An active fire detection is a satellite pixel flagged because it holds at least one fire or other strong heat source at the moment the satellite looked. The reported point is the center of that pixel, not the exact spot where flames are burning.
How satellites spot the heat
Polar-orbiting satellites carry the main global fire sensors. The Moderate Resolution Imaging Spectroradiometer (MODIS) on NASA's Terra and Aqua satellites uses pixels of about 1 km (0.6 mile).
The Visible Infrared Imaging Radiometer Suite (VIIRS) on Suomi NPP, NOAA-20 and NOAA-21 uses pixels of about 375 m (1,230 feet). NASA notes the finer pixels respond better to small fires and perform better at night.
Geostationary Operational Environmental Satellites (GOES) trade detail for frequency, scanning the contiguous United States every 5 minutes. VIIRS covers the whole globe every 12 hours or less.
What a detection does not tell you
NASA says it cannot determine the exact location or size of a fire from a detection, only that at least one fire sits inside the flagged pixel. A cluster of detections is a coarse hint at the burning area, not a fire outline.
Not every hot pixel is a wildfire. NOAA and NASA list gas flares, volcanoes, steel mills, structure fires and sun glint off bright surfaces such as solar panel arrays among the other sources.
Why fires get missed
A fire can start and end between satellite passes, and clouds, heavy smoke or tree canopy can hide it. It may also be too small or too cool to register.
MODIS routinely detects flaming and smoldering fires of about 1,000 square meters (roughly a quarter acre), and coarser data generally need larger or more intense fires.
Using detections well
Detections are quick and cover remote country, which makes them good for spotting new heat or growth. Analysts in NOAA's Hazard Mapping System review automated detections and correct errors they find.
Pair detections with mapped perimeters and official incident updates before drawing conclusions about how big a fire is or where it is heading.