Landslide Hazards, Risks, and Disasters

Landslide Hazards, Risks and Disasters Second Edition makes a broad but detailed examination of major aspects of mass movements and their consequences, and provides knowledge to form the basis for more complete and accurate monitoring, prediction, preparedness and reduction of the impacts of landslides on society. The frequency and intensity of landslide hazards and disasters has consistently increased over the past century, and this trend will continue as society increasingly utilises steep landscapes. Landslides and related phenomena can be triggered by other hazard and disaster processes – such as earthquakes, tsunamis, volcanic eruptions and wildfires – and they can also cause other hazards and disasters, making them a complex multi-disciplinary challenge.

This new edition of Landslide Hazards, Risks and Disasters is updated and includes new chapters, covering additional topics including rockfalls, landslide interactions and impacts and geomorphic perspectives. Knowledge, understanding and the ability to model landslide processes are becoming increasingly important challenges for society extends its occupation of increasingly hilly and mountainous terrain, making this book a key resource for educators, researchers and disaster managers in geophysics, geology and environmental science.

• Provides an interdisciplinary perspective on the geological, seismological, physical, environmental and social impacts of landslides
• Presents the latest research on causality, impacts and landslide preparedness and mitigation. Includes numerous tables, maps, diagrams, illustrations, photographs and video captures of hazardous processes
• Discusses steps for planning for and responding to landslide hazards, risks and disasters

• Language: English
• Publisher: Elsevier Science
• Release date: Oct 17, 2021
• ISBN: 9780128226452

Book preview

Landslide Hazards, Risks, and Disasters

Second Edition

Editor

Tim Davies

School of Earth and Environment, University of Canterbury, Christchurch, New Zealand

Editor

Nick Rosser

Department of Geography, Durham University, Durham, United Kingdom

Editor

J.F. Shroder

Emeritus Professor of Geography and Geology, Department of Geography and Geology, University of Nebraska at Omaha, Omaha

Table of Contents

Cover image

Title page

Copyright

Contributors

Editorial foreword to the second edition

Chapter 1. Landslide hazards, risks and disasters: introduction

1.1. Introduction

1.2. Understanding landslide hazards

1.3. Understanding landslide risks

1.4. Understanding future landslide disasters

1.5. Conclusion

Chapter 2. Landslide causes and triggers

2.1. Introduction

2.2. Concept of instability

2.3. Stability factors

2.4. Summary and conclusion

Chapter 3. Landslides in bedrock

3.1. Introduction

3.2. Rock materials

3.3. Mass movement characteristics

3.4. Mass movement types

3.5. Case studies

3.6. Bedrock landslide recognition and management

3.7. Risk management of rock slopes

3.8. Summary

Chapter 4. Coseismic landslides

4.1. Seismically triggered landslides

4.2. Mechanics of earthquake-induced landslides

4.3. Stability analysis and hazard assessment

4.4. Limitations of current understanding

Chapter 5. Volcanic debris avalanches

5.1. Introduction

5.2. Volcanic debris avalanches

5.3. Types of volcanic landslides

5.4. Deep-seated volcanic landslide deformation: priming and triggers

5.5. Deep-seated volcano gravitational deformation

5.6. Regional tectonic influences

5.7. Priming of volcanic landslides

5.8. Triggering volcanic landslides

5.9. The structure of volcanic landslides

5.10. Volcanic landslide deposits

5.11. Debris avalanche textures and structures

5.12. Secondary hazards of volcanic landslides

5.13. Volcanic landslide transport mechanisms

5.14. Hazards from volcanic landslides

5.15. Summary

Chapter 6. Peat landslides

6.1. Introduction and background

6.2. The nature of peat, its structure and material properties

6.3. Morphology and classification of peat landslides

6.4. Relationship between landslide type and peat stratigraphy

6.5. Impacts of peat landslides

6.6. The runout of peat landslides

6.7. Slope stability analysis of peat landslides and geotechnical properties

6.8. Historical perspective on the frequency of peat landslides

6.9. The future incidence of peat landslides

6.10. Conclusion

Chapter 7. Rock–snow–ice avalanches

7.1. Introduction

7.2. Rapid mass movements on glaciers

7.3. RSI avalanche propagation

7.4. Implications for hazard assessment

7.5. Conclusions

Chapter 8. Multiple landslide-damming episodes

8.1. Introduction

8.2. Previous work on landslide dams

8.3. Landslide-dam episodes: lessons from case studies

8.4. Discussion

8.5. Conclusions

Chapter 9. Rock avalanches onto glaciers

9.1. Introduction

9.2. Processes

9.3. Consequences

9.4. Case studies

9.5. Concluding remarks

Chapter 10. Paleo-landslides

10.1. Introduction

10.2. Significance of paleo-landslides

10.3. Recognition and mapping

10.4. Dating paleo-landslides

10.5. Temporal bias

10.6. Role in landscape evolution

10.7. Risk assessment

10.8. Conclusion

Chapter 11. Remote sensing of landslide motion with emphasis on satellite multi-temporal interferometry applications: an overview

11.1. Introduction

11.2. Brief introduction to DInSAR and Multi-Temporal Interferometry

11.3. Examples of different scale MTI applications to landslide motion detection and monitoring

11.4. Summary discussion

Chapter 12. Small landslides – frequent, costly and manageable

12.1. Introduction

12.2. Costs of small–medium landslides

12.3. Frequency of landslides

12.4. Management of landslides

12.5. Size of manageable landslides

12.6. Conclusions

Chapter 13. Analysis tools for mass movement assessment

13.1. Introduction

13.2. The computational tools available

13.3. Limit equilibrium methods

13.4. Limit analysis

13.5. Continuum numerical methods

13.6. Distinct element method

13.7. Conclusions

Chapter 14. Landslides in a changing climate

14.1. Introduction

14.2. Rockfalls, rockslides and rock avalanches

14.3. Shallow landslides and debris flows

14.4. Deep-seated landslides in soil

14.5. Coastal landslides

14.6. Landslides in the cryosphere

14.7. Regional scale landslide response

14.8. Landslide risk and economic considerations

14.9. Adaptation and mitigation

14.10. Summary

14.11. Discussion and recommendations

14.12. Concluding remarks

Chapter 15. Rockfall hazard and risk

15.1. Background

15.2. Definitions

15.3. Case study 1: assessing rockfall hazard, North Yorkshire coast, UK

15.4. Vulnerability to rockfall

15.5. Case study 2: Port Hills, Christchurch, NZ

15.6. Summary and conclusions

Chapter 16. Reducing landslide disaster impacts

16.1. Introduction

16.2. Disaster risk reduction: terminology and implications

16.3. Fundamental weakness of DRR

16.4. Disaster impacts

16.5. Landslide disaster impact reduction

16.6. Reducing the impacts of the next landslide disaster: a scenario approach

16.7. Discussion

16.8. Conclusions

Chapter 17. Geomorphic precursors of large landslides: seismic preconditioning and slope-top benches

17.1. Introduction

17.2. Slope-top benches

17.3. Mountain edifice response to coseismic shaking

17.4. Field evidence

17.5. Example of possible hazard: slope overlooking Franz Josef glacier township

17.6. Discussion

17.7. Conclusions

Index

Copyright

Elsevier

Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom

50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States

Copyright © 2022 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing-in-Publication Data

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

ISBN: 978-0-12-818464-6

For information on all Elsevier publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Candice Janco

Acquisitions Editor: Amy Shapiro

Editorial Project Manager: Emerald Li

Production Project Manager: Sruthi Satheesh

Cover Designer: Christian Bilbow

Typeset by TNQ Technologies

Contributors

F. Bovenga,     National Research Council, CNR-IREA, Bari, Italy

E.T. Bowman,     Department of Civil and Structural Engineering, University of Sheffield, Sheffield, United Kingdom

Marc-André Brideau

BGC Engineering Inc., Victoria, BC, Canada

Simon Fraser University, Burnaby, BC, Canada

John J. Clague,     Centre for Natural Hazard Research, Simon Fraser University Burnaby, Burnaby, BC, Canada

Giovanni B. Crosta,     Università degli Studidi Milano-Bicocca, Milan, Italy

Tim Davies,     School of Earth and Environment, University of Canterbury, Christchurch, New Zealand

Audray Delcamp,     Department of Geography, Faculty of Sciences, Vrije Universiteit Brussel, Brussel, Belgium

P. Deline,     EDYTEM Lab, Université Savoie Mont Blanc, CNRS, Chambéry, France

K. Hewitt,     Department of Geography and Environmental Studies, Wilfrid Laurier University, Waterloo, ON, Canada

Matthias Jakob,     BGC Engineering Inc, Vancouver, BC, Canada

Oliver Korup

Institute of Environmental Science and Geography, University of Potsdam, Potsdam, Germany

Institute of Geosciences, University of Potsdam, Potsdam, Germany

Chris Massey,     GNS Science, Lower Hutt, New Zealand

Samuel T. McColl,     Geosciences Group, School of Agriculture and Environment, Massey University, Palmerston North, New Zealand

Danilo Moretti,     WSP, Wellington, New Zealand

Bill Murphy,     Leeds University, Leeds, United Kingdom

N. Reznichenko,     University of Canterbury, Christchurch, New Zealand

Nicholas J. Roberts

Simon Fraser University, Burnaby, BC, Canada

Mineral Resources Tasmania, Hobart, TAS, Australia

Nick Rosser,     Department of Geography, Durham University, Durham, United Kingdom

D. Shugar,     waterSHED Lab, Department of Geoscience, University of Calgary, AB, Canada

Rosanna Sosio,     Università degli Studi di Milano-Bicocca, Milan, Italy

Stefano Utili,     School of Engineering, Newcastle University, Newcastle upon Tyne, United Kingdom

Benjamin van Wyk de Vries,     Laboratoire Magmas et Volcans, Univeristé Blaise Pascal, Clermont-Ferrand, France

Gonghui Wang,     Disaster Prevention Research Institute, Kyoto University, Kyoto, Japan

Jeff Warburton,     Durham University, Durham, United Kingdom

J. Wasowski,     National Research Council, CNR-IRPI, Bari, Italy

Editorial foreword to the second edition

General context of hazards, risks and disasters: In the production of an editorial foreword, as editor in chief, I generally have prior access to all the volume chapters and new ideas that are going to be presented that I can edit. This is commonly done so that any editor in chief can exert quality control, as well as alert readers to issues at hand that will have the most relevance to understanding the intents of the book series, the volume editors and their chapter authors. In this case, however, the sheer volume of chapter offerings in companion volumes, coupled with the exigencies produced by the COVID pandemic, curtailed my ability to edit in a timely fashion. Also, I did not have access to the new material, other than a table of contents (TOC). The result was that I had to use a writing form similar to the first edition, first with text on general hazards, risks and disasters, followed by more specificity on landslides. In addition, in my editorial foreword for the first edition, I had written on the first page a generalized accounting of the exponential population growth that puts so many more people in harm's way, along with the resulting climate change that exacerbates so many hazard processes plaguing humanity. I also noted that many media landscapes were saturated, in the northern hemisphere, at least, with sensational catastrophe stories and apocalyptic, ‘end-of-days’ prophets of doom and Armageddon that seem ever more present. Some existentialist discussion of human extinction as a result of overwhelming hazards and catastrophes was noted as well. In my first draft of this particular editorial foreword to the second edition, however, I originally began it with more lengthy discussion of these existentialist issues and apocalyptic extinction scenarios, but perhaps without enough emphasis on the landslide issues at hand. After all, landslides may be catastrophic locally, but will hardly threaten the planet universally. My initial approach thus was perhaps justifiably criticized by volume editors Davies and Rosser, so this new version is my attempt to better explain the situation about worldwide hazards, risks and disasters, but not to advocate any particular political point of view.

Predicting the future of anything, much less hazards and risks that might impact certain areas, is only barely possible under certain idealized conditions. Prescient thinkers such as Harari (2016) have told us repeatedly, however, about just such possible alternative futures. Utopian futures can seem possible, given certain technical advances that continue to be made, even whilst species extinction also looms more likely. Thus, threats of environmental disasters lurk as the species-terminating horsemen of famine, plague and war seem to ride on the apocalyptic horizon. In fact, scientific luminary David Attenborough (2018) noted in a speech kicking off the Climate Change Conference in Katowice, Poland (Domonoske, 2018), that ‘The collapse of our civilizations and the extinction of much of the natural world is on the horizon’.

Several ramifications of the human extinction possibilities exist that are perhaps worth further explanation in any conversation about the hazards, risks and disasters that confront the human race (Newitz, 2013). For example, one prominent answer to the Fermi Paradox about the apparent absence of intelligent alien life forms in the universe is that any intelligent life that comes into existence in the universe could be quite short lived. This might be inherently because such intelligence may be no measure of species success and instead may lead to self-induced destruction of civilization and extinction. Human extinction scenarios even include the voluntary human-extinction movement (VHEMT) by reproduction abstinence to avoid further environmental degradation. More likely scenarios have it that the threats of over-population, environmental destruction, artificial intelligence, CRISPER genetic manipulation, religious desires for ‘End Times’ and a myriad of increasingly observable other negative factors will lead inevitably to the disaster of human extinction. This is because we simply cannot continue to live as we have been doing for the past hundreds of thousands of years since the emergence of Homo sapiens on the planet.

In recognition of the great seriousness of this problem of extinction of so many organisms, including in the final phases possibly humans as well, civil disobedience has erupted in Europe and North America as the Extinction Rebellion, whose website is , and whose symbol is an eye-catching blocked X inside a circle.

The Extinction Rebellion, or XR, maintains that governments must tell the truth about the climate, and the wider ecological emergency, as well as reverse inconsistent policies and work alongside the media to communicate with citizens. In addition, XR feels that governments must enact legally binding policy measures to reduce carbon emissions to net zero by 2025, and to reduce consumption levels. To XR, it is unconscionable that our children and grandchildren should have to bear the brunt of an unprecedented disaster of our own making. Accordingly, in recognition that governments have abrogated their responsibility to protect their citizens from harm, XR feels that it is their right and moral duty to bypass governments and rebel to defend life itself from the looming catastrophe. They call for a Citizens' Assembly to work with scientists based on extant evidence and in accordance with the precautionary principle, to urgently develop a credible plan for rapid total decarbonization of the economy.

Strongly in contrast, however, many good scientists find all this apocalypse blather to be quite unhelpful (Schwartz, 2018a,b), given the well-known propensity of humanity to think and act its way out of its many prior dilemmas. Even so, the current political wrangling over climate change and possible environmental disasters can be highly aggravating and worrisome to many people (Kaplan and Fritz, 2018). For example, part of the reason many people deny human-caused climate change or the manifold disasters resulting from it is not because they are completely ignorant of climate science, but because they are on the political right (Weintraub, 2018). On the other hand, people who may accept climate change also may not understand what is causing it. In any case, Mora et al. (2018) have presented a compelling case that changing climate because of the ongoing emission of greenhouse gases (GHGs) is triggering changes to many climate-affected hazards that can impact humanity. Her research group found evidence for 467 pathways by which health, water, food, economy, infrastructure and security have been recently affected. This includes heat waves, global warming, precipitation, floods, fires, storms, drought, sea-level rise, ocean chemistry and land cover. In fact, by the next century, the world's population will be exposed concurrently to the largest magnitude in only one of these hazards if emissions are strongly reduced, or three such hazards if they are not. Some tropical coastal areas would face up to six simultaneous or cascading hazards. These findings highlight the fact that GHG emissions are seen to pose a broad threat to humanity by intensifying the multiple hazards to which humanity is increasingly vulnerable.

Furthermore, in recognition of how much climate change is being seen in action all around us now (heat waves, firestorms, massive floods, megadroughts, etc.), this book series on hazards, risks and disasters is in the process of undergoing retooling and editorship to newly emphasize direct climate change causation of these phenomena.

Landslide hazards, risks and disasters: Thus, in these days of pandemic, the other profuse potential hazards, risks and disasters that loom upon us repeatedly do tend to sometimes overwhelm our attention. However, until we become physically overwhelmed by the existential threats noted above, there remains a need to focus on less cataclysmic events. As examples of this in the landslide or mass movement arena, therefore I mention: (1) increased coastal landslides and subsidence due in part to rising sea levels as land ice melts and coastal sediments consolidate; (2) collapsing mountain peaks because of loss of buttressing support from downwasting glaciers, as well as thawing permafrost binders in rock walls; (3) increased precipitation on mountain slopes leading to more pervasive debris avalanches and (4) increasing aridity and desertification leading to ground subsidence and large crack propagation; all these mass movement processes add into the many other hazards that cause so much human trauma and possible resulting mass migration to escape such ills.

The arrival of this second edition volume on landslides offers plentiful understandings and some possible mitigations to problems with slope failures. Tim Davies and Nick Rosser have brought together a reasonably comprehensive view of landslides with the 13 chapters of the first edition either lightly or heavily updated, together with 4 new chapters that are important new additions to knowledge and dealing more effectively with landslide hazards. The new materials on rockfall hazard and risk (N. Rosser and C. Massey); landslides in a changing climate (M. Jakob); reducing landslide disaster impacts (T. Davies); geomorphic precursor of landslides: seismic preconditioning and slope-top top benches (T. Davies and D. Moretti) are especially welcome. The lengthy (77 p.) and comprehensive view of landslides in a changing climate by Jakob is replete with case studies and other highly informative material that could serve as a foundation for a new volume itself on climate change and mass movement.

In conclusion and putting possible misanthropic extinction speculations aside in favour of the more optimistic possibilities of protecting peoples and societies against hazards, risks and disasters from mass movement, this updated volume is a small but useful contribution. Our benign interests to help against outrageous fortune encompass learning as much as we can about both surficial as well as deep Earth processes that are so problematical for humanity. Only by focussing attention on some of the recurring problems with unstable slopes and paying attention to possibilities of incipient failure can we hope to avert localized disasters.

John (Jack) Shroder

Editor in Chief

September 29, 2021

References

1. Attenborough D. For the record. Time. 17 December 2018:4.

2. Domonoske C. David Attenborough Warns of Collapse of Civilizations, at U.N. Climate Meeting. 2018. https://www.npr.org/2018/12/03/672893695/david-attenborough-warns-of-collapse-of-civilizations-at-u-n-climate-meeting.

90. Harari Y.N. Homo Deus: A Brief History of Tomorrow. Random House; 2016:462.

3. Kaplan S, Fritz A. Study: Climate Change Increased Likelihood of Disasters. The Washington Post,. 11 December 2018:A3. .

4. Mora C, Spirandelli D, Franklin E.C, Lynham J, Kantar M.B, Miles W, Smith C.Z, Freel K, Moy J, Louis L.V, Barba E.W, Bettinger K, Frazier A.G, Colburn J.F, Hanasaki N, Hawkins E, Hirabayashi Y, Knorr W, Little C.M, Emanuel K, Sheffield J, Partz J, Hunter C.L.Broad threat to humanity from cumulative climate hazards intensified by greenhouse gas emissions. Nat. Clim. Change. 19 November 2018.

5. Newitz A. Scatter, Adapt, and Remember: How Humans Will Survive a Mass Extinction. NY: Doubleday; 2013:305.

6. Schwartz J. 11. Will We Survive Climate Change? New York Times; 2018 20 November 2018.

7. Schwartz J. ‘Like a Terror Movie’: How Climate Change Will Multiply Disasters. New York Times; 2018 20 November 2018.

8. Weintraub K. Steven Pinker Thinks the Future is Bright. New York Times; 20 November 2018.

Chapter 1: Landslide hazards, risks and disasters

introduction

Tim Davies¹, and Nick Rosser²     ¹School of Earth and Environment, University of Canterbury, Christchurch, New Zealand     ²Department of Geography, Durham University, Durham, United Kingdom

Abstract

Here, we provide a brief contextual introduction to landslides insofar as they are hazards that threaten society, as they are components of the risks that they pose to society and as they result in disasters that affect society.

Keywords

Disaster; Hazard; Landslides; Risk; Vulnerability

1.1. Introduction

Landslides are a ubiquitous phenomenon on any planet that, like Earth, is tectonically active. However, society is generally inclined to view landslides as exceptional events that occur very infrequently, and usually elsewhere, and their inevitable impacts on society worldwide and over extra-human timescales have hitherto been considered rarely, if at all, in societal planning. In the last decades, however, global landslide occurrence and impact have been documented (e.g., Kirschbaum et al., 2015; Froude and Petley, 2018), and the seriousness of this hazard underlined, particularly in regions that suffer recurrent impacts year on year. Interestingly, climate change is shown to have much less influence on the number of landslide fatalities than population growth and the associated disturbance to the landscape (Petley, 2010), and if population growth continues, landslide fatalities will continue to increase. This increasing focus on landsliding as an important hazard is emphasised at the time of writing by, for example, the signature into US law of the National Landslide Preparedness Act (https://blogs.agu.org/landslideblog/, 15 January 2021) which establishes a National Landslide Hazards Reduction Program in the USGS, with the aim of improving identification and understanding of landslide risks; of protecting communities; of saving lives and reducing property losses and of improving emergency preparedness.

There is thus good reason to examine the role of landslides as threats to society, and to seek for avenues, whereby this threat can be reduced in the future; this is the purpose of this volume.

1.2. Understanding landslide hazards

Landslide hazards are, in essence, landslides which have the potential to affect society detrimentally. One may debate whether or not all landslides constitute hazards to society, but in principle, any landslide anywhere is a potential hazard if there is any possibility that humankind is now or will at some time in the future make itself vulnerable to the effects of that landslide, thus generating potential impact and hence risk. There are very few if any places on Earth where this possibility is zero; thus, to a fair approximation, all terrestrial landslides can be considered to be hazards. So, on that basis are landslides on the Moon or Mars, but we have to draw a line somewhere. Landslide disasters can be extensive, such as the landsliding triggered by large continental earthquakes, and landslides individually can be disastrous at more local scales, destroying just single houses or land holdings upon which livelihoods are based.

Landslides are a crucial component of Earth's geological cycle, in which tectonic plate motion and volcanism cause parts of the crust to be continuously uplifted above a base level; these raised areas are then continually eroded down by gravity and gravity-driven water, and ice flows towards the base level. Landslides represent the directly gravity-driven component of erosion, and they occur in sizes ranging from individual rocks falling (∼< m³) to whole mountains collapsing (∼>> km³). There is increasing evidence, by way of magnitude–frequency data, that larger landslides deliver more sediment to river systems over time than do smaller ones, so that large, infrequent events appear to dominate the sediment supply spectrum (Korup and Clague, 2009) that shapes much terrestrial geomorphology. Further, since the majority of river-transported sediment originates in slope failures, this emphasises the significance of landsliding in geomorphology – including fluvial geomorphology – especially in and adjacent to active orogens. Even a mountain range such as the Southern Alps of New Zealand, which was heavily glaciated prior to 18 ka, today shows relatively little evidence of any erosion process other than mass movement (Fig. 1.1). Thus, in steep, tectonically active terrain, the progress of geology requires that landslides will continue to occur on hillslopes in the future, much as they have in the past; and the increasing presence of people and their assets on, in the vicinity of and downstream of these hillslopes means that landslide-initiated disasters will inevitably occur – and to an increasing extent – in the future.

An interesting fact discussed at some length by Korup and Clague (2009) is the fairly consistent variation of probability of occurrence for landslides of different sizes. These, irrespective of type and trigger, appear to follow a common distribution for larger events, suggesting that there is some factor constraining the frequency of occurrence of landslides of various sizes. This obviously holds relevance for assessing hazards and risks from landslides. Recently, research into complex systems has shown that distributions of this type are very common for such systems in many contexts (geomorphic, societal, financial, biological, ecological, etc.) and has in addition identified that the very largest events can depart significantly from this distribution. These largest mega-events, known as ‘dragon-king’ events (Sornette, 2009), occur much more frequently than the distribution would suggest, and may reflect the fact that these events occupy a large proportion of the space available to them. Thus, such an event has as its environment the system boundaries, which is not the case for smaller events whose environment usually does not reach these boundaries.

Figure 1.1  View eastwards towards the main divide of the Southern Alps, New Zealand, showing landscape dominated by mass movement erosion (background) and fluvial reworking of mass movement debris (foreground). 

Photo by Tim Davies.

For landslides, this is equivalent to the probability distribution of landslides from a given hillslope being limited because the physical extent of the hillslope limits the volume of the largest landslides that can occur. The tendency for still larger events to occur is constrained by the system boundaries (e.g., available topography and rock strength), so that these hypothetical larger events are in fact manifested as smaller (but still very large) events, which thus acquire correspondingly higher (but still small) frequencies. Hence, the largest events occur more frequently than the probability function for smaller events would suggest. This higher than expected frequency of the largest events is clearly a concern in anticipating future landslide disasters, as these events can give rise to the largest impacts; and, although they occur very rarely in a given place, they will inevitably occur and can occur at any time, so treating them as a low priority is not sensible. This is illustrated by the occurrence of any low-frequency event; the day before it occurred, its occurrence the next day had an extremely low probability, so prioritising preparations for it would appear irrational, but in the event would be extremely useful. At the other end of the landslide magnitude scale, the smallest landslides – even single rockfall blocks – can lead to significant harm or death and pose a day-to-day risk to the lives and livelihoods of people in many mountainous regions of the world (see Bowman, 2021). In recent years, remote sensing has been able to capture increasingly granular data providing a more complete description of the magnitude, frequency and timing of these smallest events (see Rosser and Massey, 2021). These data are starting to allow us to consider controls on localised landslide hazard and how it varies through space and time, enabling more fidelity of an otherwise often ubiquitous view of background landslide hazard in the highest risk regions.

A potentially confusing factor in identifying and quantifying landslide hazards is the geomorphic similarity between remnants of landslides and other geomorphological processes, notably moraines. In recent years, many deposits previously classified as moraines (some of which have indeed been used as paleoclimatic indicators) have been reinterpreted as deposits of large landslides (e.g., Hewitt, 1999; McColl and Davies, 2011; Reznichenko et al., 2012, 2016). Obviously, the hazard implications of such reinterpretation are significant; moraines indicate zero likelihood of repeat events until glaciers advance again, and even then the threat is minor due to the slow speed of glacial advance, but landslide deposits indicate the potential for further landslides at any time in the future, with potentially much more serious consequences.

Predicting the locations of future landslides is obviously difficult, but modern technologies allow at least the relative susceptibilities of different locations to landsliding to be estimated. For example, Kritikos et al. (2015) analysed the spatial distribution of landslides caused by the Northridge and Wenchuan earthquakes with respect to ground shaking, topography, distance from active faults, slope gradient and slope position, and were able, on this basis, to explain with >90% accuracy the relative spatial distribution of landslides resulting from the Chichi earthquake in Taiwan. This explanation used no data from Taiwan landslides, so the technique is applicable to regions with no previous landslide data or where timescales warrant a rapid estimation of landslide impact; similar analyses for rainfall-generated landslides should be feasible. Whilst providing only relative information, this technique nevertheless is useful for scenario-based planning using relative vulnerabilities of, for example, highways and utilities, so that preparation for future landslide events can take place. Applying this technique, Robinson et al. (2016) have derived landslide distributions for two major earthquakes in New Zealand, one of which has occurred (Kaikoura earthquake 2016) and one of which has not (Alpine fault). These demonstrate the value of such predictions for both emergency response to an event and for long-term resilience and response planning.

1.3. Understanding landslide risks

Risk is defined as the product of (hazard) probability and (event) consequence and so introduces as a variable the societal values potentially affected by a landslide (hazard). The consequence itself comprises products of exposure and vulnerability (fragility), so is conceptually more complex than the hazard. Nevertheless, some of its components are more readily quantifiable than many landslide characteristics – for example, the value of assets in a given area and the number of people in the area, at a given time, but these are of course also time variable at a wide range of scales into the future. In addition, the cost of a landslide disaster includes a wide range of consequential costs, such as the effect on commerce of the cessation of some activities (including transport and power supply), perhaps for a long time; the effect on commerce nationally, if lifelines and infrastructure have been badly affected and the disruption of community life and social activity, which can affect commerce even if other inputs are not affected. Landslide disasters also have intangible costs which are evidently difficult to quantify as risks, but which are nevertheless significant to society because they affect the way society functions – deaths, emotional damage, loss of quality of life and environmental damage are examples.

Probabilistic analysis of landslide risk is the commonly accepted basis for making rational decisions about choosing what measures to take to reduce or manage disaster risks. Davies (2021) examines this acceptance; a fundamental difficulty with this procedure is that in any reasonable future planning time frame (i.e., a time frame that present inhabitants perceive as relevant to their lives and livelihoods), and the number of landslide events that will occur in a given place will be very small. Thus, it is extremely unlikely that the events that actually occur in this time frame of lived experience will follow the probability distribution used, and this would be the case even if this distribution were perfectly known. Thus, the apparent precision derived from probabilistic risk analysis is to a large extent spurious when considered at the scales relevant to people. Davies (2021) presents a framework that overcomes these objections by exploring scenario-based pre-event adaptation to the next disaster that will affect a given community to develop strategies to reduce its impacts on the societal functioning of the community.

The only landslide risk management strategy with any realistic chance of being effective against major landslide events is that of avoidance. Substantial landslides – especially earthquake-generated ones – often occur without warning in apparently stable terrain and are uncontrollable once initiated, so the well-developed mitigation and response (evacuation) procedures used for other hazards such as rainstorm-induced floods are not effective. It is clear that our emerging understanding of patterns in, for example, coseismic landslides can provide first-order guidance to reduce exposure before an event occurs (e.g., Milledge et al., 2019). However, such efforts must inevitably be considered in the context of competing risks faced by a society when planning, which may mean that the threat posed by landslides is often secondary to more pressing day-to-day needs. If a potential landslide has been identified, monitoring of precursory motion may make evacuation a feasible option, but this may be inhibited by political and other considerations, as in the case of the Vajont tragedy (e.g., Davies, 2013). For small and frequent events, by contrast, prevention is more feasible (Bowman, 2021) and, if such events occur frequently in the same locality, a probabilistic risk analysis may be a realistic and useful way of designing a mitigation strategy.

Despite much progress, a significant distance persists between progress made in research into landslide hazards, risks and disasters, including much presented in this volume, and the ongoing challenges faced by residents of those areas of the world where landslide impacts are recurrently felt most acutely. Nepal, for example, harbours ca. 10% of rainfall-triggered fatal landslides each year (Froude and Petley, 2018), yet has ca. 0.4% of global population, and despite considerable international investment in disaster risk reduction, landslide impacts in rural Nepal continue to grow. The scale of annual landslide impacts is also stark. For example, in the 2020 monsoon in Nepal, there were nearly 500 recorded landslide events, resulting in almost 300 fatalities (bipadportal.gov.np; Rosser et al., 2021), numbers that exceed those from many of the widely studied historical landslide disasters in high-income countries which each have dramatically influenced our recognition of, and research into, landslide hazard and risk.

Whilst innovations in monitoring, modelling and engineering have made considerable inroads into landslide risk reduction in higher income countries, the same cannot therefore be said for lower and middle-income countries, where the resource and capacity to tackle landslide risk is commonly lacking. For example, basics, such as geological input into rural road alignment and ongoing provision for slope and drainage maintenance, mean that rural infrastructure constructed with the intention of enhancing livelihoods can scar the landscape leading to chronic landslide hazards for many years to come. Whilst there are some notable exceptions to this, there is a clear need to consider how the nature of acute landslide risk, in, for example, the mountainous regions of Asia, can be addressed. It is notable that the direct transfer of hazard and risk knowledge into practise is immensely challenging in respect to costs, sustainability, cultures and the very nature of the landslide hazard faced, a challenge that sits within the widely recognised problems of ‘development’. Whilst technology and ‘cutting edge’ research offers opportunities to make rapid progress, there is also clear value in and a need for investing in good quality baseline data and capacity, without which even the most beneficial landslide risk reduction measures, such as mapping to inform land use planning to help people avoid potentially high risk areas, are challenging. It is also clear that tackling these issues requires an interdisciplinary perspective that considers the root causes of landslide disaster risk: from politics and power to wider ongoing social and political change. Most importantly, the starting point should always be the needs of communities living with landslide risk and those tasked with supporting risk reduction.

1.4. Understanding future landslide disasters

Most of the major landslide disasters that affect society are unexpected. This statement is in fact not confined to landslides, but surprise is a particularly consistent feature of landslide disasters. It follows that landslide disaster impacts could probably be reduced substantially if they could be better expected, but how can that be achieved? Again, perhaps relying less on probabilistic information would reduce the natural tendency to under-prioritise the inevitable occurrence of major but low-probability events at some time in the future. Identification of precursory behaviour to both large and small landslides would be of great value, and in some places, this is being done very effectively (e.g., the Åknes landslide, Norway: Kveldsvik et al., 2009, where a major landslide seems possible and would generate an extremely damaging tsunami in a highly populated fiord system). Obviously, not all landslide-prone slopes can be identified and monitored; however, an alternative strategy available with modern remote sensing technologies is to search for geomorphic evidence of precursory deformation, such as slumping, developing cracks and small-scale rockfalls (e.g., Stähli et al., 2015; Wasowski and Bovenga, 2021; Rosser and Massey, 2021). There is evidence that geomorphic detection of precursory motion is also feasible for earthquake-generated landslides, based on the recent realisation that many earthquake-affected slopes may deform episodically in a sequence of earthquakes (perhaps hundreds of years apart over thousands of years) until eventually an earthquake (or, indeed, a rainstorm; Chigira et al., 2013) causes further deformation that completes a failure surface and causes catastrophic collapse. Davies and Moretti (2021) investigate this possibility.

A peculiarity of some major mountain landslides that is now becoming apparent is that a significant number of them occur with no perceptible trigger. For example, since 1991, seven major (>10⁶  m³) rock avalanches have occurred in the Southern Alps of New Zealand, none of which was associated with either earthquake or rainstorm (e.g., McSaveney, 2002; Hancox and Thompson, 2013). Furthermore, given the relatively frequent occurrence of both earthquakes and severe rainstorms in the Southern Alps, the fact that these landslides did not occur in response to any previous potential triggering event suggests that whatever processes were reducing the slope factor of safety (e.g., freeze-thaw, permafrost degradation and stress corrosion), this reduction was occurring quite rapidly. Interestingly, this temporally clustered but spatially dispersed set of landslides could, if not historically recorded, have been interpreted by future scientists as evidence of a widespread trigger earthquake or rainstorm – a potential error that cautions against simplistic paleo-interpretation of landslide deposits (Clague, 2021). In this case, a seismic trigger would appear unlikely because all of the landslide source areas are shallow and broad, indicating surficial material failure, whereas major coseismic landslides often have deep-seated source-area scars (Turnbull and Davies, 2006).

Finally, it is important in considering future landslide disasters to realise that the dominant role of landslides in mountain geomorphology means that landslides in mountains are likely to trigger consequential cascading geomorphic events of a variety of types, whose effects can propagate many tens or even hundreds of kilometers across a landscape and can persist for many decades. A notable example at the time of writing is the failure from Raunthi peak, Uttarakhand, in the Indian Himalaya on 7 February 2021, which sent a flood of debris, ice and water down the Dhauliganga river, resulting in more than 200 fatalities. Wider examples of similar chains and the challenges they pose include the following:

• A large landslide falling into a mountain valley is likely to affect, or even block, the river in the valley (Korup and Wang, 2021). A blockage forms a landslide dam that can fail when overtopped (immediately or many decades later) causing a short but severe flood to pass along the valley (e.g., Hancox et al., 2005).

• Flooding upstream before the landslide dam fails can also be troublesome, but is often less catastrophic.

• The large input of sediment to a river from a large landslide (or indeed a large number of small landslides), whether or not the river is blocked, causes the river behaviour to alter. Typically, the river will aggrade to increase its slope and sediment transport capacity, increasing flood risk to the valley floor and to downstream floodplains (e.g., Davies and Korup, 2007; Robinson and Davies, 2013).

• A landslide in a small steep catchment can substantially increase debris flow risk in that catchment and on its fan, which many regions are some of the only areas available for human habitation.

• Large landslides falling into lakes or bays can cause catastrophic tsunami damage to assets close to water level (e.g., Lituya Bay, Alaska, 1954: Weiss et al., 2009).

• A large landslide falling onto a glacier can trigger a far-reaching rock–ice avalanche (e.g., Kolka–Karmadon: Huggel et al., 2005; Sosio, 2021; and Huascaran: Pflaker and Ericksen, 1978).

• The triggering event can be difficult to identify and hence cause to be attributed, particularly where evidence of even large landslides in high mountain areas can be rapidly lost due to new snow cover (Dunning et al., 2015). As a result, assessing the true magnitude and frequency of such events, and changes in the conditions which lead to such events, remains challenging.

Thus, assessing the risks consequential on the occurrence of a landslide is no simple task (Robinson and Davies, 2013). For example, the 2008 Wenchuan earthquake caused between 56,000 (Parker et al., 2011) and 200,000 landslides (Xu et al., 2013); several hundred of these formed landslide dams that threatened downstream cities with the largest, the Tangjiashan landslide dam, threatening a total of 1.2 million people downstream with a potentially catastrophic outburst flood (Xu et al., 2009; Peng and Zhang, 2012). The largest individual landslide contained >750 million cubic meters (0.75  km³) of rock debris and affected a total area of >7  km² (Huang et al., 2012). There is also substantive evidence (Berryman et al., 2001; Davies and Korup, 2007) that past coseismic landslides in the South Island, New Zealand, have caused aggradation of several meters over hundreds of square kilometers of outwash surfaces that are today intensively utilized for farming, towns and roads. Future earthquake-generated landslides may therefore have severe impacts on society's use of these areas for up to some decades after the occurrence of the landslides. Such changes in landsliding and its consequences often play out over landscapes undergoing wider forms of similarly rapid change. For example, after the 2015 Gorkha earthquake in Nepal, landsliding 5  years after the earthquake remained at levels considerably higher than on the day of the earthquake itself (Kincey et al., 2021). Whilst much of this involved the remobilisation of slopes destabilised by the earthquake itself, a considerable proportion of this sustained risk has been anecdotally attributed to the proliferation of rural road construction, much of which is directly associated with political changes made since 2015 (Rosser et al., 2021). Distilling the influence of changing landslide hazard, the evolving nature of people's exposure and the role that society plays in both driving and then experiencing landslide disasters is therefore complex.

1.5. Conclusion

The hazards that landslides pose to society tend to be less well appreciated than the better understood hazards of floods. This is largely because landslides seldom recur in the same places, whereas floods tend to occur adjacent to water bodies where people have lived for a long time and are thus expected. Communities can nowadays often be provided with hours' or days' warning of major floods, with the opportunity to evacuate people (saving lives) and elevate possessions (reducing damage); by contrast, very few landslide locations and times of occurrence are able to be predicted with sufficient reliability to make warning and evacuation a realistic option, so identifying potentially high risk locations remains an important goal. Nevertheless, landslides are a major contributor to deaths, damage and disruption to society (e.g., Sudmeier-Rieux et al., 2013), particularly in association with earthquakes and intense rainfall. In this context, anticipating the landslide-generated effects of earthquakes and rainstorms is a glaring omission from many if not most hazard management strategies. As a general comment, perhaps the scientific input to developing such strategies has until recently been too discipline constrained, with seismologists and hydrologists not engaging sufficiently with landslide experts – or indeed with each other – to identify the full suite of geomorphic consequences of a major trigger event that can play out over months, years or decades (e.g., Kincey et al., 2021). The currently almost universal risk-based approach to reducing the impacts of specific landslide disasters has serious limitations and appears to be reliable only for minor events. Preparation for, and developing resilience to, larger events is more likely to be successful if based on event and consequence scenarios.

References

1. Berryman K, Alloway B, Almond P, et al. Alpine fault rupture and landscape evolution in Westland, New Zealand. In: Proceedings, 5th International Conference of Geomorphology, Tokyo. 2001.

2. Bowman E.T. Small landslides: frequent, costly, and manageable. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

3. Chigira M, Tsou C.Y, Matsushi Y, Hiraishi N, Matsuzawa M. Topographic precursors and geological structures of deep-seated catastrophic landslides caused by Typhoon Talas. Geomorphology. 2013;201:479–493.

4. Clague J.J. Paleo-landslides, paleolandlsides. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

5. Davies T.R.H. Averting predicted landslide catastrophes: what can be done? Ital. J. Eng. Geol. Environ. TOPIC2. In: International Conference Vajont 1963-2013, Padua, Italy. 2013:229–236.

6. Davies T. Reducing impacts of landslide disasters. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

7. Davies T.R.H, Korup O. Persistent alluvial fanhead trenching resulting from large, infrequent sediment inputs. Earth Surf. Processes Landforms. 2007;32:725–742. doi: 10.1002/esp.1410.

8. Davies T, Moretti D. Geomorphic precursors of large landslides: seismic preconditioning and slope-top benches. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

9. Dunning S.A, Rosser N.J, McColl S.T, Reznichenko N.V. Rapid sequestration of rock avalanche deposits within glaciers. Nat. Commun. 2015;6:7964.

10. Froude M.J, Petley D. Global fatal landslide occurrence from 2004 to 2016. Nat. Hazards Earth Syst. Sci. 2018;18:2161–2181.

12. Hancox G.T, McSaveney M.J, Manville V.R, Davies T.R.H. The October 1999 Mt Adams rock avalanche and subsequent landslide dam-break flood and effects in Poerua River, Westland, New Zealand. N. Z. J. Geol. Geophys. 2005;48:683–705.

11. Hancox G.T, Thompson R. The January 2013 Mt Haast rock avalanche and Ball Ridge rockfall in Aoraki/Mt Cook National Park, New Zealand. GNS Sci. Rep. 2013;2013(33):26.

13. Hewitt K. Quaternary moraines vs catastrophic rock avalanches in the Karakoram Himalaya, Northern Pakistan. Quat. Res. 1999;51:220–237.

14. Huang R, et al. The characteristics and failure mechanism of the largest landslide triggered by the Wenchuan earthquake, May 12, 2008, China. Landslides. 2012;9:131–142. .

15. Huggel C, Zgraggen-Oswald S, Haeberli W, Kääb A, Polkvoj A, Galushkin I, Strom A.The 2002 rock/ice avalanche at Kolka/Karmadon, Russian Caucasus: assessment of extraordinary avalanche formation and mobility, and application of QuickBird satellite imagery. Nat. Hazards Earth Syst. Sci. 2005;5(2).

16. Kincey M.E, Rosser N.J, Robinson T.R, Densmore A.L, Shrestha R, Pujara D.S, Oven K.J, Williams J.G, Swirad Z.M.Evolution of coseismic and postseismic landsliding after the 2015 Mw 7.8 Gorkha earthquake, Nepal. J. Geophys. Res.: Earth Surf. 2021 doi: 10.1029/2020JF005803.

17. Kirschbaum D, Stanley T, Zhou Y. Spatial and temporal analysis of a global landslide catalog. Geomorphology. 2015;249:4–15. doi: 10.1016/j.geomorph.2015.03.016.

18. Korup O, Clague J.J. Natural hazards, extreme events, and mountain topography. Quat. Sci. Rev. 2009;28:977–990.

19. Korup O, Wang G. Multiple landslide-damming episodes. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

20. Kritikos T, Robinson T.R, Davies T.R.H. Regional coseismic landslide hazard assessment without historical landslide inventories: a new approach. J. Geophys. Res. Earth Surf. 2015;120 doi: 10.1002/2014JF003224.

21. Kveldsvik V, Einstein H.H, Nilsen B, Blikra L.H. Numerical analysis of the 650,000 m² Åknes rock slope based on measured displacements and geotechnical data. Rock Mech. Rock Eng. 2009;42:689–728.

23. McColl S.T, Davies T.R.H. Evidence for a rock-avalanche origin for ‘The Hillocks’ moraine, Otago, New Zealand. Geomorphology. 2011;127:216–224.

24. McSaveney M.J. Recent rockfalls and rock avalanches in Mount Cook National Park, New Zealand. In: Evans S.G, DeGraff J.V, eds. Catastrophic Landslides: Occurrence, Mechanisms and Mobility. vol. 15. Geological Society of America Reviews in Engineering Geology; 2002:35–70.

25. Milledge D.G, Densmore A.L, Bellugi D, Rosser N.J, Watt J, Li G, Oven K.J. Simple rules to minimise exposure to coseismic landslide hazard. Nat. Hazards Earth Syst. Sci. 2019;19:837–856. doi: 10.5194/nhess-19-837-2019.

26. Parker R.N, et al. Mass wasting triggered by the 2008 Wenchuan earthquake is greater than orogenic growth. Nat. Geosci. 2011;4:449–452.

27. Peng M, Zhang L.M. Analysis of human risks due to dam break floods-part 2: application to Tangjiashan landslide dam failure. Nat. Hazards. 2012;64:1899–1923.

28. Petley D. On the impact of climate change and population growth on the occurrence of fatal landslides in South, East and SE Asia. Q. J. Eng. Geo. Hydrogeol. 2010;43(4):487–496.

29. Pflaker G, Ericksen G.E. Nevados Huascaran avalanche, Peru. In: Rockslides Avalanches. vol. 1. 1978:277–314.

30. Reznichenko N.V, Davies T.R.H, Shulmeister J, Larsen S.H. A new technique for distinguishing rock-avalanche-sourced sediment in glacial moraines with some paleoclimatic implications. Geology. 2012;40(4):319–322. doi: 10.1130/G32684.1.

31. Reznichenko N.V, Davies T.R, Winkler S. Revised palaeoclimatic significance of Mueller Glacier moraines, Southern Alps, New Zealand. Earth Surf. Process. Landforms. 2016;41(2):196–207.

32. Robinson T.R, Davies T.R.H. Review article: potential geomorphic consequences of a future great (Mw = 8.0) Alpine Fault earthquake, South Island, New Zealand. Nat. Hazards Earth Syst. Sci. 2013;13:2279–2299. doi: 10.5194/nhess-13-2279-2013. .

44. Robinson T.R, Davies T.R.H, Wilson T.M, Orchiston C. Coseismic landsliding estimates for an Alpine Fault earthquake and the consequences for erosion of the Southern Alps, New Zealand. Geomorphology. 2016;263:71–86. doi: 10.1016/j.geomorph.2016.03.033.

33. Rosser N, Massey C. Rockfall. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

34. Rosser N.J, Kincey M.E, Oven K.J, Densmore A.L, Robinson T.R, Pujara D.S, Shrestha R, Smutny J, Gurung K, Lama S, Dhital M.R.Changing significance of landslide hazard and risk after the 2015 Mw 7.8 Gorkha, Nepal earthquake. Prog. Disaster Sci. 2021:100159. doi: 10.1016/j.pdisas.2021.100159.

35. Sornette D. Dragon-kings, black swans, and the prediction of crises. Int. J. Terraspace Sci. Eng. 2009;2:1–18.

36. Sosio R. Rock-snow-ice avalanches, rock-snow-ice avalanches. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

37. Stähli M, Sättele M, Huggel C, McArdell B.W, Lehmann P, Van Herwijnen A, Berne A, Schleiss M, Ferrari M, Kos A, Or D, Springman S.M. Monitoring and prediction in early warning systems for rapid mass movements. Nat. Hazards Earth Syst. Sci. 2015;15:905–917. doi: 10.5194/nhess-15-905-2015.

38. Sudmeier-Rieux K, Jaquet S, Baysal G.K, Derron M, Devkota S, Jaboyedoff M, Shrestha S.A neglected disaster: landslides and livelihoods in central-eastern Nepal. In: Margotini C, et al., ed. Landslide Science and Practice. vol. 4. Springer-Verlag Berlin Heidelberg; 2013:169–176. doi: 10.1007/978-3-642-31337-0_22.

39. Turnbull J.M, Davies T.R.H. A mass movement origin for cirques. Earth Surf. Processes Landforms. 2006;31:1129–1148. doi: 10.1002/esp.1324.

40. Wasowski J, Bovenga F. Remote sensing of landslide motion with emphasis on satellite multitemporal interferometry applications: an overview. In: Davies T.R.H, Rosser N.J, eds. Landslide Hazard and Disasters. 2021 (in this volume).

41. Weiss R, Hermann M.F, Wünnemann K. Hybrid modeling of the mega-tsunami runup in Lituya Bay after half a century. Geophys. Res. Lett. 2009;36(9).

42. Xu C, et al. Application of an incomplete landslide inventory, logistic regression model and its validation for landslide susceptibility mapping related to the May 12, 2008 Wenchuan earthquake China. Nat. Hazards. 2013;68:883–900.

43. Xu Q, Fan X.-M, Huang R.-Q, Westen C.V. Landslide dams triggered by the Wenchuan earthquake, Sichuan Province, south west China. Bull. Eng. Geol. Environ. 2009;68:373–386.

Chapter 2: Landslide causes and triggers

Samuel T. McColl     Geosciences Group, School of Agriculture and Environment, Massey University, Palmerston North, New Zealand

Abstract

Hillslope (in)stability is governed by the balance of stability factors. If stability is lost, gradually or instantly, slope failure ensues. Assessing the causes of instability is useful for hazard analysis and mitigation, and for considering the role of landslides in landscape systems and evolution. Geological and geomorphological conditions (e.g. material type, strength and structure and hillslope geometry) predispose slopes to failure; knowledge of these conditions can help to predict the location, types and volumes of potential failures. The timing of failure, often by a specific trigger, can be anticipated by detecting and assessing movement patterns, establishing triggering thresholds or using probabilistic methods. However, predicting timing remains challenging due to the difficulty of measuring material strength degradation which can lead to failure with no readily observable trigger. This chapter describes concepts of stability and explores some of the major causes and triggers of hillslope failure and opportunities for further research.

Keywords

Factor of safety; Hillslope failure; Landslide susceptibility; Mass movement; Stability factors

2.1. Introduction

On the 8th of August 1979 in Abbotsford, New Zealand, several weeks of preliminary movements culminated in the rapid failure of an approximately five million m³ block slide in a residential area. The cost of the Abbotsford Landslide amounted to some NZ$ 10–13 million and included the loss of 69 houses, but no human casualties (Hancox, 2008). As with the Abbotsford Landslide case, when any landslide or other hazard results in financial or human losses, an element of human blame will often arise, particularly in cases where insurance companies need to make loss adjustments. This, and the desire to prevent similar disasters in the future, necessitates investigation into the causes of a landslide. Furthermore, to gain an appreciation of the role of landslides in landscape systems and the geomorphic evolution of landscapes, it is necessary to have some knowledge of the factors that determine where and over what timescales different types of landslides are likely to occur or have occurred. As with almost every other landslide, the cause of the Abbotsford Landslide cannot be put down to a single factor; its failure was the result of the convergence of a series of unfavourable conditions, both natural and human induced. Hancox (2008) reports on the results of a Government Commission of Inquiry which set out to identify the causes of this landslide. The following factors were considered to have contributed:

1. Unfavourable geology, consisting of weak clay layers in gently tilted strata.

2. Removal of toe support over the preceding millennia by natural fluvial erosion and later by artificial quarrying 10 years prior to failure.

3. A long-term rise in groundwater resulting from a leaking water main above the slide area, clearance of vegetation and a natural increase in rainfall over the preceding decade.

4. Urban development potentially adding a load surcharge.

5. Rainfall during the last few days of movement, which by itself may not have been the difference between failure and stability, but likely influenced the timing and nature of the landslide's final rapid movement.

Despite all of these causes, the disaster, rather than the landslide event, stems from the decision to build (perhaps unknowingly) on terrain susceptible to landsliding without adequate precautionary mitigation measures in place. Although land use decisions were a root cause of the disaster, the purpose of presenting this case study here is to highlight that, as with many landslides, none of the recognised factors alone would have been sufficient or necessary to cause the landslide. Assigning blame to any person, organisation or decision is therefore challenging. Yet, knowledge of the various factors that can cause landslides is invaluable for hazard mitigation. This knowledge helps to identify and avoid human activities that may affect, or be affected by, hillslope instability, and it improves the capability for predicting natural landslides. For example, providing effective advanced warnings for debris flows requires appropriate triggering thresholds to be set. At which level these thresholds are set and indeed the identification of the triggers themselves requires thorough understanding of the various factors causing and triggering debris flows. These factors may include the availability of debris to be mobilised, the geotechnical properties of the debris, antecedent soil moisture conditions, rainfall magnitude and intensity and the condition of vegetative cover. Without sufficient knowledge of contributing instability factors and the appropriate application of that knowledge, hazard mitigation efforts may be wasted.

A great deal of scientific observation and enquiry over the last few decades has contributed to more in-depth classification, rationalisation and understanding of the natural and human causes and triggers of landsliding. This has stemmed largely from the disciplines of engineering geology, soil and rock mechanics and geomorphology, but there is now a well-established field of landslide science. This chapter does not attempt to list or comment on the history of significant developments in landslide research (which has been done by others; e.g. Crozier, 1986; Petley, 2011), nor does it attempt to describe every known landslide cause. Instead, the work focuses on causes that are of high importance and remain fruitful objectives of scientific research. To begin, the concept of hillslope stability is introduced to provide a conceptual basis for understanding the effects of various hillslope conditions and destabilising processes. This is followed by a brief overview of the main types of landslide causes and triggers and the current understanding of these. Finally, a brief synopsis of landslide causes and opportunities for further research is presented.

2.2. Concept of instability

Destabilising stresses are present within all slopes. Whether or not these stresses (driving stresses) are capable of triggering failure of a given hillslope at a given moment in time will depend on the relative magnitude of stresses that resist the tendency for failure; these opposing stresses can be referred to as resisting stresses. Both driving and resisting stresses can change over time. Thus, ‘stability’ is both a relative term and one that holds meaning over a certain time period. The vast majority of existing slopes are stable at this moment in time (otherwise they would be in the process of failing), but every hillslope has the potential to fail at some future time. The magnitudes of driving and resisting stresses are the result of stability factors, which can be defined as any phenomena that control or influence the forces that determine the stability (Crozier, 1986). Some stability factors are inherent to the hillslope and unchanging (e.g. lithology), whilst others may be transient and their influence may come and go or vary in magnitude (e.g. porewater pressures).

Stability can be quantitatively assessed by measuring the balance of driving and resisting stresses. The development of shear stresses drives a tendency for failure for most landslide types (with the exception of toppling). The resisting stresses result from reactionary stresses, and can be considered as the mobilised shear strength of the hillslope with respect to the shear stresses. Mobilised strength refers to only the strength that is used to resist movement and is distinguished from the ‘total’ strength of the hillslope, which is arguably an impossible quantity to assess. For example, the movement of an intact sliding block (e.g. a rockslide) is influenced by the strength generated between the sliding surfaces, not the internal strength of the block itself; therefore, only the strength of the failure surface needs to be known to assess the stability of the sliding block. However, the assumption that landslides involving sliding are governed by shear stresses alone is an oversimplification. Other (e.g., compressional and tensional) stresses most likely also play a role, especially at the boundaries of the sliding mass and where the landslide mass moves over irregular surfaces.

The balance of strength and driving forces is often expressed as a ratio, commonly referred to as the factor of safety (FoS):

The FoS equation can be populated with a wide range of parameters which influence the resisting and/or driving stresses. The choice and determination of these parameters depends on the problem being addressed, the physical conditions and processes expected for the hillslope and the degree of simplicity or complexity being achieved. The method of assessing stability by determining the FoS is referred to as limit equilibrium analysis. Examples and explanations of this process and the selection of relevant parameters and variables are provided in numerous texts on hillslope stability (e.g. Duncan, 1996; Norrish and Wyllie, 1996; Selby, 1993).

The effect of gravity is of fundamental importance in the stability of a hillslope: The strengths and stresses operating in a slope depend on the way that gravity causes masses within a slope to interact. The weight of material can be resolved into stresses acting normal and parallel to contact surfaces (Fig. 2.1). The greater the normal stresses, the greater the frictional shear strength of the potential failure surface. The relationship between shear strength and the normal stresses is often considered to be linear and governed by the Mohr–Coulomb failure criterion. The greater the stresses acting parallel to the potential failure surfaces, the larger the shear stress. Many of the stability factors discussed below influence the stability of a hillslope by directly affecting or altering, acutely or chronically, the shear stresses and/or shear strength. Thus, it is useful to keep the relationship between these two stresses in mind.

The term ‘slope stability’ can have different meanings depending on the context of its use or the objectives of its user. It often refers to the inherent stability or FoS, as determined by the static physical hillslope properties; for example, all other stability factors being equal, a high embankment is less stable (i.e. lower FoS) than a low embankment. Landslide susceptibility, which is a measure of the inherent stability of a hillslope or distribution of hillslopes, treats the term stability in this way, without any consideration of the likelihood of failure. Alternatively, slope stability can, sometimes more usefully, be a measure of the probability of a failure of an individual hillslope, which requires consideration of both the stability factors (i.e. slope stability in the previous sense), and the likelihood of a critical failure threshold being exceeded in a given time span.¹ Assessing the stability of the above two embankments in this sense requires knowledge of the potential failure triggers and their likelihood during the time period of interest for each embankment. That knowledge must include the probability that a trigger of sufficient (critical) magnitude to induce failure will occur. It might be worth noting that the critical magnitude required is itself not static in time, but fluctuating over some range, so its determination may also warrant a probabilistic approach. The higher embankment is thus not necessarily more likely to fail than the lower embankment, because it may exist in an environment less likely to produce a trigger capable of exceeding the critical threshold for stability for that slope.

Figure 2.1  (A) The influence of slope angle (β) on the relative magnitudes of shear (τ) and normal stress (σ). (B) Stresses acting along a potential failure surface. 

Adapted from Selby (1993), by permission of Oxford University Press.

In considering the causes of mass movements, a distinction must also be made between the terms ‘failure’ and ‘movement’. Some mass movements only ever exist as a discrete event with the failure resulting in the first and only movement of that material; for example, a rockfall. Other mass movements can exist in a state of transient stability for a long time after failure is first initiated and be affected by entirely different movement triggers or controlling factors to those that initiated the first failure. These slopes may experience discrete periods of instability (movement) and periods of stability (no movement), as is often the case for slopes affected by large, slow-moving rockslides. Other mass movements may evolve from an initially very slow creeping mass and begin to accelerate as the failure surface develops, culminating in a catastrophic final failure; such mass movements are termed progressive failures (Petley et al., 2005). In such cases, it is not clear to which part of this process the term ‘failure’ applies; the final catastrophic failure only, the initiation of creep or the entire process (which may take place over an indeterminate amount of time). Perhaps it depends on the scale at which a hillslope is observed. Petley (2011) explains a distinction between a local and a global FoS for a hillslope. The global FoS applies to an entire landslide body, although this in itself may be difficult to define for landslides involving movement of multiple (nested) blocks. Failure of the entire mass occurs if the global FoS falls below unity. However, even if the landslide as an entire mass may be stable (global FoS  >  1), there may be parts of the landslide body that are locally unstable. Petley points out that those unstable parts of the landslide mass will locally undergo deformation, and cumulative deformation of these unstable parts may lead to progressively greater reductions in the global FoS (see Section 3.2).

Terms such as ‘stable’, ‘marginally stable’ and ‘actively unstable’ have been introduced in recognition of the need to classify the stability state of landslides as a function of the likelihood of failure (Fig. 2.2) (Glade and Crozier, 2005). In this context, ‘stable’ refers to a hillslope that exists in an environment which is unlikely to produce a trigger sufficient to cause failure or movement of that slope (Fig. 2.3). A