Vulnerabilities - Landslides in the past

Landslides contribution to all global natural hazards

Landslides are caused by tectonic (seismic), climatic and/or human activities. Landslides are a significant global hazard. From global data in the 1970s it was concluded that ca. 14% of total casualties from natural hazards are attributed to slope failure (19). The well-known and often used EM-DAT International Disaster Database suggests that landslides account for 4.9% of all natural disaster events and 1.3% of all natural hazard fatalities between 1990 and 2015. These lower numbers are probably underestimates. Studies have shown that the EM-DAT database underestimated the number of fatal landslide events by 1400% (20) or 2000% (21) and the number of fatalities by 331% (20) or 430% (21). For the most part this underreporting is associated with the perception of landslides as a secondary hazard, with the cause of death often being recorded in connection with the primary hazard (e.g. an earthquake) rather than the actual cause of the loss. 54% of these landslide events occurred in Asia (22).

A global dataset of fatal non-seismic landslides, covering the period from January 2004 to December 2016, shows that in total 55,997 people were killed in 4862 distinct landslide events (18). The majority of these events were triggered by rainfall (79%). South, Southeast and East Asia, and South and Central America combined contain 88% of all global rainfall-triggered landslide events. 16% of all non-seismic landslide events were not generated by rainfall but were triggered by mining, construction or illegal hill cutting. These human activities caused almost 7% of all landslide-related fatalities (18).

The data do not show a clear long-term increase or decrease in global landslide impact; rather, the record shows that the number of landslides varies strongly from one year to another on a global scale. The data show that landslide occurrence triggered by human activity is increasing, in particular in relation to construction, illegal mining and hill cutting. This supports notions that, on a global scale, human disturbance may be more detrimental to future landslide incidence than climate change. This is supported by other studies (23). Things are different in the mountains, however: there is a high confidence that glacial retreat and permafrost degradation will increase slope instabilities in high mountain areas in the long term.

No clear relationships with climate change

A global assessment
 of the impacts of climate change on landslide risk has been carried out based on a review of scientific publications on past, current, and future impacts of climate change on landslides (1). The majority (80%) of the papers in this review found causal relationships between landslides and climate change. From this review, the type, extent, magnitude, and direction of the impacts of climate change on landslide location, abundance and frequency are not completely clear, however. The effects of the warming climate on landslide risk, and particularly the risk to the population, also remain difficult to quantify. The Alps are the most investigated physiographic area, including the French, Italian, and Swiss Alps.

It is still hard to determine if and where landslide risk, and particularly risk to the population may increase (or decrease) in direct or indirect response to climate change. Whether areas are subject to (an increasing) landslide risk not only depends on temperature and rainfall, but on local geological conditions and non-climatic factors (including land use/cover, agriculture and forest practices) as well. Besides, natural and human induced drivers of landslides interact in a complex way, even in a “stable” climate (1). The direction, magnitude and effects of these non-climatic, interacting drivers may outweigh changes in landslide activity due to climate change (4). Besides, in many areas global warming will have an impact on land use and land cover, on agricultural and forestry practices, and on the economy. These changes may also change the activity and the rate of occurrence of landslides, and hence landslide hazard and risk (5). The incompleteness of old climate and landslide records also plays a role; this limits the possibility to evaluate the impact of the expected environmental and climate changes on landslide frequency, and to estimate variations in the associated risk (6).

It was concluded previously that impacts of climate change on landslides were in places likely minor, compared to impacts from human disturbance (6).

Vulnerabilities - Debris-dammed glacial lake outburst floods in the past

When glaciers retreat, the melt water of the glaciers may form lakes behind the moraines that are left behind by the receding glacier tongues. These moraines, some of which contain a melting ice core, are built from rock debris transported by glaciers. When they fail, large volumes of stored water can be released, producing glacial lake outburst floods (GLOFs). These floods have caused thousands of fatalities and severe impacts on downstream communities, infrastructure and long-term economic development (28). In addition to moraines, lakes can also form behind dams of ice and bedrock. For moraine-dammed lakes the link between climate change, glacier response, lake formation and the occurrence of glacial lake outburst floods is more straightforward compared to the range of processes driving these floods from ice- and bedrock-dammed lakes (27).

The global number of glacier lakes and their total area increased by 53 and 51%, respectively, between 1990 and 2018. Over this period, global glacier lake volume increased by around 48% (33). The risk of glacial lake outburst flooding is largest by far in Asia and in the Andes; the risk is relatively low in Europe (34).

The occurrence of glacial lake outburst floods following moraine dam failure in the past has been studied on a global scale (27). 165 moraine-dam outburst floods were identified since the beginning of the 19th century. Fourteen of them are listed from the European Alps. Three are from Austria between 1890 and 1940, five from Switzerland between 1958 and 1993, one from France in 1944, and five from Italy between 1870 and 1993.

Triggers

Glacial lake outburst floods are often triggered by ice and rock falls, rockslides or moraine failures into lakes, creating seiche or displacement waves, but also by heavy precipitation or ice melt/snowmelt events (29). While climate change plays a dominant role in the recession of glaciers, downwasting glacier surfaces debuttress valley rock walls, leading to catastrophic failure in the form of rock avalanches or other types of landslides (30). Other climatically induced triggers of moraine dam failures include increased permafrost and glacier temperatures leading to failure of ice and rock masses into lakes and the melting of ice cores in moraine dams, which leads to moraine failure and lake drainage. These triggers may occur in all seasons: warm summer weather may trigger an ice avalanche into the lake or moraine melt-through, or heavy winter snow may trigger an ice avalanche into the lake.

No clear relationship with climate change

Surprisingly, no increase in the number of glacial lake outburst floods was found in recent decades, despite a clear trend of continued glacier recession and glacier lake development in recent decades. According to the researchers this may be due to several reasons. It’s not just climate change that determines these outburst floods but non-climatic factors also play a role, such as moraine dam geometry and sedimentology, and climate-independent triggers such as earthquakes. Besides, most of these triggers (mass movements into lakes) evolve on a wide range of time scales because they are related to destabilization of mountain slopes (31). Also efforts have been successful in recent decades to stabilize moraine dams and change the ability of fluvial systems to transmit floods over time (32). All in all this results in a “lagged” response of glacial lake outburst floods to glacier perturbations following climate change. The connections between climate change and these floods are hidden in response time dynamics.

The researchers do expect a substantial increase in glacial lake outburst floods incidence throughout the 21st century and into the 22nd century, however, as glaciers and lakes respond more dynamically to anthropogenic climate warming (27).

Vulnerabilities - Ice-dammed glacial lake outburst floods in the past

At melting glaciers, lakes are being formed where meltwater gets trapped behind dams of debris or ice. Once these dams fail, large volumes of water are released that can destroy bridges, pipelines, roads, campgrounds and farmland, and kill livestock and people. Luckily, the fast majority of these floods do not cause catastrophic damage and fatalities (35). This is partly due to human interventions. Dykes, floodgates, dams, or the temporary or permanent resettlement of mountain communities have reduced the impact of these floods (36). The danger of these lakes should not be underestimated, however.

Ice-dammed lakes in particular are ‘dangerous beauties’. Over 70% of all reported glacial lake outburst floods since 1900 in six major mountain regions, globally, originated from ice-dammed lakes (37). The number and size of ice-dammed lakes are small compared with meltwater lakes dammed by moraines debris or bedrock. However, ice dams are much more prone to breaking than dams of moraines or bedrock, making ice-dammed lakes the most hazardous of all meltwater lakes (35).

Trends in outburst flooding since 1900

Global warming and the melting of glaciers can be expected to change the frequency and character of floods resulting from outbursts of ice-dammed lakes. Melting glaciers, after all, create new space for ice-dammed lakes to form and grow (38), and ice dams might store an increasing amount of meltwater, causing larger floods if they fail. These changes were studied by collating 1,569 ice-dam failures in six major mountain regions. Trends in peak discharge and total volume of these floods and in annual timing and elevation of the outburst have been assessed since 1900. In this assessment, the scientists focused on the six most glaciated mountain regions on earth, including northwest North America, High Mountain Asia, the Andes, Iceland, Scandinavia and the European Alps. They studied scientific papers and other sources, such as newspaper reports and written correspondence with local eyewitnesses and experts, and found a total of 1,569 dated outbursts from 186 ice-dammed lakes in the period 1900–2021 (35).

Volumes outburst floods have declined

They showed that extreme peak flows and total flood volumes have declined by about an order of magnitude since 1900 in five of the six regions. In this 120 years period, more outburst floods were reported since 1990 compared with earlier decades in all regions but this is probably due to a more consistent documentation of these floods in recent decades (35).

Annual timing outburst floods has advanced

The average annual timing of outbursts has shifted earlier globally by about 40 days in recent years compared with 1900. This shift in timing was also observed for the European Alps. Outbursts in Iceland, however, now occur about six weeks later in the year on average. The reason why the situation in Iceland is different from the other regions is not clear (35).

Outburst floods have shifted to higher elevations

Since 1900, outbursts from ice-dammed lakes originated at progressively higher elevations. In most regions, outbursts now emerge from lakes that are tens to many hundreds of metres higher in elevation than a century ago. In Scandinavia and Iceland, the average elevation shift was 25–50 m per decade. In the European Alps, no remarkable elevation shift was observed (35).

Floods have weakened

The study showed that the decline in peak flows and volumes and the shift of these outburst to a moment earlier in the year could signal a weakening of glacial lake outburst floods. The scientists wondered whether these changes were due to the shrinking of glaciers but could not find a clear relationship between glacier thinning and changes in outburst floods. The researchers anticipate that ice-dam formation and failure might become a phenomenon of the past within a few decades in the European Alps and Scandinavia, because of the rapid melting of the glaciers. At the global scale, this may happen in the course of one to several centuries (35).

A surprising conclusion

Remarkably, outburst floods of ice-dammed lakes have not become more extreme since 1900, on the contrary. This conclusion will come as a surprise to many readers. After all, glacial lake outburst floods are considered a growing hazard now that glaciers are melting rapidly. Apparently, these ice dams have become thinner and break at a lower water level of the lake behind the dam. These lakes fill and empty repeatedly, and this cycle has speeded up. The lakes release gradually smaller flood volumes over time. Within this cycle, the annual timing of these floods is likely to shift because thinner dams may require a shorter period of time to refill to their maximum storage capacity. This explanation illustrates that the less extreme nature of these floods is a logical consequence of climate change. Apparent, along with the melting of the glaciers the strength of the ice dams declines, thus increasing flood frequency while reducing the extreme nature of the floods. One should be aware, however, that this explanation only refers to ice-dammed lakes and not to the lakes that are formed behind dams of debris (35).

Vulnerabilities - Future projections

According to the International Disaster - Emergency Events Database (EM-DAT), over the period 1985–2014, absolute economic losses in from all flood and mass movements in the European Alps were USD 7 billion (26).

Higher risk of shallow, rapid-moving landslides

Rainfall events 

Intense rainfall events are a primary trigger of shallow, rapid-moving landslides such as debris flows, debris avalanches, rock falls, and also ice falls and snow avalanches in high mountain areas (7). These events are a primary cause of landslide fatalities (8). Given the fact that in some areas global warming is expected to increase both the intensity of rainfall events and the frequency of these events, it is to be expected that in these areas the total number of people exposed to landslide risk will increase. These areas include the Alps, the Himalayas and most of the American Cordillera, but also the Atlas Mountains in northwestern Africa, mountains and hills in southwestern Africa, the East Africa's Rift Valley and the Arabian Peninsula, the Carpathians in Eastern Europe, the Appalachians in eastern North America.

More heavy precipitation events will probably increase the number of landslides in parts of Europe that are susceptible to landslides because of their steep hills and geological subsoil. This is especially the case for a number of regions along the north side of the Alps: the Jura Mountains, the Vosges, the Black Forest, the Swabian Jura, the Bavarian Pre-alps, the foothills of the Austrian Alps and the Bohemian Forest (16). These regions are also important transport corridors for Europe’s road and rail network. More landslides may disrupt these corridors and cause a lot of damage. Heavy precipitation events can trigger landslides when they exceed a certain threshold: a certain amount of precipitation in 3 succeeding days including at least 1 day with extremely heavy precipitation (17). The potential increase of the number of these events was assessed for central Europe, based on an intermediate scenario of climate change (the so-called SRES A1B scenario) and a large number of climate model runs. The results of this study show that the frequency of landslide-triggering extreme climate events will slightly increase over time. At the end of this century, the aforementioned areas along the north side of the Alps are likely to experience substantially increased landslide activity compared to current climate conditions. The increase may be up to 14 additional landslide-triggering rainfall events per year, on average (16).

Published results on the impact of rainfall do not always point in the same direction, however. For Calabria (southern Italy) antecedent rainfall in the month before a landslide event appeared to play a key role to initiate rainfall-induced landslides (9). For the nearby Puglia region, variations in rainfall and temperature did not justify the observed increase in landslide events between 1918 and 2006, however (10).

Higher air temperature

In many high mountain areas, glacier retreat and permafrost thaw are projected to further decrease the stability of slopes, and the number and area of glacier lakes will continue to increase. Floods due to glacier lake outburst or rain-on-snow, landslides and snow avalanches, are projected to occur also in new locations or different seasons (24).

In addition to the intense rainfall events, the projected increase in air temperature is also expected to affect the stability of rock slopes at high latitudes (particularly in the northern hemisphere and at high elevations, where permafrost exists that may reduce when temperature rises (11). A number of investigators have examined the effects of air temperature on debris flows and rock falls, chiefly in the European Alps, and found an increase in landslide activity related to an increase in air temperature (12). According to IPCC, there is a “high confidence that changes in temperature, glacial retreat, and/or permafrost degradation will affect slope instabilities in high mountains, and medium confidence that temperature-related changes will influence bedrock stability” (13). In high mountain areas, not only small-sized rock falls and ice falls, but also large rock slides and rock avalanches may become more abundant (14). At high latitudes, particularly in the taiga and tundra areas in the northern hemisphere, permafrost melting can initiate ground instability processes even in low gradient terrain, producing incised gullies that transform rapidly into wide badland areas.

Global warming will also change the time required for the snow to melt, and the frequency of rain-on-snow events, two known triggers of landslides. There is “medium confidence that high-mountain debris flows will begin earlier in the year because of earlier snowmelt, and that continued mountain permafrost degradation and glacier retreat will further decrease the stability of rock slopes” (13).

Lower risk of deep-seated landslides

The degree of activity and the occurrences of new deep-seated landslides are expected to decrease (15). Extremely to moderately slow deep-seated landslides (including earthflows, mudflows, complex and compound slides) generally do not pose a serious threat to human life. Hence, their predicted reduced activity will not decrease landslide risk to the population significantly, but it is expected to contribute to reducing landslide impact and the related economic damage (1).

Climatic drivers

The literature review revealed that variations in rainfall totals influence mostly rockslides, mud flows and earth flows, at both the local and the regional scale, whereas variations in rainfall intensity affect, mostly directly, rock falls and debris flows/avalanches, in the short-term and at the local scale. Changes in the air temperature influence directly ice falls and avalanches, and have an indirect impact on rock falls (due to the formation and opening of fractures), and on deep-seated landslides (due to changes in the hydrological cycle) (1).  

Adaptation strategies

The IPCC concluded in 2019 that there is high confidence that in the context of mountain flood and landslide hazards, exposure, and vulnerability growing in the coming century, significant risk reduction and adaptation strategies will be required to avoid increased impacts (25).

Both physical constructions (“hard” measures) and “soft measures” such as sustainable land management and forest harvesting can prove cost-effective against landslides (25).

“Hard” measures

Existing single (e.g., a retaining wall, a check dam, a drainage) or multiple (e.g., a system of retaining barriers or a set of drainages in a slope, a set of check dams in a catchment) defensive structures may require modifications to adapt to the new, predicted climate conditions. Defensive structures may have been designed for a specific type of failure (e.g., a slow moving deep-seated landslide) and as a result of a change in climate a different type of landslide may be triggered (e.g., a very rapid soil slip or debris flow). In this case, the presence of structural defensive measures gives a false sense of safety (2). From a literature review of landslide risk under climate change it is recommended that all structural slope defensive measures be checked to evaluate their efficacy in the new or predicted climate conditions (1).

“Soft” measures

The effect of soft measures may be restricted because long-term land planning frequently does not consider climate change and the related environmental and societal consequences. In Italy for instance, River Basin Authorities have prepared basin-scale, landslide (and flood) hazard and risk assessment and management plans largely ignoring the effects of the predicted climate and environmental changes. This limits the future effectiveness of the plans that may even be counterproductive. In places, the plans ignore or underestimate the risk posed by specific landslide types and particularly the types that are expected to increase in response to the predicted climate changes (e.g., very to extremely rapid soil slips and debris flows). In these areas, mitigation actions and adaptation strategies based on the existing risk assessments may be misleading, inadequate, or incorrect (1). Landslide monitoring and early warning systems (3) are a different type of effective non-structural defensive measure that can greatly reduce landslide risk, and particularly the risk to the population. The ability of existing networks of meteo-hydrological sensors to measure variables relevant to landslide early warning may be reduced when climate changes more rapidly (1).

References

The references below are cited in full in a separate map 'References'. Please click here if you are looking for the full references for Europe.

1. Gariano and Guzzetti (2016)
2. Sidle and Chigira (2004), in: Gariano and Guzzetti (2016)
3. Stähli et al. (2015), in: Gariano and Guzzetti (2016)
4. Sidle and Dhakal (2002), in: Gariano and Guzzetti (2016)
5. Van Beek (2002); Wasowski et al. (2010); Lonigro et al. (2015), all in: Gariano and Guzzetti (2016)
6. Crozier (2010), in: Gariano and Guzzetti (2016)
7. Stoffel et al. (2014), in: Gariano and Guzzetti (2016)
8. Guzzetti et al. (2005b); Petley (2012), both in: Gariano and Guzzetti (2016)
9. Polemio and Petrucci (2010), in: Gariano and Guzzetti (2016)
10. Polemio and Lonigro (2015), in: Gariano and Guzzetti (2016)
11. Huggel et al. (2012, 2013); Stoffel et al. (2014); Chiarle et al. (2015); Ravanel and Deline (2015); Paranunzio et al. (2016), all in: Gariano and Guzzetti (2016)
12. Ravanel and Deline (2011, 2015); Stoffel and Beniston (2006); Paranunzio et al. (2016), all in: Gariano and Guzzetti (2016)
13. Seneviratne et al. (2012), in: Gariano and Guzzetti (2016)
14. Huggel et al. (2012, 2013), in: Gariano and Guzzetti (2016)
15. Malet et al. (2005); Coe (2012); Comegna et al. (2013); Rianna et al. (2014), all in: Gariano and Guzzetti (2016)
16. Schlögl and Matulla (2018)
17. Guzzetti et al. (2008), in: Schlögl and Matulla (2018)
18. Froude and Petley (2018)
19. Varnes and IAEG Commission on Landslides (1984), in: Froude and Petley (2018)
20. Kirschbaum et al. (2015), in: Froude and Petley (2018)
21. Petley (2012), in: Froude and Petley (2018)
22. Guha-Sapir et al. (2018), in: Froude and Petley (2018)
23. Innes (1983); Glade (2003); Soldati et al. (2004); Imaizumi et al. (2008); Borgatti and Soldati (2010); Crozier (2010); Anderson and Holcombe (2013); Lonigro et al. (2015), all in: Froude and Petley (2018)
24. IPCC (2019a)
25. IPCC (2019b)
26. Stäubli et al. (2018), in: IPCC (2019b)
27. Harrison et al. (2018)
28. Mool et al. (2011); Riaz et al. (2014); Carrivick and Tweed (2016), all in: Harrison et al. (2018)
29. Richardson and Reynolds (2000), in: Harrison et al. (2018)
30. Ballantyne (2002); Shugar and Clague (2011); Vilímek et al. (2014), all in: Harrison et al. (2018)
31. Haeberli et al. (2017), in: Harrison et al. (2018)
32. Carrivick and Tweed (2014), in: Harrison et al. (2018)
33. Shugar et al. (2020)
34. Taylor et al. (2023)
35. Veh et al. (2023)
36. Carrivick and Tweed (2016), in: Veh et al. (2023)
37. Veh et al. (2022), in: Veh et al. (2023)
38. Geertsema and Clague (2005); Rick et al. (2022), both in: Veh et al. (2023)