| Evidence | Useful for | What it does not establish |
|---|
| Historical flood footprint | Checking whether a location intersects a recorded event | Future flood probability or absence of other floods |
|---|
| Flood report or monitoring observation | Reviewing dated, source-specific evidence | Complete road closures or a guaranteed safe route |
|---|
| Published regional air-quality metric | Comparing the stated metric and reporting period | Current street-level exposure |
|---|
| Water-station catalog | Finding monitoring locations and source links | A current measurement at every station |
|---|
| Administrative boundary | Joining evidence by official area code | A travel-time catchment or a land parcel survey |
|---|
| Terrain elevation model | Understanding broad topographic context | Survey-grade elevation or a calibrated flood model |
|---|
Questions to answer before using a dataset
Does a blank flood map mean there is no flood risk?
No. It may mean no recorded observation in the selected source, period or viewport. Check the response coverage, freshness, truncation and source dates. Historical flood footprints are observations of particular events, not a complete record of every flood or a probability of future flooding.
Can current and historical environmental data be compared directly?
Only when the measurement, units, geography and period agree. An annual regional air-quality aggregate is different from a current station reading. A water-station location is different from a current water-level observation. Do not label either as a live measurement unless the response contains that measurement and its observation time.
Which date should I quote?
Quote the source observation or reference period for the fact itself. A fetch time describes when MapScale retrieved the response; a page review date describes an editorial update. Neither makes an old observation current. Versioned boundaries and terrain do not have a real-time freshness claim.
How should Thai and English administrative names be joined?
Use the administrative PCODE, then display the Thai or English name. Transliteration is not an identity: Bangkok's วัฒนา has the COD-AB English name Vadhana. A join on the commonly written Watthana would miss that district. Codes identify the geography independently of the display language.
Can this evidence determine an insurance premium or collateral value?
These datasets can supply location context for a documented review. They do not by themselves establish actuarial loss probability, an insurance premium, market value or a lending decision. Preserve the source limitations and use appropriate validated models and professional review for those decisions.
A reproducible location review
- Record the input coordinate, coordinate system and asset-matching method. Keep unmatched addresses visible.
- Choose the source geography and period; preserve the layer identifier, source links, units and observation dates.
- Check freshness, unavailable responses, empty collections, pagination and truncation before interpreting a count.
- Join administrative regions by code. Document any spatial intersection or distance method and its limitations.
- Keep observations, forecasts and simulations separate. Publish only the conclusion the evidence supports.
- Retain the request, response dates and dataset version when provided so another reviewer can reproduce the result.
Source and licence register
The catalog links each available layer to its source, licence, coverage and delivery method. Read the layer-specific limitations before using or redistributing data. Attribution stays with the data; access through MapScale does not remove an upstream licence requirement.
Open the Thailand data catalog →