Flood evidence for insurance portfolios in Thailand
Screen insured locations against historical flood observations, district terrain and source-dependent evidence, with explicit coverage limits.
Which insured locations need a closer flood review?
Compare your geocoded asset locations with published historical flood extents and record the observation date and source for each match. This produces an evidence shortlist for an underwriter, not a flood probability or premium.
Review workflow
- Prepare asset coordinates and an internal policy identifier; keep policyholder details in your own system.
- Retrieve historical extents around each asset and inspect observation periods before intersecting geometry.
- Add district terrain context and source-dependent flood evidence separately; do not merge historical and recent observations into a single risk score.
- Review unmatched assets as unknown coverage, then send the matched evidence and exceptions to an underwriter for validation.
What to produce
An asset-level review sheet with observed overlap, source, observation period, district context and a separate unknown-coverage flag.
| Dataset | Coverage and cadence | Sources and licences | Limitations |
|---|---|---|---|
| Flood history Beta | Published UNOSAT observation extents in Thailand Fixed historical archive Historical or versioned data; no real-time freshness claim | UNOSAT / United Nations Satellite Centre (Thailand flood extents) | A fixed archive of historical satellite flood-extent observations (UNOSAT). It is not a real-time flood map, and absence of an extent is not evidence that an area stayed dry. |
| Current flood evidence Experimental | Source-dependent observations in Thailand; BMA feeds cover Bangkok Source-dependent; checked about hourly Marked stale after 6 hours; unavailable after 7 days | Bangkok DDS legacy road-flood telemetry Bangkok DDS KlongMap canal water levels | Experimental, source-dependent observations (reports, road state, rainfall, water level). Coverage and cadence vary by source; an empty viewport does not mean no flooding. |
| Terrain Beta | Thailand district summaries Static historical elevation model Historical or versioned data; no real-time freshness claim | Copernicus DEM GLO-30 via Mapterhorn | District-level elevation summaries sampled from Copernicus DEM GLO-30. They are regional context, not parcel or building elevation and not a flood-probability model. |
| Administrative boundaries Stable | Thailand Versioned reference release Historical or versioned data; no real-time freshness claim | Boundaries are the COD-AB reference set for Thailand. They are a cartographic aid, not a legal or cadastral boundary. |
Questions this data cannot answer
The catalog does not supply actuarial loss curves, return periods, predictive hazard probabilities or building-level vulnerability. Historical overlap cannot establish future loss; no match cannot establish safety.
Try a bounded data request
Read-only, no key required, limited to 100 features in the stated viewport. An empty response can indicate missing coverage. Read the response timestamps, sources and status before use.
curl --fail-with-body 'https://api.mapscale.io/v1/public/atlas/layers/terrain?bbox=100.3%2C14.1%2C100.8%2C14.6&limit=100'Relevant geographic context
Inspect source-backed public examples
Use these references to inspect source identity, timestamps and limits. They do not establish coverage for every location in your project.
Prepare a pilot with your team
For a portfolio pilot, agree a small, authorized sample of asset coordinates, an observation period and an overlap rule. Review matched, unmatched and missing-source counts with your underwriter before discussing a larger integration.
Scope a data pilot with MapScale →
Workflow documentation by MapScale, using the linked dataset contracts for sources and limitations. Report source discrepancies through our contact channel. Your team remains responsible for domain review and approval.