Embed Size (px)
Citation preview
Quaternary Environmental Change in the Tropics
Quaternary Environmental Change in the TropicsEDITED BY
Sarah E. MetcalfeUniversity of Nottingham, UK
and
David J. NashUniversity of Brighton, UK
Series Editor: Ray Bradley, University Distinguished Professor in the Department of Geosciences and Director of the Climate System Research Center, University of Massachusetts Amherst.
A John Wiley & Sons, Ltd., Publication
This edition first published 2012 © 2012 by John Wiley & Sons, Ltd
Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical and Medical business to form Wiley-Blackwell.
Registered office: John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial offices: 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 111 River Street, Hoboken, NJ 07030-5774, USA
For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell.
The right of the author to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.
Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.
Library of Congress Cataloging-in-Publication Data
Quaternary environmental change in the tropics / edited by Sarah E. Metcalfe and David J. Nash. p. cm. Includes bibliographical references and index. ISBN 978-1-118-34325-8 (cloth)1. Paleoclimatology–Tropics. 2. Paleoclimatology–Quaternary. 3. Tropics–Climate. I. Metcalfe, S. E. (Sarah E.) II. Nash, David J. QC884.5.T73Q38 2012 551.6913–dc23 2012010714
A catalogue record for this book is available from the British Library.
Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.
Cover image: View of the rapidly retreating tropical glacier on Uhuru Peak, Mount Kilimanjaro, Tanzania. In the background Mount Montemero, near Arusha. © iStockphoto.com/Avatar.
Cover design: www.designdeluxe.com
Set in 9/12 pt Meridien by Toppan Best-set Premedia Limited
1 2012
v
Contents
List of contributors, xiPreface, xiiiAcknowledgements, xiv
I Global contexts, 1
1 Introduction, 3Sarah E. Metcalfe and David J. Nash1.1 Why the tropics matter, 3
1.1.1 Defining the tropics, 31.1.2 Importance of the tropics, 4
1.2 Development of ideas, 81.2.1 Early ideas about tropical environmental change, 81.2.2 The twentieth century revolution, 91.2.3 Advances in modelling, 12
1.3 Establishment of the tropical climate system, 131.4 Drivers of tropical environmental change, 171.5 The tropics as drivers of change, 20
1.5.1 The tropics and greenhouse gas concentrations, 201.5.2 Impacts of low latitude volcanic eruptions, 221.5.3 Dust emissions from the tropics and subtropics, 23
1.6 Extra-tropical forcing, 241.7 Organisation of the volume, 24Acknowledgements, 25References, 25
2 Contemporary climate and circulation of the tropics, 34Stefan Hastenrath2.1 Introduction, 342.2 Diurnal and local processes, 342.3 Planetary context, 352.4 Regional circulation systems, 36
2.4.1 Jet streams, 362.4.2 Subtropical highs and trade winds, 372.4.3 Equatorial trough zone, 372.4.4 Monsoons, 382.4.5 Equatorial zonal circulations, 38
vi Contents
2.5 Climatic variability, 392.5.1 Southern Oscillation and El Niño, 392.5.2 Indian Monsoon, 402.5.3 Northeast Brazil, 412.5.4 Sahel, 412.5.5 Timescales of variability, 42
2.6 Concluding remarks, 42References, 42
II Regional environmental change, 45
3 Tropical oceans, 47Jan-Berend W. Stuut, Matthias Prange, Ute Merkel and Silke Steph3.1 Tropical oceans in the global climate system, 47
3.1.1 Modern climatology, 473.1.2 El Niño–Southern Oscillation and its relatives, 503.1.3 Solar and volcanic radiative forcing of tropical oceans, 513.1.4 Tropical oceans and monsoons, 533.1.5 The tropical oceans as part of the global conveyor belt, 53
3.2 Reconstructing past ocean conditions, 553.2.1 Proxies for SST and SSS, 553.2.2 Reconstructing continental climate using marine archives, 57
3.3 Tropical oceans throughout the Quaternary, 573.3.1 Glacial–interglacial cycles, 573.3.2 Early Quaternary (the ‘41-kyr world’), 573.3.3 Mid-Pleistocene Transition, 583.3.4 Late Quaternary (the ‘100-kyr world’), 60
3.4 The past 20 000 years, 603.4.1 The Last Glacial Maximum, 603.4.2 Glacial termination: an active role for the tropics?, 613.4.3 History of the equatorial Pacific and the state of ENSO, 653.4.4 The Holocene, 66
3.5 Outlook, 68References, 69
4 Africa, 79David J. Nash and Michael E. Meadows4.1 Introduction, 794.2 Potential climate forcing factors, 854.3 Mediterranean North Africa, 88
4.3.1 Contemporary climate and sources of palaeoenvironmental information, 884.3.2 Longer records, 884.3.3 The Last Glacial Maximum, 894.3.4 The last glacial–interglacial transition, 924.3.5 The Holocene, 93
4.4 The Sahara and the Sahel, 944.4.1 Contemporary climate and sources of palaeoenvironmental
information, 94
Contents vii
4.4.2 Longer records, 954.4.3 The Last Glacial Maximum, 964.4.4 The last glacial–interglacial transition, 974.4.5 The Holocene, 99
4.5 Equatorial Africa, 1034.5.1 Contemporary climate and sources of palaeoenvironmental
information, 1034.5.2 Longer records, 1044.5.3 The Last Glacial Maximum, 1074.5.4 The last glacial–interglacial transition, 1094.5.5 The Holocene, 110
4.6 Southern Africa, 1134.6.1 Contemporary climate and sources of palaeoenvironmental
information, 1134.6.2 Longer records, 1134.6.3 The Last Glacial Maximum, 1184.6.4 The last glacial–interglacial transition, 1214.6.5 The Holocene, 122
4.7 Synthesis, 1274.8 Directions for future research, 129References, 129
5 India, Arabia and adjacent regions, 151Ashok K. Singhvi, Nilesh Bhatt, Ken W. Glennie and Pradeep Srivastava5.1 Introduction, 1515.2 Quaternary of India and Tibet, 153
5.2.1 Arid and semi-arid regions, 1545.2.2 Aeolian sands, 1545.2.3 Aeolian dust (loess deposits), 1595.2.4 Volcanic ash, 1595.2.5 Lacustrine records, 1615.2.6 Peat deposits, 1665.2.7 Calcretes, 1675.2.8 Coastal records, 1675.2.9 Fluvial records, 1715.2.10 Cave deposits, 176
5.3 Quaternary of the Arabian Sea and Bay of Bengal, 1765.4 Quaternary of Arabia and the Middle East, 177
5.4.1 Fluvial (wadi) systems, 1795.4.2 Lacustrine (and sabkha) records, 1805.4.3 Cave deposits, 1825.4.4 Aeolian sands, 1835.4.5 Gypsum in dunes, 1875.4.6 Late Quaternary Persian (Arabian) Gulf, 1885.4.7 Dating aeolian sediment supply, 1895.4.8 Climatic optimum and modern Arabian civilisation, 1915.4.9 Summary of environmental changes in Arabia and the Middle East, 191
viii Contents
5.5 Conclusions, 192Acknowledgements, 196References, 196
6 China and Southeast Asia, 207Dan Penny6.1 The South and Southeast Asian region as a component of the
Earth system, 2076.2 Setting the stage for the Quaternary: environmental context and controls, 2116.3 Regional syntheses, 214
6.3.1 China, 2146.3.2 Indochina, 2216.3.3 Sundaland and Wallacea, 225
6.4 The Asian tropics during the Quaternary: driver of planetary change?, 229References, 230
7 Australia and the southwest Pacific, 236Peter Kershaw and Sander van der Kaars7.1 Introduction, 2367.2 Northeastern Australia, 240
7.2.1 Orbital timescale, 2407.2.2 Suborbital timescale, 2447.2.3 Termination 1 and the Holocene, 244
7.3 Northern Australia, 2457.3.1 Orbital timescale, 2457.3.2 Termination 1 and the Holocene, 246
7.4 Northwestern Australia, 2477.4.1 Orbital timescale, 2477.4.2 Termination 1 and the Holocene, 249
7.5 Western Australia, 2497.5.1 Orbital timescale, 2497.5.2 Termination 1 and the Holocene, 251
7.6 Central Australia, 2517.6.1 Orbital timescale, 2517.6.2 Termination 1 and the Holocene, 253
7.7 Southwest Pacific Islands, 2537.8 General discussion and conclusions, 254
7.8.1 Early Quaternary, 2547.8.2 Late Quaternary cyclicity and abrupt events, 2557.8.3 Late Quaternary climate alterations, 256
Acknowledgements, 258References, 258
8 Latin America and the Caribbean, 263Mark B. Bush and Sarah E. Metcalfe8.1 Introduction, 2638.2 Precursor to the Quaternary, 264
8.2.1 Climatic consequences of closure of the Isthmus of Panama, 2678.2.2 Biotic consequences of closure of the Isthmus of Panama, 267
Contents ix
8.3 Climate mechanisms, 2678.3.1 Modern climatology, 2678.3.2 The forcing of Neotropical climates, 270
8.4 Long term climate forcings and cycles, 2718.4.1 Eccentricity, 2718.4.2 Precessional cycles and precipitation patterns, 2728.4.3 Precession as a mixed signal, 273
8.5 Records of climate change, 2748.5.1 Glacial advance and the LGM, 2748.5.2 Glacial cooling, 2768.5.3 Glacial-age precipitation, 2788.5.4 The status of the refugial hypothesis of tropical diversity, 2848.5.5 The last deglaciation, 2868.5.6 The early-mid Holocene, 2878.5.7 Late Holocene oscillations, 289
8.6 Other climate forcings, 2898.6.1 Millennial-scale oscillations, 2898.6.2 Solar cycles, 2918.6.3 El Niño–Southern Oscillation, 292
8.7 El Niño records, 2928.7.1 Archaeology, 2928.7.2 Historical records, 2938.7.3 Tree ring records, 2938.7.4 Corals, 2948.7.5 Sedimentary records, 295
8.8 Climate and societies, 2968.8.1 Early agriculture, 2968.8.2 Cultural collapse, 297
8.9 Conclusions, 298Acknowledgements, 301References, 301
III Global syntheses, 313
9 Modelling of tropical environments during the Quaternary, 315Zhengyu Liu and Pascale Braconnot9.1 Introduction, 3159.2 Tropical climate in the Holocene: response to orbital forcing, 316
9.2.1 Orbital forcing, 3169.2.2 Monsoon response, 3179.2.3 SST response and oceanic feedback, 3199.2.4 Precession forcing and obliquity forcing, 3249.2.5 Ecosystem response and feedback, 330
9.3 Tropical climate at the LGM: the roles of GHGs and ice sheet forcing, 3339.3.1 Greenhouse gases and ice sheet forcing, 3339.3.2 Temperature response and climate sensitivity, 3339.3.3 Monsoon and hydrological response, 3389.3.4 Ecosystem response and feedbacks, 339
x Contents
9.4 Tropical climate variability, 3399.4.1 ENSO and ocean–atmosphere interaction, 3409.4.2 Abrupt change of monsoon climate, 3439.4.3 Tropical variability and its interaction with high-latitude variability, 344
9.5 Summary and further discussion, 3499.5.1 Summary, 3499.5.2 Other issues in Quaternary tropical climate modelling, 3509.5.3 Climate models of intermediate complexity, 3509.5.4 Perspective of Earth system modelling of past climate, 351
References, 352
10 Historical environmental change in the tropics, 360Georgina H. Endfield and Robert B. Marks10.1 Introduction, 36010.2 Climate change and society in the tropics in the last 1000 years, 361
10.2.1 Climate variability and harvest history in China, 36510.2.2 Climate and crisis in colonial Mexico, 370
10.3 Exploring anthropogenic impacts in the tropics, 37510.3.1 Deforesting China, 37810.3.2 Exploring pre- and post-conquest land use changes in
central Mexico, 38010.4 Recent and future environmental changes in the ‘vulnerable’ tropics, 382References, 384
11 Past environmental changes, future environmental challenges, 392David J. Nash and Sarah E. Metcalfe11.1 Patterns of tropical environmental change, 392
11.1.1 Last Interglacial, 39211.1.2 Last Glacial Maximum, 39411.1.3 The last deglaciation or last glacial–interglacial transition, 39611.1.4 The Holocene, 398
11.2 Forcings, 40111.3 Future change in the tropics, 402
11.3.1 Climate responses, 40211.3.2 Water resources, 40411.3.3 Biodiversity, 406
11.4 The tropics as drivers of change, 40611.5 Conclusions, 408References, 409
Index, 412Colour plate pages fall between pp. 210 and 211
List of contributors
xi
Nilesh BhattDepartment of Geology, M.S. University of Baroda, Vadodara 390 002, India. Email: [email protected].
Pascale BraconnotIPSL/LSCE, Laboratoire Mixte CEA-CNRS-UVSQ, Orme des Merisiers bat 712, 91191 Gif-sur-Yvette CEDEX, France. Email: [email protected].
Mark B. BushBiological Sciences Department, Florida Institute of Technology, 150 W. University Blvd, Melbourne, Florida FL 32901, USA. Email: [email protected].
Georgina H. EndfieldSchool of Geography, University of Nottingham, Nottingham NG7 2RD, United Kingdom. Email: [email protected].
Ken W. GlennieSchool of Geosciences, University of Aberdeen, Aberdeen AB24 3UE, United Kingdom. Email: [email protected].
Stefan HastenrathDepartment of Atmospheric and Oceanic Sciences, University of Wisconsin-Madison, 1225 W. Dayton Street, Madison, Wisconsin WI 53706, USA. Email: [email protected].
Peter KershawCentre for Palynology and Palaeoecology, School of Geography and Environmental Science, Monash University, Melbourne, Victoria 3800, Australia. Email: [email protected].
Zhengyu LiuCenter for Climatic Research and Department of Atmospheric and Oceanic Sciences, University of Wisconsin-Madison, 1225 W. Dayton Street, Madison, Wisconsin WI 53706, USA. Email: [email protected].
Robert B. MarksDepartment of History, Whittier College, 13406 E. Philadelphia Street, Whittier, CA 90608, USA. Email: [email protected].
Michael E. MeadowsDepartment of Environmental and Geographical Science, University of Cape Town, Private Bag X01, Rondebosch 7701, South Africa. Email: [email protected].
Ute MerkelMARUM, Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany. Email: [email protected].
Sarah E. MetcalfeSchool of Geography, University of Nottingham, Nottingham NG7 2RD, United Kingdom. Email: [email protected].
David J. NashSchool of Environment and Technology, University of Brighton, Lewes Road, Brighton BN2 4GJ, United Kingdom.andSchool of Geography, Archaeology, Environmental Studies, University of the Witwatersrand, Private Bag 3, Johannesburg 2050, South Africa. Email: [email protected].
Dan PennySchool of Geosciences, University of Sydney, Sydney, NSW 2006, Australia. Email: [email protected].
Matthias PrangeMARUM, Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany. Email: [email protected].
Ashok K. SinghviGeosciences Division, Physical Research Laboratory, Navarangpura, Ahmedabad 380 009, India. Email: [email protected].
xii List of contributors
Pradeep SrivastavaWadia Institute of Himalayan Geology, GMS Road, Dehradun 248 001, India. Email: [email protected].
Silke StephDepartment of Geosciences, University of Bremen, 28334 Bremen, Germany. Email: [email protected].
Jan-Berend W. StuutNIOZ Royal Netherlands Institute for Sea Research, Den Burg, the Netherlands.andMARUM, Center for Marine Environmental Sciences, University of Bremen, Bremen, Germany. Email: [email protected].
Sander van der KaarsSchool of Geography and Environmental Science, Monash University, Melbourne, Victoria 3800, Australia. Email: [email protected].
Preface
xiii
The global climate changes that led to the expan-sion and contraction of polar ice sheets over the past 2.58 million years were associated with equally dramatic changes in tropical and subtropi-cal terrestrial and marine environments. Changes in global temperature, fluctuations in sea level and alterations of the position of the major oceanic and atmospheric circulation systems led to shifts in con-tinental vegetation zones, changes in the hydrology and ecology of tropical lake and drainage systems, and the expansion and contraction of tropical mountain glaciers and sandy deserts. Until recently, it was thought that such changes were largely a response to fluctuations in the distribution of high latitude ice cover. However, there is increasing rec-ognition that the tropics have acted as drivers of global climate change over a range of timescales. This is in part due to their importance in terms of solar radiation receipt and the resulting energet-ics of the global circulation, but also because of the role tropical oceans and ecosystems play in regulating greenhouse gases and the global hydro-logical cycle.
Despite the significance of the tropics for global climate change debates, there has not been a volume that attempts to synthesise understanding of how tropical environments as a whole have changed over the past 2.6 million years. The overall aim of Quaternary Environmental Change in the Tropics is to fill this gap and provide a readable synthesis of the large (and growing) literature on the climatic and broader environmental shifts that occurred in tropical terrestrial and marine systems during the Pleistocene and Holocene. It is mainly targeted
at final-year undergraduates and research special-ists, but will we hope provide an introduction to tropical Quaternary research for a variety of other readers. Inevitably, as with any edited volume, the authors have tackled their individual chapters in different ways, reflecting their own areas of special-ism and the key research questions that need to be addressed in different tropical regions. However, we hope that this book can provide a basic frame-work for future regional and global assessments of tropical Quaternary environments.
The idea of producing this book originated purely by chance when we met at a business meeting at the Royal Geographical Society in London in 2006. We had both had many years of experience of working on the Quaternary of tropical and sub-tropical regions – SEM in the neotropics (especially Mexico) and DJN in various parts of southern Africa. We both taught specialist courses on tropical Quaternary environments in our respective institu-tions. We had both had separate conversations with commissioning editors bemoaning the lack of books specifically concerning the Quaternary of tropical regions, but had both resisted all suggestions that we should individually write such a book. However, over afternoon tea in Lowther Lodge, we (with hindsight, foolishly) agreed that it might not be so bad to put together an edited volume on the subject. Five years later, and with the considerable goodwill of all of the authors involved, you are reading the end product.
Sarah E. MetcalfeDavid J. Nash
Acknowledgements
xiv
In addition to the editors, who reviewed all the individual chapters, numerous external referees, selected for their expertise in specific aspects of tropical Quaternary environments, provided con-structive and conscientious reviews of manuscripts. These included: Rodolfo Acuna-Soto, UNAM, Mexico; Philip Barker, Lancaster University, UK; Paul Bishop, University of Glasgow, UK; Sarah Davies, University of Aberystwyth, UK; Sherilyn Fritz, University of Nebraska, USA; Paul Hesse, Macquarie University, Australia; Dominic Knive-ton, University of Sussex, UK; Glenn McGregor, University of Auckland, New Zealand; Sharon Nicholson, Florida State University, USA; Bette Otto-Bleisner, NCAR, USA; Adrian Parker, Oxford Brookes University, UK; Chris Turney, University of New South Wales, Australia; Frank Shillington, University of Cape Town, South Africa; David Thomas, University of Oxford, UK. We would also like to thank those anonymous
reviewers who provided helpful comments on our original proposal – thank you whoever you were!
The majority of the photographs and line dia-grams within this volume are the authors’ own. We are, however, indebted to a number of publishers for their permission to either reproduce or adapt figures used in this book. These are credited within the figure captions.
Finally, our thanks go to all the authors for persevering with us during the production of this volume, and to Ian Francis and Kelvin Mat-thews at Wiley-Blackwell Publishing for their encouragement, endless patience and assistance during the long, painful gestation period leading to the publica tion of Quaternary Environmental Change in the Tropics.
Sarah E. MetcalfeDavid J. Nash
I
Global contexts
Quaternary Environmental Change in the Tropics, First Edition. Edited by Sarah E. Metcalfe and David J. Nash.
© 2012 John Wiley & Sons, Ltd. Published 2012 by John Wiley & Sons, Ltd.
3
1.1 Why the tropics matter
1.1.1 Defining the tropicsIn its strictest sense, the term ‘tropics’ refers to those parts of the world that lie between the Tropic of Cancer (23.4378 °N) and the Tropic of Capricorn (23.4378 °S). These latitudinal boundaries mark, respectively, the most northerly and southerly position at which the Sun may appear directly overhead at its zenith. Indeed, the word ‘tropical’ comes from the Greek tropikos, meaning ‘turn’, since the tropics of Cancer and Capricorn mark the latitudes at which the Sun appears to turn in its annual motion across the sky. Unfortunately, the outer boundary of the tropics sensu lato cannot be defined in such rigid astronomical terms. Certainly latitude is a major factor determining the distribu-tion of tropical climatic zones, through its control on solar radiation receipt (Fig. 1.1), but regions with distinctive climatological, physical or biologi-cal characteristics are not easily delimited by linear boundaries.
The tropics include a diverse range of environ-ments and climates (see Chapter 2). Rather than being uniformly hot and wet, the area between the tropics of Cancer and Capricorn encompasses some of the wettest regions on Earth (e.g. the rainforests of western Amazon and central Congo basins) as well as some of the driest (e.g. the Atacama Desert of northern Chile and Peru). The one feature common to all tropical climates is a relatively limited seasonal fluctuation in insolation and tem-
perature. Instead, differences in the quantity and temporal distribution of available moisture account for regional and seasonal variability (Savage et al., 1982).
Authors such as Reading et al. (1995) have pro-vided useful overviews of the various attempts to define the climates of the tropics. Some of the most widely used classifications are based directly upon meteorological parameters such as rainfall and temperature. The classic Köppen–Geiger system (Fig. 1.2), for example, centres on the concept that natural vegetation is the best expression of climate, with climate zone boundaries positioned with vegetation distribution in mind. The Köppen– Geiger scheme combines average annual and monthly temperatures and precipitation, and the seasonality of precipitation. Köppen (1936) defined tropical climates as those exhibiting a constant high temperature (at sea level and low elevations), with all 12 months of the year having average temperatures of 18 °C or higher. This classification excludes cooler highland regions (defined as areas above 900 m elevation), which comprise around 25% of the total land area within the tropics (Reading et al., 1995). These regions still receive high amounts of solar radiation and do not have a pronounced winter season, but temperatures may be sufficiently depressed to affect biological activity. Rainfall levels and the seasonal distribu-tion of precipitation are then used to subdivide tropical climates into tropical rainforest (Af), tropi-cal monsoon (Am), and tropical savanna climates
CHAPTER 1
IntroductionSarah E. Metcalfe and David J. Nash
4 Chapter 1
Fig. 1.1 Solar radiation received at the Earth’s surface assuming an atmospheric transmission coefficient of 0.60 (after McGregor and Nieuwolt (1998) Tropical Climatology, John Wiley & Sons Ltd.).
20
5
10
Meg
ajou
les/
day
15
0J F M A M J J A S O N D
20
5
10
15
45°S
20°S
0°S
10°N
30°N 20°S20°S
30°N
30°N
10°N
0°N
60°N60°N
0
(Aw). Köppen (1936) includes a range of other climate types within the tropics sensu stricto, includ-ing tropical and subtropical steppe (BSh), tropical desert (BWh) and humid subtropical climates (Cfa, Cwa). Some highland areas within the tropics also exhibit a temperate climate with dry winters (Cwb).
Working from an agricultural perspective, Jackson (1989) split the tropics into three zones (Humid, Wet and Dry, and Dry) according to the level and seasonal distribution of rainfall (Fig. 1.3). This classification recognises the importance of seasonality for agricultural productivity, and is less focused on natural vegetation zones than the Köppen–Geiger scheme. Other attempts to classify climates within the tropics are based around hydro-meteorology, with climate types defined according to the balance of precipitation inputs and evapotranspiration outputs. Garnier (1958), for example, differentiated humid tropical climates according to the number of months in which actual evapotranspiration equals potential evapotranspiration. The ratio of precipitation to potential evapotranspiration has also been used by Middleton et al. (1997), drawing upon Thorn-thwaite (1948) and Meigs (1953), to define an aridity index for categorising dry tropical climates.
In this volume, the astronomical definition of the tropics is used to broadly demarcate the geo-graphical scope of each of the substantive chapters. However, recognising that climate boundaries are fuzzy and mobile in the present day and that
climate zones shifted by many degrees of latitude during the various glacials and stadials that charac-terise the Quaternary Period, coverage in many chapters extends polewards north and south of 23.4378° into the subtropics where appropriate. The Quaternary Period is defined here as encom-passing the last 2.58 million years of the Earth’s history (Gibbard et al., 2010), the timescale ratified by the Executive Committee of the International Union of Geological Sciences in June 2009.
1.1.2 Importance of the tropicsIn comparison with the mid latitude regions of Europe and North America, our understanding of Quaternary palaeoenvironments in the tropics is, at best, patchy for some areas and extremely poor to non-existent in others. As a result, any attempt to expand our understanding of past environmental conditions in low latitude regions is likely to be a valuable contribution to knowledge. However, more significantly, understanding tropical palaeoenvironments may also be key to establish-ing the drivers of global environmental change. As discussed in section 1.5 of this chapter, the last 10–15 years have seen an increasing recognition of the significance of tropical regions in climate forcing (e.g. Kerr, 2001; Broecker, 2003). The tropical oceans and atmosphere play an important con temporary role in redistributing incoming solar radiation and would have been instrumental in transmitting past variations in radiation receipt to
Intro
du
ction
5
Fig. 1.2 The Köppen–Geiger climate classification system updated with CRU TS 2.1 temperature and VASClimO v1.1 precipitation data for 1951 to 2000 (after Kottek et al., 2006). (See Colour Plate 1)
6 C
hap
ter 1
Fig. 1.3 Classification of the tropics based on the seasonal distribution of rainfall (after Jackson (1989) Climate, water and agriculture in the tropics, Longman; Reading et al. (1995) Humid tropical environments, John Wiley & Sons Ltd.) (See Colour Plate 2)
Introduction 7
led to genetic selection and innovation. Evidence from Atlantic and Indian Ocean cores suggests that climate in the African subtropics fluctuated between markedly wetter and drier conditions in time with orbital variations. De Menocal (1995) identifies progressive shifts in African climate variability and increasing aridity after 3.0–2.6 Myr, 1.8–1.6 Myr and 1.2–0.8 Myr, coincident with the onset and intensification of high-latitude glacial cycles. Anal-ysis of well-dated mammal fossil databases suggests African faunal assemblage and, perhaps, speciation changes coincident with the appearance of more varied and open habitats at 2.9–2.4 Myr and after 1.8 Myr. These periods roughly coincide with key junctures in hominid evolution, including the emergence of the genus Homo around 2.5 Myr (de Menocal, 2004).
Environmental changes, particularly during the late Pleistocene, may also have played a role in shaping pathways for the dispersal of early modern humans around the Earth. For example, corridors formed by pluvial ‘mega-lakes’ during Marine Isotope Stage (MIS) 5 may have provided trans-Saharan pathways for humans migrating ‘out of Africa’, offering an alternative route to the Nile Valley (Drake et al., 2011). Biogeographical and palaeohydrological evidence (ibid.) suggests that similar migration pathways across the Sahara, in the form of linked lakes, rivers and inland deltas, may have existed during the early Holocene (see Chapter 4). The migration of humans into Aus-tralia, either as a single or several successive waves, also appears to have been influenced by global environmental changes. There is still much debate about the timing of the earliest arrivals; the minimum widely-accepted timeframe places this at around 45 kyr BP (e.g. O’Connell and Allen, 2005) with an upper estimate of around 60 kyr BP (e.g. Roberts et al., 1990, 1993, 1994). Regardless, this migration was achieved during the closing stages of the Pleistocene, when sea levels were much lower than they are today (see section 1.2.2 and Fig. 1.5) and an extensive land bridge existed across the Arafura Sea, Gulf of Carpentaria and Torres Strait (Lourandos, 1997).
The tropics are also highly important because they support very large numbers of species compared
other parts of the Earth system. Tropical oceans and landmasses also act as sources and sinks of green-house gases, with, for example, tropical forests acting as contemporary CO2 sinks (Cox et al., 2000) and tropical oceans (IPCC, 2007) and major river and wetland systems such as the Amazon (Richey et al., 2002) outgassing CO2 to the atmosphere. The decay of vegetation within tropical wetlands is a major source of contemporary biogenic CH4 (Lou-lergue et al., 2008). Indeed, much of the variation in CH4 concentration recorded in the Antarctic Vostok ice core coincides with fluctuations in the size and extent of tropical lakes and wetlands (cf. Raynaud et al., 1988; Chappellaz et al., 1990; Brook et al., 2000). Tropical forest ecosystems and soils are known to act as important contemporary sources for atmospheric N2O, with N2O emissions typically increasing during wet season conditions and falling during drier periods. Data from the Antarctic EPICA Dome C ice coring site suggest that biospheric changes in the low latitudes may have been instrumental in controlling emissions of N2O on glacial–interglacial timescales (Schilt et al., 2010). The precise mechanism through which this process operated is unknown, but deep water changes in the North Atlantic, and associated Dansgaard–Oeschger (D–O) events, may have had an influence on atmospheric N2O levels, either through indirect changes in low latitude ecosys-tems and soils or by a direct change in marine N2O production (Schmittner and Galbraith, 2008).
Identifying changes in tropical environments over the past 2–3 million years may have consider-able resonance for our understanding of the drivers of human evolution. Recent fossil discoveries and advances in the analysis of existing fossil collec-tions, coupled with the emergence of high resolu-tion palaeoclimatic records, have focused attention on the role that past shifts in climate variability may have had in the evolutionary history of African mammalian fauna, including early hominids (de Menocal, 2004). Although this topic is still hotly debated, the basic premise is that large-scale shifts in climate over the course of the last 5–6 million years altered the ecological composition of African landscapes, thereby generating specific faunal adaptation or speciation pressures which ultimately
8 Chapter 1
with other regions of the globe (Mace et al., 2005). This is especially true of the tropical moist forests which show the highest global levels of species and family richness and of endemism. There is increasing concern about the threat posed to tropical ecosystems by both direct human action and by future climate change (itself probably anthropogenic). Although we hear most about the tropical rainforest (e.g. Hubbell et al., 2008), it is the tropical dry forests that have been most affected to date, with about half being lost to cultivation. Mapping of species loss (mammals, birds and amphibians) since AD 1500 shows a significant concentration in tropical latitudes, especially in the tropical Americas and Australasia (Baillie et al., 2004). As well as direct loss of habitat and species, with their economic and cultural values, changes in tropical ecosystems have wider implications because of their role in the global biogeochemical and hydrological cycles. Some of these issues are discussed further in section 1.5 of this chapter.
As the chapters within this volume highlight, tropical vegetation and landscape systems have shown considerable sensitivity to climatic changes during the Quaternary Period; by inference, tropical landscapes might be expected to show a similar scale of response to future human-induced and natural environmental changes. Couplings between vegetation cover and the susceptibility of the ground surface to water or wind erosion mean that shifts in vegetation density and type in response to anthropogenic and climatic changes may act to either stabilise or destabilise land surfaces. Tropical fluvial systems, for example, are highly sensitive to external forcings in the form of short and long term shifts in effective precipitation and vegetation cover. The nature of the response within individual fluvial systems reflects the antecedent conditions, the degree and duration of the environmental change, possibly the rate of change, and whether the change is sufficient to trigger in-channel threshold-crossing events (Thomas, 2008). In northeast Queensland, Australia, for example, a long-term deterioration of the rainforest vegetation cover after 78 kyr BP, steepening after 40 kyr BP with a shift toward dry sclerophyll forest, led to widespread soil erosion and the accumulation of
fine alluvial fan deposits within fluvial systems fronting the eastern highlands of the Great Dividing Range (Nott et al., 2001; Thomas et al., 2001, 2007). Many tropical environments contain relict landforms (and their associated sediments) formed under previously wetter or drier conditions, which may be reactivated under future climatic change scenarios. Environmental modelling studies in the Kalahari Desert, for example, have suggested that large areas of presently stable and well-vegetated ‘fossil’ Pleistocene sand dunes could be reactivated if changes in wind regime and a reduction in vegetation cover (in response to warming and reduced available moisture) occur as a result of twenty- first century climate warming (Thomas et al., 2005).
1.2 Development of ideas
1.2.1 Early ideas about tropical environmental changeThe possibility that high and mid latitude regions had undergone major environmental changes was recognised as early as 1779 when the Swiss aristocrat Horace-Bénédict de Saussure identified granite boulders on the limestone slopes of the Jura ranges that had been transported some 90 km from their source in the Mont Blanc massif (de Saussure, 1779). In keeping with contemporary ideas that the Earth’s features had been shaped by the biblical Great Flood, de Saussure suggested that these ‘erratics’ had been moved by water. Bernard Friedrich Kuhn was the first to propose that the boulders had, in fact, been transported by more extensive glaciers (Kuhn, 1787; de Beer, 1953), a conclusion reached independently some eight years later by James Hutton following a visit to the Jura (Hutton, 1795). John Playfair famously extended these ideas in 1802, and, by the time of the publication of Etudes sur les Glaciers by Louis Agassiz in 1840, the concept of Die Eiszeit or large scale Ice Age in Europe was well established.
In contrast, for many years, the dominant view of the tropics was that they had seen very little climatic change, with core areas such as Amazonia remaining unaffected by the cycles of glaciation and deglaciation that drove massive environmental
Introduction 9
changes in higher latitudes (Richards, 1952). This was despite the suggestion made by Louis Agassiz, after mistaking deeply weathered bedrock for glacial diamicton during a visit to Brazil in 1865–1866, that the western Amazon basin had been glaciated (Agassiz, 1868). As early as 1850, the Scottish missionary and explorer David Livingstone had recognised that salt accumulations in the Makgadikgadi Depression of Botswana were ‘the remains of the very slightly brackish lakes of antiq-uity’ (Livingstone, 1857: 67). However, some of the main advances in our understanding of low latitude palaeoenvironments were made in the USA (see Goudie, 1999). John Strong Newberry, for example, suggested that the landscapes of the Colorado Plateau were ‘formerly much better watered than they are now’ (1861: 47). In 1863, Thomas Francis Jamieson was the first to propose that wetter conditions and higher lake levels in the southwest USA were equated with high latitude glacial episodes (a concept often termed the ‘glacial = pluvial’ hypothesis). This idea was adopted by Israel Russell (1885) and Grove Karl Gilbert (1890) to explain the origins of strandlines within the Pleistocene ‘pluvial’ lakes Lahontan and Bonneville (Fig. 1.4). The notion that low latitude pluvials were synchronous with high latitude gla-cials was widely accepted and was ultimately assumed to apply across the tropics. The corollary of this view, that post-glacial times were character-ised by desiccation, was also widely applied (Goudie, 1972), most notably in the Asian and African tropics and subtropics (Goudie, 1999). In southern Africa, for example, Schwarz (1923) proposed a grandiose scheme to divert rivers from the north to flood the Kalahari Basin as a means to ameliorate a supposed progressively drying climate.
1.2.2 The twentieth century revolutionBy the mid 1940s, challenges to the post-glacial desiccation and ‘glacial = pluvial’ hypotheses began to emerge. One of the most important conceptual advances was the recognition that some tropical areas that are now relatively moist, far from pro-gressively desiccating had been significantly drier in the past. The main evidence for this came first from the identification of ancient dunefields in
Fig. 1.4 (a) Sketch of Lake Bonneville shorelines and (b) Map of Lake Bonneville by G.K. Gilbert (from Gilbert, 1890, images courtesy of USGS).
(a)
(b)
10 Chapter 1
(1976) revealed fluctuations in the extent of the Sahara during the late Pleistocene. Some of the longest and highest resolution records available for the tropics now come from marine settings, and provide important insights into ocean palaeotem-peratures, terrestrial chemical environments and variations in the offshore transport of dust, pollen and fluvial sediments (e.g. Larrasoaña et al., 2003; Peterson and Haug, 2006) (see Chapter 3, section 3.2.2). The technology used to extract marine cores has been adapted and utilised on land, such that a number of long terrestrial records are now also available for the tropics (e.g. Trauth et al., 2003).
At the interface between land and the oceans, the Quaternary has seen major changes in relative sea level. A number of different factors may be involved, especially locally. However, at the global scale, glacio-eustatic change dominates, reflecting the volume of water locked up in ice sheets and glaciers (with glacial or stadial periods being marked by low sea levels). Although the associa-tion between changes in ice volume and sea level was put forward in the early twentieth century, major advances in reconstructing sea level were made during the 1960s and 1970s. Key sea level reconstructions (covering about the last 400 kyr) have come from tropical areas, primarily the Huon Peninsula of Papua New Guinea (Aharon and Chappell, 1986) (see Chapter 6) and Barbados (Fairbanks, 1989). Both these are based on dated sequences of coral reefs. An updated version of Fairbanks’ reconstruction for the period since the LGM is shown in Fig. 1.5. This indicates that sea level was 120–125 m lower than present at the LGM. Fairbanks identified two periods of very rapid rise (>20 m) associated with meltwater pulses 1A and B, which he dated to around 12 kyr and 9.5 kyr BP. These events were later re-dated to 14 and 11 kyr BP, following a reassessment of the record using U-Th dating (Bard et al., 1990). Bard et al. also reported sea level of +5 to +10–m in the last interglacial (MIS 5e). This and subsequent studies have confirmed the coincidence of periods of high sea level with insolation maxima, consistent with Milankovitch forcing (see section 1.4 of this chapter). The impact of these changes in sea level was particularly pronounced in areas with exten-
Texas (Price, 1944), and then vegetated ergs along the equatorward margins of the southern Sahara (Grove, 1958; Grove and Warren, 1968), northern Kalahari (Grove, 1969) and Indian deserts (Goudie et al., 1973). Prior to the advent of luminescence dating in the 1980s, the ages of these aeolian deposits could only be estimated relative to sedi-ments that could be radiocarbon dated, but their mere existence was a nail in the coffin for progres-sive desiccation.
The 1970s represented a major shift in our understanding of low latitude palaeoenvironments, as detailed records from lake basins in tropical Africa (e.g. Grove and Goudie, 1971; Grove et al., 1975; Street and Grove, 1976) and elsewhere (cf. Street-Perrott et al., 1979) started to be published. Views of the stability of the tropical rainforest also changed (Flenley, 1979). With these studies, it became apparent that the story of the tropics was much more complex than previously thought, with many areas exhibiting fluctuating rather than con-sistently high lake levels around the time of the Last Glacial Maximum (LGM). Compilations of global lake level fluctuations (e.g. Street-Perrott et al., 1979) served to demonstrate that, in many ways, what happened in the southwest USA, the home of the ‘glacial = pluvial’ hypothesis, was the exception rather than the norm. The picture that emerges today, as summarised by each of the regional chapters in this volume, is that the mag-nitude and timing of climate change in different parts of the tropics and subtropics is considerably more complex than pioneers such as Gilbert and Russell ever could have envisaged.
In parallel with the growth in knowledge about terrestrial tropical environments, our understand-ing of Quaternary stratigraphy in tropical oceans has been revolutionised since the 1950s (Imbrie and Imbrie, 1979). This has been due primarily to the introduction of new equipment for coring off-shore and deep-ocean sediments. Damuth and Fairbridge (1970), for example, used evidence from deep-sea piston cores taken in the Guiana Basin to suggest that an arid to semi-arid climate dominated large portions of equatorial South America during the Pleistocene glacial phases. Similarly, analyses of multiple cores off northwest Africa by Diester-Haas
Introduction 11
stratigraphy, electron-spin resonance, luminescence and cosmogenic radionuclide exposure dating intro-duced in more recent decades. For many of these techniques, the availability of mass spectrometry has permitted high temporal resolution dating of materials, including the micro-sampling of cave deposits (e.g. Wang X et al., 2007; Wang Y et al., 2008) and geochemical sediments such as calcrete (e.g. Candy et al., 2004).
Two examples serve to highlight the importance of the new dating tools for our understanding of tropical palaeoenvironments. First, the develop-ment of optically-stimulated luminescence (OSL) dating since the 1980s has allowed the age of depo-sition of a wide range of carbon-poor sediments to be determined, most notably those preserved within fossil dunes and other aeolian deposits (cf. Singhvi and Porat, 2008). This has led to the estab-lishment of detailed chronological frameworks for
sive continental shelves affecting marine currents, regional groundwater levels and the ease of migra-tion of terrestrial organisms including humans (see especially Chapters 6 and 7).
Advances in our understanding of tropical pal-aeoenvironments have been prompted, in part, by the availability of new avenues for environmental reconstruction, but also reflect the development of new chronological techniques (Goudie, 1999). The introduction of radiocarbon dating in the 1950s, for example, meant that it was possible, for the first time, to obtain age estimates from late Quaternary sediments and landforms rather than having to rely on stratigraphic correlation. The radiocarbon revolution was followed in the 1960s by the devel-opment of potassium-argon and uranium-series dating, dendrochronology and palaeomagnetism. These chronological tools were refined in the 1970s and 1980s, with new approaches such as amino-
Fig. 1.5 Composite record of relative sea level change over the last 32 kyr, based on data from Barbados. Data from Peltier and Fairbanks (2008) IGBP PAGES/World Data Center for Paleoclimatology, Data series 2008-101.
12 Chapter 1
reconstruction. Most early attempts to reconstruct changes in tropical flora were heavily reliant upon pollen analyses. However, it is now possible to utilise other plant remains such as macrofossils (e.g. those preserved within rodent middens; Betancourt et al., 1990; Pearson and Dodson, 1993; Holmgren et al., 2007) and phytoliths (e.g. Parker et al., 2004), not only to reconstruct terrestrial vegetation changes but also to identify shifts in CO2 concentration (Beerling and Woodward, 1993). Changes in terrestrial aquatic environments can be identified through the analysis of molluscs, diatoms and ostracods (e.g. Fritz et al., 1999; Holmes and Engstrom, 2005), whilst our understanding of changes in marine environments has been revolutionised through the analysis of foraminifera and other microorganisms such as radiolaria and coccoliths (cf. Lowe and Walker, 1997). The development of transfer functions – essentially variants on multiple linear regression models employed to establish relationships between biological data and environmental variables – now permits palaeoenvironmental parameters to be reconstructed quantitatively from fossil floral and faunal assemblages (e.g. Birks and Birks, 1980; Birks, 2005). The need for such transfer functions to reflect biologically meaningful relationships has to be borne in mind, however.
Finally, our understanding of tropical and sub-tropical environmental variability in recent centu-ries has greatly improved thanks to new efforts to tap the wealth of information contained within annually resolved proxies (e.g. corals, tree rings, speleothems). Climate chronologies derived from historical documentary materials are now availa-ble, for example, for large areas of Africa (e.g. Nicholson, 2000, 2001; Nash and Endfield, 2002, 2008; Grab and Nash, 2010; Nash and Grab, 2010) and show remarkable agreement with regional tree ring records (e.g. Therrell et al., 2006) and fossil coral (Zinke et al., 2004, 2005).
1.2.3 Advances in modellingThe application of computer modelling to palaeo-climate studies is now central to efforts to synthe-sise and understand change in climate systems and environments. The role of factors such as insolation forcing, tectonism and vegetation feedbacks have
many of the world’s desert regions (cf. Munyikwa, 2005) as well as major advances in our under-standing of how aeolian dunes evolve over time (e.g. Telfer and Thomas, 2006). Second, the rapid advances in cosmogenic radionuclide analysis in the last decade have provided a basis for exposure ‘dating′ of landforms, the quantification of erosion rates and other geologic applications in areas where opportunities for any form of chronological inves-tigation were once extremely limited. Cosmogenic radionuclide dating has been used, for example, to establish the timing of dunefield initiation in central Australia (Fujioka et al., 2005) and, alongside other techniques, to estimate residence times for ground-water in the Nubian Aquifer beneath the Western Desert in Egypt (Patterson et al., 2005).
Alongside chronological developments, there have been a number of other methodological improve-ments, including the application of an increasingly sophisticated range of field and laboratory appro-aches. These include new techniques for sedimen-tological and geochemical analysis which have offered important insights into Quaternary deposi-tional environments. Amongst the most significant of these was the advent of stable isotope analyses of sediments and biological remains in the 1950s. Oxygen isotope analysis, in particular, pioneered by Cesare Emiliani (1955), is now one of the most important tools in Quaternary stratig raphy and is routinely applied in a variety of terrestrial and marine contexts to reconstruct environmental signals such as palaeotemperature, water balance (P–E), precipitation source and amount (Leng, 2006). Stable carbon isotope analysis can be undertaken on either inorganic (authigenic calcite, biogenic carbonate), or organic C. In combination with measurements of C/N, δ13Corganic is widely used to determine the sources (C3 or C4 terrestrial veg-etation, aquatic macrophytes, algae) of organic matter coming into lacustrine systems. The applica-tion of compound specific δ13C analysis is particu-larly effective in this regard (e.g. Street-Perrott et al., 2004). More recent studies (e.g. Chase et al., 2009), have utilised variations in stable nitrogen isotope analyses as a means of establishing past rainfall levels.
A large (and growing) number of palaeoecologi-cal techniques are now available for environmental
Introduction 13
ments to keep the two elements together, but this wasn’t needed in later models. The advent of these coupled models allowed annual climatolo-gies and seasonal cycles to be reproduced (IPCC, 2001). This has been vital in efforts to model the El Niño Southern Oscillation (ENSO; see Chapter 9). The most recent development is the use of fully coupled Earth System Models such as HadGEM2-ES (dynamic vegetation response) and ECHAM5/JSBACH-MPIOM (e.g. Dallmeyer et al., 2010).
GCMs now dominate, but simpler models are still used where long time series are a key requirement (e.g. Crowley et al., 1992) and the run times of more comprehensive models would still be pro-hibitive even with the significant computing power now available. These models of intermediate complexity continue to play an important role in helping to understand long term climate change, including the role of Milankovitch cycles and tran-sitions between different climate modes (interglacial/glacial). Groot et al. (2011) use one of these models, CLIMBER (see also Chapter 9), to help interpret the arboreal pollen record from the Fuquene Basin in Colombia between 284 kyr and 27 kyr BP (see Chapter 8).
Models play a very important part in helping our understanding of tropical climate change. They have also helped us to appreciate the importance of the tropics in driving climate change, especially the role of the tropical oceans (Hostetler et al., 2006) and feedbacks from greenhouse gases (par-ticularly methane) (Loulergue et al., 2008). Unfor-tunately, there are still some parts of the world where climate models struggle to reproduce modern climate, and hence one can have little confidence in their use in palaeoclimatic studies. This is particularly the case in areas of complex terrain. The use of regional scale models and finer resolu-tion GCMs can help to address this (e.g. Hostetler et al., 1994).
1.3 Establishment of the tropical climate system
In the popular imagination, the tropics are both warm and wet, and it is the case that 56% of total global precipitation falls in the tropics (Wang and
all been explored in relation to tropical regions, with a particular emphasis on their impacts on monsoons. The application of modelling to tropical palaeoclimates is explored explicitly in Chapter 9, so this section will provide only a brief introduc-tion. The reader is also referred to a number of reviews of climate modelling, including those of McGuffie and Henderson Sellers (2001), Cane et al. (2006) and the IPCC (2007).
Climate models are derived from weather fore-casting models, originally conceived by John von Neumann who founded the GFDL (Geophysical Fluid Dyamics Laboratory). The first comprehen-sive general circulation experiments were under-taken by Smagorinsky (1963) and by 1965 it was realised that computer models could also be used to explore past climates. There are a range of model types from 1-D energy balance models, to 3-D general circulation models (GCMs). Pioneering work on the application of modelling to palaeocli-mate was carried out by Gates (1976a,b) and Manabe and Hahn (1977). This work brought climate modellers and palaeoclimatologists together, as palaeodata (e.g. CO2 concentrations, sea-surface temperatures (SSTs), ice sheet extents) were needed to set model boundary conditions. Through the 1980s, John Kutzbach and his co-authors led the way in exploring drivers of change in the monsoon using the NCAR CCM (Community Climate Model) (e.g. Kutzbach and Guetter, 1986; Prell and Kutz-bach, 1987; Ruddiman and Kutzbach, 1989). This effort was complemented by significant develop-ments in data-model comparisons through COHMAP with a particular focus on 18 k and 6 k 14C yr BP (COHMAP Members, 1988). This tradi-tion has been continued through the PMIP (Palaeo-climate Modelling Intercomparison Project). PMIP1 used CGMs with atmosphere only, or with slab ocean, while PMIP2 used coupled ocean–atmosphere (–vegetation) models (Braconnot et al., 2007). The results from PMIP are discussed in more detail in Chapter 9.
The development of fully coupled ocean–atmosphere models (see Chapter 9) represented a major challenge due to the very different response times and resolutions of these two key elements of the climate system. Early coupled models such as the UK Met Office’s HadCM2 required flux adjust-
14 Chapter 1
regional details), the monsoon climate is character-ised by a reversal of prevailing wind direction and a contrast between a wet summer and dry winter. This seasonal reversal in wind direction (conven-tionally a change of ≥120° between January and July) is driven by differential heating of oceans and continents. Evaporation and condensation pro-cesses add strength to the system and Coriolis results in the curved trajectories of monsoon winds. Traditionally monsoons were associated with Africa, Asia (India and East Asia) and Australia, being best developed in South and Southeast Asia. More recently monsoon-type systems have also been identified in the tropical Americas, although not fulfilling all the original criteria (McGregor and Nieuwolt, 1998). In this volume, we adopt this wider definition of monsoons.
It is clear that monsoons are an enduring feature of the Earth’s climate system, with monsoon cli-mates recognised in deposits from ancient super continents (e.g. Pangaea) (Clift and Plumb, 2008). It seems likely that the inception of the modern Asian monsoon dates to the construction of Asia, as it now exists, through the collision of the Indian and Asian blocks around 45 to 50 Myr. The eleva-tion of the Tibetan Plateau/Himalayas also appears to be important, and early work on the effects of mountains on monsoons was carried out by Hahn and Manabe (1975). Prell and Kutzbach (1992) linked the modern elevation of the Tibetan Plateau to the strength of the monsoon. The date that Tibet reached its present height is not clear (and may be regionally variable). Around 8 Myr has been sug-gested, but estimates vary between 35 to less than 7 Myr and a high plateau may have existed before 8 Myr. There is evidence for stronger monsoons after 8 Myr from deep sea cores in the Arabian Sea (Kroon et al., 1991) and in Chinese loess sequences which date back to 7–8 Myr. Loess itself is a proxy for the winter monsoon, and the interbedded pal-aeosols for the summer monsoon. Loess–palaeosol sequences may date back to more than 7 Myr (An, 2000) (see Chapter 6, section 6.2), but with a sig-nificant increase in loess accumulation since about 2.7 Myr. Harris (2006) questions whether the shift around 8 Myr is actually due to the monsoon itself or wider oceanographic changes associated with
Ding, 2008). As noted above, in tropical climates it is the distribution of rainfall, rather than tempera-ture, which determines the seasons, and the sea-sonality and overall amount of precipitation that distinguishes the major tropical environments: rainforest, savanna and desert (Bridgman and Oliver, 2006). The reader is referred to Chapter 2 for more on tropical climatology, but in this section some background is given on two key elements of the tropical climate system: the monsoon and ENSO.
Although the dominant dynamic controls on the tropical climate are the location of the Intertropical Convergence Zone (ITCZ) and the subtropical high pressure systems (Hadley cells), perhaps the best known feature of the tropical climate is the monsoon. The name comes from the Arabic word ‘mausim’ for a seasonal reversal of winds recog-nised in the Arabian Sea and Indian Ocean and exploited by Greek and Arab traders. The impor-tance of this seasonal change in winds and the resulting precipitation to trade (sailing ships) and livelihoods (crops etc.) was recognised early on. Failure of the monsoon rains in 1866 and 1871 led to the establishment of the India Meteorological Department in 1875 and the subsequent work of H. F. Blanford and Sir G.T. Walker to forecast and understand monsoon variability. In Walker’s case, his analysis of meteorological data from around the globe led to the recognition of the Southern Oscil-lation (identifying the importance of change in the eastern tropical Pacific) and its link to monsoon rainfall (Walker, 1924). The Southern Oscillation is discussed further below.
Although there are various delineations of monsoon areas – Wang and Ding (2008) suggest that they cover 19.4% of the Earth’s surface – monsoon rain accounts for 30.8% of total global precipitation. Given that these areas are home to more than 55% of the world’s human population (McGregor and Niewolt, 1998) and support the world’s most biologically diverse and ecologically complex terrestrial ecosystems (tropical forests) (Wilson, 1986) their significance is evident and changes in monsoon climates both in the past and into the future are important to understand. As described in Chapter 2 (and Chapters 4 to 8 for
