The Lithosphere Beneath the Indian Shield: A Geodynamic Perspective [1st ed.] 9783030529413, 9783030529420

Table of contents :
Front Matter ....Pages i-xviii
Continental Lithosphere (Ashoka G. Dessai)....Pages 1-51
Indian Shield (Ashoka G. Dessai)....Pages 53-96
Xenolith Petrology (Ashoka G. Dessai)....Pages 97-124
Lithosphere Architecture (Ashoka G. Dessai)....Pages 125-153
Geothermal Structure (Ashoka G. Dessai)....Pages 155-182
Integration of Insights (Ashoka G. Dessai)....Pages 183-253
Back Matter ....Pages 255-259

Citation preview

Modern Approaches in Solid Earth Sciences

Ashoka G. Dessai

The Lithosphere Beneath the Indian Shield A Geodynamic Perspective

Modern Approaches in Solid Earth Sciences Volume 20

Series Editors Yildirim Dilek, Department of Geology and Environmental Earth Sciences, Miami University, Oxford, OH, USA Franco Pirajno, The University of Western Australia, Perth, WA, Australia Brian Windley, Department of Geology, The University of Leicester, Leicester, UK

Background and motivation Earth Sciences are going through an interesting phase as the traditional disciplinary boundaries are collapsing. Disciplines or sub-disciplines that have been traditionally separated in the past have started interacting more closely, and some new fields have emerged at their interfaces. Disciplinary boundaries between geology, geophysics and geochemistry have become more transparent during the last ten years. Geodesy has developed close interactions with geophysics and geology (tectonics). Specialized research fields, which have been important in development of fundamental expertise, are being interfaced in solving common problems. In Earth Sciences the term System Earth and, correspondingly, Earth System Science have become overall common denominators. Of this full System Earth, Solid Earth Sciences – predominantly addressing the Inner Earth - constitute a major component, whereas others focus on the Oceans, the Atmosphere, and their interaction. This integrated nature in Solid Earth Sciences can be recognized clearly in the field of Geodynamics. The broad research field of Geodynamics builds on contributions from a wide variety of Earth Science disciplines, encompassing geophysics, geology, geochemistry, and geodesy. Continuing theoretical and numerical advances in seismological methods, new developments in computational science, inverse modelling, and space geodetic methods directed to solid Earth problems, new analytical and experimental methods in geochemistry, geology and materials science have contributed to the investigation of challenging problems in geodynamics. Among these problems are the high-resolution 3D structure and composition of the Earth’s interior, the thermal evolution of the Earth on a planetary scale, mantle convection, deformation and dynamics of the lithosphere (including orogeny and basin formation), and landscape evolution through tectonic and surface processes. A characteristic aspect of geodynamic processes is the wide range of spatial and temporal scales involved. An integrated approach to the investigation of geodynamic problems is required to link these scales by incorporating their interactions. Scope and aims of the new series. The book series “Modern Approaches in Solid Earth Sciences” provides an integrated publication outlet for innovative and interdisciplinary approaches to problems and processes in Solid Earth Sciences, including Geodynamics. It acknowledges the fact that traditionally separate disciplines or sub-disciplines have started interacting more closely, and some new fields have emerged at their interfaces. Disciplinary boundaries between geology, geophysics and geochemistry have become more transparent during the last ten years. Geodesy has developed close interactions with geophysics and geology (tectonics). Specialized research fields (seismic tomography, double difference techniques etc ), which have been important in development of fundamental expertise, are being interfaced in solving common problems. Accepted for inclusion in Scopus. Prospective authors and/or editors should consult one of the Series Editors or the Springer Contact for more details. Any comments or suggestions for future volumes are welcomed.

More information about this series at http://www.springer.com/series/7377

Ashoka G. Dessai

The Lithosphere Beneath the Indian Shield A Geodynamic Perspective

Ashoka G. Dessai Department of Earth Science Goa University Taleigao Plateau, Goa, India

Responsible Series Editor: F. Pirajno

ISSN 1876-1682 ISSN 1876-1690 (electronic) Modern Approaches in Solid Earth Sciences ISBN 978-3-030-52941-3 ISBN 978-3-030-52942-0 (eBook) https://doi.org/10.1007/978-3-030-52942-0 © Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To Arpita, Anushka, and Aniket

Foreword

Cratons are the foundations of our continents, and thus of human life; their highly depleted and buoyant Archaean roots, up to 350 km deep, are difficult to destroy and have provided a ‘life raft’ for the continental crust. However, tectonic and magmatic processes can modify the cratonic roots, thinning them and increasing their density over time, and ultimately affecting their long-term stability. These processes have attracted the attention of many petrologists over the past few decades, and the ongoing integration of petrology/geochemistry with geophysics and numerical geodynamic modelling is producing many answers and raising just as many new questions. Peninsular India—the Indian shield—has proven to be an excellent natural laboratory for such studies. Like many shields, it is composed of several (6 or 7, depending on who is counting) smaller cratons, each with its unique crustal history and an equally unique subcontinental lithospheric mantle (SCLM). Data on mantle xenoliths are available from many of these units and several time slices, establishing a basic petrological framework, and the country has an abundance of geophysical data to which petrological interpretations can be applied. The Indian shield is notable among the cratons of the world because today there is little evidence of a thick lithospheric root beneath most of the shield although geophysical interpretations remain contradictory in some areas. India is of course geodynamically interesting because of its rapid trek from the southern hemisphere to its present location: ca 6000 km northwards drift after its separation from Gondwana (120– 130 Ma) to collision with Eurasia at ca. 40 Ma, accelerating over the last 20 m.y., while collecting the Kohistan Arc and over-running the Reunion plume and other hotspots along the way. Could it move like this because it had lost some of its cratonic roots before drifting even started, or were they modified along the way? In this book, Professor Ashoka G. Dessai has summarised a lifetime of research on the lithosphere of the Indian shield and demonstrates the value of an integrated approach, which commonly is not incorporated in individual (especially older) studies. The main chapters outline the geological and tectonic history of the shield, the distribution and nature of the available xenolith suites, and the geophysical vii

viii

Foreword

studies, with a strong emphasis on seismic and geothermal data and their interpretation. These aspects are drawn together in the longest chapter ‘Integration of Insights’, which culminates in an overview of the evolutionary differences among the individual cratons of the shield. Prof. Dessai’s own words nicely sum up the essential scientific methodology underlying the book: ‘. . . an integration of detailed petrological-geochemical-geophysical data inputs can provide a better and deeper insight into the processes during the Archaean and their subsequent modification over time’. The application of this methodology to the lithospheric diversity of the Indian shield has produced a book that will become a standard reference on the subject, and hopefully will promote such integrated studies in the continuing investigation of the problems of the shield, and of other cratonic areas worldwide. GEMOC ARC, National Key Centre, Department of Earth and Planetary Sciences, Macquarie University, Macquarie Park, NSW, Australia

William L. Griffin

Preface

The continental lithosphere preserves a treasure of geological information on the evolution of the planet. Knowledge of its physico-chemical composition, thermal structure, and the mechanism of its evolution are crucial to understanding the processes responsible for the formation of the pristine crust and its subsequent growth by interaction with the mantle plumes and plate tectonics. Lithosphere studies involve a gamut of disciplines including geology, geochemistry, geophysics among others. Geophysics in particular employs various remotely sensed techniques to generate a multitude of data that are used for imaging the deep lithosphere and understanding its architectural framework. However, an interdisciplinary outlook across sciences is desirable for a comprehensive and inclusive approach to knowledge and its applications between and among disciplines. This book is a multidisciplinary perspective to comprehend the structure, composition, origin, and evolution of the lithosphere beneath the Indian shield which comprises a collage of cratons formed under varied tectonic environments over the past 3.6 billion years of natural history of the earth. The lithosphere beneath the cratons is supposedly thinner as compared to the cratons globally. Individual cratons reveal independent geotectonic and thermal histories, and this has implications in the formation of the crust and evolution of the lithospheric mantle which are petrologically varied and show considerable spatiotemporal variation in thickness both intracratons and among cratons. The book addresses in particular the following issues: (a) the petrological and geochemical variation of the deep crust and the extent of its evolution across the cratons of the Indian shield, (b) the nature of the crust-mantle and the lithosphereasthenosphere boundary in different tectonic domains, (c) evolution of the thermal structure of the lithosphere and its variability in space and time, and (d) integration of the petrological, geophysical, geochemical, and geothermal data inputs towards an improved understanding of the evolutionary processes of the Indian lithosphere in the context of global mantle dynamics and its evolution.

ix

x

Preface

The objective of this book, therefore, is to provide the reader a reference book that utilises the data and results from a combination of disciplines with emphasis on petrochemical studies on crustal and mantle xenoliths towards an updated and improved understanding of the geodynamic evolution of the Indian shield. Taleigao Plateau, Goa, India

Ashoka G. Dessai

Acknowledgements

The book has greatly benefitted from the insightful previews by W. L. Griffin, B. Lehmann, U. L. Raval, and L. Vinnik. I appreciate the keen interest they have displayed in this regard. During the preparation of the manuscript, Riffin T. Sajeev and my son-in-law Ankit Shrivastava have provided valuable assistance with the graphics and illustrations, and my former students and colleagues Amay Dashaputre, Pradip Jadhav, Makarand Kale, Lalit Kshirsagar, and Anthony Viegas have also helped in this process in multiple ways. Each of them deserves my sincere thanks. The sources of various figures and illustrations adapted in the course of this work are gratefully acknowledged. Shrimathy Venkatalakshmi and S. Amritha Varshini have helpfully spared time to go through the narrative despite their busy schedule. Notwithstanding this, if any inconsistencies, errors, or omissions have still crept into the text, they are entirely mine. It was a pleasure working with the highly competent Earth Sciences team from Springer headed initially by Petra van Steenbergen and later by Annett Büttner. Special thanks to team members, Yosuke Nishida, Taeko Sato, and Karthika Menon, who were instrumental in shaping this work to its current presentable avatar. Taleigao Plateau, Goa, India

Ashoka G. Dessai

xi

Contents

1

Continental Lithosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Geochemical Differentiation of the Earth . . . . . . . . . . . . . . . . . 1.2.1 Bimodality of the Crust . . . . . . . . . . . . . . . . . . . . . . . 1.3 Earth’s Mantle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.1 Mantle Composition . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3.2 Lithosphere Studies in India . . . . . . . . . . . . . . . . . . . . 1.4 Methods and Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.1 Seismic Methods and Properties of Rocks . . . . . . . . . . 1.4.2 Electromagnetic and Magnetotelluric Techniques . . . . . 1.4.3 Gravity Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.4 Xenolith Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.5 Thermal Structure of the Crust . . . . . . . . . . . . . . . . . . 1.4.6 Heat Flow Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .

1 1 3 4 11 12 19 21 21 31 36 37 39 42 45

2

Indian Shield . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Geotectonic Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Geology of the Peninsular Block . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Aravalli Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Bundelkhand Craton . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Singhbhum Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Bastar Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.5 Dharwar Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 Granulite Terrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.7 Chotanagpur Gneiss-Granulite Terrain . . . . . . . . . . . . . . 2.3.8 Eastern Ghat Mobile Belt . . . . . . . . . . . . . . . . . . . . . . . 2.3.9 Southern Granulite Terrain . . . . . . . . . . . . . . . . . . . . . . 2.4 Geophysical Studies in Southern India . . . . . . . . . . . . . . . . . . . .

53 54 56 56 59 61 63 65 66 71 71 72 72 78 xiii

xiv

Contents

2.4.1 Crustal Structure from Magsat and Gravity Data . . . . . 2.4.2 Crustal Configuration of the Off-Shore Region . . . . . . . 2.5 Xenolith Host Rocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.1 CITZ Dyke Swarm . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5.2 Western Continental Margin Dyke Swarm . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

79 84 86 86 87 89

3

Xenolith Petrology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Xenolith Occurrences in India . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1 Radiometric Ages of Host Rocks . . . . . . . . . . . . . . . . . . 3.3 Petrology of Xenoliths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Xenoliths from the Deccan Traps . . . . . . . . . . . . . . . . . 3.3.2 Xenoliths from Kutch . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Xenoliths in Kimberlites . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97 98 99 101 102 102 115 117 122

4

Lithosphere Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Crustal Structure: Seismic Studies . . . . . . . . . . . . . . . . . . . . . . . 4.2.1 Aravalli Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Bundelkhand Craton and the Northern Region of WDC . . 4.2.3 Central Region of WDC . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4 The Region Beneath the EDC . . . . . . . . . . . . . . . . . . . . 4.2.5 Bastar Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6 Singhbhum Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.7 Pericratonic Domain: SGT . . . . . . . . . . . . . . . . . . . . . . 4.2.8 Offshore Crust of Arabian Sea . . . . . . . . . . . . . . . . . . . 4.2.9 Offshore Crust of Bay of Bengal . . . . . . . . . . . . . . . . . . 4.3 Nature of the Subcontinental Moho . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Characteristics of Shallow Mantle . . . . . . . . . . . . . . . . . 4.3.2 Nature of the Crust-Mantle Boundary . . . . . . . . . . . . . . 4.3.3 Impact of Temperature on Seismic Interpretation . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125 126 126 128 129 131 135 136 137 138 141 145 146 146 147 148 149

5

Geothermal Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Thermal Structure of the Indian Shield . . . . . . . . . . . . . . . . . . . 5.2.1 Bundelkhand and WDC: Thermal Structure . . . . . . . . . 5.2.2 Thermal Regime of EDC and Bastar Cratons . . . . . . . . 5.2.3 Southern Granulite Terrain: Thermal Characteristics . . . 5.2.4 Xenolith-vis-à-vis Heat Flow-Geotherms . . . . . . . . . . . 5.2.5 Comparison with Cratons Globally . . . . . . . . . . . . . . . 5.2.6 Secular Variation in Thermal Structure of Cratons . . . .

155 156 156 156 158 160 160 162 163

. . . . . . . . .

Contents

xv

5.3 5.4

Lithosphere Thickness from Heat Flow and Xenolith Data . . . . . . Thermal Evolution of the South Indian Cratons . . . . . . . . . . . . . . 5.4.1 Thermal Peculiarities of WDC . . . . . . . . . . . . . . . . . . . 5.4.2 Thermal Characteristics of EDC . . . . . . . . . . . . . . . . . . 5.5 Temporal Variation in Thermal Structure . . . . . . . . . . . . . . . . . . 5.6 Heat Flow Variation among the South Indian Cratons . . . . . . . . . 5.7 Global Heat Flow Observations . . . . . . . . . . . . . . . . . . . . . . . . . 5.8 Heat Flow along Craton Margins . . . . . . . . . . . . . . . . . . . . . . . . 5.9 Influence of Composition on Heat Generation . . . . . . . . . . . . . . . 5.10 Transient Thermal Regime of Deccan Traps . . . . . . . . . . . . . . . . 5.11 Contribution to Heat Flow: Xenolith Estimates . . . . . . . . . . . . . . 5.12 Inferences from Other Parts of the World . . . . . . . . . . . . . . . . . . 5.13 Thermal Perturbation: An Effect of Crustal Thickening . . . . . . . . 5.14 Causes of Excess Heat and Estimates of Crustal Accretion . . . . . . 5.15 Overall Heat Production of the Crust . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

164 166 166 167 168 168 170 171 171 173 173 175 176 177 178 179

Integration of Insights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Assessment of Crustal Structure: Controlling Factors . . . . . . . . . 6.2.1 Western Dharwar Craton: A Synthesis . . . . . . . . . . . . . 6.2.2 Eastern Dharwar Craton . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 Aravalli Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4 Bundelkhand Craton . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.5 Bastar Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.6 Singhbhum Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.7 Pericratonic Region . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.8 Anomalous Crust of the Off-Shore Region . . . . . . . . . . 6.2.9 Nature of the Subcontinental Moho . . . . . . . . . . . . . . . 6.3 Thermal Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.1 Western Dharwar and Bundelkhand Cratons . . . . . . . . 6.3.2 Eastern Dharwar Craton and the Pericratonic Region . . 6.4 Tectonomagmatic Accretions . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Types and Causes of Crustal Thickening: WDC . . . . . . 6.4.2 Accretionary Processes Beneath the CITZ . . . . . . . . . . 6.4.3 Composition of the Lower Crust of the EDC and Bastar Craton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.4 Lower Crust of the Pericratonic Terrain . . . . . . . . . . . . 6.4.5 Origin of the Lower Crustal Granulites . . . . . . . . . . . . 6.4.6 Evolution of the Lower Crust . . . . . . . . . . . . . . . . . . . 6.4.7 Formation of the Felsic Crust: Recycling Mechanism . . 6.5 Evolution of the Lithosphere . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.1 Lithosphere Studies in India . . . . . . . . . . . . . . . . . . . . 6.5.2 Composition of SCLM: General Characteristics . . . . . .

. . . . . . . . . . . . . . . . . .

183 184 184 185 187 188 188 188 189 189 189 190 192 193 195 197 197 200

. . . . . . . .

203 204 204 206 208 210 210 211

6

xvi

Contents

6.5.3 Nature of SCLM Beneath Western India . . . . . . . . . . . . 6.5.4 SCLM Beneath the EDC . . . . . . . . . . . . . . . . . . . . . . . 6.5.5 SCLM Beneath the Bastar Craton . . . . . . . . . . . . . . . . . 6.5.6 Thickness of CBL and Compositional Variation . . . . . . . 6.5.7 Thickness of TBL . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.8 Lithospheric Mantle: Lateral Variation . . . . . . . . . . . . . . 6.5.9 Metasomatic Modification of the Lithospheric Mantle . . . 6.5.10 Continental Margins: Lithosphere Architecture . . . . . . . . 6.6 Summary and Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

213 220 225 227 228 230 232 234 237 243

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255

Abbreviations

Adr Alm Amp Apt BXP CBL Cbn CITZ CMB Cpx Cr-spnl Crt Dun Eclo EDC EGMB En Fo Fs Gar. Gran Gar. Lher Glim Gnt Grs Harz Ilm Kn LAB LC

Andradite Almandine Amphibole Apatite Between cross polarisers Chemical boundary layer Carbonate Central Indian tectonic zone Crust-mantle boundary Clinopyroxene Chrome-spinel Chromite Dunite Eclogite Eastern Dharwar Craton Eastern Ghats Mobile Belt Enstatite Forsterite Ferrosilite Garnet granulite Garnet lherzolite Glimmerite Garnet Grossular Harzburgite Ilmenite Knorringite Lithosphere-Asthenosphere boundary Lower crust

xvii

xviii

Lher LVZ Mar MBL MC Ol Opx PCS Phlg Plg PPL Prp Pyr RBL SCLM SGT Spd Spnl Sps TBL Ti-mag Tr UC Uv WDC Web Weh Wo

Abbreviations

Lherzolite Low-velocity zone MARID (Mica-Amphibole-Rutile-Ilmenite-Diopside) Mechanical boundary layer Middle crust Olivine Orthopyroxene Palghat-Cauvery shear Phlogopite Plagioclase Plain polarised light Pyrope Pyroxenite Rheological boundary layer Subcontinental lithospheric mantle Southern granulite terrain Sulphide Spinel Spessartite Thermal boundary layer Ti-magnetite Trace Upper crust Uvarovite Western Dharwar Craton Websterite Wehrlite Wollastonite

Chapter 1

Continental Lithosphere

Abstract The chapter offers a concise introduction to the constitution of the lithosphere. In the course of this discussion, the conceptual subdivisions of the lithosphere and their characteristic features are elucidated. The seismic structure and characteristics of the two main crustal types and their tectonic settings are highlighted. The significance of the more important Moho types and their interrelationship in deep continental studies is underscored. The subcontinental lithospheric mantle (SCLM) is the non-convective part of the upper mantle. The mineralogical and chemical composition of the SCLM is described and its role in the evolution of the crust and the subjacent mantle is summarised. The later part of the chapter unfolds the salient features of the methods and techniques employed to study the lithosphere. While the efficacy of the various geophysical methods, employed to study the characteristics of the deep crust and mantle, is highlighted, the limitations are also brought out into discussion. Likewise, the characteristics of the different types of seismic waves, their velocity variations, the factors that influence their propagation and which have a bearing on the interpretation of the characteristics of the deep lithologies are outlined. In addition, the chapter includes the salient features of the electrical and magnetotelluric methods used to study the deep interior. Finally the surface heat flow and heat production measurements, and their advantages and limitations are outlined. Keywords Lithosphere · Deep crust · Petrological Moho · Lithosphereasthenosphere boundary · Tectosphere · Xenolith

1.1

Introduction

The term ‘Lithosphere’ is etymologically derived from two Greek words lithos meaning ‘rocky’ and spaira meaning ‘sphere’, respectively. It is the outer hard and rigid layer of the earth about 50–300 km thick, which behaves elastically on the scale of thousands of years or more and deforms through brittle failure. It comprises the crust and a part of the non-convecting upper mantle known as the © Springer Nature Switzerland AG 2021 A. G. Dessai, The Lithosphere Beneath the Indian Shield, Modern Approaches in Solid Earth Sciences 20, https://doi.org/10.1007/978-3-030-52942-0_1

1

2

1 Continental Lithosphere

‘lithospheric mantle’ (Fig. 1.1). The lithosphere grades downwards into a highly viscous, deformable part of the mantle, to which Barrel (1914) proposed the term ‘Asthenosphere’ (from Greek asthenes meaning ‘weak’). The boundary zone between the lithosphere and the underlying asthenosphere in some regions occurs at a depth of about 100–150 km (Gutenberg 1954) and is referred to as the ‘Lithosphere-Asthenosphere Boundary’ (LAB). Geophysical studies have shown that between ~35 and 120 km depth region seismic waves show an abrupt 5–10% decrease in wave speeds forming a boundary zone known as the Gutenberg discontinuity. In many cases, the depth to the Gutenberg discontinuity approximately coincides with the depth of the LAB, leading to the suggestion that the two boundaries are closely interrelated. The actual depth of the LAB is a matter of debate, and it is known to vary depending on the environment. However, broadly the LAB is a low seismic shear wave velocity region, commonly referred to as the ‘Low-Velocity Zone’ (LVZ) which occurs between about 80 and 300 km depth. This depth range also corresponds to anomalously high electrical conductivity. Several factors such as temperature, presence of melt, water, mineral grain size among others may be responsible for the low velocity of this zone. The top of the LVZ almost coincides with the bottom of the lithospheric mantle, while its base, known as the Lehmann discontinuity (named after its discoverer, the Danish lady seismologist, Inge Lehmann 1961, 1962), occurs at a depth of ca. 220 km. This thin, yet strong layer of variable thickness (~150–450 km) commonly referred to as the ‘Subcontinental Lithospheric Mantle’ (SCLM), also known as the ‘Tectosphere’ (Jordan 1975) or the ‘mantle lid’, plays a crucial role in the evolution of both, the continental crust and subjacent upper mantle, both of which are cogenetic in nature. The SCLM is a transition zone of anisotropy characterised by

Fig. 1.1 Cartoon of the schematic section through the earth (not to scale) depicting the various layers (left). Magnified portion (right) of the crust-mantle and lithosphere-asthenosphere boundaries (modified from literature)

1.2 Geochemical Differentiation of the Earth

3

a distinctive change in mineralogical and chemical composition and transmits heat by conduction in comparison to the layer beneath, which maintains its homogeneity by a convective mechanism. The SCLM being largely rigid and non-convective has by and large retained its ancient characteristics beneath many cratons globally and thus enables to obtain an insight into the nature of the primitive lithosphere of the earth. Understanding the evolution of the continental lithosphere and getting an insight into the lithosphere growth processes has been the quest of the geological fraternity the world over since it preserves a treasure of information on the evolution of the planet. The topic has received wide attention from the geological community, with hundreds of papers and several books devoted to the topic (the reader is referred to the following general references for further reading: Taylor and McLennan (1985); Windley (1995); Condie (1997); Artemieva (2011)). The knowledge of its physicochemical composition, thermal structure, and the mechanism of lithosphere evolution are crucial for understanding the processes responsible for the formation of early lithosphere and its subsequent modification by interaction with the asthenospheric mantle and plate tectonics. With the exception of few, most studies being indirect, the information obtained through the geophysical or geochemical methods is dependent on the composition, its physical state, temperature, and pressure among others. This necessitates a multidisciplinary approach to analyse and synthesise the data obtained through various techniques such as petrological, seismic, gravity, electromagnetic, thermal among others, as they are all interrelated and aimed at probing the deep interior. The multiplicity of techniques has also led to the recognition of a variety of lithosphere types resulting in some confusion in the usage of the term lithosphere (see Artemieva 2011).

1.2

Geochemical Differentiation of the Earth

It is generally agreed that the earth and other bodies in the solar system are formed by condensation and accretion of matter from the solar nebula and that the composition of the sun roughly corresponds to the composition of the nebula. Subsequent to the condensation, the matter that accreted to form the planet got heated due to collision, compression due to gravitational pull and radioactive decay. This melted all the accreted mass and caused the first geochemical differentiation of the earth into concentric layers that have varying physico-chemical characteristics. In the molten mass, the heavier elements such as iron and nickel migrated towards the centre of the earth to form the dense hot core. The lighter elements such as silicon, oxygen, aluminium, magnesium, calcium, sodium, and potassium moved towards the surface to form the less dense rigid crust. Between these two layers there exists a layer with transitional physico-chemical characteristics, such as composition and density, between those of the crust and core. This layer is known as the mantle and consists of silicate minerals made largely of silicon, magnesium, and oxygen. Subsequent to this initial differentiation process, crustal evolution in later times has been taking

4

1 Continental Lithosphere

place broadly through tectonomagmatic processes which essentially involve the recycling of matter in various forms, primarily through mass transfer to the crust via mantle-derived magmas. It is considered to be the most significant process among others, responsible for the growth of lithosphere in general, and that of the crust in particular.

1.2.1

Bimodality of the Crust

In recent years, satellite data collected from various planets and the results of the lunar landings have shown that earth’s crust is a unique feature among the bodies of the solar system. Coupled with this, a wealth of data collected in the fields of geology, geochemistry, and geophysics in the last few decades have added much to our understanding of the physical and chemical nature of the earth’s crust and of the processes which led to its evolution. Earth is a dynamic planet in our solar system in having a bimodal topography which reflects primarily the two distinct types of crust. The oceanic crust (in older literature referred to as ‘Sima’ which stands for silica and magnesium, the most abundant constituents of oceanic crust), which is low-lying, thin, usually about 7 km in thickness, is relatively dense (~3.0 g/cm3) and is predominantly basaltic. In contrast, the continental crust (‘Sial’ refers to silica and aluminium, the most dominant constituents of continental crust) is relatively less dense (~2.7 g/cm3), hence high-standing, thicker, about 40 km on average, and is composed of a variety of lithotypes, which on an average yield an intermediate to ‘andesitic’ bulk composition (Taylor and McLennan 1985). It also contains the relicts of the oldest (4.03 Ga) rocks on earth, such as for example, the Acasta gneisses (Bowring and Williams 1999) from north Canada among others and thus preserves the record of the geological history of our planet’s evolution. The oldest rock fragment, however, ever recorded on earth, is a zircon from Jack Hills, Yilgarn Craton, Australia, which has provided an age of 4.375
6 Ga in comparison with the age of the earth which is 4.6 Ga. A third crustal division, transitional between the continental and oceanic crust, sometimes called the ‘Conrad discontinuity’ also exists, that will be discussed later. Understanding the origin of continental crust is critical to get an insight into the origin and differentiation of the planet.

1.2.1.1

The Continental Crust

The continental crust is the outer rigid shell of the earth about 40–50 km thick. Genetically, the crust is the differentiated by-product of primitive mantle of which the present-day upper mantle is the residue. Thus, the crust grades downwards into the non-convecting part of the upper mantle. The relationship between the formation of the continental crust and lithospheric mantle beneath is, however, a very complex liaison. It is a multistage process which involves generation and differentiation of

1.2 Geochemical Differentiation of the Earth

5

mafic magmas through a series of complex processes that includes recycling of continental crust to the convecting mantle at subduction zones. The process of crustal growth and the formation of lithospheric mantle being essentially the same, broadly, the composition of the average continental crust is complementary to that of the lithospheric mantle. Despite these broad similarities, neither the composition of the Archaean crust nor the composition of the mantle is well understood. The continental crust is that region which extends vertically from the earth’s surface to the Mohorovičić discontinuity (commonly referred to as Moho or Mdiscontinuity) which is a conventional boundary between the crust and the mantle defined by contrast in seismic velocity. The Moho is a seismic discontinuity, named after the Croatian lady meteorologist and seismologist Andrija Mohorovičić who identified it in 1909 beneath the European continent from the analysis of P-wave velocity data. Moho is regarded as a seismic discontinuity where a jump in compressional seismic (P) wave velocity (Vp) from 7 to 8 km/s is noticed (Bott 1982). As per the original definition, the Moho was specified as a sharp discontinuity, however, later studies have shown that the Moho or the crust-mantle boundary is a transition zone which may vary in thickness from 5–15 km where a strong positive Vp gradient is noticed. This zone therefore, marks the transition from the crust to the upper mantle. The lowest Vp expected in the mantle is 7.6 km/s; however, Vp of 7.8 is commonly encountered. Conventionally therefore, a layer with Vp > 7.6–7.8 km/s is attributed to the mantle. In terms of shear wave velocities (Vs) those greater than 4.3 km/s define the mantle (Meissner and Weaver 1989). In most crustal sections, the Moho is interpreted as a compositional boundary which delineates the crust from the mantle and is related to the planetary differentiation process. In many tectonic settings, such as in rift zones, Vp greater than 7.6 km/s is ascribed to the mantle. This leads to some confusion in data interpretation. The very high-velocity lower crust can be interpreted at times as the mantle or conversely very low-velocity uppermost mantle can be interpreted as the lower crust. The seismic Moho by definition does not specify either the crust or the mantle in terms of rock types but is merely dependent on a distinct change in seismic velocities as explained above. However, geophysical models which are simplistic commonly specify that the lower crust is andesitic and is underlain by a peridotitic upper mantle. With the improvement in geophysical techniques and data acquisition, different types of Mohos are recognised by geophysicists such as the traditional refraction Moho, the transitional Moho (Finlayson et al. 1979), the petrologic Moho (Bott 1982), and the magnetic Moho (Wasilewski et al. 1979; Mayhew and Johnson 1987). Such usage could further complicate the essential petrologic distinction between the crust and the mantle. In recent literature, it is customary to recognise the ‘seismic Moho’ which occurs at the bottom of a sequence of mafic granulites interlayered by felsic meta-igneous rocks. In such regions, the Moho appears to be a zone of transition, 5–20 km thick, in which mafic rocks are interlayered by other lithologies. The proportion of interlayered felsic rocks increases upwards whereas that of the ultramafic rocks increases downwards. The seismic Moho lies near the bottom of this zone much

6

1 Continental Lithosphere

below the petrological Moho. It is therefore necessary to know the physical properties of deep-seated rocks to place realistic constraints on geophysical models that often infer the nature of deep lithologies at times without adequate petrologic inputs. Yet, whether the Moho always represents the crust-mantle boundary is a matter of debate to which we shall return later.

1.2.1.2

Seismic Structure

As mentioned earlier, the continental crust generally ranges from 30–50 km in thickness; however, there are regions in continental settings of varied age and tectonic evolution where considerable variation in thickness from 20 to over 80 km is observed. More specifically, the thickness of the continental crust varies with the tectonic setting. In shields and platforms, it is ~38–40 km thick, in rifts it is ~32 km, and in continental margins-continental transition zones it is ~30 km. At places such as the western Indian continental margin, it is less than even 15 km, whereas in continent-continent collisional zones, such as beneath the Himalaya and Andes it is the largest, where crustal roots extend down to a depth of ~70 km. The crust displays different physico-chemical properties with depth and hence, it is differentiated into different layers. Traditionally, four layers are recognised within the continental crust (Fig. 1.2). The surface layer which generally consists of sedimentary rocks is commonly defined by Vp of 250 km) limitation on xenoliths sampled. This places limits on the interpretation of the maximum depth extent of the mantle. Similarly, the thermal state provided by the xenoliths data is usually representative of the time of eruption rather than the thermal state ‘normal’ to the continental interiors that may have existed prior to the eruption time. Spatially, localised distribution of intrusive magmas, particularly along weakness zones, also limits the efficacy of sampling. Despite limitations, xenoliths undisputedly represent the only direct samples of the deep lithologies for which there is no other substitute. The mineralogy remains largely unmodified, the ambient temperature-pressure conditions at depth are also frozen-in due to their rapid transport to the surface. Complimentary information on the deep crust provided by ophiolitic peridotites and by exhumed granulites terrains from the Archaean regions of the world, aids greatly in interpretation. This information coupled with seismic, heat flow, and other geophysical data helps constrain the composition and structure of the deep interior in an adequate manner.

1.4.5

Thermal Structure of the Crust

Among several approaches adopted to study the thermal state of the crustal rocks, two methods among others, are commonly employed to estimate the temperature at depth. One requires the determination of temperature, and pressure estimates, on

40

1 Continental Lithosphere

contemporaneous xenoliths brought up by fast erupting rocks. The other involves the extrapolation of present-day surface heat flow data to the deeper parts of the crust. Both methods have several limitations.

1.4.5.1

Thermobarometry

Boyd (1973) for the first time employed the application of mineral geothermobarometry to constrain ambient pressure-temperature conditions of mantle rocks from xenoliths. The idea was further developed by many others, including O’Neill and Wood (1979), Finnerty and Boyd (1984), Brey and Kohler (1990) among others. The principle of thermobarometry is based on the dependence of the activity of exchange reactions (element diffusion) between coexisting (equilibrated) minerals on the temperature and pressure. As mantle rocks are characterised by a restricted mineralogy (such as olivine, orthopyroxene, clinopyroxene, spinel, and garnet) the exchange reactions are confined to a limited number of phases. The reactions are experimentally calibrated for exchange equilibria. However, experimentally measured values of diffusion rates for cations in mantle peridotites vary greatly at any given temperature and are not well constrained for certain cations, e.g. Al, Cr, and some others. Reliable geothermometers should comply with the most fundamental assumption that the equilibrium has been attained in the mineral phases in terms of major elements. Secondly, the P-T array should satisfy the petrological constraints imposed by experimentally calibrated phase relations. Geothermobarometers offer an opportunity of estimating the equilibration temperature and pressure (i.e. depth) of individual mineral grains in deep crustal xenoliths at the time of their entrainment in the host magma. The P-T array of xenoliths so determined is referred to as xenoliths geotherm. The latter is able to provide an estimate of the rate of temperature increase in deep crust and its thermal thickness at the time of emplacement of the xenoliths in the carrying intrusive rock. The xenoliths P-T data are commonly compared with a set of conductive geotherms of Pollock and Chapman (1977) which offer a reference frame for comparing the deep thermal regime of different tectonic settings. The conductive geotherms are based on the assumption that 60% of the surface heat flow is derived from the mantle. However, it must be noted that average crustal heat production decreases with age of the crustal block in question and is lowest in Archaean crust (e. g. Rudnick et al. 1998) where barely 30% of the heat flow is generated. Geochemical studies of deep crustal xenoliths from different tectonic settings provide information on the chemistry of the deep crust. In most regions, the deep crust exhibits significant variation in composition with depth. In cratonic environments, the low temperature xenolith assemblage is often associated with a change in major element composition with siliceous (acidic, felsic) rocks which have a higher concentration of heat-producing elements (K, Th, U). The high temperature assemblage from deeper levels is characteristic of basic (mafic) rocks which are relatively depleted in heat-producing elements. The boundary between the felsic and mafic lithologies is usually defined by the Conrad discontinuity that usually separates the

1.4 Methods and Techniques

41

middle crust from the lower crust. There are instances, however, where acidic and basic rocks coexist in the deep crust.

1.4.5.2

Limitations

Interpretation of P-T estimates obtained from xenoliths has been controversial. The cooling of the lithosphere after thermal perturbation is relatively rapid, with time constants according to some of 10 Ma (e.g. Sass and Lachenbruch 1979), whereas according to others over 100 Ma (e.g. Chapman and Furlong 1992). It has been suggested that the exchange of elements between the silicates from xenoliths might be ‘frozen in’ at some point during their cooling history. The temperature and pressure calculated from the analyses of these phases would then bear no direct relationship to the ambient geotherm at the time the xenoliths entrainment (Harte et al. 1981; Frazer and Lawless 1978; Harte and Freer 1982). This effect is clearly important at low temperatures; the most obvious evidence is the preservation in surface granulite outcrops of apparent palaeotemperatures of the order of 700–800  C. Similarly, the preservation of symplectites, coronas, and other reaction microstructures in many xenoliths demonstrates that equilibrium has not been maintained during cooling, such samples are not suitable for thermobarometry. For intercrystalline cation, distribution to be ‘frozen in’ the cooling rate should be greater than the two-way diffusion of, for example, Fe and Mg. It has been shown (e. g. Harte and Freer 1982) that measured diffusion rates are so slow that such disequilibrium is likely. However, cooling that is rapid relative to diffusion will result in the creation of compositional zoning towards interphase grain boundaries. The minerals of xenoliths selected for thermobarometry therefore, should be checked for homogeneity especially near their rims. If complete homogeneity is established, the xenoliths may be assumed to have been in equilibrium at the ambient P-T at the time of entrainment. Within the accuracy of the various thermobarometers, such xenoliths may give points on the ambient geotherm. The approach to evaluate the thermal state and the depth regimes of crustal rocks is via the geothermobarometry of lower crustal xenoliths. This approach is not less problematic, than that of upper mantle xenoliths, which is often controversial. The problem arises partly from the low temperatures involved and partly from the general scarcity of low-variance mineral assemblages in the dominant rock types. The most commonly used geothermometers are based on pyroxene solvus (Wood and Banno 1973; Wells 1977; Lindsley et al. 1981) and the Fe/Mg partitioning between garnet and clinopyroxene or orthopyroxene (Ellis and Green 1979; Harley and Green 1982; Harley 1984). The pyroxene solvus thermometers are hampered by the insensitivity of pyroxene compositions to temperatures below about 900  C. The temperature calculations are also beset with difficulties. For instance, the garnet-pyroxene temperatures are very dependent on estimation of Fe3/Fe+2 in pyroxenes. The calculation of this ratio from microprobe data commonly leads to overestimation of Fe+3 and underestimation of temperature, whereas the common assumption that all Fe ¼ FeO

42

1 Continental Lithosphere

is equally erroneous. Single-crystal X-ray diffraction studies can improve this situation by providing more precise estimates of Fe+3 (Griffin et al. 1984). Pressure is commonly estimated by the partition of Al between orthopyroxene and garnet (Wood 1974; Harley and Green 1982); or of Ca between garnet and plagioclase in the assemblage garnet + plagioclase + quartz + Al2SiO5. However, the common difficulty encountered is garnet-bearing assemblages are found in a minority of localities and aluminosilicates are even rarer. Temperature and pressure estimates obtained for the same rock by different geothermobarometric methods may provide different results of pressure and temperature. However, the results can be checked for consistency against experimental studies on phase equilibria in systems of similar composition. The accuracy of results within the range of P-T used in experimental calibration is fairly good and the interpretation of results could be relied upon.

1.4.6

Heat Flow Studies

The earth can be viewed as a massive heat engine, with various energy sources and sinks. Insights into its evolution can be obtained by quantifying the various energy contributions in the context of the overall energy budget. Temperature distribution within the earth provides crucial information on its evolution through geological history by enabling to place constraints on the process of planetary accretion and differentiation. Heat energy that the earth receives from the sun is about two orders of magnitude greater than the heat loss from the interior of the earth. Most of the solar energy is reradiated and the temperature inside the earth is controlled by the internal heat. About 80% of the heat from the deep interior comes from radiogenic heat production, the remaining 20% is due to the secular cooling of the earth which includes latent heat from the solidification of the core. Physical properties of the crustal and mantle rocks determined by geophysical techniques are temperature dependent and hence information on the thermal regimes of the crust and mantle are crucial in the interpretation of geophysical data. For example, seismic velocity, mantle viscosity (on which depends the vigour of mantle convection), electrical conductivity, and density of rocks are all temperature dependent.

1.4.6.1

Heat Generation

In the crust, heat is largely generated either due to radioactive decay or advection. The radiogenic heat production of rock may be calculated from the concentration of K, U, and Th concentration in the rocks. The heat productivity from radioactive decay of long-lived isotopes such as 232Th, 238U, 40K, and 235U are listed in order of decreasing importance for the present-day thermal balance. The relative abundance of radioactive isotopes in the earth, K:U:Th is 10,000:1:3.7. The heat generated

1.4 Methods and Techniques

43

provides the internal energy and largely defines the temperatures in the interior. The concentration of these isotopes can be determined by counting the natural radioisotopes with a gamma-ray spectrometer. Radiogenic heat generation (A) is expressed in 10–13 cal/cm3 s or as μW/m3. One heat generation unit (1HGU) is defined as 10–13 cal/cm3 s and is equivalent to 0.0239 μW/m3. The concentration and the depth distribution of radiogenic elements in the crust and mantle rocks are not satisfactorily known. Despite this, information on radioactive heat generation in the crust is crucial for thermal modelling since it contributes a significant quantity of heat to the surface heat flow. Most estimates suggest that the contribution from crustal heat production can vary from less than 20% to over 80%. Several different attempts and approaches have been adopted to address these questions without any single technique being satisfactory for estimating the crustal radiogenic heat production. Therefore, in majority of thermal studies the distribution of crustal heat production remains more of an assumption.

1.4.6.2

Mechanism of Heat Transfer

Primarily there are three mechanisms by which heat is transferred from the interior of the earth to the surface. These are conduction, convection, and advection. The fourth transfer mechanism, namely radiation, is minor. As far as crustal environments are concerned the transfer of heat by conduction is one of the prime mechanisms of heat transfer. Conduction, also called as diffusion, is the direct microscopic exchange of kinetic energy of particles through the boundary between two systems. Conduction occurs if there is a temperature gradient in a solid or stationary fluid medium. In this process, energy is transferred from more energetic to less energetic molecules when neighbouring molecules collide. Heat flows in the direction of decreasing temperature since higher temperatures are associated with higher molecular energy. Temperature difference between rocks leads to heat flow from the hotter to the cooler, so that they attain the same temperature at which point they are said to be in thermal equilibrium. Such spontaneous heat transfer always occurs from a region of higher temperature to that of lower temperature. Heat flow from the deep interior is transferred upwards by mantle convection. In the upper layers of the earth, heat transfer by conduction predominates. Thus, it is the conductive heat flow that is measured in geophysics and is commonly mentioned as the ‘terrestrial heat flow’. However, locally convection and advection may also be significant for transfer of heat. For example, tectonically driven convection along mid-oceanic ridges and convection due to water circulation, especially in the oceanic crust, can be significant mechanisms of heat transfer. Similarly, convection in the mantle and in the liquid outer core plays an important role too in regulating the thermal regimes of the planet. Advection of heat takes place due to thermal diffusion in regions of mass transfer from depth, such as in the vicinity of magmatic intrusions.

44

1.4.6.3

1 Continental Lithosphere

Heat Flow Measurements

Heat flow determinations involve two types of measurements. One is of thermal gradient (dT/dx) and the other of thermal conductivity (K ). From these measurements, heat flow (q) is calculated as follows: q ¼ –K (dT/dx). It is expressed as μcal/ cm2 s or as mW/m2 where 1 μcal/cm2 s is defined as one heat flow unit (1 HFU) and 1 HFCU ¼ mW/m2. Thermal gradient is measured with thermistors. Surface heat flow in continental and oceanic crust is governed by several factors and it often decreases with time. As mentioned above, one of the prime sources of heat loss from the earth is from hydrothermal circulation. Approximately 25% of the total global heat flow could be accounted for by hydrothermal circulation (Davis 1980). Heat flow of continental areas varies as a function of the age of the last magmatic event, the distribution of heat-producing elements, and the level of erosion. Accounting for all sources of heat loss, the total heat loss from the earth today is about 1013 cal/s which consists of 3  1012 cal/s from the continents and 7  1012 cal/s from the ocean basins (Sclater et al. 1980). Heat flow models indicate that 70% of the average surface heat flow is lost from convection in the mantle, about 20% by conduction, and the remainder by radioactive decay of heat-producing elements in the crust. Precambrian shields display the lowest continental heat flow values varying between 0.9 and 1.1 HFU. The plat formal areas lie between 1 and 2 HFU. The difference in heat flow between the shields and the platforms is most likely because the measurements in the latter case are from areas with a thicker crust. Heat flow measurements from continental rift systems are high and variable. High heat flow is generally associated with thin crust, low P-wave velocities, shallow depth to the zone of seismic low-velocity, low q and high electrical conductivity in the upper mantle. Current total heat flow at the earth’s surface- 46
3 (1012 J/s1)-involves contributions from heat entering the mantle from the core, as well as mantle cooling, radiogenic heating of the mantle from the decay of radioactive elements, in addition to various other minor extraneous processes including gravitational heating. Surface heat flow measurements are useful to place constraints on the nature of underlying crust. Heat production indirectly provides an approximate measure of the thickness and geochemistry of the lithologies at depth.

1.4.6.4

Constraints

Heat flow measurements are not without limitations. The extrapolation of heat flow is strongly model dependent. For example, in cratonic areas with low heat flow of about 30–40 mW/m2, conductive heat transport models imply temperatures as low as 400  C at the base of the crust. However, if same models are applied to high heat flow areas (80–100 mWm2), such as continental rifts, it would be seen that the temperatures encountered would be sufficient to imply complete melting of the crustal rocks at a shallow depth. In such areas, contribution from the advective

References

45

heat transport to the regional heat flow is important. However, the large number of parameters involved in advective heat transport makes it difficult to model the process unambiguously (for discussion, refer Sass and Lachenbruch 1979). Measurements of in situ temperature distribution are usually done in bore holes, where the conditions are disturbed by drilling. Measurements therefore, need to be taken when the temperature distribution equilibrates to the steady state. This requirement is at times not fulfilled consequently affecting the quality of heat flow data (Jaupart et al. 1982). Groundwater circulation, particularly in shallow bore holes, considerably affects the temperature measurements, affecting the quality of data. Temperature variation associated with palaeoclimate changes, such as glaciationaffected areas in the northern latitudes of America and Europe, may affect the temperature distribution and gradients at shallow crustal levels impacting heat flow measurements in continental areas. The other factor which affects the heat flow is the thermal conductivity measurements. When conductivity measurements are done in the laboratory, particularly on low-porosity crystalline rocks, they tend to provide values lower than those undertaken in in situ conditions.

1.4.6.5

Reduced Heat Flow

As seen above, heat flow has contributions both from the mantle and the crust. The difference between the surface heat flow and the heat generated within the upper crustal layer enriched in heat-producing elements is referred to as Reduced Heat Flow and is represented by Qr. Thus, Qr gives a rough approximation of the nonradiogenic component of the heat flow. It is at times equated with heat flow at the Moho, which means the mantle heat flow. However, this is true for certain specific assumptions on depth variation of heat production. By and large, Qr does not include heat generated by radioactive decay in the middle and lower crust. The latter can amount to 10–15 mW/m2. An analysis of heat flow–heat production relationship shows a trend of increasing Qr with age from the Archaean through the Proterozoic to Palaeozoic regions. A decrease in Qr with age implies a decrease in mantle heat flow beneath the ancient continental regions.

References Agarwal BNP, Thakur NK, Negi JG (1992) Magsat data and Curie depth below Deccan Flood Basalt (India). Pageoph 138:678–691 Alaard O, Griffin WL, Lorand JP, Jackson SE, O’Reilly SY (2000) Non-chondritic distribution of the highly siderophile elements in mantle sulphides. Nature 407:891–894 Ammon CJ, Randall GE, Zandt G (1990) On the non-uniqueness of receiver function inversions. J Geophys Res 95:15303–15318 Anderson DL (1995) Lithosphere, asthenosphere, and perisphere. Rev Geophys 33:125–149

46

1 Continental Lithosphere

Artemieva IM (1996) The dependence of transport properties of in situ rocks on pore fluid composition and temperature. Surv Geophys 17:289–306 Artemieva IM (2011) The lithosphere. Cambridge University Press, Cambridge, p 773 Atchuta Rao D, Ram Bau HV, Sivakumar Sinha GDJ (1992) Crustal structure associated with Gondwana graben across the Narmada-Son lineament in India: an inference from aeromagnetics. Tectonophysics 212:163–172 Barrel J (1914) The strength of the Earth’s crust. J Geol 22:28–48 Barruol G, Kern H (1996) Seismic anisotropy and the shear-wave splitting in lower crustal and upper mantle rocks from the Ivrea zone- an experimental and calculated data. Phys Earth Planet Inter 95:175–194 Barton G (1986) The relationship between seismic velocity and density in the continental crust- a useful constraint? Geophys J R Astron Soc 87:195–208 Bass JD, Anderson DL (1984) Composition of the upper mantle: geophysical test of two petrological models. Geophys Res Lett 11:229–232 Bell DR, Rossman GR (1992) Water in the earth’s mantle: the role of nominally anhydrous minerals. Science 255:1391–1397 Berckhemer H, Kapfmann W, Aulbach E, Schmeling H (1982) Shear modulis and Q of forsterite and dunite near partial melting from forced oscillation experiments. Phys Earth Planet Inter 29:30–41 Bodine JH, Steckler MS, Watts AB (1981) Observation of flexture and the rheology of the oceanic lithosphere. J Geophys Res 86:3695–3707 Bott MHP (1982) The mechanism of continental splitting. Tectonophysics 81:301–309 Bowring SA, Williams IS (1999) Priscoan (4.00–4.03) orthogneisses from northwestern Canada. Contrib Mineral Petrol 134:3–16 Boyd FR (1973) A pyroxene geotherm. Geochim Cosmochim Acta 37:2533–2546 Boyd FR (1989) Compositional distinction between oceanic and cratonic lithosphere. Earth Planet Sci Lett 96:15–26 Brey GP, Kohler T (1990) Geothermobarometry in four phase lherzolite II: new thermobarometers and practical assessment of existing thermobarometers. J Petrol 27:1353–1358 Chapman DS, Furlong KP (1992) Thermal state of the continental lower crust. In: Fountain DM, Arculus RJ, Kay RW (eds) The continental crust. Elsevier, Amsterdam, pp 129–199 Christensen I (1965) Compressional wave velocities in metamorphic rocks at pressures to 10 kbars. J Geophys Res 70:6147–6164 Christensen I, Fountain DM (1975) Constitution of the lower continental crust based on experimental studies of seismic velocities in granulite. Geol Soc Am Bull 86:227–236 Christensen I, Mooney WD (1995) Seismic velocity structure and composition of the continental crust: a global view. J Geophys Res 100:9761–9788 Clitheroe G, Gudmundsson O, Kennett BLN (2000) The crustal thickness of Australia. J Geophys Res 105:13697–13713 Condie KC (1997) Plate tectonics and crustal evolution. Butterworth Heinemann, Oxford, p 282 Constable S, Shankland TJ, Duba A (1992) The electrical conductivity of an isotropic olivine mantle. J Geophys Res 97:3397–3404 Cooper CM, Lenardic A, Moresi L (2004) The thermal structure of stable continental lithosphere within a dynamic mantle. Earth Panet Sci Lett 222:807–817 Crampin S (1977) A review of the effects of anisotropy layering on the propagation of seismic waves. Geophys J R Astron Soc 45:9–27 Crampin S (1987) Geological and industrial implications of extensive-dilatancy anisotropy. Nature 328:491–496 Dahl-Jensen T, Larsen TB, Woelbern I, Bach T (2003) Depth to Moho in Greenland: receiver function analysis suggests two Proterozoic blocks in Greenland. Earth Planet Sci Lett 205:379– 393 Davis GF (1980) Review of oceanic and global heat flow estimates. Rev Geophys Space Phys 18:718–722

References

47

Dessai AG (1985) Ultramafic xenoliths(?) in lamprophyre dykes from Murud-Janjira, Raigarh district. Maharashtra Curr Sci 54:1235–1238 Dessai AG (1987) Geochemistry and petrology of xenolith bearing alkaline lamprophyres from Murud-Janjira, Raigad district. Maharashtra J Geol Soc Ind 30:61–71 Dessai AG, Vaselli O (1999) Petrology and geochemistry of xenoliths in lamprophyres from the Deccan Traps: implications for the nature of the deep crust boundary in western India. Min Mag 63:703–722 Dessai AG, Knight K, Vaselli O (1999) Thermal structure of the lithosphere beneath the Deccan Traps along the western continental margin: evidence from xenoliths data. J Geol Soc India 54:585–598 Dessai AG, Markwick A, Vaselli O, Downes H (2004) Granulite and pyroxenite xenoliths from the Deccan trap: insight into the nature and composition of the lower lithosphere beneath cratonic India. Lithos 78:263–290 Dessai AG, Peinado M, Gokarn SG, Downes H (2009) Structure of the deep crust beneath the central Indian tectonic zone: an integration of geophysical and xenolith data. Gondwana Res 17:162–170 Du Frane WL, Roberts JJ, Toffelmier DA, Tyburczy JA (2005) Anisotropy of electrical conductivity in dry olivine. J Geophys Res 32:L24315. https://doi.org/10.1029/2005GL023879 Durrheim RJ, Mooney WD (1991) Archaean and proterozoic crustal evolution: evidence from crustal seismology. Geology 19:606–609 Dziewonski AM, Anderson DL (1981) Preliminary reference earth model. Phys Earth Planet Inter 25:297–356 Ellis DJ, Green DH (1979) An experimental study of the effect of Ca upon garnet-clinopyroxene FeMg exchange equilibrium. Contrib Mineral Petrol 71:13–22 Faul UH, Jackson I (2005) The seismological signature of temperature and grain size variation in the upper mantle. Earth Planet Sci Lett 234:119–134 Finlayson DM, Prodehl C, Collins CDN (1979) Explosion seismic profiles and implications for crustal evolution in southeastern Australia. Bur Miner Res J Aust Geol Geophys 4:243–252 Finnerty AA, Boyd FR (1984) Evaluation of thermobarometers for garnet peridotites. Geochim Cosmochim Acta 48:15–27 Frazer D, Lawless P (1978) Palaeogeotherms: implications of disequilibrium in garnet lherzolite xenoliths. Nature 273:220–222 Ganguly J, Battacharya PK (1987) Xenoliths in Proterozoic kimberlites from southern India: petrology and geophysical implications. In: Nixon PH (ed) Mantle xenoliths. Wiley, New York, pp 249–265 Garnero EJ, Revenaugh J, Williams Q, Lay T (1998) Ultralow velocity zone at the core-mantle boundary. Geodynamics 28:319–334 Goetze C, Kolstedt DL (1973) Laboratory study of the dislocation climb and diffusion in olivine. J Geophys Rev 18:1–75 Grad M, Bruckl CE, Majdanski M, Behm M (2009) Crustal structure of the eastern Alps and their foreland: seismic model beneath the CEL10/Alp04 profile and the tectonic implications. Geophys J Int 177:321–333 Gribb TT, Cooper RF (1998) Low-frequency shear attenuation in polycrystalline olivine: grain boundary diffusion and the physical significance of the Andrade model for viscoelastic rheology. J Geophys Res 103:27267–27279 Griffin WL, Wass SY, Hollis JD (1984) Ultramafic xenoliths from Bullenmerri and Gnotuk Maars, Victoria, Australia: petrology of a subcontinental crust–mantle transition. J Petrol 25:53–87 Gupta S, Rai SS, Prakasam KS, Srinaguesh D, Chadda RK (2003) The nature of the crust in southern India: implications for Precambrian crustal evolution. Geophys Res Lett 30(8):1419. https://doi.org/10.1029/2002GL016770 Gutenberg B (1954) Seismicity of the earth and associated phenomena. Hafner, New York, p 310

48

1 Continental Lithosphere

Haak V, Hutton VRS (1986) Electrical resistivity in the continental lower crust. In: Dawson JB, Carswell DA, Hall J, Wedepohl KH (eds) The nature of the lower continental crust, Geological Society, London, Special Publications, vol 24, pp 35–49 Hanson GN (1980) Rare earth elements in petrogenetic studies of igneous systems. Annu Rev Earth Planet Sci 8:371–406 Harley SL (1984) An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contrib Mineral Petrol 86:359–373 Harley SL (1989) The origin of granulites: a metamorphic perspective. Geol Mag 126:215–247 Harley SL, Green DH (1982) Garnet-orthopyroxene barometry for granulites and peridotites. Nature 300:697–701 Harte B, Freer R (1982) Diffusion data and their bearing on the interpretation of mantle nodules and the evolution of mantle lithosphere. Terra Cognita 2:273–275 Harte B, Jackson PM, McIntyre RM (1981) Age of mineral equilibria in granulite facies nodules from kimberlites. Nature 281:147–155 Hatton C, Gurney J (1987) Robert victor eclogites and their relation to mantle. In: Nixon PH (ed) Mantle xenoliths. Wiley, New York, pp 453–464 Hirsch LM, Shankland TJ (1993) Electrical conduction and polaron mobility in Fe-bearing olivine. Geophys J Int 114:36–44 Hirth G, Evans RL, Chave AD (2000) Comparison of continental and oceanic mantle electrical conductivity: is Archean lithosphere dry? Eochem Geophys Geosyst 1. https://doi.org/10.1029/ 2000GC000048 Hofmann AW (1997) Mantle geochemistry: the message from oceanic volcanism. Nature 385:219– 229 Ito K, Kennedy GC (1967) Melting and phase relations in a natural peridotite to 40 kilobars. Am J Sci 265:519–538 Ito E, Harris DM, Anderson AT (1983) Alteration of oceanic crust and geologic cycling of chlorine and water. Geochim Cosmochim Acta 47:1613–1624 Iyer HM, Hirahara K (eds) (1993) Seismic tomography: theory and practice. Chapman and Hall, London, p 842 Jackson I (1993) Progress in the experimental study of seismic wave attenuation. Annu Rev Earth Planet Sci 21:375–406 Jaupart C, Mann JR, Simmons G (1982) A detailed study of the distribution of heat flow and radioactivity in New Hampshire (USA). Earth Planet Sci Lett 59:267–287 Jones AG (1992) Electrical properties of the lower continental crust. In: Fountain DM, Arculus R, Kay RW (eds) Continental lower crust. Elsevier, The Netherlands, pp 81–144 Jones AG (1999) Imaging the continental upper mantle using electromagnetic methods. Lithos 48:57–80 Jordan TH (1975) The continental tectosphere. Rev Geophys Space Phys 13:1–12 Kaila KL (1988) Mapping the thickness of Deccan Trap flows in India from DSS studies and inferences about a hidden Mesozoic basin in the Narmada–Tapti region. In: Subbarao KV (ed) Deccan flood basalts, Geological Society of India Memoirs, vol 10, pp 91–116 Karato S-I (1986) Does partial melting reduce the creep strength of the upper mantle. Nature 319:309–310 Karato S-I (1992) On the Lehmann discontinuity. Geophys Res Lett 19:2255–2258 Katayama I, Jung H, Karato S (2004) New type of olivine fabric at modest water content and low stress. Geology 32:1045–1048 Kelemen PB, Hart SR, Bernstein S (1998) Silica enrichment in the continental upper mantle via melt/rock reaction. Earth Planet Sci Lett 164:387–406 Kennett BLN (2006) On seismological reference models and the perceived nature of the heterogeneity. Phys Earth Planet Inter 159:129–139 Kern H (1978) The effect of high temperature and high confining pressure on compositional wave velocities in quartz-bearing and quartz-free igneous and metamorphic rocks. Tectonophysics 44:185–203

References

49

Kern H (1993) P-wave and S-wave anisotropy and shear-wave splitting at pressure and temperature in possible mantle rocks and their relation to the rock fabric. Phys Earth Planet Inter 78:245–256 Kind R, Yuan X, Saul J, Nelson D (2002) Seismic images of crust and upper mantle beneath Tibet: evidence for Eurasia plate subduction. Science 288:1219–1221 Knittle CE, Jeanloz R (1991) Earth’s core-mantle boundary results of experiments at high pressure and temperature. Science 251:1438–1443 Kumar P, Yuan X, Kind R, Kosarev G (2005) The lithosphere-asthenosphere boundary in the Tien Shan-Karakoram region from S receiver function: evidence for continental subduction. Geophys Res Lett 32:L07305 Kumar P, Yuan S, Ravi Kumar M, Kind R, Li CX, Chadha K (2007) The rapid drift of the Indian tectonic plate. Nature 449:894–897 Kushiro I, Syono Y, Akimoto S (1968) Melting of a peridotite nodule at high pressure and high water pressures. J Geophys Res 73:6023–6029 Kuskov OL, Kronrod VA, Annersten H (2006) Inferring upper mantle temperatures from seismic and geochemical constraints: implications for Kaapvaal craton. Earth Planet Sci Lett 244:133– 154 Lee CTA (2003) Compositional variation of density and seismic velocities in natural peridotites at STP conditions: implications for seismic imaging of compositional heterogeneities in the upper mantle. J Geophys Res 108(B(9)):2441. https://doi.org/10.1029/2003JB002413 Lehmann I (1961) S waves and the structure of the upper mantle. Geophys J R Astron Soc 4:124– 138 Lehmann I (1962) The travel time of longitudinal waves of the Logan and Blanca atomic explosions and their velocities in the upper mantle. Bull Seismol Soc Am 52:519–526 Lindsley DH, Grover JE, Davidson PM (1981) The thermodynamics of the Mg2Si 2O 6-CaMgSi2O 6 join: a review and new model. In: Newton RC, Navrotsky A, Wood BJ (eds) Thermodynamics of minerals and melts, Advances in Physical Chemistry. Springer-Verlag, New York, pp 149– 175 Matsukage KN, Nishihara Y, Karato S-I (2005) Seismological signature of chemical differentiation of the earth’s upper mantle. J Geophys Res 110:B12305. https://doi.org/10.1029/2004JB003504 Mayhew MA, Johnson BD (1987) An equivalent layer magnetisation model for Australia based on Magsat data. Earth Planet Sci Lett 83:167–174 McCulloch MT, Arculus RJ, Chapell KW, Ferguson J (1982) Lower continental crust: inferences from isotopic and geochemical studies of nodules in kimberlites. Nature 300:166–169 McKenzie D, Jackson J, Priestly K (2005) Thermal structure of oceanic and continental lithosphere. Earth Planet Sci Lett 23:337–349 Meissner R, Strehlau J (1982) Limits of stress in the continental crust and their relation to depth frequency distributions of shallow earthquakes. Tectonics 1:73–89 Meissner R, Weaver T (1989) Continental crustal structure. In: James ED (ed) The encyclopaedia of solid earth geophysics. Van Nostrand Reinhold, New York, pp 79–89 Misra DC, Ravi Kumar M (2008) Geodynamics of Indian plate and Tibet: buoyant lithosphere, rapid drift and channel flow from gravity studies. Geol Soc Ind Mem 68:151–172 Montagner J-P, Guillot L (2002) Seismic anisotropy and global geodynamics. In: Karato S, Wenk H-R (eds) Plastic deformations of minerals and rocks. MSA, Washington, DC, pp 353–385 Montagner J-P, Tanimoto T (1990) Global anisotropy in the upper mantle inferred from the regionalization of phase velocities. J Geophys Res 95:4797–4819 Mooney WD, Brocher TM (1987) Coincident seismic reflection/refraction studies of the continental lithosphere: a global review. Rev Geophys 25:723–742 Mooney WD, Meissner R (1992) Multi-genetic origin of crustal reflectivity: a review of seismic reflection profiling of the continental lower crust and Moho. In: Fountain DM, Arculus R, Kay RW (eds) Continental lower crust. Elsevier, Amsterdam, pp 45–79 Mukherjee AB, Biswas S (1988) Mantle-derived spinel lherzolite xenoliths from the Deccan Volcanic Province (India): implications for the thermal structure of the lithosphere underlying the Deccan traps. J Volcanol Geotherm Res 35:269–276

50

1 Continental Lithosphere

Murase T, Kushiro I, Fuji T (1977) Compressional wave velocity in partially molten peridotite. Annual Report of the Director Geophysical Laboratory, Carnegie Inst., Washington, DC, pp 414–416 Nehru CE, Reddy AK (1989) Ultramafic xenoliths from Wajrakarur kimberlite, India. In: Ross J et al (eds) Kimberlites and related rocks, Special Publications, Geological Society, Australia, vol 14, pp 745–759 Nguuri TK, Gore G, James DE, Webb SJ, Wright C (2001) Crustal structure beneath southern Africa and its implications for the formation and evolution of the Kaapvaal and Zimbabwe cratons. Geophys Res Lett 28:2501–2504 Nicolas A, Christensen NI (1987) Formation of anisotropy in the upper mantle peridotites- a review. In: Fuchs K, Froidevaux C (eds) Composition, structure and dynamics of lithosphere-asthenosphere systems, AGU Geodynamic Series, vol 16. Wiley, New York, pp 111–123 Nixon PH, Boyd FR (1973) Petrogenesis of the granular and sheared ultrabasic nodule suite in kimberlites. In: Nixon PH (ed) Lesotho Kimberlites. Lesotho National Development Corporation, Lesotho, pp 48–56 Nolasco R, Tarits P, Filloux JH, Chave AD (1998) Magnetotelluric imaging of the Society Island hotspot. J Geophys Res 103(B12):30287–30309 Nolet G (1987) Seismic tomography. Reidel, Dordrecht, The Netherlands, p 386 Nolet G (2008) A Breviary of seismic tomography: imaging the interior of the Earth and Sun. Cambridge University Press, Cambridge, pp 339–344 Nur A (1971) Effect of stress on velocity anisotropy in rocks with cracks. J Geophys Res 76:2022– 2034 O’Neill HSC, Wood BJ (1979) An empirical study of Fe-Mg partitioning between garnet and olivine and its calibration as a geothermometer. Contrib Mineral Petrol 70:59–70 Oreshin S, Vinnik L, Peregoudov D, Roecker S (2002) Lithosphere and asthenosphere of the Tien Shan imaged by S receiver functions. Geophys Res Lett 29:1191. https://doi.org/10.1029/ 2001GL014441 Pollock CH, Chapman DH (1977) On the regional variation of heat flow, geotherms and the thickness of the lithosphere. Tectonophysics 38:279–296 Qureshy MN (1970) Relation of gravity to elevation, geology and tectonics in India. Proc. 2nd Symp. on Upper Mantle Project, NGRI, 1–20 Raval UL (2003) Interaction of mantle plume with Indian continental lithosphere since the cretaceous. Geol Soc Ind Mem 53:449–479 Reddi AGB, Mathew MP, Singh B, Naidu PS (1988) Aeromagnetic evidence of crustal structure in the granulite terrain of Tamil Nadu-Kerala. J Geol Soc India 32:368–381 Ringwood AE (1962) A model for the upper mantle. Geophys Res Lett 67:857–867 Rogers NW, Hawkesworth CJ (1982) Proterozoic age and cumulate origin of granulite xenoliths, Lesotho. Nature 299:409–413 Rudnick RL, McDonough WF, O’Connel RJ (1998) Thermal structure, thickness and composition of continental lithosphere. Chem Geol 145:395–411 Santos FAM, Soares A, Nolasco R, Rodrigues H, Luzio R, Palshin N (2003) Lithosphere conductivity structure using the CAM-1 (Lisbon-Madeira) submarine cable. Geophys J Int 155:591– 600 Sass JH, Lachenbruch AH (1979) Thermal regime of the Australian continental crust. In: McElhinny MW (ed) The earth its origin, structure and evolution. Academic Press, London, pp 311–352 Sclater JG, Jaupart C, Galson D (1980) The heat flow through oceanic and continental crust and the heat loss of the earth. Rev Geophys Space Phys 18:269–311 Shearer PM (1999) Introduction to seismology. Cambridge University Press, Cambridge, p 260 Singh BP, Rajaraman M (1990) Magsat studies over the Indian region. Proc Indian Acad Sci (Earth Planet Sci) 99:619–637 Sumino Y, Anderson OL (1982) Elastic constants in minerals. In: Carmichel RS (ed) CRC handbook of physical properties of rocks. CRC Press, Boca Raton, FL, pp 29–138

References

51

Taylor SR, McLennan SM (1985) The continental crust: its composition and evolution. Blackwell, Oxford, p 312 Ufen RJ (1952) A method of estimating the melting-point gradient in the earth’s mantle. Trans Am Geophys Union 33:893–896 Verma RK (1985) Gravity field, seismicity and tectonics of the Indian peninsula and the Himalaya. Allied Publ. Pvt. Ltd., New Delhi, p 203 Walter MJ (1999) Melting residue of fertile peridotite and the origin of cratonic lithosphere. In: Fei Y, Bertka C, Mysen BO (eds) Mantle petrology: field observations and high pressure experimentation, vol 6. Geochemical society, Houston, TX, pp 225–239 Wasilewski PJ, Thomas HH, Mayhew MA (1979) The Moho as a magnetic boundary. Geophys Res Lett 6:541–544 Wass SY, Hollis JD (1983) Crustal growth in southern Australia—evidence from lower crustal eclogitic and granulitic xenoliths. J Met Petrol 1:25–45 Wells PRA (1977) Pyroxene thermometry in simple and complex systems. Contrib Mineral Petrol 62:129–139 White RS, McKenzie D, O’Nions RK (1992) Oceanic crustal thickness from seismic measurements and rare earth element inversions. J Geophys Res 97:19683–19715 Windley BF (1995) The evolving continents, vol 489. Wiley, Chichester, New York, Brisbane, Toronto-Singapore Wittlinger G, Vergne J, Tapponnier P et al (2004) Teleseismic imaging of subducting lithosphere and Moho offsets beneath western Tibet. Earth Planet Sci Lett 221:117–130 Wood BJ (1974) Solubility of alumina in orthopyroxene coexisting with garnet. Contrib Mineral Petrol 46:1–15 Wood BJ, Banno S (1973) Garnet-orthopyroxene and orthopyroxene-clinopyroxene relationships in simple and complex systems. Contrib Mineral Petrol 42:109–121 Wyllie PJ (1995) Experimental petrology of upper mantle materials, process and products. J Geodyn 20:429–468 Xu YS, Shankland TJ, Duba AG (2000) Pressure effect on electrical conductivity of the earth’s mantle. J Geophys Res 105:27865–27875

Chapter 2

Indian Shield

Abstract The Indian shield is divisible into two blocks along the major ENE-WSW trending Central Indian Tectonic Zone (CITZ) which separates the Aravalli and Bundelkhand cratons to the north from the Singhbhum, Bastar, Western- and Eastern-Dharwar cratons to the south. The Aravalli Craton consists of 3.3–2.5 Ga gneisses, migmatites, metasediments, and amphibolites which make up the Banded Gneissic Complex. This cratonic block was stabilised by about 2.5 Ga. The Bundelkhand Craton to the east of Aravalli block was stabilised around 1.8 Ga. It consists of gneiss-greenstone enclaves within granitic plutons and associated quartz reefs and mafic dyke swarms, intruded by the Bundelkhand granite. The Singhbhum Craton to the south of CITZ consists of a core of 3.5 Ga TTG gneisses intruded by three pulses (3.4, 3.3, and 3.1 Ga) of Singhbhum granite and its equivalents. The granite contains enclaves of Older Metamorphic Group and Older Metamorphic Tonalite Gneisses. The granite is fringed by Proterozoic BIF basins. The craton was stabilised between 2.5 and 3.0 Ga. The Bastar Craton comprises 2.6–2.2 Ga granite which contains vestiges of 3.5– 3.0 Ga TTG gneisses. The supracrustals consist of conglomerates with BIF and subordinate basic and ultrabasic rocks as enclaves within granites. The EDC consists of a remobilised basement of 2.7–2.5 Ga granodioritic gneisses and granitoids which also corresponds to the stabilisation of the craton. The supracrustals consist of high-Mg, low-K mafic volcanics of tholeiitic affinity with tuffs, and minor BIF which lack true conglomerates and sedimentary intercalations. The WDC comprises 3.3–3.1 Ga TTG gneisses intruded by the Closepet Granite at 2.51 Ga. The supracrustal rocks consist of two sequences assigned to the Sargur Group (~3.3–3.1 Ga) and the greenstones of the Dharwar Supergroup (2.9–2.4 Ga). The craton was stabilised around 3.3–3.0 Ga. The WDC and the EDC are brought in contact with the Southern Granulite Terrain, primarily along the Palghat-Cauvery shear zone. Keywords Indian shield · Dharwar Craton · Singhbhum Craton · Peninsular gneiss · Greenstones · Granulites

© Springer Nature Switzerland AG 2021 A. G. Dessai, The Lithosphere Beneath the Indian Shield, Modern Approaches in Solid Earth Sciences 20, https://doi.org/10.1007/978-3-030-52942-0_2

53

54

2.1

2 Indian Shield

Introduction

The Indian shield forms a part of the Indo-Australian plate which is one of the major plates of the globe. Within this plate, the Indian plate is recognised as a minor (secondary/subplate) plate. For the sake of convenience, the boundaries of the Indian plate that are of direct relevance to the geotectonic framework of the Indian shield are described in more detail. The Indian plate is bounded in the north by a major dislocation known as the Indus-Tsangpo Suture Zone (early Cretaceous to Eocene) along which it is juxtaposed against the Tibetan block. To the east, in the eastern Himalaya, the suture is expressed as a continent-continent collisional boundary (‘Cordilleran-type’ or ‘Hercyno-type’) between India and Nepal. In contrast, in the west in the Ladakh Himalaya, it is an island-arc type (‘Alpino-type’) of margin as the Indus Suture here bifurcates and reunites further west in Pakistan. The northern bifurcation is known as the Shyok Suture, whereas the southern one is called the Indus-Tsangpo Suture. These two sutures are also known, respectively, as the Northern Suture and the Main Mantle Thrust (Searle et al. 1987). These two sutures together delimit the Dras-Kohistan Island Arc. The arc comprises a granodioritic batholith, the Ladakh Granitic Complex (late Cretaceous to early Eocene age), along with arc-volcanics, cherts, agglomerates, dunites, and serpentinites tectonically emplaced in the radiolarian cherts and limestones of late Cretaceous to early Eocene age. The eastern boundary of the Indian plate is a dynamic boundary marked by an island-arc known as the Sunda (Java) Arc which is a seismically active zone that has experienced several devastating earthquakes and recent volcanic activity on the Barren Island in the Bay of Bengal. Such active margins are known as the ‘Pacific-type’ margins; they represent potential sites for the development of future orogenic belts. The northern part of this arc is referred to as the Andaman-Nicobar island-arc. The associated Bay of Bengal back-arc is expressed as an ocean basin with an extensional setting marked by the 90
E ridge. The island-arc region consists of basalts dating from 80 Ma in the north and becoming successively younger to the south, implying that the 90
E ridge has been formed by the northward movement of the Indo-Australian plate over the Kerguelen hotspot (Fig. 2.1). The back-arc contains a thick pile of sediments that cover the northern end of the 90
E ridge. A smaller ridge is known as the 87
E ridge. The southwestern boundary of the Indo-Australian plate is defined by the Central Indian Ridge. Its northwestern extension is known as the Carlsberg Ridge. A NESW trending transform fault across the Carlsberg Ridge, known as the Owens Fracture, continues northeastwards, and can be traced on land in Pakistan, where it is known as the Chaman Fault. It delimits the northwestern boundary of the Indian plate against the Arabian plate. This fault zone is seismically active and has experienced several earthquakes in recent times. In comparison with the eastern margin, the western margin of India is one of relative tectonic stability and could be referred to as a ‘passive margin’. It is

2.1 Introduction

55

Fig. 2.1 The Indian shield depicting the kimberlite occurrences and the various hotspot tracks which have impacted it after the breakup of Gondwana until it reached its present location to the north of the Equator (modified after Mandal et al. 2018)

characterised, for the most part by a broad, low-gradient continental shelf that displays an apron of shallow water sediments. Westwards the shelf plunges steeply into the continental slope which has a thick pile of sediments known as the continental rise, gradational into the abyssal plains. Such passive margins are sites of prolonged sedimentation and are referred to as the ‘Atlantic-type’ continental margins. They are the repositories of precious mineral resources such as petroleum and natural gas. Several oil fields along the west coast, including the ‘Bombay High’, are hosted by these sediments.

56

2.2

2 Indian Shield

Geotectonic Framework

India can be divided into three geotectonic provinces. The northern part which includes the Himalaya is referred to as the Foreland Block and comprises largely the Extra-peninsular region. To the south of it lies the Indo-Gangetic Foredeep (trough) which includes the extensive plains of the Indus, Ganga, and Brahmaputra rivers. To the south of the foredeep lies the Peninsular Block which includes peninsular India. The basement of the Indo-Gangetic foredeep consists of horsts and grabens and it deepens towards the Himalaya. It is covered by the sub-recent to recent alluvial sediments that attain a thickness of more than 2 km and form the plains of the Punjab, Haryana, Uttar Pradesh, Chhattisgarh, and Bihar. Several NE-SW trending basement ridges, e.g. the Delhi ridge, Haridwar ridge, Faizabad ridge, from the Aravalli and Bundelkhand cratonic blocks abut against the foredeep. These two cratonic blocks along with the Himalayan belt constitutes the Foreland Block of the Indian subcontinent. Peninsular India, to the south of the foredeep, is a collage of six cratons variously bounded by palaeo-rifts and mobile belts. A seventh craton known as the Shillong Plateau, is an outlying entity in the NE, which according to some, forms a part of the northern cratonic block; however, it is not considered here for further discussion.

2.3

Geology of the Peninsular Block

Broadly peninsular India can be divided into two blocks (Fig. 2.2) along a major ENE-WSW trending tectonic zone which has been active since the Proterozoic and is referred to in modern literature as the Central Indian Tectonic Zone (CITZ) (Radhakrishna and Ramakrishnan 1988). In older literature, the CITZ was referred to as the Satpura mobile belt. The southern limit of this zone is marked by a suture known as the Central Indian Suture (CIS). It corresponds to the satellite imagery lineament known as the Narmada-Son lineament (Fig. 2.2) which controls the courses of the Narmada and Son rivers. The CITZ separates the Aravalli-Bundelkhand cratons in the north from the Singhbhum and Bastar cratons in the southeast and the Dharwar Craton in the south. The Bundelkhand Craton comprises the Bundelkhand granite and contains enclaves of Aravalli (early Proterozoic) and Delhi (middle to late Proterozoic) metasediments. A granitic massif from Rajasthan known as the Berach Granite (2.5 Ga) is correlated with Bundelkhand granite. The Berach-Bundelkhand granite is brought in contact with late Proterozoic to early Palaeozoic Vindhyan sediments along a NE-SW trending major fault zone referred to as the Great Boundary Fault in western Rajasthan. The CITZ comprises the Narmada-Tapti rift (Jain et al. 1995) and the intervening Satpura block which is expressed by a Bouguer anomaly interpreted to correspond to a horst-block associated with the Tapti graben in the southwest and the Narmada

2.3 Geology of the Peninsular Block

57

Fig. 2.2 Geotectonic framework of peninsular India depicting the configuration of Archaean cratons, the Deccan Traps Flood Basalt Province, gravity highs and major lineaments that represent Proterozoic and Phanerozoic continental rift zones; CUD: Cuddapah Basin, EGMB: Eastern Ghat Mobile Belt, PCS: Palghat-Cauvery Shear

graben in the south (Tewari et al. 2001). A N-S schematic section across the CITZ is depicted in Fig. 2.3 (Murthy and Misra 1981). This tectonic belt has had a more complex evolution than the peninsular shield to the south. A major portion of this belt is covered variously by alluvium, Deccan Traps and Gondwana sediments. Three E-W trending supracrustal belts, namely the Mahakoshal in the north, the Sausar belt in the south, and the Betul in between, are enclosed within the gneisses and are separated from each other by ductile shear zones. The Mahakoshal and Betul belts are separated by the Son-Narmada South Fault whereas the Betul and Sausar belts are delimited by the Tan Shear Zone. Lenticular granulite belts are associated with the shear zones and from north to south are referred to as the Makrohar belt, Ramakona-Katangi belt and Balaghat-Bhandara belt, respectively. The Precambrian terrain to the south of the CITZ comprises southern India. As per the long-standing concept the Precambrian terrain is a single entity (Fermor

58

2 Indian Shield

Fig. 2.3 N-S schematic section across the CITZ depicting the disposition of the various lithological units, F1–F5 depict the different faulting episodes witnessed by the region (modified after Murthy and Misra 1981)

1936; Narayanaswamy 1975). It comprises the granite-greenstone terrain of the Dharwar Craton in the north which is gradational into the massif charnockite terrain in the south. However, recent isotope geochronology has revealed that the Precambrian terrain of southern India can be divided into the Archaean amphibolite to granulite facies transition in the north from granulite facies massif charnockites to the south along the Palghat-Cauvery shear belt. This shear belt represents a boundary that delimits the southern terrain characterised by crustal reworking and high-grade metamorphism during the Pan-African, from the Archean terrain to the north which does not exhibit Pan-African overprint (Harris et al. 1994). The massif charnockite terrain itself comprises of several granulite blocks amalgamated at different times during the course of the evolution of the terrain (Harris and Santosh 1993; Harris et al. 1994). Broadly, the terrain to the south of the CITZ is an agglomeration of three cratonic blocks, namely the Singhbhum, Bastar, and Dharwar cratons. The last in itself consists of two blocks, the Western Dharwar Craton (WDC) and the Eastern Dharwar Craton (EDC). These cratonic blocks consist largely of the granitic gneisses, greenstones, and granulites. They are overthrust in the SE by the (NESW) Eastern Ghat Mobile Belt. The Singhbhum Craton is delimited in the SW by the NW-SE trending Mahanadi graben and in the N it is in contact with the Chotanagpur gneissic terrain along the EW arcuate Singhbhum Thrust also known as the Copper Belt Thrust or the Singhbhum Shear Zone. The craton is bounded in the south jointly by the WNWESE Sukinda Thrust and the Mahanadi graben which contains Cretaceous to Tertiary sediments. These separate the Singhbhum Craton from the Bastar Craton to the southwest, and the Eastern Ghat Mobile Belt (EGMB) in the east and southeast.

2.3 Geology of the Peninsular Block

59

The Bastar Craton is overthrust by the Eastern Ghat Mobile Belt (~1.6/1.4 Ma) along a NE-SW trending thrust. The craton is intersected in the SW by the NW-SE trending Godavari graben (with Permo-Carboniferous and Cretaceous to Cenozoic sediments) that separates it from the EDC in the west. The craton comprises middle to late Archaean granitic gneisses which are overlain by Proterozoic supracrustal rocks. The EDC is bounded by the ENE-WSW trending Eastern Ghat Mobile Belt. Its western boundary is defined by the N-S trending Closepet Granite (2.6 Ga) that forms a stitching pluton with the WDC, which is characterised by a great profusion of linear greenstone belts. Both of these cratonic blocks are juxtaposed against the southern granulite terrain primarily along the E-W arcuate Moyar-Bhavani and Palghat-Cauvery shear belt system (Drury and Holt 1980). These belts separate a structurally coherent amphibolite to granulite facies transition in the north from the charnockite massifs in the south. In the transition zone, the transformation of hornblende-biotite tonalite gneiss (Peninsular Gneiss) to charnockite is observed in patches. The massive charnockites to the south have been interpreted as a CO2-flushed terrain (Harris et al. 1982). The three crustal blocks, namely the WDC, EDC, and SGT (itself an ensemble of four blocks) have behaved as one composite unit since about 2.5 Ga and constitute one of the extensively studied regions of the Indian shield. The E-W transition zone within which the orthopyroxene ‘isograd’ is located, is called the ‘Fermor Line’. It serves to distinguish the amphibolites facies gneissic terrain to the north from the granulite facies charnockite terrain to the south. The granulites yield Nd model ages from 2.65–2.90 Ga (Peucat et al. 1989) that have led to models of a synaccretional granulite metamorphic event. The granulites from the transition zone provide protolith ages of 3.47 Ga (Nd model ages) (Jayananda and Peucat 1996). They were metamorphosed at 2.52–2.513 Ga (Friend and Nutman 1992) and were exhumed at 2.31 Ga (Jayananda and Peucat 1996) by a late shearing event.

2.3.1

Aravalli Craton

The Aravalli Craton is an areally extensive entity which consists of an amalgamation of two cratonic blocks, namely the Mewar Craton in the east and the Marwar Craton in the west, separated by a lineament known as the Phulad Lineament which also marks the western boundary of the Delhi fold belt. In the east, the Aravalli Craton is bounded by the Great Boundary Fault and the Vindhyan basin separating it from the Bundelkhand Craton in the east. It is delimited in the south by the Narmada-Son lineament that marks the northern boundary of the CITZ. Towards the north, several NE-SW trending ridges of basement rocks, e.g. the Delhi ridge, Haridwar ridge, and Faizabad ridge, concealed beneath the Indo-Gangetic alluvium, abut against the Himalayan foredeep.

60

2 Indian Shield

Fig. 2.4 Geological sketch map of the Aravalli Craton depicting major lithologies and the bounding dislocations (modified after Roy and Jakhar 2002; Shekhawat et al. 2007; Kilaru et al. 2013)

The craton consists of an Archaean basement of grey gneisses referred to as the Banded Gneissic Complex (BGC) which is unconformably overlain by early Proterozoic sequences of the Aravalli fold belt to the east and the mid-late Proterozoic Delhi fold belt to the west (Goodwin 1996). The BGC (e.g. Heron 1953) consists of the Mewar Gneiss dated to 3.3–2.5 Ga, it contains a variety of supracrustal enclaves and is intruded by granitoids (Roy and Jakhar 2002) (Fig. 2.4). The Geological Survey of India (Gupta et al. 1980) renamed the Banded Gneissic Complex as the Bhilwara Supergroup which now comprises three tectono-stratigraphic units, namely Hindoli Group, Mangalwar Complex, and Sandmata Complex. The relationships between these and the Aravalli and Delhi sequences are controversial (Ramakrishnan and Vaidhyanadhan 2008). The Berach Granite dated to 2.61–2.45 Ga by various radiometric methods occurs within the Hindoli Group; however, the contact relationships between the two are controversial. The group comprises a turbidite sequence intercalated with basic, intermediate, and acid volcanic rocks and intruded by dolerite sills and dykes of varying ages. The Hindoli belt is at times equated with the Mahakoshal belt of central India. Both belts are considered by some to be interlinked to form a part of the CITZ or the Satpura mobile belt.

2.3 Geology of the Peninsular Block

61

The Mangalwar Complex consists of a variety of migmatitic and composite gneisses with enclaves of amphibolites, mica schists, quartzites, and crystalline limestones. Geochronological data are sparse, but it may correspond to a time band of 2.9–2.6 Ga. The Sandmata Complex is largely a granulite belt whose contact with the gneissic rocks of the Mangalwar Complex is marked by the Delwara lineament. The contact is transitional and encloses the amphibolites-granulite facies transition. There are enclaves similar to the migmatitic gneisses of the Mangalwar Complex and the Mewar Gneiss. The granulites consist of charnockites, enderbites (intermediate charnockite), and two-pyroxene granulites along with amphibolites, norites, and dolerites. The pelitic facies of the granulitesis represented by garnet-sillimanitefeldspar gneisses and cordierite-hypersthene-spinel-kornerupine gneisses (Sharma 1990). These rocks are intruded by granitoids which have yielded ages of 2.83 Ga. Several belts of lead-zinc mineralisation hosted by a quartzite-pelite-carbonate association with black shales and BIF unconformably overlie the Mangalwar Complex and the Hindoli Group. These were included by Heron (1953) within the Raialo ‘Series’ which was included in turn in the pre-Aravalli sequence of the Bhilwara Supergroup (Gupta et al. 1997).

2.3.2

Bundelkhand Craton

The Bundelkhand Craton in central India and to the east of Aravalli Craton is a triangular region with a semi-circular southern boundary and is prominently seen on the geological map of India. The northern fringes of the craton are covered by the Indo-Gangetic alluvium. The southern and western boundaries of the craton, respectively, with the CITZ/Satpura mobile belt and the Aravalli Craton are concealed beneath the Proterozoic Vindhyan basin. The latter is brought in contact with the CITZ along the Son-Narmada North Fault and with the Aravalli Craton along the Great Boundary Fault. The craton is largely occupied by the Bundelkhand granite which contains enclaves of grey and pink banded to streaky gneisses (Fig. 2.5). These comprise high-Al trondhjemites, which have been dated to 3.2–3.3 Ga (Pb-Pb ages on zircon) and low-Al trondhjemites that provide ages of 2.50–2.55 Ga possibly indicating two episodes of magmatism. The basic enclaves provide ages of 3.25 Ga. The enclaves of supracrustal rocks occur primarily as two E-W belts, the northern and the southern, both show northerly dips. The enclaves consist of amphibolites with BIF, ultramafites with subordinate quartzites, metapelites, and marbles. The ultramafites are represented by peridotites, dunites, pyroxenites, and gabbros, which at places form palaeosome in migmatites. Most ultramafites are Mg-rich, however, no detailed work on them is available. The granites are intruded by giant quartz reefs (NE-SW) along brittle- to ductile-shear zones. They contain pyrophyllite and diaspore mineralisation at places. Hydrothermal sulphide mineralisation is also exhibited by the granites. Mafic intrusive rocks post-date the reefs and primarily occur as swarms

62

2 Indian Shield

Fig. 2.5 Geotectonic framework of the Bundelkhand Craton (after Mohanty 2012)

of NW-SE trending dykes. Three episodes of dykes are recognised of which the majority are dolerites and gabbros with subordinate felsites, keratophyres, and rare lamprophyres and/or kimberlites. The Palaeoproterozoic Gwalior basin occurs to the north and the Bijawar and Sonrai basins are located to the south. The Vindhyan basin surrounds the craton in the west, south, and southeast. The Bundelkhand Craton and the adjacent Aravalli Craton both located to the north of CITZ were considered in older literature to form a single entity. The Berach Granite from the Aravalli Craton was treated as equivalent of the Bundelkhand granite. These two cratonic blocks are considered by some authors to form the Bundelkhand Protocontinent whereas the cratons to the south of CITZ, namely Dharwar, Bastar, and Singhbhum makeup the Southern Protocontinent, also referred to as the Deccan Protocontinent of the Indian shield. However, later work has shown that the similarities between the two cratons are confined to the Archaean. The Proterozoic is expressed by extensive sedimentation solely confined to the Aravalli Craton. The latter is also unique since it does not contain typical greenstone sequences such as those developed prominently in the southern cratons. The Proterozoic sequences of the Aravalli Craton consist of quartzites, marbles, pelites, greywackes, and volcanics and comprise two prominent fold belts, namely the Aravalli and Delhi, which unconformably overlie the Archaean grey gneisses and constitute the Banded Gneissic Complex. It is also a craton that hosts one of the largest felsic volcanic provinces (Malani Igneous Province) of India, and possibly

2.3 Geology of the Peninsular Block

63

the third largest globally. Most modern works consider these two cratons to be separate entities, and therefore the tectonostratigraphic framework of each is provided separately.

2.3.3

Singhbhum Craton

The Singhbhum Craton is bounded in the north by the Chotanagpur gneiss terrain which is considered by many as an extension of the CITZ or the Satpura mobile belt. However, the relationship between the two is controversial. The gneissic terrain is considered by some as a mobile belt (Ghose 1983; Mukhopadhyay 1988; Ghose and Chaterjee 2008) whereas others regard it to be a cratonised mobile belt (Naqvi and Rogers 1987; Sharma 2009; Srivastava et al. 2012). The craton consists of a core of 3.5 Ga TTG gneisses intruded by the Singhbhum granite (Saha 1994) in three pulses, the first two between ~3.4 and 3.3 Ga and the third at ~3.1 Ga. The Nilgiri granite, Chakradharpur granite, and Bonai granite are its equivalents. The Singhbhum granite contains enclaves of metasediments assigned to the Older Metamorphic Group and of Older Metamorphic Tonalite Gneiss. The northern margin of the Singhbhum granite is defined by the Singhbhum or Copper Belt shear zone. Three Proterozoic basins of metasediments with BIF (banded iron formations) and metavolcanics, namely the Noamundi-Koira, Gorumahisani-Badampahar, and Tonka-Daiteri basins fringe the Singhbhum granite to the west, east, and south, respectively. They have been grouped by most workers under the Iron Ore Group (Fig. 2.6). Volcanic rocks of varied ages, at times regarded as greenstone belts, overlie the Singhbhum granite and are referred to as Simlipal and Dhanjori in the east, Jagannathpur and Malangtoli in the west, Dalma volcanics in the north, and the Ongarbira volcanics in the northwest. The Simlipal and Dhanjori volcanics are similar to the Bababudan volcanics of the Dharwar Supergroup of south India and to Hamersley-Nabberu basin of Western Australia, the Circum-Superior basin of Canada and Witwatersrand-Triad of South Africa (Ramakrishnan and Vaidhyanadhan 2008). To the north, the North Singhbhum orogen comprises a younger sequence of sediments that rests on the Singhbhum granite with a basal conglomerate that is overlain by low-grade metasediments. The contact zone between the North Singhbhum orogen and the CITZ or the Satpura mobile belt is represented by the Chotanagpur gneiss. The gneissic terrain comprises composite Archaean to Proterozoic granitoids with enclaves of sedimentary rocks, granulites, and mafic to ultramafic rocks. Intrusive rocks often referred to as Newer Dolerites, intrude this sequence, culminating in the stabilisation of the craton. Mildly deformed Proterozoic sediments of the Kolhan Group unconformably overlie the Singhbhum granite.

64

2 Indian Shield

Fig. 2.6 Geological framework of the Singhbhum Craton (modified after Iyengar and Murthy 1982; Saha 1994; Sengupta et al. 1997; Misra 2006)

2.3 Geology of the Peninsular Block

2.3.4

65

Bastar Craton

The Bastar Craton comprises predominantly granites ranging in age from 2.6– 2.2 Ga. These contain vestiges of older TTG gneisses with ages of 3.5–3.0 Ga (Mazumder and Eriksson 2015) as evident from the ages provided by zircons from a gneissic suite in the south Bastar region. Supracrustal rocks comprising quartz pebble conglomerates with banded iron formations (BIF) and subordinate basic and ultrabasic rocks occur as enclaves within the granitic rocks. The metasediments are assigned to the Sukma Group in the south and to the Amgaon Group in the north. These are overlain by rare quartzites at the base which are followed by alternating metabasalts and andalusite schists that belong to the WNW-ESE trending Bengpal Group. These sediments are intruded by granitic rocks in the 2.6–2.5 Ga age bracket and hence the Bengpal Group is considered to be of Neoarchaean age. The supracrustals of the Sukma and Bengpal groups exhibit low-pressure regional metamorphism which has been attributed to the dominant granitic activity (Fig. 2.7). The granitic gneisses and the supracrustals are overlain by the Palaeoproterozoic metasediments of the Bailadila Group that host some of the richest iron ore deposits of this region. The banded iron formations occur as N-S trending belts that consist of basal quartzites and greywackes with pelitic tuffs that contain intercalations of

Fig. 2.7 Geological map of Bastar Craton (modified after Naqvi and Rogers 1987; Ramachandra 2004; Meert et al. 2010)

66

2 Indian Shield

polymictic conglomerates. The BIF is categorised as the ‘Superior type’ and was deposited in a long-lived rift-basin known as the Kotri rift that ultimately culminated/developed into the Kotri-Dongargarh orogeny. The BIF is followed by felsic to mafic volcanics and pyroclastic rocks that comprise the Nandgaon Group which is intruded by the Dongargarh and Nandgaon granites (~2.3 Ga). The Khairagarh Group unconformably overlies the Nandgaon Group. An association of greywackes-pelites-tuffs interbedded with chert-limestone-manganese beds comprises the Chilpi Group which overlies the Khairagarh Group. Two isolated belts almost parallel to each other occur to the east and west of the Kotri-Dongargarh orogen. These are designated as the Sonakhan and Sakoli belts. The former consists of acidic-basic bimodal volcanics, BIF, and a greywackeargillite association with subordinate conglomerates. It is intruded by ~2.4 Ga granites and hence may be tentatively correlated with the Nandgaon Group. The Sakoli Group (~1.8 Ga) comprises metasediments and subordinate volcanics, BIF, cherts, and tourmalinites and could be homotaxial with the Chilpi Group, both of which were deflected parallel to the CITZ during the Satpura orogeny (Sm-Nd protolith age: ca. 2.0–1.7 Ga; metamorphic age: ~846–986 Ma). Three prominent belts of granulites occur within the craton, viz. Bhopalpatnam, Bhandara-Balaghat, and Kondagaon. The Mesoarchaean Bhopaltnam belt occurs on the shoulder of the Godavari rift. The Kondagaon belt is a horst block within the gneisses; it forms a part of the CITZ at the craton boundary and may have been exhumed during Neoarchaean. The Balaghat-Bhandara belt occurs as boudins of granulites which may have been exhumed during the evolution of the CITZ. The Bastar Craton is an ensemble of rift-basins that possibly stabilised by 1.8 Ga, prior to which periodic exhumation may have brought up narrow granulite belts from mid-crustal levels. Late-Archaean-early Proterozoic metasediments of the Chhattisgarh basin overlie the basement rocks and have a thickness of about 2.5–3.0 km. Several episodes of orogeny and metamorphism have been identified based on geochemical and geochronological data. Continental rifting and dyke emplacement followed, with at least six intrusive episodes between 2.9 and 1.42 Ga period, within which several petrological associations have been identified including High-Ti and Low-Ti basic dykes (Srivastava et al. 2014). The evolution of the craton bears many similarities with Western Australia and South Africa (Mazumder and Eriksson 2015).

2.3.5

Dharwar Craton

The Dharwar Craton as mentioned above consists of two cratonic blocks, the WDC and the EDC. The WDC is delimited in the west by the Arabian Sea which represents the western marginal rift. In the east, it is bounded jointly by the Chitradurga Shear zone and the Closepet Granite Batholith along which it is welded to the EDC. Both cratonic blocks are juxtaposed against the Middle to Late Proterozoic Eastern Ghat Mobile Belt (Naqvi and Rogers 1987) which has a discordant relationship with the

2.3 Geology of the Peninsular Block

67

general NNW-SSE grain of the metamorphic terrain of the cratons to the N. The northern boundary is concealed beneath the Deccan Traps flood basalts. It occurs about 400 km to the north and is marked by the Central Indian Tectonic Zone (CITZ), a major tectonic lineament also known as the SONATA (Son-NarmadaTapti) rift system that represents the eastern arm of the Cambay triple junction (Burke and Dewey 1973). The WDC and EDC form a part of the non-Charnockitic (low grade) region (Fermor 1936) and exhibit the classical Archaean granite-greenstone assemblage (Naqvi and Rogers 1987). The greenstones comprise a volcano-sedimentary sequence also referred to as schist belts (e.g. Radhakrishna and Naqvi 1986) in older literature and as supracrustals (Chadwick et al. 1992) in modern works (Fig. 2.8). The evolution of the WDC and EDC although appears to be broadly similar, like most Archaean cratons of the world, yet they differ significantly from each other in several respects and show a distinctly independent identity displayed in tectonic, lithological, and thermal characteristics even during Archaean and particularly so, during Cenozoic (post-Palaeocene). The supracrustal rocks show profuse development in the WDC in particular and exhibit distinct variation from those of the EDC, reflected in magmatism, lithology, grade of metamorphism, and thermal characteristics. The differences in the characteristics of both have significant tectonic implications which has prompted the division of the craton into two blocks, namely the WDC and EDC. In comparison with these, the adjacent cratons, namely the Bastar and Singhbhum, consist largely of granitic gneisses and granite with a very subordinate component of greenstones. A concise summary of the characteristics of the two cratonic blocks is provided below. The northern portion of the Dharwar and Bastar cratons and the southern part of the Bundelkhand Craton in west-central India is covered by lavas of Deccan Traps. They variously overlie a basement of granitic gneisses, granites, greenstones, Proterozoic sediments of the Kaladgi, Vindhyan, and Gondwana Supergroups, and the late Cretaceous Lameta sediments.

2.3.5.1

Western Dharwar Craton

Swami Nath and Ramakrishnan (1981) proposed the division of the Dharwar Craton into two crustal blocks, namely the WDC and EDC based on their distinctive tectonic evolution, the petrotectonic assemblage, the metamorphic zonation, the distinct environmental conditions of formation of the supracrustal assemblages, and the characteristic thermal evolution of the two blocks. The WDC consists of 3.3–3.0 Ga TTG gneisses some of which are migmatitic which evolved under high temperature-intermediate pressure metamorphic conditions. The associated greenstones consist of two greenstone sequences. The older (~3.3–3.1 Ga) Sargur Group supracustals occur as tectonically transported enclaves within the gneisses. They evolved largely as a subaerial, platformal assemblage of clastics deposited in shallow water, in evaporite environment of an ensialic basin along with continental, extrusive

68

2 Indian Shield

Fig. 2.8 Regional geological framework of south India depicting the configuration of the Western and Eastern Dharwar Cratons with the Southern Granulite Terrain, separated by the Moyar-Bhavani and Palghat-Cauvery and Achankovil shear belts (bold black lines). The white line represents the ‘Fermor Line’—the orthopyroxene isograd which separates the amphibolites facies rocks in the north from those of the granulite facies in the south (modified after Ramakrishnan 1994)

komatiitic ultramafites. The younger sequence (~2.9–2.4 Ga) comprises the greenstones of the Dharwar Supergroup (Fig. 2.9) that largely evolved in intra-cratonic ensialic rift basins (e.g. Chadwick et al. 1989, 1992) or passive continental margin and were intruded by the 2.6–2.5 Ga Closepet Granite.

2.3 Geology of the Peninsular Block

69

Fig. 2.9 Geological sketch map of the Western Dharwar Craton (modified after Meert and Pandit 2015)

In recent years, the greenstone belts have been regarded as tectonic slices of oceanic and island-arc crust that have been thrust together to form tectonic assemblages similar to those in belts found in the present-day Pacific Ocean. Such assemblages are believed to have formed along subduction zones and are seldom brought to the surface with least change. The greenstone belts are economically important being the repositories of several types of deposits including ferrous-, base, and precious metals. The evolutionary style of this block is analogous to the Superior- and Barbertontype greenstone belts (Dessai and Deshpande 1979). The volcaniclastic assemblage consists of subaqueous pillow basalts, and subaqueous volcaniclastic rocks which host gigantic concentrations of banded iron formations and manganese ores. The

70

2 Indian Shield

supracrustals were profusely intruded by dyke swarms referred to as the Newer Dolerites during Proterozoic time. Ramakrishnan (1988) places the boundary of the Dharwar Craton along the Palghat-Cauvery shear zone.

2.3.5.2

Eastern Dharwar Craton

The EDC consists of a remobilised gneissic basement which comprises granodioritic gneisses and granitoids that vary in age between about 2.7–2.5 Ga. The supracrustals are represented by a volcano-sedimentary assemblage that gives the impression of an oceanic intra-arc tectonic setting (Krogstad et al. 1991). They consist of high-Mg mafic volcanic rocks of low-Ktholeiitic affinity with tuffaceous sediments and minor iron formations and contain gold mineralisation (Fig. 2.10). The greenstone assemblage locally contains subordinate ultramafic members; no extrusive ultramafites exist and this is characteristic of oceanic environments. It is characterised by a high temperature, low-pressure metamorphism of middle- to upper-amphibolites grade. This assemblage is analogous to the Keewatin-type greenstones of Canada and lacks true sedimentary conglomerates and sedimentary intercalations. This sequence is overlain by the mid-Proterozoic, intra-cratonic sequence of the Cuddapah

Fig. 2.10 Geological sketch map of the Eastern Dharwar Craton (modified after Meert and Pandit 2015)

2.3 Geology of the Peninsular Block

71

Supergroup with a maximum thickness of ~600 m (Goodwin 1996). The rocks are intruded by kimberlites and lamproites with radiometric ages of 1.02–0.4 Ga. These are succeeded by the emplacement of kimberlites ~90 Ma (Chalapathi et al. 2016). This is the only evidence of magmatism in the craton after late Proterozoic and especially during the Cenozoic.

2.3.6

Granulite Terrain

Granulites have always attracted attention from geologists as they are believed to be the representatives of the lower crust. As the latter is not directly accessible for observation, it is either the granulite terrains or the xenoliths brought up by rapidly intruding magmas that are often used to get an insight into the composition of the deep crust. They are lower crustal felsic rocks typically characterised by the presence of anhydrous minerals such as orthopyroxene and garnet. In the south Indian peninsula, the regional grade of metamorphism progressively increases from the north to south. In the southern part, the rocks have been metamorphosed to the highest grade of granulite facies as a result of high temperature and pressure. The traditional view is that the granulite terrain in general represents the continuation of the low-grade granite-greenstone terrain in the north. Three belts of granulite facies rocks are recognised in the shield, namely (a) the Chotanagpur Gneiss-Granulite Terrain (b) the Eastern Ghat Mobile Belt (EGMB) delineating in general the eastern borders of the EDC, the Bastar Craton, and the Singhbhum Craton and (c) the Southern Granulite Terrain (SGT) comprising the terminal portion of the Indian peninsula to the south of the Moyar-Bhavani, PalghatCauvery, and Achankovil shear belts. These rocks were first studied by Holland (1907) who designated them as ‘Charnockites’ after Job Charnock the founder of Calcutta, whose tombstone is made of this rock type. The granulites may either represent dry felsic magmas that were emplaced in the lower crust or granitic intrusions that have been dehydrated during a subsequent granulite facies metamorphic event. In the former case, post-magmatic high temperature recrystallisation may result in the development of granulitic microstructures superimposed on the magmatic textures recognisable from the relict igneous assemblages. For both types, fluid mineral interactions at grain boundaries during retrogression, are documented by microstructures, including K-feldspar microveins and myrmekites.

2.3.7

Chotanagpur Gneiss-Granulite Terrain

The E-W trending Chotanagpur gneiss-granulite terrain occupies an area 500 km long and 200 km wide in eastern India. The belt is bounded in the north by IndoGangetic alluvium and in the south by the South Purulia shear zone which separates

72

2 Indian Shield

it from the North Singhbhum orogen. In the east, it is delimited by the Rajmahal Traps and the Bengal basin, and in the west it is separated from the CITZ by the Mahanadi Gondwana graben. The belt comprises a variety of gneisses and granites along with supracrustal enclaves and discontinuous bands of granulites. These rocks are intruded by anorthositic gabbros, norites, and dolerites. Some of the dolerite dykes contain coronas of olivine around plagioclase and garnet around pyroxene (Ramakrishnan and Vaidhyanadhan 2008).

2.3.8

Eastern Ghat Mobile Belt

The Proterozoic (~1.6–1.4 Ga) Eastern Ghat Mobile Belt also referred to as the Eastern Ghat Granulite Belt, overthusts the eastern margin of the EDC, and the Bastar and Singhbhum cratons. It extends NE for over 1000 km and is about 300 km wide. It is an ultrahigh temperature metamorphic belt in which rocks were reequilibrated at temperatures of ~1000
C and pressures of 0.9–1.0 GPa corresponding approximately to depths of 30–35 km. The belt is dissected by two almost NW-SE trending major palaeo-rifts, the Godavari rift towards the south, and the Mahanadi rift further northeast. The Godavari rift-graben contains Permo-Carboniferous and Cretaceous to Tertiary sediments. The Mahanadi graben hosts Cretaceous to Tertiary sediments. Charnockites (granulites), khondalites, and lepynites dominate this belt. The eastern boundary of the belt is concealed under the Quaternary sediments of the east coast of India. The main phase of granulite metamorphism occurred 0.95– 0.93 Ga, followed by decompression at 0.78–0.75 Ga. A late thermal overprint occurred at 0.52–0.52 Ga. Between 0.78–0.75 Ga (monazite U-Th-total Pb and zircon U-Pb SHRIMP geochronology, Chatterjee et al. 2017) the break-up of the supercontinent Rodinia took place. An older age of 1.18 Ga has been recently attributed to the ultrahigh temperature metamorphism (Das et al. 2017). The isobarically cooled granulites were exhumed ca. 0.98–0.93 Ga (Bose and Gupta 2018). Alkaline plutons with nepheline syenite occur in the vicinity of the cratonmobile belt margin to the west and north. Layered anorthosite complexes occur in the south while massive anorthosite are found in the north. The belt is intruded by the Kandra Igneous Complex interpreted as an ophiolite sequence which is characterised by a mélange zone near Kanigiri.

2.3.9

Southern Granulite Terrain

These rocks, in general, extend from the southern border of Karnataka in the north to Sri Lanka in the south. Holland (1907) recognised that charnockite is a suite of rocks that varies from acid to ultrabasic types. Two transitional facies of granulites are recognised, namely the ‘cratonic charnockites’ and the ‘massif granulites’ which

2.3 Geology of the Peninsular Block

73

occur further south of the above and comprise the Southern Granulite Terrain (SGT) which includes the Pandyan Mobile Belt (e.g. Ramakrishnan and Vaidhyanadhan 2008). The rocks are characterised by a greasy appearance and a granular texture made up of equidimensional minerals with rectilinear boundaries that meet in 120
triple-point junctions, in contrast to the schistose texture of the rocks in the north, and are hence known as ‘Granulites’. As mentioned above, these rocks were first described as charnockites. The term charnockite sensu stricto is, however, restricted to a rock of ‘granitic’ composition consisting of quartz, microcline, plagioclase, orthopyroxene (hypersthene), and biotite often with garnet. Unfortunately, the term has been loosely used by later workers to include any greasy-looking hyperthene-bearing rock. The term is, however, losing its importance in preference to the term ‘granulite’ which is widely used in modern literature. These rocks were originally proposed by Holland (1907) to be magmatic, but are now known to be metamorphic/metasomatic in origin (Fyfe 1973; Newton and Hansen 1983; Clemens 2006). The rocks attain considerable importance worldwide since they are the sole representatives of the lower crust, available for observation. They are among the only other samples of lower crustal rocks, but for the xenoliths accidentally entrained by kimberlites, lamproites, lamprophyres, and other alkaline rocks. The SGT was originally considered to be the exhumed lower crust of the Dharwar Craton. A more modern view, however, suggests that it is an amalgamation of several blocks of granulite terrains (Meert et al. 2010) welded together by collisional tectonics (Chadwick et al. 2000; Santosh et al. 2009; Raith et al. 2016). Four blocks are identifiable; from north to south they are (a) the Nilgiri domain (also referred to as the south Karnataka block inclusive of Tamil Nadu), (b) Coimbatore-Salem domain, (c) Madurai domain, and (d) Trivandrum-Kanniyakumari domain. These domains are variously delimited by major tectonic dislocations, namely the MoyarBhavani, Palghat-Cauvery, and Achankovil shear belts. Granulite xenoliths, entrained by lamprophyres from the Deccan Traps, have been recently reported from Murud-Janjira to the south of Mumbai and further north along the CITZ (Dessai and Vaselli 1999; Dessai et al. 1999, 2004, 2009).

2.3.9.1

Granulite Types in South India

Recent work has shown that granulites from south India occur as two distinct types that probably belong to two different episodes of granulite formation. One is a massif-type granulite that constitutes the Southern Granulite Terrain and is widespread in the southern part of Karnataka, Kerala, and north Tamil Nadu. The other, more frequently found in Karnataka, in areas further to the north from the above, is a metasomatic type (‘cratonic charnockite’) which is patchy and is often seen to grade into the surrounding gneisses. The massif granulites form homogeneous masses covering hundreds of square kilometres and whose protoliths are believed to be Archaean in age. The metasomatic type which develops across the foliation of the

74

2 Indian Shield

host gneiss is considered to reflect a later event that occurred around the close of the Archaean. The metasomatic granulites occur in a 30–60 km wide zone that marks the transition from the low-grade metamorphic terrain in the north to the high-grade granulite terrain in the south. A number of E-W and N-S trending shear zones define the regions of granulite transformation. Within this transitional belt intermediate charnockites referred to as enderbites and mafic types called mafic granulites are widely distributed. These rocks are interspersed with amphibolites facies gneisses and metasedimentary granulites referred to as khondalites. A typical acid granulite referred to as “charnockite” is a dark rock greasy in appearance due to the presence of bluish quartz. The latter may also be greenish or brownish and is often studded with acicular inclusions. Orthopyroxene (hypersthene) is strongly pleochroic, and garnet is common. K-feldspar is generally represented by microcline which also has a bluish tinge. Plagioclase varies in composition from oligoclase to andesine, with variable proportions of biotite may be present. Accessories are represented by zircon, apatite, graphite, and monazite. Opaques, such as magnetite and ilmenite occur in fairly large proportions. In metasomatic granulites, the conversion of hornblende to diopside is often seen. Quartz and felspar are invariably crowded by minute inclusions. Biotite is often brownish red.

2.3.9.2

Igneous Versus Metasomatic Granulites

Early workers favoured an igneous origin and considered the granulites to be magmatic. This was supported by (a) their occurrence as large massifs, (b) crosscutting relationships exhibited by granulite dykes, (c) presence of contact metamorphic effects with the development of garnet, kyanite, or sillimanite, (d) presence of xenoliths of older rocks, e.g. enclaves of Sargur-type supracrustals consisting of hypersthene-bearing banded iron formations and cordierite granulites, (e) variation in composition from acid to ultrabasic types taken to indicate magmatic differentiation, and (f) evidence of assimilation producing hybrid compositional types. However, more recent studies, however, support a metamorphic origin evident from (a) close association of granulites with metasediments such as khondalites (garnetsillimanite-graphite schist/gneiss), (b) absence of intrusive relationship, (c) confinement to areas of high-grade metamorphism, (d) occurrence of schists/gneisses and granulites as exclusive outcrops indicating transformation of gneisses to granulites and (e) petrographic evidence of the gneiss to granulite transformation (Newton et al. 1980; Janardhan et al. 1982). Field evidence broadly supports two processes for the formation of granulites in south India. High temperature regional metamorphism of impure sediments and igneous rocks such as norites, pyroxenites, and peridotites is suggested to account for the formation of granulites. The metasomatic types are believed to have formed by metamorphism under high P-T conditions followed by dehydration brought about by fluids, primarily CO2, driven from the deeper parts of the crust. The role of fluids in the transformation of the gneiss to granulite is supported by fluid inclusion studies

2.3 Geology of the Peninsular Block

75

(Hansen et al. 1987). Dehydration during this transformation was responsible in bringing about partial melting of continental rocks and formation of granitic melts that possibly resulted in the emplacement of the anatectic Closepet Granite Complex. The other view, however, suggests that partial melting itself is sufficient to bring about dehydration of rocks and that the fluxing of CO2 may not be a necessary prerequisite for the transformation process. As discussed above, the granulite terrain, bordering the Meso- to Neoarchaean province of the Dharwar Craton, to the south was exhumed by 2.3 Ga (Jayananda and Peucat 1996). It is a region of palaeo-lower crust which has had a complex lithotectonic architecture characterised by several episodes of metamorphism, magmatic activity, deformation, and exhumation. Certain sections of this terrain have been studied in great detail in recent years. The mafic-ultramafic intrusives from these sections are of critical importance in elucidating the deep crustal structure of this region, in terms of depth estimates that are better than those provided by the geothermobarometry of granulites. The geology of these tectonic domains is, therefore, described in greater detail in the following paragraphs.

2.3.9.3

Nilgiri Domain

The northernmost part of the granulite terrain consists of felsic orthogranulites (enderbitic granulites) comprising Biligirirangan Hills, Male Mahadeshwar Hills, Krishnagiri domain, and Shevaroy Hills, which are considered by some to represent the extensions of the Dharwar Craton to the north (Fermor 1936; Narayanaswamy 1975). The magmatic protoliths of these were emplaced during two magmatic pulses at ~3.4 and ~2.55 Ga (Peucat et al. 1989; Raith et al. 1999) followed by granulite facies metamorphism and migmatisation at ~2.52 Ga (Harlov et al. 1997; Raith et al. 1999). These rocks were overprinted by a tectonothermal event around 2.48 Ga (Raith et al. 1999; Meissner et al. 2002; Saitoh et al. 2011). A felsic paragranulite dominated unit forms the Nilgiri Hills as a part of the Moyar-Bhavani suture zone that marks a thrust contact with the WDC to the north. A narrow belt of granulite facies mafic to ultramafic rocks intruded by 2.52 Ga non-garnetiferous enderbitic orthogneisses underlies the para-autochthonous granulites of the Nilgiri Hills. The mafic-ultramafic rocks occur as dismembered igneous complexes and bodies of ferroan metagabbros. Eastwards, enclaves of mafic-ultramafic rocks can be traced to the south of the Eastern Dharwar province (Raith et al. 1999; Mukhopadhyay et al. 2003). The mafic-ultramafic rocks have been interpreted to represent late Archaean (2.9– 2.54 Ga) fragments of the former oceanic crust (Bhaskar Rao et al. 1996; Raith et al. 1999) that later evolved into the Moyar-Bhavani-Cauvery suture. Along this suture, the Madurai granulite province to the south was accreted to the Western Dharwar Province and the Nilgiri Granulite Terrain during Palaeoproterozoic (~2.48 Ga) by collision and coeval high-pressure granulite facies metamorphism (Raith et al. 1990, 1999; Sengupta et al. 2009).

76

2 Indian Shield

The Moyar-Bhavani shear zone separates the Nilgiri granulite terrain from the WDC including the Biligirirangan Hills to the north. The pervasive N-S grain of the WDC is truncated by the E-W trending structure of the Moyar-Bhavani shear zone. Within the shear zone extremely sheared mylonitic to ultramylonitic lithologies of the Dharwar Craton and strongly deformed and annealed high-pressure basic granulites and enderbitic granulites show northerly lineation with a steep plunge of 80– 90
, and almost no significant strike-slip movement (Srikantappa 1996). Due to the northerly upward thrusting of the Nilgiri granulite block along the Moyar shear zone, the N-S trending Dharwar lithologies have been rotated to E-W along the northern margin of the Moyar shear zone. However, no such structural disturbance is observed in the east and the N-S grain continues up to the Bhavani shear zone. The Nilgiri highland massif exposes a strongly tilted section of granulites from the lower crust. Palaeopressure estimates document a deep level of exposure (0.9– 1.0 GPa, ~35 km) along Moyar-Bhavani shear zone and progressively higher levels towards southwestern part (0.6–0.7 GPa, ~22 km palaeo-depth) (Raith et al. 1990). The crustal section is essentially composed of garnetiferous enderbitic granulites with minor kyanite bearing gneisses, quartzites, and banded magnetite quartzites. In the north, these lithologies are intruded by intermediate and basic magmatic rocks now represented by non-garnetiferous enderbites, garnetiferous basic granulites, gabbroic to anorthositic two-pyroxene granulites and pyroxenites. The igneous nature of the protolith is evident from their clear intrusive relationship with the gabbroic granulites and by their geochemical features (Raith et al. 1999). To the south, the Nilgiri granulites are truncated by the NW-SE trending Bhavani shear zone with intensely deformed polyphase association of amphibolites and hornblendic gneisses and deformed basic granulites (ferroan gabbros), deformed stratiform peridotites, pyroxenites/hornblendites, gabbronorites, and enderbitic granulites (Selvan 1981). The dominant NE-SW planar fabric in the Bhavani shear zone dips steeply towards the SE and exhibits stretching lineation that plunges 30–40
towards the SE. Northerly directed movement and uplift of the southern block against the Nilgiri block is apparent. Details of the basic rocks are added in the next chapter.

2.3.9.4

Coimbatore-Salem Domain

This tectonic block is delimited in the north by the Moyar-Bhavani shear belt and in the south by the Palghat-Cauvery shear zone. It is about 60 km at its widest and broadly trends E-W. Petrologically, the belt comprises fissile gneisses (also referred to as Bhavani Gneiss) with enclaves of high-grade rocks assigned to the Satyamangalam Complex, with slices of dunite, peridotites, websterites, garnetiferous gabbros, and anorthosites along with rocks of eclogitic affinity, charnockites, pyroxene-granulites, migmatites, and rare pink granites (Viswanathan et al. 1990). The western part of the belt comprising ultramafites and layered pyroxenitesgabbros-anorthosites with chromites and associated cobalt-nickel-copper sulphide

2.3 Geology of the Peninsular Block

77

mineralisation and metasediments form a part of this belt referred to as the Udagamandalam sub-block.

2.3.9.5

Madurai Domain

Further south, in northeastern Madurai (10
, 530 , 4400 N, 78
, 220 , 8800 E), in the vicinity of the Ayyarmalai temple hills, the enderbitic charnockites are retrogressed to hornblende-biotite gneiss. In a polyphase basement exposure remnants of older (~2.67 Ga) rift-related alkaline rocks intrude granitoid intrusions of an arc complex (~2.60 Ga) (Raith et al. 2016). Locally, the charnockites contain inclusions and enclaves of older basement rocks (ibid.). Isoclinally folded and compositionally banded, extended bodies of meta-troctolite, the remnants of a stratiform intrusive complex occur among the protoliths of thebiotite gneiss. The banding in the metatroctolite is attributed to former cumulate layering of the magmatic protolith (Raith et al. 2010). It was exhumed to mid-crustal conditions (~750
C, 0.5–0.6 GPa) during Neoarchaean reworking. In the west of the Ayyarmalai quarry, syenitic biotite gneiss which represents an older basement fragment (Raith et al. 2016), shows a discordant relationship with host charnockite. The syenitic bodies contain mafic lenses forming boudins that represent dismembered mafic dykes (ibid.). The mafic dykes were emplaced coevally with the emplacement of precursor syenites but prior to the deformation of the host gneiss. E-W trending mafic dyke swarms that intruded around 2000–1500 Ma mark the end of mafic magmatism in the Dharwar Craton. In older literature, these dykes are variously described as newer dolerites and metadolerites. In the granulite terrain to the south of the WDC, the enderbitic charnockites of the Nilgiri massif and those from Madurai contain inclusions of mafic and ultramafic rocks now represented by numerous extended bodies, lenses, and pods of gabbroic to anorthositic two-pyroxene plagioclase rocks, ferroan garnetiferous pyroxene-plagioclase rocks and pyroxenites. These rocks are cut across by undeformed but metamorphosed late dolerite dykes. In the Madurai block further south deformed bodies, isoclinally folded and compositionally banded meta-troctolite, occur as remnants of a stratiform intrusion. In the western parts of the terrain, syenites contain mafic lenses forming boudins of dismembered mafic dykes (Raith et al. 2016). The presence of mafic intrusives is significant to obtain information on the palaeodepth of derivation of the progenitors of granulites. Reliable estimates of the depth of derivation of the protoliths is otherwise difficult to get hold of, as the latter have undergone re-equilibration during emplacement and thus often tend to provide lower P-T estimates.

78

2.4

2 Indian Shield

Geophysical Studies in Southern India

The crust beneath the Indian shield has received fairly wide attention in recent years, particularly post-1990s, through multidisciplinary approaches largely based on remote sensing applications, with several papers and books devoted to the topic (the reader is referred to the collection of papers in Geological Society of India Memoirs 50 and 53 and a review on the Deep Continental Structure of India by Mahadevan 1994). A brief summary of the important findings which are of relevance primarily to deep crustal investigations are outlined below. India in general, reveals a distinct zonation in terms of gravity distribution (Misra and Ravi Kumar 2008). The southern part of the shield is characterised by a negative gravity anomaly (70 to 130 mGal), the central region consists of denser material (70 to +10 mGal) whereas the Himalayan region again displays a negative gravity field (Verma 1985; Raval 2003) suggesting distinct density variation in the lithospheric mantle. Geophysical studies of the Aravalli Craton have shown a ~80 mGal Bouguer anomaly corresponding to the Aravalli mountains representing the horst of Fermor (1930). Seismic reflection and electromagnetic profiles between Nagaur and Jhalawar have shown several reflector zone that may represent palao-subduction zones and/or collisional zones. Moho has been identified at depths varying from 38 to 48 km. A high-velocity zone (Vp: 7.3 km/s) above the Moho may define the lower crust in this region. Magnetotelluric studies (NW-SE profile) have identified a midcrustal conductor (Gokarn et al. 1995) prominently observed in the west is disrupted around Jahazpur (approx. 26o N, 76
E) where it occurs at shallower depths and has been attributed to the Jahazpur thrust. The southern margin of the Bundelkhand Craton is marked by steep gravity gradient which is considered to be an expression of the Son-Narmada Fault. The northern contact is defined by a low gravity gradient. The extension of the Bundelkhand granite beneath the Vindhyan sediments is indicated by aeromagnetic investigations. The northern extension of Hirapur-Mandla DSS profile (NW-SE) identified a Vp: 6.5–6.7 km/s layer below the Vindhyan sediments. This may correspond to the exhumed lower crust (Mahadevan 1994) which served as the basement for the Vindhyan sediments. The Singhbhum Craton exhibits considerable variation in crustal thickness. In the west, Moho is located at a depth of around 37 km whereas in the east it occurs at a much shallower depth of 28 km. Deepening of the Moho is also noticed in N-S profile, from 35 km in the south to 38 km in the north. Prasad et al. (2005) have suggested dense (3.3 g/cm3) mafic lower crust beneath the craton with Moho discontinuity at a depth of 40 km. Ps and Sp receiver function data from the Bastar Craton show that Moho is located at a depth of 35–42 km (Sharma and Ramesh 2013). On the basis of heat flow studies, the thickness of the lithosphere beneath the Bastar cratons is estimated to be at least ~175 km (Gupta et al. 1993; Roy and Mareschal 2011).

2.4 Geophysical Studies in Southern India

79

Surface wave dispersion studies beneath the Bastar cratons indicate a 155 km thick lithosphere which has a 120 km thick high-velocity lid underlain by the LVZ where shear wave velocity is 4.4 km/s up to a depth of 230 km (Mitra et al. 2006). A zone of anisotropy is also indicated with SH velocity of 4.8 km/s and SV velocity of 4.58 km/s between 160 and 280 km depth with low S-wave velocity of 4.55 km/s (Battacharya 1974). The Dharwar Craton has been subjected to several geophysical investigations. Three deep seismic sounding (DSS) transects (E-W) across the craton and one NNESSW profile have provided a wealth of information. East dipping thrusts have been identified which define the boundary of the Shimoga and Chitradurga schist belts and of the Cuddapah Basin. Moho is identified at a depth of 38–42 in the WDC and at 34–38 km in the EDC. Teleseismic tomography, receiver function analysis, and electromagnetic studies along the Goa-Raichur (E-W) transect have corroborated these findings. The upper mantle velocities beneath the WDC are slightly lower (Vp: 8.0–8.2 km/s) than those (Vp: 8.5–8.6 km/s) beneath the EDC. Receiver function studies using Rayleigh and Love waves along the eastern fringes (Western Ghats) of the WDC show a 60 km thick high-velocity lithospheric lid along the continental margin; its thickness increases to 155 km in the interior of the shield and is underlain by an LVZ (Vs: 4.5 km/s) (Battacharya 1974). A thick lithospheric keel (165–180 km) was suggested for Dharwar Craton and northern block of Southern Granulite Terrain while a thin lithosphere was proposed beneath Bangalore (EDC) from body wave analysis (Singh et al. 2014). Gupta et al. (2003) suggested a thickness of 200 km for the EDC and 260–280 km for the WDC. Similarly, surface waves (Mitra et al. 2006) and magnetotelluric models (Nagarjunneyulu and Santosh 2012) also suggested thick lithosphere. Maurya et al. (2016) proposed variable depth for the LAB which varied from 120 km in the south to 250 km in central India. Magnetotelluric studies along three E-W transects across the craton provided for a thickness of 80–100 km (Gokarn et al. 2004, and in press). Vs beneath the WDC vary from 3.9–4.0 km/s clearly reflecting on the nature and composition of the mantle. The N-S Kuppam-Palani DSS transect, which is almost parallel to the tectonic grain of the terrain and which extends into the SGT, has identified a four-layered velocity structure (discussed later) with a low-velocity zone at mid-crustal levels. The crustal thickness has been estimated to be 41–45 km along the extent of the transect. A Moho upwarp has been identified at Nagarsapatti in the Dharmapuri alkaline complex.

2.4.1

Crustal Structure from Magsat and Gravity Data

The thickness of the magnetic crust over the WDC is about 36 km, whereas over the CITZ it 34–36 km. These generalisations are made despite the limitations of interpreting the thickness of Curie crust from Magsat data (Meissner 1986). The

80

2 Indian Shield

thickness of 36 km is higher than the DSS and gravity estimates which are 18–30 km in the vicinity of Mumbai (Mahadevan 1994). Despite the intrinsic limitations of gravity data in estimating the crustal thickness, Qureshy (1970) made an attempt to determine the crustal thickness of the Indian subcontinent based on a regression analysis of Bouguer anomalies and elevation. A crustal thickness map based on a 150  150 square grid has been prepared (Mahadevan 1994 after Qureshy 1970). The highest elevation in each grid has been used to estimate crustal thickness as these are the likely areas of maximum crustal thickness at their respective locations. Broadly, the thickness of the crust ranges from 35 to 40 km over a large part of the subcontinent. In the high relief regions, such as the Himalaya, Aravallis, and Nilgiris among others, the crust attains maximum thicknesses ranging from 40 km to over 80 km, such as in the Himalaya. In the south, maximum thickness is found at Nilgiris (~50 km), 44 km in the Eastern Ghats, and 40–50 km over the Western Ghats. On the basis of Vp, a density column has been established to a depth of 70 km. The average crustal density of the Indian shield is estimated to be 2.88 g/cm3 whereas the sub-crustal density varies from 3.45 to 3.49 g/cm3 (Qureshy 1970). This density distribution agrees fairly well with density estimates determined in the Laccadive region by Francis and Shor, (Francis and Shor Jr. 1966). A Bouguer anomaly map of India (Verma and Subramanyam 1984) delineates a gravity high field in Central India. In the south, the crust thins gradually towards the off-shore region along the west coast whereas the crustal transition is abrupt along the east coast (ibid.). Negi et al. (1989) have used free-air gravity data to generate a model of apparent relative density contrast model assuming a uniform crustal thickness 40 km. Most of the country displays low density contrast values. The palaeo-rifts such as the Narmada, Godavari, and others show considerable variation in relative density which is a reflection of the deep crustal structure as will be discussed later.

2.4.1.1

Seismic Structure of WDC

Western India is traversed by several major rifts. The Narmada-Tapti rift system (Naqvi et al. 1974) is characterised by a Bouguer anomaly of +40 mGal. The crust here is attenuated by 8–15 km (Kaila 1988). The Cambay graben forms the N–S western margin of the rift and shows a linear Bouguer anomaly (+70 mGal, over an area of 300  60 km) at Mumbai (Takin 1966). Starting from the CITZ in the north, integration of DSS profiles across the Narmada-Son lineament reveals convergence of dips of the reflectors on either side of the CITZ. The zone of convergence has been interpreted as a thrust zone along which the Amgaon Gneiss in the south has been thrust on the Sausar-Tirodi gneiss/granulite terrain in the north. A five-layer velocity structure is proposed along the transect with the crust varying in thickness from 44– 46 km with two low-velocity layers (Reddy and Rao 2003). DSS profiles suggest shallowing of the Moho from 31 km at the coast (Mahad) to about 18 km at Billimora (Kaila 1988). Crustal thinning of 12–25 km is observed

2.4 Geophysical Studies in Southern India

81

along the trend of the gravity anomaly. The thinning is supported by DSS profiles which suggest that Moho lies at a depth of about 31 km near the coast at Mahad and at 18–21 km near Mumbai (at Billimora, about 150 km north of Mumbai) indicating an upwarped (arched) Moho (ibid.) where the upper crustal thickness is estimated to be only 6 km. A crustal thickness of 36–40 km was reported from Koyna I and II profiles (Kaila and Krishna 1992). Agarwal et al. (1992) identified a magnetic interface at 40  4 km whereas Negi et al. (1986b, 1987) have estimated a Curie depth of 43 km for this region. Interpretation of aeromagnetic data over the WDC suggests that the magnetic crust extends to a depth of 36 km corresponding to the Moho depth (Mahadevan 2003) whereas in the SGT the Curie isotherm is confined to a shallow depth of 22 km. This depth corresponds to the low-velocity zone encountered in the Kuppam-Palani DSS transect (Reddy et al. 2003a, b). The lithospheric thickness varies from approximately 101 km beneath the Deccan Traps in general (ibid.) to a suspected 70 km near Valsad (e.g. Mahadevan 1994).

2.4.1.2

Structure of the Crust Beneath EDC

The crust beneath the EDC is considered to be volcano-sedimentary in nature and may have evolved in an intra-arc tectonic setting (Krogstad et al. 1991) of compressional environments between the EDC batholiths and the WDC (Chadwick et al. 1996). Krogstad et al. (1989) suggest a Phanerozoic-type plate model for the Kolar metavolcanics whereas Zachariah et al. (1996) propose an island-arc type of crustal model for the Ramaguiri Belt, both of which occur within the EDC. The EDC displays a comparatively thinner crust 32–35 km with 10–15 km upper crust (Vs ~ 3.5 km/s) and a lower crust with Vs 3.8–3.9 km/s up to the Moho; these thicknesses are consistent with DSS profiles (Kaila and Krishna 1992). Receiver function S-wave velocity depth profiles suggest that a layer of velocity 4.1–4.3 km/s occurs beneath the deep crust. The bulk of Vp/Vs ratio for this layer range 0.25–0.28 suggesting that it is felsic to intermediate in composition (Prakasam and Rai 2003). Low heat flow value 35 mWm2 implies LAB at least 175 km deep (Gupta et al. 1993; Roy and Mareschal 2011). Surface wave data supports a mantle lid at least 200 km thick (Oreshin et al. 2011). Similar thickness for the lithospheric keel is proposed by Gupta et al. (2003). From joint inversion of receiver function and surface waves (Bodin et al. 2013) place the LAB at ~150–200 km. As against, from joint inversion of Ps and Sp receiver functions Kiselev et al. (2008) have argued that high-velocity keel is completely lacking beneath the Dharwar Craton. High resolution Ps imaging near Hyderabad area of the EDC obtained a depth of merely 65 km for the lithosphere-asthenosphere boundary (Reychert and Shearer 2009). Xenolith P-T of kimberlites and chemical tomography on garnet xenocrysts from the EDC provide a thickness of >200 km (Ganguly and Battacharya 1987; Nehru and Reddy 1989) and 190 km (Griffin et al. 2009) respectively.

82

2.4.1.3

2 Indian Shield

Crustal Structure Beneath the Bastar Craton

The Hirapur-Mandla (N-S) DSS profile across the Narmada towards the south extends into the Bastar Craton. The crust comprises three velocity layers. The upper crustal layer has a thickness of 8 to 90 tr–10 – tr tr tr–10 apt+sld.

– Ca0.8–1.4 Mg77–83 Fe22–15 Mg # 0.83–0.85 Ca45–47 Mg46–42 Fe12–7 Mg # 0.83–0.92

Mafic granulite 2–10.0 cm Melagabbro Meta-igneous/ layered

– 2–67 tr–36 – 2–54 tr–33 tr rtl+kst+apt+sld.

– Ca1 Mg65–83 Fe33–15 Mg # 0.82–0.92 Ca46 Mg39–46 Fe14–7 Mg # 0.82–0.92

– Ca1.4–1.5 Mg72–88 Fe26–10 Mg # 0.75–0.91 Ca45–48 Mg49–40 Fe6–12 Mg # 0.85–0.99

tr >80 tr–15 – – – tr–05 ol+apt+sld.

Pyroxenite 0.3–2.0 cm Websterite Hypidiomorphic granular

Table 3.2 Summary of Petrological Characteristics of Xenoliths in dykes from WDCa

Fo90–91.5 Ca0.7–2.0 Mg89–90 Fe8.0–8.6 Mg # 0.90–0.94 Ca48–49 Mg47–48 Fe2–3 Mg # 0.93–0.94

85–62/76–92 05–20/6–20 08–15/1–2 0.5–3.0/0.5–1.0 – – – amp+cbn

Peridotite 0.4–3.0 cm Spinel lherzolite/Harzburgite-Dunite Allotriomorphic granular/Protogranularporphyroclastic

104 3 Xenolith Petrology

Ca12–15 Mg45–52 Fe38–33 Ca46–68 Na52–31 K2–1 Ca30–36 Na68–64 K2 Ca75 Na24 K0.5 –

–

–

Ca20–15 Mg30–39 Fe50–46 Ca30–35 Na69–63 K0.4–2.0 – – – – – – –

–

–

Ca19–20 Mg23–30 Fe48–53 – – – – – – – – – TiO2 wt.% 11.59–13.01 FeO wt.% 81.81–83.55 – –

–

Ca11 Mg50–72 Fe35–16 – – – – – – – – – –

Mg # 0.66–0.77

Cr2O3 wt.% 16–37

– – – – – – – – – – – – Al2O3 wt.% 32–53

ol: olivine, opx: orthopyroxene, cpx: clinopyroxene, gnt: garnet, plg: plagioclase, phlg: phlogopite, amp: amphibole, cbn: carbonate, apt: apatite, spd: sulphide, spnl/Ti-mag:spinel/Ti-magnetite, ilm: ilmenite, rtl: rutile, Fo: forsterite, tr: trace (aafter Dessai and Vaselli 1999; Dessai et al. 1999; 2004; unpublished data)

Spnl./Ti-mag.

Plg. III

Plg. II

Plg. I

Gnt.

3.3 Petrology of Xenoliths 105

106

3 Xenolith Petrology

Fig. 3.3 Olivine porphyroclast armoured by fine granular aggregates of same mineral species along with orthopyroxene, showing stretching, undulose extinction, deformation lamellae, and development of neoblasts along grain boundary (BXP). In many samples, olivine shows gradational features: strain shadows, deformation lamellae/kink bands, and at places a mosaic of neoblasts (ol: olivine, opx: orthopyroxene)

Fig. 3.4 Exsolutions of chromian spinel in orthopyroxene porphyroclast with curvilinear grain boundaries (polariser rotated by 22 ). Elongated, flattened grains of orthopyroxene at places, with exsolutions of clinopyroxene, define the foliation. Porphyroclasts of orthopyroxene and rare olivine are armoured by fine granular olivine and orthopyroxene

3.3.1.2

Crustal Xenoliths

The mafic xenoliths consist of clinopyroxene, orthopyroxene, plagioclase, and garnet as the main mineral phases with accessory proportions of apatite, spinel, rutile, quartz, and phlogopite. A plot of modal proportions of xenoliths in pyroxeneplagioclase-garnet triangular diagram is presented in Fig. 3.10.

3.3 Petrology of Xenoliths

107

Fig. 3.5 Foliated peridotite defined by stretched and deformed olivine and orthopyroxene porphyroclasts and recrystallised neoblasts (BXP). The neoblasts are polygonal at places they too show deformation effects in the form of strain shadows suggesting synkinematic recrystallisation

Fig. 3.6 Foliated dunite with recrystallised olivine neoblasts largely with rectilinear grain boundaries meeting in 120 triple point junctions and sub-idiomorphic chromites (BXP). Most samples are coarse grained and show equigranular textures. Chromites occur as individual crystals or segregations or lenticels along the foliation or occur at an angle to it (ol: olivine, crt: chromite)

Most xenoliths display compositional layering on a scale of mm to cm (Fig. 3.11). A photomicrograph depicting modal layering is presented in Fig. 3.12. A few exhibit foliation defined by mafic minerals primarily pyroxene/garnet and plagioclase. Most are equigranular. Plagioclase-rich samples exhibit xenomorphic granular or polygonal granoblastic microstructures, the result of annealing in a static environment. The clinopyroxene dominated xenoliths exhibit porphyroclastic and meta-igneous textures. The porphyroclasts are armoured by granular neoblasts (Fig. 3.13). Some samples display well-developed foliation defined by preferred alignment of grain clusters. Still others show well-developed micro-layering defined by felsic and mafic minerals (Fig. 3.14). Plagioclase within the layering shows stretching, bending, granulation, and deformation lamellae. It is in places replaced by scapolite. Clinopyroxene is granular, mosaic-textured, pale green, and weakly pleochroic. Orthopyroxene occurs as exsolution lamellae in clinopyroxene porphyroclasts (Fig. 3.15) and as discrete grains.

108

3 Xenolith Petrology

Fig. 3.7 Layering in dunite defined by subhedral chromites (PPL). The grain aggregates tend to form pod-shaped segregations which could be lenticular or sack-form. The chromites exhibit occluded silicate texture

Fig. 3.8 Wehrlite cumulate least deformed partly recrystallised in places. The concentration of opaques is variable in different domains with portions completely free from mafics (BXP)

Orthopyroxene contains exsolution lamellae of spinel (Fig. 3.16). Garnet is pale pink and invariably altered along borders and fractures to cryptocrystalline kelyphite at places with double rims (Fig. 3.17). Garnet exhibits double rims of cryptocrystalline kelyphite, the inner rim is feathery and finer grained than the outer one (Dessai et al. 2004) that consists of a symplectic intergrowth of orthopyroxene, plagioclase, and spinel. Clinopyroxene and garnet contain exsolutions of rutile that show inclined extinction (e.g. Griffin et al. 1979). Feldspar is represented by calcic plagioclase; in a few cases, it contains randomly oriented needles of sillimanite. Many samples exhibit layering of mafic and felsic domains on a millimetre scale wherein individual layers display granular microtextures. Spinel, rutile, magnetite, and apatite occur in trace amounts. Apatite occurs as inclusions in most minerals. Phlogopite, kaersutite, K-feldspar, apatite, and sulphides vein the granulites. Rarely, granulites xenoliths display gneissose banding due to deformation, defined by granular clinopyroxene and plagioclase (Fig. 3.18).

3.3 Petrology of Xenoliths

109

Fig. 3.9 Heteradcumulate texture of clinopyroxenite (BXP). Large, irregular, amoeboid oikocrysts of orthopyroxene poikilitically enclose a number of irregular to rounded clinopyroxene, rare olivine, and chromite

Fig. 3.10 Modal compositions of xenoliths in the pyroxene-plagioclase-garnet triangular diagram

Most mineral phases display spongy rims suggesting partial melting due to reheating during transport (Fig. 3.19).

110

3 Xenolith Petrology

Fig. 3.11 Megascopic modal layering defined by pyroxene and plagioclase feldspar in mafic granulite xenoliths in lamprophyre

The xenoliths consist of clinopyroxene, orthopyroxene, plagioclase, and garnet as the main mineral phases with accessory proportions of phlogopite, apatite, and opaques. Felsic granulites exhibit gneissic texture with bands/layers rich in plagioclase alternating with those of clinopyroxene and garnet. Amphibole is present in a few xenoliths and is in equilibrium with the anhydrous phases. The plagioclase is annealed and anhedral whereas clinopyroxene shows stretching and elongation within the plane of layering. Rarely, the xenoliths show protomylonitic texture with stretched and bent porphyroclasts of plagioclase in a finer matrix of plagioclase. It is less calcic than that in mafic granulites. In a few xenoliths, it is replaced by scapolite (Dessai and Vaselli 1999). Compound xenoliths are less abundant with interlayering of mafic granulite and clinopyroxenite. At places, clinopyroxenite cross cuts the granulite layering. The latter is granoblastic whereas the clinopyroxenites exhibit cumulate textures distinctly seen in the xenoliths from the CITZ.

3.3.1.3

Mineral Chemistry

Pyroxenes (Fig. 3.20) are aluminian ferroan diopside (Morimoto et al. 1988), with generally higher Alvi/Aliv and lower Ti/Al ratios than those in clinopyroxenites and

3.3 Petrology of Xenoliths

111

Fig. 3.12 A photomicrograph-mosaic of a microsection of mafic granulite depicting the modal layering defined by pyroxene and plagioclase (PPL)

Fig. 3.13 Stretched and bent porphyroclasts of plagioclase armoured by granular neoblasts (BXP)

websterites which is characteristic of high-pressure environments. In most samples from CITZ (e.g. Nrd/19, Nrd/20, Nrd/22), the clinopyroxene is fairly high in Jd which varies from >6–8% with Jd/Ts ratio > 0.5. The orthopyroxene is an aluminian ferroan enstatite. Most garnets from the xenolith suite are rich in pyrope (Fig. 3.21). Many show similarities to those from eclogites in kimberlites and in gneisses (e.g. Coleman et al. 1965).

112 Fig. 3.14 Micro-layering in mafic granulite defined by alternating layers (highlighted by dashed lines) of pyroxene (coloured) and plagioclase (colourless/grey lamellae) which shows deformation due to stretching and bending of porphyroclasts (BXP)

Fig. 3.15 Exsolutions of orthopyroxene in a deformed clinopyroxene porphyroclast (opx orthopyroxene, cpx clinopyroxene) (BXP)

Fig. 3.16 Exsolutions of brown spinel in orthopyroxene from mafic granulite (BXP). (Note resorbed grain boundaries)

3 Xenolith Petrology

3.3 Petrology of Xenoliths Fig. 3.17 Double rims of kelyphyte on garnet in garnet granulite (BXP). The inner rim is cryptocrystalline whereas the outer one exhibits symplectic intergrowth between orthopyroxene and plagioclase/spinel

Fig. 3.18 Gneissic texture formed due to deformation of mafic granulite (PPL)

Fig. 3.19 Resorption of minerals along grain boundaries as a result of decompression partial melting during transport in the host (BXP)

113

114

3 Xenolith Petrology

Fig. 3.20 Compositions of pyroxenes from the xenoliths in the Wo-En-Fs face of the pyroxene quadrilateral

The mafic xenoliths consist of two pyroxene granulites with or without garnet. Most xenoliths display compositional layering on a scale of mm to cm. A few exhibit foliation defined by mafic minerals and plagioclase. Most are equigranular. The dominant texture is polygonal granoblastic, the result of annealing in a static environment. The xenoliths consist of clinopyroxene, orthopyroxene, plagioclase, and garnet as the main mineral phases with accessory proportions of phlogopite, apatite, and opaques. Pyroxene compositions are aluminian ferroan diopside (Morimoto et al. 1988), with generally higher Alvi/Aliv and lower Ti/Al ratios than those in clinopyroxenites and websterites which is characteristic of high-pressure environments. In most samples from CITZ (e.g. Nrd/19, Nrd/20, Nrd/22), the clinopyroxene is fairly high in Jd which varies from >6–8% with Jd/Ts ratio > 0.5. The orthopyroxene is an aluminian ferroan enstatite.

3.3 Petrology of Xenoliths

115

Fig. 3.21 Atomic proportions of Ca-Mg-Fe of garnets from xenoliths. Group A: eclogites in kimberlites, Group B: eclogites in gneisses and Group C: eclogites in blueschists (modified after Coleman et al. 1965)

3.3.2

Xenoliths from Kutch

Peridotite xenoliths in nephelinites from Kutch were first reported by De (1980) although it was a mention en passant. A fairly detailed account of spinel lherzolites from Kutch was provided by Mukherjee and Biswas (1988). The xenoliths range in size from 5–30 mm. The mineralogy is dominated by four phases, namely olivine, orthopyroxene, clinopyroxene, and aluminous spinel in order of abundance. Olivine varies from 45–62%, orthopyroxene from 10–20%, clinopyroxene 3–12%, and reddish brown aluminous spinel from 0.5–5.0%. The rocks are fine grained and unfoliated. They are generally allotriomorphic granular, however, at places the texture is porphyroclastic and are akin to chrome diopside (Type-I) lherzolites of Wilshire and Shervais (1975) and Type-I xenoliths of Frey and Prinz (1978). The olivine porphyroclasts contain trails of fluid inclusions. They occur as anhedral, inequigranular crystals (1–4 mm) and show strain shadows, undulose extinction, polygonised bands, and bent cracks. The prophyroclasts are associated with secondgeneration olivine which are polygonal and densely clustered neoblasts. Often these are found bordering orthopyroxene. Orthopyroxene is subhedral (0.9–3.0 mm) with serrated borders. Clinopyroxene is represented by chrome diopside and occurs as pale green anhedral grains (1–3 mm) which contain tiny bubble-like inclusions possibly of melt. The grain

116

3 Xenolith Petrology

boundaries are often cuspate and fritted whereas the central core of the grains is relatively clear. Both pyroxenes are devoid of exsolution lamellae. Reddish brown aluminous spinel occurs in association with orthopyroxene and clinopyroxene at places showing ‘holly-leaf’-like form (Mercier and Nicolas 1975). The minerals are traversed by veins of pale brown glass. Most minerals show spongy, fritted margins suggesting decompression melting and xenoliths–magma interaction during ascent. Unlike xenoliths from other localities in India, the ones from Kutch are devoid of amphibole and phlogopite.

3.3.2.1

Mineral Chemistry

Diopsides from Kutch xenoliths contain HREE and MREE two to four times the Primitive Mantle (PM) values and the LREE and related elements show a spread from one to ten times PM (Karmalkar et al. 2005). One group of xenoliths is characterised by Ti, Zr, and/or Sr anomalies. A second group shows flat HREE and MREE patterns but is enriched in LREE, Sr, U, and Th and has negative Ti, Zr, and Nb anomalies but no Sr anomalies. A third group shows a decrease from HREE to MREE and variable enrichment of LREE, Sr, U, and Th. The negative Nb, Zr, and positive Th anomalies of xenoliths are complementary to positive Nb, Zr, and negative Th anomalies exhibited by the host basanites (Fig. 3.22). The elemental ratios such as Zr/Nb, Th/La, Th/Nb, and La/Nb of the clinopyroxenes are higher than those of the host basanites and are characteristic of EMI OIB source (ibid.).

Fig. 3.22 Primitive mantle normalised trace element patterns of clinopyroxenes from Kutch xenoliths compared with those of host basanites (after Karmalkar et al. 2005)

3.3 Petrology of Xenoliths

3.3.3

Xenoliths in Kimberlites

3.3.3.1

Xenoliths from Bundelkhand Craton

117

Xenoliths from Bundelkhand Craton are hosted by two different intrusive rocks; the kimberlites which are Proterozoic and the tholeiitic dykes of Deccan Traps age. Within the latter, xenoliths occur within alkaline caught-up blocks entrained by tholeiitic dykes. The kimberlites at Majhgawan and Hinota (25 km SW of Panna, Madhya Pradesh) contains rare xenoliths and fragments of igneous rocks (e.g. Scott-Smith 1989) along with macrocrysts of olivine pseudomorphs set in a fine grained (brown) groundmass. The xenoliths contain abundant olivine pseudomorphs. Macrocrysts of olivine range in size from 30 vol.%), garnet (4–8 vol.%)  phlogopite (5–7 vol.%)  sulphides. Orthopyroxene occurs as exsolution blebs in clinopyroxene. The pyroxenites belong to the A-augite group equivalent to the Group II xenoliths of Frey and Prinz (1978). The pyroxenites exhibit cumulate textures, rarely clinopyroxene shows preferred orientation defining a weak layering. Secondary minerals have formed by reaction with silicate melt. Glass veins and melt pockets contain tiny secondary clinopyroxene. Phlogopite is anhedral. Sulphides are represented by pyrite, chalcopyrite, and pyrrhotite, which occur as megacrysts (>3 mm). The opaques are represented by magnetite and titanomagnetite. Websterites (0.3–1.5 cm) consist of clinopyroxene (~60 vol.%), orthopyroxene (30 vol.%), opaques (6–10 vol.%), garnet (trace-4 vol.%)  phlogopite (2–5 vol. %)  sulphides. They are veined by sulphides, carbonates and contain apatite. Clinopyroxene exhibits weak planar fabric along with spinel. The mafic granulite xenoliths can be described as garnet bearing plagioclase-poor two pyroxene granulites. Most show modal layering defined by alternations of clinopyroxene + garnet and plagioclase. Texturally, they vary from granoblastic to rare meta-igneous types. Rarely, the texture is porphyroclastic with bent porphyroclasts of plagioclase with the development of subgrains indicative of deformation. These textures are superimposed on the overall granoblastic fabric of the rock. Most minerals display abundant fluid inclusions. The garnet granulites are composed of clinopyroxene (2–67 vol.%), orthopyroxene (trace-36 vol.%), plagioclase (2–54 vol.%), and garnet (trace33 vol.%)  rutile (trace). The rocks are veined by phlogopite, apatite, and sulphides. Garnet is invariably fractured and altered along fracture and borders to

118

3 Xenolith Petrology

cryptocrystalline kelyphite. It contains inclusions of clinopyroxene and is also veined by it. Plagioclase is anhedral and at times deformed due to stretching, elongation, and is granulated along borders. The second-generation plagioclase is free from deformation. Orthopyroxene occurs both as exsolutions in clinopyroxene and as recrystallised grains. Phlogopite is anhedral, intergranular, and its proportion varies considerably in different samples. Spinel is chromian and anhedral. Composite xenoliths are very rare. They consist of mafic granulites with lenticular segregations and veins of clinopyroxene. The granulitic portion is equant granoblastic whereas the clinopyroxenite veins are xenomorphic granular. The contact between the two is generally irregular yet sharp. The clinopyroxene in veins is compositionally similar to that in clinopyroxenite xenoliths described above.

3.3.3.2

Mineral Chemistry

Clinopyroxene from pyroxenite is an aluminian ferroan diopside (Ca45–43Mg43–46Fe10–9, Mg #: 0.84–0.82). Most clinopyroxenes display Jd/Ts ratio > 0.5 with Jd component varying from 5–6% indicative of their eclogitic affinity. Clinopyroxenes contain abundant fluid inclusions. Orthopyroxene is aluminian ferroan enstatite (Ca1–43Mg80–81Fe18–17, Mg #: 0.84–0.82), CaO is 6–8% with Jd/Ts ratio > 0.5. Garnet (Ca14–15Mg43–47Fe26–24) shows close compositional similarity to that in crustal eclogites. Orthopyroxene which occurs as exsolutions in clinopyroxene and as discrete grains has the composition represented by Ca1Mg67–79Fe31–20, (Mg #: 0.76–0.60) and therefore can be identified as ferroan enstatite. Unaltered garnet (Ca14–15Mg48–47Fe26–24) shows close similarity to garnet in crustal eclogites. Plagioclase is calcic and shows moderate compositional variation (Or1.6–0.1Ab50–0.5An48–99).

3.3.3.3

Lherzolite Xenoliths from the EDC

The xenoliths from the EDC are represented by garnet harzburgite, garnet lherzolites, garnet wehrlite, and olivine clinopyroxenites. Detailed petrography and mineral chemistry are provided by Ganguly and Battacharya (1987). The modal mineralogy is represented by olivine, orthopyroxene, clinopyroxene, garnet, and spinel (Table 3.3) and most of the xenoliths contain phlogopite. In addition to richterite, serpentine group minerals and grains of ilmenite are the other associated phases. The xenoliths exhibit coarse granoblatic texture; some are more equilibrated with adjoining orthopyroxenes showing grain boundaries which meet in 120 triple points, suggesting annealing in the absence of significant stress. At places, the

3.3 Petrology of Xenoliths

119

Table 3.3 Summary of petrological characteristics of peridotite xenoliths in nephelinites from Kutch, and in kimberlites from Bundelkhand, EDC, and Bastar cratons Peridotite (Kutch, Aravalli)a,b 0.5–2.0 mm Spinel Lherzolite

Peridotite (Bundelkhand)c >5–10 mm Harzburgite (?)

Allotriomorphic granular 50–60 15–20 02–13 0.5–4.0 – – Cr-spnl Fo89–91

Allotriomorphic granular (?) 43.3& – – – – – 56.6 Fo91–92

Opx.

Ca0.7–1.0 Mg89–88 Fe9–10 Mg # 0.89–0.91

– – – –

Cpx.

Ca47–49 Mg48–49 Fe5–4 Mg # 0.90–0.91

– – – –

Gnt

–

–

–

–

–

–

–

–

–

–

– – Al2O3 wt.% Cr2O3 wt.% –

– – – – –

Xenolith type/ group Size Protolith Texture Ol Opx Cpx Spnl. Gnt Phlg Others Ol.

Spnl.

Peridotite (EDC)d 0.5–2.0 mm Garnet Lherzolites Granoblastic P P P/A P P P/A P/A Fo91–94 Ca0–0.5 Mg93 Fe6 Mg # 0.92–0.93 Ca45–38 Mg49–54 Fe4–6 Mg # 0.91–0.93 Adr 0.8–0.36 Grs 9.61–11.83 Prp 75.41–74.96 Sps 0.47–0.78 Alm 13.89–9.94 Uv 0.37–2.10 Kn 0.04–0.01 15.83–20.19 49.14–53.86 –

Peridotite (Bastar)e 2–10 mm Garnet Lherzolites Allotriomorphic/ Protoclastic 30 12 25 03 15 – – – – – – – Mg # 0.89 – – – – – – – – – – – – – –

ol: olivine, opx: orthopyroxene, cpx: clinopyroxene, gnt: garnet, spnl: spinel, Cr-spnl: chromespinel, P: present, A: absent, Fo: forsterite, adr: andradite, alm: almandine, grs: grossular, uv: uvarovite, sps: spessatite, prp: pyrope, kn: knorringite, tr: trace, &: xenoliths/igneous fragment3 (?) (aDessai et al. 1999; bKarmalkar et al. 2000; 2005; cScott-Smith 1989; dGriffin et al. 2009; eBabu et al. 2009)

120

3 Xenolith Petrology

xenoliths are foliated and in these the olivines show shearing effects possibly related to forceful injection along conduit walls (e.g. Mercier 1979). Rare xenoliths show blasto-mylonitic fabrics with the formation of olivine neoblasts (Ganguly and Battacharya 1987). Olivines invariably occur as granular aggregates and rarely as recrystallised neoblasts. By and large, they are free from strain effects, except the larger ones which may show undulose extinction. They are invariably sepentinised except where they are included in orthopyroxene. Mg/(Mg + Fe) (Mg #) varies from 0.94 to 0.91, those from the matrix show lower Mg #s, as low as 0.82. Ni content of olivines varies from 0.30 to 0.35 wt.%. Olivines from kimberlite matrix display rims of serpentine around them whereas the ones with lower Mg #s are surrounded by groundmass minerals. This led Ganguly and Battacharya (1987) to suggest that serpentinisation took place prior to their incorporation in the kimberlite magma, thus preventing chemical exchange. Orthopyroxenes exhibit a small range of Mg # from 0.94 to 0.91. Larger grains of orthopyroxene contain inclusions of olivine and garnet. Orthopyroxenes from the kimberlite matrix have similar Mg # as those from the xenoliths but contain lower concentration of Al2 O3 and Cr2O3. Clinopyroxenes occur as smaller grains and their modal proportion is less as compared to the other mafic phases. Majority of them are anhedral, their shape appears to have been defined by the intergranular of adjacent minerals. They are free from inclusions and do not contain any other mineral phase. This suggests that their crystallisation is post olivine, orthopyroxene, and garnet formation. Chemically, they could be categorised as chrome diopsides. Phlogopites are confined to the kimberlite matrix or at times are found to mantle garnet and orthopyroxene. Fractures within garnet are at times replaced by phlogopite. They could be therefore, classified as secondary and are formed by reaction with kimberlite host or may have even crystallised from kimberlite magma concomitant with infiltration into the xenoliths. Majority of phlogopites show reverse pleochroism. Reverse phlogopites are considered to have crystallised in situ in mantle peridotites whereas the normal ones are formed from kimberlite magma (Farmer and Boettcher 1981). However, it has been shown that the higher Fe of reverse phlogopites cannot be reconciled with mantle origin in in situ peridotites (Ganguly and Battacharya 1987). Spinel occurs as minute grains within the intergranular of other phases such as olivine, orthopyroxene, and garnet. At places, it also occurs as veinlets and in reaction coronas around garnet. Most spinels appear to be post-xenoliths phases and may have formed as a breakdown product of garnet during ascent and emplacement of kimberlite.

3.3.3.4

Eclogite Xenoliths from EDC

In addition to the ultramafic xenoliths described above, the kimberlite also hosts eclogite xenoliths which have been described from the Lattavam kimberlite pipe.

3.3 Petrology of Xenoliths

121

They consist of garnet, omphacite, phlogopite, corundum, kyanite, rutile and occasionally quartz, orthopyroxene, and opaques. Detailed mineragraphy of these is provided by Ganguly and Battacharya (1987), Patel et al. (2009).

3.3.3.5

Xenoliths from Bastar Craton

Rare mantle and crustal xenoliths occur within a kimberlite (150–200  70–100 m) exposed to the south of Mainpur. Mantle xenoliths vary in size from 2–10 cm and can be categorised as garnet lherzolites. They are porphyroclastic, but some of them show development of strain-free neoblasts. in places, relict porphyroclastic texture is seen within a recrystallised dunite. Some of them comprise olivine, pyrope, chrome diopside, and spinel. Large subgrains of garnet occur along the contacts of the peridotite xenoliths. Garnet invariably shows kelyphytic alteration along grain boundaries. Stumpy amphibole occurs within the reaction aureole between the xenoliths and the glassy kimberlite groundmass. A megacrysts of olivine and diopsidic clinopyroxene which have a reaction relation with kimberlite, also occur. They show the development of kink bands and undulose extinction due to deformation. This suggests that they are xenocrystic phases from the wall rock which have been carried by the erupting kimberlite magma. Deformation-free chrome diopside in the form of polygonal grains occur along cracks and grain boundaries. Olivine commonly occurs both as xenocrysts and also as euhedral microphenocrysts along with subordinate chrome diopside, phlogopite, and rare spinel. The olivine xenocrysts are subhedral to anhedral and show resorption along grain boundaries referred to as pelloidal-texture olivine (Lehmann et al. 2010). Phlogopite is strongly pleochroic in shades of brown and poikilitically encloses apatite, perovskite, and wadeite. Spinel and perovskite occur as constituents of the groundmass. Felsic xenoliths are relatively less common and vary in size from 7.0 km/s) in seismic profiles. The pressure estimates for pyroxenites (0.57–1.05 GPa) provide a spatial context for these rocks. The pressures are consistent with the lower crustal depths. This relationship could be interpreted to be due to magmatic additions to the deep crust. The interstratified zones are characterised by seismic reflectors with high Vps (7.0–7.5 km/s) as for example along the Hirapur-Mandla DSS transect (Fig. 4.3b). In certain sections along this transect, especially between Damoh and Katangi, the upper crust is missing and the granulites are overlain by the late Proterozoic Vindhyan sediments (Mahadevan 1994). This has been suggested to be the result of transport of lower crustal rocks as slivers along thrust faults that run oblique to the general trend of the rift (Mahadevan 1994; Dessai et al. 2009). From several seismic profiles across the CITZ, it was concluded that the crust in this region is about 40 km thick and the sub-Moho velocities are in the range of 7.8–8.2 km/s (Kaila and Krishna 1992). An E-W horst was identified between the Narmada north and south faults (Tewari et al. 2001). A uniform lower crust with Vp of 6.8–6.9 km/s has also been suggested (Reddy et al. 1997; Tewari et al. 2001; Murthy et al. 2004). The RF investigations of the Bundelkhand Craton reveal a two-layer crust with a 15 km thick upper crust (Vs: 3.5 km/s) followed by the lower crust which extends

Fig. 4.3 Schematic cross-section of the continental crust and Vp profile along the (a) Mehmadabad-Billimora (Kaila et al. 1981a) and (b) Hirapur-Mandla DSS transects (Kaila 1988) beneath northern region of the WDC, based on the xenolith data and the seismic velocities of the respective DSS transects (Kaila et al. 1979, 1981a, b; Reddy and Rao 2003)

130 4 Lithosphere Architecture

4.2 Crustal Structure: Seismic Studies

131

Fig. 4.4 Joint inversion model for the JHN station from the Bundelkhand Craton. Bold black line: model velocity depth profile, grey shade: 2σ confidence limit (modified after Julia et al. 2009)

down to 37.5 km (Vs: 3.8 km/s). Sub-Moho velocities (Vs) are relatively low, about 4.4 km/s. A representative velocity profile of station JHN (Julia et al. 2009) is presented in Fig. 4.4.

4.2.3

Central Region of WDC

In the interior of the craton at middle latitudes (e.g. Kelsi-Loni DSS transect, Kaila et al. 1981a; Fig. 4.5a), the crust was suggested to consist of two layers with the Conrad discontinuity at a depth of 26 km. However, a re-examination of the DSS data (Dessai et al. 2004) suggested that the crust beneath the craton in general, can be divided into three velocity layers with two distinct breaks in Vp at 12 and ~26 km depth. The supra-Conrad layer can be divided into the upper crust (Vp: 5.7–6.2 km/s, from 2 to 12 km) and the middle crust (Vp: 6.4–6.6 km/s, from 12 to 25 km). Xenolith data permit interpretation of this layer as consisting of felsic granulites. Both the Vp values and the pressure estimates (0.6–0.8 GPa) on xenoliths support this deduction. DSS profiles between Udupi and Kavali revealed a change in crustal thickness across the Closepet Granite from 38–41 km to the west to 34–36 km to the east. A crustal thickness of 40 km was proposed on the basis of the N-S (Kuppam-Palani) transect (Reddy and Rao 2003). Krishna et al. (1991) noted complex high- and low-velocity layering from the Koyna I and Koyna II profiles with a crustal thickness of ~40 km. The sub-Conrad layer (depth 26–40 km) along both the Kelsi-Loni and UdupiKavali transects (Fig. 4.5b) shows Vps (6.76–7.26 km/s) characteristic of mafic granulites (e.g. Rudnick 1992). The lower values of Vp < 0.7 suggest the presence of felsic granulites.

Fig. 4.5 Schematic cross-section of the continental crust and Vp profile along the (a) Kelsi-Loni (Kaila et al. 1981b) and (b) Udupi-Kavali DSS transects (Kaila et al. 1979) constructed on the basis of the xenolith data and the seismic velocities along the DSS transects

132 4 Lithosphere Architecture

4.2 Crustal Structure: Seismic Studies

133

The dominance of mafic lithologies below this level is substantiated by the higher pressure estimates (0.7–1.6 GPa) on the mafic granulite xenoliths and is supported by the higher values of Vp in the lower part, characteristic of these lithologies (e.g. Rudnick 1992; Rudnick and Gao 2003, 2014). The composite xenoliths with pyroxenites intercalated with the granulites, permit to suggest its interstratified nature. Strong reflectors in this zone are observed in a recent seismic transect further south in the craton, between Chikmagalur and Perur. From the xenolith and seismic data, it is clear that the lower crust beneath the western Dharwar Craton contains a significant component of ultramafic rocks. In this respect, the lower crust beneath the craton differs considerably from the model of the continental crust presented by Wedepohl (1995) although gross layering and the total thickness of the crust are similar. However, the seismically estimated crustal thickness in this western part of the Dharwar Craton diminishes nearly by one third and at places, by half (of the total thickness) towards the continental margin. In the vicinity of Mumbai, for example, the geophysically estimated crustal thickness is only about 18–20 km (Kaila 1988).

4.2.3.1

Structure from S-wave Velocity Function

Inversion of S-wave velocity depth profiles from receiver functions and surfacewave dispersion velocities (e.g. Passyanos 2005) reveal that a layer of high S-wave velocity (Vs: 4.1–4.3 km/s) occurs in the deep crust under the Proterozoic areas and some Archaean terrains of the Indian shield (Julia et al. 2009). These velocities are higher than those (3.8–3.9 km/s) in adjacent cratons of Bundelkhand in the north and the EDC in the south. This high-velocity layer could be attributed to mafic cumulates or restites, on the basis of the Proterozoic granulite xenoliths entrained by the late alkaline dykes from the Deccan Traps (Dessai et al. 2004). Lithologies with S-wave velocities in the mafic range are proposed at a depth of ~15 km by Kiselev et al. (2008) and in their opinion, the WDC crust is an exception to the norm (Durrheim and Mooney 1994) which stipulates a thin crust of felsic composition for the Archaean terrains. Crustal thicknesses have been estimated (Julia et al. 2009 and references therein) using the receiver-function approach. A crustal thickness of 50 vol.%) along with clinopyroxene and garnet that vary in proportion from trace to as much as 25 vol.% along with amphibole  epidote  sillimanite that may be present to the extent of up to 10 vol.%. The fabric of the rocks is, however, more crucial as is apparent from the comparison of Vp values observed in the two transects almost at right angles to each

4.2 Crustal Structure: Seismic Studies

139

Fig. 4.11 Schematic cross-section of the continental crust and Vp profile along the Kuppam-Palani DSS transect (Reddy and Rao 2003)

other. In the E-W DSS transect (Udupi-Kavali) the Vp of felsic granulites/gneisses is almost constant at 6.15 km/s (from 12 to 23 km depth). However, along the N-S transect the Vp for the same depth zone shows values of 6.0–6.2 km/s. The fabric of granitic gneisses and felsic granulites is primarily responsible for this seismic anisotropy. For instance, the calculated Vp values of 6.2 km/s correspond to granitic orthogneisses. The measured seismic velocities correlate with density, and modal mineralogy especially to the presence of garnet and sillimanite (e.g. den Berg et al. 2005) both of which are fairly abundant in the gneisses of the transition zone. The presence of mica, amphibole and sillimanite which are strongly anisotropic minerals have also aided in this process. It is apparent therefore, that the regional N-S grain (regional trend of foliation) of the terrain is paralleled by the strong N-S foliation of the granitic gneisses/felsic granulites at depth. The preferred alignment of clinopyroxene and amphibole within the strong penetrative foliation of the felsic metapelitic granulites (Dessai et al. 2004) and the fissility of the amphibole ( epidote) gneisses (e.g. Mahadevan 2003) is responsible for the observed decrease of Vp in felsic lithologies despite the high

140

4 Lithosphere Architecture

Fig. 4.12 Generalised schematic longitudinal cartoon of the continental crust beneath the western Indian shield constrained by integrating arguments based on the xenolith data and the seismic velocities for the DSS transects Mehmadabad-Billimora, Kelsi-Loni, Udupi-Kavali, and KuppamPalani (Fig. 4.1). Note slivers of lower crust transported along thrust faults to account for granulites exhumed in certain sections of CITZ, e.g. Damoh-Katangi section of the Hirapur-Mandla DSS transect. The supracrustal rocks have been omitted

abundance of garnet (e.g. Rudnick and Fountain 1995). The seismic anisotropy is of more relevance in the transition zone gneisses and metapelites due to the presence of strongly anisotropic minerals such as mica, amphibole and sillimanite. It has been shown how Vp is related to single-crystal velocities and modal compositions (e.g. ibid.). The combined effect of the fabric and the modal composition of the rocks (e.g. presence of anisotropic minerals) have overcome the expected increase of Vp in felsic lithologies containing garnet. The presence of fluids such as CO2 as the agents responsible for the high conductivity of this zone, does not seem likely as there would have been a drop in Vp (e.g. Durrheim and Mooney 1994) instead of the observed increase. In fact, the abundance of CO2 is high enough to influence Vp and the charnockites would have higher Vp than their precursors (Griffin per. com.). Thus, the proposition of an additional layer for the deep crust merely based on the N-S DSS transect is not warranted. Lithologically, the rocks of this low-velocity zone are contiguous with the rocks below and are petrologically and chemically similar in composition. This is also supported by the pressure estimates (0.6–0.8 GPa) of the felsic granulite xenoliths (Dessai et al. 2004). Integration of geophysical and xenolith data reinforces a three-layered structure for this terrain and suggests that the seismic anisotropy within the same lithology has a strong fabriccontrol. The visualised structure is depicted by means of a cartoon (Fig. 4.12). In the RF studies, SGT shows ~45 km thick crust which in certain stations reduces to ~35 km, possibly due to the proximity to the continental margin. The velocity depth profile for the station KOD (Fig. 4.13) (Julia et al. 2009) depicts the broad velocity picture of the SGT. The crustal section can be divided into three velocity layers. The upper crust from 12.5–15.0 km (Vs: 3.5 km/s), the middle crust up to 25 km depth (Vs: 3.9 km/s) followed by the lower crust up to the Moho (Vs:

4.2 Crustal Structure: Seismic Studies

141

Fig. 4.13 Joint inversion model for the KOD station from the SGT. Bold black line: model velocity depth profile, grey shade: 2σ confidence limit (modified after Julia et al. 2009)

4.1 km/s). The significant point here is the low-velocity layer (Vp: 6.15 km/s) encountered in active seismic profile by Reddy et al. (2001) is not observed in the passive seismic data. The sub-Moho velocities vary from 4.70–4.54 km/s.

4.2.8

Offshore Crust of Arabian Sea

The western Indian continental shelf and the region off it, namely the Arabian Sea abyssal plain, (~25
N to ~10
S) in general, reveal a crustal structure which is distinctly different from that observed either on the continent or in a typical ocean basin (Fig. 4.14). Naini and Talwani (1982) distinguish two basins on either side of the Chagos-Laccadive Ridge. The western basin is characterised by oceanic crust with sea-floor spreading magnetic anomalies. The eastern basin has no distinctive magnetic features indicative of its non-oceanic origin. Beneath a layer of sediments (here referred to as oceanic layer 1) and volcanic rocks (oceanic layer 2), there is no velocity layer corresponding to the continental crust (i.e. layer of Vp: 5.2–5.8 km/s). The absence of this crustal layer has been attributed to the transitional nature of the crust (Closs and Hinz 1967; Harbison and Bassinger 1973; Naini and Talwani 1982; Chekunov et al. 1984). Chekunov et al. (1984) further suggested that the density distribution is similar to sub-oceanic crust and that the crust was different than that of the Indian peninsula to the northeast. However, neither the lithology nor the crustal types have been identified. Specific details of the crustal structure are discussed below.

142

4 Lithosphere Architecture

Fig. 4.14 LaccadiveChagos Ridge in the Arabian Sea (L: Laxmi Ridge, LC: LacccadiveChagos Ridge, S: Seychelles, M: Madagascar, NER: Ninety East Ridge, AUST: Australia) (modified after Dietz and Holden 1970; Whitemarsh 1974)

4.2.8.1

Laxmi-Laccadive Ridge Section

At the Laxmi Ridge in the north and the northern part of the Laccadive Ridge further south, beneath the oceanic layer 2, a 4.5 km thick crustal layer displays a Vp of 6.2 km/s. Such a layer does not occur in typical oceanic crust. Xenolith data from the region in the vicinity of Mumbai to the east enable to interpret this velocity layer as representing topmost part of the middle crust, observed beneath continental crustal sections (Fig. 4.15). Alternatively, it could also be interpreted as a diffuse transition between the upper and middle crust. The latter possibility is more likely as seen from Vp values. The measured Vp values of 6.2 km/s permit interpretation of this layer to consist of felsic granulites or paragranulites (e.g. Christensen 1996). The low Vp values are consistent with the high geothermal gradient of the region (Biswas 1989; Pandey et al. 2015). A crustal layer with Vp of 5.2–5.8 km/s, normally present below oceanic layer 2, in most oceanic plateaux, oceans, forearc basins, continental shelves, and even in regions of extended crust, is conspicuous by its absence. This layer also occurs beneath most platforms and shields. The 4.5 km crustal layer is underlain by a layer with Vp of 7.2 km/s which is typical lower crustal velocity displayed by mafic granulites; restites after extraction of a partial melt (intermediate to felsic in composition). Coupled with the absence of a layer with Vp more than 7.6–8.0 km/s suggests that the so-called oceanic crust of the Laxmi-LaccadiveChagos Ridge is a transitional type of continental crust which is dominated by basic lithologies such as gabbro-derived mafic granulites, overlain by more felsic lithologies such as the felsic granulites, so abundant in the xenoliths population beneath the craton. This model is also supported by a large negative gravity anomaly over the LaxmiChagos Ridge. The structure of the crustal layers, their characteristic velocities, thicknesses, and the non-existence of a ‘typical’ Moho of velocity >7.6 km/s at

4.2 Crustal Structure: Seismic Studies

143

Fig. 4.15 Schematic crosssection of the transitional crust and Vp profile along the Laxmi-Laccadive section of the ridge

shallow levels as expected in an oceanic crust, all suggest an extended transitional type of continental crust. The total crustal thickness of the Laxmi-Laccadive Ridge comprises a thin middle crustal layer ( 7.8 km/s) has been interpreted as mantle, composed of restite peridotite. The reason for attributing the low-velocity zone to the mantle is that typical Moho with Vp > 7.8 km/s has been identified in the immediate vicinity in the adjoining back-arc and hence the low-velocity zone beneath the arc is attributed to low-velocity upper mantle and not to lower crust.

4.2.9

Offshore Crust of Bay of Bengal

Considerable work has been carried out on the eastern margin of the shield, off the east coast in the Bay of Bengal. Five structural units have been identified in this area. These are the marginal high, followed by a marginal basin, the graben succeeded by the central basin which has the 90
E ridge as its eastern boundary. The graben defines a pronounced gravity low and a large amplitude magnetic anomaly. The gravity low has been suggested to mark the boundary between the continental and oceanic crust (Rao and Rao 1986). The average seismic velocity of the basement layer of the graben is 6.22 km/s which was tentatively attributed to continental crust (Naini and Leyden 1973). Beneath a variable thickness of sediments (ca. 6–8 km), there is a thin volcanic layer (oceanic layer 2). This is followed by another thin layer with Vp 6.2–6.7 km/s which has been variously interpreted. Naini and Leyden (1973) attributed it to continental crust which appears to be nearer to reality. However, Curry (1991) referred to it as high density metasedimentary rock. As observed along the western margin, this layer is analogous to the middle crust. Such a layer does not occur in typical ‘oceanic crust’. It is probably made up of gneissic rocks—the remnants of the EGMB. It represents the transitional crust of this region subsequent to the break-up of India from Antarctica (Gopal Rao et al. 1997) during Cretaceous times. In this crustal section, the 5.2–5.8 km/s velocity layer, which normally occurs below the oceanic layer 2, is again missing. The middle crustal layer mentioned above is underlain by a thin prominent layer with Vp: 7.3–7.7 km/s which again has been variously interpreted except Mall et al. (1999) who have interpreted it as the lower crust. This layer is underlain by a layer which displays a Vp: > 8.1 km/s (by and large 8.3–8.5 km/s) which signifies the Moho. Its depth varies from 11–16 km up to17o N and increases to 32 km further north towards the Bangladesh shelf (Radhakrishna et al. 2010). The steep deepening of the basement rocks and in particular of the LAB, from 77 km in the south to 127 km in the north (Srinivas Rao et al. 2016), it is tempting to suggest that the off-shore crust has undergone considerable thinning attendant to extension. It is of a transitional nature, very similar to what is observed along the western margin. The observed crustal velocity structure suggests that the crust beneath the eastern Indian margin is not much different from that encountered along the western margin.

146

4 Lithosphere Architecture

The velocity of the crustal layer corresponds to mid-crustal granitic gneisses which directly overlie the Moho without any distinct lower crustal layer. This situation thus points to a transitional type of crust, similar to that observed along the western margin. The only exception between the two margins is the present-day configuration. Genetically, the western is an extensional margin now passive, whereas the eastern is a back-arc with an extensional setting within an overall active compressional regime.

4.3

Nature of the Subcontinental Moho

The Mohorovičić (Moho) discontinuity as originally defined is a sharp seismic discontinuity marked by a distinct increase in compressional wave velocity (Vp) from 7 to 8 km/s. Later, this discontinuity was considered to represent a compositional boundary between the lower crustal rocks above, and peridotitic ultramafic rocks of the mantle below (e.g. Ringwood 1975). Other interpretations for the Moho are also available. It has been regarded to represent a phase change from gabbroic rocks above to eclogites below (Ito and Kennedy 1970). Yet another interpretation suggests that it is a structural discontinuity that separates differently deformed rocks that vary in fabric (Nicolas et al. 1980). More commonly, however, the term Moho has been used both for seismic discontinuity and the base of the crust, i.e. the boundary that separates the silicic crustal rocks above from ultrabasic rocks below. These two concepts are commonly referred, respectively, as the ‘seismic Moho’ and the ‘petrological Moho’ (e.g. Griffin and O’Reilly 1987). Often in nature these two do not coincide, the petrological Moho generally lies above the seismic Moho.

4.3.1

Characteristics of Shallow Mantle

Although overall the mantle is considered to be ultramafic in composition, mineralogically it differs with tectonic environment. In intra-plate tectonic settings which have been subjected to young volcanism particularly during Cenozoic (Wilshire and Shervais 1975; Upton et al. 1983), the lower crustal xenoliths assemblage is invariably associated dominantly with spinel lherzolites and rarely with garnet lherzolites (Sutherland and Holis 1982). The situation is completely different when it comes to old cratonic areas. For instance, lower crustal xenoliths in kimberlite from cratonic areas are mostly associated with garnet lherzolites (Dawson 1980; Carlswell et al. 1984). It has been shown that garnet peridotites are stable under pressures as low as 1.5 GPa at temperatures of 900
C. Geothermobarometry on xenoliths as well as heat flow data indicate that such low temperatures are possible in cratonic regions wherein garnet lherzolite may be the dominant lithology beneath a crustal thickness of at least 45 km.

4.3 Nature of the Subcontinental Moho

147

In contrast, regions subjected to young basaltic volcanism display anomalously high geothermal gradients. In these regions, spinel lherzolites are believed to be the stable lithology up to depths of at least 55–60 km, as seen from the xenolith assemblages such as that in SE Australia (Griffin and O’Reilly 1987). The pyroxene-chrome spinel symplectites in garnet-free lherzolites (Carlswell et al. 1984) provide evidence in support. These symplectites are considered to have formed from reaction between garnet (originally present) and olivine during the course of heating. Such changes cannot be conclusively distinguished from those that are produced during diapiric uprise of garnet lherzolites to shallower levels (Smith 1977). In course of time, as the regions cool and the spinel lherzolite transition moves up, the Moho is seismically better expressed.

4.3.2

Nature of the Crust-Mantle Boundary

The Moho has been traditionally regarded as a sharp discontinuity by most seismologists. Across this discontinuity, the Vp shows an increase from about 6.5–7.0 to >8.0 km/s. However, studies using refraction seismology have shown that reflectivity of the Moho requires a more complex structure. Many workers (Meissner 1973; Hale and Thomson 1982) have opined that the Moho consists of an interstratification of high- and low-velocity rocks which form a 3–6 km zone with a strong positive Vp gradient. The depth of the occurrence of the Moho varies regionally. It may occur at 30–40 km beneath cratons and at more than 50 km beneath young orogenic belts. The Vp in the crust above the Moho varies from about 6.5 to >7.5 km/s. This variation is due to the heat flow of the region in question, but it could also reflect differences in composition of the lower crust. Areas of active volcanism, subduction zones and rift zones often fail to show a clear Moho discontinuity although velocity gradients or reversals are recorded. This usually occurs in areas of high heat flow where Vp may not reach a value of 8.0 km/s until a depth of 50–60 km (Finlayson 1983). This has often been interpreted as existence of thick crust, but petrological and magnetic data are contrary to this view and would suggest that the crust in these areas should be thin (Wasilewski and Mayhew 1982). Recent developments in reflection profiling has shown that the lower crust may be strongly layered in areas where refraction studies show a strong positive Vp gradient. For example, the young crust of Europe, northern America, eastern Australia, and even the northern part of the Indian shield shows a strong increase in the density of reflectors in the lower crust, above seismically homogeneous upper mantle. In areas such as eastern Australia, the layered lower crust is overlain by a middle crust which is seismically homogeneous, whereas in Germany a broad transition from crust to mantle is noticed. The individual reflectors are less pronounced and are discontinuous on a kilometre scale. In rifted areas such as the Rhine graben and North Sea graben, the lower crust displays few discontinuous reflectors and the

148

4 Lithosphere Architecture

Moho is not clearly expressed on reflection profiles (Meissner et al. 1983; Barton et al. 1984). Deep crustal and mantle sections constructed from data on xenoliths in basaltic rocks show that mafic lithologies are interstratified within the lherzolites in the upper mantle. Such mafic xenoliths may include cumulates of mafic magmas such as pyroxenites and websterites, mafic granulites (gabbro protoliths), garnet pyroxenites, and eclogites. The xenoliths provide P-T estimates that place them at depths greater than those typical of crustal rocks (Griffin et al. 1984). The geobarometry on composite xenoliths enables us to place the xenoliths at mantle depths. These depth estimates are substantiated from the tectonically emplaced slivers of continental crust, such as the Ivrea-Verbano zone in north-western Italy (Mehnert 1975). The field association of ultramafic rocks with granulites and pyroxenites supports the compound nature of the xenoliths assemblage (Wilshire and Pike 1975). Thus, it is seen that the crust-mantle boundary in such areas is a gradational transition where a mixture of felsic and mafic granulites towards the top is gradational into ultramafic rocks and pyroxenites below. The maximum concentration of mafic rocks occurs in the lowermost part of the lower crust and the uppermost part of the upper mantle. It is the concentration of the mafic rocks of the lower crust that distinguishes the nature and the depth of the seismic Moho.

4.3.3

Impact of Temperature on Seismic Interpretation

In seismic reflection studies of the deep crust, the layered thickness is interpreted as the ‘crust’ and the bottom of these portions is interpreted as the crust-mantle boundary (Griffin and O’Reilly 1987) which is also regarded as the ‘petrological Moho’. However, wherever refraction data are available, they suggest that Vp increases to >8.0 km/s near about the bottom of the layered portion. This goes to suggest that the ‘seismic Moho’ coincides with the ‘petrological Moho’. However, it needs to be examined whether this argument really always holds good, as discussed below. The ‘Seismic Moho’ by definition is that region below which the mantle displays a Vp of 8 km/s. Both xenoliths and exposed deep crustal sections suggest that lower crust in many regions consists of mafic to intermediate granulites. At temperatures of 500–800
C, normally prevalent at these depths, these rock types will have Vp ranging from about 6.6 to 7.5 km/s (Christensen and Fountain 1975; Jackson and Arculus 1984). If the crust-mantle boundary consists of interstratifications of peridotites with granulites/pyroxenites, then this layered zone will have Vp less than 8 km/s and by definition therefore, it cannot be regarded as the seismic Moho. Thus, the seismic Moho will not correspond to petrological Moho. The former will occur at some level where the proportion of mafic rocks becomes so low and the proportion of ultramafic rocks becomes so high that the Vp becomes 8 km/s. In such cases therefore, the seismic Moho invariably occurs below the petrological Moho.

References

149

An added complication in placement of the seismic Moho occurs due to the high geothermal gradient prevalent in some regions where the crust-mantle boundary is transitional. The high temperature lowers the Vp of all rock types and the granulite/ eclogite transition in mafic rocks, and the spinel lherzolite/garnet lherzolite transition in ultramafic rocks will be pushed to greater depths. As in areas of high heat flow, Vp  8 km/s will not be attained until the depth where the spinel lherzolite/garnet lherzolite transition is crossed, whereas the crust-mantle boundary as defined by dominance of ultramafic rocks is at relatively shallow depths. Conversely therefore, as the area cools towards attaining steady-state temperature conditions, the rocks will display higher seismic velocities and the position of the granulite/eclogite transition or the spinel lherzolite/garnet lherzolite transition will be shifted to shallower levels. Thus, a situation may arise wherein due to the reduction of Vp as a result of cooling, the seismic Moho may move to shallower levels into the crust-mantle transition zone. As regards the magnetisation of deep crustal rocks, the mafic granulites are strongly magnetised due to the presence of magnetite or titanomagnetite, whereas mantle peridotite xenoliths are essentially non-magnetic (Wasilewski and Fountain 1982; Wasilewski and Mayhew 1982). Mafic granulites have Curie temperature of 560–570
C, in some cases the temperature may be as low as 300
C. Thus, the lower crust appears to be the most magnetic layer and the petrological Moho could be regarded as the magnetic boundary. Thus, the long-wavelength magnetic anomalies seen on MAGSAT data could be interpreted as variation in crustal thickness and/or the regional heat flow (Mayhew 1982). The velocity inversion zone encountered in DSS transect across the SGT almost coincides with the depth of Curie point isotherm at 22 km. This has been attributed to high mantle heat flow (Mita 2003). However, the more likely reason should be the relatively high heat flow of the region which varies from 28–58 mWm2. The mean heat flow (38 mWm2) is almost equal to that of the EDC and is more than that of WDC. Presuming that the heat flow values are robust the high heat flow should be due to heat generated by the granulites; however, this is not ably corroborated by the concentrations of Th and U in granulites. The concentrations of HPE in these rocks are low. The other possibility appears to be the contributions to the overall heat flow from mafic boudins within the granulites (Raith et al. 2016); however, more data are required to substantiate this suggestion. The mafic lower crust per se therefore, does not contribute to the magnetic anomaly due to high heat flow and consequent loss of magnetisation. The textural and mineralogical characteristics of the country rocks seem to be the valid reason for the velocity inversion as discussed above.

References Agrawal PK, Pandey OP (2004) Unusual lithospheric structure and evolutionary pattern of the cratonic segment of the south Indian shield. Earth Planets Space 56:139–150

150

4 Lithosphere Architecture

Barton P, Matthews D, Hall J, Warner M (1984) Moho beneath the North Sea compared on normal incidence and wide-angle seismic records. Nature 308:97–105 Biswas SK (1989) Hydrocarbon exploration in western off-shore basins of India. Geol Surv India Pub No 24: 185–194 Carlswell DA, Griffin WL, Kresten P (1984) Peridotite nodules from the Ngopetsoeu and Lipelaneng kimberlite. In: Kornprobst J (ed) Kimberlite II: the mantle and crust-mantle relationships. Elsevier, Amsterdam, pp 229–243 Chekunov AV, Sollugub VB, Starostenko VI, Kharetchiko GE, Rusakov VG, Kosttukovich AS (1984) Structure of the earth’s crust and upper mantle below Hindustan and the northern part of the Indian Ocean from geophysical data. Tectonophysics 101:63–73 Christensen I (1996) Poisson’s ratio and crustal seismology. J Geophys Res 101:3139–3156 Christensen I, Fountain DM (1975) Constitution of the lower continental crust based on experimental studies of seismic velocities in granulite. Geol Soc Am Bull 86:227–236 Closs H, Hinz K (1967) Refraction seismic measurements in the northern Arabian Sea. Paper presented at the symposium on Upper Mantle Project, Hyderabad, India, Jan 4–8, 1967 (quoted by Mahadevan, 1994) Curry JR (1991) Possible greenschist metamorphism at the base of a 22 km sediment section, Bay of Bengal. Geology 19:1097–1100 Dawson JB (1980) Kimberlite and their xenoliths, vol 224. Spinger Velag, Berlin, Heidelberg, New York den Berg R, Stephan Daly J, Salisbury M (2005) Seismic velocities of granulite facies xenoliths from Central Ireland: implications for lower crustal composition and anisotropy. Tectonophysics 407:81–99 Dessai AG, Markwick A, Vaselli O, Downes H (2004) Granulite and Pyroxenite xenoliths from the Deccan Trap: insight into the nature and composition of the lower lithosphere beneath cratonic India. Lithos 78:263–290 Dessai AG, Peinado M, Gokarn SG, Downes H (2009) Structure of the deep crust beneath the central Indian tectonic zone: an integration of geophysical and xenolith data. Gondwana Res 17:162–170 Dietz RS, Holden JC (1970) Reconstruction of Pangea break-up and dispersion of continents, Permian to present. J Geophys Res 75:4939–4956 Durrheim RJ, Mooney WD (1994) The evolution of the Precambrian lithosphere; seismological and geochemical constraints. J Geophys Res 99:15359–15374 Finlayson DM (1983) The midcrustal horizon under the Eromanga Basin, Eastern Australia. Tectonophysics 100:199–214 Foulger GR, Natland JH, Anderson DL (2005) Genesis of the Iceland melt anomaly by plate tectonic processes. In: Foulger GR, Natland JH, Presnall DC, Anderson DL (eds) Plates, plumes and paradigms, vol 388. Geological Society of America, Boulder, Colorado, pp 595–626 Francis TJG, Shor G Jr (1966) Seismic refraction measurements in the north-West Indian Ocean. J Geophys Res 71:427–449 Gokarn SG (2003) Electrical conductivity patters along transects over the Indian lithospheric domains of differing temporal evolution: a review. In: Madevan TM, Arora BR, Gupta KR (eds) Indian continental lithosphere, vol 53. Geological Society of India, Memoirs, Bangalore, pp 129–147 Gopal Rao D, Krishna KS, Sar D (1997) Crustal evolution and sediment history of Bay of Bengal since Cretaceous. J Geophys Res 102:17747–17768 Griffin WL, O’Reilly SY (1987) The composition of lower crust and the nature of the continental Moho-xenoliths evidence. In: Nixon PH (ed) Mantle xenolith. Wiley, New York, pp 413–430 Griffin WL, Wass SY, Hollis JD (1984) Ultramafic xenoliths from Bullenmerri and Gnotuk Maars, Victoria, Australia: petrology of a subcontinental crust–mantle transition. J Petrol 25:53–87 Hale LD, Thomson GA (1982) The seismic reflection character of the continental Mohorovicic discontinuity. J Geophys Res 87:4625–4635

References

151

Harbison RN, Bassinger BC (1973) Marine geophysical study of Western India. J Geophys Res 78:432–440 Harley SL (1984a) The solubility of alumina in orthopyroxene coexisting with garnet in FeO-MgOAl2O3-SiO2 and CaO-FeO-MgO-Al2O3-SiO2. J Petrol 25:53–87 Harley SL (1984b) An experimental study of the partitioning of Fe and Mg between garnet and orthopyroxene. Contrib Mineral Petrol 86:359–373 Ito K, Kennedy GC (1970) Melting and phase relations in a natural peridotite to 40 kilobars. Am J Sci 265:519–538 Jackson I, Arculus RA (1984) Laboratory wave velocity measurements on lower crustal xenoliths from Calcutteroo, South Australia. Tectonophysics 101:185–197 Jagadeesh J, Rai SS (2008) Thickness, composition, and evolution of the Indian Precambrian crust inferred from broadband seismological measurements. Precambrian Res 162:4–15. https://doi. org/10.1016/j.precamres.2007.07.009 Julia J, Jagadeesh S, Rai SS, Owens TJ (2009) Deep crustal structure of the Indian shield from joint inversion of P wave receiver functions and Rayleigh wave group velocities: implications for Precambrian crustal evolution. J Geophys Res 114. https://doi.org/10.1029/2008JB006261 Kaila KL (1988) Mapping the thickness of Deccan Traps flows in India from DSS studies and inferences about a hidden Mesozoic basin in the Narmada–Tapti region. In: Subbarao KV (ed) Deccan flood basalts, vol 10. Geological Society of India, Memoirs, Bangalore, pp 91–116 Kaila KL, Krishna VG (1992) Deep seismic sounding studies in India and major discoveries. Curr Sci 62:117–154 Kaila KL, Roy Choudhury K, Reddy PR, Krishna VG, Narayan H, Subbotin SI, Sollogub VB, Chekunov TV (1979) Crustal structure along Kavali-Udupi profile in the Indian peninsular shield from deep seismic sounding. J Geol Soc India 20:307–333 Kaila KL, Murthy PRK, Rao VK, Kharatchko GE (1981a) Crustal structure from deep seismic sounding along Koyna II (Kelsi-Loni) profile in the Deccan Trap, India. Tectonophysics 73:365–384 Kaila KL, Reddy PR, Dixit MM, Lazarenko MA (1981b) Deep crustal structure at Koyna, Maharashtra, indicated by deep seismic soundings. J Geol Soc India 22:1–16 Kaila KL, Reddy PR, Mall DM, Venkateswarlu N, Krishna VG, Prasad ASSSRS (1992) Crustal structure of the west Bengal basin, India from deep seismic sounding investigations. Geophys J Int 111(1):45–66 Kiselev S, Vinnik L, Oreshin SI, Gupta S, Rai SS, Singh A, Ravi Kumar M, Mohan G (2008) Lithosphere of the Dharwar craton by joint inversion of P and S receiver functions. Geophys J Int 173:1106–1118 Kosarev GL, Oreshin SI, Vinnik LP, Kiselev SG, Dattatrayam RS, Suresh G, Baidya PR (2013) Heterogeneous lithosphere and the underlying mantle of the Indian subcontinent. Tectonophysics 592:175–186 Krishna VG, Kaila KL, Reddy PR (1991) Low velocity layers in the subcrustal lithosphere beneath the Deccan Traps region of western India. Phys Earth Planet Inter 67:288–302 Mahadevan TM (1994) Deep continental structure of India: a review. Geol Soc Ind Mem 28:569 Mahadevan TM (2003) Kuppam-Palani transect programme and new insights into continental evolution. In: Madevan TM, Arora BR, Gupta KR (eds) Indian continental lithosphere, vol 53. Geological Society of India, Memoirs, Bangalore, pp 99–114 Mahadevan TM (2008) Southern high grade domain a differentially transported Archaean lithospheric segment. Geol Surv Ind 74:89–99 Mall DM, Rao VK, Reddy PR (1999) Deep crustal features in the Bay of Bengal: seismic signatures for plume activity. Geophys Res Lett 26:2545–2548 Mandal B, Vijaya Rao V, Sarkar D, Bhaskar Rao YJ, Raju S, Karuppannan P, Sen MK (2018) Deep crustal seismic reflection images from the Dharwar craton, South India-evidence for Neoarchaean subduction. Geophys J Int 212:777–794. https://doi.org/10.1002/2016JB012948 Mayhew MA (1982) Application of satellite magnetic anomaly data to Curie isotherm mapping. J Geophys Res 87:4846–4854

152

4 Lithosphere Architecture

Mehnert KR (1975) The Ivrea Zone: a model of the deep crust. N Jb Mineral (Abh) 125:156–199 Meissner R (1973) The Moho as a transition zone. Geophys Surv 1:195–216 Meissner R, Lüschen E, Flüh ER (1983) Studies of the continental crust by near vertical reflection methods: a review. Phys Earth Planet Inter 31:363–376 Mishra DC (1984) Magnetic anomalies — India and Antarctica. Earth Planet Sci Lett 71(1):173– 180 Mita R (2003) Multi-platform imaging of lithospheric magnetic anomalies. In: Madevan TM, Arora BR, Gupta KR (eds) Indian continental lithosphere, vol 53. Geological Society of India, Memoirs, Bangalore, pp 233–245 Mitra S, Priestley K, Gaur VK, Rai SS (2006) Shear wave structure of the south Indian lithosphere from Rayleigh wave phase velocity measurements. Geophys J Int 164:88–98. https://doi.org/10. 1111/j.1365-246X.2005.02837.x Mooney WD, Laske G, Masters TG (1998) CRUST 5.1: a global crustal model at 5ox5o. J Geophys Res 103:727–747 Murthy ASN, Tewari HC, Reddy PR (2004) 2-D crustal velocity structure along Hirapur-Mandla profile in Central India: an update. Pure Appl Geophys 161:165–184 Nagarjunneyulu K, Santosh M (2012) Nature and thickness of lithosphere beneath the Archaean Dharwar craton, southern India: a magnetotelluric model. J Asian Earth Sci 49:349–361 Naini BR, Leyden R (1973) Ganges Cone, a wide angle seismic reflection and refraction study. J Geophys Res 78:436–439 Naini BR, Talwani M (1982) Structural framework and evolutionary history of the continental margin of western India. Am Assoc Pet Geol Mem 34:167–191 Nickel KG, Green DH (1985) Empirical thermobarometry of garnet peridotites and implications for the nature of lithosphere, kimberlite and diamond. Earth Planet Sci Lett 73:158–170 Nicolas A, Boudier F, Bouchez JL (1980) Interpretation of peridotitic structures from ophiolitic and oceanic environments. Am J Sci 280:192–210 Oreshin SI, Vinnik LP, Kiselev SG, Rai SS, Prakasam KS, Treussov AV (2011) Deep seismic structure of the Indian shield western Himalaya, Ladakh and Tibet. Earth Planet Sci Lett 307:415–419 Pandey DK, Clift PD, Kulhanek DK (2015) The expedition 355 scientists. Expedition 355 preliminary report: Arabian Sea monsoon. Int Ocean Discov Program. https://doi.org/10.2204/iodp.pr. 355.2015 Prasad ASSSRS, Venkateswarlu N, Reddy PR (2005) Crustal density model along Gopali- Port Canning profile, West Bengal basin using seismic and gravity data. J Indian Geophys Union 9:235–239 Radhakrishna M, Subramanyam C, Damodharan T (2010) Thin oceanic crust below Bay of Bengal inferred from 3-D gravity interpretation. Tectonophysics 493:93–105 Rai SS, Vijay Kumar T, Jagadeesh S (2005) Seismic evidence for significant crustal thickening beneath Jabalpur earthquake, 21 May 1997 source region in Narmada-Son lineament, central India. Geophys Res Lett 32:L22306. https://doi.org/10.1029/2005GL023580 Raith MM, Brandt S, Sengupta P, Berndt J, John T, Srikantappa C (2016) Element mobility and behaviour of zircon during HT metasomatism of ferroan basic granulite at Ayyarmalai, South India: evidence for polyphase Neoarchaean crustal growth and multiple metamorphism in the northeastern Madurai province. J Petrol 57:1729–1774 Rao TCS, Rao VB (1986) Some structural features of the Bay of Bengal. Tectonophysics 124:141–153 Reddy PR, Rao IBP (2003) Deep seismic studies in central Indian shield-a review. Geol Soc Ind Mem 53:79–98 Reddy PR, Mall DM, Prassad ASSSRS (1997) Subhorizontal layering in the lower crust and its tectonic significance in the Narmada-Son region, India. Pure Appl Geophys 149:525–540 Reddy PR, Misra DC, Sarma SVS, Harinarayana T, Divakara Rao V, Narayana BL, Singh SB (2001) Modelling the tectonic evolution of the southern granulite belt of the Indian shield using coincident seismic reflection/refraction, geological/geochemical, geochronological, gravity/

References

153

magnetic, magnetotelluric and deep resistivity studies along the southern geotransect. Technical Report NGRI Exp-317 Category B 1–47 Ringwood AE (1975) Composition and petrology of the earth’s mantle. Mcgraw Hill, New York, p 618 Rudnick RL (1992) Xenolith-samples of the lower continental crust. In: Fountain DM, Arculus RJ, Kay RW (eds) The continental crust. Elsevier, Amsterdam, pp 269–316 Rudnick RL, Fountain DM (1995) Nature and composition of continental crust: a lower crustal perspective. Rev Geophys 33:267–309 Rudnick RL, Gao S (2003) Composition of the Continental Crust. In: Holland HD, Turekian KK (eds) Treatise on geochemistry, 3: the crust, 1st edn. Elsevier-Pergamon, Oxford, UK, pp 1–64 Rudnick RL, Gao S (2014) Composition of the Continental Crust. In: Holland HD, Turekian KK (eds) Treatise on geochemistry, 3: the crust, 2nd edn. Elsevier-Pergamon, Oxford, UK, pp 1–51 Shalivahan SS, Bhattacharya BB, Chalapathi Rao NV, Maurya VP (2014) Thin lithosphereasthenosphere boundary beneath eastern Indian craton. Tectonophysics 612-613:128–133 Smith D (1977) The origin and interpretation of spinel pyroxene clusters in peridotite. J Geol 85:476–482 Srinivas Rao G, Radhakrishna M, Sreejit KM, Krishna KS, Bull JM (2016) Lithosphere structure and upper mantle characteristics below Bay of Bengal. Geophys J Int 206:675–695 Sutherland FL, Holis JD (1982) Mantle-lower crust petrology from inclusions in basaltic rocks in eastern Australia-an outline. J Vocanol Geotherm Res 14:1–29 Tatsumi Y, Suzuku T (2009) Tholeiitic vs calc-alkaline differentiation and evolution of the arc crust: constraints from melting experiments on a basalt from the Izu-Bonine-Mariana Arc. J Petrol 50:1575–1603 Tewari HC, Vijaya Rao V (2003) Structure and tectonics of the Proterozoic Aravalli-Delhi geological province, NW Indian peninsular shield. In: Madevan TM, Arora BR, Gupta KR (eds) Indian continental lithosphere, vol 53. Geological Society of India, Memoirs, Bangalore, pp 57–78 Tewari HC, Murthy ASN, Kumar P, Sridhar AR (2001) A tectonic model of the Narmada region. Curr Sci 80:873–878 Upton BJG, Aspen P, Chapman MA (1983) The upper mantle and deep crust beneath the British Isles: evidence from inclusions in volcanic rocks. J Geol Soc 140:105–122 Wasilewski P, Fountain DM (1982) The Ivrea Zone as a model for the distribution of magnetization in the continental crust. Geophys Res Lett 9:333–336 Wasilewski P, Mayhew MA (1982) Crustal xenoliths magnetic properties and long wave length anomaly source requirements. Geophys Res Lett 9:329–332 Wedepohl KH (1995) The composition of the continental crust. Geochim Cosmochim Acta 59:1217–1232 Whitemarsh RB (1974) Summary of general features of the Arabian Sea and Red Sea Cenozoic history. Leg 23 cores. Init Rep DSDP 23:115–1123 Wilshire HG, Pike JEN (1975) Upper mantle diapirism: evidence from analogous features in alpine peridotites and ultramafic inclusions in basalts. Geology 3:467–470 Wilshire HG, Shervais JW (1975) Al-augite and Cr-diopside xenoliths in basaltic rocks from western United States. Phys Chem Earth 9:252–272 Wood BJ (1974) Solubility of alumina in orthopyroxene coexisting with garnet. Contrib Mineral Petrol 46:1–15

Chapter 5

Geothermal Structure

Abstract The shield comprises cratons, both with steady-state and anomalous thermal structures. The WDC and Bundelkhand craton display thermal structure expressed by perturbed geotherms. The difference between the two is stark, whereas the WDC shows intra-craton variation in thermal structure and exhibits a Proterozoic geotherm overprinted by the Cenozoic thermal anomaly, the Bundelkhand craton has a Cenozoic thermal imprint with contributions from crustal accretion. These changes are the result of advective heat from magmas ponded at the crust-mantle boundary. The distribution of granulites and the eclogitic rocks provides an indication of the rate of thermal equilibration and its spatial distribution. The thermal structure of the EDC and that of the adjoining Bastar Craton are different. They are characterised by cratonic geotherms distinct from the other two. The thermal thickness of the EDC lithosphere is similar to Archaean cratons elsewhere, but that of the WDC is analogous to the Mesozoic-Cenozoic terrains worldwide. The heat flow picture is distinct from that of the geothermal structure. The WDC registers the lowest heat flow (29–32 mWm 2) with 35–40 km crustal thickness whereas the EDC has a heat flow of 25–51 mWm 2 (crustal thickness 34–36 km), Bastar 51–64 mWm 2 and Bundelkhand the least, 32–41 mWm 2. In the interior of the shield, the heat flow is 41–55 mWm 2 (31–18 km) whereas at the continental margin in the west, near Cambay basin it varies from 55–96 mWm 2 (crustal thickness 8–15 km). The high concentration of radioactive elements in the felsic granulites from the lower crust could contribute significantly to the surface heat flow. In the SGT, the heat flow is 28–58 mWm 2 (crustal thickness 41–45 km). It is likely that this high heat flow is related to the metasomatic imprint. Both the western and eastern margins of the shield display gravity lows, negative magnetic anomalies, and are characterised by high geothermal gradients. Keywords Thermal structure · Geotherm · Dharwar Craton · Heat flow · Felsic granulite · Advective heat

© Springer Nature Switzerland AG 2021 A. G. Dessai, The Lithosphere Beneath the Indian Shield, Modern Approaches in Solid Earth Sciences 20, https://doi.org/10.1007/978-3-030-52942-0_5

155

156

5.1

5 Geothermal Structure

Introduction

Earth’s processes and properties are both governed by its thermal structure. An understanding of the thermal state of the earth through time is therefore, essential to get a peep into the evolution of the planet in general, and that of its lithosphere in particular. Most processes are controlled by internal heat. Major contribution to it comes from radiogenic heat production in crustal rocks. Concentration of radiogenic elements in crustal rocks influences the heat supply which governs the temperature distribution in the crustal and mantle rocks. Most remotely sensed parameters such as seismic velocity, electromagnetic conductivity, viscosity among others, which are used for the assessment of thermal structure of the deep interior are temperature dependent. Hence, an estimate of thermal state of the crust and mantle rocks attains paramount importance in lithospheric studies. It is a known fact that the transfer of matter involves physical transfer of energy of a hot or cold body from one place to another. This leads to the movement of heat energy from hotter to the colder regions. For instance, magma emplacements into the crustal rocks lead to transfer of thermal energy to the crust. This is known as the advection of heat. Thickening of crust also leads to addition of heat and is known as the crustal inflation.

5.2

Thermal Structure of the Indian Shield

Temperature is one of the key elements required for thermal modelling of the physico-chemical processes operating at depth. Two methods are commonly employed among others to estimate the temperature in the lower part of the crust. One involves the determination of equilibration temperatures and pressures of minerals phases from xenoliths. The other deals with extrapolation of surface heat flow data to depth. Both methods have several limitations which have been discussed earlier. We shall examine first the geothermobarometry of xenolith and their efficacy in determining the thermal state of the Indian shield. Xenolith constraints on temperature distribution are of importance to ascertain the deep thermal regime particularly in tectonically active regions where steady-state thermal models are of limited significance or are invalid.

5.2.1

Bundelkhand and WDC: Thermal Structure

A cursory comparison of the thermal structure of the Bundelkhand and WDC cratons makes it appear that both cratons are quite similar, particularly as far as the mean heat flow (~40 mWm 2) of the craton interior is concerned. However, a closer look reveals that they differ significantly both as regards heat generation and heat flow.

5.2 Thermal Structure of the Indian Shield

157

This is apparent from a comparative assessment of the geotherms of the WDC and that of the CITZ—the junction between the two cratons, namely WDC and Bundelkhand. Thermobarometric estimates on granulite xenoliths from the Deccan Traps indicate equilibration conditions between 650 and 1100  C, 0.6–1.6 GPa suggestive of lower crustal environments. The spinel lherzolites provide equilibration temperature from 1020–1045  C by Ellis and Green (1979) geothermometer while those by Brey and Kohler (1990) computation are lower by 100 C. High T values are also provided by the Sachtleben and Seck (1981) and Witt-Eickschen and Seck (1987) geothermometers. The calibration of Wells (1977) gives T values that are lower by 90–100 oC than those by Wood and Banno (1973) thermometer. These values are treated as significant considering the mineral-micromorphology and the earlier high temperature history (high Al/Ca in core of pyroxenes) of spinel peridotites. Equilibration pressures using the updated calibration of Putirka (2008) provided values of 0.8–1.3 GPa. These T-P estimates suggest that the granulites are derived from near the crust-mantle boundary and that there is no apparent sudden temperature change across the boundary. The empirical palaeogeotherms (Fig. 5.1) constructed from P-T estimates on xenoliths from WDC (western India geotherm) (modified after Dessai et al. 1999, 2004), and from the Bundelkhand Craton, (Central Indian geotherm) (Dessai et al. 2009), enable to evaluate the thermal state of the two regions. Both geotherms are strongly perturbed indicative of temperatures higher than normally expected from conductive cooling conditions of the cratons. The Central Indian geotherm representative of the Bundelkhand Craton is a highly perturbed geotherm characteristic of the Cenozoic thermal imprint. High T-P of spinel lherzolites also shows an overprint on the Proterozoic thermal structure expressed by the Western Indian geotherm. The elevated geotherms of the WDC and Bundelkhand cratons are more likely due to advective heat from the intrusion of ultramafic magmas into the lower crust and upper mantle. Such thermal perturbations are characteristic of regions undergoing thinning of the lithosphere while being subjected to extension. Under the influence of the elevated geotherm, the mafic composition will equilibrate to granulites at relatively greater depths. As the magmatic activity ceases, the geotherm decays to a conductive geotherm. During this cooling, the mafic rocks may get transformed to eclogites subject to the favourable kinetic factors. The presence of granulites and eclogitic rocks at Murud-Janjira indicates a slow cooling after the heating event. Whereas in the north, beneath Kutch the absence of granulites and eclogites may be attributed to relatively rapid cooling that did not allow sufficient time for the rocks to equilibrate to the eclogite assemblage. This also indicates the scale of alkaline magmatism in the two regions. It should be noted that passage of Kutch over the plume was much earlier than that of Murud-Janjira consistent with the northward movement of the Indian plate subsequent of the break-up of Gondwana (Kent et al. 1992). It is quite likely that the extent of alkaline igneous activity in regions such as Murud-Janjira was deeply derived, more varied, widespread, and attendant to more crustal disturbance as compared to that at Kutch.

158

5 Geothermal Structure

Fig. 5.1 Empirical geotherms for the WDC and Bundelkhand cratons constructed from geothermobarometry on mantle and crustal xenoliths. P-T data calculated following the Wood and Banno (1973), Ellis and Green (1979), Brey and Kohler (1990), Battacharya et al. (1992) geothermometers and Newton and Perkins (1982), Brey and Kohler (1990), Battacharya et al. (1992) and Putirka (2008) calibrations [modified after Dessai et al. (1999); Dessai et al. (2004; 2009)]. The Southeast Australian (SEA) (O’Reilly and Griffin 1985), Lesotho (Griffin et al. 1979), and Spitsbergen (Amundsen et al. 1987) geotherms are shown for comparison. The shield geotherm is schematic

5.2.2

Thermal Regime of EDC and Bastar Cratons

On the basis of the studies on eclogite xenoliths in Proterozoic kimberlites (Ganguly and Battacharya 1987) from EDC and from the few average values of heat flow data

5.2 Thermal Structure of the Indian Shield

159

Fig. 5.2 P-T path (using calibration of Ryan et al. 1996) based on the chemical tomography of garnets from lherzolites is presented merely as a representative illustration of the thermal state of EDC (modified after Griffin et al. 2009)

from the Indian shield (e.g. Verma 1970), it is contended that the Indian shield in general, has a steady-state thermal structure (Nehru and Reddy 1989; Gupta 1995). Chemical tomography on garnets from heavy mineral concentrates and garnet lherzolite xenoliths in kimberlites from EDC (Griffin et al. 2009), however, provide a better picture of the thermal state. It is more representative of the thermal state of the craton than that by earlier workers (op. cit.) Thermobarometric estimates on garnets indicate equilibration conditions between ca. 1100 and 1150  C, 2.4–5.2 GPa (Griffin et al. 2009) suggestive of their derivation from subcontinental lithospheric mantle (SCLM). A geotherm constructed from chemical tomography of garnet compositions from lherzolites is presented in Fig. 5.2 (ibd.) as a representative illustration of the thermal structure of EDC. The garnet geotherm does not constrain pressures independently as in classical geothermobarometry, but instead the temperatures from Ni content in garnet (TNi) are projected on to a conductive geotherm to estimate the pressure (Ryan et al. 1996). The xenolith data conform to 35–37 mWm 2 cratonic geotherm expect for the metasomatised part of the mantle. Griffin et al. (2009) present a kinked geotherm for EDC which indicates a transient thermal event that heated the base of the lithosphere. The garnet lherzolite xenoliths in kimberlites from the Bastar Craton show P-T range of 4.3–4.7 GPa and 1140–1270 C (Babu et al. 2009) consistent with a cratonic geotherm of 45 mWm 2.

160

5.2.3

5 Geothermal Structure

Southern Granulite Terrain: Thermal Characteristics

In several terrains, rocks from mid- and lower-crust are tectonically exhumed and an entire crustal section is exposed for observation and sampling (Salisbury and Fountain 1990). The upper 15–25 km of crust is thus available for direct studies such as in the Vredefort structure in Kaapvaal Craton of South Africa, the WawaKapuskasing and the Pikwitonei structure in Superior Province of Canada, the Lewisian terrain from Scotland, and the Sveconorwegian-Svecofennian Province of Norway along with metamorphic core complexes from Arizona (USA). The SGT of south India is also an exhumed section of the deep crust; however, the exposed part largely corresponds to the mid-crustal levels meaning thereby that a considerable portion (~15–20 km) of the lower crust is not available for direct study. As a result, the heat production of the lower crust of this region is poorly constrained. Moreover, how representative these exposed cross-sections are, is also a matter of speculation. Archaean-Proterozoic terrain of SGT which is in juxtaposition with WDC and EDC in the southern part of India is tectonically different and displays a distinct thermal palaeoimprint. The off-craton inclusions (boudins) of mafic and ultramafic dyke rocks within granulites have preserved the relicts of the mineral assemblage of granulite facies stage of the rocks. The P-T on this mineral assemblage is more representative of the palaeo-P-T conditions of their precursors than the P-T determinations on granulites which provides lower P-T values possibly due to re-equilibration attendant to exhumation and emplacement. The mineral assemblage is an anhydrous assemblage consisting of garnet + clinopyroxene + feldspar. The P-T estimates on these provide a more realistic estimate of the equilibration conditions at the time of metamorphism. The cores of minerals from the mafic boudins (emplacement age ~ 2.67 Ga) from granulites yield temperatures of 800  C, 0.8 GPa for the granulite facies metamorphism (M1) (Raith et al. 2016) that affected the terrain ~2.6 Ga ago. This P-T range accounts for a lithospheric palaeothickness of ~30 km for the Madurai granulite terrain prior to metamorphism (M2) of charnockite protoliths ~2.48 Ga and subsequent Pan-African overprint (~610 Ma, 520 Ma), and exhumation of deep-seated granulites to mid-crustal levels. Thus, it is seen that each lithospheric block has a different thermal history both intra-craton and among the cratonic collage.

5.2.4

Xenolith-vis-à-vis Heat Flow-Geotherms

The rate of temperature increase with depth provides an indication of the prevalent geothermal gradient. The latter varies with the tectonic setting of the region and therefore, depends on the lithological variation with depth which governs the concentration of heat-producing elements (radioactive mineral concentration) and on the advective heat contribution. Generally, the shields exhibit lower geothermal

5.2 Thermal Structure of the Indian Shield

161

Fig. 5.3 Comparison of P-T arrays of xenoliths from the Deccan Traps covered region of the WDC with the kimberlite xenoliths from the EDC in reference geotherms of Pollack and Chapman (1977) and typical continental geotherms constrained by heat flow data (modified after Artemieva 2006)

gradient. Higher gradient is exhibited by rifts and areas affected by recent volcanism. However, there are exceptions to these generalities, depending upon the characteristics of the region in question. It is widely accepted practice in petrological studies based on xenoliths to compare the xenoliths geotherms with the conductive continental geotherms commonly referred to as ‘reference geotherms’ (e.g. Pollock and Chapman 1977) constrained by surface heat flow data. Xenolith constraints on deep crustal temperature are of significance in evaluating the thermal regimes of the regions affected by Cenozoic volcanism where steady-state models are invalid. Most of the xenolith data from the Deccan Traps covered region cluster around 90 mWm 2 model geotherm (Fig. 5.3) of Pollock and Chapman (1977). Whereas the kimberlite xenoliths (Ganguly and Battacharya 1987; Nehru and Reddy 1989) from EDC define approximately 50–55 mWm 2 conductive geotherm. A more recent

162

5 Geothermal Structure

study suggests a 35–37 mWm 2 (Griffin et al. 2009) geotherm which indicates steady-state conductive cooling rather than significant thermal anomalies in the upper mantle beneath EDC during Proterozoic for the last 1.0 Ga. So also the xenoliths from Bastar Craton define a 45 mWm 2 geotherm. In a similar manner, xenoliths from Kaapval Craton of South Africa define a 45–50 mWm 2 geotherm, except for the high temperature xenoliths which could perhaps be attributed to transient thermal effect (e.g. Sleep 2003).

5.2.5

Comparison with Cratons Globally

A comparison of Deccan Traps xenoliths from WDC with P-T arrays from different regions of Cenozoic extension, rifting and volcanism worldwide such as Australia, China, Siberia, and Spitsbergen, however, indicates a trend of abnormal heating. Assuming the MORB adiabat to be representative of the convective upper mantle, the xenolith barometry provides a uniform lithosphere thickness of 60–80 km in tectonically active regions. In contrast, cratonic geotherms constrained by xenolith thermobarometry normally fall between 35 and 50 mWm 2 conductive geotherms, with lower estimates of temperatures for the Archaean and Palaeoproterozoic regions. For instance, the geotherms of the Kaapvaal Craton (Rudnick and Nyblade 1999), South Africa; South America and Superior Province broadly display hotter cratonic conductive (40–45 mWm–2) geotherms (except for high temperature region) than those of the Slave, Fennoscandia (Baltic Shield) and Siberian cratons which exhibit relatively colder (35–38 mWm–2) conductive geotherm. Broadly the ages of the continental lithosphere and their thermal thicknesses show a significant correlation irrespective of the variation in lithospheric thicknesses (Artemieva 2006). Generally, Archaean and Proterozoic cratons from Gondwana such as South Africa, Western Australia, South America, and India have a thermal thickness of 200–220 km, (perhaps with the exception of one or two), whereas those from the Northern Hemisphere, e.g. Siberian platform, Baltic Shield, West Africa, and Canadian shield have relatively thicker roots of 300–350 km (Artemieva and Mooney 2001). The variation in thicknesses is of course related both to the tectonic evolution of the terrain in question and the uncertainty associated with the thermal model. By and large, the thickness of continental lithosphere shows an increase with the tectonothermal age. Lithosphere thicknesses of terrains older than Palaeozoic, in general, are more than 100–120 km. The Mesozoic-Cenozoic terrains also show a lithosphere thermal thickness of 60–90 km. The Archaean terrains affected by Phanerozoic tectonomagmatic events, such as the Wyoming and Sino-Korean cratons, also show much lower lithosphere thicknesses of 60–80 km. The Archaean WDC clearly falls in this category as far as the thermal structure and the lithosphere thickness are concerned. It is a case of a reworked Archaean craton wherein the lithosphere thermal thickness during Proterozoic is estimated to be about 125–145 km.

5.2 Thermal Structure of the Indian Shield

163

The lithospheric thicknesses of the Archaean cratons computed from the xenoliths data differ considerably, the hotter geotherm for instance of Bastar Craton yields 200–230 km thickness, whereas the relatively colder one of the EDC craton provides about ~300–320 km thickness following Pollock and Chapman (1977). Heat flowbased geotherm of the SGT provides for a thickness of >250 km presuming average heat flow of 38 mWm 2. These thickness ranges for the depleted Archaean cratons tally by and large with adequate provisos, with the estimates of thermal modelling (Artemieva and Mooney 2001). The calculated temperatures of the WDC and the Bundelkhand xenolith suites are remarkably higher than those normally expected from surface heat flow, thereby invoking the question whether the calculated temperatures represent the ambient conditions, that is at the time of the impact of the presumed mantle plume beneath western India, or are fossil, contemporaneous with high temperature metamorphism and late Proterozoic magmatism. High temperatures at relatively shallow depths imply contributions from advective heat transfer from magmas ponded at the base of the crust. This is supported by the composite xenoliths in which the granulites are interlayered by pyroxenites and websterites (Dessai et al. 1999, 2009) and is also corroborated by the gravity high near Mumbai (Glennie 1951; Takin 1966). Post-Palaeocene, the WDC has been affected by tectonomagmatic activity, the last being due to the impact of the presumed Reunion plume (associated with the Deccan Traps). The xenolith assemblage from the two regions, namely the WDC and Bundelkhand regions (i.e. CITZ) belongs to two different thermal regimes. The xenoliths from the WDC represent Precambrian frozen melts and depict a Proterozoic thermal impress superimposed by a Cenozoic imprint whereas the Bundelkhand xenoliths are cumulates of Deccan Traps magmas and exhibit a Cenozoic thermal imprint (Dessai et al. 2004). This enables reconstruction of the palaeolithospheric temperature variation across the craton and also offers an opportunity to compare the intra-craton secular variation in thermal regimes. A comparison of the two geotherms reveals that a significant change in thermal regime was witnessed by the WDC and Bundelkhand regions during Cenozoic compared to that during Proterozoic, due to two successive impacts, first from Marion plume (~84 Ma, tholeiitic basalts of Madagascar) which led to the breakup of Greater India from Madagascar (Storey and Mahoney 1995), followed by the Reunion plume (65 Ma, Deccan Traps) that resulted in the splitting of Greater India (India-Seychelles split at ~84 Ma) and opening of the Arabian Sea (Norton and Sclater 1979).

5.2.6

Secular Variation in Thermal Structure of Cratons

A comparison of Archaean xenolith geotherms from several cratons of the world reveals crucial information on the thermal state of the Archaean cratons and the extent of lithosphere involvement in their evolutionary process. By and large, cratonic geotherms constrained by xenoliths thermobarometry fall between 35 and

164

5 Geothermal Structure

50 mWm 2 conductive geotherms and show lower estimates for mantle temperatures in Archaean-Palaeoproterozoic regions. Some cratons have been affected by several pulses of tectonomagmatic activity, the contained xenolith assemblages permit to reconstruct the variation in palaeolithospheric temperature regimes. For instance, deep crustal xenoliths from the Kaapvaal Craton, South Africa, have been affected by two thermal events at two different times during the Mesozoic. The xenoliths from the older eruptives follow a 40 mWm 2 conductive geotherm whereas those from the younger eruptives from Lesotho follow a flatter trajectory between 150 and 160 km depth with higher temperatures of 1300–1500 C. Such high temperatures (above mantle solidus) are generally interpreted to represent reduced viscosity and asthenospheric flow which is a pointer to the thickness of the TBL that may not have exceeded about 160–170 km. Likewise, two episodes of kimberlite eruptions in SW São Francisco Craton, Brazil, also show significant change in the thermal regime of the lithospheric mantle. The kimberlites of 85 Ma follow a 40–50 mWm 2 conductive geotherm and temperatures of ~250 C higher than the xenoliths from the older (~89 Ma) kimberlite which follow a 35–40 mWm 2 conductive geotherm. Read et al. (2003) attribute this anomaly to the late Cretaceous opening of the south Atlantic and the break-up of Gondwana.

5.3

Lithosphere Thickness from Heat Flow and Xenolith Data

A comparison of lithosphere thicknesses based on xenoliths P-T and surface heat flow data of Indian cratons with cratons globally indicate that among cratons with adequate data solely EDC shows thickness that matches with both the techniques (Fig. 5.4). The WDC exhibits its unique identity whereas the Bastar Craton has insufficient data. The Kaapvaal Craton of South Africa is possibly the only craton where lithospheric thickness estimates from xenoliths, seismic, electromagnetic, and heat flow data all agree with each other pointing to a thickness of 175–190 km. In most of the Indian cratons except the EDC, there is considerable variation in lithospheric thickness between and among methods and cratons as well. Thermal modelling of cratonic regions also points to large differences in lithosphere thickness which varies from 200–220 km beneath Archaean cratons of South Africa, Western Australia, and south-central part of Canada. The estimated thermal thickness of the south Indian Archaean lithosphere is 200–220 km (Artemieva and Mooney 2001). A thickness of 300–350 km is observed in West Africa, Baltic Shield, Siberia, and northern part of the Superior Province (Artemieva and Mooney 2001) as noted previously. Notwithstanding the overall thermal thickness of 200 km for the south Indian shield (ibid.), considerable variation in lithosphere thermal thickness is exhibited by the Indian cratons.

5.3 Lithosphere Thickness from Heat Flow and Xenolith Data

165

Fig. 5.4 A comparison of lithosphere thicknesses of Indian cratons with those from other parts of the world, based on xenoliths P-T- and surface heat flow data detailed in Table 6.1. The EDC and Bastar (?) cratons show thickness variations in proximity with other well-studied cratons. The WDC clearly plots far removed from most others indicating its unique thermal behaviour (DT: Deccan Traps, WDC: Western Dharwar Craton, EDC: Eastern Dharwar Craton (modified after Artemieva 2006)

The recent estimates by seismic tomography (Maurya et al. 2016) place the LAB of the WDC at ~160 km, of EDC at ~140–200 km, and that of Bastar Craton at ~160–200 km. Singh et al. (2014) and Mitra et al. (2006) from body waves and surface waves, respectively, arrived at much thicker continental roots for the Indian cratons. These estimates are at variance with those from xenoliths, MT studies, and heat flow estimates and will be discussed later. Similarly, MT of the EDC points to a thickness of 175 km (Gokarn et al. 2004) whereas xenoliths and xenocrysts suggest a thickness of 160–190 km (Griffin et al. 2009). The Bastar Craton also appears to be similar, however, caution need be exercised since the xenoliths P-T data are very meagre. Conflicting thickness estimates are also noticed in other parts of the world too such as the Baltic Craton. Here, asthenosphere is not encountered in MT studies even at a depth of 220 km, in high resolution surveys (Jones 1999). Similarly, seismic tomography and thermal models indicate a thickness of ~300 km whereas bodywave tomography suggests 400 km thick lithosphere. Likewise, even the deepest of

166

5 Geothermal Structure

the xenoliths from ~240 km, have depleted composition and granular textures which are uncharacteristic of rocks subjected to flow. These variations are due to the models and assumptions employed in geophysical estimations as well as due to the heterogeneous nature of the base of the lithosphere (Artemieva and Mooney 2002). The techniques too also tend to be depth specific, for instance, seismic tomography tends to depict the lowermost part of the lithosphere base whereas thermal data tends to provide an average of lithosphere thickness midway between xenolith and seismic estimates. Xenolith thickness naturally depends on the depth where from they have been picked up, and therefore, cannot be truly representative of the overall lithospheric thickness, although by no means it can be denied they are the only direct representative of the lithosphere.

5.4

Thermal Evolution of the South Indian Cratons

Although the WDC, EDC, and SGT may have behaved as a consistent unit since about 2.5 Ga, prior to this, individually they display tectonic and thermal evolution which have a distinct identity and which should not be overlooked/by treating them collectively merely for the sake of convenience of fitting them in the regional/global framework. More specifically, of the three crustal blocks, the evolution of the WDC and EDC appears to be broadly similar, like most Archaean cratons the world over, yet when observed carefully it is seen that they differ significantly from each other in several respects and show a distinctly independent identity displayed in tectonic, lithological, and thermal evolution even during the Archaean and particularly so, during the Cenozoic (post-Palaeocene).

5.4.1

Thermal Peculiarities of WDC

The supracrustal rocks of the WDC in particular, exhibit distinct variation from those of the EDC, reflected in magmatism, lithology, grade of metamorphism and thermal characteristics. The differences in characteristics of both have significant tectonic implications which have prompted the division of the craton into two blocks. Tectonically, the WDC consists of 3.3–3.0 Ga TTG and greenstones which evolved initially as subaerial, continental komatiitic ultramafites and later as an intra-cratonic ensialic rift-basin-assemblage and was intruded by the 2.6 Ga Closepet Granite. A three-stage evolutionary model for the WDC is suggested by Chadwick et al. (1989, 1992). The volcanic facies of the Bababudan Group are suggestive of intra-cratonic rifting or passive continental margin settings. An active continental margin or microcontinental arc environments with transpression are suggested for the deposition of clastics of the Chitradurga Group (ibid.). The rocks display intermediate pressurehigh temperature metamorphism. This was followed by the profuse intrusion of the

5.4 Thermal Evolution of the South Indian Cratons

167

newer dolerite dyke swarms during Proterozoic. The WDC is characterised by an elevated geotherm during Proterozoic (Dessai et al. 1999, 2004). It was again overprinted by thermal episode during the Cenozoic (Dessai et al. 2009). Depending on the influence of the impact of the asthenospheric plume on the different parts of the craton, diverse regions display distinct thermal histories unlike regions alongside. For instance, the western region of the craton is hotter than the craton interior whereas the northern region is hotter than the western. Thus, intra-craton thermal structure itself is significantly variable making allowances for the uncertainties of temperature estimates from the xenoliths assemblage. The thermotectonic evolution of both the western and northern regions of the WDC exhibits profuse intrusive activity during the Cenozoic.

5.4.2

Thermal Characteristics of EDC

The EDC, on the contrary, consists of a reworked and remobilised gneissic basement of granodioritic gneisses and granitoids that vary in age between about 2.7–2.5 Ga and exhibits a volcano-sedimentary assemblage that may have evolved in an oceanic intra-arc tectonic setting (Krogstad et al. 1991). The rocks exhibit low pressure, high temperature metamorphism. The Proterozoic magmatism, although restricted, is of the alkaline type represented by kimberlites and other alkaline variants. It does not exhibit any significant intrusive activity post-late Proterozoic and thus has not experienced any major magmatic activity during Cenozoic, save the emplacement of kimberlites (Chalapathi et al. 2016). The impact of the asthenospheric plume on the evolution of EDC does not appear to be of any consequence. Its thermal structure is also uniform over much of the craton except for the sporadic kimberlite magmatism, unlike that of the WDC. Moreover, it has a shield-like thermal structure (e.g. Ganguly and Battacharya 1987; Griffin et al. 2009) similar to the Cretaceous thermal structure of Lesotho. The apparent variation in the thermal gradient 14–16  C/km (Karmalker et al. 2009) is not very significant and does not indicate any abnormal heating of the mantle. Additionally, kimberlite magmatism is a localised phenomenon restricted both in space and time. Therefore, kimberlite emplacement cannot reflect the equilibrium P-T conditions within the Mesoproterozoic mantle lithosphere beneath the EDC, but rather points to some unusual localised thermal event during the tectonic evolution of the region. Furthermore, the present-day thermal structures of the WDC and EDC are also not commensurate with their past thermal states as discussed later. Even if all peculiarities between the two cratons outlined above are ignored, yet two differences still stand out; one, the differences in crustal thickness and two, the differing thermal gradients. Thus, to discuss the evolution of the WDC on the basis of either petrological or thermal characteristics of the xenoliths assemblage in kimberlites from EDC is completely erroneous for understanding the evolution of either of the cratons and misleading when applied to the shield as a whole.

168

5.5

5 Geothermal Structure

Temporal Variation in Thermal Structure

The xenolith-derived palaeogeotherm (~30  5  C/km; Dessai et al. 2004) of western India (Fig. 5.1) gives a rough idea of the Proterozoic thermal state of the lithosphere as the equilibration temperatures of xenolith have been shown to be metamorphic temperatures (ibid.). Low Sm/Nd and Rb/Sr ratios and time-integrated LREE enrichment in xenoliths also support their Proterozoic lineage. This proposition is valid as the craton was not affected by any major tectonomagmatic activity after the intrusion of newer dolerite dyke swarms, i.e. from late Proterozoic to Tertiary. A crustal thickness of about 30 km can be safely assumed as the craton was already stable by 2500 Ma. The temperature at the Moho may have been not less than 750–850  C. This range of temperatures indicates the peak of metamorphism (ibid.) and is common for lower crustal granulites. The pyroxenites and websterites provide temperatures of 890–1207  C whereas the spinel lherzolite xenoliths yield temperatures of 1020–1045  C. The deep crust was then overprinted by a thermal perturbation that is evident from the symplectic outer kelyphitic coronas on garnet, fracturing of plagioclase and garnets, and higher CaO on the rims of orthopyroxenes (Dessai et al. 2004). This event, however, was not penetrative and of a shorter duration as most minerals do not show appreciable compositional zoning that could indicate prolonged reheating. The heating event may have occurred in late Cretaceous to Palaeocene times (~65 Ma), contemporaneous with the Deccan magmatism. This is supported by granulite xenoliths from the Narmada Valley (CITZ) wherein websterite occurs as fracture fillings in garnet granulites. The websterites are cumulates of the Deccan Traps magmas (Dessai et al. 2009) unlike the pyroxenites/websterites from the western region as discussed earlier. The heating event is also corroborated by the lower temperatures yielded by the rims of the minerals than the cores. It can be surmised therefore, that the granulites from western India have undergone non-penetrative thermal shock attendant to the impact of the Marion plume followed by the Reunion.

5.6

Heat Flow Variation among the South Indian Cratons

As mentioned previously extrapolation of surface heat flow data to deep crustal lithologies is yet another method to estimate the temperature of the lower part of the crust. Surface heat flow observations offer additional constraints on bulk heat production in the crust (Artemieva and Mooney 2001). It is, however, presumed that the mantle heat flow is not more than the least measured surface heat flow and that the surface heat flow measurements are unidirectional, mainly vertical, and ignore the horizontal component of heat conduction. The surface heat flow variations are able to offer significant information on the distribution of radiogenic elements. However, contributions from deeper layers to the surface heat production are usually nullified by lateral heat conduction, thus, effectively it is the near-surface heatproducing rocks that have an impact on surface heat flow variations (Jaupart 1983).

5.6 Heat Flow Variation among the South Indian Cratons

169

Fig. 5.5 Heat flow distribution map of India (modified after Roy and Mareschal 2011)

Fairly good amount of surface heat flow data have been generated in the south Indian shield (Roy and Rao 2000; Roy and Mareschal 2011 and references therein). Most of these data are interpreted merely based on surface geology and not much importance is attached to lithologies at depth, partly because of the non-existence of deep information. Similarly, there is an inadvertent tendency to compare cratons both within the subcontinent and globally and draw parallels with them at times ignoring the characteristics and peculiarities of one from the other. Likewise, lateral heterogeneity of the surface heat flow is also at times neglected in characterising heat flow regions for simplistic interpretations. In Precambrian terrains, in general, crustal contribution to surface heat flow varies from 20–44 mWm 2 (ibid.) and contributes more than 50% to the observed heat flow. Present-day thermal structure of western India defined by the heat flow-based geotherm can provide further insight into the thermal state of the craton since the Cretaceous, despite potential limitations of such models. The surface heat flow alone does not indicate the thermal conditions at depth (e.g. Rudnick and Nyblade 1999) yet a comparative approach can be useful in deducing the temporal variation in the thermal state of the deep crustal lithologies. Available surface heat flow data of the south Indian shield broadly show large variations from 27–107 mWm 2 across terrains. The overall average of data is 56 mWm 2 which compares well with the global average. The Archaean cratonic regions show heat flow in the range of 27–46 mWm 2. The Mesoarchaean (3.3–3.0 Ga) WDC registers the lowest heat flow of 29–32 mWm 2 (Mean 31 mWm 2) among the south Indian cratons and has a crustal thickness of 35–40 km (mean thickness 37 km) (Sarkar et al. 2001). In contrast, the Neoarchaen (2.7–2.5 Ga) EDC, the craton to the east of the 350 km (N-S) long Closepet Granite Batholith (2.6 Ga) has lesser crustal thickness (34–36 km) but records higher heat flow varying from 25–51 mWm–2 (Mean 40  3 mWm 2). A heat flow distribution map of India is presented in Fig. 5.5 (after Roy and Mareschal 2011).

170

5 Geothermal Structure

The potential reason for higher heat flow in the EDC despite lesser crustal thickness could be that the EDC crust has a larger component of granodioritic gneisses (relatively acidic) and granitoids that vary in age from ca. 2.7–2.5 Ga in contrast with the 3.3–3.0 Ga TTG gneisses of the WDC which are intermediate in composition. It is well known in south India, as is the case globally, that granitic rocks emplaced around Archaean-Proterozoic boundary are more potassic in comparison to those of the Archaean (from WDC) which are sodic (e.g. Martin 1993). The compositional difference between the two has a bearing on the concentration of radiogenic elements. However, in the pericratonic region to the south, the heat flow picture is different. In the SGT mobile belt, heat flow varies from 28–58 mWm–2 (Mean 38 mWm 2) and crustal thickness 41–45 km (Mahadevan 2003). Considerable variation in heat flow is noticed within the granulite terrain itself. This variation is related to the geochemical differences of the granulites from the different crustal blocks. Moreover, due to strong lateral heterogeneity of surface heat flow, the difference in heat flow values between provinces of different ages can be different. In the Mesoarchaean WDC, mean heat flow is 31 mWm 2, in the Neoarchaen EDC it is 40 mWm 2 whereas in the Proterozoic mobile belt in the south although it decreases to a mean value of 38 mWm 2, it is higher than in the WDC. The granulites also display a shallow Curie point isotherm and a velocity inversion in DSS corroborating high heat generation as discussed earlier. Although the heat flow has been attributed to the mantle in general, there is no evidence of any specific lithology which could host the radioactive elements. Surprisingly, the granulites, which have been suggested to have undergone widespread metasomatism by different workers to account for the high heat flow, contain low concentrations of Th and U. In the absence of any xenolith record from the granulite terrain, it is difficult to pinpoint the exact source of heat-producing elements, provisionally, however, the mafic boudins in granulites (Raith et al. 2016) could be the potential candidates for the generation of higher heat flow. An incisive search for xenoliths from SGT is clearly warranted.

5.7

Global Heat Flow Observations

An opposite situation to the one described above is encountered in some parts of the world. For instance, in the Archaean Kaapvaal Craton of South Africa and the surrounding Neoproterozoic mobile belt (Ballard and Pollack 1987) the heat flow is higher than that suggested by global averages; 25 mWm 2 in cratons versus 8 mWm 2 in the mobile belt. Nyblade and Pollack (1993) have analysed the global heat flow data from Precambrian terrains. They have noticed that although the transition from Archaean to Proterozoic terrains is generally marked by an increase in heat flow, the observed change is not necessarily due to the surface expression of the craton margin, but due to their dissimilar age. Converse of this is also observed, where a decrease in heat flow in Proterozoic terrains within a 200–400 km wide zone along the margin of Archaean cratons has

5.9 Influence of Composition on Heat Generation

171

also been noticed, for instance, the Kalahari Craton of South Africa (Ballard et al. 1987), Superior Province of North America, and Baltic Shield in Europe (op. cit.). A better global heat flow data base is clearly warranted. At present, heat flow measurements on continents are uneven, with some areas densely sampled (e.g. Europe, North America, Asia, South Africa, and Australia) than others, where the measurements are sparse (e.g. Arctic North America and north-central South America) and still other areas (north and central Africa, Antarctica, and Greenland) where measurements are non-existent.

5.8

Heat Flow along Craton Margins

The elevated heat flow patterns observed along the craton margins have been attributed to higher radioactivity of the crustal rocks in Proterozoic terrains (Jaupart and Mareschal 1999). This appears to be true as far as the WDC and the adjoining Proterozoic terrain is concerned as shown below. Nyblade and Pollack (1993), however, ascribe it to overthrusting of the Proterozoic crust on to the Archaean lithosphere. The inclination (dip) of the thrust plane, the age of the crust and that of the underlying mantle can be different within a wide zone along the boundaries. Although this could be a possibility in the Indian case, the more likely reason for which there is strong evidence, is definitely the radioactivity, as discussed below.

5.9

Influence of Composition on Heat Generation

The granodioritic gneisses of the EDC display higher heat production (2.4  2 μWm 3) than the TTG gneisses of the WDC (1.0  0.1 μWm 3) in keeping with the potassic composition of the former (Roy and Mareschal 2011). A broad range of heat flow and heat production of the crustal rocks for the different cratons is presented in Fig. 5.6. It is apparent from the heat flow and heat generation data that there is a wide variability in the estimates of both these parameters (Roy and Rao 2000; Ray et al. 2003; Roy and Mareschal 2011). The granulites of the SGT exhibit heat production in the range of 0.20–2.94 μWm 3 (Mean 0.5  3 μWm 3). The unaltered metabasic boudins within the granulites of the Madurai block of SGT have been investigated by Raith et al. (2016), which show interesting results that have a bearing on heat flow/production and support the proposition above. The mafic boudins consist of a core domain A made up of an anhydrous granulite assemblage of garnet, clinopyroxene, and minor plagioclase. This grades through a narrow mineralogical transition zone (domain B) into a broad amphibole-rich domain C consisting of ferropargasite, K-feldspar  plagioclase, and quartz. The amphibolitisation of the ferroan basic granulite was caused by fluid infiltration during formation of the boudinage of the basic dyke. The boudins have undergone metasomatism under static conditions (ibid.). The domains

172

5 Geothermal Structure

Fig. 5.6 Broad range of heat flow and heat production for various Indian cratons. The mean values for the individual cratons are not depicted due to the wide variability both in heat flow and heat production (modified after Roy and Rao 2000; Ray et al. 2003; Roy and Mareschal 2011)

A to C exhibit a progressive increase in radiogenic element (K, Th, and U) concentration brought about by infiltrating fluids (ibid.). Heat generation computed for domain A to C shows an increase from 8.2 μWm 3 in domain A to 28.24 μWm 3 in domain B to 15.18 μWm 3 in domain C. The occurrence of these mafic boudins is an indication of dyke swarms within the granulites. The high values of heat generation is a pointer to metasomatism perpetrated by the fluids and the enrichment in radiogenic elements brought about both, in the country rock granulites and the associated intrusives. However, such enrichments could be surely overlooked as localised, which they are not, in support of the low heat generation of the granulite terrain. Likewise, the Archaean granulites from the Jequie-Bahia terrain in Brazil show high values of heat generation, up to 10 μWm 3 (av. ~ 2.5 μWm 3). One should be cautious, however, that detailed studies of heat flow in various continental locations, such as Western Australia, Canada, Baltic Shield, and China, have shown significant scatter of heat production. Heat flow values may be variable both laterally and vertically not only within a geological province but even on the scale of an individual pluton. For example, heat production in gneisses of the Yilgarn Craton, Western Australia, varies from almost zero to more than 6 μWm 3 (av. 2.6  1.9 μWm 3). Whereas the gneisses from the Gawler Craton, Central Australia, show average heat generation values of 3.7  3.6 μWm 3. Similar variations are also noticed elsewhere such as in the Baltic Shield where heat generation varies from zero to more than 12 μWm 3. Despite these variations, the basic granulite terrains do provide indication that they can potentially generate more heat due to the introduction of radiogenic elements by infiltrating fluids attendant to metasomatism.

5.11

5.10

Contribution to Heat Flow: Xenolith Estimates

173

Transient Thermal Regime of Deccan Traps

In the Deccan Traps covered areas of the craton to the north, the heat flow is 41–55 mWm 2 (Mean 46 mWm 2) with a crustal thickness 31–18 km. Further north, in Cambay basin including CITZ, in particular, heat flow varies from 55 to 96 mWm 2 (Mean 78 mWm 2) (in Roy and Mareschal 2011); crustal thickness 38–41 km in the eastern parts of the rift zone and varies from 8–15 km in the west. The low heat flow values in the Deccan Traps supposedly confirm that the thermal perturbation was localised and the transients have decayed (ibid.). In younger regions affected by tectonomagmatic activity less than ~250 Ma ago, however, surface heat flow measured in boreholes may reflect the past thermal regime and not the present thermal state of the mantle due to the slow rate of conductive heat transfer. The thermal perturbation in the mantle may not have yet reached the surface fully and is hence not reflected in the surface heat flow. However, considering the thinned crust along the rifted margins of the craton and the high heat flow noticed along the CITZ, it is unreasonable to believe that either the heat flow has not yet reached the surface or that it has decayed. The lithosphere in these tectonically young regions typically exhibits a transient thermal regime (as distinctly seen in xenolith geotherms). The steady-state thermal approximation in such regions is questionable and may even be invalid. If at all, the time delay in thermal front propagation from a mantle thermal anomaly can be explained with reference to regions such as the Grenville Province. It passed over the Great Meteor and Monteregian hotspots at 180 and 100 Ma ago, respectively (Morgan 1983). Yet, the present-day surface heat flow increase is only 5–8 mWm 2 which is near the uncertainty limit associated with heat flow measurements (Nyblade 1999). Therefore, to suggest that the thermal perturbation due to the impact of successive massive plumes such as the Marion, followed by the Reunion, was either localised or the transients have decayed within less 65 Ma is an oversimplification (Cox 1988). For instance, according to Chapman and Furlong (1992), the relaxation time following the cessation of thermotectonic event required to attain a steady state is ~100 Ma or longer.

5.11

Contribution to Heat Flow: Xenolith Estimates

Gupta (1995) suggests that heat flow differences between the EDC and WDC may be due to higher crustal heat generation, and that the heat flux from the lower crust may be more or less equal for these areas. Caution need to be exercised in dealing with terrains which have had a complex tectonomagmatic evolution as will be shown later. The heat flow is somewhat higher in Proterozoic terrains and varies from 44–75 mWm 2. The Deccan Traps, excluding the Cambay basin, show values of 41–47 mWm 2 which are reportedly similar to those encountered in the Precambrian terrain. The Cambay graben registers the highest heat flow of 55–93 mWm 2. On the

174

5 Geothermal Structure

basis of these data, the thermal structure of the shield is considered characteristic of a stable continental area (Gupta 1982, 1995; Gupta et al. 1991). In contrast, it has been shown to be abnormally hot and characterised by a thinner lithosphere by Rao (1976), Singh (1985), and Negi et al. (1986, 1987). Based on the available data Rao (1976), Singh and Negi (1982) and Singh (1984, 1985) concluded that the Moho temperature in southern India is about 550  C whereas to the north it is about 850  C. They further suggest that the Curie point isotherm is around the depth of the Moho in the south and less than the Moho in the north (Mahadevan 1994). Singh and Negi (1982) further conclude that such high thermal regimes may have signatures of Deccan magmatism and the Himalayan orogeny. The present-day surface heat flow of the WDC is 35–50 mWm 2 whereas that along the western continental margin it varies from 63  5 mWm 2 in the south to 83  7 mWm 2 in the north (Gupta 1995). The heat flow along the continental margin (e.g. Morgan 1989) is nearly two to three times higher than that of the craton interior. From DSS studies, it is inferred that the crust of the Dharwar Craton in general, is thicker to the west of Closepet Granite than to the east of it. The Conrad discontinuity is at a depth of about 25 km in the west whereas it is at 20 km in the east. This is expressed by a gravity—low in the west and a high in the east. The heat flow distribution, however, is not commensurate with what should be normally expected. Heat flow is low (av. 31 4 mWm 2) to the west than to the east (av. 40  3 mWm 2) of Closepet Granite despite the greater crustal thickness in the west than to the east. Heat flow is largely related to advective and radioactive heat contributions. Advective heat in the interior of the craton cannot be very significant as has been shown earlier that the under-plating is relatively less here than at the margin of the craton. Therefore, the major contributor to the heat flow is radioactive heat. The latter depends on the concentration of heat-producing elements (HPE) in the crustal rocks. The concentration of these elements in the upper crustal rocks cannot make a significant difference to the heat flow. The other possibility therefore, is the concentration of HPE in deep crustal rocks. The mafic granulite xenoliths contain low concentrations of HPE (K2O: average 0.38 wt.%, Th: av. 2.9 ppm and U: 0.93 ppm) that are similar to the granulite xenoliths worldwide (Rudnick 1992). The heat production thus works out to be 0.54 μWm 3 which may contribute at most ~9 mWm 2 to the total heat flow. This cannot significantly contribute to the radioactive heat. Heat production in mafic granulites globally varies from 0.06 to 0.4 μWm 3, which is the range of possible variations in the lower crust (Rudnick and Fountain 1995). These authors have noted that for a 13–16 km thick continental lower crust, a 0.3 mWm 3 difference in assumed values of lower crustal heat production results in a 4–5 mWm 2 difference in estimates of mantle heat flow, and about 20–70 km difference in estimates of thermal thickness of lithosphere. The felsic granulite xenoliths, on the other hand, have higher concentration of HPE, for example, K2O: av. 1.35, Th: 4.5 ppm, and U: 1.2 ppm (Dessai et al. 2004). This is largely due to the presence of apatite which occurs as inclusions in most minerals and is also found to vein them. It is mostly of metasomatic derivation. It

5.12

Inferences from Other Parts of the World

175

usually has high concentrations of U and Th to the extent of 60 and 200 ppm, respectively (O’Reilly and Griffin 2000). The heat production from felsic granulites therefore works out to be 3.72 μWm 3 which is much higher than the average crustal value of 0.71 μWm 3 (e.g. McLennan and Taylor 1996). One of the prime contributors to the heat flow in the WDC may be the thicker middle crust which primarily consists of felsic granulites as compared to the EDC. Moreover, the felsic granulites contain more than trace amounts of apatite which is known to have high concentration of U (~60 ppm) and Th (~200 ppm). These may have a significant effect on the heat budget of the deep crust and crustal temperatures. Thus, a 12–13 km thick middle crust of felsic granulites could contribute about 44–48 mWm 2 to the heat flow.

5.12

Inferences from Other Parts of the World

Several continental areas from parts of the world provide a unique opportunity to measure directly heat production in the mid- crustal rocks (Fountain and Salisbury 1981; Ashwal et al. 1987; Fountain et al. 1987). The upper 15–30 km of the crust is available for direct studies in the exposed sections of Precambrian terrains such as the Vredefort structure (South Africa), the Wawa-Kapuskesing and Pikwitonei structure (Superior Province, Canada), the Lewisian terrain (Scotland), Sveconorwegian-Svecofennian Province (South Norway), and metamorphic core complexes from Arizona (USA). However, the lower portion (~15–20 km) of the crust in some of these sections such as Vredefort, Wawa-Kapuskesing, and Pikwitonei are not available for direct study and therefore, the heat production of the lower crust is poorly constrained. Laboratory studies on rocks from the exposed crustal cross-sections show that heat production in middle crustal rocks ranges from 0.2 to 0.4 μWm 3. However, higher heat production values of 0.4 to 0.5 μWm 3 have been determined for Precambrian granulite terrain of the Canadian Shield along the Wawa-Kapuskesing transect. Similarly, in the Indian shield high values of heat production (8–25 μWm 3) are obtained from granulites from Madurai, which are interpreted as mid-crustal rocks from the equilibration pressures on minerals from mafic intrusives (Raith et al. 2016). Likewise, amphibolite facies rocks representative of the mid-crustal levels of the Sveconorwegian-Svecofennian Province provide average heat production of 1.6 μWm 3 whereas the associated granulites yield 0.4 μWm 3. Is the present-day heat flow entirely a reflection of the past thermal state? Temperatures higher than 250–450  C have been predicted to prevail in the lower crust (Krishna et al. 1989). Moho temperature beneath the Dharwar Craton has been estimated to be 550  C (Singh 1984). However, if the present heat flow is anything to go by, the Moho temperature along the continental margin should be much higher (~800–900  C). The present-day geotherm is hotter than the one during the Cretaceous. Thus, the thermal evolution of regions within a craton need not always be

176

5 Geothermal Structure

similar, leave alone adjacent cratons in close proximity with each other, this is explicitly seen in the case of south Indian cratons.

5.13

Thermal Perturbation: An Effect of Crustal Thickening

Xenolith P-T data are sometimes employed to constrain the thickness of the lithosphere in a variety of tectonic settings. The lithosphere beneath rifts is generally associated with anomalous structures. These include among others thinning of the lithosphere due to asthenospheric upwelling and regions of anomalously high heat which influences the seismic wave velocity, density, and elastic moduli of the minerals and rocks of the lithosphere. These features are generally related to the transient thermal phenomenon associated with the process of lithospheric extension. The strong and distinct curvature of the Central Indian T-P profile (Fig. 5.1) in the 10–30 km depth range is indicative of the perturbed nature of the geotherm. The elevated geotherm is a reflection of both under- and intra-plating of the lower crust, as is also evident from the seismic reflectors that are observed from a depth of about 8–12 km down to the Moho (e.g. Reddy et al. 1997) as well as mantle upwelling beneath the rift (e.g. Kaila et al. 1985). An indirect assessment of the thermal structure of the region can be made on the basis of the surface heat flow of the region. However, a note of caution is essential as the variation observed in the heat flow need not necessarily be due to the advective heat from the mantle. The surface heat flow along the palaeo-rift zone ranges from >70–120 mWm 2. Assuming that the mean heat flow is ~90 mWm 2 and presuming, as a first approximation, that the entire heat flow is contributed by the mantle, the temperature at the Moho should be at least about 900 C. However, as per the xenolith-derived geotherm the temperature at the Moho could be in the range of 1050–1100 C. This suggests that the thermal structure of the region has not yet completely decayed/equilibrated despite the fact that the region experienced magmatic activity some 65 Ma ago. As mentioned earlier, Chapman and Furlong (1992) suggest a relaxation time of at least 100 Ma to modify the thermal structure since the cessation of tectonomagmatic activity. The low P-wave velocities of deep crustal rocks are also suggestive of high temperatures at lower crustal depth. High geothermal gradient would also imply high advective heat and therefore, high surface heat flow (>150 mWm 2). However, a much lower mean heat flow of about 90 mWm 2 (max. 120 mWm 2) has been recorded in subsurface measurements along the extent of the rift. This suggests that the thermal anomaly has indeed decayed but not to the extent of attaining a quasi-steady state. The xenolith-derived geotherm further implies that it is also capable of generating temperatures sufficient to bring about wet-melting of granitic rocks at a relatively shallow depth of about 20 km. However, if one goes by the present-day geotherm it is unlikely that any melting would occur. Yet the depth region shows low Vps. This apparently implies

5.14

Causes of Excess Heat and Estimates of Crustal Accretion

177

that the minerals are at melting point temperatures though not laced by melt formation. This has considerably affected the Vp at lower crustal depths. A 2 km thick low-velocity layer encountered beneath the Cambay basin (Vp: 5.5 km/s) between depths of 10.5 and 12.5 km, sandwiched between a layer above of Vp 6.2 km/s and the one below with Vp 6.6 km/s (e.g. Kaila et al. 1990) supports this deduction. This low-velocity layer is most likely due to the presence of minerals near melting point temperatures unlike the one in the southern part of the craton (e.g. Kuppam-Palani DSS transect) which is controlled by the fabric of the rocks as discussed earlier. It is well known from laboratory investigations that seismic velocities strongly depend on temperature. Although most measurement were made at low temperatures of about 600–700  C, limited high temperature (above 1000  C) experimental data exists. These studies have shown that at near-solidus temperatures seismic velocities decrease rapidly even prior to actual melting. Both Vp and Vs show a significant drop with increasing percentage of embryonic melt (Sato et al. 1989; Jackson 1993). These studies thus provide important constraints both on the interpretation of seismic data and the thermal state of the lithosphere and consequently influence the inferred composition of the lithologies at depth. Apart from temperature, compositional variations such as Fe content of peridotites in the upper mantle also play a crucial role in generating seismic velocity variation. Other factors that have a strong influence on seismic velocities are water content, other type of fluids including partial melt, and anisotropy.

5.14

Causes of Excess Heat and Estimates of Crustal Accretion

Perturbation to the thermal regime beneath the CITZ has led to heat generation in two ways, one by crustal inflation and, the other by under- and intra-plating. Both processes have operated beneath CITZ. It is suggested that the seismic reflectors especially in the shallower parts of the deep crust (reflectors occur from a depth of 8–12 km and beyond) are the result of crustal inflation which can contribute to the thermal structure that in turn can affect the velocity of seismic waves due to the increase in crustal temperature. However, the inflational heating effect diminishes over time and is negligible after a few million years. Therefore, although the CITZ crust has experienced inflation, the heat loss from it may not have significant effect on the present-day seismic velocities. As against this, the heat contributed by underand intra-plating is significant and can have considerable effect on the seismic velocities over a very long period of time (~100 Ma). This is apparent from a comparison of the heat flow of the craton interior and that along the CITZ. The present-day heat flow of the Bundelkhand region (north of CITZ) is 41 mWm 2, and it may have been the same prior to the commencement of rifting as the craton may have had a quasi-steady-state thermal regime like most other cratons in its vicinity

178

5 Geothermal Structure

prior to fragmentation of Gondwana. This would mean that at a depth of 30 km the temperature would have been 300 C. The present-day mean heat flow along CITZ is ~90 mWm 2 which accounts for a temperature of about 750 C at a depth of 30 km meaning an effective increase of 450 C. The increase in heat has produced a transient thermal perturbation to the steady-state thermal regime that has not yet decayed. The low velocities of P-waves (Vp: 6.7–7.2) at lower crustal depths despite being largely composed of garnet granulites and pyroxenites are consistent with excess heat (Christensen 1979). The inferred temperature increase also provides a measure of the thickening that the crust has undergone. An effective temperature increase of 450  C pre-supposes that the crustal thickness has been almost doubled. The present-day lower crustal thickness at the CITZ rift zone varies from 26–30 km. This implies that the lower crustal thickness at the CITZ prior to rifting and magmatism may have been not more than 10–15 km. There is no reason to believe that the present-day crustal structure beneath CITZ is unlike that of a rift zone.

5.15

Overall Heat Production of the Crust

Continental crust consists of barely 0.5% of the mass of the earth, yet it contains a maximum of 70% of incompatible elements K, Th, and U (Rudnick et al. 1998) which contribute to the bulk of the heat production. Bulk of global chemical budget is based on the concentration of heat-producing elements in the continental crust and the mantle. Estimates of bulk heat production in the continental crust vary from 0.74–0.86 μWm 3 (O’Nions et al. 1979; Allègre et al. 1988). A 40 km thick crust contributes about 30–35 mWm 2 of heat as crustal contribution to the surface heat flow. This obviously means that in Precambrian regions which have an average heat flow of 41–49 mWm 2 (Nyblade and Pollack 1993) should have a mantle heat flow less than ~20 mWm 2. These values when compared with the south Indian cratons, for instance, the WDC which by and large has low heat flow (350 km. The presence of direct samples of the mafic xenoliths over such a large region cannot be overlooked; it shows the dominance of these rock types in the lower crust. Finally, a note of caution in the application of Vp/Vs ratios to estimate the composition of the lithologies at depth would not be out of place. The Vp/Vs ratio depends on many factors such as temperature, composition, and increases with

6.4 Tectonomagmatic Accretions

203

decreasing silica content. It is doubtful how effective the resolution of this technique will be in determining the composition of a heterogeneous mixture of rock types such as felsic and mafic granulites along with spinel lherzolites, on scales that vary from submicroscopic to tens of metres. There is no simple relation between the Vp/Vs ratio and Vp or Vs as shown by studies on igneous and metamorphic rocks (Christensen 1996). Moreover, temperature influences elastic modulii and hence Vp and Vs (ibid.). In a crustal section that ranges in age from Archaean to Proterozoic and where there is a strong lithological heterogeneity and temperature gradient with depth, how far the efficacy of this technique will hold good to determine the composition of a crustal layer, is debatable.

6.4.3

Composition of the Lower Crust of the EDC and Bastar Craton

The EDC crust distinguishes itself from the crust in the WDC in not having experienced significant magmatic accretions during Phanerozoic. Unlike the western craton, it experienced two episodes of alkaline igneous activity, first during the Proterozoic (~1.0 Ga) followed by another during the Tertiary (Chalapathi Rao et al. 2016). However, the latter was a localised activity that did not have a widespread impact either on its physical or thermal structure, as evident from xenolith studies (Ganguly and Battacharya 1987; Nehru and Reddy 1989; Griffin et al. 2009) on a large number of kimberlites both Proterozoic and Phanerozoic. In a primarily gneiss-granulite terrain such as the EDC, the several kimberlite intrusives have not sampled granulite xenoliths. However, there is large proportion of eclogites reported from the xenolith population (Ganguly and Battacharya 1987; Nehru and Reddy 1989; Babu et al. 2009; Griffin et al. 2009). The presence of these xenoliths and the reflective lower crust in the depth range of 34–43 km (Mandal et al. 2018) suggest that the lower crust is layered and consists of mafic lithologies, although granulites sensu stricto as in the WDC, may be lacking. However, the observed density contrast across the crust-mantle boundary and the lower crustal velocities (Vs) of 3.8 km/s and Vp/Vs ratios in the range of 1.70–1.75 are more compatible with a felsic composition (Julia et al. 2009). The S-wave velocity layer of similar velocity range accompanied by elevated Vp/Vs values of 1.8–2.1 (standard ratio of IASP91 is 1.73) are recorded by Oreshin et al. (2008). The depth region from 15 to 35 km, presumably representing the top portion of the lower crust shows S-wave velocity of 3.8–3.9 km/s. The most likely rock type in the lower crust with these attributes is supposedly a dry metastable granulite (Priestley et al. 2008). The laboratory measured Vs and Vp/Vs of mafic granulites are 3.85 km/s and 1.83  0.004, respectively (Christensen 1996). Estimates in this range thus conform to mafic granulites (Oreshin et al. 2008) although they can also be due to gabbro/ norite/troctolite.

204

6 Integration of Insights

Surprisingly, there is no layer with velocity of ~4.0 or 4.1 km/s at lower crustal depths, as observed in some other cratons. Two possibilities exist to explain the absence of this layer. The low Vs could reflect the presence of fluids or melt, but given the low geotherm of the area (Griffin et al. 2009) this is precluded. The other possibility for the low Vs is the effect of high T on the anelastic attenuation of olivine (e.g. Afonso et al. 2008) which in other words means anelastic attenuation of seismic waves in the peridotites below. In comparison with the EDC, the WDC lower crust has a velocity of 3.9–4.0 km/s despite being mafic. The slightly lower velocities observed here are due to high geothermal gradient which affects the receiver function velocities. The Bastar Craton to the northeast also displays a mafic lower crust as is indicated by bulk Vp/Vs of around 1.84 which has been attributed to cumulates under-plating the region (Rai et al. 2005). This is corroborated both by the presence of xenoliths and by receiver function studies (Vs: 3.9–4.1 km/s; Julia et al. 2009).

6.4.4

Lower Crust of the Pericratonic Terrain

The lower crust beneath the SGT is not adequately characterised by petrological investigations although considerable amounts of geophysical data and information on surface geology exist. The terrain itself is an exhumed section of the deep palaeocrust which corresponds to mid-crustal depths according to the P-T estimates of exposed granulites. The exposed geology shows subordinate components of mafic lithologies within the granulites. The majority of the exposed granulites correspond to intermediate type of rocks which have resulted from sialic (metasediments) precursors making allowance for a significant proportion of mafic lithologies. Gupta et al. (2003a, b) from RF studies suggest a felsic to intermediate composition for the lower crust. Ravi Kumar et al. (2001) also observed lower S-wave velocities in the lower crust. These observations are in conflict with the proposition of Julia et al. (2009) who suggest a mafic lower crust from 25 to 45 km depth with Vs ranging from 3.96 to 4.15 km/s. Although there is no direct evidence from xenoliths of lower crustal derivation, the possibility of mafic nature of the lower crust beneath the SGT exists as indicated by the presence of mafic granulites and the occurrence of mafic boudins (the remnants of mafic dykes ca. 2.6 Ga) within the granulite protoliths (Raith et al. 2016 and references therein).

6.4.5

Origin of the Lower Crustal Granulites

In continental shield areas the world over, the lower crust is generally believed to be intermediate (andesitic) in bulk composition as shown by a number of studies on granulite facies terrains (Taylor and McLennan 1985; Wedepohl 1995) and xenoliths from parts of Europe (Kempton et al. 1997). However, most interpretations of

6.4 Tectonomagmatic Accretions

205

seismic data and the petrology of xenoliths from most parts of the world suggest a larger proportion of mafic rocks and a more mafic average composition. Although a lot of information is available on exhumed granulite terrains from India, little is known regarding the lower crustal xenoliths per se. It is therefore, intended here to provide a concise summary of the lower crust through the xenolith-window and draw a comparison with the granulite terrains. The textural characteristics, mineralogy dominated by mafic compositions, and complete absence of aluminous silicates in the xenolith suite indicate that the granulite protoliths from the WDC were igneous in origin. Modal and normative mineralogy permit classification of magmatic protoliths as clinopyroxenites and gabbros. The mafic xenoliths, in general, may represent cumulates/frozen magmas of the Deccan Traps activity (Dessai et al. 2004). However, this does not seem likely except for the xenoliths from CITZ since the Deccan Traps are tholeiitic in composition and contain less Al2O3 than the mafic granulite xenoliths (ibid.) from the western margin. The latter have higher Al2O3 (>17.5 wt.%), lower TiO2 ( 0.2 wt.% and Zr/Nb ratios vary from 8 to 14. Such a comparison, however, need to be attempted with caution since most xenoliths represent cumulates rather than frozen melts. The xenoliths from the regions covered by the Deccan Traps display characteristics similar to Precambrian xenoliths from other parts of the world such as Lesotho (Griffin et al. 1979) and Kola (Kempton et al. 1995). In contrast, the xenoliths from the CITZ have geochemical characteristics that appear to be analogous to cumulates of Deccan Traps magmas. Therefore, the xenoliths from these two localities from the Indian shield indicate that the rocks from these suites are not genetically related to one another as members of a single igneous suite. The trace element data of xenoliths from the western Indian margin suggest that the garnet granulite suite has been depleted in LIL elements, most likely during highgrade metamorphism and anatexis (e.g. Dessai et al. 2004). Low Sm/Nd (0.1–0.2) and Rb/Sr (0.02) and the enrichment of light rare earths in mafic granulites are features characteristic of Precambrian granulites the world over. It seems reasonable to believe that the lower crustal xenolith suite of the shield as a whole is a polygenic complex with a prolonged history of intrusion, metamorphism, and anatexis. These processes have resulted in the formation of a residuum enriched in refractory basic lithologies (Dessai et al. 2004) and depleted in felsic rocks relative to the normal continental lower crust. This generalisation is applicable more for the lower crust beneath the craton than that at the northern margin.

6.4.5.1

Exhumation of SGT

The mobile belt bordering the Dharwar Craton to the south is believed to represent the vestiges/roots of a neighbouring craton (Newton 1990); Kroner (1977) suggests that these granulites formed through multistage reworking of the cratonic rocks and were subsequently exhumed. The granulite facies rocks were reworked and intruded

206

6 Integration of Insights

by basic dykes during extension of the deep crust as is evident from a distinct Conrad discontinuity below 25 km along the Kuppam-Palani (N-S) DSS transect across the SGT. The dykes exhibit syn-boudinage infiltration metasomatism coeval with highgrade metamorphism of host granulites (Raith et al. 2016) at ~2.6 Ga leading to enrichment of high field strength- (HFS) and rare earth-elements (REE) in the palaeo-deep crust. In contrast, at the western and northern margins of the shield, the dykes have entrained granulite xenoliths which are largely dominated by metaigneous rocks. Thus, the distinction between the exhumed granulites and the granulite xenoliths is stark. However, caution must be exercised as xenoliths normally tend to be more mafic than the exhumed granulites (e.g. Rudnick 1992; Rudnick and Gao 2003, 2014). It follows therefore, that very few granulite terrains truly represent lower crustal rocks. Two scenarios are possible: (a) withdrawal/subtraction of felsic rocks as a result of melt generation during decompression consequent on delamination (e.g. Hacker et al. 2015). The partial melts so formed have relaminated portions of the lower crust, so that mafic and felsic lithologies coexist within the deep crust; (b) emplacement of mafic sill complexes into the lower crust during tectonomagmatic episodes. Both mechanisms are consistent with field evidence of mafic intrusives into the granulites and the xenolith petrology.

6.4.6

Evolution of the Lower Crust

Durrheim and Mooney (1991, 1994) suggested that Archaean terrains in general, have a thinner lower crust than their Proterozoic counterparts due to the absence of mafic cumulates in the Archaean terrain. Although this may be valid globally, it does not appear to be true in the Indian context. The Archaean crust of the Indian shield is indeed thin, not due to the absence of mafic cumulates, but because of the destruction of that layer due to repeated/successive impacts of several asthenospheric plumes especially during the Phanerozoic. Broadly, the crustal thickness in the two southern cratons, namely the WDC and the EDC is not much different but a mafic lower crust is better developed beneath the WDC. Kiselev et al. (2008) from joint inversion of P and S receiver functions (PRFs and SRFs resp.) and teleseismic P and S travel time residuals have identified a mafic layer beneath the WDC. The xenoliths from the Deccan Traps confirm this layer to be of mafic granulites (Dessai et al. 1999, 2004). The granulitic layer of the WDC has been largely removed by thermal and magmatic erosion (Dessai et al. 1999) as a consequence of the rise of the Deccan plume and attenuation and foundering of the crust attendant to rifting. Crustal thinning of this nature is common in rifted continental margins and western Indian margin is not an exception in this regard. However, there is agreement with Durrheim and Mooney (1991, 1994) to the extent that beneath the WDC the mafic layer is made of cumulates of Proterozoic age (Dessai et al. 2004). In this context, the observation of Julia et al. (2009) that the

6.4 Tectonomagmatic Accretions

207

mafic layer is absent beneath the Archaean terrains of the WDC and EDC needs to be modified. Absence of evidence need not be taken as evidence for absence. Seismic refraction and wide-angle reflection data can be used to decipher the sub-horizontal layering within mafic granulites as seen in several investigations. Evidence of this nature indicates that the deep crust in many regions is the result of episodic emplacement of basaltic magmas. Intrusive features may vary in type and size, as small sills, lenses, dykelets, and veins that may occur intrusive in the host lower crustal or upper mantle rocks. Attendant to magma emplacement there is contribution of advective heat to the deep crustal rocks. This is capable of bringing about partial melting of the lower crustal rocks with the formation of more felsic types. For instance, in the lower crustal xenoliths beneath the WDC there is evidence that felsic and mafic lithologies coexist and the former are seen to cross-cut the mafic layering. Evidence of such occurrences is also available in upthrust lower crustal sequences (e.g. Shervais 1979; Wilshire 1984). Large-volume melts emplaced into the lower crust may form layered intrusions which may equilibrate at granulite facies conditions. Layered basic intrusions such as the Giles Complex in Australia (Good and Moore 1975), the Sittampundi anorthosite complex from the SGT in south India and similar intrusions in the Ivrea-Verbano zone in Italy (Rivalenti et al. 1981) are examples of the type. It is believed that the high rate of growth of the continents ca. 1.95–1.80 Ga was responsible for the widespread anorthosite emplacement. Such intrusive bodies bring about under-plating of the crust-mantle boundary. The underplating is facilitated due to the density contrast between magmas and wall rocks. Herzberg et al. (1983) opine that the density variation serves as a density-filter which enables ponding of the magmas at the ‘petrological Moho’ resulting in deep crust dominated by mafic rocks. Rheological properties of the crust compel magmas to spread laterally (Meissner 1980) in the lower crust forming sill-like intrusive bodies. As these rocks re-equilibrate to eclogite facies assemblages, they become denser than the mantle rocks (Herzberg et al. 1983). They thus tend to sink into the relatively less dense mantle. Such detached lithospheric blocks sink to the base of the upper mantle and become potential source rocks for hotspot volcanism (e.g. McKenzie and O’Nions 1983). In the case of older cratonic terrains such as those of Archaean age, the existence of a mafic lower crust is less forthcoming as the under-plating may have been less important. The reason for this appears to be the early stabilisation of cratonic portions due to which possibly limited under-plating occurred and which was also not subjected to subsequent partial melting of the lower crustal rocks. However, Griffin et al. (2014) from a worldwide compilation of isotope and trace element data on zircons and Re-Os model ages on sulphides and alloys in mantle-derived rocks suggest that almost until the close of Palaeoarchaean the earth’s crust was essentially stagnant and mafic in composition. On the contrary, in regions that experienced Proterozoic and Phanerozoic tectonic activity, basaltic under-plating at the crustmantle boundary, as well as partial melting of the lower crust, may have occurred episodically, often leading to the formation of more mafic residual lower crust.

208

6 Integration of Insights

The Indian shield in general, and the WDC and EDC in particular are comparable to the Baltic shield in some respects. Both these terrains witnessed Proterozoic tectonomagmatic episodes. The WDC in addition was affected by several tectonomagmatic episodes during Cenozoic. All these episodes have left variable impact by way of under-plating to different extents in different crustal blocks. Distinct under-plating is seen beneath WDC, but evidence of under-plating during the Cenozoic is lacking in the EDC possibly an artifact of natural sampling, although sporadic intrusive activity is indicated by the Proterozoic kimberlite emplacements. The Bastar Craton experienced alkaline magmatism (e.g. Lehmann et al. 2010) almost synchronous with the Deccan magmatism along the western Indian margin. In the SGT, scanty evidence of under-plating during Proterozoic exists, but there is no proven record of magmatism during Cenozoic. Although most xenoliths from CITZ (Bundelkhand Craton) are Cenozoic cumulates, there are indeed rare xenoliths of Proterozoic age, which support under-plating of Proterozoic crust as evident from dyke emplacement.

6.4.7

Formation of the Felsic Crust: Recycling Mechanism

From the aforesaid, it becomes apparent that the xenolith evidence suggests that the lower crust is largely mafic in composition (Rudnick and Gao 2003, 2014). Seismic studies suggest that 10–20% of the lower third of the continental crust is mafic. Heat flow and wave speeds, however, imply that lower crust elsewhere should contain rocks in which SiO2 should vary from 49 to 62 wt.%, meaning thereby that the lower crust should be largely intermediate in composition. Thus, contrary to the widely held belief, the lower crust could be relatively felsic with SiO2 content identical to andesites and dacides (Hacker et al. 2015). The origin of the continental crust is a mystery. The main issue that requires a convincing explanation is the formation of andesitic continental crust when the compositions of the mantle-derived magmas that are responsible for crust formation are basaltic. If the compositions of the bulk continental crust and calc-alkaline arc andesites are compared, there is a similarity between the major- and trace element compositions of both. This has led to the belief that the continental crust has a high proportion of arc andesites and their plutonic equivalents. However, if one analyses the most primitive of the arc lavas, with Fe/Mg close to equilibrium with mantle peridotites, it is clear that they are basalts (Kelemen et al. 2003, 2014). If seismic velocities of arc lava-dominated lower crust and continental crust are compared, the velocities of former are faster than those of the latter (Calvert 2011; Hayes et al. 2013). So is the case with the composition. The arc lava lower crust is depleted in highly incompatible elements compared to the continental lower crust (Hacker et al. 2015). It is fairly well recognised that upward moving diapir-like bodies of anomalously hot mantle when interacted with the lithosphere can lead to thinning of the lithosphere in two different ways: (a) thermal thinning due to increase of mantle heat flow

6.4 Tectonomagmatic Accretions

209

at the base of the lithosphere (Crough and Thomson 1976; Doin et al. 1997) and (b) mechanical thinning of the lithosphere due to density inversion (Bird 1978; Houseman et al. 1981; Drazin 2002) which brings about gravitational instability leading to delamination. Both processes bring about the thinning of the lithosphere, however, the efficacy of one over the other is dependent on the viscosity contrast between the lithosphere and the anomalous mantle (Sleep 1994; Molnar et al. 1998). The solution to the above issues is explained by suggesting a process such as delamination (sensu Hacker et al. 2015) and relamination in the mantle to produce differentiated crust in the island-arcs. Following Hacker et al. (2015), ‘delamination’ is here meant to include a variety of processes that bring about an instability in the dense lower crust, so that it descends into the less dense mantle. Delamination takes place when the lower crust and/or the underlying mantle suffer gravitational instability at high temperatures necessary for viscous flow. It is a mechanism envisaged for refining of the crustal material from an island-arc type lava, since the latter is more mafic than the continental crustal rocks (Kelemen et al. 2003, 2014). Therefore, it is implied that by this refinement the relatively more felsic material accretes to the crust, whereas the more mafic material is returned to the mantle. An igneous differentiation process visualises that delamination could occur if magma emplaced into the crust forms a buoyant differentiate that then forms a part of the crust, whereas the dense ultramafic residue that could sink into the mantle (Arndt and Goldstein 1989; Herzberg et al. 1983). At depths of more than 35 km, an alternative process of metamorphic differentiation could lead to delamination if sufficient garnet is formed in mafic lithologies (Ringwood and Green 1966; Kay and Kay 1991). Such a possibility could be envisaged for the lower crust beneath the EDC. It has also been observed that delaminated crust in several arcs is different from the continental crust and hence the need to invoke alternative mechanism to aid in the refinement process. Relamination is a consequence of delamination. In this process, the more buoyant felsic material is subducted, isolated from the downgoing plate and made to return to the plate above the subduction zone, whereas the dense, ultramafic material is converted to eclogite and subducted into the mantle (Hacker et al. 2011). Several scenarios of relamination have been envisaged to provide alternative mechanisms by which felsic material is brought into the lower crust. The relatively denser ultramafic components can be eliminated from the crust and at the same time, volatile enriched melts can be separated and can penetrate the middle and upper crustal rocks making the latter more felsic whereas the lower crust more mafic. The delamination-relamination process can to some extent explain the formation of felsic rocks from mafic precursors, as well as explaining the coexistence of felsic and mafic rocks in the lower crust as observed in the xenolith assemblage from the WDC.

210

6.5

6 Integration of Insights

Evolution of the Lithosphere

In the previous section, the geophysical and thermal structure of the crust and the crust-mantle transition were analysed based on the seismic, magnetotelluric, thermal, and petrological characteristics of the shallow lithosphere. The following section focuses on the geochemical and geophysical evolution of the deep lithosphere, including the generation of lithospheric mantle and its modification and recycling in later times in response to the tectonic processes. The modification and recycling processes affect among other aspects the thickness of the lithospheric mantle; the determination of which is of considerable interest as a part of the evolutionary process. Estimates of lithosphere thickness being indirect, no single estimate by an individual technique can be held valid and reliable, due to the limitations of the technique, assumptions of crustal structure and density distribution with depth, coupled with heterogeneity and anisotropy of the interior. It is the combined picture drawn from various techniques that would be better suited and arguably nearer to the reality. Often therefore, a multipronged approach is adopted to know the thickness of the lithosphere. It is convenient to estimate the thickness from seismic, electromagnetic, xenolith, and thermal data. All these data sets are available for the Indian shield to enable a critical discussion on the thickness of the lithosphere. Lithosphere characterisation is model dependent and varies according to the geophysical technique employed. Although a variety of proxies of individual geophysical (seismic, magnetotelluric, thermal) methods are employed, most of them are contentious and poorly understood (Eaton et al. 2009). Different techniques provide variable estimates of lithosphere thickness and the depth of the LAB. The results from different techniques are not comparable with one another, since each geophysical technique addresses a different property of the lithosphere. Intertechnique comparisons are difficult and need to be attempted, with utmost caution. However, comparison of estimates between and among regions/cratons by a single technique is possible and more appropriate. For instance, propagation of S-waves is temperature dependent; therefore, variation observed in S-wave velocities beneath regions can be employed to estimate the depth variation of the thermal boundary layer of the regions in question. So is the case with P-waves which can be used to study the density variation of the mantle of several cratons. However, the geophysical estimates of the lithosphere by themselves are unable to provide a realistic petrological picture of the deep interior.

6.5.1

Lithosphere Studies in India

It is fairly well established that continental shields display a tectosphere keel which likely extends to depths of ca. 300 km (Jordan 1975). The cratons that have not been subjected to thermotectonic modifications post-Precambrian are believed to retain a

6.5 Evolution of the Lithosphere

211

thickness (TBL) of about 200–220 km (Rudnick and Nyblade 1999) based on the intersection of the 40–42 mWm–2 xenolith geotherm with the MORB adiabat, as discussed previously. Such thickness estimates to be realistic should be done on the basis of a steady-state geotherm. The estimates become redundant when the xenolith assemblage used for thickness determination has been subjected to advective heat overprint. They also do not tally with the present state of the mantle determined from seismic and electromagnetic studies. A variety of geophysical techniques has been employed to study the lithosphere beneath peninsular India to know its thickness, nature, composition, and the extent of modification, if any, undergone by it during its evolution since the Archaean. These techniques include seismic tomography (Srinaguesh and Rai 1996); receiver function analysis (Kumar et al. 2007; Ramesh et al. 2010), seismology (Mitra et al. 2006); magnetotellurics (Gokarn 2003; Nagarjunneyulu and Santosh 2012); geothermics (Pandey and Agrawal 1999; Agrawal and Pandey 2004), mantle xenoliths (Dessai et al. 1999; 2004; 2009; Griffin et al. 2009); a combination of heat flow, shear wave velocity, and mantle xenoliths (Priestley and McKenzie 2006); topography, gravity, geoid, and heat flow data (Kumar et al. 2013); and heat flow, shear wave velocities, and mantle xenoliths (Roy and Mareschal 2011). Results of these studies are not complementary owing to differences in resolution and assumptions made in employing the preferred model. We shall first evaluate the mineralogical and chemical composition of the SCLM and its evolutionary changes attendant to modification. The thickness estimates of the SCLM are dealt with in the following section under the determination of the LAB of the respective craton.

6.5.2

Composition of SCLM: General Characteristics

The subcontinental lithospheric mantle (SCLM) is now well known to be residual in character after melt extraction. These melt-generation events in continental settings may have occurred either at pressure usually 2–3 GPa which leads to Al2O3 depletion, or at higher pressures > 3 GPa wherein FeO is impoverished in the residue with a concomitant increase in Al2O3 and SiO2. Partial melting of fertile mantle peridotite leaves behind a residue which has high Mg # [Mg/(Mg+Fe)], low SiO2, Mg/Si, Cr/Al, and Ca/Al ratios. Partial melting at pressures of >~3 GPa generates a residue depleted in magmaphile elements and enriched in Mg. The degree of mantle depletion is usually expressed by the Mg # or Fo content of olivine in the residual peridotite. Depleted-mantle peridotites broadly show compositions of olivine in the range of Mg # >0.92, while Mg # ~90 is typical of fertile continental mantle. In terms of mineralogy, the cratonic peridotites are typified by the presence of enstatite and are enriched in olivine with Mg # in the range of 0.915–0.935 in comparison with most others, particularly oceanic peridotites which are rich in olivine and have comparatively lower Mg # of 0.905–0.915 (Artemieva 2011).

212

6 Integration of Insights

Petrochemical studies of mantle peridotites from oceanic settings (mantle slivers from modern ocean basins) and from cratonic terrains indicate significant differences in mantle composition, which is the result of partial melting taking place at differing pressures. The composition of cratonic peridotites is characteristic of deeper melting at pressures of more than 3 GPa, whereas abyssal peridotites can form from low-pressure (1–3 GPa) mantle melting. Melt extraction at high pressures generates residues depleted in basaltic components, such as a decrease in FeO, CaO, Al2O3, TiO2, Na2O, and K2O and an increase of MgO, Cr2O3 (Walter 1998). The degree of melt extraction and depletion of mantle can be quantified by the iron content of olivine (expressed by Mg #). Cratonic mantle which is strongly depleted by melt extraction has Mg # >0.92, (with olivine Mg # ~0.92 and orthopyroxene Mg # ~0.93). The fertile mantle has Mg # ~0.90, whereas convective mantle (asthenospheric mantle) has Mg # varying from ~0.88 to 0.89. The large ion lithophile elements, Ti, Zr, and Y, show lower concentrations in depleted mantle peridotites than in fertile ones.

6.5.2.1

Archaean vs Proterozoic SCLM

Characterisation of the continental lithospheric mantle on the basis of xenolith composition holds the key to identification of the Archaean SCLM. The continental lithospheric mantle has remained distinct from the convecting mantle for billions of years (Menzies and Murthy 1980; Richardson et al. 1984). The Re-Os depletion ages broadly place it between 2.5 and 3.5 Ga; TRD ages peak around 3.0 Ga and no age is older than 3.5 Ga, thereby suggesting that the mantle beneath Archaean cratons is of ‘Archaean’ age (Griffin and O’Reilly 2018). It is primarily on this premise that characterisation of peridotite xenoliths as of Archaean- or Proterozoic-mantle depends. However, there have been instances where peridotites derived from younger lithospheric mantle provided Archaean ages that were much older than the lithosphere in which they have been included. This provided an indication that SCLM in these regions could be relict of the Archaean times (Griffin et al. 1998; Gao et al. 2002). The xenolith localities from India, as elsewhere, can be categorised as cratonic and circum-cratonic. In the first category can be included xenoliths in kimberlites from the EDC, Bastar, and some from Bundelkhand cratons. The paucity of xenoliths and the state of preservation has been a matter of concern. The second category pertains to those from the WDC, (Murud-Janjira, Kutch), and some from Bundelkhand Craton. The cratonic xenoliths are generally characterised by predominance of garnet peridotites along with spinel peridotites (Nixon 1987) whereas the circumcratonic xenoliths invariably comprise spinel peridotites. The peridotites from most cratons are sub-calcic garnet harzburgites (Boyd et al. 1993) which have a paucity of modal clinopyroxene as distinct from the circumcratonic xenoliths which contain larger proportions of this phase. In practice, the former group makes up a small portion of SCLM that is available for sampling. Textural characteristics of xenoliths also aid to some extent in the separation of two

6.5 Evolution of the Lithosphere

213

groups of xenoliths. The coarse textures gradational into porphyroclastic are generally characteristic of cratonic peridotites. They are commonly attributed to the barometric constraints imposed on them. In the light of the features discussed above, it is possible to place SCLM beneath the Indian cratons in a better perspective. Accordingly, the xenoliths from the WDC and Kutch are discussed here separately due to their similarity in tectonic environment, as compared to those from the EDC, Bastar, and some from the Bundelkhand Craton which clearly belong to a different tectonic setting.

6.5.3

Nature of SCLM Beneath Western India

The xenoliths from the WDC and Kutch region exhibit similarity in physicochemical characteristics. The compositions of mantle xenoliths from the WDC show that they vary between harzburgites and spinel lherzolites. They are dominated by olivine which varies from 62 to 85 vol.% with Mg # >0.90–>0.92 and orthopyroxene (Mg #: 0.90–0.94) in most xenoliths varies from 5 to 20 vol.%. Clinopyroxene shows a limited variation from 8 to 15 vol.% (Table 6.1), in this respect they show similarities with Proterozoic xenoliths from North American Craton (Schmidberger and Francis 1999). Based on their orthopyroxene content, the xenoliths show compositional similarity with those from Siberia (Boyd 1989). Most xenoliths contain higher abundance of clinopyroxene indicating that they are relatively less depleted. The lherzolite xenoliths from Kutch contain olivine in the range of Fo89–91 with Mg # varying from >0.89 to >0.92. Orthopyroxene shows a limited variation with Ca0.87–1.0Mg89–91Fe10–8 with low Al2O3 (3–3.5 wt.%) except that it is rarely between 1.5% and 2.0% (Karmalkar and Duraiswami 2010). Clinopyroxene is a chrome diopside with composition Ca47–49Mg48–49Fe5–2. CaO and Al2O3 show negative correlation. Although a clear-cut categorisation either as high-Mg # or low-Mg # type is not possible, a few of the xenoliths have low CaO and Al2O3 than typical primitive mantle compositions (e.g. McDonough and Sun 1995) suggesting fairly high degree of depletion in basaltic components (e.g. Bernstein et al. 1998). The xenoliths from Kutch show similarities to both Archaean and Phanerozoic lherzolites. It is possible to depict the relationship between melt depletion and residual peridotite composition in the Mg # versus modal olivine diagram (e.g. Boyd 1989), in which the lherzolite xenoliths from the WDC define a crude trend whereas those from Kutch display a scatter broadly within the Proterozoic lherzolites (Fig. 6.3). Mineral compositions show their moderately depleted nature compared to primitive mantle, with some having low CaO and Al2O3 and olivine Mg # < 0.92. The xenoliths, however, show some similarities with garnet lherzolites from Kaapvaal Craton and with low-clinopyroxene (70 km) levels. The MT results (Abdul Azeez et al. 2015) and the S-wave velocities complement each other indicating low Vs and low resistivity both occurring broadly at a depth 200 km. Y-depleted garnets (Y  10 ppm, Griffin et al. 1999) as mentioned above are common up to a temperature of 1050  C

222

6 Integration of Insights

Fig. 6.5 Lithospheric section sampled by xenoliths and xenocrysts in kimberlites from the southwestern end of the NE-SW traverse (Fig. 6.4) across EDC is presented merely as an illustration of the depleted and remetasomatised nature of the lithosphere (modified after Griffin et al. 2009)

corresponding to a depth of about 190 km. This depth is taken as the lithosphereasthenosphere boundary, or the base of the depleted lithospheric mantle (ibid.). Eclogite xenoliths are strongly concentrated in the 175–190 km range in the SW of the craton whereas they are scattered in the 95–160 km range in the NE. This is similar to the occurrence of eclogites in many other SCLM sections (Griffin and O’Reilly 2007) worldwide. The concentration of eclogites has been attributed to the ponding of mafic melts from which the eclogites formed as magmatic cumulates (op. cit.). Heat flow-based estimates are a little different from these. Low heat flow values, 35 mWm–2 and attendant low mantle heat flow component (~17 mWm–2) imply that the lithosphere/asthenosphere boundary is at least 175 km deep (Gupta et al. 1993; Roy and Mareschal 2011). The depth estimates of the LAB approximately tally with the lithospheric thickness of Gupta et al. (2003a) and Mitra et al. (2006). These observations are consistent with the models derived from surface wave data which support a mantle lid at least 200 km thick (Oreshin et al. 2011). Likewise, combined results of Ps and Sp wave receiver function data (Sharma and Ramesh 2013) also agree with the above observations. From the joint inversion of receiver function and surface waves, Bodin et al. (2013) place the LAB at ~150–200 km. The above depth estimates are, however, not consistent with MT results of Gokarn et al. (2004) who suggest a conductive asthenosphere at a depth of 80–100 km in agreement with the seismic receiver function data of Kumar et al. (2007). A conductive horizon at a depth of 180 km is suggested by Nagarjunneyulu and Santosh (2012). It is quite likely that the lithospheric keel that existed until about 1.1 Ga (the eruption age of the kimberlites) has been partly eroded and partly subjected to metasomatic modification. For a typical cratonic geotherm (Griffin et al. 2009) as is encountered beneath the EDC, the graphite-diamond transition occurs at a depth of 120–170 km (Kennedy

6.5 Evolution of the Lithosphere

223

and Kennedy 1976). Diamonds are fairly abundant in the cratonic roots of the EDC. It is therefore not unreasonable to consider graphite as a potential candidate to account for the higher conductivity of the upper mantle beneath this region. A shallow conductive layer is also noticed in the Slave Craton which is attributed to low iron content of olivine (Fo92–94 vs the mean value Fo91.5) along with low concentrations of TiO2, Y, and Zr in garnets (Griffin et al. 1999). The depleted nature of mantle peridotite xenoliths and low concentrations of incompatible elements in garnets (subsequently enriched by re-fertilisation) indicate that a similar phenomenon may have occurred beneath the EDC. As mentioned before, the observations of Kiselev et al. (2008) are a little different. They point out that as per one of the upper mantle models employed by them for the arrival of P to S converted phases, the high-velocity keel (Vs: ~4.7 km/s) beneath the craton could be present solely at depths of less than 150 km. At greater depth, it is replaced by a low-velocity layer (Vs: ~4.3 km/s) possibly related to additions (under-plating) from a mantle plume. Although no present-day heat flow anomaly exists (e.g. Roy and Rao 2000) to corroborate the evidence for a plume, it is likely that the bottom part of the high-velocity keel might have been lost (Kumar et al. 2007; Kiselev et al. 2008) due to magmatic and thermal erosion attendant to the impact of the asthenospheric plume, as was suggested for the WDC (Dessai et al. 1999). Alternative explanations for the low-velocity layer have been provided by Kiselev et al. (2008) and Griffin et al. (2009) who attribute it to metasomatismrelated re-fertilisation which led to iron enrichment, addition of clinopyroxene and garnet, and attendant velocity reduction. The garnet- and Fe-contents of peridotites need to be taken note of. Garnet is quite abundant beneath the EDC and hence S-wave velocities should have been higher. However, Fe-enrichment by metasomatic modification may be responsible for the observed lowering of velocities. The other possibility invoked is that the uppermost mantle beneath the Dhawar Craton was never subjected to depletion (Kiselev et al. 2008). However, as noted above mantle peridotite xenoliths from circum-cratonic regions at least show high Mg # (>0.90 to 2.0 wt.% (Babu et al. 2009) are characteristic of the depleted lithospheric mantle. Compositional similarity of the source rocks of these kimberlites to those from the EDC, is indicated by Sr-Nd isotope studies

226

6 Integration of Insights

(Dongre et al. 2020). The mantle may resemble that beneath EDC in certain respects particularly as far as the depleted nature is concerned. As regards the source rocks for the magmas of Tokapal kimberlite, Chalapathi Rao et al. (2014) suggest that as the rocks are high in Ni (2970 ppm), Cr (1720 ppm), with Mg # >0.86, low Al2O3 < 3.9 wt.% imply involvement of garnet-bearing mantle, analogous to depleted harzburgite (e.g. Mitchel and Bergman 1991; Beard et al. 2000). However, going by the Mg # and the Al2O3 content the compositions do not adhere strictly to depleted peridotites discussed above. As proposed by the authors (op. cit.), lower normalised HREE relative to LREE necessitates a source that has experienced a previous melt depletion (e.g. Tainton and McKenzie 1994). This could also imply depletion and subsequent enrichment as is commonly observed in peridotite xenoliths. Broadly, two processes are visualised for crust formation, namely accretion (a) through modern plate tectonic processes and (b) via plume activity (Stein and Hoffman 1994). Even if it is assumed for the sake of simplicity, that subcontinental lithosphere may have formed from cooling of the thermal boundary layer, its thickness should show abnormal variation with depth. This is not observed. Thus, it is obvious that the lithosphere must have been subjected to periodic destruction by convective processes in the mantle, maintaining the lithospheric thickness within observed limits. Several mechanisms are envisaged, such as convective erosion, mechanical delamination, destruction by mantle convective drag, mechanical recycling along subduction zones, among others.

6.5.5.1

LAB Beneath Bastar Craton

The Bastar Craton appears to have retained its lithospheric keel even in postCretaceous times. Garnet lherzolite xenoliths provide equilibration pressures of 4.3–4.7 GPa, with equilibration temperatures in the range of 1140–1270 C (Babu et al. 2009) consistent with a model mantle geotherm 45 mWm–2 (e.g. Pollock and Chapman 1977) which corresponds to a depth of 130–140 km (Lehmann et al. 2010). Estimates of lithospheric thickness from heat flow data provide values of 180–200 km (Artemieva and Mooney 2001; Artemieva 2006) whereas seismic tomography indicates a thickness of 140–200 km (Maurya et al. 2016). The craton displays a prominent gravity high (+40 mGal) over the Chhattisgarh region suggestive of dense lithologies underneath. In the adjoining Singhbhum Craton, S-wave velocity reduction (5–10% relative to IASP91) is noticed, and it has been attributed to high radiogenic heat (Roy and Rao 2000) and anomalously hot upper mantle (Kosarev et al. 2013) as a result of the interaction of the Indian plate with asthenopheric plumes.

6.5 Evolution of the Lithosphere

6.5.6

227

Thickness of CBL and Compositional Variation

An evaluation of the composition of the lithospheric mantle has been attempted through several studies of cratonic peridotites from Siberia, South Africa, Canada, Australia, and EDC from India (Griffin et al. 1996, 1999, 2004, 2009; Gaul et al. 2000). These studies tried to relate the major element compositions of peridotite garnets with temperature estimates by the Ni-in-garnet geothermometer (Ryan et al. 1996). The results of olivine compositions were projected on to the regional geotherm to obtain the depth of origin. This method provides a fairly good estimate of the depth of distribution of Fo content of olivine within the limitations of the employed geothermometer. It should be noted that the depth estimates of CBL by this method indirectly depend on the thickness of the TBL since the P-T estimates are employed to determine the depth of xenolith mineral assemblages. Similar attempts to relate Fo content to the depth of lithosphere have been undertaken by independently determining the equilibration pressures of xenoliths using Al/Cr ratios of spinels especially in shallow regions of spinel peridotite stability fields. Such an exercise for the xenoliths from the WDC indicated that the tentative depth of CBL is less than about 100 km (Fig. 6.6) whereas in other cratons such as EDC and Bastar the thickness of the CBL is much greater, extending to more than 150 km. The depth estimates for the WDC need to be taken with caution since they are not based on garnet facies peridotites as in the neighbouring cratons. It has been observed globally that olivine Mg # composition in general, decrease with depth and also exhibit an almost identical pattern for all Archaean and Proterozoic cratons. Within the constraints of the data, it could be suggested that a similar decrease is also seen in the Indian cratons. Broadly, in the Archaean cratons at depths of ~100 km Fo content of olivine shows maximum values of 93 which decreases progressively to 92 at ~120–170 km and attains a value of Fo 88–90 at ~210–240 km. The depth regions where olivines attain primitive mantle values of

Fig. 6.6 Comparative variation in Fo content of mantle olivine as a function of depth in different Indian, Chinese and Australian cratons (modified after Dessai et al. 1999; 2004; Karmalkar et al. 2005; Griffin et al. 2009; Chinese and Australian data courtesy W. L. Griffin)

228

6 Integration of Insights

Fo: 88–90 could be interpreted as the base of the CBL or the base of the petrologic mantle (Artemieva 2011). This change is noticed at ~210–240 km beneath the Archaean regions and at much shallower depths of ~160 km in Proterozoic terrains. In this respect, the WDC mantle qualifies to be categorised under the latter class. The pattern of decreasing Fo within a narrow depth range is, however, observed only in certain terrains such as the Kaapvaal and Siberian cratons. In the Kaapvaal Craton, for instance, the younger kimberlites (emplaced after 90 Ma) exhibit different Mg # below a depth of ~175–200 km than the older kimberlites (emplaced prior to ~110 Ma). It can be inferred therefore, that subsequent to the intrusion of the older kimberlites the mantle beneath the Kaapvaal Craton was subjected to metasomatic modification. Such a possibility is expressed for the EDC (Griffin et al. 2009) and it also exists in the case of the WDC, as discussed later.

6.5.7

Thickness of TBL

The xenoliths P-T array enables us to make an estimate of the thermal lithosphere at the time of the emplacement of the xenoliths-carrying intrusive rock. The base of the thermal boundary layer could be defined as the transition from a conductive to an adiabatic (isentropic) geotherm. If the two Indian geotherms (Fig. 5.1) are extrapolated to the MORB adiabat, presumed to be representative of the convective mantle, the ‘central Indian geotherm’ yields a lithospheric thickness of about 65–85 km for the central Indian Bundelkhand region. The estimates are broadly in conformity with the estimates arrived at from the reference geotherms of Pollock and Chapman (1977). The ‘western Indian geotherm’, however, suggests a lithospheric thickness of ~80–100 km for western India (Dessai et al. 2004) somewhat similar to the lithosphere thickness estimates (101 km) of Negi et al. (1986, 1987) and Mukherjee and Biswas (1988). However, caution needs to be exercised in interpreting the lithospheric thicknesses as both are perturbed geotherms. The heat flow-based estimates are, however, greater by nearly 100 km (Artemieva 2006) A comparison of the WDC and EDC xenoliths with xenolith-geotherms from the Kaapvaal and Slave Craton allows an evaluation of the thickness of the TBL of these regions vis-a-vis Archaean cratons in South Africa and Canada. The WDC xenoliths occupy a field which is analogous to regions of Cenozoic extension, rifting, and volcanism such as China, Siberia, Spitsbergen, and indicate high mantle temperatures along a 80–90 mWm–2 conductive geotherm indicating transient thermal effects (Fig. 6.7). The geotherm intersects the mantle adibat at about 60–80 km (Fig. 5.1). This depth region signifies a transition, in which a distinct change in the mode of heat transfer occurs, from conductive in shallower regions to convective below ~100 km. This observation also finds support in some geophysical investigations except that the geophysically estimated thermal boundary layer is deeper by about 50 km than the xenolith-defined layer.

6.5 Evolution of the Lithosphere

229

Fig. 6.7 A tentative comparison of xenoliths from WDC and EDC with those from the Kaapvaal Craton-pale grey-shaded field (Rudnick and Nyblade 1999), bold dark grey line shows the best-fit for low-T peridotites from Slave Craton (Kopylova et al. 1999; Kopylova and McCaammon 2003). Thin black lines are conductive geotherms of Pollock and Chapman (1977); the figures (35, 40) indicate surface heat flow values in mWm–2 (modified after Artemieva 2011)

The EDC xenoliths on the contrary indicate a much cooler conductive geotherm of 35–38 mWm–2 indicating intersection with the mantle adiabat at ~250 km depth. Griffin et al. (2009) have, however, estimated the thickness of TBL to be about 175–190 km. This indicates a large difference in the TBL thicknesses of the two adjacent cratons. It is also worth noting that various techniques provide for different thicknesses for the different conceptual layers of the Indian lithosphere (Table 6.2). Such a comparison of lithosphere thickness estimates from xenolith, seismic, thermal, and electromagnetic data should be made with caution. As noted earlier like in the case of the Kaapvaal Craton, the EDC is the only Indian craton whose lithosphere thickness estimates (175–190 km) by all four data sets almost tally with one another. A careful examination of the WDC data suggests that the chemical, thermal, MT, and some of the seismic estimates of lithospheric thickness show some proximity to one another although most point out to a shallow LAB. These observations need to be further investigated and confirmed with additional data. For the rest of the cratons either all the data sets are not available or wherever they do exist, estimates by one technique do not agree with those of the other.

6.5.7.1

Contrasting Thermal Structures

It is tempting to suggest here that it is unreasonable to draw parallels as regards the thermal structure or lithosphere thicknesses of cratons in juxtaposition with each other on the basis of either xenoliths solely from one of the cratons or by other techniques. Such analogies should be treated with utmost caution, both for the

230

6 Integration of Insights

Table 6.2 Depth estimates of LAB for the chemical-, thermal-, seismic-, and electrical-lithosphere beneath the Indian shield based on thermal and non-thermal data

Precambrian craton WDC EDC Bastar Bundelkhand Aravalli Singhbhum SGT Western Margin Eastern Margin

Cratonisation age (Ga)15,16,17 3.3–3.0 2.7–2.5 ~1.8 2.6–2.5 2.6–2.5 3.0–2.5 2.6–2.5, 2.3 –

Chemical lithosphere (Xenoliths)1,2,7 60–801a 175–1902 130–1407 65–851 – – – –

Thermal lithosphere (Xenolith P-T) ~80–100a ~1753,10 ~1753,10 – 160–200 160–200 180–230 –

Seismic lithosphere (RF function)4 160–2008,18 175–2004,9 140–2008 200–2508 – 70–1006 165–18012 80–10013

Electrical lithosphere (MT)5 70–1605,19 ~1805 – – – 58–956,11 – –

–

–

–

77–12714

–

1

Dessai et al. 2004, 2Griffin et al. 2009, 3Artemieva 2006, 4Gupta et al. 2003a, 5Gokarn 2003, Gokarn et al. 2004, 6Shalivahan et al. 2014, 7Babu et al. 2009, 8Maurya et al. 2016, 9Oreshin et al. 2011, 10Gupta et al. 1993, 11Manglik and Mandal 2016, 12Singh et al. 2014, 13Kumar et al. 2007, 14 Srinivas Rao et al. 2014, 15Meert et al. 2010, 16Jayananda and Peucat 1996, 17Krogstad et al. 1991, 18Borah et al. 2014b, 19Abdul Azeez et al. 2015, aDessai (unpublished data)

spatially related erstwhile Gondwana cratons but even for others presently juxtaposed cratons such as those from the Indian shield. It is observed that each craton has a peculiar evolutionary history, thermal structure, and therefore its own thermal identity. This implies that the post-Archaean tectonomagmatic events in the Indian shield have been quite selective and varied both in time and space to considerably modify/alter the thermal structure, obviously the impact on cratons is also variable. This is explicitly seen not only in the case of the Bundelkhand and Dharwar cratons, as discussed previously, but also in the case of the WDC and EDC which apparently have contrastingly different thermal structures and hence evolutionary patterns.

6.5.8

Lithospheric Mantle: Lateral Variation

Several studies have been carried out globally to ascertain the lateral compositional variation of the lithospheric mantle. Among the better studied cratons, the Slave, Kaapvaal, Siberian, and North China figure fairly prominently. Systematic studies from a large xenoliths population have shown that the individual cratons display a distinct compositional variation of the lithospheric mantle which could be related to their geodynamic evolution. This is also true of some of the Indian cratons as seen from their thermal and geochemical characteristic discussed above. Estimation of the compositional variation of the mantle beneath the WDC needs to be done with caution due to a smaller data set (of mantle xenoliths). Within the

6.5 Evolution of the Lithosphere

231

data constraints, it could be suggested that a distinct difference in composition exists in the lithospheric mantle. The Archaean mantle beneath the southcentral and southeastern part of WDC is more depleted than the mantle towards the northwest beneath Mumbai-Kutch as discussed above. In the vertical sense, the mantle is stratified with Archaean lithosphere in shallower regions which is overlying and is partly replaced by the Proterozoic mantle. The moderately depleted sub-calcic harzburgites and lherzolites predominate in the southeast whereas they coexist with primitive mantle compositions in the northwest. The mantle beneath the northwestern region was pervasively metasomatised at least in the depth range of 40–70 km possibly twice during its long history, first during the Proterozoic followed by the Cenozoic thermal imprint (Fig. 6.8). The variation in mantle composition beneath the EDC has been discussed with the help of composition of garnet concentrates from kimberlites (Griffin et al. 2009). They have observed variation of lithospheric mantle along a NE-SW direction with highly depleted mantle in the shallower parts in the SW of the EDC. The mantle here is dense with abundant eclogites and sub-calcic harzburgites; considered to be typical Archaean lithosphere. These lithologies form a subordinate component of the mantle in the NE (ibid.). The lithospheric mantle beneath the Bastar Craton consists of depleted sub-calcic harzburgites. They are underlain by depleted garnet-bearing harzburgites/lherzolites (Chalapathi Rao et al. 2014).

Fig. 6.8 Compositional variation in the lithospheric mantle beneath the Indian cratons is depicted by means of cartoons which are based on information from the xenolith. The WDC mantle shows temporal evolution from the Archaean to the Cenozoic. The EDC has the thickest present-day lithosphere, whereas Bastar has the thickest crust, in comparison to all others (Ar: Archaean, Pr: Proterozoic, Lherz: Lherzolite, Sp/Gnt: Spinel/Garnet, G/D: Graphite/Diamond)

232

6.5.9

6 Integration of Insights

Metasomatic Modification of the Lithospheric Mantle

The majority of clinopyroxene xenoliths from the western Deccan Traps (WDC) have resorbed rims indicating reaction with infiltrated melt. The melt could be (a) trapped host liquid, (b) partial melt formed due to decompression during ascent or (c) a trapped melt residual after crystallisation of pyroxenes. The xenoliths contain hydrous phases such as phlogopite, kaersutite, and apatite. The phlopopite in xenoliths has higher Mg # than that in the lamprophyre host. Therefore, it could be surmised that xenolith-phlogopite crystallised from an infiltrated melt that brought about compositional variation in clinopyroxenes (Dessai and Vaselli 1999) which is a case of patent (modal) metasomatism (e.g. Harte 1983). The granulite xenoliths associated with spinel lherzolites contain scapolite again indicating metasomatic modification. The xenoliths also exhibit LREE (light rare earth elements) enrichment by cryptic metasomatism (Dessai et al. 1990). It has also been shown (op. cit.) that the clinopyroxenite xenoliths host high concentrations of compatible elements, several times more than the host lamprophyre and they are positively correlated with Mg #, consistent with a cumulative origin. Dessai and Vaselli (1999) have shown that the xenoliths belong to an older cycle which predates the Deccan magmatism. Peninsular India has witnessed several magmatic episodes of particular relevance are the ones ca. late Proterozoic, and the one ~84 Ma which resulted in the Indo-Madagascar split (Storey and Mahoney 1995), and continued into the 65 Ma episode which resulted in the Deccan Traps activity. The older event coincides broadly with the Pan-African event in other parts of Gondwana (Kroner 1981). This indicates that the mantle beneath the Dharwar Craton underwent pervasive metasomatism much prior to the Deccan magmatic episode and this metasomatism served as a precursor to the generation of alkaline melts (e.g. Menzies and Murthy 1980).

6.5.9.1

Source Characteristics

Several studies have shown that metasomatised mantle is more favourable for the generation of alkaline magmas (Edgar 1987; Zhang et al. 2001). Nb peaks in LIL (large ion lithophile) spider diagrams and Nb/La ratios of xenoliths are characteristic of incompatible-element-enriched mantle (e.g. Sun 1980). The clinopyroxenes in xenoliths from Kutch, for instance, have heavy rare earth element (HREE) concentrations that can be attributed to low degrees of partial melting, but their higher Middle REE and Light REE may indicate introduction of incompatible elements into clinopyroxenes which had a prior depletion history. The chemical features of Kutch clinopyroxenes, such as enrichment in LREE, Nb enrichment, low Ti/Eu, high La/Yb, Nb/La, can be explained by the interaction of residual mantle peridotites with carbonatitic melts (Karmalkar et al. 2005). Ti, Zr, Y, and Nb-negative anomalies observed in Kutch xenoliths suggest their depleted nature.

6.5 Evolution of the Lithosphere

233

Petrological and experimental studies have shown that metasomatic fluids, either silicates or carbonatitic, could carry a variety of incompatible elements along with them. Such low-viscosity fluids bring about modal or cryptic metasomatism in the invaded peridotites leading to the formation of, e.g. amphibole, rutile, phlogopite, sphene, ilmenite apatite, zircon, and titanite, which may host otherwise incompatible elements (Meen et al. 1989; Rudnick et al. 1993). It has been shown that most Kutch xenoliths could be produced by 2–5% melting of the primitive mantle source; however, more depleted samples would require 8–20% melting. Modelling for LREE shows that nearly half of the xenoliths plot near the modelled melt-extraction curve with the contents of LREE and MREE significantly higher than predicted (Karmalkar et al. 2000).

6.5.9.2

Depth of Partial Melting

Kutch spinel peridotite data lie close to the trend of garnet facies implying small degree of partial melting in the garnet peridotite facies probably between the depths of 80–100 km (ibid.). It has also been shown that the most likely phase to host P is apatite rather than garnet as seen from the negative correlation between P and Al2O3/ TiO2. The latter mineral is stable up to 2.5 GPa at 1350 C (Watson 1980; Dupuy et al. 1992; O’Reilly and Griffin 2000). The estimated thickness of the mantle (~100 km) implies the source region to be at the base of the SCLM rather than in the asthenopheric mantle. Similar spinel peridotite xenoliths from the Ethiopian rift have been suggested to be from plume-derived magmas (Bedini et al. 1997) which possibly brought about thermo-mechanical erosion of the lithospheric mantle. Calculated P-T of the xenoliths 950  C, 0.9–1.2 GPa indicate that the metasomatism has possibly occurred within the subcontinental lithospheric mantle above the carbonatite field. The host magmas were generated at deeper levels perhaps at pressures of 2.2 GPa which is within the depth required for the generation of carbonatite magmas (2.1–3.0 GPa). Such a mechanism of simultaneous generation of nephelinitic and carbonatitic magmas has been proposed for the xenoliths from Canary Islands (Coltorti et al. 1999). As far as the Dharwar Craton as a whole is concerned, a zone of metasomatised (fertile) SCLM between the WDC and EDC is suggested by Griffin et al. (2009) on the basis of their work on garnets from Anumapalle and Wajrakarur clusters of kimberlite intrusives. They propose metasomatism of the mantle at varying depths, almost across the whole EDC (ibid.) along a NE-SW transect. Most of the high-T garnets have high contents of Ti and Zr consistent with metasomatism. Melt-related metasomatism is widely distributed at different depths at diverse places throughout the section, but is primarily concentrated towards the basal portion. Thus, metasomatic modification of the mantle is noticed in most of the lithospheric sections beneath many cratons of the Indian shield, particularly from the peninsular block. It is but expected that a plate subjected to repeated asthenopheric impacts would not be free from metasomatic modification though the extent and the

234

6 Integration of Insights

degree of metasomatism may vary depending on the tectomagmatic severity experienced by the region in question. Griffin et al. (2005) have opined that magmatic events are capable of modifying the composition of the lithospheric mantle more than several hundred kilometres from the magmatic centre.

6.5.10 Continental Margins: Lithosphere Architecture The western and northern margin of the WDC has played a significant role in modifying the architecture of the mosaic of Indian cratons. The multiple types of magmatism documented along both the margins and the accretions that have taken place illustrate the role of the craton margins in guiding and channelising mantle fluids and low-degree partial melts, e.g. lamprophyres, nephelinites, carbonatites, and tephriphonolites (Dessai 1987; Dessai et al. 1990; Karmalkar et al. 2000). The disposition of these rocks is controlled by linear belts coinciding broadly with craton margins. Burke et al. (2003) suggest that the craton boundary zones represent erstwhile tectonic sutures which could become reactivated and potentially participate in lithosphere accretion. These rift-related magmatic additions have been responsible to bring about modification of SCLM. Just as in the case of the Kaapvaal Craton (Griffin et al. 2003), the WDC was subjected to accretionary modifications during late Proterozoic and the ~140–65 Ma period during which two episodes of kimberlite emplacement occurred in adjacent cratons in addition to the ‘newer dolerites’. The second episode of kimberlite emplacement is evident in the EDC and Bastar Craton, which carries some of the depleted xenoliths. The older Proterozoic kimberlites from the Bastar Craton carry xenoliths which are not characteristic of Archaean SCLM. However, the high Mg # olivine in lherzolite xenoliths from the EDC and Bastar represent geochemically depleted, rigid, and tough lithosphere which extends to the depth of about 200 km. Elsewhere as in South Africa the cratonic roots extend to more than 300 km (Begg et al. 2009). The remnants of cratonic roots are also available as detachments and relicts of craton subsequent to break-up of Gondwana and separation of first Madagascar from Greater India (Storey and Mahoney 1995) and later of Seychelles from Indian platform (Norton and Sclater 1979). These repeated break-up events, especially along the western margin, have been responsible for metasomatic modification of SCLM rendering it geochemically less depleted and rheologically weak. This finds parallels with the areas around the Afar Triangle (Dessai and Bertrand 1995) and those along the East African Rift System (Begg et al. 2009). Proterozoic tectonism and magmatism in Africa is largely restricted to craton margins which are presumably confined to modified Archaean crust (ibid.). Such an observation is not definitive in the Indian context although quite a few kimberlite occurrences from Bundelkhand and Bastar cratons are localised to marginal zones of cratons. Griffin et al. (2009) also opine that kimberlite occurrences from EDC exhibit

6.5 Evolution of the Lithosphere

235

a proximity to the craton margin. In their opinion, the mantle beneath these areas is a modified Archaean SCLM as distinct from a juvenile Proterozoic SCLM.

6.5.10.1

Unroofing of the Continental Lithosphere: Offshore Evidences

Along the western Indian continental margin, the shelf is wide, about 350 km in the north around Mumbai and narrowing to 150 km around Cochin in the south. The region extending approximately from ~25o N to ~10o S, in general, from Laxmi Ridge in the north to Laccadive-Chagos Ridge in the south, does not exhibit marine magnetic anomalies but these are well displayed in the basin to the west of the ridge. The oceanic crust of the western basin was generated in the southern hemisphere ~45–65 Ma ago (Naini and Talwani 1982). The Laxmi and the Laccadive-Chagos ridges on the contrary are relict fragments of continental lithosphere with no oceanic crust over the topographic features, consistent with the absence of magnetic stripes. Furthermore, the analogue of the west coast dyke swarm (Dessai and Viegas 1995) has been identified on the Mahe and North Islands of Seychelles (Devey and Stephens 1991, 1992). This is a pointer to the opening of the Arabian Sea which involved attenuation, extension, thinning, and foundering of continental crust from the subjacent mantle along a system of extensional faults (Dessai and Bertrand 1995) which locally may have a listric configuration as has been observed in several parts of the world. In seismic reflection profiles, most of the shallow faults in the Tertiary sedimentary apron over the basement rocks in the Arabian Sea are interpreted as listric faults. However, the listric configuration of faults cannot be ascertained from field data, since no master fault along the western margin is either identifiable or reported, although there is strong evidence of faulting all along the west coast of India. Most faults that are available for observation are, however, extensional faults (Dessai and Bertrand 1995; Dessai and Viegas 1995). The sub-Moho region beneath the Laxmi-Laccadive and Chagos-Maldive ridges exhibits low seismic velocities suggesting the presence of hot mantle (e.g. Braile et al. 1995; Baldridge et al. 1995). The situation is analogous to that observed in some of the active Cenozoic rifts in other parts of the world such as the Kenyan rift and the Rio Grande rift and differs significantly from that prevalent beneath passive rifts such as the Rhine Graben and the Baikal rift where faster mantle is encountered beneath the rift zones. Palaeo-rifts in stable continents such as the Proterozoic mid-continental rift in North America, the Dnieper-Donets Palaeozoic rift in Russia, the Oslo Graben (Palaeozoic) in the Baltic shield also do not exhibit low-velocity anomalies in the upper mantle. The sub-Moho mantle beneath the Arabian Sea ridges is quite peculiar. It is more akin to that beneath active rifts rather than the passive ones or the palaeo-rifts on the continents. The Arabian Sea, from the east coast of Africa to the west coast of India, is a region characterised by low shear wave velocities (Nataf et al. 1984). The region around Mumbai is conceived as a depleted tectosphere composed of olivine and pyroxene, resulting in 1% density reduction and an increase in S-wave velocity due to lower iron content. The region characterised by a 35 mGal gravity low is a

236

6 Integration of Insights

relatively high-velocity region (1–4% faster) in the depth range of 100–400 km and is underlain by a very diffuse low velocity in the depth range 400–600 km. The highvelocity region is interpreted to form the root of the lithosphere as the continental tectosphere (Iyer et al. 1989). This is countered by recent studies (Kiselev et al. 2008; Oreshin et al. 2011; Kosarev et al. 2013) which suggest that high-velocity cratonic lithosphere is practically absent beneath the Indian shield in general. The upper mantle S-wave velocities at depths of 70 mGal) and negative residual topography (