The ice-rock avalanche that tore off Langtang Lirung and sent a wall of debris into the Lhende Khola last week was not, in any meaningful scientific sense, a surprise. It was a catastrophe.
At least 1000 people are dead and thousands are still remain missing along the Nepal–China border.
Water in the Bhotekoshi rose seven to nine metres in half an hour. But the mechanism, heat weakening the ice that binds fractured high-mountain rock, a glacier slumping, a torrent racing down a steep valley is precisely the sequence climate scientists have been describing for years.
That is the uncomfortable fact this disaster forces into the open. The event is being treated as a shock. For those who study the cryosphere, it is closer to a confirmation.
What the Warnings Actually Said:
The 2019 Intergovernmental Panel on Climate Change (IPCC) Special Report on the Ocean and Cryosphere was already blunt. Glacier retreat and permafrost thaw have decreased the stability of mountain slopes.
Resulting landslides, floods and cascading events would appear even where there was no historical record. The 2023 IPCC Synthesis Report went further; every additional increment of warming multiplies hazards from ice-bound regions. Floods, landslides and freshwater shortages from glacier retreat “pose a serious threat to mountain regions across the world.”
Dr Richard Waller, a physical geographer at Keele University who studies mountain ice and rock, offered the most useful image: high mountains as shattered bedrock glued together by ice-filled joints. When that glue thaws, the rock does not politely stay put.
Dr Hamish Pritchard, a glaciologist at the British Antarctic Survey, put the same process in operational terms high temperatures weaken the snowpack, fill crevasses with water and thaw the bonds between ice and rock that hold glaciers in place. More avalanches and more floods, he said, are inevitable as lakes grow in front of retreating ice. Early warning systems are urgently needed.
None of this is new. Alpine researchers, including Christian Huggel of the University of Zurich, Wilfried Haeberli, a leading permafrost specialist, and Stephan Gruber of Carleton University, were arguing more than a decade ago that permafrost degradation would produce slope failures unexpected in location, magnitude and timing.
The same logic applies from the Andes to the Caucasus to High Mountain Asia. Glacier “debuttressing”, the ice that once propped up a valley wall simply disappearing. This adds another, slower form of destabilisation. In the Langtang catchment itself, glacier loss rates have already increased more than fourfold since 1964.
Why the Next Events Eill Be Worse, Not Rarer
Two things are happening at once. First, the physical ingredients of disaster are becoming more abundant; more and larger glacial lakes, more thawed permafrost, more rain falling where snow once fell, more water in already fractured rock. Second, people and infrastructure have moved into the runout zones. Exposure has risen even as the hazard has intensified.
The numbers attached to Glacial Lake Outburst Floods (GLOFs) make the scale plain. Around ten million people already live in the path of such events, concentrated in High Mountain Asia, Alaska and Iceland. Without deep emissions cuts, the frequency of these floods in Asia is projected to roughly triple by the end of the century.
Many of the coming failures will not look like a textbook lake burst. They will be hybrid events ice and rock collapsing together, a debris flow becoming a flood, a flood becoming a dam-break further down the valley. Those cascades are harder to forecast and harder to evacuate.
There is a longer, quieter danger behind the sudden ones. After a period of increased meltwater, many glacier-fed rivers will pass “peak water” and then decline. Communities that now fear floods will, in time, face water shortages.
The IPCC has already identified hard adaptation limits in high-mountain and downstream regions if warming substantially exceeds 1.5°C. Mountains warm faster than the global average. For the Hindu Kush Himalaya, 1.5°C globally is closer to 2°C on the peaks. That is not a distant scenario. It is the neighbourhood we are already in.
Pakistan, And the Rest of the Third Pole
The Langtang collapse sits in one part of a much larger system. The Hindu Kush–Karakoram–Himalaya (HKH) — sometimes called the Third Pole, stores more ice than anywhere outside the Arctic and Antarctic and feeds rivers on which nearly two billion people depend. Pakistan sits on the western flank of that system.
Its northern parts Gilgit-Baltistan, Hunza, Chitral, Swat, Upper Dir contain thousands of glacial lakes. Dozens have been classed as potentially dangerous. Historical outburst floods in Hunza Nagar alone number in the dozens over recent decades. The Shisper glacier lake flood of 2022 was a preview, not an outlier.
The vulnerability is not only physical. Settlements have expanded along streams fed by those lakes. Early-warning coverage is incomplete. Pakistan contributes well under one per cent of global greenhouse-gas emissions, yet it is repeatedly ranked among the countries most exposed to climate-driven disasters. The 2022 monsoon floods already showed how glacier and snow-melt signals can combine with extreme rainfall.
A large GLOF or ice-rock avalanche on a major tributary of the Indus would not stay a mountain story. It would become a plains story — of washed-out roads, damaged hydropower, silted canals and interrupted water supply.
The same map of risk includes Nepal and the Tibetan Plateau, Sikkim and Uttarakhand in India, Bhutan, and parts of Afghanistan. It also includes the European Alps, where permafrost thaw has already brought down peaks and buried a Swiss village; the tropical Andes, which have lived with GLOFs since the 1940s; and smaller ranges from the Caucasus to New Zealand. The geology differs. The physics of thawing ice in fractured rock does not.
What Follows From a Predicted Disaster
An honest assessment should not pretend that every rockfall is “caused by climate change” in the simple sense of a single trigger. Earthquakes, monsoon saturation and the inherent instability of steep Himalayan slopes all matter.
Formal attribution of any one collapse takes time. That is not the same as saying the trend is mysterious. The trend of more heat, less ice, weaker slopes, larger lakes, more people in the way is the one scientists described.
The practical response has two layers that should not be played off against each other. The first is adaptation that can start now; satellite monitoring of lakes and slopes, community sirens that actually work, restrictions on building in obvious runout paths, and transboundary data-sharing on rivers that do not respect borders.
Switzerland’s evacuation before the 2025 Blatten collapse showed that monitoring can save lives when the political will and the instruments exist. Most of the HKH does not yet have that combination.
The second layer is the one that determines how many such systems will be needed. Every fraction of a degree that is still avoidable reduces the number of slopes that lose their ice glue, the number of lakes that overtop, and the number of valleys that will have to live with a new, more violent hydrology.
The Langtang disaster is not proof that warning was impossible. It is proof that warning without action is a kind of negligence- scientific, political and moral.
The mountains have not suddenly become dangerous. They have become less glued. Scientists said that would happen. The question now is how many more valleys have to learn it the hard way.
