Two British researchers recently published the first global inventory and damage assessment of the societal consequences incurred by glacial lake outburst floods (GLOFs). They revealed that glacial lake outburst floods (GLOFs) have been declining in frequency since the mid-1990s, with the majority released by ice dam failures.
Glacial hazard specialists Jonathan Carrivick and Fiona Tweed spent 18 months scouring the records of over 1,348 GLOFs, determining that such floods have definitely claimed over 12,400 lives since the medieval period. Their work stems from a need to strengthen data on glacier lakes.
“There was very very little quantitative data out there on the importance of glacier lakes, from a societal point of view,” Carrivick said in an interview with GlacierHub. He explained that this recent paper was a natural progression from his earlier research, which focused on modelling hydrological, geological and geomorphological processes.
Based purely on frequency, Carrivick and Tweed found that north-west North America (mainly Alaska), the European Alps (mainly Switzerland), and Iceland are the “most susceptible regions” to GLOFs. However, the impacts of these events have have often been minimal, as they occur in sparsely populated, remote regions, and in places where resilience is high.
The greatest damage has been inflicted upon Nepal and Switzerland — respectively accounting for 22 percent and 17 percent of the global total damage reported. When Carrivick applied the normalized ‘Damage Index,’ which considered GDPs of the affected country (used as a crude proxy for ability to mitigate, manage and recover), he found that Iceland, Bhutan and Nepal have suffered the “greatest national-level economic consequences of glacier flood impacts.”
Historically, Asian and South American GLOFs have been the deadliest, taking the lives of 6,300 and 5,745 individuals since 1560 respectively. However, these figures are dominated by only two catastrophes, which accounted for 88 percent of the 12,445 fatalities confirmed by Carrivick and Tweed. The first, in December 1941, saw over 5,000 Peruvians perish in Huaraz, when a landslide cascaded into the glacial Lake Palcacocha. The second event, swept away more than 6,000 Indians from across Uttarakhand in June 2013, as torrential rains triggered outburst floods and landslides.
The study’s authors adopted a method for normalizing damage assessments new to GLOF hazard analysis, striving to fairly compare the cataclysmic impacts of outburst flooding on communities around the world.
They found that there has actually been a decline in number of floods since the 1990s, which was surprising to the researchers, given that a 2013 study which they had conducted found that the number and size of glacial lakes has increased, as the world’s ice masses have wasted. Carrivick stated that he was “very interested in the fact that, apparently, so few glaciers have lakes that have burst [0.7% of the total], on a global scale.” He added, “it beggars belief that there isn’t a higher percentage of those lakes that have burst at some point.”
In their paper, the pair suggest that the “apparent decline” could be attributed to improved successful stabilisation efforts, natural resilience, greater awareness and preparedness in threatened communities, or declined number of GLOFs from ice-dammed lakes.
An additional factor may be that some glacial floods are missing from the English-language record. Carrivick revealed, “We have a contact in China who says that there’s a lot of unpublished floods…that individual has not been able to send us the data yet.” Government restrictions on the flow of potentially sensitive information has contributed to this partial release of data.
Carrivick also noted that new data is continually being published, in many cases in foreign languages. He referenced a recent issue of the Geological Journal, which released “a whole heap of extra data,” translated from Russian.
Academics have been actively studying GLOFs since at least 1939. But it was not until 1996 that the first relatively comprehensive, global-scale inventory was compiled and published by Joseph Walder and John Costa, who recognized the “flood hazards posed by glacier-dammed lakes.” Carrivick and Tweed found the failure of this type of dam was the leading cause of GLOFs, accounting for 70 percent of events around the world.
Earlier this year, GlacierHub wrote about an alternate database, which has been compiled under the oversight of the International Programme on Landslides — glofs-database.org. The project has been led by Adam Emmer, a PhD working with Vít Vilímek at Charles University in Prague. Three years ago, Emmer, Vilímek, and their team sought to compile a comprehensive global database, identifying over 500 events since the mid-1800s.
The work of Emmer and Vilímek’s team, like Walder, Costa and many others, predominantly focused on physical processes, such as the mechanisms which set off GLOFs, flood routes and distance, volume, as well as the quantity of debris carried by the floodwaters. Documentation of the socioeconomic impacts has remained been relatively less developed in glacial hazards research.
Noting this shortcoming, Carrivick and Tweed decided their study should focus specifically on the societal consequences of GLOFs. They included the number of deaths, injuries, evacuees, displaced, structural damage, financial loss, and called for the inclusion of less tangible social impacts in future studies, including Post-Traumatic Stress Disorder (PTSD). They also acknowledged potentially positive effects of floods, such as increased power generation at hydropower facilities.
They developed a ‘Damage Index,’ which allowed them to conduct standardised assessments of the impacts each GLOF had on downstream communities. This was by no means easy or straightforward. As Carrivick noted, “A footbridge going down in Bhutan has a very different impact to a footbridge going down in Alaska. One is absolutely vital to the functioning of society, and the other one probably receives ten tourists in a year.” They sought a methodology for normalising the heterogeneous impacts of GLOFs around the world, ultimately choosing the ‘Natural Disaster Impact Assessment’ (NDIA), developed by Olga Petrucci of the Italian National Research Council.
The authors decided that the damage investigation should be conducted by Carrivick alone, who assigned a “relative score” to each event, as they sought to “provide a quantitative comparison.” Carrivick spent six months trawling through the records of 332 GLOFs (24 percent of the total) for which the societal impacts were known.
Carrivick emphasised that he and Tweed were “indebted” to the teams that have established the various comparable databases, which provided them with a “running start.” However, in reviewing their data they found that “whilst several natural hazards databases purport long-term records, they are in reality biased towards more recent events.”
The researchers note the reality that GLOF-related research and mitigation activity at potentially hazardous sites is costly. Lack of funds has plagued efforts around the world. Both proactive (i.e. glacial lake research, continuous monitoring, mitigation works), and retroactive (i.e. repairs, reparations) initiatives are often low on national to-do lists, especially where resources are limited.
Carrivick and Tweed are hoping that their latest paper will establish an important foundation, upon which affected nations and colleagues can build. “It’s not wagging the finger at all, and saying ‘You can’t cope’ or ‘You can’t manage,’ but it’s identifying where we might strategically invest scientific work, and invest international collaborative efforts,” said Carrivick.