Skip to content
worldfires
Smoke · CAMS forecast, 96 hours

Smoke today in Ambanja

Today the average level of fine particles (PM2.5) in Ambanja is 20 µg/m³, 1.3 times the WHO daily guideline (15 µg/m³). The peak of the coming days arrives on Saturday 7 at 21:00, with 32 µg/m³. It most likely comes mainly from scattered hotspots upwind that are not part of any fire page (68 % of the weight), the nearest 60 km to the south. Among fires with a page, the one that weighs most is Mandritsara fire, 12 October 2026, 193 km away (12 %).

Wildfire smoke episode CAMS global, 0.4° cell Local time: UTC+3
Today’s average
20 µg/m³
×1.3 the 24-h WHO guideline
Peak in 96 h
32 µg/m³
Sat 7 at 21:00
Likely source
Hotspots without a fire page
68 % of the upwind weight · 60 km to the south
Best time to air the house
09:00
Friday 6 · 13 µg/m³ on average over 2 h

The next 4 days, hour by hour

Hourly forecast of fine particles in the Ambanja cell, coloured by the WHO steps. The numbered green bands are the windows with the least smoke.

Where it most likely comes from

At 10:00 on Thursday 5, air reaches Ambanja at 850 hPa (about 1,500 m) from the south, at 16 km/h. Move the hour on the strip above and the sector turns with the wind.

Upwind hotspots (24 h) Other hotspots Fire page box Upwind sector (850 hPa)

VIIRS hotspots from the last 24 hours; the latest at 5 Nov 2026, 02:42 UTC

Candidates

Upwind fires within 250 km, ranked by their weight in the smoke arriving at that hour.

  1. 1Most likely: Scattered hotspots without a fire page68 %
    417 hotspots and 3,116 MW FRP in 24 h · the nearest 60 km to the south · outside any active fire page
  2. 2Mandritsara fire, 12 October 202612 %
    9,363 ha · since 12 October 2026 · about 34 hotspots and 666 MW FRP in 24 h inside its box · 193 km to the south
    Open the fire page
  3. 3Analalava fire, 25 October 20268 %
    ≈ 2,352 ha · since 25 October 2026 · about 35 hotspots and 172 MW FRP in 24 h inside its box · 77 km to the south
    Open the fire page
  4. 4Port-Berge (Boriziny-Vaovao) fire, 25 October 20265 %
    ≈ 2,086 ha · since 25 October 2026 · about 41 hotspots and 311 MW FRP in 24 h inside its box · 184 km to the south
    Open the fire page

There is no dispersion model behind this: it is an estimate. We add up the radiative power (FRP) of the last 24 hours of VIIRS hotspots lying against the wind at 850 hPa, weighted by distance. That is why we say “most likely” and show the whole list.

The 3 windows with the least smoke to air the house or go out

  1. 1 · Thursday 5
    11:00–13:00
    16 µg/m³ on average · ×1.1 the WHO guideline
  2. 2 · Friday 6
    09:00–11:00
    13 µg/m³ on average · within the WHO guideline
  3. 3 · Sunday 8
    10:00–12:00
    16 µg/m³ on average · ×1.0 the WHO guideline

Between 7:00 and 22:00, future hours only, 2-hour blocks at least 4 hours apart. The WHO guideline is a 24-hour average: here it is only a reference.

The last time there was smoke like this

We are still gathering the history for this cell: the archive loads at night, a little at a time.

We do not give medical advice. To know what to do at these levels, follow the health authority: World Health Organization · Air quality and health. Our figures are a model forecast, not a measurement in your street.

How this page works

Sources. Air quality: Copernicus Atmosphere Monitoring Service (CAMS global and CAMS Europe) served by our own Open-Meteo instance, Open-Meteo.com data under CC BY 4.0. Wind: ECMWF IFS 0.25° via Open-Meteo.com (CC BY 4.0). Active fires: NASA FIRMS, VIIRS (Suomi NPP, NOAA-20 and NOAA-21). Fire pages: EFFIS and GWIS (Copernicus, European Commission). Reference: 2021 WHO air quality guidelines. Open-Meteo.com (CC BY 4.0) · CAMS · NASA FIRMS

Contains modified Copernicus Atmosphere Monitoring Service information (2026). Neither the European Commission nor ECMWF is responsible for any use that may be made of this information.