The disaster on the Nepal–Tibet border this week began with a phrase most people had never needed before: a glacier collapse. It is worth understanding what that means, because the term covers several different things, and the differences decide whether anyone could have been warned.

Three different disasters with similar names

The first is a glacial lake outburst flood, usually shortened to GLOF. Meltwater collects in a lake held back by loose rock and ice left behind by a retreating glacier. The barrier gives way, and the lake empties downhill in minutes. This is the best-studied hazard of the three, because the lakes are visible from orbit and can be inventoried and ranked by risk.

The second is a rock-ice avalanche: a section of glacier, together with the rock it is frozen to, detaches from a steep face and falls. There is no lake and no gradual build-up. Scientists studying this week’s event, including researchers writing for the American Geophysical Union, have described it in these terms rather than as a lake outburst.

The third is what happens next. When millions of tonnes of ice and rock hit a river valley, the mass picks up water, mud, boulders and trees and becomes a debris flow — something closer to wet concrete moving at speed than to a flood of water. Debris flows are what destroy bridges. Water alone rarely does.

Why it registered as an earthquake

Seismometers thousands of kilometres away recorded a signal from the collapse, and early reports described it as an earthquake of around magnitude 4 to 5. Officials in the region repeated that.

Landslide specialists say this is a familiar misreading. A seismometer measures ground shaking; it does not know what caused it. A very large mass falling several thousand metres and striking a valley floor produces a signal that looks, at first glance, like a tectonic event. The waveform is different on closer inspection — it builds and fades over a longer period rather than arriving as a sharp break — but that analysis takes hours, and the automatic catalogue entry appears in minutes.

The practical consequence is that the first official explanation of a disaster like this is often wrong, and is corrected within a day or two. That is not a cover-up. It is the difference between an automated instrument reading and an interpretation.

Is this climate change?

This is where care is needed, because the honest answer sits between two confident ones.

Glaciologists who study the Langtang area have published work showing those glaciers are warming and retreating, and that the rate of retreat has been accelerating. Across the wider Himalaya, researchers have documented both warming at high altitude and the thawing of permafrost — the frozen ground that, in effect, glues steep rock faces together. The general mechanism by which a warming mountain range becomes less stable is well established in the literature and is not seriously disputed.

What has not been established is that this particular collapse, on this particular day, was caused by climate change. Researchers investigating the site have said a direct link has not yet been demonstrated. Individual slope failures also happen for reasons that have nothing to do with temperature: the geometry of the rock, an unusually wet monsoon, or simple accumulated fatigue in a face that would eventually have failed anyway.

Attribution studies exist precisely to answer this question, and they take months. Anyone offering a definitive answer this week — in either direction — is ahead of the evidence.

Why there was no warning

Early warning systems for glacial lakes exist and work: sensors at the lake detect a sudden drop in water level and sirens sound in the valleys below. Nepal has installed such systems on several high-risk lakes.

A rock-ice avalanche defeats that model in two ways. There is nothing to monitor in advance — no lake filling up, no measurable trend — and the warning time is close to zero. The mass in this case travelled roughly 100 kilometres. The settlements nearest the source had minutes at most.

Specialists in the field argue that the realistic protections are therefore not sirens but land use: knowing which valley floors are debris-flow paths, and being careful about what gets built on them. That is a slower and less satisfying answer than a warning system, and it runs into the fact that in steep country the flat ground beside the river is usually the only place to build. Researchers who study the trade-off are generally clear that mapping hazard zones is easier than persuading anyone to move out of them.

Our running coverage of the disaster: Nepal’s flood toll passes 500 as a new lake forms above the valley.

Sources

United States Geological Survey landslide hazards programme; Eos (American Geophysical Union) landslide blog; published research on Langtang glacier retreat; Al Jazeera; BBC; EarthSky.