Watching the Lakes While the Mountains Fall: Why Himalayan Early Warning Is Built for Yesterday's Disaster

The Himalaya's Warning Successes Have Always Been Lake Stories and the Next Disaster May Not Have a Lake In September 1893, a

massive landslide blocked the Birahi Ganga in Garhwal and created what came to be called the Gohna Tal. The colonial administration could see the new lake filling and could estimate roughly when the debris dam would fail.

It strung a telegraph line down the valley and moved people off the riverbanks. When the dam gave way in August 1894 and the flood tore down the Alaknanda, the warning had already done its work. This is not new. For more than a century, the Himalayan disasters that have been successfully warned about have shared one feature: a body of water in a known place that could be watched.

Modern GLOF monitoring works the same way. A glacial lake can be mapped from satellite, modelled, gauged and in principle drained. The October 2023 South Lhonak outburst in Sikkim, which destroyed the Teesta III dam at Chungthang, was devastating.

It was also a lake event, the kind that current systems are designed to understand.

The disaster type the system was not built for Chamoli (2021) and Rasuwa (2026) were not lake events.

Each began high on a peak (Ronti in Uttarakhand, Langtang Lirung in Nepal) where thawing permafrost and retreating ice left a rock-and-ice mass unstable. It collapsed, hit the valley floor with enough force to register as an earthquake and turned into a fast-moving debris flow. There was no lake to model, no dam to gauge and no water level to watch. Dharali's 2025 debris flow showed the same destructive power arriving from above.

Here the Gohna parallel breaks down. In 1893 the danger was a single lake and people watched it fill for eleven months. Today the potential sources are hundreds of thousands of steep, deglaciating slopes. Detecting that some are moving may be possible.

Predicting which one will fail catastrophically is not yet reliable and a false alarm can cost as much trust as a missed one.

Why this is a land-use and trust problem, not only a sensing problem A denser seismic network could detect these flows as they rumble down valleys, but it would need cross-border cooperation and tens of millions of dollars. Even then, it would give downstream villages minutes of warning and upstream settlements almost none. Warning works only where communities trust it and know what to do. That leads to the harder question: who is in the path? In both Chamoli and Rasuwa, many of the dead and missing were hydropower workers. Raini, the village where the Chipko movement took shape in the 1970s, was hit again in 2021 when the Rishiganga project upstream was swept away.