Skip to content
worldfires
BR · Brazil · São Paulo · Local fire danger

Is today like the days when Lucélia burned?

Fires with a WorldFires page and a starting FWI in Lucélia since 2024: 2. That is fewer than 20, so the forecast is read with the rule for São Paulo (fires: 98). The fires of Lucélia are shown, but they do not set the percentile.

The chosen day, read with two rules

FWI from ECMWF IFS 0.25° (Open-Meteo)
16.0
Forecast FWI
Wednesday 18 November

FWI 16.0 today: EFFIS calls it “moderate”, and 1% of the fires in São Paulo (1 of 98) started with an FWI this low or lower. It is below almost every fire start in São Paulo.

Rule 1 · the EFFIS scale, the same everywhere

Fixed classes: with an FWI below 5.2, everything reads “very low”, wherever it is.

Rule 2 · the FWI at which fires here started

Tap a dot: each one is a real fire, with the date it started and the FWI on that day.

The next 14 days against its history

ECMWF IFS 0.25°

Forecast FWI (ECMWF IFS 0.25°) EFFIS “very low” limit

When fires start here, and at what FWI

  1. Jan
  2. Feb
  3. Mar
  4. Apr
  5. May
  6. Jun
  7. Jul
  8. 1 Aug
  9. Sept
  10. 1 Oct
  11. Nov
  12. Dec

Fires in Lucélia with a starting FWI, by month of start. The busiest month is August (1).

Area Fires p25p50p75p90
Lucélia municipality · too few: goes up a level 2 58.659.560.360.8
São Paulo region · rule in use 98 42.452.564.268.6
Brazil country 7,616 36.348.156.964.7

Rows of the table rebuilt every night. With fewer than 20 fires, the row of the level above is used.

Read this carefully

“Similar to”, never a warning. We compare the forecast weather with the weather on the days when real fires started here. We do not know whether there will be a fire or how it would start.

  • Sample: São Paulo, fires with a starting FWI: 98. With fewer than 20, we go up a level and say so.
  • Fires that entered Lucélia count, even if they started in a neighbouring area. Of the fires shown here, started elsewhere: 2 of 2.
  • The starting FWI comes from the ERA5 reanalysis at the fire’s location on the day it started. Recent fire pages are provisional and may still change.
  • Threshold and forecast come from the same calculation: the Canadian Fire Weather Index (Van Wagner, 1987) with local noon weather, the same as on the fire pages.
  • Series not comparable from October 2025 onwards: GWIS changed its near-real-time perimeters. Fires from that date onwards are left out of the calculation.

For official warnings, check with your local emergency or civil protection service. Monitoring maps:

NASA FIRMS GWIS · Copernicus Emergency: 193

Which days would pass a threshold?

Pick a percentile of the fires here and see which of the next 14 days match or exceed it.

Where this data comes from

Forecast: Open-Meteo.com (CC BY 4.0) on our own server, with ECMWF IFS 0.25° and AIFS (ECMWF, CC BY 4.0), GFS (NOAA, public domain) and GEM (data source: Environment and Climate Change Canada). Starting FWI: ERA5 reanalysis (Copernicus Climate Change Service) via Open-Meteo. Perimeters: EFFIS and GWIS (Copernicus, CC BY 4.0). Administrative boundaries: geoBoundaries/Instituto Brasileiro de Geografia e Estatística (IBGE), OCHA ROLAC (Creative Commons Attribution 3.0 Intergovernmental Organisations (CC BY 3.0 IGO)). FWI calculated by WorldFires with the Van Wagner method; EFFIS danger classes. This is not an official warning. Open-Meteo.com