Gustau Camps-Valls - Deep learning for the Earth Sciences: A Comprehensive Approach to Remote Sensing, Climate Science and Geosciences

1. Chapter 4
2. Chapter 5
3. Chapter 6
4. Chapter 7
5. Chapter 9
6. Chapter 10
7. Chapter 11
8. Chapter 12
9. Chapter 13
10. Chapter 15
11. Chapter 16
12. Chapter 18

1. Chapter 2
2. Chapter 3
3. Chapter 4
4. Chapter 5
5. Chapter 6
6. Chapter 7
7. Chapter 8
8. Chapter 9
9. Chapter 10
10. Chapter 11
11. Chapter 12
12. Chapter 13
13. Chapter 14
14. Chapter 15
15. Chapter 16
16. Chapter 17
17. Chapter 18
18. Chapter 19
19. Chapter 20
20. Chapter 21
21. Chapter 22
22. Chapter 23

Edited by

Gustau CampsValls

Universitat de Valncia, Spain

Devis Tuia

EPFL, Switzerland

Xiao Xiang Zhu

German Aerospace Center and Technical University of Munich, Germany

Markus Reichstein

Max Planck Institute, Germany

This edition first published 2021

2021 John Wiley & Sons Ltd

Chapter 14 2021 John Wiley & Sons Ltd. The contributions to the chapter written by Samantha Adams Crown copyright 2021, Met Office. Reproduced with the permission of the Controller of Her Majesty's Stationery Office. All Other Rights Reserved.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at ;http://www.wiley.com/go/permissions.

The right of Gustau CampsValls, Devis Tuia, Xiao Xiang Zhu, and Markus Reichstein to be identified as the authors of the editorial material in this work has been asserted in accordance with law.

Registered Offices

John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA

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

Editorial Office

The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by printondemand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress CataloginginPublication Data

Name: CampsValls, Gustau, editor.

Title: Deep learning for the earth sciences : a comprehensive approach to

remote sensing, climate science and geosciences / edited by Gustau

CampsValls [and three others].

Description: Hoboken, NJ : Wiley, 2021. | Includes bibliographical

references and index.

Identifiers: LCCN 2021012965 (print) | LCCN 2021012966 (ebook) | ISBN

9781119646143 (cloth) | ISBN 9781119646150 (adobe pdf) | ISBN

9781119646167 (epub)

Subjects: LCSH: Earth sciencesStudy and teaching. | AlgorithmsStudy and

teaching.

Classification: LCC QE26.3 .D44 2021 (print) | LCC QE26.3 (ebook) | DDC

550.71dc23

LC record available at https://lccn.loc.gov/2021012965

LC ebook record available at https://lccn.loc.gov/2021012966

Cover Design: Wiley

Cover Image: iStock.com/monsitj, Emilia Szymanek/Getty Images

To Adrian Albert, in memoriam

Earth science, like many other scientific disciplines, is undergoing a data revolution. In particular, a massive amount of data about Earth and its environment is now continuously being generated by Earth observing satellites as well as physicsbased earth system models running on largescale computational platforms. These informationrich datasets offer huge potential for understanding how the Earth's climate and ecosystem have been changing, and for addressing societal grand challenges relating to food/water/energy security and climate change.

Deep learning, which has already revolutionized many disciplines (e.g., computer vision, natural language processing) holds tremendous promise to revolutionize earth and environmental sciences. In fact, recent years have seen an exponential growth in the use of deep learning in Earth Science, with many amazing results. Deep learning also faces challenges that are unique to earth science data: multimodality; high degree of heterogeneity in space and time; and the fact that earth science data can only provide an incomplete and noisy view of the underlying ecogeophysical processes that are interacting and unfolding at different spatial and temporal scales. Addressing these challenges requires development of entirely new approaches that can effectively incorporate existing earth science knowledge inside the deep learning learning framework. Success in addressing these challenges stands to revolutionize deep learning itself and accelerate discovery across many other scientific domains.

The book does a fantastic job of capturing the state of the art in this fast evolving area. It is logically organized in to 3 coherent parts, each containing chapters written by experts in the field. Each chapter provides an easily to understand introductory material followed by an indepth treatment of the applications of deep learning to specific earth science applications as well as ideas for future research. This book is a must read for the students and researchers alike who would like to harness the data revolution in earth sciences to address pressing societal challenges.

We would like to acknowledge the help of all involved in the collation and review process of the book, without whose support the project could not have been satisfactorily completed. A further special note of thanks goes also to all the staff at Wiley, whose contributions throughout the whole process, from inception of the initial idea to final publication, have been valuable. Special thanks also go to the publishing team at Wiley, who continuously prodded via email, keeping the project on schedule.