Talamone Atlas
Retrospective analysis · Gironde, France — July 2026

The vegetation was already dry before the fire started.

Measured backwards from satellite time series alone, over the ground that burned around Saumos and towards the Cap Ferret peninsula.

Measured, not modelled Vegetation condition No network data
// basemap
// map layer
// overlays
// legend
// click the map
No point selected.

The seasonHow the moisture fell away, week by week

Where the departure stood on 21 July
Share of the area below its normal level

Nine yearsThe same week of July, one year after another

Which cover dried out most
How hard it burned

Cross-checkDoes the measurement hold up against the record?

Burn scar — measured here vs officially reported
What burned, by land cover

What this is — and what it is not

Retrospective, not predictive. We did not forecast this fire. The ignition near Saumos was reported as accidental; a dry canopy is a condition, not a cause. What the measurement shows is that the drying was already visible from orbit, weeks in advance, on the ground that later burned.

One dimension only. Vegetation condition and dryness, on a relative scale — not the calibrated, height-based clearance assessment used for operational work.

Computed here, end to end. No third-party fire product, burned-area map or land-cover database is used; the land-cover split is derived from the imagery itself.