Réf. Huggel & al. 2012 - A

Référence bibliographique complète

HUGGEL, C., ALLEN, S., DELINE, P., FISCHER, L., NOETZLI, J., RAVANEL, L. 2012. Ice thawing, mountains falling—are alpine rock slope failures increasing? Geology Today, Vol. 28, No. 3, 98–104. DOI

Abstract: Many high-mountain environments of the world have seen dramatic changes in the past years and decades. Glaciers are retreating and downwasting, often at a dramatically fast pace, leaving large amounts of potentially unstable debris, moraines and rock slopes behind. Although in the main invisible to the eye of an observer, permafrost, i.e. rock and debris with permanent zero or subzero temperatures, is thawing. Several slopes have become unstable and landslides potentially related to permafrost degradation have received wide-ranging attention from both scientists and the media. A number of those landslides can be related to the effects of recent changes in the cryosphere, which are ultimately driven by changes in climatic parameters, in particular temperature and precipitation.



Organismes / Contact

• Department of Geography, University of Zurich, Switzerland
• Climate and Environmental Physics, Physics Institute, University of Bern, Switzerland
• Laboratoire EDYTEM, Université de Savoie, CNRS, Le Bourget-du-Lac, France
• Geological Survey of Norway (NGU), Trondheim, Norway

Provision of temperature data for Switzerland by MeteoSwiss is acknowledged. Part of the studies described here was funded by the Swiss National Science Foundation.

(1) - Paramètre(s) atmosphérique(s) modifié(s)
(2) - Elément(s) du milieu impacté(s)
(3) - Type(s) d'aléa impacté(s)
(3) - Sous-type(s) d'aléa
Temperature, precipitation Permafrost Landslides Rockslides, rockfalls

Pays / Zone
Massif / Secteur
Site(s) d'étude
Période(s) d'observation
- World high mountains
- Swiss Alps
- French Alps
- High mountain regions of Switzerland and border regions with Italy and France
- Aiguilles de Chamonix area, Mont Blanc massif

(1) - Modifications des paramètres atmosphériques








Informations complémentaires (données utilisées, méthode, scénarios, etc.)


(2) - Effets du changement climatique sur le milieu naturel








Sensibilité du milieu à des paramètres climatiques
Informations complémentaires (données utilisées, méthode, scénarios, etc.)



(3) - Effets du changement climatique sur l'aléa



In view of the link between climate parameters and slope stability, concern is increasing that current and future climate change could adversely affect related landslide hazards. Here we will first depart from a basic magnitude–frequency concept and in the context of landscape models, review recent assessments of climate change in terms of temperature and precipitation, and then examine whether we can find changes in the activity of rock slope failures in alpine regions over the past decades and century. Eventually we will look at the role of extreme temperature events for slope stability in high mountains.

As with many other hazards, mass movements occurring in high-mountain regions have characteristic magnitude–frequency relationships. [Figure 1 specifies several freeze and thaw processes in high mountains and indicates their characteristic frequency and magnitude.] In line with existing empirical and theoretical concepts, the frequency–magnitude relationship is non-linear and typically follows a power-law. Large rock and ice avalanches form the upper end of a magnitude continuum starting with block and rock falls. Consequently, these are extreme events of rare occurrence but potentially enormous physical and human impact.

Changes in the magnitude or frequency of such extreme landslides can have a tremendous effect in terms of the hazard they represent. Their frequency of occurrence can be modified either by gradual degradation of slope stability, or by changes in trigger conditions. Typically, precipitation or earthquakes are recognized as triggers of slope failures, whereas geology (i.e. lithology, faulting systems, tectonics, etc.) and topography are considered as factors determining the disposition of a slope to failure. Lithology, faulting and discontinuity systems are static over large time scales and can be considered as constant in the context of our topic. Other components, such as the properties or kinematics of fractures, can change over much shorter time scales. In addition, topography is usually considered as a slowly changing, basic landscape element, e.g., in terms of hillslope adjustment to erosive processes such as fluvial bedrock incision, on rather larger time scales of thousands of years or more. Large-scale climate variations can therefore have an important influence on topography. Studies in tectonically active mountain systems such as Alaska, Karakorum or the Andes have also suggested a close dynamic coupling between climate, tectonics and topography, over similarly large time scales. Even though the strength of coupling and feedback between climate, tectonics and topography is still under intensive discussion, it is without doubt that landscape evolution research over the past twenty years has made essential contributions to questions on the interplay of climate, geology, topography and slope stability.

Climate changes can influence slope stability, both in terms of disposition and trigger conditions. For instance, long-term climate change (over century to millennia) can result in increased erosion and in a changing topographic disposition. Similarly, long-term changes in subsurface thermal regimes can alter the disposition of a rock slope to failure. On the other hand, short-term changes in heavy precipitation can modify the trigger conditions for slope failures.

Shorter time intervals

We will now primarily address those shorter time periods (i.e. decades to centuries) of climatic changes with an emphasis on trigger factors of slope failures, that are arguably more relevant for current and future hazards in high mountains. An assessment of recent and contemporary climate change, such as that performed by the Intergovernmental Panel on Climate Change (IPCC), provide evidence that air temperature is rising at a scale and rate without historical precedence, amounting to a global average of about 0.8 °C air temperature increase over pre-industrial levels, yet with much higher values for some regions, including many mountain regions. Not only are average air temperatures on the rise but also their extremes, which are commonly expressed as the 90th or 95th percentile of the long-term record. Although there is a lack of studies specifically focusing on trends of these extremes in high-mountain regions, it has been robustly demonstrated that the frequency of hot days and heat waves have increased over most land areas since the mid-twentieth century. The IPCC Fourth Assessment Report (AR4 published in 2007), as well as several post-AR4 studies, also showed that increasing trends in precipitation extremes have been observed in many parts of the world for the past decades. However, changes vary seasonally and spatially in different regions of the world, and high-mountain areas mostly lack detailed information on heavy rainfall. In some mountain regions of eastern and southern Europe, both increasing and decreasing trends of extreme precipitation have been found, while in parts of central South America an increase of precipitation extremes for the second half of the twentieth century has been observed. Recent studies also confirmed increasing trends of extreme rainfall in many regions of North America and Asia, yet not in a consistent way across these continents.

Detection and attribution

In this context one of the questions often asked, is whether the corresponding changes in mass movement and landslide activity in mountain regions can be identified over the past years and decades, and if so, whether this trend can be attributed to the effects of climate change. In climate sciences, as defined in the IPCC Assessment Reports, ’detection’ is the process of demonstrating that climate has changed in some defined statistical sense, without providing a reason for that change, whereas ‘attribution’ establishes the most likely causes for the detected change with some defined level of confidence. While detection and attribution are well established in physical climate sciences they are much less well developed in the field of climate impacts, including landslide research. In climate sciences different methods of attribution to climate change are distinguished, and include single-step and multi-step attributions that explicitly model the response of a system or variables to external forcing and drivers, such as increased greenhouse gas emissions and their related air temperature increases. Single- and multi-step attribution studies on the level of climate change impacts are difficult to achieve due to a still-incomplete understanding of how environmental systems respond to climate change and the many confounding factors that complicate this response. However, some studies have recently advanced formal attribution methods for flood risks in the UK. More common in the field of climate change impacts is an associative pattern approach, where spatial patterns of observed impacts are compared with observed climate trends using statistics on large numbers of data series.

In landslide research the first detection and attribution studies are just about to emerge. A traditionally strong focus on geological and geomorphological aspects of landslides, lag effects with respect to contemporary climate change (e.g. debuttressing effects, or lagged changes in the subsurface thermal regime), and a general difficulty of detecting changes in landslide occurrence and its relation to climate change, may be reasons that have somewhat prevented an earlier onset of research in this field. A recent study has thereby outlined a conceptual model as to how detection and attribution to climate change could be approached in landslide research, and distinguished five methods: (1) analysis of event inventories, (2) damage and loss data, (3) case studies, (4) identification of causative and trigger factors, and (5) development of process models simulating climate change impact chains.

Anecdotal evidence for a change in the frequency of large rock slope failures comes from several high-mountain regions of the world. In fact, recent years have seen a number of remarkable slope failures, with several events being among the largest rapid mass movements that have been observed during the past few centuries. Noteworthy are several rock and ice avalanches that occurred in the twenty-first century in Alaska and Western Canada with volumes of between 10 and 50 million cubic metres. In 2002, for instance, an ice-rock avalanche in the Russian Caucasus attracted much attention due to its enormous size (>100 million m3), and the path of complete destruction it left behind over a distance of more than 30 km. Similarly, in the European Alps a considerable number of large slope failures occurred in the vicinity of rock and glacier ice in the 1990s and 2000s, although their dimensions did not quite reach that of their gigantic counterparts in other parts of the world.

Common to these slope failures is that they all occurred in environments shaped by the cryosphere, i.e. glaciers, permafrost and snow. The cryosphere components in mountain regions, and of these most visibly the glaciers, have undergone tremendous changes due to global warming over the past few decades. Such changes can affect topography (downwasting and retreat of glaciers) but also cause changes to the thermal regime of frozen rock and debris at and below the surface. Due to the thermal sensitivity of ice, temperature changes play a more important role for slope stability in high-mountain environments than in mountain regions without perennial surface and subsurface ice. Permafrost degradation can affect the stability of steep alpine rock walls through a variety of thermo-mechanical processes of which not yet all are understood in detail. For example, the degree of fracturing and discontinuity systems influence infiltration and hydrostatic pressure which results from the vertical height of the interconnected saturated zone. In frozen rock slopes in high-mountain regions infiltration from melting of snow and ice or liquid precipitation during warm periods can affect slope deformation and stability. Recent measurements in the Swiss Alps for instance showed that infiltration from snow melting into frozen rock fractures represents an effective transport of latent heat that can result in permafrost thaw processes and eventually in increased slope deformation. However, each slope has an individual combination and history of factors determining slope stability, and their gradual or sudden change ultimately determine a slope failure event.

In the following we present results for one of the five approaches for detection of changes in rock falls, as listed above, by looking at different published studies based on event inventories of rock slope failures in the Alps over the twentieth and early twenty-first centuries. Thereby, we analyse studies on two spatial scales (i.e. local and regional) and evaluate whether there exist trends in these data series, and if so, whether these trends are consistent over the different spatial scales.

Regional case studies

The regional study is based on an inventory of 52 rock slope failures that occurred between 1900 and 2010 in the high mountain regions of Switzerland and border regions with Italy and France. The data set has been developed and extended over the few past years and is based on scientific publications, newspaper articles, field observations and personal comments and observations, and is now being maintained by the Swiss Permafrost Monitoring Network (PERMOS). Only significant rock slope failures with volumes larger than about 1000 m3 and with failure zone higher than 2000 m asl (i.e., above the tree line) have been considered for this study (the PERMOS inventory includes in total > 150 events, but here only events are used for which volume estimates could be made). The data set used here certainly represents a minimum number of rock slope failures, as not all rock slope failures that occurred in nature were observed and documented. This is especially true for earlier periods of the twentieth century and more remote areas of the Alps. The recently increasing interest of both science and the media on the relation of climate change and slope stability, and more advanced monitoring technology, has led to increased documentation activity. (...)

Although the sample of rock slope failures is relatively small—an inherent limitations in all such studies—some trends in the occurrence of events over the twentieth and twenty-first centuries can be distinguished. First of all, a strong increase in the frequency of events was documented from the 1980s and 1990s to the present day. It is without doubt that this trend can partly be attributed to an increased level of documentation during more recent years, as compared to earlier periods. However, it is interesting that the trend of increasing rock slope failures also holds true if we consider only large events (> 100 000 m3). The latter are assumed to show a more consistent level of documentation over the period of observation, and therefore are less documentation- biased than small-volume failures. Furthermore, the analysis shows that an earlier peak of increased rock slope failure activity occurred around the 1940s. This is noteworthy because this period was characterized by higher air temperatures.

If we analyse the rock types of the slope failures of the inventory data used here we see that 40 per cent of the slope failure zones lie in granite or diorite, 30 per cent in gneiss and 20 per cent in limestone, while a minor proportion is located in mafic metamorphic rocks and schists. Interestingly, the slope failures are not equally distributed over altitude and rock type. Although 90 per cent of the study area lies between 2000 and 3000 m asl, only 40 per cent of the failures occurred from this elevation range, and 60 per cent of the failures occurred at elevations higher than 3000 m asl. Similarly, the failures do not correspond to the distribution of rock type classes. For instance proportionally more events were observed in limestone and granite, and proportionally fewer in gneiss. It is furthermore interesting to note that the largest rock slope failures occurred in limestone, followed by gneiss, while most small failures were documented from areas with granite. None of the failures had its origin in conglomerate, which generally also does not exist at elevations > 3000 m asl. Since only slope failures at altitudes of 2000 m asl and higher were considered for this data set, it is not surprising that many events had their origin from areas with permafrost.

While about one-tenth of the events are located outside permafrost areas, around one-quarter originated from areas of continuous and cold permafrost. The majority of the events failed from transition areas of frozen to non-frozen bedrock. Although there exist considerable uncertainties related to the distribution of permafrost resulting from local climatic or three-dimensional heat flow effects, it is interesting to note that the clustering of slope failures in the transition zone corresponds to theoretical considerations backed by lab experiments and theoretical considerations that regard areas of warm permafrost (with temperatures just below 0 °C) as particularly prone to an alteration in the conditions of slope stability.

Local case study

The local study stems from the Mont Blanc massif, and more specifically the Aiguilles de Chamonix, a number of rocky peaks that extend between Mer de Glace and Glacier des Bossons at elevations between about 3300 and 3800 m asl. Thanks to the long mountaineering tradition in Chamonix, at the foot of these mountains, a rich archive of historical photographs exists that goes back 150 years from present and based on which a detailed local inventory of rock fall events has been established. Sources of the photographs include albums from mountain guides, postcards, private collections and museum collections. Only photographs of adequate quality were used, and only rock slope failures > 500 m3 were considered in order to reduce systematic errors of documentation. Additional oral and written sources were used to support dating of the rock slope failures identified in the photographs. To estimate the failure volumes a terrestrial laser scanning device was used, supported by analysis of detailed digital elevation models. As with most volume estimates, the error range may be quite high and amount to 10–50 per cent. Forty-nine rock slope failures with volumes of up to 65 000 m3 have been documented.

Several aspects of the analysis of the rock slope failures at the Aiguilles de Chamonix over the twentieth century and into the first decade of the twentyfirst century are remarkable. First, a strong increase in the frequency of events can be identified in the 1990s and the 2000s. A second peak of increased activity is observed for the 1940s, just as in case of the regional rock slope failure inventory. A study using a similar methodology in the west face of the Drus, located further north on the other side of the Mer de Glace, also confirmed this trend. The analysis of failure volumes furthermore reveals that the frequency of events during the past two decades has increased over the entire range of volumes as compared to the period prior to 1990.

The local Chamonix study has a great advantage over the regional study in that it reasonably avoids the documentation bias which is inherent in many event and disaster inventories that span periods of several decades, stemming from a generally increased level of documentation in more recent times. The use of photographs allows for a quasi-objective representation of rock slope failure events throughout the twentieth century that is otherwise very difficult to achieve.

A fundamental conclusion of the analysis of the local and regional studies is that they are consistent in terms of observed trends of rock slope failure activity over the past ~100 years. This is remarkable because the data sources, collection and analysis methods are different and completely independent. Consistency in terms of geographic location is maintained between the two study areas as they are part of the same Alpine region. Because the local study represents a quasi-complete inventory and confirms the signal of change found for the regional level, the evidence for an actual trend of increasing rock slope failures can be regarded as more robust for the central Alps.

Studies such as presented here are still very rare, maybe due to the fact that the documentation of highmountain rock slope failures in other less populated high-mountain regions of the world is limited. However, a study in western Canada analysed 38 large landslides in rock and soil and found an increase from 1.3 to 2.3 landslides per year over 30 years, although the relation to potential effects of climate change over that period has not yet been clarified.

In New Zealand, an inventory of over 100 late Quaternary large rock avalanches and rockfalls has been recorded in the central region of the Southern Alps, including 22 failures observed since the mid-twentieth century. This inventory is thought to be incomplete and hence, despite an apparent cluster of slope failures observed over the past two decades, related studies have been unable to conclude with any confidence regarding longer term changes in landslide frequency and magnitude over time. Nonetheless, two robust conclusions have been drawn from this region, which parallel findings from the European Alps regarding slope stability and cryospheric change: (1) Recent rock failures (i.e. since the mid-twentieth century) have predominated from slopes located within the transitional or warm zone of permafrost, and (2) there is a strong association of recent slope failures from slopes that have been directly influenced by glacial recession over the past 100–150 years.





Paramètre de l'aléa
Sensibilité des paramètres de l'aléa à des paramètres climatiques
Informations complémentaires (données utilisées, méthode, scénarios, etc.)




(4) - Remarques générales


(5) - Syntèses et préconisations

Research perspectives

In combination, the evidence accumulating from geographically distinct high mountain regions across the globe points strongly towards the role of global warming, related thawing and the recession of surface and subsurface ice, and slope stability. For the Alps we have seen that an increased frequency of rock slope failures in the 1940s and especially since the 1990s coincides with higher mean annual air temperatures (MAAT). We have outlined the difficulties of interpreting slope failure inventories and we therefore should consider the apparent relationship with MAAT with caution based on currently available data. However, while the signal of warming and higher slope failure activity is relatively weak for the 1940s it is much stronger for both variables for the period since the 1990s. The physical processes that are likely to be responsible for more landslides under higher air temperatures are not yet entirely clear though. Glacier retreat and debuttressing can have an effect on the decade/century time scale. A greater production of melt water with infiltration into rock joints can be another effect of elevated mean air temperatures, going back to the 1940s. It is still not understood since when over the past century and to what degree, permafrost degradation started to play a more important role in the change of slope stability. Field observations would point to the 1990s and 2000s. Furthermore, over longer time scales it is interesting to note that a number of recent studies in the Alps and Himalayas have found evidence of higher landslide (debris flows and large rock slides) frequency during warmer and wetter periods of the Holocene.

In addition to the influences of longer-term changes in mean air temperatures, it was recently found that many large slope failures in the 1990s and 2000s in the high-mountain regions of the Alps, Alaska and western Canada, and New Zealand were preceded by periods of extremely high maximum temperatures in the days to weeks before slope failure. Field observations and new measurements of rock slope deformation suggest that the triggering effect of warm and hot periods is primarily through the rapid generation of significant amounts of melt water from snow, glacier ice and permafrost. In the Alps it is projected that such warm periods with air temperatures several degrees above freezing up to elevations of 3500–4000 m asl will occur 1.5–4-fold more frequently by 2050 than during 1950–2000. Although clearly more research is needed in this field it is likely that increasing mean and maximum air temperatures, with further strong glacier retreat and permafrost degradation will be paralleled by an increase in slope failures in many high-mountain regions of the world. Some of these slope failures may impact transportation, energy or tourism infrastructure and in some cases population centres. Most critical are situations where rock slope failures transform into highly mobile debris flows or generate flood waves by impacting natural or artificial lakes. Under these conditions much longer distances of potential destruction must be anticipated. Regular monitoring, early warning systems, structural interventions and other risk reduction measures should be considered to prevent damage to life and property.

Références citées :

Allen, S.K., Cox, S.C. & Owens, I.F. 2010. Rock avalanches and other landslides in the central Southern Alps of New Zealand: a regional study considering possible climate change impacts. Landslides, v.8, pp.33–48.

Bookhagen, B., Thiede, R.C. & Strecker, M.R. 2005. Late Quaternary intensified monsoon phases control landscape evolution in the northwest Himalaya. Geology, v.33, pp.149–152.

Borgatti, L. & Soldati, M. 2010. Landslides as a geomorphological proxy for climate change: a record from the Dolomites (northern Italy). Geomorphology, v.120, pp.56–64.

Deline, P. 2009. Interactions between rock avalanches and glaciers in the Mont Blanc massif during the late Holocene. Quaternary Science Reviews, v.28, pp.1070–1083.

Fischer, L., Purves, R.S., Huggel, C., Noetzli, J. & Haeberli, W. 2012. On the influence of topographic, geological and cryospheric factors on rock avalanches and rockfalls in high-mountain areas. Natural Hazards Earth System Science, v.12, pp.241–254.

Gruber, S. & Haeberli, W. 2007 Permafrost in steep bedrock slopes and its temperature-related destabilization following climate change. Journal of Geophysical Research, v.112, p.F02S18.

Huggel, C., Allen, S., Clague, J.J., Fischer, L., Korup, O. & Schneider, D. 2011. Detecting potential climate signals in large slope failures in cold mountain regions. Proceedings of the Second World Landslide Forum, 3–7 October 2011, Rome.

Huggel, C., Clague, J.J. & Korup, O. 2012. Is climate change responsible for changing landslide activity in high mountains? Earth Surface Processes and Landforms, v.37, pp.77–91.

Huggel, C., Salzmann, N., Allen, S.K., Caplan-Auerbach, J., Fischer, L., Haeberli, W., Larsen, C., Schneider, D. & Wessels, R. 2010. Recent and future warm extreme events and high-mountain slope stability. Philosophical Transactions of the Royal Society A, v.368, pp.2435–2459.

IPCC (Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M. and Miller, H.L., eds.) 2007 Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge & New York.

Montgomery, D.R., Balco, G. & Willett, S.D. 2001. Climate, tectonics, and the morphology of the Andes. Geology, v.29, pp.579–582.

Noetzli, J., Gruber, S., Kohl, T., Salzmann, N. & Haeberli, W. 2007. Three-dimensional distribution and evolution of permafrost temperatures in idealized high-mountain topography. Journal of Geophysical Research, v.112, p.F02S13.

Pall, P., Aina, T., Stone, D.A., Stott, P.A., Nozawa, T., Hilberts, A.G.J., Lohmann, D. & Allen, M.R. 2011. Anthropogenic greenhouse gas contribution to flood risk in England and Wales in autumn 2000. Nature, v.470, pp.382–385.

Ravanel, L. & Deline, P. 2011. Climate influence on rockfalls in high-Alpine steep rockwalls: The north side of the Aiguilles de Chamonix (Mont Blanc massif) since the end of the’Little Ice Age’. The Holocene, v.21, pp.357–365.