Rasuwa Flood · 26 August 2026 · Bhote Koshi–Trishuli corridor

figures 01 Sep 2026 · WGS 84
08:37 NPT+0 min
Source zone — collapse

Damage at this reach

Legend

Water
Flood flow path
Observed extent · HOT
Modelled peak inundation
Trigger
Collapse origin
Barrier lake (GLOF)
Damage
Bridge washed out
Bridge damaged
Bridge standing
Hydropower
Detected disturbance
Glacier / rock collapse → flash flood

The flood, minute by minute

Press ▶ on the map — the surge travels the modelled geo-pera arrival times and each stop lights up below with a satellite view of the place.

Flood-wave timing (modelled)

SitekmNPT+minPeak

geo-pera 2D shallow-water model (V ≈ 100 Mm³).

Use the slider below the image to wipe between Vantor / Maxar Open Data imagery from before and after the flood. Same frame, same scale.

After the flood Before the flood
Before After
Before After

Imagery released for this event under CC BY-NC 4.0 through the Vantor (Maxar) Open Data Program, via OpenAerialMap. Pre-flood frames are the most recent archive pass for each site.

People & response

Exposed population by district

2021 census population as an exposure proxy — Rasuwa & Nuwakot are the severely-hit corridor.

Infrastructure (HOT survey)

Economic loss

Five bodies assessed this flood with different data and methods, on different footprints. They mostly agree — where numbers differ it is because they measure different things.

Why the numbers differ

    Strongest result: bikal3's flood corridor, built from a 30 m elevation model alone, contains 90.8% of HOT's independently-mapped flood extent and 100% of the washed-out bridges.
    Hazard
    Flood flow path
    Observed extent — HOT
    Modelled peak inundation
    Terrain HAND corridor (bikal3)
    Collapse zone / barrier lakes
    Damage & exposure
    Satellite-detected damage
    Damaged / destroyed buildings
    Damaged roads
    Bridges (HOT survey)
    Hydropower facilities
    Impact zones Z1–Z5
    Settlements (OSM)
    Base map (pick one)
    Light gray
    Streets
    OpenStreetMap
    Topographic
    Terrain + hillshade
    Satellite
    Dark
    Administrative
    District boundaries

    What this is

    A rapid multi-method situational assessment of the 26 August 2026 Rasuwa (Bhote Koshi → Trishuli) flood. It runs no new analysis; it collects five independent published assessments, puts their vectors on one map in WGS 84, and reconciles their headline numbers. Secondary data only, no field survey, not affiliated with NDRRMA, ICIMOD, UNOSAT, HOT or the UN.

    Sources, per layer

    • Collapse origin, 1.96 km² detachment, barrier lakes — UNOSAT / UNITAR (Landsat-9 26 Aug, Cartosat-3 28 Aug). Exposure figures (~43 km² mudflow, ~5,000 buildings, ~130 km roads, ~1,300 ha cropland) from the UNOSAT ArcGIS Experience.
    • Observed flood extent (31.7 km²), bridges, buildings, roads — Humanitarian OpenStreetMap Team, hot_flood_npl via HDX (ODbL).
    • Flow path, modelled peak inundation, wave-arrival times — geo-pera/bhotekoshi-2026-reconstruction (MIT), calibrated 2D shallow-water model.
    • Terrain HAND corridor, detected damage, impact zones, validation — bikal3/rasuwa-flood. No licence stated — attribution only; regenerate from its pipeline to redistribute. Live: rasuwaflood.bikal3.com.np. Site pattern also draws on bikal3/himalaya-glof.
    • Casualties, camps, economic loss — NDRRMA situation reports; IFRC / Nepal Red Cross.
    • District boundaries — OCHA/HDX Nepal Common Operational Dataset (admin-2), simplified.
    • Basemaps — CARTO Positron; Esri World Imagery.

    Reproduce

    The working folder holds every source under data/sources/ with a data/MANIFEST.md. scripts/fetch_sources.sh re-pulls them; scripts/build_data.py reprojects, simplifies and writes web/app-data.js, which this page reads. No figure is typed into the page.

    Limitations

    The five assessments cover different footprints. bikal3 is unlicensed and its zone coordinates past Z1 are approximate. geo-pera's depths are model output. HOT's observed extent covers only the surveyed corridor. Casualty / damage figures are provisional (snapshot 2026-09-01). Simplification adds ≤ ~30 m edge error. The animation interpolates the modelled arrival keyframes — it is an illustration of timing, not a hydrodynamic playback.

    Key references

    ⚠ Provisional. Compiled from public secondary sources for situational awareness and analytical demonstration. Verify against the latest NDRRMA SitRep before operational use. Layer licences: OSM/HOT ODbL; geo-pera MIT; bikal3 unlicensed (attribution); UNOSAT & Copernicus free re-use with attribution.