Day two changed the diagnosis before it settled the toll. USGS now says event us7000tbwb was a glacial collapse and debris flow, with no tectonic earthquake behind the seismic signal [E1]. Analysis of nearby stations, long-period waves and satellite imagery drove that reclassification [E1]. The working sequence now starts with the mountain giving way; the seismic pulse records that failure [E1].
By 16:00 Thursday, Nepali authorities had recovered 289 bodies from the flood-hit corridor [E2]. NDRRMA listed 826 people missing, including 579 tourists and personnel from the police, army, customs office and Armed Police Force [E3]. Those figures describe different administrative states: recovered dead and people still unaccounted for [E2][E3]. Combining them into a single death toll would overstate what Nepal had confirmed at that cutoff [E2][E3].
Across the border, Gyirong kept its own ledger. Xizang authorities had three confirmed dead and 558 missing at 08:00, including 260 foreign nationals [E4]. That count was compiled for the port area on a different jurisdictional sheet from Nepal’s NDRRMA tally [E3][E4]. No reconciled cross-border list had been published in these updates, so the two counts remain separate [E3][E4].
Mechanism now lines up with the river geography. Satellite analysis places the debris in the Lhende, where it produced a debris-laden flood after a large avalanche near Rasuwagadhi [E5]. On 26 August, Galchhi on the Trishuli rose as much as nine metres in 30 minutes, while Malekhu gained about seven metres over the same interval [E6]. Taken together, the reclassified seismic trace, debris path and river rise support a cascade: high-altitude ice and rock collapse, possible temporary blockage, then a sediment-heavy surge through the Bhote Koshi into the Trishuli [E1][E5][E6].
Thursday’s damage accounting sharpened the downstream ledger. The Department of Roads put preliminary damage to swept roads and bridges at Rs 20 billion [E7]. Among the 27 August public updates cited here, none supplied a revised NEA generation-loss total [E7]. The hydro loss therefore remains carried at the roughly 430 MW class from 26 August coverage.
Upstream risk has not frozen with the first surge. A barrier lake near the Chotchen–Purupe Tsangpo confluence in Tibet held about 2 million cubic metres by Thursday morning, with water already flowing out [E8]. Another 3 million cubic metres could enter over the following three days [E9]. Rescue and repair crews are therefore working beneath a second hydrological clock while the original collapse sequence is still being reconstructed [E8][E9].
Search crews are working a corridor whose trigger is better constrained than its casualty ledger [E1][E3]. The 289 recovered bodies are a 16:00 count, and the 826 missing are another live sheet [E2][E3]. Gyirong still carries its separate 558-person missing list [E4]. The event first labeled an earthquake is now, in the USGS record, the seismic signature of a mountain coming down [E1].