USGS Called It an Earthquake. It Was a Glacier Collapsing.

The NDMA's GLOF Programme Was Built for the Wrong Hazard India's National GLOF Risk Management

Programme was designed after the 2021 Chamoli disaster to do one specific job: monitor glacial lakes for signs of an imminent moraine-dam breach and warn downstream populations before the water arrives. That is a coherent response to a Glacial Lake Outburst Flood in the textbook sense - a lake fills, its debris dam weakens and it fails.

The Nepal flash flood that killed at least 389 people this week did not happen that way and the confusion around its cause is the more important story. On Wednesday, the US Geological Survey said a 4.4-magnitude earthquake had triggered the flood.

By Thursday, it had reversed that finding entirely: "additional analysis of long period seismic waves indicate that the seismic energy was instead generated by a glacial collapse and debris flow," now measured at a 5.2-magnitude equivalent.

Glaciologist Anil Kulkarni of IISc's Divecha Centre described a mass of ice and rock falling 1.5-2 km into a deglaciated valley - ground once covered by ice, now exposed moraine and debris - with the impact acting, in his words, like "a high-intensity bomb blast." This is not a footnote.

It means the trigger mechanism was not a lake failing but a glacier itself failing - snapping off, dropping onto loose debris and generating enough force and displaced material to dam and then release a river within hours.

A monitoring system built to watch lake levels and moraine dams is not built to watch hanging ice masses for structural failure. Those require different sensors, different terrain models and different warning thresholds.

India's own record shows this gap already exists in overlapping form. Chamoli in 2021 involved a rock-and-ice avalanche, not a classic lake breach. The 2023 South Lhonak Lake disaster in Sikkim was closer to the textbook GLOF the programme was designed around.

Nepal's event sits in between - closer to Chamoli than to Sikkim. Three events, three different physical mechanisms, one monitoring programme still oriented primarily toward the third type. NDMA member Krishna Swaroop Vatsa has confirmed the authority will study the Nepal event "in detail," alongside parallel efforts on landslide forecasting.

That is the right instinct and it should be taken further: the GLOF programme's early warning architecture across Arunachal Pradesh, Sikkim, Uttarakhand, Himachal Pradesh, Jammu and Kashmir and Ladakh needs an explicit hanging-glacier and ice-avalanche monitoring layer, not an assumption that lake-focused sensors will catch every cascading hazard upstream of them.

ISRO's satellite data already shows 676 of India's large glacial lakes have expanded since 1984 - the lake-monitoring case is well established. What is not yet established is equivalent tracking of unstable hanging ice masses feeding into those same valleys.

For an aspirant, the exam-relevant point is not "glaciers are melting"

• it is that disaster-risk governance built around one failure mode can leave a structurally similar but mechanically different hazard uncovered, even inside a programme that looks comprehensive on paper.