Nepal buildings and infrastructure data

Part of the Humanitarian OpenStreetMap Team response activation for the 26 August 2026 Bhote Koshi and Trishuli flash flood in Nepal. AI-scored building damage per OpenStreetMap footprint across the upper Trishuli and Bhote Koshi corridor, from Rasuwagadhi to the Narayani confluence. Post imagery is a 26-scene Vantor open-data composite (2026-08-27 to 2026-09-08, roughly 0.4 m/px, about 9.8% cloud). Pre imagery is ESRI World Imagery. Building footprints come from OpenStreetMap and cover 20,367 features in the corridor; 8,421 fall inside the analyzed area and carry an AI damage class. Each feature carries `osm_id`, `osm_type`, `building`, a `damage` label (`no-visible-damage`, `minor-damage`, `major-damage`, `destroyed`) and a `damage_confidence`. Evaluated and validated against 3,896 OSM manually tagged affected buildings (3,185 destroyed + 711 damaged, 2026-09-12): destroyed precision 0.91, recall 0.78, F1 0.84; severe (major + destroyed) recall against manually tagged affected buildings 0.75. The reference layer is included in this dataset. Post-event imagery TMS: ``` https://titiler.hotosm.org/cog/tiles/WebMercatorQuad/{z}/{x}/{y}?url=https://production-raw-data-api.s3.amazonaws.com/ISO3/NPL/buildings/hot_flood_npl_buildings_damage_post.tif ``` For the wider corridor context, see Nepal Flood 2026 Flood Affected Area. - Source: OpenStreetMap, Vantor open data, ESRI World Imagery - Part of HOT disaster response exports on HDX by HOT

Education facilities: schools, kindergartens, colleges, and universities tagged via `amenity` or `building`. Useful for school-access mapping, education planning, and resource allocation. Coverage reflects where volunteer mappers have been active. Urban areas are usually well represented, remote areas may be sparse. Cross-check critical decisions with local field knowledge. - Source: OpenStreetMap, a free, collaborative world map - Tags: amenity, building - Part of the country exports on HDX by HOT

Building footprints across the area: every feature tagged `building=*`, from residential homes to hospitals, schools, and places of worship. Useful for population estimation, damage assessment, and urban planning. Coverage reflects where volunteer mappers have been active. Urban areas are usually well represented, remote areas may be sparse. Cross-check critical decisions with local field knowledge. - Source: OpenStreetMap, a free, collaborative world map - Tags: building - Part of the country exports on HDX by HOT

Educational facilities and critical educational sector infrastructure in the Kathmandu Valley in Nepal. Data downloaded from the Open Cities Project Educational Facilities page.