimating the relative magnitudes of coastal change that are likely to occur during hurricanes along the US coast should be highlighted (e.g. Stockdon et al., 2007). This framework is based on the storm impact scale developed by Sallenger (2000) to categorize tropical and extratropical storm impacts on natural barrier islands. This conceptual model has also been used by other authors to derive storm-induced risk indicators for characterizing coastal vulnerability under non-hurricane conditions. Thus, following a similar approach, Mendoza and Jiménez (2008, 2009) developed a methodology based on the use of coastal indicators that separately evaluate flood and erosion vulnerability to storm impacts in the Mediterranean. This framework was applied at a regional scale to evaluate the vulnerability of the Catalonian coast (Spain) to typical storms of the area. Other examples of using this conceptual model for coastal risk assessment to storm impacts can be seen in Jiménez et al. (2009), Almeida et al. (2012) and Armaroli et al. (2012), among others. At a small spatial scale, Villatoro et al. (2014) analyzed inundation and erosion risk for open beaches in different European environments. They presented a risk assessment framework for storm-induced coastal flood and erosion. Recent reviews on existing methodologies for risk assessment for natural hazards in coastal areas can be seen in Cirella et al. (2014) and Rangel-Buitrago and Anfuso (2015). Ferreira et al. (2009), after analyzing different coastal storms risk assessment strategies in Europe, classified them in strategic and operational approaches. The strategic approach is essentially designed for long-term planning, and is usually based on probabilistic hazard descriptions. On the other hand, the operational approach is based on real-time predictions, which are used for emergency plans. Thus, the two approaches can be considered as offline and online for storm-induced risk management, respectively. In offline approaches, the objective is to anticipate potential damages along the coast for a given hazard (or a combination of them), and to decide where to concentrate efforts in preventing, fighting or counteracting their consequences. Overall, this approach’s aim is to optimize the use of resources for risk management. The objective of online approaches is to allow coastal managers to make real-time decisions regarding actions that will minimize the real impact of a given hazard. An example of a general approach is the one presented by Barnard et al. (2009), who describe a modeling framework for forecasting the impact of storms on the Pacific coast developed by the USGS, which can be used online or offline. Considered hazards are: coastal flooding, inundation, erosion and cliff failure. These two approaches are not exclusive, and in fact, they are especially efficient for storm-risk management when they are combined (nested). Thus, an ideal combination should be to first apply a regional scale analysis in which the entire coast is investigated, in order to identify sensitive spots to the impact of storms, and then to develop an online system for the identified spots, where the specific impact of any storm can be analyzed in detail. In essence, the two approaches anticipate the impact of storms in two different but complementary ways. The first approach anticipates the most probable locations for coastal damage caused by the overall maritime climate, and the second one anticipates the impact of a given storm in a specific (sensitive) location. This nested approach has been adopted in the EU-funded research project RISC-KIT (Van Dongeren et al., 2014). Within this context, the main aim of this work is to address the impact of storms on the coast from the coastal manager’s standpoint, that is, how to use all the scientific

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

219

knowledge about coastal storms presented in the different chapters of this book in a practical way. Thus, the aim is to help managers make decisions on the design of DRR measures, and also to allocate resources for such purposes. To this end, this work presents two different and complementary ways to anticipate the impact of coastal storms. The first one consists of a Coastal Vulnerability Assessment Framework to be applied at a regional scale (∼100 km), whereas the second one consists of an Early Warning System to be applied locally (1 km). In both cases, the main aspects to be considered are presented in detail and are illustrated with real cases.

11.2

Coastal vulnerability assessment framework

11.2.1 General framework In the context of this work, coastal vulnerability is defined as the potential of the coast to be harmed by storms. Its assessment enables the quantification of the difference between the impact, which is characterized by the intensity of storm-induced hazards, and the adaptive capacity of the coastal environment to cope with such hazards. Thus, areas with a high vulnerability should be considered as susceptible to significant damage. It should be considered that in this work, when we refer to vulnerability, we do not include socio-economic aspects of the coastal zone; we primarily refer to the geomorphic component. Figure 11.1 shows an example of a methodological framework proposed by Bosom and Jiménez (2011) to assess the coastal vulnerability to storm impacts. It consists of three main blocks: 1. 2. 3.

Characterization of the storm climate (forcing definition) Assessment of the storm-induced hazards (hazard quantification) Assessment of the coastal vulnerability (vulnerability quantification)

Although this kind of analysis can be applied by using different ways of characterizing the storm forcing, for example real wave conditions (Prasad et al., 2009), or storm classes (Mendoza & Jiménez, 2009), a probabilistic approach is highly recommended (e.g. Jiménez et al., 2009; Bosom & Jiménez, 2011) so that the probability of occurrence of the storm-induced hazards along the coast is estimated for the entire storm climate. This approach will allow ‘fair’ comparison of the vulnerability of different coastal stretches in order to identify the most sensitive areas to be prioritized for Disaster Risk Reduction (DRR) measures. In this case, the probability of occurrence is the normalization factor; once this is selected by the decision maker, the corresponding vulnerability at any point along the coast is compared.

11.2.2 How to characterize storm-induced hazards One of the critical points within any vulnerability assessment framework is determining how to properly assess the storm-induced hazards. The first task to be accomplished is the selection of coastal hazards to be considered in the analysis. When an extreme storm impacts on a sandy coast, it produces different morphodynamic responses, which

220

COASTAL STORMS: PROCESSES AND IMPACTS

Waves

Water levels

Storm time series Forcing definition

Beach morphology

Hazard assessment

Erosion

Inundation

Storm-induced hazards time series

Fiting extreme value distribution

Erosion

Inundation

Hazards probability distributions

Vulnerability calculator

Decision-maker's choice Risk level time period of concern Return period selection

Vulnerability to erosion

Vulnerability to inundation Vulnerability assessment

Figure 11.1 Coastal vulnerability assessment framework to storm impacts.

are controlled by both storm characteristics and coastal geomorphology (e.g. Morton, 2002; Morton & Sallenger, 2003). An extensive analysis of these processes for different coastal environments can be seen in various chapters of this book (e.g. Chapters 4, 5, 6 and 7). In general terms, and using an open sedimentary coast as a prototype, the simplest version of the framework should include the two most important storm-induced coastal harmful processes: inundation and erosion. When assessing the magnitude of the hazards associated with the impact of an event of a given probability of occurrence, one of the points introducing uncertainty in the analysis is the assignment of the probability of occurrence. Due to the nature of the problem in which storm-induced hazards depend on more than one single storm variable (e.g. wave height, period, duration), different combinations of wave and water level conditions (events) will result in similar magnitudes of hazard. Due to

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

221

this fact, we strongly recommended using the so-called response approach (see Garrity et al., 2006). In this approach, wave and water level time series are used to calculate the hazard parameters of interest, such as shoreline retreat, runup and overtopping (see Chapters 2, 4 and 9 of this book). A probability distribution of extremes is then fitted to the obtained hazard dataset. From here, the hazard parameter of interest (associated with the selected given probability) will be directly calculated from its probability distribution. This method is especially recommended when storm wave and water level variables (e.g. Hs , Tp and duration) are partially or poorly correlated. This is the approach recommended by the Federal Emergency Management Agency (FEMA) guidelines for flood studies (Divoky & McDougal, 2006). Its use in storm-induced erosion can be seen in Callaghan et al. (2008, 2013) and Corbella and Stretch (2012), whereas Jiménez et al. (2009) and Bosom and Jiménez (2010, 2011) fully applied it to characterize storm-induced erosion and inundation within a vulnerability assessment framework.

11.2.3 How to measure the vulnerability Once the storm-induced hazards are known, the remaining part of the assessment is the determination of the coastal vulnerability, which takes into account the ability of the coast to cope with the induced impacts. To this end, the response capacity of the beach is characterized by considering the beach width, that is, the wider the beach is, the smaller the probability will be of full erosion. In the case of flooding, a good proxy for such a purpose will be the beach and/or dune height: the higher the beach, the smaller the inundation. This balance between impact and capacity of response can be formulated in terms of simple intermediate erosion and inundation variables EV and IV, which are given by: EV = Δx∕BW IV = 𝜉∕B max

(11.1) (11.2)

where Δx is the storm-induced shoreline retreat, BW is the beach width, 𝜉 is the storm-induced total water level, and Bmax is the beach/dune height. The resulting vulnerability to each hazard is formulated as a function of these intermediate variables by defining the optimum and failure states of the beach for the considered hazard. These states define the limits of a vulnerability scale corresponding to zero and maximum vulnerability, respectively. For the erosion hazard, the optimum state (zero vulnerability) is given by a beach width significantly larger than the expected storm-induced erosion associated with a given probability of occurrence. At the other extreme, the failure state (maximum vulnerability) corresponds to a beach narrower than the storm-induced erosion, resulting in an exposed hinterland (e.g. Jiménez et al. 2011). For the inundation hazard, the optimum state corresponds to a beach significantly higher than the total water level reached during the storm, whereas the failure state is associated with a beach/dune lower than the target water level, so significant overtopping will occur during the event. Figure 11.2 shows an illustration of potential functional relationships linking vulnerability values to these intermediate variables. Obtained vulnerability values are scaled in

222

COASTAL STORMS: PROCESSES AND IMPACTS

Vulnerability

1.0

Very high

0.8

High 0.6

Medium 0.4

Low 0.2

Very low 0.0

0

Optimum state

Failure state

ξ / Bmax Δx / BW

Figure 11.2 Functional relationship to compute vulnerability values.

a range from a minimum value of 0 (safe beach with an optimum state) to a maximum value of 1 (extremely vulnerable beach in a failure state).

11.2.4 How to select the probability to be analyzed The main objective of this framework is to assess the magnitude of coastal vulnerability to the impact of an extreme event (the storm). This will allow mapping of local vulnerabilities along the coast, associated with given probabilities specified by stakeholders according to the target safety level. Within this context, it is of interest to discuss how to define extreme events and how to select relevant probabilities of occurrence for the analysis. In a simple way, extremes can be defined and/or quantified based on the characteristics of the events (e.g. Beniston & Stephenson, 2004): • How rarely they occur, which involves notions of frequency of occurrence. • How intensely they occur, which involves notions of threshold of exceedance. • The impacts they exert (e.g. in social, economic and/or environmental terms). In the context of this work, an extreme event has the potential to cause morphological and/or socio-economic consequences along the coast. As one might expect, different sites experience different exposure to storms, and thus are likely to be characterized by different return periods (Tr). The definition of the extreme event is, therefore, site dependent. With respect to the selection of probabilities to be considered in the analysis, one possibility is, in spite of the site dependency, to analyze given common probabilities of exceedance. This is the approach adopted in the EU Floods Directive (EC, 2007), which specifies that flood hazard maps and flood risk maps will identify areas with a medium likelihood of flooding (at least a 1 in 100-year event), as well as extremes or low-likelihood events. Another approach is to select Tr based on the concept of the lifetime or design life of a coastal structure. In this case, we consider the beach as a coastal protection measure

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

223

protecting the hinterland against the impact of a storm. In this context, the lifetime is the period over which the beach is expected to continue providing protection against the ‘design’ condition, which in this case corresponds to the target storm (e.g. Reeve, 2010). Thus, we can make use of the relationship predicting the probability of exceedance, P, the lifetime, L, and the return period: )L ( 1 P=1− 1− Tr

(11.3)

To select appropriate Tr values, we can fix L as the desired minimum lifetime of the beach and P as the accepted probability of occurrence of the event within the said lifetime as a function of the importance of the site. For high-interest areas, where the exceedance of protection capacity provided by the beach against the storm would have significant consequences, relatively long lifetimes and low probabilities should be adopted. From the practical standpoint, the selection of the lifetime and the accepted probability of exceedance determines the return periods to be analyzed. The first one, the lifetime, makes reference to the expected time horizon of the analysis: For how long do we assume that the coast will provide the current protection level? Before giving a typical conservative answer, for example a very long time period, we have to consider that sedimentary coasts are usually subjected to littoral processes, which affect their stability, and consequently, the current beach configuration (and the corresponding level of provided protection) will not necessarily be steady. Making an analogy between the beach and coastal protection measures, Table 11.1 shows expected lifetimes for such works as a function of the importance of expected consequences of their failure (Puertos del Estado, 2001). The second variable, the probability of exceedance, is also dependent on the importance of the implications of the hazard. Table 11.2 shows recommended values of maximum allowable probabilities of failure for coastal protection works as a function of the expected consequences.

11.2.5 The Catalonia coastal vulnerability assessment framework The general methodology presented above has been adopted to identify sensitive coastal spots to storm-induced hazards along the Catalonian coast (NE Spanish Table 11.1 Recommended minimum lifetime for coastal protection works (Puertos del Estado, 2001). Type of work

Importance

Defence against big floods* Margins protection and defence Beach nourishment and protection

High Medium Low

Min. lifetime (yrs) 50 25 15

∗ It refers to defence works that in the case of failure may cause an important inundation of the hinterland.

224

COASTAL STORMS: PROCESSES AND IMPACTS

Table 11.2 Recommended maximum values of failure probability for coastal protection works as a function of their importance (Puertos del Estado, 2001). Importance Very high High Medium Low

Maximum probability 0.0001 0.01 0.10 0.20

Mediterranean coast, Figure 11.3). This area can be considered as a good example of most of the Mediterranean coastlines, with a high geodiversity and subjected to significant human pressure. Its 600-km-long coastline is comprised of various environments, such as cliffs, sandy beaches, coastal plains and river deltas. Coastal administrative units (comarcas) concentrate about 62% of the population and 65% of the Gross Domestic Product (GDP) in just 22.8% of the territory. A detailed description of the Catalonian coast in socio-economic and environmental terms can be seen in Brenner et al. (2006). Erosion is the dominant process along the Catalonian sedimentary coast, with about 70% of the beaches being eroded during recent decades at an average rate of about 0.7 m/year (CIIRC, 2010). Direct implications have been a decrease in beach recreational carrying capacity (Valdemoro & Jiménez, 2006) and an increase in the exposure of the hinterland against the impact of storms (Jiménez et al., 2011). This behaviour, combined with the increasing urbanization of the Catalonian coastal zone, serves to explain the observed significant increase in coastal damage during the past 50 years, in spite of the absence of any increasing trend in storm-induced hazards (Jiménez et al., 2012). Bosom and Jiménez (2011) proposed a framework to assess coastal vulnerability to storm impacts with a probabilistic approach, using individual beaches as the basic assessment unit. The magnitude of storm-induced hazards, such as erosion and inundation, are calculated following the response approach, and their extreme distribution at every location/beach along the coast is thus computed (see Bosom & Jiménez (2011) for details). The erosion hazard was calculated as the maximum storm-induced shoreline retreat, using an adapted version of the parametric erosion model of Mendoza and Jiménez (2006). The inundation hazard was parameterized as a function of the wave-induced runup, which was calculated by using the Stockdon et al. (2006) model. The selection of runup as the main flood hazard indicator was due to the fact that the contribution of storm surges to the total water level during storms can be considered as secondary when compared to the runup magnitude (Mendoza & Jiménez, 2008, 2009). Both hazards were calculated for all storms occurring in the area during a period of about 50 years, by using existing hindcast data (Reguero et al., 2012). The local extreme probability distributions were calculated by fitting obtained values by means of a Generalized Extreme Value (GEV). Finally, the coastal vulnerability for each hazard is assessed by comparing its magnitude with a variable, characterizing the beach capacity of response following relationships (1) and (2). After obtaining the probability distribution of storm-induced

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

225

Mediterranean Sea

Figure 11.3 The Catalonian coast.

hazards, it is possible to assess the vulnerability associated with any probability of occurrence. Thus, for relatively infrequent hazards with a return period of 50 years (which can be considered representative for exposures of low and medium importance), the percentage of the coast classified as highly and very highly vulnerable is nearly the same for storm-induced erosion and for inundation (about 28% and 31%, respectively). However, this does not necessarily imply that coastal areas are equally vulnerable to both hazards. In fact, the methodology used permitted the detection of a high spatial variability in vulnerability as a consequence of the combination of wave climate conditions and geomorphological heterogeneity. Thus, for the above-mentioned 50-year return period, the northern part of the Catalonian coastline was identified as the most vulnerable to erosion, mostly due to the combination of the intense storm climate of this area and the existence of relatively narrow beaches. On the other hand, the highest vulnerability to inundation is observed in the northern and central areas of the Catalonian coast, as a consequence of the moderate to intense storm climate, combined with characteristically steep beach slopes. The small-scale spatial variation in coastal vulnerability is illustrated in Figure 11.4 for three comarcas for different return periods (10, 50 and 100 years). There is a direct relationship between observed vulnerability and return periods, in such a way that the

226

COASTAL STORMS: PROCESSES AND IMPACTS

Tr 10y Tr 50y Tr 100y

Erosion vulnerability Very low Low Medium High Very high

Tr 10y Tr 50y Tr 100y

Inundation vulnerability Very low Low Medium High Very high

Figure 11.4 Vulnerability of the Catalonian coast to storm-induced erosion (top) and inundation (bottom) associated with different return periods.

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

227

more frequent hazards (low return periods) lead to lower vulnerability than the less frequent hazards (higher return periods). This is the expected situation for relatively narrow coastal stretches, where the magnitude of storm-induced hazards exceeds the coastal resilient capacity. In the case of erosion, the percentage of highly and very highly-vulnerable coastlines increases from 12% for a return period of 10 years to 26% for the highest return period (100 years). These values are similar in the case of inundation, ranging from 12% to 24% of the coastline for the 10 and 100-year return periods, respectively. Looking at the spatial distribution of the calculated vulnerability indexes, existing hotspots can be easily identified as those stretches whose values are significantly higher than neighbouring areas. Figure 11.4 shows the presence of hotspots for both hazards, whereas some areas behave as hotspots for just a single storm-induced hazard. Such an analyisis will permit managers to make decisions on risk-reduction measures, taking into account the specific expected damage. These results have been obtained for current maritime climate and geomorphological conditions of the study area (2010). However, the methodological framework permits an easy update of such characteristics when necessary in order to consider coastal evolution and/or potential changes in the wave climate due to the effects of climate change.

11.3

Coastal early warning systems

11.3.1 Generalities The common saying ‘forewarned is forearmed’ clearly explains the importance of warning systems in the field of prevention of coastal disasters. This is formally expressed in the Hyogo Framework for Action 2005–2015 (UNISDR, 2007) as the second priority for action, which includes the identification, assessment and monitoring of risks and the enhancement of Early Warning Systems (EWSs). The document states that the priority is: ‘to develop early warning systems that are people-centred, in particular systems whose warnings are timely and understandable to those at risk’. Therefore, among DRR measures, EWSs play a vital role (Lavell et al., 2012). A people-centred warning system should include: 1. 2. 3. 4.

Risk knowledge (systematically collect data and undertake risk assessments) Monitoring and warning service (the development of hazard monitoring and early-warning service) Dissemination and communication (communicate risk information and early warning) Response capability (the development of a national and community response capability)

The general aim of an EWS is to provide timely information on hazards, so that the exposed individuals can undertake the necessary actions to save their lives and to minimize the impact (Basher, 2006). EWS information has to be sufficiently clear and accurate in order to be translated into effective actions.

228

COASTAL STORMS: PROCESSES AND IMPACTS

11.3.2 Coastal EWSs Until now, most of coastal EWSs have focused mainly on ‘hydrodynamic’ hazards. Some well-known examples of these systems around the world mainly predict the surge component, such as the acqua alta surge forecast system for the Venice lagoon (Ferrarin et al., 2013; Mariani et al., 2015), the UK joint Met Office/EA Flood Forecasting Centre (Stephens & Cloke, 2014), the storm surge in the US National Hurricane Centre forecast system (Morrow et al., 2015) and the Bangladesh storm surge EWS (Dube et al., 2009). Different authors have highlighted the capability of these early warning systems to significantly reduce impacts and loss of life (e.g. Paul, 2009; Spencer et al., 2014; Stephens & Cloke, 2014). Coastal storm hazard forecasts and their related warnings, seldom include the morphological component that is the beach and dune response to storm waves and currents (Vousdoukas et al., 2012; Harley et al., 2016). Along sandy shores, additional protection from flooding of landward areas is represented by the interplay between beach characteristics (slope, elevation and width) and by the amount of sediment and the elevation of dune systems (Armaroli et al., 2012). The impact of a storm is largely influenced by beach and dune behaviour and how both morphologies are able to absorb the wave energy. Therefore, a complete hazard assessment should take into account the forcing components, along with the morphological response. A first attempt to implement operational warning systems for coastal areas, including the beach response during storms, was done within the FP7 MICORE project, where nine EWSs prototypes were created for different European coastal stretches (Ciavola et al., 2011). The EWSs were designed to provide reliable predictions of the morphological impact of coastal storms in support of civil protection mitigation strategies. The systems were designed to include a set of useful information for decision makers, which consists of five modules (Haerens et al., 2012): 1. 2. 3. 4. 5.

An observation module that includes the forcing components, derived from meteorological and oceanographic models A forecast module that simulates beach response to waves and currents A decision-support module, which includes tools and results that help with decision making A warning module based on specific thresholds A visualization module to display warnings and other information for decision makers

11.3.3 The Emilia-Romagna coastal early warning system The Emilia-Romagna region is located in the north Adriatic Italian coast (Figure 11.5) and is composed of sandy beaches stretching over 130 km. The area is characterized by low elevations above MSL and by a high level of human occupation. The coastal zone experienced an intensive exploitation after World War II, which had its peak around the 1970s, when urban settlements grew and tourism became one of the most important economic activities of the region. This region is a micro-tidal environment, with a maximum spring tidal range of about 90 cm (Armaroli et al., 2012). The wave climate of the coast is of low energy (91% of

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

229

Figure 11.5 The Emilia-Romagna coast.

the significant wave height (Hs) is below 1.25 m; Ciavola et al., 2007). The main storm directions are from E-NE (bora wind) and SE (scirocco wind). The one-in-one year return period storm is characterized by an Hs of 3.3 m and by a wave peak period of 7.7 s. Surge levels are an important element controlling total water levels measured during storms, with the ten-year return period surge being about 1 m (Masina & Ciavola, 2011).

230

COASTAL STORMS: PROCESSES AND IMPACTS

Protected beaches currently represent nearly 60% of the coastline (protected by groins, breakwaters, submerged barriers, artificial embankments, dikes and rubble mound slopes). The Emilia-Romagna coastline is largely exposed to storms, and in fact, nearly 80% of the regional beaches are occupied by permanent concessions that are directly exposed to wave action and to high surge levels during storms. This is a highly eroding coast, where erosion problems are caused by a chronic sediment shortage due to the decrease in riverine sediment supplies, and by the presence of numerous groins and jetties, which interrupt the alongshore sediment transport. In addition, the low-lying nature of the coast renders the area very vulnerable to high water levels, which can induce extensive inundation of inland villages. Armaroli et al. (2012) and Perini et al. (2016) found that the coast is particularly vulnerable, even to storms with low return periods (i.e. ten years). Therefore, the improvement of the existing prevention, mitigation and preparedness measures are key issues for regional managers. To this end, an EWS was designed and implemented in the Emilia-Romagna region, with the end goal of supporting civil protection mitigation strategies for coastal storm impacts. The system adopts the structure of five modules developed within the MICORE project (Figure 11.6), which are described as follows. The observation module comprises a series of nested numerical models, which include the meteo-marine forecasting system for Emilia-Romagna (Russo et al., 2013), characterizing the forcing during the storm and its propagation to the coastline. It includes (Harley et al., 2016): 1.

Atmospheric model COSMO-I7 (7 km resolution, www.cosmo-model.org), which provides the atmospheric forcing. 2. SWAN model (grid resolution from 25 km for the Mediterranean Sea, 8 km for the Italian region and, finally, to 800 m for the Emilia-Romagna coastline (Valentini et al., 2007), which generates and propagates waves. 3. Ocean model ROMS (Haidvogel et al., 2008) of the entire Adriatic Sea, with a regular grid resolution of 2 km, which provides the water level response of the atmospheric forcing. The system provides three-day forecasts of waves and water levels. A complete description of the meteo-marine forecasting system for Emilia-Romagna can be seen in Russo et al. (2013). The forecast module is designed by using the morphodynamic numerical model XBeach (Roelvink et al., 2009), running in 1-d mode along selected coastal profiles. This module is fed by outputs of the observation module (waves and water level) at specific grid nodes located close to each selected site. At present, the system is applied in 22 profiles along the Emilia-Romagna coastline, of which 11 are situated in the Lido di Classe area, where the system was first tested. XBeach was extensively tested and calibrated by using field data gathered in periodic surveys done in the Lido di Classe beach (see detailed information on the model settings in Harley et al., 2016). One of the points to be considered is that this module needs to have a realistic representation of the actual beach morphology, which is usually a limiting condition. Thus, beach profiles are composed by combining topographic data extracted from a Lidar survey in 2010 and bathymetry data derived from a Lidar survey in 2006. The exception is Lido di Classe, where the 11 profiles are regularly surveyed (nearly every two months and after major storms).

Observation module

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

231

Measures of waves, winds and tide Measures of beach profiles

Forecast module

Weather forecast model

Surge forecast model

Wave forecast model

Morphological forecast Results: Hs, water level, currents, beach profile response

Visualisation module

Warning module

Decision support module

Translation module

Prevent loss of life due to hazardous maritime conditions (SCW) Storm Impact Indicators

Prevent marine flooding of buildings due to high water level and/or run-up conditions (BWD)

Storm Impact forecast

Code: green, orange, red Issue alerts, ...

Emergency actions

Hazard maps GIS maps

XBeach plots

SIIs

Weather forecast

Wave forecast

Surge forecast

Figure 11.6 Scheme of the modules included into the Emilia-Romagna EWS described in section 11.3.3.

The decision support module includes a set of Storm Impact Indicators (SIIs), which are used to transform the model-chain outputs into clear information useable for decision making in emergency responses (Ciavola et al., 2011). Due to the characteristics of the Emilia-Romagna coast, two indicators were designed. The first is called the Safe Corridor Width (SCW), which characterizes the available subaerial beach width between the dune foot and the shoreline. Simply put, the SCW represents the escaping corridor available for beach users in the case of a storm impact. The second indicator used is the Building Waterline Distance (BWD), which characterizes the available subaerial beach width, measured between the seaward limit of structures located on the rear part of the beach and the shoreline. This indicator was defined as a proxy

232

COASTAL STORMS: PROCESSES AND IMPACTS

of inundation/damage to buildings during the storm. The shoreline position, which is included in both indicators, evolves with time during the storm, and is derived from XBeach outputs. The warning module sets alerts on the basis of pre-defined thresholds, defined through a benchmarking procedure. The thresholds are related to SII values, and they include a no-alert condition (‘code green’) when both the SCW and BWD are greater than 10 m; a medium hazard (‘code orange’) when the SCW is greater than 5 m and the BWD is greater than 0 m; and a ‘code red’ (highest hazard) when the SCW is less than 5 m and the BWD is less than 0 m (i.e. the inundation of buildings is predicted). Finally, the visualization module is the web interface where the outputs of the model chain are displayed in terms of ‘codes’ (Figure 11.7). The three levels of alert are represented with coloured pins, which are geo-localized on a map of the coastal area in the profile lines used to run the morphological model. The web-interface displays a series of useful information, such as: 1. 2. 3. 4. 5. 6. 7.

SII names, along with their definition and a table that includes the target objectives and methods for each SII Sites where the model train is run Various base maps (e.g. aerial photographs, Lidar DEMs and land use maps) Hazard maps on the basis of the works of Ciavola et al. (2008) and Armaroli et al. (2012) SWAN and ROMS forecasts Measured tides and waves Help menu

The Emilia-Romagna Coastal EWS was first tested in Lido di Dante and Classe, and then extended to another seven sites along the coast (Harley et al., 2016), since it was recognized as a useful tool for Civil Protection purposes. The EWS outputs are evaluated, together with meteorological and sea state forecasts, within the Functional Centre. The ER Functional Centre is composed of personnel from three different regional agencies – ARPA-SIMC, Geological Service and Civil Protection. ARPA-SIMC provides hydro-meteorological and sea state (wave and water level) forecasts. When a forecast is generated, an alert is issued according to several criteria such as: 1. Expected water level 2. Foreseen duration of the storm 3. The state of the coastal area (if it was already affected by previous storms) 4. Outputs of the Early Warning System. Once an alert is issued, all information is sent to all regional offices involved in coastal risk management. The performance of the system was tested against a storm that impacted the coast on 31 October 2012, which caused extensive damage along the regional coastline (Harley et al., 2016). The analysis was based on the skill assessment of the entire operational chain. The test identified that the system would have failed to predict the storm impact due to the under-prediction of surge by the ROMS model, as well as the over-prediction of the morphodynamic impact by the XBeach model. In other words, although EWSs

CH11

Figure 11.7

PREPARING FOR THE IMPACT OF COASTAL STORMS

Screenshot of the Emilia-Romagna Coastal EWS web page.

233

234

COASTAL STORMS: PROCESSES AND IMPACTS

are powerful tools for civil protection purposes, they need to be carefully calibrated over a number of events. Moreover, each component of the chain should be evaluated, both separately and as a whole in order to strengthen the reliability of the overall predictions. As an example of the required continuous calibration, the regional Geologic, Seismic and Soil Service in Italy performed a qualitative evaluation of the regional EWS after the impact of one of the strongest storms ever recorded in the area, which took place on 5 February 2015. The whole coastal area was affected by erosion and/or inundation, and several families were evacuated from their homes (Perini et al., 2015). The performance of the system was quite effective, such that an EWS alert was issued for six out of eight studied sites (Figure 11.7). This improved system performance has permitted regional end-users to consider the EWS as a key component of the early intervention chain, as it undergoes continuous improvements of model predictions.

11.4 Conclusion The two-step approach presented here to address the impact of storms in coastal areas permits coastal managers to anticipate potential damages and to optimize resource allocations for DRR. It combines a large spatial scale- and probability-based analysis with small, spatial and real-time forecasting. The first step in this approach consists of identifying the impact of storms to sensitive stretches along the coast in a robust manner, that is, using a common base of comparison. In order to achieve this, we propose using the probability of occurrence of a hazard of a given intensity as the basis of the analysis. The probability will be selected by the manager, depending on the characteristics of the coast to be managed, as well as the target safety level. The framework presented here can be used to deal with any storm-induced hazard. It has been illustrated here with the two most important ones in sedimentary coasts: erosion and inundation. Separate assessments are needed, because each hazard has a different dependence on storm characteristics, and also because each hazard induces distinct damages. The main result from this phase of the assessment is the identification of stretches along the coast that are sensitive to the impact of storms for a given probability of occurrence. This will permit coastal managers to identify those locations where a priori larger storm-induced damages are anticipated (at least from a geomorphic viewpoint), and to make more robust decisions regarding resource allocations for managing storm risks. This information will also allow for making a preliminary selection of those areas where DRR measures should be considered. The second step in the approach consists of providing timely information on hazards and expected damages at a given location and for a given event. In this sense, it is a DRR in and of itself. This is going to be achieved by the development of EWSs for previously-identified sensitive stretches, with the aim of helping affected individuals to undertake actions that will minimize the impact. An EWS is a chain of processes, combining observation, forecasting, decision making, warning and visualization modules. As in any chain, the robustness of the system is controlled by the weakest link, and as a consequence, this axiom implies that all system components must be carefully calibrated, both as individual elements and as parts of

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

235

the whole system. The principal result from the application of this phase of the EWS is the creation of forecasting information that will aid in the making of real-time decisions regarding the management of storm-induced risks. Both approaches have been implemented in different types of coastal environments, demonstrating their utility and versatility, provided they are suited for relevant hazards. In any case, decision makers need to be aware that the proper management of storm-induced risks along a coast requires real data for these systems, and consequently, this data must be combined with a proper data acquisition/monitoring network.

Acknowledgements This work has been done in the framework of the RISC-KIT (Grant No. 603458) research project funded by the European Union. The work of JJ and EB was also funded in the framework of the PaiRisClima (CGL2014-55387-R) research project funded by the Spanish Ministry of Economy and Competitiveness.

References Alcántara-Araya, I. (2002) Geomorphology, natural hazards, vulnerability and prevention of natural disasters in developing countries. Geomorphology, 47, 107–124. Almeida, L.P., Vousdoukas, M.V., Ferreira, Ó., Rodrigues, B.A. & Matias, A. (2012) Thresholds for storm impacts on an exposed sandy coastal area in southern Portugal. Geomorphology, 143–144, 3–12. Armaroli, C., Ciavola, P., Perini, L., Calabrese, L., Lorito, S., Valentini, A., et al. (2012) Critical storm thresholds for significant morphological changes and damage along the Emilia-Romagna coastline, Italy. Geomorphology, 143–144, 34–51. Barnard, P.L., O’Reilly, B., van Ormondt, M., Elias, E., Ruggiero, P., Erikson, L.H., et al. (2009) The framework of a coastal hazards model: A tool for predicting the impact of severe storms. US Geological Survey Open-File Report 2009–1073,19 p. Basher, R. (2006) Global early warning systems for natural hazards: Systematic and people-centred. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 364, 2167–2182. Beniston, M. & Stephenson, D.B. (2004) Extreme climatic events and their evolution under changing climatic conditions. Global and Planetary Change, 44, 1–9. Bosom, E. (2014) Coastal vulnerability to storms at different time scales. PhD thesis, Universitat Politècnica de Catalunya⋅BarcelonaTech, Barcelona. Bosom, E. & Jiménez, J.A. (2010) Storm-induced coastal hazard assessment at regional scale: Application to Catalonia (NW Mediterranean). Advances in Geosciences, 26, 83–87. Bosom, E. & Jiménez, J.A. (2011) Probabilistic coastal vulnerability assessment to storms at regional scale – application to Catalan beaches (NW Mediterranean). Natural Hazards and Earth System Sciences, 11, 475–484. Brenner, J., Jiménez, J.A. & Sardá, R. (2006) Definition of homogeneous environmental management units for the Catalan coast. Environmental Management, 38, 993–1005. Callaghan, D.P., Nielsen, P., Short, A. & Ranasinghe, R. (2008) Statistical simulation of wave climate and extreme beach erosion. Coastal Engineering, 55, 375–390.

236

COASTAL STORMS: PROCESSES AND IMPACTS

Callaghan, D.P., Ranasinghe, R. & Roelvink, D. (2013) Probabilistic estimation of storm erosion using analytical, semi-empirical, and process based storm erosion models. Coastal Engineering, 82, 64–75. Ciavola, P., Armaroli, C., Chiggiato, J., Valentini, A., Deserti, M., Perini, L., et al. (2007) Impact of storms along the coastline of Emilia-Romagna: The morphological signature on the Ravenna coastline (Italy). Journal of Coastal Research, SI 50, 540–544. Ciavola, P., Armaroli, C., Perini, L. & Luciani, P. (2008) Evaluation of maximum storm wave run-up and surges along the Emilia-Romagna coastline (NE Italy): A step towards a risk zonation in support of local CZM strategies. Integrated Coastal Zone Management – The Global Challenge, 505–516. Research Publishing Services, Singapore. Ciavola, P., Ferreira, O., Haerens, P., Van Koningsveld, M. & Armaroli, C. (2011) Storm impacts along European coastlines. Part 2: Lessons learned from the MICORE project. Environmental Science and Policy, 14, 924–933. CIIRC (2010) Estat de la zona costanera a Catalunya. Resum Executiu. Generalitat de Catalunya, Barcelona. Cirella, G.T., Semenzin, E., Critto, A. & Marcomini, A. (2014) Natural hazard risk assessment and management methodologies review: Europe. In: I. Linkof (Ed.) Sustainable Cities and Military Installations. Springer, pp. 329-358. Corbella, S. & Stretch, D.D. (2012) Predicting coastal erosion trends using non-stationary statistics and process-based models. Coastal Engineering, 70, 40–49. Divoky, D. & McDougal, W.G. (2006) Response-based coastal flood analysis. Proceedings of the 30th International Conference. on Coastal Engineering, ASCE, 5291–5301. Dube, S., Jain, I., Rao, A. & Murty, T. (2009) Storm surge modelling for the Bay of Bengal and Arabian Sea. Natural Hazards, 51, 3–27. EC (2007) Directive 2007/60/EC of the European Parliament and of the Council of 23 October 2007 on the assessment and management of flood risks. Official Journal L., 288, 06/11/2007, 27–34. Ferrarin, C., Roland, A., Bajo, M., Umgiesser, G., Cucco, A., Davolio, S., et al. (2013) Tide-surge-wave modelling and forecasting in the Mediterranean Sea with focus on the Italian coast. Ocean Modelling, 61, 38–48. Ferreira, O., Ciavola, P., Armaroli, C., Balouin, Y., Benavente, J., Del Rio, L. et al. (2009) Coastal storm risk assessment in Europe: Examples from 9 study sites. Journal of Coastal Research, 1632–1636. Gaddis, E. B., Miles, B., Morse, S. & Lewis, D. (2007) Full-cost accounting of coastal disasters in the United States: Implications for planning and preparedness. Ecological Economics, 63, 307–318. Garrity, N.J., Battalio, R., Hawkes, P.J. & Roupe, D. (2006) Evaluation of the event and response approaches to estimate the 100-year coastal flood for Pacific coast sheltered waters. Proceedings of the 30th International Conference on Coastal Engineering, ASCE, 1651–1663. Haerens, P., Ciavola, P., Ferreira, Ó., van Dongeren, A., Van Koningsveld, M. & Bolle, A. (2012) Online operational early warning system prototypes to forecast coastal storm impacts (CEWS). Proceedings of the 33rd International Conference on Coastal Engineering, ASCE, management, 45. Haidvogel, D.B., Arango, H., Budgell, W.P., Cornuelle, B.D., Curchitser, E., Di Lorenzo, E., et al. (2008) Ocean forecasting in terrain-following coordinates: Formulation and skill assessment of the Regional Ocean Modeling System. Journal of Computational Physics, 227, 3595–3624.

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

237

Harley, M., Valentini, A., Armaroli, C., Perini, L., Calabrese, L. & Ciavola, P. (2016) Can an early-warning system help minimize the impacts of coastal storms? A case study of the 2012 Halloween storm, northern Italy. Natural Hazards and Earth System Sciences, 16, 209–222, doi:10.5194/nhess-16-209-2016. Jiménez, J.A., Ciavola, P., Balouin, Y., Armaroli, C., Bosom, E. & Gervais, M. (2009) Geomorphic coastal vulnerability to storms in microtidal fetch-limited environments: Application to NW Mediterranean & N Adriatic Seas. Journal of Coastal Research, SI 56, 1641–1645. Jiménez, J.A., Gracia, V., Valdemoro, H.I., Mendoza, E.T. & Sánchez-Arcilla, A. (2011) Managing erosion-induced problems in NW Mediterranean urban beaches. Ocean & Coastal Management, 54, 907–918. Jiménez, J.A., Sancho, A., Bosom, E., Valdemoro, H.I. & Guillén, J. (2012) Storm-induced damages along the Catalan coast (NW Mediterranean) during the period 1958–2008. Geomorphology, 143–144, 24–33. Kron, W. (2012) Coasts: The high-risk areas of the world. Natural Hazards, 66, 1363–1382. Lavell, A., Oppenheimer, C., Hess, J., Lempert, R., Li, J., Muir-Wood, R., et al. (2012) Climate change: New dimensions in disaster risk, exposure, vulnerability and resilience. In: C. Field, V. Barros, T. Stocker, D. Qin, D. Dokken, K. Ebi, et al. (Eds) Managing the risks of extreme events and disasters to advance climate change adaptation. A special report of Working Groups I and II of the Intergovernmental Panel on Climate Change (IPCC), 25–64, Cambridge University Press, Cambridge. Mariani, S., Casaioli, M., Coraci, E. & Malguzzi, P. (2015) A new BOLAM-MOLOCH suite for the SIMM forecasting system: Assessment over two HyMeX intense observation periods. Natural Hazards and Earth System Sciences, 15, 1–24. Masina, M. & Ciavola, P. (2011) Analisi dei livelli marini estremi e delle acque alte lungo il litorale ravennate. Studi Costieri, 18, 87–101. McFadden, L., Nicholls, R.J. & Penning-Rowsell, E. (2007) Managing coastal vulnerability. Elsevier. Mendoza, E.T. & Jiménez, J.A. (2006) Storm-induced beach erosion potential on the Catalonian coast. Journal of Coastal Research, SI 48, 81–88. Mendoza, E. T. & Jiménez, J.A. (2008) Clasificación de tormentas costeras para el litoral Catalán (Mediterráneo NO) Ingeniería Hidráulica en México, 23, 2, 21–32. Mendoza, E. & Jiménez, J.A. (2009) Regional geomorphic vulnerability analysis of Catalan beaches to storms. Proceedings of the Institution of Civil Engineers: Maritime Engineering, 162, 127–135. Meur-Férec, C., Deboudt, P. & Morel, V. (2008) Coastal risks in France: An integrated method for evaluating vulnerability. Journal of Coastal Research, 24, 178–189. Morrow, B.H., Lazo, J.K., Rhome, J. & Feyen, J. (2015) Improving storm surge risk communication: Stakeholder perspectives. Bulletin of the American Meteorological Society, 96, 35–48. Morton, R.A. (2002) Factors controlling storm impacts on coastal barriers and beaches: A preliminary basis for near real-time forecasting. Journal of Coastal Research, 18, 486–501. Morton, R.A. & Sallenger, A.H. (2003) Morphological impacts of extreme storms on sandy beaches and barriers. Journal of Coastal Research, 19, 560–573. Paul, B.K. (2009) Why relatively fewer people died? The case of Bangladeshs Cyclone Sidr. Natural Hazards, 50, 289–304. Perini, L., Calabrese, L., Lorito, S. & Luciani, P. (2015) Costal flood risk in Emilia-Romagna (Italy): The sea storm of February 2015. Proc. Coastal and Maritime Mediterranean Conference, Theme 3: Risk management in the Mediterranean, 3, 25–27 November 2015, Ferrara, Italy,

238

COASTAL STORMS: PROCESSES AND IMPACTS

44, 225–230, doi:10.5150/cmcm.2015.044, http://www.paralia.fr/editions/index.php/CM2/issue/ view/26. Perini, L., Calabrese, L., Salerno, G., Ciavola, P. & Armaroli, C. (2016) Evaluation of coastal vulnerability to flooding: Comparison of two different methodologies adopted by the Emilia-Romagna region (Italy). Natural Hazards and Earth System Sciences, 16, 181–194, doi:10.5194/nhess-16-181-2016. Prasad, K.V.S.R., Arun Kumar, S.V.V., Venkata Ramu, Ch. & Sreenivas, P. (2009) Significance of nearshore wave parameters in identifying vulnerable zones during storm and normal conditions along Visakhapatnam coast, India. Natural Hazards, 49, 347–360. Puertos del Estado (2001) ROM 0.0. General Procedure and Requirements in the Design of Harbor and Maritime Structures. Spanish Ministry of Public Works, Madrid. Rangel-Buitrago, N. & Anfuso, G. (2015) Review of the existing risk assessment methods. In: N. Rangel-Buitrago & G. Anfuso, Giorgio (Eds) Risk Assessment of Storms in Coastal Zones: Case Studies from Cartagena (Colombia) and Cadiz (Spain), Springer International Publishing, pp. 7–13. Reeve, D. (2010) Risk and Reliability: Coastal and Hydraulic Engineering. Spon Press, London. Reguero, B.G., Menéndez, M., Méndez, F.J., Mínguez, R. & Losada, I.J. (2012) A Global Ocean Wave (GOW) calibrated reanalysis from 1948 onwards. Coastal Engineering, 65, 38–55. Roelvink, D., Reniers, A., van Dongeren, A.P., de Vries, J.V.T., McCall, R. & Lescinski, J. (2009) Modelling storm impacts on beaches, dunes and barrier islands. Coastal Engineering, 56, 1133–1152. Russo, A., Coluccelli, A., Carniel, S., Benetazzo, A., Valentini, A., Paccagnella, T., et al. (2013) Operational models hierarchy for short term marine predictions: The Adriatic Sea example. Proceedings of OCEANS-Bergen 2013 MTS/IEEE, 1–6, IEEE. Sallenger, A.H. (2000) Storm impact scale for barrier islands. Journal of Coastal Research, 16, 890–895. Sardá, R., Avila, C. & Mora, J. (2005) A methodological approach to be used in integrated coastal zone management process: The case of the Catalan Coast (Catalonia, Spain). Estuarine, Coastal and Shelf Science, 62, 427–439. Spencer, T., Brooks, S. M. & Möller, I. (2014) Floods: Storm surge impact depends on setting. Nature, 505, 26. Stephens, E. & Cloke, H. (2014) Improving flood forecasts for better flood preparedness in the UK (and beyond). The Geographical Journal, 180, 310–316. Stockdon, H.F., Holman, R.A., Howd, P.A. & Sallenger, A.H. (2006) Empirical parameterization of setup, swash, and runup. Coastal Engineering, 53, 573–588. Stockdon, H.F., Sallenger, A.H., Holman, R.A. & Howd, P.A. (2007) A simple model for the spatially variable coastal response to hurricanes. Marine Geology, 238, 1–20. UNISDR (United Nations – International Strategy for Disaster Reduction) (2007) Hyogo Framework for Action 2005–2015: Building the Resilience of Nations and Communities to Disasters. Valdemoro, H.I. & Jiménez, J.A. (2006) The influence of shoreline dynamics on the use and exploitation of Mediterranean tourist beaches. Coastal Management, 34, 405–423. Valentini, A., Delli Passeri, L., Paccagnella, T., Patruno, P., Marsigli, C., Cesari, D., et al. (2007) The sea state forecast system of ARPA-SIM. Bollettino di Geofisica Teorica e Applicata, 48, 333–350. Van Dongeren, A., Ciavola, P., Viavattene, C., De Kleermaeker, S., Martinez, G., Ferreira, O., et al. (2014) RISC-KIT: Resilience-Increasing Strategies for Coasts – toolkit. Journal of Coastal Research, SI 70, 366–371.

CH11 PREPARING FOR THE IMPACT OF COASTAL STORMS

239

Villatoro, M., Silva, R., Méndez, F.J., Zanuttigh, B., Pan, S., Trifonova, E., et al. (2014) An approach to assess flooding and erosion risk for open beaches in a changing climate. Coastal Engineering, 87, 50–76. Vousdoukas, M., Ferreira, O., Almeida, L.P. & Pacheco, A. (2012) Toward reliable storm-hazard forecasts: XBeach calibration and its potential application in an operational early-warning system. Ocean Dynamics, 62, 1001–1015. Zhang, K., Douglas, B.C. & Leatherman, S.P. (2000) Twentieth-century storm activity along the US east coast. Journal of Climate, 13, 1748–1761.

12 Assessing Storm Erosion Hazards Roshanka Ranasinghe1 and David Callaghan2 1 UNESCO-IHE 2 School

Institute, Delft, The Netherlands of Civil Engineering, University of Queensland, Brisbane, Australia

12.1

Introduction

The storm erosion hazard on coasts is usually expressed as an erosion volume and/or associated episodic coastline retreat (Hoffman & Hibbert, 1987; Li et al., 2014a; Woodroffe et al., 2014; Wainwright et al., 2015) . The accurate assessment of present-day and future storm erosion volumes is a key task for coastal zone managers, planners and engineers. These assessments are a precursor to coastal risk assessments, determination of coastal setback lines, design of coastal protection measures and the development of short and long-term coastal zone management/planning strategies. With the anticipated climate change driven variations in storm wave characteristics and occurrence frequency (IPCC, 2013), and ever expanding coastal communities (and associated developments and infrastructure), the accurate determination of the storm erosion hazard is now more important than ever before. The first issue that needs to be addressed when undertaking a storm erosion hazard study for coastal management/planning purposes is exactly how to define the storm erosion hazard. In practice, most managers/planners would require estimates of storm erosion volumes of pre-defined return periods (e.g. 1 in 100 years, 1 in 50 years, etc.) or exceedance level (e.g. 1%, 10%, 99% probability of being exceeded). The definition of the appropriate design return period should at least take into account the design life of structures and, if relevant, the business cycle. An accurate knowledge of the return period (or the exceedance probability) distribution of storm erosion volumes is crucial for making risk-informed (if not risk-based) on-the-ground decisions. However, most of the approaches traditionally adopted to determine the storm erosion hazard, in fact, return the storm erosion volume that would result from a storm wave event of a given return period or exceedance level. Thus, there is a significant difference between what is required and what is calculated. This aspect is discussed in the next section of this chapter. The second issue that needs to be addressed in a storm erosion hazard study is how to quantify storm erosion volumes as required. This is mostly done via some type of Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

242

COASTAL STORMS: PROCESSES AND IMPACTS

modelling exercise; the sophistication level of this will depend on a number of variables such as available data, funds, models, modelling skills, etc. Much of this chapter is devoted to summarising the various approaches (and their limitations/strengths) that are presently available to quantify storm erosion (section 12.3). Finally, the issue of how to use the storm erosion volume estimates thus obtained to develop effective coastal management/planning strategies, especially in view of contemporary risk informed decision making frameworks, needs to be tackled. This is addressed in section 12.4 of this chapter.

12.2 The diagnostic conundrum Historically, most coastal management/planning decisions have been based on the erosion volume that would result from a given return period (RP) storm wave height. This thinking was probably ’borrowed’ from the older and more established field of flood management where, for flood mitigation measures that could withstand, for example, a 100-year RP, flood height is designed based on the measured/modelled flood heights that result from the same RP riverflow, which is not incorrect as flood heights are dominantly governed by riverflow. However, this is not the case for storm erosion volumes. The coastal response to wave forcing is a highly non-linear, multi-variate process. The volume of sand that may be eroded due to a storm event with a given wave height may depend on many other things, including: the duration of the storm, direction of wave incidence, wave period, the antecedent coastal profile shape (i.e. accreted vs eroded profile), concurrent storm surge levels, tidal state when the storm occurs, offshore bathymetric features, such as reefs/islands. In short, two storms with the same RP wave height are more likely to result in very different erosion volumes than not. For example, the storm that occurred at Sydney, Australia in May 1997 had a storm wave height RP of 10 years, but the RP of the associated erosion volume was 4 years (Figure 12.1). This is one example of a situation where management/planning strategies based on a 10-year RP storm wave height would fail to mitigate the negative effects of the same RP storm erosion event. Indeed, there is now an emerging recognition among practitioners that this historical way of designing coastal protection measures is inappropriate and may lead to under-designed protection measures (Wainwright et al., 2014). A better approach to designing adaptation/protection measures to withstand a given RP storm erosion event is to directly use the storm erosion volumes of that RP as the Beach erosion above MSL and bounded by the 2 m contour (m3/m)

Hs,max (m)

12 10 8 6 100

101 Return period (years)

102

100

50

0 100

101 Return period (years)

102

Figure 12.1 May 1997 storm at Narrabeen Beach, in which the offshore significant wave height peaked at 8.1 m or 10-year RP wave height (left), resulting in 4-year RP erosion volume of 73 m3 /m (right).

CH12 ASSESSING STORM EROSION HAZARDS

243

diagnostic. However, this is easier said than done. In places where there is a long and high temporal resolution (at least monthly) historical record of coastal profile surveys, measured storm erosion volumes may be directly subjected to statistical analysis to determine the RP distribution of storm erosion volumes (Callaghan et al., 2008a). This could then be extrapolated to determine higher design RP erosion volumes. However, apart from a handful of locations around the world (Narrabeen, Moruya Beaches in SE Australia; Hasaki Beach, Japan; Duck, NC, USA) such data does not exist. The only other option is to undertake multiple simulations of a coastal profile model forced with a large number of different storm conditions and post-process the predicted erosion volumes to calculate the RP distribution of storm erosion volumes. The difficulty in this approach lies in the computational power required to execute the thousands of individual simulations of a commonly used coastal profile model such as XBeach (Roelvink et al., 2009) or SBEACH (Larson, 1988; Wise et al., 1996). Nevertheless, ways to circumnavigate this problem are slowly emerging and are further discussed in sections 12.3.3 and 12.3.4.

12.3

Quantifying storm erosion volumes for coastal management/planning

Storm erosion volumes are usually determined using coastal profile models (see also Chapter 10). For coastal management purposes, it is necessary to calculate not only the eroded volume in the submerged profile, but also that in the subaerial profile, usually also including the dune (if present). Therefore, profile models such as UNIBEST-TC (Reniers et al., 1995), LITPROF (Brøker-Hedegaard et al., 1991) and COSMOS (Nairn & Southgate, 1993), which only simulate the submerged profile, are of limited use in this respect. Commonly used generic models for calculating storm erosion (including the subaerial profile) include empirical models (Kriebel & Dean, 1993), semi-empirical models (SBEACH, Larson, 1988; Wise et al., 1996) and fully process-based models (XBeach, Roelvink et al., 2009). In addition, there are some very site specific empirical models used for coastal management purposes in some parts of the world, for example DUROS (Vellinga, 1986), DUROSTA (Steetzel, 1993), DUROS+ (van Gent et al., 2008), and DUNERULE (van Rijn, 2009) in the Netherlands. The rest of this section provides a summary of modelling approaches that have been used to quantify storm erosion volumes, in a more or less chronological order.

12.3.1 Coastal profile model application with Extrapolated Wave Exceedance Characteristics (EWEC) In this approach, first, available offshore wave data is subjected to a statistical analysis to obtain extrapolated storm wave height RPs of storm events with different durations as follows: 1.

From the offshore wave data time series, identify independent wave storms where the significant wave height (Hs ) exceeds a pre-determined threshold (usually around 2.5–3 m, depending on local conditions)

244

2.

COASTAL STORMS: PROCESSES AND IMPACTS

Estimate wave height exceedance curves by: a)

for each storm, determining the wave height that was exceeded for 1 h, 6 h, 12 h, 24 h, 48 h and 72 h; b) sorting the wave heights into ascending order for each duration and assign empirical return periods (the largest wave height for each duration is assigned the empirical return period of N where N is the record length in years, with the second and subsequent events assigned N/2, N/3 and so on); c) using exponential functions (straight lines on a log-linear plot) to extrapolate storm information to extreme event level; 3. Estimate storm wave heights at particular RPs by using the extrapolated wave height-frequency-duration curves (Figure 12.2). The 50-year or 100-year RP wave heights, which are generally utilised in storm erosion hazard studies, are then extracted for the desired storm duration (which is usually decided via expert judgment) from the above figure. The next step is to apply a coastal profile model to calculate the storm erosion volume that would result from the design storm event. Prior to application, however, the model should be calibrated and validated, ideally for storm conditions that have RPs as close as possible to the RP of the design storm event. As a minimum, sufficient wave, water level and before/after profile surveys would be required for three separate storm events (one for calibration, two for validation). If the required data exists, this approach is relatively straightforward to implement. The main limitations of this approach are: (1) the aforementioned issues related to the disparity between the storm wave heights of a given RP and the RP of the associated erosion volume; (2) providing only a single deterministic estimate, thus precluding risk-informed decision making; (3) reliability of model prediction when the RPs of

100 Significant wave height (m)

Return period 1000-year 100-year 50-year 10

20-year 10-year

1 0

50

100 Duration (hour)

150

200

Figure 12.2 Extrapolated wave height-frequency-duration curves for Narrabeen Beach, Sydney, Australia (from Callaghan et al., 2009).

CH12 ASSESSING STORM EROSION HAZARDS

245

storm events used for model calibration/validation are much lower than that of the design event; (4) non-consideration of the enhanced cumulative erosion effect when two or more storms occur in quick succession (i.e. storm clustering effect); and (5) non-inclusion of the effect of wave direction on storm erosion.

12.3.2 Coastal profile model application with the Synthetic Design Storm (SDS) approach The SDS approach (Carley & Cox, 2003) is an improved version of the above described EWEC approach. The main difference being that the SDS approach accounts for wave height variation during the storm event with the design RP wave height. In this method, the statistical analysis described above is continued on from step 2 as follows, prior to model application: Estimate the synthetic design storm wave height time series at particular RPs by: 1.

2. 3.

Using the extrapolated wave height-frequency-duration curves to estimate the wave height-duration estimates at 10-year, 20-year, 50-year, 100-year and 1000-year return periods (Figure 12.2). Subsequently, using exponential functions (straight lines on a linear-log plot) to extrapolate storm information to longer exceedance durations. For each return period, estimating the temporal evolution of the synthetic design storm using the fitted functions assuming that the temporal evolution of the storm events is symmetrical around the time that the maximum storm wave height occurs (Figure 12.3).

After identifying the design RP storm event as above, coastal profile modelling is executed in the same way as described in section 12.3.1. 10

1000-year ARI 100-year ARI 50-year ARI 20-year ARI 10-year ARI

Significant wave height (m)

9 8 7 6 5 4 3 2 1 0 0

1

2

3

4 Time (day)

5

6

7

8

Figure 12.3 Temporal evolution of various RP design storms at Narrabeen Beach, Sydney, Australia, obtained via the SDS approach (from Callaghan et al., 2009).

246

COASTAL STORMS: PROCESSES AND IMPACTS

While all of the limitations inherent to the EWEC approach are also shared by the SDS approach, it is a more scientifically defendable approach owing to its representation of the wave height variation during the design storm event. This approach has, however, been shown to under-estimate erosion volumes of less than 10-year RP (Callaghan et al., 2009). Shand et al. (2011) refined Carley and Cox’s (2003) SDS approach for Australian conditions, with particular attention to forcing distributions (e.g. wave height) and storm shapes (building in spatial and probability variations using wave measurements around Australia).

12.3.3 The Joint Probability Method (JPM) approach Callaghan et al. (2008a) trialled several statistical approaches to model coastal erosion at Narrabeen Beach, Sydney Australia, including the EWEC approach, the structural variable method, joint probability method (JPM) and full simulation of the JPM (referred to the JPM approach hereon). From this assessment, Callaghan et al. (2008a) selected the full simulation of the JPM, primarily due to its ability to include beach recovery between storms. This approach requires three major steps as follows (see also Figure 12.4): 1.

Statistical modelling of the storm wave and storm surge climate to derive a time series of storm wave/surge events. 2. Translating the storm wave/surge climate via a coastal profile model (i.e. structural function) to quantify beach erosion/accretion for each storm wave/surge event in the time series generated in (1) (this may also have to include wave transformations and adding in tidal variations where required). 3. Post-processing the time series of storm erosion volumes thus calculated to establish their RPs (as outlined in section 12.3.1). The initial application of the JPM approach to Narrabeen Beach by Callaghan et al. (2008a) adopted the empirical storm erosion model proposed by Kriebel and Dean (1993) (KD93 hereon) as the structural function, mainly for computational efficiency. With this arrangement, a thousand-year JPM simulation takes about two seconds on a contemporary single processor (e.g. Intel-Xeon L5520). The RP distribution of storm erosion volumes given by the JPM approach (Figure 12.5) can be directly used in risk informed coastal management/planning (see section 12.4 below). Since its introduction in 2008, the JPM approach has been extended in several ways including: estimation of confidence limits via bootstrapping (Callaghan et al., 2008b), inclusion of tidal variations and nonlinear wave propagation into the structural function (Callaghan & Wainwright, 2013), inclusion of the semi-empirical model SBEACH and the fully process-based model XBeach into the structural function (Callaghan et al., 2013), and inclusion of a process-based accretion model into the structural function (Pender et al., 2014). Alternative versions of the JPM approach have also been developed and applied by Vuik et al. (2015), Li et al. (2014b), Corbella and Stretch (2012), and den Heijer et al. (2012), with each application following the above three-step framework introduced by the JPM approach. The JPM approach has several strengths compared to the EWEC and SDS approaches. These include: (1) providing, as direct output, a continuous RP

CH12 ASSESSING STORM EROSION HAZARDS

247

Figure 12.4 Operational structure of the JPM approach (from Callaghan et al., 2013). Hs is significant wave height, D is storm duration, Tp is peak wave period, R is storm surge, 𝜃 m is mean storm wave direction and 𝛿t is time between individual storm events.

distribution of storm erosion volumes, thus circumnavigating the diagnostic issue discussed in section 12.3.1; (2) including the effects of enhanced cumulative erosion (storm clustering) and wave direction on storm erosion; and (3) direct facilitation of risk informed coastal management/planning via the provision of probabilistic storm hazard estimates. Despite these strengths, the JPM approach also has a few drawbacks including: (1) numerical complexity and computational power required; (2) simulation of beach recovery; and (3) the unreliability of model predictions when calibration/validation storm erosion volumes have small RPs compared to higher RP erosion events modelled within the 1000-year Montecarlo simulation. The issue of numerical complexity was sufficiently overcome by Callaghan et al. (2008a) via the use of the KD93 empirical storm erosion mode as the structural function. With this arrangement, a 1000-year JPM simulation bootstrapped 2000 times takes about one hour on a single processor (Intel-Xeon L5520). However, the required computational effort increases to about 40 days when either of the more sophisticated model SBEACH or XBeach is used as the structural function (Callaghan et al., 2013). How to simulate post-storm beach recovery accurately is still a significant outstanding question within the coastal engineering research community, with solutions (of limited accuracy) ranging from exponential recovery rates using timescales derived from medium-term measurements

248

COASTAL STORMS: PROCESSES AND IMPACTS

Beach erosion above MSL and bounded by 2 m contour (m3/m)

200

150

100

50

0

1

2

5

10

20

50

100

Return period, RP (year)

Figure 12.5 Return period distribution for Narrabeen Beach, Sydney, Australia using the JPM Approach. In this application KD93 was used as the structural function. The 95% confidence limits (shaded grey) were obtained by bootstrapping the 1000-year Montecarlo simulation 2000 times. The triangles indicate the RP distribution of measured storm erosion volumes (modified from Callaghan et al., 2013).

(Ranasinghe et al., 2012) through to beach state models (e.g. Yates et al., 2011; Splinter et al., 2014a). On the third drawback listed above, Callaghan et al. (2013) demonstrated that, especially for JPM-XBeach, the reliability of model predictions significantly increase when XBeach is calibrated to achieve the best possible match between the full RP distributions of measured and modelled storm erosion volumes, rather than following the traditional approach of calibrating against individual storm events. This is not unexpected as, due to the lack of very long-term profile measurements, the RPs of calibration events (RP < 30 years, at best) are likely to be much lower than a majority of the RPs of erosion events modelled within the Montecarlo simulation (RPs ranging from 1 to 100 years, at least).

12.3.4 Corbella and Stretch (CS) approach Corbella and Stretch (2012) presented a probabilistic storm erosion forecasting method that estimates the non-stationary impacts of a storm event of which the RP is known at the start of the forecast period. This method involves modeling the wave and water level climate using extreme value distributions with time dependent parameters. A statistical model is then used to estimate storm parameters moving forward through the forecast period from time zero. For example, if the 100-year RP storm in 2010 is known and is chosen as the design storm event, the CS approach can determine how this storm will look, due to non-stationary forcing, in 2060, that is, a 50-year forecast of the 100-year RP storm. Corbella and Stretch (2012) applied this approach to obtain a 100-year forecast of the 31-year RP storm event (based on storm wave height, period and storm duration)

CH12 ASSESSING STORM EROSION HAZARDS

249

35%

Percentage increase (%)

30% 25% 20% 15% 10% 5% 0% 0

20

40 60 Forecast (years)

80

100

Figure 12.6 The 100-year forecasted percentage increases of the significant wave height (dotted line), water level (dashed line) and erosion volume (solid line) associated with a 31-year RP storm event at one cross-shore profile in the Durban Bight, South Africa as calculated by the CS approach (from Corbella & Stretch, 2012).

in the Durban Bight, South Africa (Figure 12.6). The main steps of the Corbella & Stretch (2012) approach are: •

• • • •

•

Statistically model the storm forcing parameters (peak wave height during the storm, storm duration, typical wave period and wave direction, peak water level), including non-stationary effects through time-dependent distribution parameters. Estimate the time-dependent non-exceedance probabilities (across the forecast period) corresponding to a known (at the start of the forecast) RP storm event. Through simulation of the forcing statistical model, estimate time-dependent wave and water level forcing parameters (across the forecast period). Translate forcing parameters, if required, to relevant nearshore locations (e.g. using SWAN). Using various shape functions that are scaled by the forcing parameters, develop input time-series (e.g. wave height, period and direction and water level) for use in storm erosion models (e.g. peak wave height for a 10-year forecast is combined with a triangular shape function with base duration provided by the statistical model). Estimate time-dependent beach erosion across the forecast period (e.g. using XBeach).

This approach has the advantage of including non-stationary forcing and if applied to a number of different RP storm events, would facilitate probabilistic design (although the application of Corbella & Stretch (2012) was limited to a deterministic application). Thus, the CS approach may be modified and applied in a way that is useful

250

COASTAL STORMS: PROCESSES AND IMPACTS

for risk informed coastal management/planning. However, this approach, similar to deterministic EWEC and SDS approaches, does not overcome the disparity between the RPs of storm wave height and storm erosion volumes described in section 12.2. Furthermore, the CS approach does not account for storm clustering, which might be of crucial importance at some locations (Karunarathna et al., 2008; Callaghan et al., 2009; Splinter et al., 2014b) (see also Chapter 8). Further development of this procedure will be required to enable applications in regions where storms are generated by several different meteorological features of different spatial and temporal scale (which leads to storm duration being very weakly dependent on wave height). As is the case with all the other approaches discussed above, the reliability of predictions obtained with the CS approach will be questionable when model calibration/validation storms have significantly shorter RPs compared to the design event RP.

12.4 Application of storm erosion volume estimates in coastal management/planning Coastal managers/planners generally use storm erosion volume estimates to determine coastal setback lines or storm buffer zones. Single deterministic erosion volume estimates which may be obtained by using any of the methods described in sections 12.3.1 to 12.3.4 in conjunction with the dune stability schema presented by Nielsen et al. (1992) for this purpose (Figure 12.7). Here, in addition to the estimated storm erosion volume, consideration is also given to the process of dune slumping and the associated landward zone of reduced foundation capacity to arrive at a conservative and safe location for the coastal setback line, seaward of which developments should be restricted or prohibited. Probabilistic estimates of storm erosion obtained from the JPM approach (section 12.3.3) may be directly used within contemporary risk-informed coastal management/planning frameworks. Along a single coastal profile, the JPM approach will provide the full RP distribution of storm erosion volumes. If confidence limits are Zone of reduced foundation capacity Stable foundation zone

Zone of slope adjustment Zone of wave impact α

Slumped dune escarpment

i

Pre-storm deach-dune profile

Top of swash (~R.L. 2.0 m) 1:10

1:10

Scour level (~R.L. 1.0 m)

Angle of repose of dune sand: i ~ ϕ ≈ 34° Safe angle or repose of dune sand: α = tan−1 {(tan ϕ)/1.5} ≈ 24° All levels to AHD

0m (datum for reference volume calculations)

Figure 12.7 Dune stability schema proposed by Nielsen et al. (1992) for converting predicted storm erosion volumes to coastal setback lines (or storm buffer zones). AHD is Australian Height datum, which is approximately at mean sea level (i.e. MSL is AHD = 0).

CH12 ASSESSING STORM EROSION HAZARDS

Without sea level rise Exceedance probability

251

Property value AUS$/m2

0.001

0 – 500

0.01

501 – 1000

0.02

1001 – 1500

0.1

1501 – 2000 2001 – 2500 2501 – 3000 3001 – 3500 > 3500

0

125

250

500

750

1000 Meters

Figure 12.8 Erosion contours of different exceedance probabilities obtained by applying the JPM approach along a number of cross-shore profiles at Narrabeen Beach, Sydney, Australia. The green shaded areas indicate 2010 property values.

required, the JPM Montecarlo simulation could be bootstrapped (Figure 12.5). When considering a coastal cell (e.g. an embayed beach), the JPM approach could be applied along a number of cross-shore profiles, and the results could be combined to obtain erosion contours at different RPs or exceedance probabilities along the beach (Figure 12.8). Economic risk estimates (in terms of $ per m2 ) may also obtained by combining JPM predicted storm erosion volume exceedance probabilities with property values (Figure 12.9). Such risk assessments can be used in conjunction with economic considerations to obtain the position of the economically optimal coastal setback line (see Figure 12.9) using the economic modelling techniques described by Jongejan et al. (2011).

12.5

Conclusions and recommendations

Assessment of storm erosion volumes is a key task for coastal engineers/managers/ planners worldwide. Storm erosion volumes are generally used for the determination of coastal setback lines or storm buffer zones, design of coastal protection structures and development of emergency evacuation plans. One of the key issues that has been routinely ignored is whether the erosion volume due to a design RP storm wave height or the design RP storm erosion volume itself is more appropriate for use in coastal management/planning efforts. The former approach has been in wide use to date, primarily because wave data is more easily found

252

COASTAL STORMS: PROCESSES AND IMPACTS

Without sea level rise Yearly risk (AUS$/m2) 0.0 – 500.0 500.1 – 1000.0 1000.1 – 1500.0 1500.1 – 2000.0 2000.1 – 2500.0 2500.1 – 3000.0 3000.1 – 3500.0 3500.1 – 4000.0 4000.1 – 4500.0 4500.1 – 5000.0 Economically optimal SBL

0

125

250

500

750

1000 Meters

Figure 12.9 Risk map and economically optimal coastal setback line (blue line) for Narrabeen Beach, Sydney Australia, obtained by using the probabilistic output of the JPM approach in conjunction with property values and economic modelling methods described by Jongejan et al. (2011).

than morphological data and/or the lack of modelling methods to derive probabilistic estimates of storm erosion volumes that lend themselves to the determination of the RP distribution of storm erosion volumes. As measurements indicate that the RPs of the erosion volume and the causative storm wave height can indeed be rather different, with the former RP being much smaller than the latter, and as modelling approaches that are capable of providing probabilistic estimates of storm erosion volumes are now available, we strongly recommend the use of the RP of the storm erosion volume itself in future coastal management/planning efforts. At present, there are four main approaches that can be used to assess storm erosion volumes for coastal management/planning purposes. Of these, the EWEC and SDS approaches can only return a single deterministic estimate of storm erosion volume associated with a design storm wave height. However, in situations where a hazard assessment study is hindered by budgetary, computational or modelling skill constraints, these approaches may still be used to support coastal management/planning strategies if the user is acutely aware of the caveats associated with thus calculated storm erosion volume estimates. Nevertheless, these approaches do not serve the needs of contemporary risk informed coastal management/planning frameworks. The CS approach, as applied in its single application to date, also returns a single deterministic estimate of storm erosion volume (albeit at different times) resulting from a design RP storm event, which is calculated using highly sophisticated statistical modelling that takes into account the wave height, wave period and storm duration. The CS

CH12 ASSESSING STORM EROSION HAZARDS

253

approach can be applied differently, likely at significant computational cost, to obtain a range of storm erosion volumes with different RPs, both at present and in the future. The only approach that has a demonstrated capability to provide probabilistic estimates of storm erosion volumes, and thereby a full RP distribution of erosion volumes, is the JPM approach. While, for computational efficiency, the JPM approach was initially developed with the empirical KD93 model as the structural function, subsequent applications of the JPM framework (den Heijer et al., 2012; Li et al., 2014b; Vuik et al., 2015) have successfully used other more site specific empirical erosion models (DUROS+ and DUNERULE, respectively) as the structural function. Callaghan et al. (2013) compared JPM results given with three different structural functions with varying levels of sophistication (KD93, SBEACH and XBeach, see Figure 12.10). This application at Narrabeen Beach, Sydney, showed that JPM-SBEACH provided the best results, with a good agreement between measured and modelled storm erosion volumes at all RPs from 1 year to 100 years. When accounting for the uncertainty associated with modelled erosion volumes by bootstrapping the JPM Montecarlo simulations, Callaghan et al. (2013) found that 53%, 97% and 90% of data points fell within the 95% confidence limits of the JPM-KD93, JPM-SBEACH and JPM-XBeach predictions, respectively. These results show that despite the computational efficiency that a simple structural function such as KD93 affords (1 hour on a single processor for 2000 Montecarlo simulations of 1000 years duration each), a more process-based structural function such as SBEACH or XBeach (40 days on a single processor for 2000 Montecarlo simulations of 1000 years duration each) is desirable where accuracy is concerned (note: the 40-day computational time estimate for JPM-XBeach is applicable when offline XBeach simulations are used to develop a look-up table of erosion volumes for subsequent JPM simulations). However, it is indeed possible that the relative accuracy levels associated with the different structural functions may vary at sites that have significantly different

Beach erosion above MSL and bounded by 2 m contour (m3/m)

300 XBeach

250

SBeach

200

KD93 150 100 50 0

1

2

5 10 20 Return period, RP (year)

50

100

Figure 12.10 JPM application to Narrabeen Beach, Sydney, Australia with three different structural functions with varying levels of complexity (from Callaghan et al., 2013). The solid lines with symbols indicate measured storm erosion volumes calculated in two different ways (for details, please see Callaghan, 2013).

254

COASTAL STORMS: PROCESSES AND IMPACTS

dynamics to Narrabeen Beach. Furthermore, all applications of the JPM framework to date have been based on distributions fitted to storm wave characteristics derived from long data sets (20–30 years of continuous data). It is worthwhile to investigate whether using distributions fitted to freely available wave hindcast data (ERA40, WaveWatch III) instead would have much of an effect on the end diagnostic, that is, on the RP distribution of storm erosion volumes, or taking it even a couple of steps further, on the risk maps and/or the location of optimal coastal setback lines. If not, the generic applicability of the JPM framework would increase many-fold, making it an approach suitable for quantitative risk assessments even in data poor environments. Finally, what government regulatory authorities will do with the quantitative risk assessments and/or optimal coastal setback lines arising from coastal hazard assessments to a large extent governs how property values, coastal risk and the socioeconomic demography of the coastal zone will evolve in time. If, as is usually done in the contemporary imperfect economic market conditions, relevant government authorities invest in protecting high risk coastal areas (e.g. by constructing coastal protection structures or through buy-back schemes), the value of coastal properties will continue to rise as government interventions aimed at reducing societal risk will indirectly result in an enhanced reduction of (perceived) individual risk. In this scenario, particularly with foreshadowed coastal climate change impacts, economic risk will keep increasing along the coastal zone, while optimal setback lines will continue to move landward, placing more and more developments in high risk zones. On the other hand, if the government does not intervene, thus creating an economic market without imperfections, it is very likely that insurers will significantly increase premiums for properties in high risk zones, or in extreme cases, refuse to insure high risk properties (McNamara & Werner, 2008). This will have the immediate follow-on effect of increasing individual risk, with the inevitable consequence of properties located in high risk coastal zones becoming less sought after, greatly driving down their highly inflated prices. In turn, presently high risk zones will slowly become lower risk zones and optimal coastal setback lines will automatically move seaward, placing more and more coastal properties in lower risk zones. In essence a ’do nothing’ decision by governments will result in natural adaptation to coastal risk that would be auto-regulated by the insurance industry. The ultimate choice that property owners, investors, insurers and regulatory authorities have to make is straightforward but extremely difficult: How safe is safe enough?

Acknowledgments RR is supported by the AXA Research fund and the Deltares Harbour, Coastal and Offshore engineering Research Programme Bouwen aan de Kust.

References Brøker-Hedegaard, I., Deigaard, R. & Fredsøe, J. (1991) Onshore/offshore sediment transport and morphological modelling of coastal profiles. Coastal Sediments ’91, 643–657. Callaghan, D.P., Nielsen, P., Short, A.D. & Ranasinghe, R. (2008a) Statistical simulation of wave climate and extreme beach erosion. Coastal Engineering, 55 (5), 375–390.

CH12 ASSESSING STORM EROSION HAZARDS

255

Callaghan, D.P., Ranasinghe, R., Nielsen, P., Larson, M. & Short, A.D. (2008b) Processdetermined coastal erosion hazards, Proceedings of the 31st International Conference on Coastal Engineering. World Scientific, Hamburg, pp. 4227–4236. Callaghan, D.P., Ranasinghe, R. & Short, A. (2009) Quantifying the storm erosion hazard for coastal planning. Coastal Engineering, 56 (1), 90–93. Callaghan, D.P. & Wainwright, D.J. (2013) The impact of various methods of wave transfers from deep water to nearshore when determining extreme beach erosion. Coastal Engineering, 74, 50–58. Callaghan, D.P., Ranasinghe, R. & Roelvink, D. (2013) Probabilistic estimation of storm erosion using analytical, semi-empirical, and process based storm erosion models. Coastal Engineering, 82, 64–75. Carley, J.T. & Cox, R.J. (2003) A methodology for utilising time-dependent beach erosion models for design events. Proceedings of the 16th Australasian Coastal & Ocean Engineering Conference, Auckland, New Zealand, CD-ROM. Corbella, S. & Stretch, D.D. (2012) Predicting coastal erosion trends using non-stationary statistics and process-based models. Coastal Engineering, 70, 40–49. van Gent, M.R.A., van Thiel de Vries, J.S.M., Coeveld, E.M., de Vroeg, J.H. & van de Graaff, J. (2008) Large-scale dune erosion tests to study the influence of wave periods. Coastal Engineering, 55 (12), 1041–1051. den Heijer, C., Baart, F. & van Koningsveld, M. (2012) Assessment of dune failure along the dutch coast using a fully probabilistic approach. Geomorphology, 143–144, 95–103. Hoffman, J. & Hibbert, K. (1987) Collaroy/Narrabeen beaches, coastal process hazard definition study, 1. Public Works Department, Coastal Branch, NSW, PWD Report No. 87040, Sydney. IPCC (2013) Summary for policymakers. In: T.F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. Allen, J. Boschung, et al. (Eds) Climate change 2013: The physical science basis. Contribution of working group to the fifth assessment report of the intergovernmental panel on climate change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Jongejan, R.B., Ranasinghe, R., Vrijling, J.K. & Callaghan, D.P. (2011) A risk-informed approach to coastal zone management. Australian Journal of Civil Engineering, 9 (1), 47–59. Karunarathna, H., Reeve, D. & Spivack, M. (2008) Long-term morphodynamic evolution of estuaries: An inverse problem. Estuarine, Coastal and Shelf Science, 77 (3), 385–395. Kriebel, D.L. & Dean, R.G. (1993) Convolution method for time-dependent beach-profile response. Journal of Waterway, Port, Coastal and Ocean Engineering, 119 (2), 204–226. Larson, M. (1988) Quantification of beach profile change, Lund University, Lund, Sweden. Li, F., van Gelder, P.H.A.J.M., Ranasinghe, R., Callaghan, D.P. & Jongejan, R.B. (2014a) Probabilistic modelling of extreme storms along the Dutch coast. Coastal Engineering, 86, 1–13. Li, F., Gelder, P.H.A.J.M.v., Vrijling, J.K., Callaghan, D.P., Jongejan, R.B. & Ranasinghe, R. (2014b) Probabilistic estimation of coastal dune erosion and recession by statistical simulation of storm events. Applied Ocean Research, 47, 53–62. McNamara, D.E. & Werner, B.T. (2008) Coupled barrier island–resort model: 1. Emergent instabilities induced by strong human-landscape interactions. Journal of Geophysical Research: Earth Surface (2003–2012), 113.F1 Nairn, R.B. & Southgate, H.N. (1993) Deterministic profile modelling of nearshore processes. Part 2. Sediment transport and beach profile development. Coastal Engineering, 19 (1–2), 57–96. Nielsen, A.F., Lord, D. & Poulos, H.G. (1992) Dune stability considerations for building foundations. Australian Civil Engineering Transactions, Institution of Engineers Australia, CE34 (2), 167–174.

256

COASTAL STORMS: PROCESSES AND IMPACTS

Pender, D., Callaghan, D. & Karunarathna, H. (2014) An evaluation of methods available for quantifying extreme beach erosion. J. Ocean Eng. Mar. Energy, 1–13. Ranasinghe, R., Holman, R., de Schipper, M., Lippmann, T., Wehof, J., Duong, T.M., et al. (2012) Quantifying nearshore morphological recovery time scales using Argus video imaging: Palm Beach, Sydney and Duck, North Carolina, Proceedings of 33rd International Conference on Coastal Engineering, Santander, Spain. Reniers, A.J.H.M., Roelvink, J.A. & Walstra, D.J.R. (1995) Validation study of Unibest-TC model, Report H2130, Delft Hydraulics, Delft, The Netherlands. van Rijn, L.C. (2009) Prediction of dune erosion due to storms. Coastal Engineering, 56 (4), 441–457. Roelvink, D., Reniers, A., van Dongeren, A., van Thiel de Vries, J., McCall, R. & Lescinski, J. (2009) Modelling storm impacts on beaches, dunes and barrier islands. Coastal Engineering, 56 (11–12), 1133–1152. Shand, T.D., Mole, M.A., Carley, J.T., Pierson, W.L. & Cox, R.J. (2011) Coastal storm data analysis: Provision of extreme wave data for adaptation planning. Water Research Laboratories, University of New South Wales, School of Civil and Environmental Engineering, Manley Vale, Australia. Splinter, K.D., Turner, I.L., Davidson, M.A., Barnard, P., Castelle, B. & Oltman-Shay, J. (2014a) A generalized equilibrium model for predicting daily to interannual shoreline response. Journal of Geophysical Research: Earth Surface, 119 (9), 1936–1958. Splinter, K.D., Carley, J.T., Golshani, A. & Tomlinson, R. (2014b) A relationship to describe the cumulative impact of storm clusters on beach erosion. Coastal Engineering, 83, 49–55. Steetzel, H.J. (1993) Cross-shore transport during storm surges. Doctoral thesis, Delft University of Technology, the Netherlands. Vellinga, P. (1986) Beach and dune erosion during storm surges, PhD thesis, Delft Hydraulics Communications No. 372, Delft, the Netherlands. Vuik, V., van Balen, W. & van Vuren, S. (2015) Fully probabilistic assessment of safety against flooding along the Dutch coast. Journal of Flood risk management. Doi: 10.1111/jfr3.12153. Wainwright, D.J., Ranasinghe, R., Callaghan, D.P., Woodroffe, C.D., Cowell, P.J. & Rogers, K. (2014) An argument for probabilistic coastal hazard assessment: Retrospective examination of practice in New South Wales, Australia. Ocean & Coastal Management, 95, 147–155. Wainwright, D.J., Ranasinghe, R., Callaghan, D.P., Woodroffe, C.D., Jongejan, R., Dougherty, A.J., et al. (2015) Moving from deterministic towards probabilistic coastal hazard and risk assessment: Development of a modelling framework and application to Narrabeen Beach, New South Wales, Australia. Coastal Engineering, 96, 92–99. Wise, R.A., Smith, S.J. & Larson, M. (1996) SBeach: Numerical model for simulating storm-induced beach change. Report 4. Cross-shore transport under random waves and model validation with supertank and field data. US Army Corps of Engineers, Waterways Experiment Station, Technical Report CERC-89-9, United States. Woodroffe, C.D., Callaghan, D.P., Cowell, P.J., Wainwright, D.J., Rogers, K. & Ranasinghe, R. (2014) A framework for modelling the risks of climate-change impacts on Australian coasts. In: J. Palutikof, S. Boulter, J. Barnett & D. Rissk (Eds) Practical studies in climate change adaptation: Applied climate change adaptation research. John Wiley & Sons, Ltd. Yates, M.L., Guza, R.T., O’Reilly, W.C., Hansen, J.E. & Barnard, P.L. (2011) Equilibrium shoreline response of a high wave energy beach. Journal of Geophysical Research, 116 (C4), C04014.

Conclusions and Future Perspectives Giovanni Coco and Paolo Ciavola

The chapters presented in this book offer a state-of-the-art, in-depth overview of the effects of storms on a variety of coastal systems, of our understanding of the physical processes operating during storms and of our ability to predict and manage the impact of extreme events of a marine origin. In the comprehensive overview presented in this book the reader can only find the marine aspects of the storm processes, mainly waves, surges and the contribution that wind can bring to these phenomena. But additional coastal impacts can be generated by purely meteorological (e.g. extreme wind speeds, torrential rainfall, hailstones and snow) as well as hydrological processes (flash floods or river flooding). The advances in this field over the past decades have been remarkable: our ability to measure hydrodynamic and sediment transport processes under extreme conditions has increased, and we are now capable of obtaining concurrent small- and fast-scale hydrodynamic and sediment transport data. At the same time we are now able to measure in detail large-scale bathymetric changes and also to contextualize an individual storm within a climatological framework. Equally remarkable is the ability to feed fast and small-scale processes into numerical models that attempt to predict storm-driven changes over large areas and that at present are a key component of risk-assessment studies. However, the increased knowledge and the development of novel numerical models has not readily translated into improved predictions of storm impacts. The reasons for this are manifold and, in general terms, can be attributed to the many spatial and temporal scales involved, and to the many nonlinearities that further complicate the problem. More specifically, for example, the input of energy into coastal systems occurs over a variety of temporal scales ranging from waves to tides and the evolution of atmospheric fronts. But what further complicates the matter is that the frequency of storms and the ‘memory’ of the previous beach conditions can effectually determine storm impacts. At the same time, a storm is essentially a collection of nonlinear processes that include, just to mention a few, the breaking of waves and the transport of sediment. Research continues to advance our knowledge of these complicated processes but increased understanding has yet to translate into detailed prediction of storm impacts. The chapters presented in this book illustrate these points vividly, as in many cases we Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

258

CONCLUSION

still struggle to predict if a storm will be erosive or accretionary on a sedimentary coast. This should not surprise the reader as small differences in water levels can, for example, change the nature of overwash in barrier islands from constructive to destructive. Clearly, each natural system analyzed in this book has its own peculiarities and complications. On cliffed coastlines and coral reefs, the lithology and biology of the systems provide such a defined characterization of the systems that any aspiration to develop universal parameterizations is elusive and site-specific. The search for reliable, realistic and robust numerical models is at present approached from two distinctive directions. A deductive approach (e.g. the application of a numerical model based on conservation equations), which has the advantage of addressing physical processes and interactions within a clear theoretical context and associated explanatory power. At the same time, approaches with a strong inductive component (e.g. data-driven models of shoreline position based on equilibrium principles) can be stronger predictors, especially over intermediate to long timescales. Tackling the problem of coastal erosion from different methodological approaches, and possibly developing novel ones that more directly account for anthropogenic drivers and feedbacks, will, we hope, allow research in this area to progress fast. Improved understanding and prediction capability, particularly now that sea level rises are further exposing our coasts to storm attacks, will directly address the societal need to improve coastal hazard assessments and possibly develop early warning systems.

Index

Locators in italic refer to figures; those in bold to tables Adour Estuary, France, 28 Albufeira, Portugal, 178 Annie, tropical cyclone, 144 Arcachon Lagoon, France, 28, 31 archives. see records of past storms area morphological models, 198–201, 199, 200 ARPA-SIMC (Hydro-Meteo-Climate Service of the Environmental Agency of Emilia-Romagna), 232 Assateague Island, USA, 66, 185, 186 assessment, coastal storminess, 8. see also erosion hazard assessment; risk assessment/management atmospheric model, early warning systems, 230 atmospheric pressure gradients, hydrodynamics, 25, 38 atolls, 127, 128, 141 143, 145, 206. see also Funafuti Atoll Australia. see also Great Barrier Reef assessment of coastal storminess, 8 clustered storms, 151, 152, 154, 160 erosion hazard assessment, 242, 244 high energy environments, 6 morphological modeling, 203 Pasha Bulker storm, 11, 11 statistical approaches to storm classification, 14 backreefs/backreef storms, coral, 128, 130, 133, 139–144 backshore, topography, 7 Bangladesh, 3, 228 BARDEX (Barrier Dynamics Experiment), 182, 182–183, 184, 187, 188, 189 Barreta Island, Portugal, 176, 178, 184, 185, 186 barrier islands, 201, 218. see also Chandeleur Islands barrier reefs, 127. see also Great Barrier Reef barrier rollover, morphological modeling, 205

barrier spits. see spits bars clustered storms, 159–160 sandy beaches, 50, 53, 58 bathymetry cliff retreat, UK, 115 Gulf of Mexico, 29 hydrodynamics, 24 Bay of Bengal, hydrodynamics, 23 Bay of Biscay, hydrodynamics, 28 Bayonne, France, 31, 31 beaches, coral reefs, 128. see also sandy beaches Beaufort Scale, 5, 118 Beaufort, Sir Francis, 5 Bebe, tropical cyclone, 138, 141–142, 142 bed return flows, hydrodynamics, 32–33 bed shear stress, 48, 60 Belgium, North Sea storm surge, 3 Belize barrier reef, hurricane Hattie, 140–141, 141 berm building, morphological modeling, 205 Bewick Island, Great Barrier Reef, 138 Bhola cyclone, Bangladesh, 3 biological diversity, coral reefs, 127 Blaavands Huk, Denmark, 54, 54–58, 55, 56, 57 Bonnaire, coral reefs, 138 boulders, coral, 138, 142, 143 Boussinesq wave model (COULWAVE), 204 breaking wave processes, sandy beaches, 46 breakpoints, wave, hydrodynamics, 34–35 Bruun rule, cliff retreat, 116 buildings, coastal, 4 building waterline distance (BWD), early warning systems, 231–232 C2Shore morphological model, 200 Cadiz coast, Spain, 14, 164 Camille, hurricane, 16 Camp Osborne, Brick, USA, 209, 209

Coastal Storms: Processes and Impacts, First Edition. Edited by Paolo Ciavola and Giovanni Coco. © 2017 John Wiley & Sons Ltd. Published 2017 by John Wiley & Sons Ltd.

260

INDEX

Canada, statistical approaches to storm classification, 14 Catalonian coast, Spain, 14, 218, 223–227, 225, 226 Chandeleur Islands, USA, 65–66, 66, 68, 76–77 Assateague Island, 66 elevation changes, 68, 70, 70, 71, 71, 72, 74, 75, 76 modeling, 202, 202–203 quantifying changes, examples, 70–74, 71, 72, 73, 74, 75 resilience, 75–76 response to storms, 66–70 Sallenger storm-response model, 68–70, 69 shoreline change histograms, 75 Changjiang River delta, China, tidal flats, 81–83, 82, 96 conceptual model of morphodynamics, 95, 96–97 conceptual model of mud shore equilibrium shape, 91 erosion-deposition processes, 88–96, 89, 90, 91, 92, 93, 94, 95 sand/mud-dominated layers, 83–87, 84, 85 sedimentologic characteristics, 83–87, 84, 85, 86, 87, 88 settling/scouring lag, 88, 90 time-velocity asymmetry, 89 Tonglu rhythmites, 85, 87, 88 zonation, 93 chronic erosion, coastal, 45–46 cliffs, UK, 99–102, 119, 258 accumulated excess energy concept, 102 cliff retreat, 100–102, 101, 105, 106–107, 109–110, 113, 114 cliff retreat models, 115–117 cliff types, 103 future storm impact models, 117–119 past storm records, 106–110, 108 soft-rock cliff case study, 110–115, 111, 112, 113, 114 cliff-top storm deposits (CTSDs), 106–107 climate change, 3, 8, 153–154 cliff retreat, UK, 118–119 climate patterns, 8. see also El Niño/Southern Oscillation; North Atlantic Oscillation clustered storms, 7, 151–153, 152, 167 bar dynamics, 159–160 data collection, 156–157, 158 definitions, 154–156, 155, 156 dynamic equilibrium concept, 162–164, 163 genesis of, 153–154 morphologic feedback, 155, 160–162, 161 numerical models, 157–159 recovery period, 155, 156, 165–167 storm cluster characteristics, 156 storm impact scale model, 163, 164–165, 165, 167 coastal backshore, topography, 7 coastal environments, low/high energy, 5–6 coastal erosion. see erosion coastal management/planning, 250, 250–251, 251. see also risk assessment/management coastal plains, coral reefs, 128

coastal storms (general information), 1–4, 18–19, 257–258. see also specific storm types and storms by name assessment, 8. see also erosion hazard assessment; risk assessment classification, 8, 12–19, 13, 14, 16, 17, 18 clusters. see clustered storms definitions, 3–7 duration, 12–13, 14, 15, 18 hazard losses, 1, 2 impact indicators, early warning systems, 231 impact scales, 164–165, 165, 167, 218 ramparts, coral reefs, 134, 137, 138, 142, 142 records. see records of past storms synoptic systems, 9–12, 10, 11 threshold, 12, 14, 15. see also wave height coastal vulnerability, 3, 217, 219–223, 220, 222, 223, 224. see also risk assessment/management coastal zone, 1–3 Cocos (Keeling) Islands, 134 collision regime, Sallenger storm-response model, 68–69, 69 combined storms. see clustered storms constructive processes, coral reefs, 133, 135–139, 141–144, 142, 143, 144 convolution model, morphological modeling, 196 coral boulders, 138, 142, 143 coral coasts, morphological modeling, 206 coral reefs, 127–129, 145, 258 backreef storms, 139–144 Belize barrier reef, 141 destructive and constructive processes, 133 ecomorphodynamics, 130, 130–132 effect of storm waves, 132–134, 133 forereef storms, 134–136 Funafuti Atoll, Tuvalu, 138, 142 geomorphology, 128, 129–134 One Tree Reef, Australia, 128, 134, 135, 137, 139, 140 reef flat storms, 136–139, 138 reef islands, 139–144, 141, 144 reef island shoreline dynamics, 144 washover deposits, Maldives, 143 world distribution, 131 Corbella and Stretch (CS) approach, erosion hazard assessment, 248–250, 249, 252–253 Coriolis effect, hydrodynamics, 28, 30. see also Ekman transport/setup COSMO-I7 atmospheric model, 230 COULWAVE (Boussinesq wave model), 204 Covehithe, Suffolk, UK, 103, 111, 114 crest overtopping. see overtopping cross-shore sediment transport, 45, 46, 54, 58, 60 CSHORE morphological model, 198 CTSDs (cliff-top storm deposits), 106–107. see also cliffs cumulative storms. see clustered storms currents, sandy beaches, 48, 49, 49, 50, 51 cyclones, coral reefs, 127, 131, 131–132. see also extra-tropical cyclones; hurricanes; tropical cyclones; typhoons; and see specific storms by name

INDEX

Dauphin Island, USA, 187 deaths from natural disasters worldwide, 1, 2 decision support module, early warning systems, 231, 231 definitions clustered storms, 154–156, 155, 156 coastal erosion, 45 overwash processes, 175–177, 178 storm erosion hazards, 241 storms, 4 tidal flats, 81 Delaware Bay, USA, 6 Denmark Skallingen barrier spit, 54, 54–58, 55, 56, 57 statistical coastal storm classification, 16 depth-integrated shallow water equations, storm surges, 24 design life, coastal structures, 222–223, 223, 224 destructive processes, coral reefs, 133, 135–136, 139–141 Differential Global Positioning System (DGPS), 157 Digital Shoreline Analysis System (DSAS), 105 disaster risk reduction (DRR) measures, 219, 227, 234 dissipative sandy beaches, 47, 50, 51, 58 Dolan and Davis storm intensity classification, 17–18, 18 Douglas Scale, 5 DRR (disaster risk reduction) measures, 219, 227, 234 DSAS. see Digital Shoreline Analysis System Duck, USA, 159–160 dune-height variability, barrier islands, 69 dune stability schema, coastal management, 250, 250–251 Dunwich-Sizewell Bank, Suffolk, UK, 115 duration of a storm event, 7 DUROSTA morphological model, 198 dynamic equilibrium. see equilibrium states early warning systems (EWS), 227–234, 229, 231, 233, 258 ECE (estuarine and coastal ecosystem) protection, 119 ecomorphodynamics, coral reefs, 130, 130–132 ecosystem services, 127, 195 Ekman transport/Ekman setup, hydrodynamics, 26–28 elevation changes, shore. see also profile changes Chandeleur Islands, 68, 70, 70, 71, 71, 72, 74, 75, 76 Changjiang River delta, China, 85–86, 86, 94 overwash processes, 187 El Niño – Southern Oscillation (ENSO) clustered storms, 153 overwash processes, 183 past storm records, 107, 109–110 Emilia-Romagna early warning system, 228–234, 229, 231, 233 empirical models, morphological modeling, 196

261

equilibrium states clustered storms, 162–164, 163 wave conditions, 6, 7 erosion, coastal. see also specific examples/case studies barrier islands, 7 clustered storms, 153 definitions, 45 morphological modeling, 205 remote sensing, 157 risk management, 220, 221 sandy beaches, 60 tidal flats, China, 88–96, 89, 90, 91, 92, 93, 94, 95 erosion hazard assessment, 241–242. see also risk assessment/management coastal management/planning, 250, 250–251, 251 Corbella and Stretch approach, 248–250, 249, 252–253 extrapolated wave exceedance characteristics, 243–245, 244, 250, 252 joint probability method, 246–248, 247, 248, 250–254, 251, 252, 253 recommendations, 251–254 return period diagnostic indicator, 242, 242–252, 248 synthetic design storm approach, 245, 245–246, 250, 252 erosion volume, 151, 152, 241–248, 249, 250, 251–254, 253 Esbjerg Harbour, Wadden Sea, Denmark, 55, 56 estuarine and coastal ecosystem (ECE) protection, UK, 119 ETCs. see extra-tropical cyclones Eulerian currents, sandy beaches, 50, 58–59, 59 Europe, risk assessment/management, 218 European Union Floods Directive, 222 EWS (early warning systems), 227–234, 229, 231, 233, 258 extrapolated wave exceedance characteristics (EWEC), 243–245, 244, 250, 252 extra-tropical cyclones (ETCs), 12 statistical approaches to classification, 13, 16, 18 synoptic systems, 10–11, 11 extreme storms, 7 assessment, 8 coral reefs, 131 historical examples, 3 risk management, 222 failure state, risk management, 221 Fanø barrier island, Denmark, 58–59, 59 Federal Emergency Management Agency (FEMA), 221 field experiments clustered storms, 157 overwash processes, 180–181 financial implications of storms (hazard losses), 1, 2 Floods Directive, European Union, 222 Flood Forecasting Centre, UK, 228 forecast module, early warning systems, 230, 231 forereef, coral reefs, 128, 129, 133, 134–136

262

INDEX

France assessment of coastal storminess, 8 clustered storms, 160, 155, 162–164, 163 hydrodynamics, 28, 31, 31 overwash processes, 185, 186 storm classification, 14 French Polynesia, 135, 138 frequency/magnitude of coastal storms, 3, 8, 9 Freshwater, USA, 28, 30 fringing coral reefs, 127 ‘fully-developed sea’ concept, 5 Funafuti Atoll, Tuvalu, 138, 141–142, 142 Galveston, USA, 3, 28, 30 General Pareto Distribution (GPD) framework, 224 geology, cliff case study, 112 Geological Service and Civil Protection, 232 Gironde coast, France, clustered storms, 163 Glossary of Geology (Bates & Jackson), 81 Gold Coast, Australia, 203 Google Earth image, Skallingen, 54 GPS surveying, overwash processes, 181 gravel beaches, morphological modeling, 204–205, 205 Great Barrier Reef, Australia, 127, 128, 134, 135, 137, 138, 139, 140, 141 Greenland, past storm records, 107, 108 groove systems, coral reefs, 134–136, 137 grouped storms. see clustered storms Gulf of Mexico, 28, 29, 75. see also Chandeleur Islands Gustav, hurricane, 70, 74, 74, 75, 75, 76–77 Haiyan, typhoon, 23 Hamish, cyclone, 139 harbor resonance, hydrodynamics, 25, 38 hard coastal structures, modeling, 206, 207–209, 208, 209 Hattie, hurricane, 140–141, 141 hazard assessment/management. see erosion hazard assessment; risk assessment/management high energy environments, 6, 7, 93 historical extreme storms, 3. see also records of past storms Hunstanton, UK, chalk cliffs, 100, 101 hurricanes, 9, 16. see also tropical cyclones; and see specific storms by name hydrodynamics, 23, 38, 257 morphological modeling, 200 overwash processes, 183–185, 184 sandy beaches, 48–53, 49, 52 sea-surface drag coefficient, 26 storm surges, 23–31, 26, 27, 29, 30, 31 surf zone during storms, 31–38, 33, 34, 37 surface stress, 27 hydrostatic models, morphological modeling, 200, 200 Iceland, past storm records, 107, 108 Ike, hurricane, 28

Indian Ocean, tsunami, 143 Indian region, tropical cyclones, 9 infragravity waves (IG) waves hydrodynamics, 33, 33–35, 34, 38 morphological modeling, 200 sandy beaches, 46, 51, 52, 53, 57–58, 60 infrastructure, coastal, 4. see also hard coastal structures intertidal zone, 47–48, 54, 81. see also Changjiang River delta, tidal flats inundation, risk assessment/management, 220 inundation regime, Sallenger storm-response model, 69, 69, 70–71, 73 inverse barometric effect, 24, 25 IPCC (Intergovernmental Panel on Climate Change), 119 islands, coral (reef islands), 128, 133, 139–144, 141, 144 Italy assessment of coastal storminess, 8 early warning systems, 228–234, 229, 231, 233 morphological modeling, 209 overwash processes, 185, 186 storm classification, 16 Ivan, hurricane, 201, 202 Japan, tropical cyclones, 9 jet stream, 153, 154 Joachim, cyclone, 26, 27 joint probability method (JPM), erosion hazard assessment, 246–248, 247, 248, 250–254, 251, 252, 253 Katrina, hurricane, 3 Chandeleur Islands, 70–72, 72, 73, 74–76, 75 hydrodynamics, 23 overwash processes, 187 severity indicators, 16 Kelvin waves, 107 Kirby model, tidal flats, 91, 91 Klaus, storm, hydrodynamics, 28 laboratory experiments morphological modeling, 196, 208–209 overwash processes, 181–183, 182 Lady Elliot Island, Great Barrier Reef, 141 Lagrangian currents, sandy beaches, 50, 58–59, 59 landfall, tropical cyclones, 9–10 large benthic foraminifera (LBF), coral reefs, 139 LiDAR surveying, 181, 230 Lido di Dante, Italy, 185, 186 lifetime of coastal structures, 222–223, 223, 224 Lili, hurricane, 70, 70, 71, 72, 73, 74, 75, 76 Long Island, USA, hurricane Sandy, 202 Long Reef Point, Australia, clustered storms, 151, 152 longshore currents, hydrodynamics, 31–32 longshore sediment transport, 45–46, 54 Louisiana coastline, USA, hurricane Katrina, 3 low energy environments, 6 Lowestoft, UK, cliff case study, 113

INDEX

Maldives, coral reefs, 129, 143, 143, 145 management, erosion hazard assessment, 250, 250–251, 251. see also risk assessment/management Marshall Islands, coral reefs, 141 marsh, coastal, 81, 83. see also tidal flats Matagorda Peninsula, USA, 185 Mehby Rule, Changjiang River tidal flats, 91, 91 meteorological independence criterion, storm classification, 13, 14, 14 Met Office, UK, 228 Meyer-Peter model, sandy beaches, 49 MICORE project, 209, 228, 230 monetary losses, storms (hazard losses), 1, 2 morphological changes, shore. see elevation changes; profile changes morphological modeling, 195–196, 209–210 empirical models, 196 operational models, 209 process-based models, 197–201, 199, 200 process-model applications, 201–209, 202, 205, 207, 208, 209 mortality, from natural disasters worldwide, 1, 2 mud-dominated layer (MDL), Changjiang River delta, 83–87, 84, 85 Müller models, sandy beaches, 49 Nadikdik atoll, 141 Narrabeen Beach, Australia, 11, 11 erosion hazard assessment, 244, 246–248, 248, 251, 252, 253, 253, 254 natural disasters, deaths worldwide, 1, 2 Netherlands, 3, 199, 199 Net Offshore Migration (NOM) cycles, 160 New South Wales, Australia, 203–204 nor’easters, extra-tropical cyclones, 11 North Atlantic oscillation (NAO), 107, 153 North Sea storm surge (1953), 3 observation module, early warning systems, 230, 231 ocean model, early warning systems, 230, 232 oceanographic conditions, overwash processes, 183 Oly, cyclone, 138 one dimensional (profile) morphological models, 197–198 One Tree Reef, Australia, 128, 134, 135, 137, 139, 140 Ontong Java, Solomon Islands, 144 operational models, morphological modeling, 209 optically stimulated luminescence (OSL), 107 optimum state, risk assessment/management, 221 Oregon, USA, 58 oscillatory transport, sediment, 48, 49, 49, 51–52, 53, 57–58 overtopping morphological modeling, 205 overwash processes, 177, 178, 179, 188 overwash processes, 175, 189–190. see also washover Chandeleur Islands, 69, 69, 70, 72, 74, 75

263

coral reefs, 143–144 definitions, 175–177, 178 field experiments, 180–181 hydrodynamics, 183–185, 184 impact, 180 laboratory experiments, 181–183, 182, 187, 188, 189 occurrences, 176, 177–180, 179 Paibian, typhoon, 95 Pam, tropical cyclone, 4 Papua New Guinea, Takuu atoll, 143, 145 Pasha Bulker, storm, 11, 11 past storm records. see records of past storms peaks-over-threshold (POT) method, 12–14, 13, 14 peak wave time series, clustered storms, 155 Philippines, 9, 23 planning, erosion hazard assessment, 250, 250–251, 251. see also risk assessment/management Poisson clusters, clustered storms, 154 Portugal assessment of coastal storminess, 8 clustered storms, 154 overwash processes, 176, 178, 184, 185, 186 statistical approaches to storm classification, 14 POT (peaks-over-threshold) method, 12–14, 13, 14 probabilistic models clustered storms, 159 morphological modeling, 203 risk assessment/management, 222, 223 process-based morphological models, 197–201, 199, 200 applications, 201–209, 202, 205, 207, 208, 209 profile changes, shore. see also elevation changes Chandeleur Islands, 65 clustered storms, 160–162, 161 overwash processes, 185–189, 186, 187, 188 sandy beaches, 45, 46–48, 46 profile (one dimensional) morphological models, 197–198 punctuated erosion, 45, 46, 54–55 radiation stress gradients, 24, 28, 32, 33, 207 rain bands, tropical cyclones, 9 real time kinematic (RTK) instrumentation, 105 records of past storms cliffs, United Kingdom, 106–110, 108 clustered storms, 156, 157 recovery period, 7, 155, 156, 165–167 reef crest, coral reefs, 129 reef flats, coral reefs, 128, 129–130, 133, 133 morphological modeling, 206 storms, 136–139, 138 reef islands, coral reefs, 128, 133, 139–144, 141, 144 reef-lined coasts, morphological modeling, 206 reef platforms, coral reefs, 127 reflective sandy beaches, 47, 48 remote sensing, coastal erosion, 157 resilience, barrier islands, 75–76 resonance/multi-resonant processes, hydrodynamics, 25, 38

264

INDEX

response approach, risk management, 221 retreat, coastal cliffs, 100–102, 101, 105, 106–107, 109–110, 113, 114 erosion hazard assessment, 241 models, 115–117 return period (RP) erosion hazard assessment, 242, 242–252, 248 severity indicators, 13, 16–17 rip currents, sandy beaches, 50, 53 ripples, seabed, 48–49, 83 RISC-KIT risk assessment/management tool, 218 risk assessment/risk management, 217–219, 234–235. see also erosion hazard assessment Catalonia vulnerability assessment, 223–227, 225, 226 early warning systems, 227–234, 229, 231, 233 vulnerability framework, 219–223, 220, 222, 223, 224 Risk Index (RI), storm intensity classification, 18 rocky platform coasts, 206. see also hard coastal structures Roi-Namur Island, Republic of the Marshall Islands, 176 ROMS Ocean model, 230, 232 root mean squared error (RMSE), hydrodynamics, 28, 31 Rossby wave breaking, clustered storms, 153 RP. see return period RTK (real time kinematic) instrumentation, 105 rubble deposits, coral reefs, 129–130, 133, 134, 135–137, 138, 140, 142, 142 safe corridor width (SCW), early warning systems, 231 Saffir-Simpson hurricane wind scale, 9, 10, 16 St. George Island, USA, 186, 187 Saint-Venant depth-integrated shallow water equation, 24 Sallenger storm-response model, 68–70, 69 Salthouse, UK, overwash processes, 179 sand aprons, coral reefs, 128, 139 sand-dominated layer (SDL), Changjiang River delta, China, 83–87, 84, 85 Sandy, hurricane Chandeleur Islands, 77 Long Island barrier, 202 morphological modeling, 208, 209, 209 storm surges, 23 sandy beaches. see also sediment transport on sandy beaches classification, 47 morphological modeling, 195–196, 201–204, 202 Sandy Duck experiment, 32–33 Santa Rosa Island, USA, 185, 186, 201, 202 SBEACH model morphological modeling, 197–198 storm erosion hazard assessment, 243, 247, 253 SCAPE (Soft Cliff And Platform Erosion) model, 117 Scotland, United Kingdom, 106–107 scouring lag, Changjiang River delta, China, 88, 90

SDS. see synthetic design storm sea levels, 6–7 Chandeleur Islands, 65 clustered storms, 163, 164–165, 165 hydrodynamics, 24, 25 past storm records, 107 storm classification, 15–16, 16, 17, 19 total water level, 6–7, 15–16 United Kingdom, 117 seasonality, storminess, 153 sea-surface drag coefficient, hydrodynamics, 26 sedimentary archives, clustered storms, 156, 157 sediment transport/deposition, tidal flats, 81–82, 83–87, 84, 85, 86, 87, 88 sediment transport on sandy beaches, 45–46, 60, 257 lower shoreface, 58–59, 59 morphologic consequences of storms, 45, 46, 46–48 offshore-directed, 53 onshore-directed, 50–51, 53, 58 transport processes during storms, 48–53, 49, 52 upper shoreface, 53–58, 54, 55, 56, 57 see-saw effects, climate, 107, 154 SELFE modeling system, hydrodynamics, 28 settling lag, Changjiang River delta, China, 88, 90 severity indicators, storm classification, 16–18, 18, 19 shear stress coral reefs, 133 hydrodynamics, 26, 32, 38 sandy beaches, 48, 49, 51, 60 tidal flats, 88, 89 Shetland, UK, past storm records, 106–107, 108 shoaling wave processes morphological modeling, 201 sandy beaches, 46, 46, 47, 48 shore profiles. see elevation changes; profile changes short wave dissipation, hydrodynamics, 28, 31 Sillon de Talbert, France, 185, 186 SIIs (storm impact indicators), early warning systems, 231 Skallingen barrier spit, Denmark, 54, 54–58, 55, 56, 57 South Africa, storm classification, 14 Southwold, soft-rock cliff case study, 113, 113 Spain assessment of coastal storminess, 8 clustered storms, 164 risk assessment/management, 218, 223–227, 225, 226 storm classification, 14, 17–18, 18 spectral wave model (STWAVE), 200 spits, barrier Chandeleur Islands, 65, 66 Skallingen, 54, 54–58, 55, 56, 57 spur systems, coral reefs, 134–136, 137 statistical techniques clustered storms, 159 storm classification, 8, 12–19, 13, 14, 16, 17, 18 St. George Island, USA, 186, 187 Stokes drift, 32, 50, 58

INDEX

storm surges, 1, 3, 11–12 Chandeleur Islands, 68, 69, 70, 71, 71, 72, 74, 74, 76 cliff retreat, UK, 118 coral reefs, 133, 133–134 deaths worldwide, 2 hydrodynamics, 23–31, 26, 27, 29, 30, 31 storms, definition, 4. see also coastal storms STWAVE (spectral wave model), 200 suction responses, soft-rock cliff case study, 112, 113 Suffolk coast, United Kingdom, 105, 106 soft-rock cliff case study, 110–115, 111, 112, 113, 114 super storm Sandy. see Sandy, hurricane surface stress, hydrodynamics, 25–26, 26, 27 surf zone clustered storms, 164 hydrodynamics, 31–38, 33, 34, 37 morphological modeling, 201 sandy beaches, 46, 47, 50, 53 SWAN model, 230, 232, 249 SWASH model, morphological modeling, 201 swash regime, Sallenger storm-response model, 68, 69, 72 swash zone clustered storms, 161, 164 hydrodynamics, 35–38, 37 sandy beaches, 46, 47 swell, 5, 6 Sydney, Australia, erosion hazard assessment, 242 synoptic climatological approaches, assessment, 8 synoptic systems, coastal storms, 9–12, 10, 11 synthetic design storm (SDS) approach, 245, 245–246, 250, 252 Taiwan, tropical cyclones, 9 Takuu atoll, Papua New Guinea, 143, 145 Tasmania, 6, 6 tidal flats, Chinese distribution 82. see also Changjiang River delta tidal wave asymmetry, Changjiang River delta, 88–89, 89 tide effects, clustered storms, 163, 164–165, 165 tide forcing, Changjiang River delta, China, 83 time exposure images, clustered storms, 157, 158 timing of storm events, 7, 8 Tomas, cyclone, 139 Tonglu rhythmites, Changjiang River delta, 85, 87, 88 topography, influence of, 7 total water level (TWL), 6–7, 15–16 tropical cyclones, 9–10, 10, 12. see also specific storms by name clustered storms, 154 hydrodynamics, 25 storm classification, 13, 16 tropical storms, 9. see also above Truc Vert beach, France, clustered storms 155,, 162–164 tsunami, Indian Ocean, 143

265

two dimensional (area) morphological models, 198–201, 199, 200 TWL (total water level), 6–7, 15–16 typhoons, Changjiang River delta, China, 86, 91, 95. see also specific storms by name undertow, sandy beaches, 50, 51, 53, 54, 57, 57–59 UNIBEST-DE morphological model, 198 United Kingdom. see also cliffs, UK Met Office/EA Flood Forecasting Centre, 228 North Sea storm surge (1953), 3 storm classification, 14, 16 United States. see also Chandeleur Islands, USA assessment of coastal storminess, 8 clustered storms, 159–160 extra-tropical cyclones, 11 hazard losses, 1, 2 hurricane Katrina, 3 hydrodynamics, 23 low energy environment, 6 morphological modeling, 201, 202, 209, 209 overwash processes, 185, 186, 187 past storm records, 107, 109–110 storm classification, 14, 16, 18 Unites States Geological Survey (USGS), 70, 217–218 unmanned aerial vehicles (UAVs), clustered storms, 157 urbanized coasts, morphological models, 207–209, 208, 209 USGS (Unites States Geological Survey), 70, 217–218 Van Thiel-Van Rijn sediment transport formulation, 202 Vanuatu island, tropical cyclone Pam, 4 vegetated coasts Changjiang River delta, China, 93 morphological modeling, 206–207, 207 visualization module, early warning systems, 231, 232 volume, erosion. see erosion volume vulnerability, coastal, 3, 217, 219–223, 220, 222, 223, 224. see also risk assessment/management Wadden Sea, Denmark, 55, 56 warning module, early warning systems, 231, 232 washover. see also overwash processes coral reefs, 142–143, 143, 145 deposits, 175, 177, 179, 180, 181, 185, 187, 189 water-level time-series, storm classification, 15–16, 16, 17, 19. see also sea-level; total water level wave forcing, 242 Changjiang River delta, China, 82, 83, 89 morphological modeling, 196 risk assessment/management, 219, 220 storm clusters, 159, 161 wave heights Changjiang River delta, China, 93 clustered storms, 155 storm classification, 12–15, 13, 14, 19 tidal flats,

266

waves, coastal, 5, 6, 6 hydrodynamics, 24, 26, 27, 28, 31 sandy beaches, 46, 48 tidal flats, 83 wetlands, coastal, 81. see also tidal flats wind clustered storms, 153 hydrodynamics, 24, 26

INDEX

XBeach model clustered storms, 151–152, 152, 160 early warning systems, 230, 232 morphological modeling, 200–206, 205, 208, 209, 209 storm erosion hazard assessment, 243, 247–248, 253 Xynthia, storm, 25–26, 27