KATHMANDU: The Himalayan disaster that has killed hundreds in Nepal and left thousands missing has exposed a dangerous blind spot in how the region prepares for glacier-related disasters. Scientists and governments have increasingly focused on identifying dangerous glacial lakes, while some of the most destructive events may begin without a lake at all.
The August 26 catastrophe along the Nepal-Tibet border was not a conventional Glacial Lake Outburst Flood (GLOF), according to climate scientist Miriam Jackson, a glaciologist with extensive Himalayan experience. It began with an ice-rock collapse that rapidly transformed into a debris flow before entering a river and producing a devastating downstream flood.
“This event was not a GLOF, but an ice-rock collapse that fluidised and became a debris flow and then went into a river,” said Jackson, IPCC lead author in the High Mountain Areas chapter in the IPCC Special Report on Oceans and Cryosphere in a Changing Climate (SROCC) and former ICIMOD cryosphere lead, to TNIE.
The distinction is critical because GLOFs are comparatively better understood and monitored than sudden glacier collapses and ice-rock avalanches. “Such events are a lot harder to predict. They are relatively unknown, so not well understood,” Jackson said.
Farooq Azam, Senior Cryosphere Specialist and Intervention Manager at ICIMOD, said, “This was an unprecedented and unexpected disaster. There was simply no detection and no time to issue a warning.”
“The event was a sudden-onset hazard that could not be captured by conventional warning systems. However, if there had been an early warning system, it could have saved lives to some extent. But in such rock-ice avalanches, establishing a warning system is very difficult to ascertain the place of the breakpoints, which can sometimes be below the bedrock or glaciers and not easily visible with satellite,” Azam added.
The latest disaster therefore raises a question far more fundamental than whether Nepal has mapped enough dangerous glacial lakes: Is monitoring glacial lakes alone enough to prepare for the next Himalayan disaster?
The answer may lie partly in Nepal’s own flood history.
An analysis of the Global Glacier Flood Database shared with The New Indian Express identifies 41 glacier-lake or glacier-related flood records in Nepal. But the database also reveals how incomplete the historical understanding of these hazards remains. Only 16 of the Nepal records have a reasonably identifiable exact year, while many others are poorly dated and, in several cases, the trigger is unknown.
Of the recorded events, 27 involve moraine-dammed lakes and 13 involve supraglacial lakes. The records are concentrated in major Himalayan river basins, including Humla, Tamor and Dudh Koshi.
The gaps are not necessarily evidence that scientists failed to investigate the events. Many occurred in remote and inaccessible terrain, making it difficult to determine what actually triggered them.
Jackson said determining the trigger of a GLOF can itself be difficult.
A lake may gradually fill until water pressure overwhelms the opposing pressure from ice or a moraine, but an external event, such as a landslide, rockfall, avalanche, heavy rainfall or thawing permafrost, can suddenly initiate a flood.
“Many of these events happen in remote places, so it is not always possible to go up and take a look,” she said. Even with high-resolution satellite imagery, she added, determining exactly what happened can be difficult.
The historical record is therefore “most likely incomplete”. But the problem exposed by the August 26 disaster is different. A dangerous glacial lake can be mapped. A mountain that is about to collapse is much harder to identify.
Jackson said glacier collapses and ice-rock falls are less common than GLOFs, but potentially much harder to predict. The 2021 Chamoli disaster in Uttarakhand and the 2025 Blatten event in Switzerland are examples of similar glacier-collapse hazards.
Blatten also demonstrates what monitoring can achieve when warning signs become visible. There, increased rockslides and rockfalls around a glacier prompted intensified monitoring and the evacuation of people living nearby before the major ice-rockfall occurred.
But replicating that model across the Himalaya is extremely difficult. “It is hard enough to try and monitor glacial lakes, but to monitor all mountains for increased landslides, would be impossible,” Jackson said.
Some glaciers may show signs of acceleration before collapsing, including the development of additional crevasses. But those signs have to be detected and continuously monitored.
“There just isn’t the resources to do this,” she said.
That resource problem is becoming more urgent as the Himalayan cryosphere changes.
The database does show a history of repeated glacier-related flooding. It records, for instance, repeated events associated with the Halji glacier in Humla between 2004 and 2011, while other documented Nepal events have been linked to ice avalanches and rockfalls.
Yet Jackson cautions against reading the historical record as straightforward evidence that such floods are becoming more frequent. Part of the apparent increase, she said, is simply because scientists are now much better at detecting, attributing and recording such events.
In the past, “a flood was just a flood”; it was often impossible to determine whether it had been caused by a glacial lake, a glacier collapse or another mountain process.
But that does not mean the underlying risk is static. “Measurements certainly show that there are more glacial lakes and they are getting bigger,” Jackson said. “So reasonable then to believe that there will be more GLOFs.”
That creates a paradox for Himalayan disaster management. The region needs better monitoring of glacial lakes even as the latest disaster demonstrates that monitoring lakes alone is not enough.
Meanwhile, Nepal has renewed calls for stronger real-time information sharing with China, including better access to information on glacier movement and potential avalanches. The disaster has exposed the importance of transboundary early-warning systems because a hazard originating high in the mountains can cross a national border and reach communities downstream within hours. In fact, in this case, it was a matter of minutes from when the seismic signal generated by the ice-rock fall was recorded until it reached the Rasuwagadhi-Gyirong crossing.
(This reporter is a resident fellow of the Climate Change Media Hub at the Asian College of Journalism, Chennai.)