Lozère, France
Fires in Lozère
Is there a fire in Lozère today? Satellite detections from the last 24 hours, plus five years of NASA FIRMS archive: 906 detections, season, worst year and the largest recorded fires.
Fires burning now in Lozère
Reading the latest satellite pass…
Window: last 24 hours, NASA FIRMS
See it on the mapThe official vigilance in Lozère
The level published by Météo-France in “Météo des Forêts”. This is the authoritative source, and the one the prefecture acts on — not a calculation of ours.
Reading the official bulletin…
The level covers the whole department: it does not say where something is burning, it says under what conditions a fire would start. It never replaces a prefectoral order or an instruction from the emergency services.
See the vigilance on the mapFive years of archive
2020 to 2026 · through 2026-07-31
The years up to 2024 come from NASA’s finalized yearly archives. What follows, through 2026-07-31, comes from its near-real-time interface: 2026 is therefore a partial year, and its most recent weeks have not yet received the classification that removes volcanoes, flares and industrial sources. A very recent fire counts here, with a little more noise around it.
- Satellite detections
- 906
- Distinct fires
- 339
- Busiest month
- March
- Worst year
- 2021
- Strongest signal
- 269 MW
- Seen at night
- 5.6%
- Detections per 1,000 km²
- 175.4
Detections per year
| Year | Detections |
|---|---|
| 2020 | 83 |
| 2021 | 193 |
| 2022 | 170 |
| 2023 | 89 |
| 2024 | 98 |
| 2025 | 107 |
| 2026 | 166 |
The season
Share of detections by month, over the five years together.
| Month | Detections |
|---|---|
| Jan | 112 |
| Feb | 241 |
| Mar | 292 |
| Apr | 135 |
| May | 1 |
| Jun | 7 |
| Jul | 41 |
| Aug | 8 |
| Sep | 10 |
| Oct | 5 |
| Nov | 22 |
| Dec | 32 |
The largest recorded fires
The largest fires recorded here over five years — when each was first seen from orbit, and where.
40 detections · 14.3 km across · 48 MW
32 detections · 17.6 km across · 52 MW
27 detections · 13.5 km across · 14 MW
19 detections · 10.5 km across · 137 MW
18 detections · 4.3 km across · 269 MW
A first detection is not an ignition. A polar satellite passes only a few times a day, so a fire is seen on the next overpass — sometimes hours after it started, in a pixel about 375 m across. The point given is where it was first seen, not where it began. The span is the straight-line distance between detections: a floor on how far the fire reached, never a burned area. The named municipality is the one nearest the centre of the detections.
Elsewhere in Occitanie
| Area | Detections |
|---|---|
| Pyrénées-Orientales | 1,119 |
| Gard | 969 |
| Hérault | 922 |
| Aude | 900 |
| Hautes-Pyrénées | 780 |
| Ariège | 510 |
| Aveyron | 403 |
| Haute-Garonne | 294 |
The questions people ask
- Are there any wildfires in Lozère right now?
- The “Right now” block at the top of this page counts the satellite detections that fell inside Lozère over the last 24 hours, and gives the time of the most recent one. It refreshes about every ten minutes, the rate at which NASA republishes its files. Zero detections means nothing was seen, not that nothing is burning.
- How do we know there is a fire in a given place?
- We do not know it, we infer it. Polar-orbiting satellites measure infrared radiation from the ground and flag every abnormally hot pixel. A VIIRS pixel is about 375 metres across, a MODIS pixel a kilometre. A detection therefore says “something very hot in this square, at this minute” — not “a forest fire started here”.
- When is the fire season in Lozère?
- Across the five archive years, the busiest month in Lozère is March. The “The season” curve below gives the full distribution. Note that a winter or spring peak usually means agricultural or pastoral burning rather than forest fire.
- What was the worst year in Lozère?
- 2021, with more satellite detections than any other year in the archive. This ranks by how many hot pixels were seen, not by burned area or damage: a cloudy year can be understated, and one large well-observed fire weighs heavily.
- Can the water-bombing aircraft and helicopters be seen?
- Yes, when they transmit ADS-B and a receiver hears them. The map draws their tracks, and the catalog lists verified aircraft country by country. A missing track proves nothing: radio coverage, a transponder switched off, an aircraft not in the catalog — three ordinary reasons not to see an aircraft that is flying.
- Can this map replace an official alert?
- No. Fire Maps shows observations and models, with their timestamps and their limits; it replaces neither the authorities, nor official alerts, nor incident command. In immediate danger, call the emergency services and follow official instructions.
How to read this page
A detection is not a fire. It is a pixel the satellite saw radiating like fire. One fire lights several pixels and comes back on every overpass: the unit counted here is evidence of heat, not the incident. That is why the page also gives a count of distinct fires, obtained by chaining detections that are close in space and in time.
The archive has gaps, and they are not random. A polar satellite passes a few times a day; under cloud it sees nothing; a fire smaller than the pixel can go unnoticed; a fire smouldering under canopy radiates little. An area with no detection is an area where nothing was seen — which is not the same as an area where nothing burned.
Two filters were applied, and this is what they removed. First NASA's own classification: 38,727 rows flagged as volcano, fixed ground source or offshore flare were discarded. Then persistence: 250 cells of about 500 metres lit in six different months across at least three distinct years behave like an industrial installation rather than a season; their 4,992 detections were removed.
Not everything that burns is a wildfire. Stubble burning, pastoral burning and the disposal of agricultural residue all produce perfectly real detections, often outside the summer season. A February or March peak on this page is almost always that.
The administrative outlines are generalized to a few hundred metres. A detection falling right on a boundary may therefore be counted on the neighbouring side. For a count of hot pixels the difference is immaterial; for a question of jurisdiction, use the official source.
Data sources
Archive built on July 31, 2026, covering 2020 to 2026.