Statistical Methods for Trend Detection and Analysis in the Environmental Sciences E-Book

Richard Chandler

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 841

Veröffentlichungsjahr: 2011

Beschreibung

The need to understand and quantify change is fundamental throughout the environmental sciences. This might involve describing past variation, understanding the mechanisms underlying observed changes, making projections of possible future change, or monitoring the effect of intervening in some environmental system. This book provides an overview of modern statistical techniques that may be relevant in problems of this nature. Practitioners studying environmental change will be familiar with many classical statistical procedures for the detection and estimation of trends. However, the ever increasing capacity to collect and process vast amounts of environmental information has led to growing awareness that such procedures are limited in the insights that they can deliver. At the same time, significant developments in statistical methodology have often been widely dispersed in the statistical literature and have therefore received limited exposure in the environmental science community. This book aims to provide a thorough but accessible review of these developments. It is split into two parts: the first provides an introduction to this area and the second part presents a collection of case studies illustrating the practical application of modern statistical approaches to the analysis of trends in real studies. Key Features: * Presents a thorough introduction to the practical application and methodology of trend analysis in environmental science. * Explores non-parametric estimation and testing as well as parametric techniques. * Methods are illustrated using case studies from a variety of environmental application areas. * Looks at trends in all aspects of a process including mean, percentiles and extremes. * Supported by an accompanying website featuring datasets and R code. The book is designed to be accessible to readers with some basic statistical training, but also contains sufficient detail to serve as a reference for practising statisticians. It will therefore be of use to postgraduate students and researchers both in the environmental sciences and in statistics.

Leseprobe

Table of Contents

Statistics in Practice

Title Page

Copyright

Preface

Contributing authors

Part I: Methodology

Chapter 1: Introduction

1.1 What is a Trend?

1.2 Why Analyse Trends?

1.3 Some Simple Examples

1.4 Considerations and Difficulties

1.5 Scope of the Book

1.6 Further Reading

References

Chapter 2: Exploratory Analysis

2.1 Data Visualisation

2.2 Simple Smoothing

2.3 Linear Filters

2.4 Classical Test Procedures

2.5 Concluding Comments

References

Chapter 3: Parametric Modelling—Deterministic Trends

3.1 The Linear Trend

3.2 Multiple Regression Techniques

3.3 Violations of Assumptions

3.4 Nonlinear Trends

3.5 Generalised Linear Models

3.6 Inference with Small Samples

References

Chapter 4: Nonparametric Trend Estimation

4.1 An Introduction to Nonparametric Regression

4.2 Multiple Covariates

4.3 Other Nonparametric Estimation Techniques

4.4 Parametric or Nonparametric?

References

Chapter 5: Stochastic Trends

5.1 Stationary Time Series Models and Their Properties

5.2 Trend Removal via Differencing

5.3 Long Memory Models

5.4 Models for Irregularly Spaced Series

5.5 State Space and Structural Models

5.6 Nonlinear Models

References

Chapter 6: Other Issues

6.1 Multisite Data

6.2 Multivariate Series

6.3 Point Process Data

6.4 Trends in Extremes

6.5 Censored Data

References

Part II: Case Studies

Chapter 7: Additive Models for Sulphur Dioxide Pollution in Europe

7.1 Introduction

7.2 Additive Models with Correlated Errors

7.3 Models for the SO2 Data

7.4 Conclusions

7.5 Acknowledgement

References

Chapter 8: Rainfall Trends in Southwest Western Australia

8.1 Motivation

8.2 The Study Region

8.3 Data Used in the Study

8.4 Modelling Methodology

8.5 Results

8.6 Summary and Conclusions

References

Chapter 9: Estimation of Common Trends for Trophic Index Series

9.1 Introduction

9.2 Data Exploration

9.3 Common Trends and Additive Modelling

9.4 Dynamic Factor Analysis to Estimate Common Trends

9.5 Discussion

9.6 Acknowledgement

References

Chapter 10: A Space–Time Study on Forest Health

10.1 Forest Health: Survey and Data

10.2 Regression Models for Longitudinal Data with Ordinal Responses

10.3 Spatiotemporal Models

10.4 Spatiotemporal Modelling and Analysis of Forest Health Data

10.5 Acknowledgements

References

Index

Statistics in Practice

Statistics in Practice

Series Advisors

Human and Biological Sciences

Stephen Senn

University of Glasgow, UK

Earth and Environmental Sciences

Marian Scott

University of Glasgow, UK

Industry, Commerce and Finance

Wolfgang Jank

University of Maryland, USA

Statistics in Practice is an important international series of texts which provide detailed coverage of statistical concepts, methods and worked case studies in specific fields of investigation and study.

With sound motivation and many worked practical examples, the books show in down-to-earth terms how to select and use an appropriate range of statistical techniques in a particular practical field within each title's special topic area.

The books provide statistical support for professionals and research workers across a range of employment fields and research environments. Subject areas covered include medicine and pharmaceutics; industry, finance and commerce; public services; the earth and environmental sciences, and so on.

The books also provide support to students studying statistical courses applied to the above areas. The demand for graduates to be equipped for the work environment has led to such courses becoming increasingly prevalent at universities and colleges.

It is our aim to present judiciously chosen and well-written workbooks to meet everyday practical needs. Feedback of views from readers will be most valuable to monitor the success of this aim.

A complete list of titles in this series appears at the end of the volume.

This edition first published 2011

2011 John Wiley & Sons, Ltd

Registered office

John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ

United Kingdom

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.

The right of the author to be identified as the author of this work has been asserted in accordance with the 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.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books.

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

Chandler, R. E. (Richard E.)

Statistical methods for trend detection and analysis in the environmental sciences /

Richard Chandler, E. Marian Scott.

p. cm.\emdash (Statistics in practice ; 90)

Includes bibliographical references and index.

ISBN 978-0-470-01543-8 (hardback)

1. Environmental sciences\emdash Statistical methods. I. Scott, E. Marian. II. Title.

GE45.S73C43 2011

577.01′1 — dc22

2010051076

A catalogue record for this book is available from the British Library.

Print ISBN: 978-0-470-01543-8

ePDF ISBN: 978-1-119-99156-4

oBook ISBN: 978-1-119-99157-1

ePub ISBN: 978-1-119-99196-0

Preface

The life story of this book has been long and rather involved. It started with general conversations about the ubiquitous nature of the questions ‘What is changing and by how much?’ within environmental sciences, moved on to discussions about how some of the newer, potentially more powerful and almost certainly more realistic statistical approaches to answering such questions had not yet been translated over to the applied sciences communities and then ultimately reached the ‘Wouldn't a book be a good idea?’ stage. Many of those initial conversations were held around meetings organised by the Environmental Statistics section of the Royal Statistical Society. Momentum was maintained subsequently through a series of training courses for environmental science PhD students, loosely titled ‘Quantifying the environment’ and sponsored by the Natural Environment Research Council. As always, having formulated our grand and ambitious plans, the task of actually putting them into practice took much longer than we either hoped or expected. So, now that we have finally reached our objective, we owe a huge debt of thanks to all the staff at Wiley for their patience over more years than we care to admit.

Many people have contributed to the discussions above. Others have helped by responding constructively to our unsolicited requests for preprints or queries regarding technical details of their research. With apologies that we cannot mention all of these people by name, we take the opportunity here to thank all of them for their time and input. In particular, some of the work described herein has developed from past PhD projects co-supervised with Adrian Bowman at Glasgow, notably those of Andrew McMullan, Marco Giannitrapani and especially Claire Ferguson: the review chapters in Claire's PhD thesis formed the nugget for some of the sections in Chapter 4. We are also grateful to Rebecca Wilson at the University of Leicester, to Job Verkaik at KNMI and to Gavin Simpson and Don Monteith at UCL, for providing the ozone, wind speed and alkalinity data used in Part I of the book and for dealing with our queries with patience and good grace.

The book is split into two parts: the first a textbook-style introduction to the area and the second a collection of extended case studies demonstrating the practical application of modern statistical approaches to the analysis of trends in real environmental studies. These case studies have been chosen to span different environmental application areas as well as to highlight different methodologies, statistical issues and styles of analysis. To present this range and depth of material would not have been possible without the help of our many collaborators and contributors, who have been most helpful and forgiving throughout the long gestation of the book. To them we offer our warmest thanks and appreciation. We hope that the final result of this collaborative effort will appeal both to environmental scientists and to statisticians, and perhaps in a small way will encourage further interaction and engagement between the two communities.

This book includes an accompanying website. Please visit www.wiley.com/go/trend_detection for more information.

Marian Scott and Richard Chandler

Glasgow and London

Contributing authors

Bryson C. Bates

Climate Adaptation Flagship

CSIRO Marine and Atmospheric

Research, Wembley, Western Australia

Australia

Adrian Bowman

School of Mathematics and Statistics

University of Glasgow

Glasgow G12 8QW

UK

Richard E. Chandler

Department of Statistical Science

UCL, Gower Street

London WC1E 6BT

UK

Stephen P. Charles

Climate Adaptation Flagship

CSIRO Land and Water

Wembley, Western Australia

Australia

Ludwig Fahrmeir

Department of Statistics

Ludwig-Maximilians-University Munich

Ludwigstraße 33

D-80539 München

Germany

Carla Rita Ferrari

ARPA Emilia-Romagna

Struttura Oceanografica Daphne, V le

Vespucci 2, 47042 Cesenatico (FC)

Italy

Marco Giannitrapani

Formerly of the School of

Mathematics and Statistics,

University of Glasgow.

Now at Novartis Farmaceutica S.A.,

E-08013 Barcelona, Spain

Elena N. Ieno

Highland Statistics

Suite N 226, Av Finlandia 21

CC Gran Alacant Local 9

03130 Santa Pola

Spain

Thomas Kneib

Department of Mathematics

Carl von Ossietzky University Oldenburg

D-26111 Oldenburg

Germany

Cristina Mazziotti

ARPA Emilia-Romagna

Struttura Oceanografica Daphne, V le

Vespucci 2, 47042 Cesenatico (FC)

Italy

Giuseppe Montanari

ARPA Emilia-Romagna

Struttura Oceanografica Daphne, V le

Vespucci 2, 47042 Cesenatico (FC)

Italy

Attilio Rinaldi

ARPA Emilia-Romagna

Struttura Oceanografica Daphne, V le

Vespucci 2, 47042 Cesenatico (FC)

Italy

E. Marian Scott

School of Mathematics and Statistics

University of Glasgow

Glasgow G12 8QW

UK

Ron Smith

CEH Edinburgh

Bush Estate, Penicuik

Midlothian EH26 0QB,

UK

Alain F. Zuur

Part I

Methodology