Landslide Hazards, Risks, and Disasters

Tim Davies

• Publisher: Elsevier
• Year: 2014
• eBook ISBN: 9780123964755
• Topic: Biological Sciences
• Subtopic: Environmental Science
• Index: Biological Sciences

Landslides are the most costly geo-hazard in the world, and they're often the cause or the result of other hazards and disasters such as tsunamis, earthquakes, wildfires, and volcanic eruptions. Landslide Hazards, Risks, and Disasters makes a close and detailed examination of major mass movements and provides measures for more thorough and accurate monitoring, prediction, preparedness, and prevention. It takes a geoscientific approach to the topic while also discussing the impacts human-induced causes such as deforestation, blasting, and building construction—underscoring the multi-disciplinary nature of the topic.- Contains contributions from expert geologists, seismologists, geophysicists, and environmental scientists selected by a world-renowned editorial board- Presents the latest research on causality, economic impacts, fatality rates, and landslide and problem soil preparedness and mitigation- Numerous tables, maps, diagrams, illustrations, photographs, and video captures of hazardous processes- Discusses steps for prevention and treatment of problem soils, the most expensive geo-hazard in the world

Chapter 1

Landslide Hazards, Risks, and Disasters

Introduction

Tim Davies Geological Sciences, University of Canterbury, Christchurch, New Zealand

Abstract

Consideration of the nature and occurrence of landslide hazards leads to perspectives on the dominant role of landsliding in the geomorphology of active orogens, and consequently the major role that landsliding plays in determining hazard- and riskscapes both within and downstream of these areas. Discussion of landslide risks suggests that probabilistic analyses are only likely to be reliable in planning location-specific landslide risk management strategies for small, frequent events, and the potential for identifying sites of future landslides—both rainfall generated and coseismic—is examined. Finally the role of landslides in triggering consequential hazards, such as tsunami, river flooding, and debris flows, is emphasized.

Keywords

Landslides; Active landscapes; Hazards; Risks; Disasters

1.1. Introduction

Landslides are a ubiquitous phenomenon on a planet that, like Earth, is tectonically active. However society is generally inclined to view them 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 decade, however, global landslide occurrence and impact has been better documented (e.g. Petley, 2012), and the seriousness of this hazard underlined. Interestingly, climate change is shown to have much less influence on the number of landslide fatalities than population growth (Petley, 2010), and if this trend continues, landslide fatalities will continue to increase. 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.

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, and whether 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 a 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, but we have to draw a line somewhere…

Landslides are a crucial component of Earth's geological cycle, in which tectonic plate motion causes parts of the crust to be continuously uplifted above a base level; they are then continuously eroded down again by gravity and gravity-driven water flow toward base level. Landslides represent the directly gravity-driven component of erosion, and they occur in sizes ranging from individual rocks falling to whole mountainsides collapsing. 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 dominate the sediment supply spectrum (Korup and Clague, 2009); and, since the majority of river-transported sediment originates in slope failures, this emphasizes 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 little evidence of any erosion process other than mass movement (Figure 1.1).

Thus, in steep, active terrain, the progress of geology requires that landslides will continue to occur on hillslopes in the future; and the increasing presence of people and their assets on, in the vicinity of and downstream of these hillslopes, means that landslide-generated 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 (Figure 1.2), suggesting that there is some factor constraining the frequency of occurrence of landslides of various sizes. This obviously has 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 mega-events, known as “dragon-king” events (Sornette, 2009), occur much more frequently than the distribution would suggest (indicated by the red line in Figure 1.2), and appear to 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 include these boundaries. In landslide terms 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, so that these larger events are in fact manifested as smaller (but still very large) events, which thus acquire correspondingly higher (but still small) frequencies. 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 disasters; 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. Hence the largest events occur more frequently than the probability function for smaller events would suggest.

FIGURE 1.1 The Southern Alps, New Zealand from the west. All of the landforms in the upper part of the picture are mass movement related, with a complete absence of glacial landforms (in spite of a small glacier being visible). The prominent high terraces in the center of the picture are glacial and/or tectonic in origin, while the sediment being reworked by the river at bottom is largely landslide derived.

FIGURE 1.2 Landslide area—probability density distribution. (After Malamud et al., (2004).) The red curve indicates “dragon-king” events (see text).

To better deal with landslide hazards, it is first necessary to know where landslides are likely to occur; and, second, how big they will be. These two steps can lead to an event scenario for the hazard. Estimating a probability for an event of given size and location is much more difficult, and is in fact of lesser value from a disaster reduction perspective, because we are interested predominantly in reducing the impacts of the next disaster that will affect a locality, and probabilities give no reliable information about when that will occur or how big it will be, even in an ideal world with perfect magnitude–frequency information. Thus a landslide susceptibility map, together with an event scenario, provides local people and their governments with realistic information about what can happen there at any time; this can then be used, in conjunction with information on the location of societal assets, to develop a societal consequence scenario that forms the basis for designing hazard avoidance and/or damage reduction (assuming that preventing a major landslide is an unrealistic ambition) at all scales from personal to societywide and for thinking about disaster recovery frameworks.

The tricky bit in this train of logic lies in choosing the magnitude of the event scenario. A case can be made for selecting the worst-case scenario, on the basis that a community that has thought through how to cope with this scenario can also cope with anything smaller (although here it needs to be recognized that calculated worst-case scenarios—“maximum credible events”—have recently been dramatically exceeded in a number of earthquake disasters); however, this is likely to be criticized as scare mongering, on the grounds that the maximum possible event occurs incredibly rarely. Thus a somewhat less catastrophic scenario is likely to gain broader acceptance, although here it is easy to set off down the slippery slope of associating probabilities with scenarios. Recent geo-disasters have been spoken of by scientists as having return periods of tens of thousands of years, so it is quite clear that a useful disaster event scenario needs to be substantially worse than the commonly used “100-year” event. People who experience major disasters learn that statistical improbability does not prevent rare events from happening at any time.

Developing a consequence scenario from an event scenario depends on having (or developing) information about where vulnerable assets are located, and where they will be located in the future. A distinct benefit of consequence scenarios is that if (as is often the case) the landslide location (event scenario) is uncertain, then the areal extent of the consequence scenario can be expanded accordingly to accommodate this uncertainty. Foreseeing landslide effects is clearly difficult; one needs to know the volume of the landslide and its location to predict the extent of its deposit and what...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Contributors
• Editorial Foreword
• Preface
• Chapter 1. Landslide Hazards, Risks, and Disasters: Introduction
• Chapter 2. Landslide Causes and Triggers
• Chapter 3. Mass Movement in Bedrock
• Chapter 4. Coseismic Landslides
• Chapter 5. Volcanic Debris Avalanches
• Chapter 6. Peat Landslides
• Chapter 7. Rock–Snow–Ice Avalanches
• Chapter 8. Multiple Landslide-Damming Episodes
• Chapter 9. Rock Avalanches onto Glaciers
• Chapter 10. Paleolandslides
• Chapter 11. Remote Sensing of Landslide Motion with Emphasis on Satellite Multitemporal Interferometry Applications: An Overview
• Chapter 12. Small Landslides—Frequent, Costly, and Manageable
• Chapter 13. Analysis Tools for Mass Movement Assessment
• Index