TRACE THE
FAILURE
CASCADING NATURAL HAZARDS

Nepal’s mystery earthquake.
It wasn’t an earthquake.

A seismic signal detected the movement. It took further analysis to identify the landslide and the flood that followed.

Follow the evidence
Illustrative snow-covered Himalayan peaks beneath clouds
AI-generated Himalayan landscape from the episode. Not an image of the confirmed failure location or an exact reconstruction of the landslide.
THE COMPANION FILMNepal's 2026 Flood Disaster: What Really Happened 4,000m Up Watch episode 5:08

What happened in Nepal?

On August 26, 2026, a rapid slope failure involving glacier ice near the Nepal–China border developed into a debris flow and flood. USGS reports that the flow traveled approximately 100 kilometres along the Lende Khola and Trishuli river system. Although its seismic signal was equivalent to magnitude 5.2, the reviewed USGS record classifies the event as a landslide, not a tectonic earthquake. USGS findings; reviewed event record.

The seismic signal came from moving mountain material; it did not prove that an earthquake had triggered the flood. USGS identified a glacial collapse and debris flow after further analysis of long-period seismic waves and satellite imagery. USGS event record.

Watch the explanation

Follow the mountain-to-river sequence in the original episode. The companion article includes a September source update.

What the September update adds

Our video reflects the developing August record. This written version also incorporates ICIMOD’s September 16 account, which describes bedrock collapse triggering an ice-rock avalanche on Langtang Lirung’s north-facing slope. ICIMOD proposed calling it the Lirung ice-rock avalanche. September update.

USGS’s separate response page still contains its August 28 uncertainty about whether rock failed first or glacier ice collapsed. The page says it will no longer be updated. The reviewed seismic-event record uses the more specific glacial-collapse-and-debris-flow description, while ICIMOD’s September account adds later analysis. USGS response page; event record.

These later accounts do not provide a complete, measured reconstruction of every stage, including an exact released volume, dam height or flood speed.

How rock and ice became a far-traveling flood

USGS describes debris containing water and rapidly melting ice. Moving downhill, it picked up material and water from streams and river channels. This process, called entrainment, added to the mixture beyond the material in the original release. USGS findings.

The approximately 100-kilometre figure describes the downstream debris-flow and flood corridor, rather than the journey of one intact glacier block. As the mixture changed along its route, a localized slope failure became a hazard far downstream.

ICIMOD’s resource page identifies an ice avalanche, rather than a glacial-lake outburst, and includes an expert account of temporary river blockage. It does not establish the blockage’s exact dimensions or a timed sequence of filling and breaching. ICIMOD resource page.

What the evidence means for warning systems

A seismic network, satellite imagery and river gauges describe different parts of a cascading hazard. Responders need to connect those observations to understand what a mountain failure may send along the rivers.

ICIMOD’s September account describes an international meeting focused on reconstructing the event, monitoring mountain hazards and improving warning and response. It anticipated an action plan. That announcement is evidence of coordination work, not proof that a new warning system has already been installed. ICIMOD, September 16.

The available record describes a rock-and-ice failure that became a channelized debris flow and flood. The seismic signal helped reveal the event, while understanding its downstream effects required other observations. These accounts are not a final inquiry report.

The source record

USGS: Nepal debris avalanche and flash flood

Findings as of August 28, 2026. The page explicitly says it is provisional and is no longer being updated.

USGS reviewed event record: us7000tbwb

Event type: landslide. General-text and header products identify a glacial collapse and debris flow; retrieved September 21 locally.

ICIMOD: September expert meeting on the Lirung ice-rock avalanche

September 16, 2026 · Opening account, 15–17 September meeting and proposed action plan.

ICIMOD: Gyirong–Rasuwa flood resource page

Dated resource collection and attributed expert summaries; not a final forensic investigation.

Research checked September 21, 2026. Illustrative episode images are not primary evidence.