Warming is changing the risk landscape in the Himalayas
Preliminary analyses indicate that a massive rock-and-ice avalanche triggered the devastating flood along the Nepal–Tibet border on 26 August 2026. “The precise sequence of events has yet to be established. Even so, the disaster shows how warming, unstable high-mountain terrain and growing development in downstream valleys can combine to increase risk,” says Arnkjell Løkke, Head of Dam Safety at NGI.
A violent surge of water, mud and rock reaches the border facilities at Gyirong Port on the Tibetan side of the border on 26 August 2026. Preliminary analyses indicate that the flood was triggered when enormous volumes of rock and ice collapsed from Langtang Lirung in Nepal. ( Still image from automated CCTV footage / Wikimedia Commons, public domain.)
Arnkjell Løkke has explained the catastrophic event in Nepal to Norwegian media including NRK and Aftenposten. For decades NGI has worked on landslides, floods and unstable rock and glacier systems in Norway and internationally, including in the Himalayas. The institute develops hazard assessments, monitoring and early warning for areas where natural forces meet settlements and infrastructure.
“What has now happened is far more than an ordinary flood. The most likely explanation is a rapid cascade in which an enormous volume of rock and ice broke away, continued down the valley at high speed and mobilised huge volumes of water and debris. When processes like these interact, the consequences can unfold extremely quickly,” says Løkke.
Within minutes, a mountain valley was struck by a violent surge of ice, water, rock and mud. Villages, roads and other infrastructure were destroyed, with a severe loss of life.
Satellite and drone imagery points to the collapse of large volumes of rock and ice high on Langtang Lirung. Scientists around the world are still investigating the size of the avalanche and the exact mechanism behind the flood.
Langtang Lirung seen from Tsergo Ri in Nepal. The steep slopes, glaciers and fractured high-mountain terrain illustrate how rock, ice and meltwater combine to create a complex risk landscape. The photograph was taken in 2000 and does not show the avalanche area after the disaster. ( Photo: Ahtih / Wikimedia Commons, CC BY-SA 3.0)
A changing high-mountain landscape
Glacier-related disasters are not new. The Himalayas, the Andes, the Alps and the Caucasus all have historical records of sudden glacial-lake floods and large ice avalanches.
What is changing are the conditions in which such events occur. The Himalayas are warming faster than the global average. Glaciers are retreating and thinning, while meltwater is accumulating in a growing number of lakes held back by natural dams of ice and loose debris. High-altitude permafrost is also thawing. As the ice that helps bind fractured rock together disappears, and meltwater penetrates deeper into cracks in the rock mass, steep mountain slopes can become less stable.
Climate change is part of the risk picture
“We cannot conclude that climate change triggered this particular disaster. That would require a dedicated study, and the immediate cause is still being investigated. But the long-term trend is well documented: warming is changing the underlying conditions in high mountains and may increase both the frequency and the magnitude of events like this,” says Løkke.
Risk is not determined only by how often an avalanche or flood might occur. It also depends on what lies in its path. More people now live in exposed valleys, while roads, hydropower facilities and other critical infrastructure have been built along rivers.
“The hazard may grow as glaciers and mountains change. At the same time, the consequences may become greater because more of society is exposed. We need to understand both sides of this in order to map the risk picture,” says Løkke.
Blatten in the Lötschental valley in Switzerland, around two months after a massive rock-and-ice avalanche struck the village on 28 May 2025. The debris dammed the River Lonza and flooded part of the settlement. Close monitoring and early evacuation enabled around 300 residents to reach safety. One person outside the evacuation zone was killed. ( Photo: Beat Ruest / Wikimedia Commons, CC BY 4.0)
Early warning can save lives, but not everywhere
Many warning signs can be detected using existing technology. Satellites can track glacier movement and identify new or expanding glacial lakes. Radar can measure slow deformation in mountain slopes, while seismic networks can detect large avalanches shortly after they begin.
When large volumes of rock and ice threatened the Swiss Alpine village of Blatten in May 2025, close monitoring, local observations and cooperation between scientists and public authorities enabled residents to be evacuated. Much of the village was later buried, but all of the roughly 300 evacuated residents survived. One man was nonetheless killed: a shepherd who was in his barn outside the evacuation zone when the avalanche proved larger and travelled further than the hazard assessment had assumed. NGI researcher Benjamin Bellwald, who is from Blatten, has highlighted clear communication and close coordination between local, regional and national stakeholders as decisive in getting the evacuation carried out in time.
Conditions in the Himalayas are far more challenging. The mountain range covers a vast area, many potential hazards are extremely remote, and continuous monitoring is both technically demanding and costly. Even where an early-warning system exists, an extreme event may develop so rapidly and with such force that the system is unable to function as intended.
“Technology can give us valuable time, but it cannot offer a guarantee. Monitoring must be combined with hazard maps, local knowledge, preparedness plans and clear lines of responsibility. A warning only saves lives when people understand it, trust it and know how to respond,” Arnkjell Løkke concludes.
The disaster in the Himalayas therefore demonstrates both the potential and the limits of modern early-warning systems. The aim cannot be to monitor every slope or predict every event. The priority is to identify where the consequences could be greatest, track changes over time and ensure that knowledge leads to action.
Arnkjell Løkke
Head of Section Dam Safety Dam Safety and Hydrodynamics arnkjell.lokke@ngi.no+47 480 48 843
