Terrain: OpenTopoMap (open-source, OSM + SRTM elevation, CC-BY-SA). None: the original flat look, real Bihar geometry with no external basemap. Remembered for next visit.

Real Bihar administrative hierarchy, State > District (38) > Block (534, the smallest unit shown here). This only changes which outline is drawn on the map; it does not filter which stations or river reaches are shown, those are controlled by Basin and Rivers below.

Real river course geometry (72 named reaches, Ganga/Gandak/Kosi/Sone and tributaries), not the illustrative straight-line station chains used elsewhere in PRAVAH. The moving dashes show real flow direction, confirmed by search against real sources, but only on the 7 major rivers where it's well-established (Ganga, Gandak, Kosi, Burhi Gandak, Mahananda, Ghaghara, Sone); the source data's own point order for other reaches wasn't reliable enough to trust for direction, so those stay un-animated.

This is a direct current-level-vs-threshold read, not the forecast-based risk classification shown elsewhere in PRAVAH (see Basin Dashboard): those are probability-based forecasts, this is where the level sits right now.

Districts containing at least one station currently above its own warning level, ranked by how many and how severe (real point-in- district-boundary match on each station's own coordinates, not a pre-computed list).

Every real Bihar district shaded by the worst live threshold status of any PRAVAH station whose real coordinates fall inside it (same point-in-boundary match used by Top risk districts above), not a separate model. Switches the boundary layer to District level while on.

Loading…

Real inundation-extent polygon published by CWC's own Advisory Flood Forecast portal, matched to Hajipur's current forecast level (not interpolated). Hajipur is the only PRAVAH station this exists for — CWC has never published one for any Kosi-basin station, confirmed against their own 44-station index, not a gap on PRAVAH's side. The slider explores CWC's own 15 discrete pre-computed extent polygons at other levels (their real "what-if" library, not interpolated by PRAVAH); each level's own area figure is our own planar estimate for relative comparison across levels only, and won't exactly match CWC's single headline number for the current level.

Named tehsils and villages around Hajipur, ranked by the live fraction of each place's own area sitting inside CWC's current flood extent. Click a row to locate it on the map. Sorted worst-affected first.

126 real gauge-blended rainfall stations within Bihar's own bounding box, CWC's own "past 3–6 hrs" product. District average interpolates each district's real gauges (nearest gauge used where none fall inside), shading the boundary layer instead of points.

Real flood embankments and dykes within Bihar's own boundary, from OpenStreetMap (man_made=embankment / man_made=dyke / waterway=dyke), including named ones like the Kosi Eastern Embankment where OSM has a name tag. Community-mapped, so coverage isn't guaranteed complete — a real, sourced sample, not an exhaustive asset register.

Basins (pick any number, or "All")

Real, live water-level and rainfall stations from Nepal's own Department of Hydrology and Meteorology, across the 5 basins that actually feed Bihar (Gandak is Narayani in Nepal). This is Bihar's real upstream signal - a rise here reaches the border hours to a day before it reaches these stations' own downstream counterparts. Refreshed once per PRAVAH live cycle, not a static snapshot. Full Nepal Upstream table →

Pick a single basin (Gandak or Kosi) above to use time travel — it plays back one basin's own real recorded state history at a time.

Real recorded state history for the current basin (backed by PRAVAH's own pravah_station_state table via /state/at and /state/range), not a simulation. Playback moves every marker to that real moment's own values; Compare lists the real change between two moments, largest rise first.