USGS says a glacial collapse, not an earthquake, set off the Nepal-Tibet flood

The United States Geological Survey has said the disaster on the Nepal-Tibet border was a glacial collapse and debris flow, not an earthquake that then caused a landslide. The falling mass itself produced a magnitude 5.2 seismic signal at 08:37 local time on Wednesday, 26 August. First reports from Kathmandu had pointed to a tremor. The sequence, on the evidence now published, runs the other way.

Daniel Shugar of the University of Calgary, reading Planet Labs images, described a piece of ice of about 0.2 square kilometres that broke from roughly 5,200 metres and fell about 1,200 metres to the valley floor. Jakob Steiner of the University of Graz put the drop from about 5,100 metres to 3,000 metres and said the rock under the glacier tongue failed first, leaving the ice with nothing to sit on. Simon Cook of the University of Dundee placed the failure on the lower part of a glacier near Langtang Lirung, some 15 kilometres west of the worst flood damage, with the debris travelling northwest and then west.

ICIMOD, the Kathmandu-based mountain centre, described an ice-rock avalanche from high ground that dumped a large volume of ice and rock into the Lende Khola, a tributary of the Bhote Koshi. Water, sediment and boulders then moved down the Trishuli-Bhotekoshi system. River gauges recorded a rise of seven to nine metres within 30 minutes. Several monitoring stations were wrecked or washed away. Prakash Pokharel of Nepal's National Disaster Risk Reduction and Management Authority called it a glacier-inclusive flood caused by an ice-rock avalanche on the Nepali side.

Chinese satellite images provided to Kathmandu show that the debris briefly dammed the Lhende Khola. When that dam broke, the stored water added a second pulse. That is why the flood could travel more than 100 kilometres and why a second barrier lake now sitting near the Chhochen Khola and Purepu Tsangpo confluence is the immediate fear. China has warned that another three million cubic metres could flow into that lake over three days, with peak risk around 1 September.

Death and missing counts have moved every few hours. By Friday morning public tallies put the dead near 390 to 469 and the missing above 1,400, including hundreds of foreign visitors and at least 288 to 290 Indian nationals whom New Delhi still could not reach. The numbers will change again as roads open. The cause is less likely to change. USGS, ICIMOD and the university groups are describing the same mechanics with different instruments: seismometers, optical satellites and river gauges.

The Himalaya is warming faster than the global mean. Cirque and hanging glaciers on steep faces fail when the rock below them loses strength or when meltwater lubricates a plane of weakness. Those failures do not need a monsoon cloudburst on the same morning. Wednesday's sky over the border was described as clear. That fact confused early official comments. It fits a slope failure better than a rain flood.

Warning systems in this catchment were built around earthquakes and around known glacial lakes. A bedrock-and-ice collapse that creates a lake in minutes sits between those two plans. The next useful public product is a map of which hanging glaciers on the Nepali and Tibetan sides have similar geometry. Until that map exists, the second dammed lake is the only alert that matters this week.