Every figure on GeoDwel comes from an official public source — a national statistics office, land registry, weather service or risk agency. We aggregate, we map, we cite.
We never interpolate. Where a municipality has no reported figure we show “no data”, never a value guessed from its neighbours.
Not every official figure is a direct measurement, and we say which is which. The climate layer is an official warming projection, not recorded weather. Natural hazards is a score we build by combining published hazard classifications. Healthcare access is an accessibility index modelled by the French health ministry’s statistics service. Every layer names its own source on screen, beside the figure.
When a source publishes no figure for a municipality — often to protect privacy in small places — we show grey and say so. We also grey out a figure of our own accord when it rests on too few records to mean anything: a property price drawn from one or two sales is noise, not a market. Either way we never fill the gap with an estimate.
Each legend and detail row shows its source and reference year, e.g. “INSEE Filosofi 2020”. When sources release new vintages, layers update and the citation changes with them.