HotSpotters

How this works

The problem. MODIS has watched fires since November 2000 with 1 km pixels. VIIRS, flying since January 2012, has 375 m pixels and reports several times more hotspots for the same fires. Put side by side, the record seems to jump in 2012 although nothing changed on the ground.

The data. from NASA FIRMS, counted per month in 0.25 degree cells. The three VIIRS satellites are merged first: sightings in the same 390 m square on the same date and day/night pass count once.

The harmonized series is expressed in "one VIIRS satellite" units:

• From February 2012: hotspots seen by S-NPP (NOAA-20 after S-NPP stopped on 2 Nov 2026). Using one satellite keeps the count from rising just because more satellites were launched.
• Up to January 2012: MODIS counts multiplied by a local factor: how many S-NPP hotspots there were per MODIS hotspot in that 5 degree region while both flew (Feb 2012 to Dec 2025). World average: .
• Before July 2002 only Terra was flying, so those months are also scaled by the local (Terra + Aqua) / Terra ratio. World average: .

How well it holds.

Limits. Areas are matched to 0.25 degree cells (about 28 km), so small boxes are approximate. Years up to 2025 come from NASA's per-country files, which leave out detections outside any country's borders; 2026 includes them. The most recent months are near-real-time data and may change. November 2026 covers 7 days.

Built for the 2026 NASA Space Apps Challenge, "Harmonization of MODIS and VIIRS Hot Spots". Base map: CARTO, OpenStreetMap contributors.