Methodology & sources

Every report is a free, automated estimate built entirely from public data. It's designed to give you a starting point before you buy or rent — not a professional flood survey, an insurance assessment, or legal advice. Always confirm anything decision-critical with a licensed surveyor or local authority.

The overall score (0–100, lower is safer) combines three factors: elevation (40%), flood history (35%), and distance to the nearest canal or river (25%). These weights are expert-assigned starting points — not yet calibrated against confirmed flood zones — and we're upfront about that rather than presenting them as scientifically proven.

Elevation (weight 40%): ground height relative to sea level, from 30m-resolution SRTM data (OpenTopoData). Bangkok's average elevation is about 1.5m, so the bands are calibrated to that context — at or below sea level scores the highest risk (100), above 10m scores the lowest (5).

Flood history (weight 35%): number of years with a recorded flood event at that location, from GISTDA's public flood-recurrence data. Zero recorded years scores 10 (low), 3+ years scores 90 (high).

Distance to nearest canal/river (weight 25%): closer waterways score higher risk — under 100m scores 80, over 1km scores 15. Search radius is 3km; nothing found within that radius is treated as low risk for this factor specifically, not as missing data.

Two of our data sources (GISTDA and OpenStreetMap/Overpass) are free public APIs without guaranteed uptime. When a factor's data is unavailable for a given point, we never fabricate a value — we exclude that factor entirely and redistribute its weight across the remaining factors, so the score always stays on a consistent 0–100 scale. Every report lists exactly which factors were excluded and why.

Local drainage capacity and pumping-station coverage meaningfully affect real flood risk — a well-maintained canal with active pumps can offset what its proximity alone would suggest. We don't yet have a validated public data source for this, so it's listed as an excluded factor on every report rather than guessed at.

GISTDA — historical flood recurrence data (Open Data Common license)

OpenTopoData / SRTM 30m — ground elevation

OpenStreetMap — canals, rivers, and other waterways, via the Overpass API

OpenGISData-Thailand — administrative boundaries (tambon/district/province)

This is a young, actively developed tool. The fixed weights above are a reasonable starting point, not a scientifically calibrated model — we plan to refine them against known flood-prone and flood-safe reference addresses over time. Programmatic pages (Bangkok districts, tourist destinations) use a single representative coordinate for that area, not every address within it — for a specific building or plot, always search its exact address.

Flood history relies on satellite-detected flooding (GISTDA) — reliable for large, sustained flooding, but it can miss brief, localized events, like a road flooded for a few hours after heavy rain, that don't leave a water body large or lasting enough to be visible from space. A location with zero recorded years means no satellite-detected flooding was registered there — not a guarantee the area has never flooded.