How the map is made
- Terrain. Elevation from SRTM (about 30 m) is resampled to a 360 × 360 grid of about 68 m cells covering roughly 24 × 24 km.
- Drainage. Mapped OpenStreetMap rivers are burned 5 m into the surface, pits are filled (priority-flood), and every cell drains to its steepest neighbour (D8). Counting the cells upstream gives each cell's catchment.
- Height above river (HAND). Each cell follows its flow path to the first river: a mapped river, open water, or a channel draining > 10 km². Its height above that river's surface is the HAND value. The 3D water and the river-rise scenario flood every cell whose HAND is below the chosen rise.
- Factors. Slope, curvature, wetness index, drainage density and distances to rivers and roads are computed from the same grid. Land cover comes from Sentinel-2 at 10 m, rain from the 7-day Open-Meteo forecast.
- Overlay. Each factor is split into five classes and rated 1 (least flood-prone) to 5 (most). The weighted mean of the ratings is the susceptibility score, then grouped as very low, low, moderate, high, very high at 2, 2.6, 3.2, 3.8.
| Factor | Weight | Most flood-prone when | Why it matters | Source |
|---|---|---|---|---|
| Elevation | 10% | low (classes from this window) | Low ground collects water first. | SRTM via AWS Terrain Tiles |
| Slope | 12% | low | Flat land drains slowly; steep land sheds water. | Derived from elevation (Horn method) |
| Distance from river | 14% | low | Rivers overflow onto the land beside them. | OpenStreetMap rivers + drainage traced from elevation |
| Rain forecast, 7 days | 8% | high | More rain means more runoff to carry. | Open-Meteo forecast, 5 × 5 points, interpolated |
| Drainage density | 6% | high (classes from this window) | Many channels nearby mean water gathers quickly. | Channel length within 1 km |
| Land cover | 8% | built-up, flooded vegetation or crops | Roofs and roads shed rain; forests soak it up. | Esri / Impact Observatory Sentinel-2 10 m |
| Wetness index (TWI) | 10% | high (classes from this window) | Where a big area drains onto flat ground, soil stays wet. | ln(upslope area ÷ tan slope) |
| Height above river (HAND) | 22% | low | Land only a metre or two above the river floods first. | Elevation minus the river it drains to |
| Upstream area | 4% | high | Bigger catchments deliver bigger floods. | D8 flow accumulation |
| Distance from main road | 3% | low | Paved corridors speed runoff and often follow low ground. | OpenStreetMap main roads |
| Curvature | 3% | low | Hollows (negative) gather water; ridges (positive) shed it. | Laplacian of smoothed elevation |
Limitations
- This is a terrain-based screening map, not an official hazard map. For official flood hazard maps see Project NOAH and the GeoRisk HazardHunterPH portal.
- SRTM elevation dates from 2000 and measures rooftops and tree tops in places, so city blocks and dense forest read a few metres too high. Embankments and drainage pipes are not represented.
- The river-rise scenario raises every river by the same height. Real floods depend on how much water arrives, how long it lasts, and levees; the live river outlook above shows flow, which cannot be converted to a height without a local rating curve.
- Rivers that enter the window from outside carry more water than the upstream area counted here. Cells whose flow leaves the window before reaching a river show "no river in view" for HAND.
- Weights follow the usual ordering in published flood-susceptibility studies (HAND and distance to river first). They are expert judgement, not fitted to past floods.
- Soil type and vegetation index (NDVI) are not included: no keyless open source returns them at this scale. Land cover stands in for vegetation.
Data sources
- Elevation: SRTM and other open DEMs via AWS Terrain Tiles (Mapzen).
- Land cover: Esri | Impact Observatory Sentinel-2 10 m Land Use/Land Cover, CC BY 4.0.
- Rivers and roads: © OpenStreetMap contributors, ODbL. Preset areas use a saved copy; other spots are fetched live from the Overpass API.
- Rain: Open-Meteo forecast, CC BY 4.0. River outlook: GloFAS v4 via Open-Meteo.
- Satellite imagery: Esri World Imagery (Esri, Maxar, Earthstar Geographics).