Author's personal copy
7.26 Mass-Movement Hazards and Risks
M Crozier, Victoria University of Wellington, Wellington, New Zealand
r 2013 Elsevier Inc. All rights reserved.
7.26.1
7.26.2
7.26.2.1
7.26.2.2
7.26.2.3
7.26.3
7.26.4
7.26.5
7.26.5.1
7.26.5.2
7.26.5.3
7.26.5.4
7.26.6
7.26.6.1
7.26.6.1.1
7.26.6.1.2
7.26.6.1.3
7.26.6.1.4
7.26.6.2
7.26.6.3
7.26.6.4
7.26.7
References
Introduction
The Physical Context
The Energy of Processes
Mass Movement
Mass Movement in Mountains
The Human Context
Social and Physical Environmental Change
Concepts: Hazard, Risk, and Susceptibility
Hazard
Risk
Susceptibility
Vulnerability
Assessing Hazard and Risk
The Nature of Mass-Movement Hazard
Dynamic or static
Site specific or nonsite specific
Landslide type
Intensity of landslides
Where Landslides Occur: Susceptible Terrain
Frequency of Occurrence
Risk
Conclusion
Glossary
Elements at risk These include people, their well-being,
buildings, infrastructure, economic activity, and all other
things valued by the community.
Hazard (physical) A hazard is a potentially damaging
process or condition, for example, an earthquake above a
certain intensity or a landslide of sufficient size, depth, or
displacement to cause damage or disruption or, as an
example of a condition, the presence of weak foundation
material.
Hazard (temporal condition) The probability of a
potentially damaging event (a landslide) occurring
in a unit of time. This probability varies with the
magnitude of the event (generally small landslides occur
more frequently than large landslides). Consequently,
hazard is often expressed as the probability of occurrence
of a given magnitude of event. Defined in this way,
hazard represents a state or condition and is assessed and
applied to a particular place, for example, site, unit area
of land surface, region, or object, such as lifelines,
hydrodams, etc.
250
250
250
250
250
251
251
253
253
253
253
253
253
254
254
254
254
254
255
255
256
256
256
Risk Expected consequences emanating from a hazard,
expressed as the probability and severity of loss to the
elements at risk for a unit area, object, or activity, over a
specified period of time. Risk is a function of the magnitude
and frequency of hazard, the elements exposed to the
hazard and their vulnerability.
Susceptibility The propensity of a designated area to
experience a particular physical hazard, based on the
presence of indicative conditions, stability analysis, or
precedence established from historical records or other
proxy evidence. Susceptibility assessments do not provide a
measure of hazard (temporal condition). However,
susceptibility of an area is commonly ranked (or zoned)
from high to low based on the strength of indicative
parameters.
Vulnerability The degree of damage expected from given
magnitude of hazard, usually expressed as a ratio of the
existing value. For example, a structure that receives damage
amounting to 50% of its value would record a vulnerability
of 0.5, whereas complete destruction would be recorded
as 1.0.
Crozier, M., 2013. Mass-movement hazards and risks. In: Shroder, J. (Editor
in Chief), Marston, R.A., Stoffel, M. (Eds.), Treatise on Geomorphology.
Academic Press, San Diego, CA, vol. 7, Mountain and Hillslope
Geomorphology, pp. 249–258.
Treatise on Geomorphology, Volume 7
http://dx.doi.org/10.1016/B978-0-12-374739-6.00175-5
249
Author's personal copy
250
Mass-Movement Hazards and Risks
Abstract
Mountains and steeplands offer conditions conducive to high magnitude and frequent mass-movement hazards, most
dramatically in the form of landslides. These conditions in many instances are being exacerbated by global environmental
change emanating from both physical and human systems. The concepts of susceptibility, hazard, vulnerability, and risk
from landslides are examined in the light of these changes. In addition, the specific information requirements and approaches for assessing hazard and risk are explored. In particular, emphasis is placed on understanding the different
characteristics of landslide processes, including intensity, magnitude and frequency, and location. Assessment methodologies and decisions based on their outcomes need to acknowledge the rapid evolution of hazard and risk in response to
unprecedented rates of environmental change.
7.26.1
Introduction
Mountains have played an important part in understanding
the geomorphology of the Earth and have a long-held fascination for geomorphologists. They are the most obvious
source of the principal fluxes of sediment and water that travel
over the Earth’s surface as rivers, intermittently depositing and
re-entraining until they reach the ultimate sink of the ocean.
The great debates on the evolutionary relationship between
land form and process along this pathway, stimulated by luminaries such as Gilbert (1877), Davis (1899), Penck (1953),
and Hack (1960), have been seminal for geomorphology.
However, today much steepland research is focused on
understanding process rather than interpreting form (Clague,
2009; Korup et al., 2010) – a shift in emphasis that has been
stimulated by concern over the increasing impact of natural
hazards associated with human activity (Cendrero et al., 2006;
Remondo et al., 2005; Marston, 2008) and the far-reaching
consequences of global warming for mountain regions (Haeberli et al., 1997; Huggel, 2009; Crozier, 2010). Mountains and
steeplands are not only important and emphatic morphological terrains of grand scale but also the domain where
gravity can dramatically demonstrate its power over earth
materials, with sometimes catastrophic results (Evans and
DeGraff, 2002). The intersection of human activity with these
high-energy geomorphic systems, together with the accelerating pace of social and environmental change, enhances
hazard and risk, and exacts an increasing toll on mankind
(Slaymaker, 2010).
7.26.2
7.26.2.1
The Physical Context
The Energy of Processes
The potential energy of geomorphic systems can be represented by the mass of slope-forming material or water together
with its elevation in the landscape. When material and water
(solid or liquid) are released to move downslope, that potential energy is converted into kinetic energy, which is capable of carrying out geomorphic work and impacting human
systems. The kinetic energy of a moving body is equal to half
of the product of mass times velocity squared – the power on
the velocity term indicating its preeminence in determining
the energy of a geomorphic process. For any given material in
motion down a mountain slope, velocity, in turn, is a function
of the slope gradient and resistance mobilized by slope conditions such as surface roughness and vegetation cover. With
such factors controlling the energy of geomorphic processes, it
is not surprising that mountain lands provide examples of
some of the highest energy and most destructive geomorphic
processes observed on the Earth.
7.26.2.2
Mass Movement
In the context of high-energy, active geomorphic processes in
mountain lands, a range of different processes and influencing
factors operate at different scales (Glade and Crozier, 2005;
Slaymaker, 2010). However, here the focus is on mass-movement hazards and risk and as such requires some specific
definitions in order to constrain the boundaries of the discussion. Mass movement is defined as the outward or downward movement of slope-forming material under the influence
of gravity. Whereas this definition precludes water as a
hydraulic transportational agent, water is commonly a critical
agent in reducing strength and allowing ambient gravity
conditions to promote movement. As a geomorphic process,
mass movement (involving blocks or aggregates undergoing
shearing and brittle or plastic strain) represents one end
of a spectrum of transportational processes that, with
increasing water content, grade initially into slurry flow (e.g.,
debris flow). As water content further increases, debris becomes transported as hyperconcentrated flow and ultimately
hydraulic flow which operate as turbulent Newtonian fluids.
Some mass-movement processes, such as soil creep, are
almost imperceptibly slow and diffuse, while others are discrete, fast moving, and with clearly identifiable boundaries,
commonly in the form of shear surfaces. It is these discrete
processes, referred to as landslides, that represent the most
hazardous of all mass-movement processes (Crozier, 1999a).
Comprehensive accounts of mass-movement hazards in
mountain lands have been provided by Eisbacher and Clague
(1984), Korup and Clague (2009), and Alcántara-Ayala and
Goudie (2010).
7.26.2.3
Mass Movement in Mountains
Most forms of mass movement, particularly landslides, are
episodic in occurrence and their onset is commonly (but not
always) associated with a climatic (Glade et al., 2000; Guzzetti
et al., 2008) or seismic trigger (Keefer, 1984; Keefer and
Wilson, 1989; Hancox et al., 1997). Mountains generate their
own climates, in places displaying extreme rainfall, dramatic
altitudinal and aspect gradients in temperature, humidity,
precipitation, and associated vegetation and soil conditions,
which in turn affect the type and magnitude and frequency
Author's personal copy
Mass-Movement Hazards and Risks
of geomorphic processes (Starkel and Sarkar, 2002). Although
some mountain ranges and steeplands exhibit low levels of
seismicity, most of the world’s dominant mountain ranges are
associated with compressional plate boundaries and are subject to earthquakes, active faulting, folding, and crushing – all
of which are conducive to the development of a range of massmovement processes from widespread frequent downslope
movements to rarer high-magnitude catastrophic failure
(Evans and DeGraff, 2002).
Mountain lands, therefore, are subject to processes that
both lower stability by reducing rock strength and provide
triggering mechanisms for failure, producing a set of potent
conditions highly conducive to the onset of rapid mass
movement. To give one example, the Alpine ranges of New
Zealand that show complex response to plate convergence at
rates of up to 50 mm yr1 exhibit strain by buckling, faulting,
rock shattering, mylonitization, and orogenic uplift. The rates
of uplift occurring on the Australian Plate in the North Island
are about 4 mm yr1, whereas, in the South Island, uplift
of the Pacific Plate is in excess of 10 mm yr1 (Cooper and
Norris, 2008). Although plate convergence and uplift are
ongoing, these mountains are considered to have reached an
equilibrium altitude where uplift is matched by denudation
(King, 2008). To achieve that equilibrium mass movement,
processes need to be exceedingly efficient. Orographically
enhanced rainfall in the Southern Alps of New Zealand in
places reaches 15 m yr1 and together with frequent earthquakes in excess of M7 along range bounding and intermountain faults (Figure 1), sufficient triggers exist to induce
catastrophic mass movement (Korup et al., 2004).
A notable example from the Southern Alps is the Green
Lake Landslide – one of the largest terrestrial landslides ever
recorded. The landslide has a volume of ~27 billion cubic
meters, covers an area of ~45 km2, dropped 700 m vertically,
and moved 2.5 km laterally. A debris dam 800 m in height
retained a lake 11 km in length. The landslide is thought to
have been triggered by an earthquake about 12–13 K years BP;
however, de-buttressing of slopes by glacier wasting may have
also been a factor (Hancox and Perrin, 2009).
Figure 1 Multiple rock and debris slides triggered by an Mw 7.2
earthquake, 22 August 2003 Fiordland, New Zealand. Photo by M.J.
Crozier.
7.26.3
251
The Human Context
Although landslides represent a natural geomorphic process
that in the absence of people or property is of little concern, it
is the juxtaposition of landslides, human use, and habitation
that exacts a cost. The potential loss from processes in such
situations defines both the hazard and the risk. Of all the areas
subject to mass-movement hazards, mountain lands are
clearly the most hazardous. In some areas of the world, for
example, New Zealand, apart from a few tourist centers, the
mountains are virtually uninhabited and consequently risk is
low. By contrast, in the mountainous country of Tajikistan
with about 6 million inhabitants, rockfalls and rockslides have
accounted for over 100 000 deaths during the twentieth century (Slaymaker, 2010). From a hazard and risk perspective,
Slaymaker (2010) classified mountains into four categories
reflecting demography and land use: polar mountains
(population density of o0.1 km2), for example, Svalbard;
low-population-density temperate mountains (population
density of 0.1–25 km2), for example, Southern Alps of New
Zealand; high-population-density temperate mountains
(population density of 25–75 km2), for example, Europe and
Japan; and tropical mountains (population density of
50–100 km2), for example, Indonesia.
7.26.4
Social and Physical Environmental Change
There is little doubt that increasing impact of hazards is driven
by accelerating population growth, associated increased demand for land resources, urbanization, and the capability of
modern technology to drastically change land form and processes (Figure 2). Wright (2004) has noted that, since the year
1900, the world has experienced a four-fold increase in
population and a 40-fold increase in economic activity. Rapidly accelerating urbanization and globalization mean that
transportation routes and other lifelines for transmission of
water, sewage, telecommunications, electricity, gas, and oil are
being expanded, commonly through terrain subject to massmovement hazards. In many parts of the world, population
pressures or poverty has seen habitation expand into areas
otherwise considered unsuitable for occupation (Figure 3).
Increased exposure to hazards can also be voluntary, as indicated by the surge of new housing in coastal locations, where
benefits of proximity to the sea and views have been the attraction. Cendrero et al. (2006) argued that many exploitative
activities resulting from increased population and wealth not
only increase exposure of the population to hazard but also,
by alteration of the landscape through such activities as deforestation and mining (Figure 4), exacerbate the onset of
hazards.
Numerous international databases indicate the increasing
loss from natural hazards and the disproportionate impacts
in less-developed countries. It is clear that improvement
in risk-reduction procedures and technology, although successful in reducing loss in advanced economies, has not
been able to stem the rising global impact. Although much
of this can be attributed to increasing population and its
consequences, the physical environment also appears to be
undergoing radical change. If climate changes to the extent
Author's personal copy
252
Mass-Movement Hazards and Risks
Elements at risk
Vulnerability
Exposure to
hazard
Increased landslide risk
and losses
Landslide activity
Triggering
activity
Susceptibility
Figure 2 Social and physical factors contributing to increased landslide risk and loss.
Figure 3 High-density residential tower blocks in proximity to steep
landslide prone terrain, Lantau Island, Hong Kong. Photo by M.J.
Crozier.
predicted by intergovernment agencies, then many areas
of the world can expect intensification of the thermal regime
and changes in precipitation, with some areas receiving
significant increases in both total rainfall and intensity.
Crozier (2010) has outlined how such changes may influence the geotechnical and geomorphic controls of mass
movement.
Currently, the possible impacts of global climatic change
are being pursued in a number of approaches: general circulation models (GCMs), downscaling of GCMs to assess local
climate, modeling slope stability using downscaled GCM
scenarios (Dehn et al., 2000), paleoenvironmental interpretation (Borgatti and Soldati, 2010), and observations (Haeberli
et al., 1993; Haeberli et al., 1997). Although GCMs appear to
be producing verifiable results, downscaled data still lack
sufficient resolution to predict changes in mass-movement
activity with any certainty. Nevertheless, observational results
are becoming increasingly indicative of climate change,
Figure 4 Rock avalanche deposits in the Ok Tedi River valley,
triggered by mining activity, August 1989. The landslide totaled 170
million tonnes and traveled 3.5 km downstream. Photo by M.J. Crozier.
whereas paleoenvironmental evidence, using a wide range of
proxy data, suggests a strong correlation between climatic
conditions and phases of mass-movement activity throughout
the Quaternary.
Author's personal copy
Mass-Movement Hazards and Risks
7.26.5
Concepts: Hazard, Risk, and Susceptibility
Mankind resides within and ultimately gains its livelihood
from the natural environment – a relationship that has never
been without risk. The concept of natural hazard recognizes
that some aspects of that natural environment can be
dangerous.
The concepts discussed below, although generic and applicable to all hazardous situations and processes, are discussed with reference largely to landslides, the most dangerous
form of mass movement.
might include not just the probability of a landslide occurring,
but also the probability that it could run out into the subdivision, the probability that it is large enough to impact the
building of concern (spatial probability), the number of
people occupying the building, constrained by a temporal
probability, usually reflecting the difference between daytime
and nighttime occupancy, and a vulnerability factor reflecting
the structural integrity of the building and commonly the
location of the occupant within the building.
7.26.5.3
7.26.5.1
Hazard
In natural hazard literature, two accepted meanings of the
term hazard occur. The first refers to an actual physical entity
(process or situation) that has the potential to cause damage
(e.g., a large rockslide or a long-runout debris flow). This is the
common nontechnical understanding of hazard. However,
this meaning of the term hazard is also used in some legal and
statutory documents, with statements such as notification to
local authorities: ‘‘to record the date and location of hazards,
including landslide, debris flow, surface flooding, subsidence,
etc.’’ The second meaning of the term hazard is more technical
and refers not to a process but rather to a threatening condition resulting from the activity (magnitude–frequency) of
that process, expressed as the probability of occurrence of a
damaging landslide (Crozier and Glade, 2005). Probabilities
are sometimes expressed as recurrence intervals in years but
for risk-assessment purposes they are generally given in the
form of annual probabilities. For example, for a landslide of
given magnitude with a recurrence interval of 20 years, the
reciprocal (1/20) represents the annual probability, that is,
0.05, expressed as a 5% chance of occurring in any year.
7.26.5.2
Risk
The consequences related to hazard occurrence can be great or
small, as well as direct or indirect; the latter linked to the
primary impact by a chain of dependent reactions that may be
manifest at some distance in time and space from the initial
occurrence. Clearly, the consequences depend on the context
in which they occur, the particular elements and attributes
affected, their value, and level of importance.
In simple generic terms, the important concept of risk can
thus be seen as having two components: the likelihood of
something adverse happening and the consequences if it
happens. Risk thus results from the intersection of hazard with
the value of the elements at risk by way of their vulnerability
(Crozier and Glade, 2005).
Varnes (1984) was one of the first to apply the generic
formulation of risk to the phenomenon landslides as
Hazard Elements at risk Vulnerability ¼ Risk
253
Susceptibility
Susceptibility is the ability or the propensity of an area to
produce landslides. With reference to triggering factors, susceptibility is sometimes construed as terrain sensitivity or
physical vulnerability. It can be determined by a range of
techniques, including geomorphic mapping, stability analysis,
or even expert opinion. Depending on the spatial resolution
of susceptibility assessment, the terrain may be ranked into
areas of greater or lesser susceptibility. Although susceptibility
can be determined independently for different types of landslides, it indicates very little about the magnitude and frequency,
that is, temporal probability of occurrence. Nevertheless, susceptibility assessments commonly represent an important first
step in hazard and risk assessment.
7.26.5.4
Vulnerability
The concept of vulnerability is widely debated, especially
among social scientists. Essentially, it is the degree of damage
that can be expected from a given magnitude of landslide.
Where definitions diverge, it depends largely on the element at
risk of concern. In the case of a building or fixed asset, vulnerability depends on the structural integrity of the feature, for
example, mud brick houses are much more vulnerable than
those constructed with reinforced concrete. In this sense,
vulnerability is often referred to as the damage ratio. However,
if the element of risk is more complex, for example, an individual or a community, then it is the integrity of social
structures as influenced by factors such as social networks,
demographics, awareness, and preparedness that may determine the level of vulnerability.
A related but different concept applied to individual and
communities is resilience. Whereas vulnerability refers to how
easily wounded an element may be, resilience refers to its
ability to bounce back or recover. Although many of the factors affecting community vulnerability also influence resilience, others specifically aid resilience such as robust transport
networks, access to finance, or medical facilities. Given these
influencing factors, it is not surprising that hazards have a
disproportionately high impact in less-developed countries.
(Hufschmidt and Glade, 2010).
½1
This fundamental hazard risk formulation can be further
modified for risk assessment in specific situations. For example, the risk of people being killed in a particular building
in a subdivision at the base of an unstable slope would need
to take into account a number of additional factors. These
7.26.6
Assessing Hazard and Risk
In simple terms, in order to make rational decisions with respect to living or carrying out activities in a hazardous area it is
essential to know ‘what’ might happen (nature of the hazard
Author's personal copy
254
Mass-Movement Hazards and Risks
faced), ‘where’ and ‘when’ (or at least how frequently) it might
happen, and ultimately what damage it might do – in short,
all the components of the risk. With this information, those
that have the options and resources (and many people in the
world do not), choices can be made and priorities set which
may involve, avoidance, education, the adoption of best
practice, or other mitigating measures such as warning systems
and engineering works.
7.26.6.1
7.26.6.1.1
The Nature of Mass-Movement Hazard
Dynamic or static
The character, intensity, and type of hazard have a major
bearing on the damaging capability of the process, methods of
assessment, and options for treatment. In terms of hazard
character, most mountain and steepland hazards are dynamic
rather than static. In other words, they are episodic and in
time and space their occurrence conforms to observed magnitude–frequency laws (Stark and Hovius, 2001; Malamud
et al., 2004). Static hazards, on the other hand, are represented
by conditions such as weak foundation material (e.g., bentonitic clays, peat deposits subject to shrinkage, sink holes, or
permafrost conditions). Static hazards generally cannot be
removed, may be mitigated by engineering design but are best
avoided. The calculation of risk posed by a static hazard has to
be approached differently from that outlined in eqn [1] and is
a much more tenuous undertaking, requiring estimates of the
probability of human encounter and the magnitude of effect
when that occurs. Encounter probabilities are also used in the
case when there is not a fixed population at risk and elements
of risk are predominantly dynamic. For example, encounter
probabilities employed in estimating the risk to road traffic
from snow avalanches or debris flow involve not only the
frequency of the physical event and the extent of road covered
but also factors such as traffic volume, speed, and stopping
distance (Crozier and Glade, 2010).
7.26.6.1.2
Site specific or nonsite specific
The other important aspect of hazard character is whether
hazards are site specific or nonsite specific. The difference
between these characteristics is easily appreciated in the case of
earthquakes. Fault rupture hazards are clearly site specific,
causing deformation at or near the seat of rupture, whereas
earthquake shaking is widespread and nonsite specific. In
terms of mass movement, creep and, to a certain extent, deepseated gravitational failure on a mountain slope scale are
widely distributed and nonsite specific, whereas debris flow
and lahar regularly occupy the same source and movement
tracks and are therefore site specific. In the case of multiple
occurrences, regional landslides events (Crozier, 2005), which
generally consist of hundreds to thousands of shallow slope
failures triggered by rainstorms or earthquakes, there is also an
element of nonspecificity in terms of location (Figure 5).
Whereas the specific occurrence site of these landslides might
be anticipated, the density of landslides and whether those
sites are activated are most closely correlated with intensity
regime of the triggering agent, for example, related to the
isohyetal pattern and location of the rainfall-producing weather cell. The vagaries of weather systems and the large tracts of
Figure 5 The multiple occurrence regional landslide event, triggered
by intense rainfall, 2002, Gisborne, New Zealand. Photo by M.J. Crozier.
land involved, in most cases, preclude the implementation of
expensive mitigating measures. Conversely, in the case of sitespecific hazards, several mitigation options are presented, including, avoidance, warning systems, and engineering works
(Bromhead, 2005).
7.26.6.1.3
Landslide type
The type of landslide also provides some indication of how
potentially damaging the landslide might be. Landslides have
been classified by many systems but the types represented in
the scheme developed by Cruden and Varnes (1996) are the
most widely accepted. Their scheme recognizes the mechanisms of, fall, topple, translational slide, rotational slide,
flow, lateral spread, and complex movements. These mechanisms are differentiated principally by the nature of the
material involved in the initial failure: earth, debris, and rock.
Landslide types are also broadly correlated with the velocity of
movement; falls, topples, and rock avalanches (Korup et al.,
2004) exhibit the highest velocities, whereas other mechanisms can display a wide range of velocities. This scheme also
recognizes that there can be different degrees of disruption of
the displaced mass. Given the parameters used in this classification, identification of landslide types can give an approximate indication of the destructive qualities or intensity of
landslides.
From the few records that are available (Alexander, 2005),
it appears that debris flows and mudflows account for the
most number of fatalities from landslides over short periods of
time, whereas over longer time frames individual large rockslides and rock avalanches can exact huge death tolls.
7.26.6.1.4
Intensity of landslides
Whereas hazard is generally defined as the frequency of a
given magnitude of event, this construction assumes that
magnitude, generally represented by the volume of the landslide, reflects the destructive qualities of the event. Although
this may be realistic, for example, for earthquake magnitude or
wind speed, it is not so clear in the case of mass-movement
hazards. In reality, the destructive qualities (or intensity) of a
landslide relate to a range of characteristics. These include
Author's personal copy
Mass-Movement Hazards and Risks
principally volume and velocity but importantly qualities such
as degree of disruption, extent of dislocation, depth of
movement, and run-out length. Regrettably, many of these key
intensity characteristics are not recorded in landslide inventories – consequently establishing return periods for intensity
is not always possible.
7.26.6.2
Where Landslides Occur: Susceptible Terrain
There are sufficient geomorphic indicators to suggest that with
respect to mass-movement hazards certain hot spots can be
identified in the landscape. The most obvious ones include
alpine couloirs and avalanche chutes, debris flow fans, fault
scarps, coastal cliffs, slopes recently exposed by glacial wasting,
and erosional reaches of rivers, especially outside bends of
rivers encroaching on steep topography. In terms of development, these sites should be treated with extreme caution.
These locations, together with the morphologic, sedimentologic, or dendrogeomorphic evidence of previous landslides,
can be included in geomorphic maps to provide an indication
of landslide susceptibility. The advent of light detection and
ranging (LiDAR) technology with its ability to see through the
vegetation cover greatly enhances this approach. With sufficient dating resolution, these techniques can also provide
information of magnitude and frequency and hence hazard.
Thus, comprehensive geomorphic maps represent the first
order of landslide susceptibility mapping.
Alternative approaches to mapping susceptibility may involve geotechnical slope stability analysis, where areal units,
from pixels to natural slope elements, can be analyzed in
terms of their factors of safety (Crozier and Glade, 2005)
pertaining to a range of climatic and hydrological conditions.
This approach requires data on slope geometry, slope material,
strength and stress, and hydrological or seismic conditions
likely to reduce the factors of safety. Such methods are
amenable to distributed slope modeling within the geographical information system (GIS) environment (van Westen,
2010) – the units once assessed then need to be aggregated and
classified to indicate varying degrees of susceptibility to mass
movement.
A more traditional approach to assessing susceptibility
is based on slope stability theory, and involves mapping
all those factors considered conducive to promoting mass
movement, such as rock type, rock structure, slope morphology, vegetation type, and hydrological conditions. In many
cases, experience allows certain factors to be weighted more
heavily than others. This approach has also been facilitated by
the development of GUSs whereby various parameter or thematic maps (layers) can be superimposed (intersected) to
define clusters of parameter combinations that can then be
ranked in terms of influence on susceptibility.
Landslide inventories provide an extremely important
empirical basis for susceptibility mapping and with sufficient
spatial reference can be used as indicators of susceptibility in
their own right. Maps of landslide distribution derived from
such inventories have also been combined with theoretical
parameter maps in order to discriminate those parameters that
are exclusively or predominantly associated with the presence
of landslides from those associated with stable terrain.
255
Discriminant analysis, logistic regression, hierarchical factor
analysis, and other statistical techniques are commonly used
to test the strength of relations before consequent susceptibility models are developed.
Landslide inventories as well as continued monitoring are
of particular value in validating susceptibility mapping. Stateof-the-art methodology on susceptibility mapping and zonation has been documented by Fell et al. (2008).
Essentially, however, susceptibility mapping only ranks the
terrain in terms of where certain landslides are able to occur.
Although this is an important first step, land and resource
managers need to know more than that – they need to have
these maps converted into indications of hazard and risk.
7.26.6.3
Frequency of Occurrence
To convert susceptibility maps into hazard maps, estimations
of magnitude and frequency of the expected events are required (Crozier, 1996; Crozier, 1999b). This represents one of
the most intractable problems in hazard and risk assessment,
especially with respect to reactivation of movement. Essentially, an extensive database is needed on the magnitude
and frequency of landslides in the area of concern, including
reactivations. Although some jurisdictions have databases
or historical written records that can assist, this is not the
case in most areas and relict landslide forms and deposits
need to be assessed and dated to establish a suitable record. In
many cases, this requires resort to proxy records from landslide deposits, for example, in lake sediments or from other
records of landslide impact on vegetation. However, as with
any historical data, application to future predictions needs to
take into account the changes that have occurred, or are likely
to occur in the environmental conditions conducive to landsliding. This consideration is particularly important today in
the face of unprecedented rates of global environmental
change.
An alternative approach to determining frequency requires
the identification of the magnitude of the triggering agent
required to initiate landslides and then the application of this
threshold to the long-term record of the triggering agent in
order to determine the frequency of threshold exceedance.
This method has been used extensively for rainfall-triggered
landslides (Glade et al., 2000) as well as for earthquaketriggered landslides (Wilson and Keefer, 1985; Hancox et al.,
1997).
Much of the above discussion is predicated on there being
a presence of landslides or evidence of former landslides in the
area under concern, and that hazard, as defined here, can only
be assessed with resort to magnitude and frequency derived
from landslide evidence. However, the concept of first-time
failure is an important consideration, particularly where there
are critical elements of risk exposed. The development of
reservoirs is a case in point. Even though there may be no
evidence of former landslides in the vicinity, changes in the
slope hydrological regime associated with dam construction
should address the possibility that landslides may be initiated.
In this case, susceptibility is generally established through
geotechnical stability analysis, and probabilities are thereby
assigned on the basis of factors of safety, and the natural
variability in parameters used in the stability assessment.
Author's personal copy
256
Mass-Movement Hazards and Risks
7.26.6.4
Risk
In a comprehensive study of world landslide hazards carried
out between 1993 and 2002, Alexander (2005) recorded over
40 000 deaths from landslides alone, and of these the majority
of deaths was caused by debris flows and mudflows resulting
from largely intense and prolonged rainfall. His data set recorded a much lower death toll from slumps, debris avalanches, and rock falls, and only few were attributed to
triggering by earthquakes and volcanic activity.
However, longer records indicate that massive death tolls
can arise from a range of different sources. For example, the
earthquake-triggered Huascaran rock and ice avalanche in
Peru of 31 May 1970 killed 2000 people in the village of
Ranranhirca and 19 000 in the town of Yungay (Browning,
1973). Similarly, an earthquake in 1920 in Kansu, China,
caused landslides in loess that resulted in 10 000 deaths. In
another instance, intense rainfall produced debris flows and
floods on 16 December 1999 in Venezuela, resulting in a
death toll of 30 000. More recently, the Szechwan earthquake
of 12 May 2008 produced hundreds of landslides, many creating temporary dams, and resulted in 75 000 deaths. These
events are essentially the realization of risk (Figure 6).
Ultimately, risk is determined, as outlined earlier, through
determining the probability of loss based on the value of
elements at risk, their vulnerability, and exposure to hazard.
Although mass-movement hazards may be one aspect generating risk to an individual or community, they are often accompanied simultaneously by other hazards such as
earthquake shaking or flooding. Risk assessments need to
factor in the probability of multiple hazard occurrences so that
emergency management and risk-reduction planning can respond accordingly. The nature of that response, however, is
dependent not only on how those risks are evaluated but also
on the ability and resources available to the community.
7.26.7
Conclusion
Hazard and risk are not static. Their assessments in any region
are subject to becoming redundant with time. Consider
the causative factors of landslides. These include preconditions, preparatory factors, and triggering factors. Preconditions such as stratigraphic disposition, slope gradient,
and slope material are generally considered constant. Nevertheless, slope conditions and material are constantly being
modified by urban development, mining, and quarrying.
Preparatory factors relate to changes that lower the stability of
the slope, such as deforestation or groundwater changes
alongside reservoirs. These are increasingly being changed
by accelerating demand for resources. Triggering factors that
initiate movement on unstable slopes are primarily the result
of rainfall, earthquakes, and slope under-cutting. Global
environmental changes are likely to enhance both rainfall
intensity and slope modification. The bottom line is that
historical assessments of hazard are subject to dramatic
changes in the future. The capability to predict, cope with, and
mitigate such changes requires the geomorphologist to realize
and anticipate the rates and impact of climate and social
change.
References
Figure 6 The Abbotsford landslide that destroyed 69 houses,
8 August 1979, Dunedin, New Zealand. Photo by W. Brockie.
Alcántara-Ayala, I., Goudie, A. (Eds.), 2010. Geomorphological Hazards and Disaster
Prevention. Cambridge University Press, Cambridge, 291 pp.
Alexander, D., 2005. Vulnerability to landslides. In: Gade, T., Anderson, M., Crozier,
M.J. (Eds.), Landslide Hazard and Risk. Wiley, Chichester, pp. 175–198.
Borgatti, L., Soldati, M., 2010. Landslides and climatic change. In: Alcántara-Ayala,
I., Goudie, A. (Eds.), Geomorphological Hazards and Disaster Prevention.
Cambridge University Press, Cambridge, pp. 87–96.
Bromhead, E.N., 2005. Geotechnical structures for landslide risk reduction.
In: Glade, T., Anderson, M., Crozier, M.J. (Eds.), Landslide Hazard and Risk.
Wiley, Chichester, pp. 549–594.
Browning, J.M., 1973. Catastrophic rock slide, Mount Huasacaran, north-cenral
Peru, May 31, 1970. Bulletin of the American Association of Petroleum
Geologists 57, 1335–1341.
Cendrero, A., Remondo, J., Bonachea, J., Rivas, V., Soto, J., 2006. Sensitivity of
landscape evolution and geomorphic processes to direct and indirect human
influence. Geographia Fisica e Dinamica Quaternaria 29, 125–137.
Clague, J.J., 2009. Climate change and slope stability. In: Sassa, K., Canuti, P.
(Eds.), Landslides – Disaster Risk Reduction. Springer, Berlin, pp. 557–572.
Cooper, A., Norris, R., 2008. The great divide. In: Graham, I.J. (Ed.), A Continent
on the Move: New Zealand Geosciences into the 21st Century. The Geological
Society of New Zealand, Wellington, pp. 118–122.
Crozier, M.J., 1996. Magnitude/frequency issues in landslide hazard assessment.
In: Maeusbacher, R., Schulte, A. (Eds.), Beitrage zur Physiogeographie. Barsch
Festschrift, Heidelberger Geographische Arbeiten. Heft, 104, pp. 221–236.
Crozier, M.J., 1999a. Landslides. In: Alexander, D.E., Fairbridge, R.W. (Eds.),
Encyclopedia of Environmental Science. Kluwer, Dordrecht, pp. 371–374.
Author's personal copy
Mass-Movement Hazards and Risks
Crozier, M.J., 1999b. Frequency and magnitude of geomorphic processes.
In: Crozier, M.J., Maeusbacher, R. (Eds.), Magnitude and Frequency in
Geomorphology, Zeitschrift fuer Geomorphologie Supplementband, vol. 115, pp.
35–50.
Crozier, M.J., 2005. Multiple-occurrence regional landslide events: hazard
management perspectives. Landslides 2(4), 245–256.
Crozier, M.J., 2010. Deciphering the effect of climate change on landslide activity.
Geomorphology, doi: 10.10.16/j.geomorph.2010.04.009.
Crozier, M.J., Glade, T., 2005. Landslide hazard and risk: issues, concepts and
approach. In: Glade, T., Anderson, M., Crozier, M.J. (Eds.), Landslide Hazard
and Risk. Wiley, Chichester, pp. 1–40.
Crozier, M.J., Glade, T., 2010. Hazard assessment for risk analysis and risk
management. In: Alcántara-Ayala, I., Goudie, A. (Eds.), Geomorphological
Hazards and Disaster Prevention. Cambridge University Press, Cambridge, pp.
221–232.
Cruden, D.M., Varnes, D.J., 1996. Landslide types and processes. In: Turner, R.K.,
Schuster, R.L. (Eds.), Landslides: Investigation and Mitigation. Special Report
247. US Transportation Research Board, National Academy Press, Washington,
DC, pp. 36–75.
Davis, W.M., 1899. The geographical cycle. Geographical Journal 14, 481–504.
(Also in: Johnson, D.W. (Ed.), 1954. Geographical Essays, Dover, New York, NY,
pp. 249–278).
Dehn, M., Burger, G., Buma, J., Gasparetto, P., 2000. Impact of climate change
on slope stability using expanded downscaling. Engineering Geology 55,
193–204.
Eisbacher, G.H., Clague, J.J., 1984. Destructive mass movements in high mountains:
hazards and management. Geological Survey of Canada, Paper 84-16.
Evans, S.G., DeGraff, S.G. (Eds.), 2002. Catastrophic Landslides: Effects,
Occurrence, and Mechanisms. Reviews in Engineering Geology. Geological
Society of America, Boulder, CO, vol. XV.
Fell, R., Corominas, J., Bonnard, C., Cascini, L., Leroi, E., Savage, W.Z., 2008.
Guidelines for landslide susceptibility, hazard and risk zoning for land use
planning. Engineering Geology 102, 85–98.
Gilbert, J.K., 1877. Geology of the Henry Mountains (Utah); US Geographic and
Geological Survey of the Rocky Mountain Region. U.S. Govt. Printing Office,
Washington, DC.
Glade, T., Crozier, M.J., 2005. A review of scale dependency in landslide hazard
and risk analysis. In: Glade, T., Anderson, M., Crozier, M.J. (Eds.), Landslide
Hazard and Risk. Wiley, Chichester, pp. 75–138.
Glade, T., Crozier, M.J., Smith, P., 2000. Establishing landslide-triggering rainfall
thresholds using an empirical antecedent daily rainfall model. Journal of Pure
and Applied Geophysics 157, 1059–1079.
Guzzetti, F., Peruccacci, S., Rossi, M., Stark, C.P., 2008. The rainfall intensityduration control of shallow landslides and debris flows: an update. Landslides
5(1), 3–18.
Hack, J.T., 1960. Interpretation of erosional topography in humid temperate regions.
American Journal of Science 258A, 80–97.
Haeberli, W., Guodong, C., Gorbnov, A.P., Harris, S.A., 1993. Mountain permafrost
and climate change. Periglacial Processes 4, 165–174.
Haeberli, W., Wegmann, M., Vonder Mühll, D., 1997. Slope stability problems
related to glacier shrinkage and permafrost degradation in the Alps. Ecolgae
Geologica Helvetica 90, 407–414.
Hancox, G.T., Perrin, N.D., 2009. Green lake landslide and other giant and very
large postglacial landslides in Fiordland, New Zealand. Quaternary Science
Reviews 28, 1020–1036.
257
Hancox, G.T., Perrin, N.D., Dellow, G.D., 1997. Earthquake-Induced Landsliding in
New Zealand and Implications for MM Intensity and Seismic Hazard
Assessment. Institute of Geological and Nuclear Sciences, Wellington.
Hufschmidt, G., Glade, T., 2010. Vulnerability analysis in geomorphic risk
assessment. In: Alcántara-Ayala, I., Goudie, A. (Eds.), Geomorphological Hazards
and Disaster Prevention. Cambridge University Press, Cambridge, pp. 243–333.
Huggel, C., 2009. Recent extreme slope failures in glacial environments: effects of
thermal perturbation. Quaternary Science Reviews 28, 1119–1130.
Keefer, D.K., 1984. Landslides caused by earthquakes. Geological Society of
America Bulletin 95(4), 406–421.
Keefer, D.K., Wilson, R.C., 1989. Predicting earthquake-induced landslides with
emphasis on arid and semi-arid environments. Publication of the Inland
Geological Society 2, 118–149.
King, P.R., 2008. Shifting ground: plate tectonic evolution of Zealandia. In: Graham,
I.J. (Ed.), A Continent on the Move: New Zealand Geosciences into the 21st
Century. The Geological Society of New Zealand, Wellington, pp. 123–127.
Korup, O., Clague, J.J., 2009. Natural hazards, extreme events and mountain
topography. Quaternary Science Reviews 28, 977–990.
Korup, O., Densmore, F., Schlunegger, F., 2010. The role of landslides in
mountain range evolution. In: Glade, T., Crozier, M.J. (Eds.), Landslide
Geomorphology in a Changing Environment. Geomorphology (Special Issue),
120, pp. 77–90.
Korup, O., McSaveney, M.J., Davies, T.R.H., 2004. Sediment generation and delivery
from large historic landslides in the Southern Alps, New Zealand.
Geomorphology 61, 189–207.
Malamud, B.D., Turcotte, D.L., Guzzetti, F., Reichenbach, P., 2004. Landslide
inventories and their statistical properties. Earth Surface Processes and
Landforms 29, 687–711.
Marston, R.A., 2008. Land, life and environmental change in mountains. Annals of
the Association of American Geographers 98, 507–520.
Penck, W., 1953. Morphological Analysis of Landforms. Translated by Czech, H.
and Boswell, K.C. Macmillan, London.
Remondo, J., Soto, J., González-Dı́ez, A., Dı́az De Terán, J.R., Cendrero, A., 2005.
Human impact of geomorphic processes and hazards in mountain areas in
northern Spain. Geomorphology 66, 69–84.
Slaymaker, O., 2010. Mountain hazards. In: Alcántara-Ayala, I., Goudie, A. (Eds.),
Geomorphological Hazards and Disaster Prevention. Cambridge University Press,
Cambridge, pp. 33–47.
Stark, C.P., Hovius, N., 2001. The characterisation of landslide size distributions.
Geophysical Research Letters 28, 1091–1094.
Starkel, L., Sarkar, S., 2002. Different frequency of threshold rainfalls transforming
the margin of Sikkimese and Bhutanese Himalaya. Studia Geomorphologica
Carpatho-Balcanica 36, 51–67.
Varnes, D.J, 1984. Landslide Hazard Zonation: A Review of Principles and Practice.
UNESCO, Paris.
van Westen, C.J., 2010. GIS for the assessment of risk from geomorphological
hazards. In: Alcántara-Ayala, I., Goudie, A. (Eds.), Geomorphological Hazards
and Disaster Prevention. Cambridge University Press, Cambridge, pp. 205–219.
Wilson, R.C., Keefer, D.K., 1985. Predicting areal limits of earthquake-induced
landsliding. In: Ziony, J.I. (Ed.), Evaluating Earthquake Hazards in the Los
Angeles Region. United States Geological Survey Professional Paper 1360,
Washington, DC, pp. 317–493.
Wright, R., 2004. A Short History of Progress. Text Publishing, Melbourne, 211 pp.
Author's personal copy
258
Mass-Movement Hazards and Risks
Biographical Sketch
Michael Crozier is Professor Emeritus at Victoria University of Wellington where he held a personal chair in
geomorphology from 1998 to 2009 in the School of Geography, Environment and Earth Sciences. He is president
of the International Association of Geomorphologists and immediate past president of the New Zealand Geographical Society. He has received a number of Academic awards including a Fulbright Scholarship to USA, a
Leverhulme fellowship to University of Bristol, and a distinguished visiting professorship at the University of
Durham. Professor Crozier has been a member of the New Zealand Conservation Authority with responsibility
for National Parks. He is a geomorphologist with research interests in landslides, natural hazards, and human
impact on natural systems.