Réf. Frey & al. 2010 - A

Référence bibliographique complète

FREY, H., HAEBERLI, W., LINSBAUER, A., HUGGEL, C., PAUL, F. 2010. A multi-level strategy for anticipating future glacier lake formation and associated hazard potentials. Natural Hazards and Earth System Sciences, 10, 339-352.

Abstract: In the course of glacier retreat, new glacier lakes can develop. As such lakes can be a source of natural hazards, strategies for predicting future glacier lake formation are important for an early planning of safety measures. In this article, a multi-level strategy for the identification of overdeepened parts of the glacier beds and, hence, sites with potential future lake formation, is presented. At the first two of the four levels of this strategy, glacier bed overdeepenings are estimated qualitatively and over large regions based on a digital elevation model (DEM) and digital glacier outlines. On level 3, more detailed and laborious models are applied for modeling the glacier bed topography over smaller regions; and on level 4, special situations must be investigated in-situ with detailed measurements such as geophysical soundings. The approaches of the strategy are validated using historical data from Trift Glacier, where a lake formed over the past decade. Scenarios of future glacier lakes are shown for the two test regions Aletsch and Bernina in the Swiss Alps. In the Bernina region, potential future lake outbursts are modeled, using a GIS-based hydrological flow routing model. As shown by a corresponding test, the ASTER GDEM and the SRTM DEM are both suitable to be used within the proposed strategy. Application of this strategy in other mountain regions of the world is therefore possible as well.


Organismes / Contact


(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
  Glacier retreat Glacial hazards Glacier lake outburst (flood waves, debris flows), ice avalanches

Pays / Zone
Massif / Secteur
Site(s) d'étude
Période(s) d'observation
Switzerland - Aletsch region (canton of Valais)
- Bernina region (canton of Grisons)
Bernina region: Morteratsch Glacier and Pers Glacier / Villages: Pontresina and Samedan      

(1) - Modifications des paramètres atmosphériques

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

(2) - Effets du changement climatique sur le milieu naturel
For the Aletsch and Bernina regions, scenarios of potential future lakes are presented [see details in the study].

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

Glacier retreat:

In proximity to melting conditions, snow and ice react sensitively to climate change (Haeberli and Beniston, 1998). As a consequence, high-mountain landscapes that are dominated by glacial and periglacial processes are influenced by the governing climatic conditions and their changes. The continuous rapid retreat of glaciers in mountain ranges all over the world is one of the most obvious and reliable indicators of global warming (GCOS, 2004; Lemke et al., 2007).

The authors chose test sites in the Swiss Alps for this study due to the good data availability, including a high-quality DEM and various historical maps since the middle of the 19th century [see data in the study]. The methods for the detection of overdeepenings in the glacier bed are verified at Trift Glacier in the Bernese Alps. Since the mid 1980’s, the tongue of this glacier experienced strong down-wasting (Paul and Haeberli, 2008) and finally collapsed between 2000 and 2005. A new lake (length 1 km, max. width 450 m) formed, which contributed to the rapid disintegration of the glacier tongue. The location is easily accessible and a suspension bridge over the gorge at the end of the lake attracts many tourists in summertime.

They applied the assessment strategy on data from before the start of the lake formation (i.e., before 2000), to analyze whether the formation of this lake could have been anticipated with this approach. For the Aletsch and Bernina regions, scenarios of potential future lakes are presented. The testing of the different DEMs and the preliminary hazard assessment are performed in the Bernina region.

The assessment strategy integrates the following four levels:
– Level 1: selection of parts of the glacier surfaces below a slope threshold. Application to regions of 104 km2 to 105 km² to quickly obtain an overview over large regions.
– Level 2: manual application of three criteria (distinct slope increase, reduction of glacier width, and crevasse-free part followed by heavily crevassed part), to detect potential overdeepenings in the glacier bed. Application to regions of 103 km² to 104 km² to detect regions and situations of special interest.
– Level 3: use of more detailed tools, which require more input data (e.g. digitized central flow lines) to model the ice thickness distribution (e.g., Farinotti et al., 2009; Linsbauer et al., 2009). The region of application ranges from individual glaciers up to several 104 km², depending on the model and the availability of required input data.
– Level 4: in-situ geophysical investigations (e.g., radio-echo sounding) or drilling in the field for detailed information about the bed topography and bed properties at individual points or transects. Local application for situations that are considered to be critical. This level 4 is not further discussed in this article.

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

Potential future lake outbursts at Morteratsch Glacier and Pers Glacier [results of the MSF modeling]:

According to the modeled bed topography, the lakes are all hydrologically connected, in other words, many of the outburst floods reach other lakes located further downstream. This is inherent to the applied model approach, because the largest depths are modeled along the central flow line. However, this is also realistic from a glaciological point of view, because the depressions were shaped by the same glacier. The slope values from the modeled glacier bed exceed 8° in the surrounding of most future lakes, hence, loose material could be entrained by outburst floods from these lakes. The average slope values of the glacier beds are around 25°, but below the glacier tongues the terrain levels out. Even in worst-case scenarios with runout distances for debris flows according to an 11° average slope threshold (Haeberli, 1983; Huggel et al., 2002), debris flows from future glacier lakes could reach lakes situated below, but in most cases the affected areas are confined to regions close to the lakes; they do not reach presently existing buildings, settlements or other expensive infrastructure.

Because it is highly uncertain how much loose sediment will be available in the future, the authors also modeled potential flood waves, which, compared to debris flows, have a much lower sediment concentration. For the maximum runout distance they chose again a worst-case approach and used a maximum average slope of 3°.

In the combined modeling of the whole region in a single model run, the MSF model calculates the average slope always from the lowest starting zone. In other words, if a lake outburst flood reaches a lower lake, the modeling of the flood from the upper lake is aborted and only the flow path from the lower lake is calculated. This is reasonable because a debris flow or flood wave from an upstream lake can trigger an outburst of a lower lake, but the kinetic energy of the incoming flow mass will be used to displace the water from the lower lake. However, in the case of modeling outbursts from potential future lakes, it is uncertain which lakes will form at all and when. It is possible that a lower lake is filled with sediments, drained already earlier, or did never form at all. For this reason, the authors modeled flood waves for all potential future lakes individually to avoid undesirable interactions.

The maximum runout paths of the flood waves are shown for each glacier individually. According to the model, the floods could reach existing settlements and infrastructure. Flood waves from all these potential future lakes could affect the roads and the railroad in the Bernina Valley. Floods originating from the bed of Morteratsch Glacier would not reach Pontresina, but all expected future lakes of Pers Glacier have the potential to cause flood waves that could reach the village of Samedan.

The influence of lakes located in the runout of an upstream lake can be shown with the example of the lowest lake on Morteratsch Glacier. If this depression is empty of water during an outburst of Pers Glacier lakes, the potential destructive reach of such a flood wave could reach the flood plain of Samedan, where a regional airport is located. But if there is a full lake, it could absorb the energy from an outburst of a smaller lake upstream. Of course this depends strongly on the volume and geometry of the retention basin and the magnitude of the outburst flood from Pers Glacier lakes. [...]

The modeling of potential debris flows and flood waves originating from anticipated future lakes of the Morteratsch- and Pers glaciers indicated that a growing number of lakes may not in all cases imply a higher regional hazard potential.


After the [modeled] formation of a lake at [an] overdeepening, Pers Glacier will terminate in a steeper part of the bed and thus favor the formation of ice avalanches. The new lake in front of this tongue might be in the runout distance of such ice avalanches. After continued retreat of Pers Glacier, the situation might become less critical.

The influence of lakes located in the runout of an upstream lake can be shown with the example of the lowest lake on Morteratsch Glacier. If this depression is empty of water during an outburst of Pers Glacier lakes, the potential destructive reach of such a flood wave could reach the flood plain of Samedan, where a regional airport is located. But if there is a full lake, it could absorb the energy from an outburst of a smaller lake upstream. Of course this depends strongly on the volume and geometry of the retention basin and the magnitude of the outburst flood from Pers Glacier lakes. From these finding the hypothesis arises, that a lake of a sufficient size, located at the transition from steep terrain to a flat valley bottom reduces the hazard potential of other lakes located higher up in steep terrain. This is only the case, if the lower lake is large enough and has a sufficiently stable dam to absorb the kinetic energy of an incoming flood wave from an upstream lake outburst.

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

Potential future lake area, runout distances, slope thresholds, probability of affection by a flooding event (pixel)

In this study, the temporal evolution of glacier retreat and thus also the temporal evolution of lake formation is not considered. Instead, runout paths for worst-case scenarios of glacier lake outburst are modeled: all glaciers have retreated strongly, most parts of the beds are exposed, and all overdeepenings modeled on level 3 are completely filled with water. Based on this setup, potential future lake outbursts are modeled for the Morteratsch- and Pers Glaciers, also to demonstrate the potential of simple flow-routing models like the MSF model (Huggel et al., 2003) that can be used in such cases with highly uncertain conditions. Based on the results of such a preliminary assessment, sites that require more detailed investigations on level 4 can be selected. Nevertheless, the temporal evolution of glacier retreat and lake formation has a direct influence on the hazard situation.

Fast changes of the high-mountain cryosphere affect various phenomena, processes and process interactions and can lead to hazard situations beyond historically known conditions. Associated with glacier retreat, pro-glacial lakes can form behind moraine dams or in overdeepened parts of the exposed glacier bed. [...] Glacier lakes can pose a potential threat to the population and infrastructure in the valleys below as they are forming in an environment dominated by interacting, rapidly changing and highly dynamic processes. [...]

The most severe glacier catastrophes often result from a combination and chain reaction of different processes (Huggel et al., 2004; Kääb et al., 2005; Reynolds Geo- Science Ltd, 2003). In glaciated environments, the following changes affect the hazard situation of glacier lakes:
– Glacier retreat can lead to the development of new potential ice avalanche starting zones (Clague and Evans, 2000; Kääb et al., 2005; Richardson and Reynolds, 2000a; Kershaw et al., 2005).
– Transitions in the thermal regime of hanging glaciers, from cold to polythermal or temperate, can destabilize them due to reduced basal friction (Alean, 1985; Fischer et al., 2006; Huggel, 2009).
– The retreat and thinning of glaciers since the Little Ice Age has a destabilizing effect on steep adjacent rockwalls due to the related debuttressing effect (Augustinus, 1995; Ballantyne, 2002). Such rock walls can be situated close to newly forming glacier lakes (e.g. the Schlossplatten rock fall and -toppling at the tongue of the Lower Grindelwald Glacier) (Oppikofer et al., 2008).
– Glacier retreat often exposes morainic and unconsolidated material (Evans and Clague, 1994) that, if situated in steep terrain, is prone to be the starting zone for landslides and debris flows (Huggel et al., 2004; Haeberli et al., 1991; Hubbard et al., 2005).
– The formation of new glacier lakes itself is a potential threat because an outburst of such a new lake can initiate a cascade-like chain reaction involving other lakes located further downstream (cf. Gruben Glacier, Haeberli et al., 2001).
– Regions with warm permafrost conditions are expected to show the most critical stability conditions of frozen rock walls (e.g., Gruber and Haeberli, 2007). With increasing air temperatures, permafrost degradation and thus deep long-term warming of such rock walls will increase the probability of rock falls. Glacier lake outbursts triggered by rock fall will thus be a scenario of increasing probability in the future.
– Permafrost degradation also reduces the stability of moraine dams that contain dead ice or other subsurface ice (Richardson and Reynolds, 2000b). [...]

Reynolds (2000) found a slope gradient of 2° to be the critical threshold for supraglacial lake formation on debris-covered glaciers in the Himalayas. Quincey et al. (2007) confirmed this finding and concluded that debris-covered glacier parts with low flow velocities are most likely sites with a potential for supraglacial lake formation.

In this study, the authors present a multi-level strategy to anticipate the formation of future glacier lakes at different scale levels. At all levels, local overdeepenings in the glacier bed are estimated by analyzing the current glacier surface characteristics based on digital elevation models (DEMs), digital glacier outlines and satellite imagery. The focus of this study is on the quick and qualitative assessment of potential future lake formation in regions, where only sparse data are available.

In view of the obvious difficulty of verifying predictions of future conditions, the approaches are validated using historic data at a test site in the Bernese Alps (Trift Glacier), where a proglacial lake formed recently [see validation in the study]. Scenarios of future lake formation are then derived for the two regions Bernina and Aletsch. In addition, the likely runout path of a flood from potential outbursts of expected future lakes are modeled in the Bernina region. [...]

In order to detect critical situations and cases, which require in-situ investigations on level 4, a hazard assessment concerning the expected lakes is performed based on the results. of level 3. Because these future scenarios have large uncertainties about the exact location, lake volume and properties of the environment surrounding the lakes, such an assessment can only be of preliminary nature. [...]

Huggel et al. (2003) presented a simple but robust debris flow model that includes these empirical runout distances and slope thresholds, and that can be applied with reasonable assumptions for the required input data. The so-called modified single flow model (MSF) is a hydrological flow routing model that calculates for each cell of a DEM a probability of affection by a flooding event. It is based on a hydrological algorithm that calculates the flow direction from one DEM cell to another according to the steepest downward gradient between one cell and its eight neighbours. In addition, it allows also a flow spreading of up to 45° from the main flow direction. The input data are starting cells and a DEM. In this study the modeled potential future lake areas were used as starting zones and the DEM without glaciers obtained on level 3 for flow path modeling. The MSF model can be applied to larger regions, i.e. several lake outbursts can be modeled within one model run. The somewhat speculative and uncertain assumptions about the volume and the sediment concentration of outburst floods are not considered in the model.

(4) - Remarques générales

The presented strategy to detect sites with potential future lake formation on four different levels proved to be feasible and the results of the different levels are consistent. In the Bernina region all the overdeepenings in the modeled bed topography at level 3 could be identified on the previous levels already. Regarding an application to other mountain regions of the world, where only sparse input data may be available, the methods of these first two levels can quickly provide qualitative information without much computational effort.

For such an approach, designed to anticipate future scenarios, it is essential that all overdeepenings are identified, even though not all of them may cause lake formation. [...]

So far, glacier sedimentation was not considered in this study. The sedimentary properties of the glacier beds that will be uncovered in the future have a twofold impact on potential lake formation: on the one hand, continued sedimentation can fill overdeepenings in the bedrock and thus reduce the lake volume or prevent lake formation at all. On the other hand, sedimentary glacier beds can provide loose material for debris flow formation and lead to moraine dammed lakes with a potential for mechanical dam failures and related extreme discharges. [...]

The methods of this study were developed and tested for glaciers without extensive debris cover. However, supraglacial lakes can also develop on debris-covered glaciers of the Alps, e.g. on the Belvedere Glacier in Italy or on the tongue of the Lower Grindelwald Glacier in Switzerland. [...]

The formation of a lake that is in contact with a glacier can have a strong impact on glacier dynamics. In general, a lake at the front of a glacier will accelerate glacier retreat due to increased ablation by calving and thermokarst effects (Kääb and Haeberli, 2001). For this reason, length changes of glaciers that terminate into a lake, are to a certain degree decoupled from the direct forcing of climate change and should thus not be used as indicators of climatic changes. Chinn et al. (2008) estimated an increase in the rate of ice loss by about one order of magnitude for glaciers terminating in a lake compared to other glaciers with ice loss only due to surface melt. These authors also state, that the inclined glacier surfaces are replaced by horizontal water surfaces which results in steeper upslope glacier surfaces and, hence, in faster ice flow. Such impacts of glacier lake formation on glacier dynamics are not commonly taken into account for physically based, dynamic modeling of future glacier development (e.g., like in the studies from Jouvet et al., 2009 or Le Meur et al., 2007). In combination with such models of transient glacier evolution, the temporal component could be included in the assessment of future lake formation. So far, only relative predictions of sequence of lake formation for individual glacier branches are possible by locating the bed-overdeepenings (first, the lake at the lowest overdeepening will form). As soon as multi-tributary glaciers start to separate into individual glaciers, the order of formation may change. For example, Morteratsch and Pers Glacier will separate in the near future and it is possible, that a lake will form earlier than the lower lake due to different retreat rates caused by different ice thicknesses of the individual glacier branches.

In this study outburst floods were modeled, but other important components for the determination of the hazard potential of future lakes were not considered. Dam characteristics for example strongly influence the risk of failure, and external factors like impact waves from mass movements can trigger catastrophic lake outbursts. To allow a more detailed and integrated hazard assessment of lakes expected to form in the future, such components must be anticipated as well. More detailed assumptions about the lake size, dam properties and lake surrounding would allow a more sophisticated modeling of potential lake outbursts.

(5) - Syntèses et préconisations

The presented integrated four-level strategy for the detection of sites with potential glacier lake formation is a suitable approach to anticipate future situations in high-mountain environments. Starting with the first level to gain a rough overview over a large region, the strategy subsequently focuses on smaller regions in more detail, down to in-situ geophysical measurements in the field. This multiscale procedure allows to rapidly identifying the most critical sites in a large region. Applications of the strategy to two test regions showed, that the newly presented, qualitative approaches (levels 1 and 2) proved to be consistent with the results from more detailed/laborious models that are applied on the third level. In both test regions, ten or more overdeepenings were detected and the formation of several glacier lakes has to be expected in the future. Due to the irregular longitudinal bed profiles caused by glacier erosion (Hooke, 1991), glacier lakes will form in the course of continued glacier retreat in cold mountain regions. Tests of the presented strategy with the nearly globally available SRTM and ASTER DEMs revealed satisfying results. This indicates that the approach can be applied worldwide, even in remote regions with less detailed data coverage.

The hazard situation of such glacier lakes needs to be (re-)assessed regularly and at short time intervals, because these lakes form in a rapidly changing environment and their hazard potential is determined by factors that are themselves subject to changes. Modeling of potential glacier lake outbursts is an important part of such a hazard assessment and can also be performed for glacier lakes that are expected to form in the future. The modeling of potential debris flows and flood waves originating from anticipated future lakes of the Morteratsch- and Pers glaciers indicated that a growing number of lakes may not in all cases imply a higher regional hazard potential.

To predict the time of lake formation, the here-presented strategy needs to be coupled with simple but transient models that determine glacier retreat over large regions. Such a combination is in the focus of future research and will further increase the benefit of the presented approaches.

Références citées :

Alean, J.: Ice avalanches: some empirical information about their formation and reach, J. Glaciol., 31, 324–333, 1985.

Augustinus, P.: Glacial valley cross-profile development: the influence of in situ rock stress and rock mass strength, with examples from the Southern Alps, New Zealand, Geomorphology, 14, 87– 97, 1995.

Ballantyne, C.: Paraglacial geomorphology, Quaternary Sci. Rev., 21, 1935–2017, 2002.

Chinn, T., Salinger, J., Fitzharris, B., and Willsman, A.: Glaciers and climate, Bulletin of the Federated Mountain Clubs of NZ, 171, 1–15, 2008.

Clague, J. and Evans, S.: A review of catastrophic drainage of moraine-dammed lakes in British Columbia, Quarternary Sci. Rev., 19, 1763–1783, 2000.

Evans, S. and Clague, J.: Recent climatic change and catastrophic geomorphic processes in moutnain environments, Geomorphology, 10, 107–128, 1994.

Farinotti, D., Huss, M., Bauder, A., Funk, M., and Truffer, M.: A method to estimate the ice volume and ice-thickness distribution of alpine glaciers, J. Glaciol., 55, 1–9, 2009.

Fischer, L., KäÄb, A., Huggel, C., and Noetzli, J.: Geology, glacier retreat and permafrost degradation as controlling factors of slope instabilities in a high-mountain rock wall: the Monte Rosa east face, Nat. Hazards Earth Syst. Sci., 6, 761–772, 2006, http://www.nat-hazards-earth-syst-sci.net/6/761/2006/. [Fiche Biblio]

Frey, H., Huggel, C., Paul, F., and Haeberli, W.: Automated detection of glacier lakes based on remote sensing in view of assessing associated hazard potentials, in: Proceedings of the 10th International Symposium on High Mountain Remote Sensing Cartography. Grazer Schriften der Geographie und Raumforschung, edited by: Kaufmann, V. and Sulzer, W., vol. 45, in press, 2010.

GCOS: Implementation plan for the Global Observing System for Climate in support of the UNFCCC, GCOS – 92, WMO, Geneva, 2004.

Gruber, S. and Haeberli,W.: Permafrost in steep bedrock slopes and its temperature-related destabilization following climate change, J. Geophys. Res., 112, F02S18, doi:10.1029/2006JF000547, 2007. [Fiche Biblio]

Haeberli, W.: Frequency and characteristics of glacier floods in the Swiss Alps, Ann. Glaciol., 4, 85–90, 1983.

Haeberli, W. and Beniston, M.: Climate change and its impacts on glaciers and permafrost in the Alps, Ambio, 27, 258–265, 1998. [Fiche Biblio]

Haeberli, W., Rickenmann, D., Zimmermann, M., and Rösli, U.: Murgänge, in: Ursachenanalyse der Hochwasser 1987, Ergebisse der Untersuchungen, vol. 4, Mitteilungen des Bundesamtes f¨ur Wasserwirtschaft, 77–88, 1991.

Haeberli, W., Kääb, A., Mühll, D. V., and Teysseire, P.: Prevention of outburst floods from periglacial lakes at Grubengletscher, Valais, Swiss Alps, J. Glaciol., 47, 111–122, 2001.

Hooke, R.: Positive feedbacks associated with erosion of glacial cirques and overdeepenings, Geol. Soc. Am. Bull., 103, 1104– 1108, 1991.

Hubbard, B., Heald, A., Reynolds, J., Quincey, D., Richardson, S., Luyo, M., Portilla, N., and Hambrey, M.: Impact of a rock avalanche on a moraine-dammed proglacial lake: Laguna Safuna Alta, Cordillera Blanca, Peru, Earth Surface Processes and Landforms, 30, 1251–1264, 2005.

Huggel, C.: Recent extreme slope failures in glacial environments: effects of thermal perturbation, Quaternary Sci. Rev., 28, 1119– 1130, doi:10.1016/j.quascirev.2008.06.007, 2009.

Huggel, C., Kääb, A., Haeberli, W., Teysseire, P., and Paul, F.: Remote sensing based assessment of hazards from glacier lake outbursts: a case study in the Swiss Alps, Can. Geotech. J., 39, 316–330, 2002.

Huggel, C., Kääb, A., Haeberli, W., and Krummenacher, B.: Regional-scale GIS-models for assessment of hazards from glacier lake outbursts: evaluation and application in the Swiss Alps, Nat. Hazards Earth Syst. Sci., 3, 647–662, 2003, http://www.nat-hazards-earth-syst-sci.net/3/647/2003/.

Huggel, C., Haeberli, W., Kääb, A., Bieri, D., and Richardson, F.: An assessment procedure for glacial hazards in the Swiss Alps, Can. Geotech. J., 41, 1068–1083, 2004.

Jouvet, G., Huss, M., Blatter, H., Picasso, M., and Rappaz, J.: Numerical simulation of Rhonegletscher from 1874 to 2100, J. Comput. Phys., 228, 6426–6439, 2009.

Kääb, A. and Haeberli, W.: Evolution of a high-mountain thermokarst lake in the Swiss Alps, Arctic, Antarctic, and Alpine Research, 33, 385–390, 2001.

Kääb, A., Huggel, C., Fischer, L., Guex, S., Paul, F., Roer, I., Salzmann, N., Schlaefli, S., Schmutz, K., Schneider, D., Strozzi, T., and Weidmann, Y.: Remote sensing of glacier- and permafrost related hazards in high mountains: an overview, Nat. Hazards Earth Syst. Sci., 5, 527–554, 2005, http://www.nat-hazards-earth-syst-sci.net/5/527/2005/. [Fiche Biblio]

Kershaw, J., Clague, J., and Evans, S.: Geomorphic and sedimentological signature of a two-phase outburst flood from morainedammed Queen Bess Lake, British Columbia, Canada, Earth Surface Processes and Landforms, 30, 1–25, 2005.

Le Meur, E., Gerbaux, M., Schäfer, M., and Vincent, C.: Disappearance of an Alpine glacier over the 21st Century simulated from modeling its future surface mass balance, Earth Planet. Sci. Lett., 261, 367–374, 2007.

Lemke, P., Ren, J., Alley, R., Allison, I., Carrasco, J., Flato, G., Fujii, Y., Kaser, G., Mote, P., Thomas, R., and Zhang, T.: Observations: Changes in Snow, Ice and Frozen Ground, in: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K., Tignor, M., and Miller, H., Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA 2007.

Linsbauer, A., Paul, F., Hoelzle, M., Frey, H., and Haeberli, W.: The Swiss Alps Without Glaciers – A GIS-based Modelling Approach for Reconstruction of Glacier Beds, in: Geomorphometry 2009 Conference Proceedings, edited by: Purves, R., Gruber, S., Straumann, R., and Hengl, T., University of Zurich, Zurich, 243– 247, www.geomorphometry.org, 2009.

Oppikofer, T., Jaboyedoff, M., and Keusen, H.-R.: Collapse at the eastern Eiger flank in the Swiss Alps, Nature Geosci., 1, 531– 535, 2008.

Paul, F. and Haeberli, W.: Spatial variability of glacier elevation changes in the Swiss Alps obtained from two digital elevation models, Geophys. Res. Lett., 35, L21502, doi:10.1029/2008GL034718, 2008.

Quincey, D., Richardson, S., Luckman, A., Lucas, R., Reynolds, J., Hambrey, M., and Glasser, N.: Early recognition of glacial lake hazards in the Himalaya using remote sensing datasets, Global Planet. Change, 56, 137–152, 2007.

Reynolds, J. M.: On the formation of supraglacial lakes on debriscovered glaciers, in: Debris-Covered Glaciers. Proceedings of a workshop held at Seattle, Washington, USA, September 2000, edited by: Nakawo, M., Raymond, C., and Fountain, A., 153– 161, IAHS Publication, 2000.

Reynolds Geo-Science Ltd: Guidelines for the management of glacial hazards and risks, Reynolds Geo-Science Ltd, Mold, UK, 2003.

Richardson, S. and Reynolds, J.: An overview of glacial hazards in the Himalayas, Quaternary Int., 65–66, 31–47, 2000a.

Richardson, S. D. and Reynolds, J. M.: Degradation of ice-cored moraine dams: implications for hazard development, in: Debris- Covered Glaciers, edited by: Nakawo, M., Raymond, C. F., and Fountain, A., Proceedings of a workshop held at Seattle, Washington, USA, September 2000, Oxford, IAHS Publication, 2000b.