Why a fourth fire sensor
FIRMS does not carry Sentinel-3. The Copernicus service GWIS does feed SLSTR detections into its hotspot layers and it is from there that samoEKO takes Sentinel-3, and only Sentinel-3.
That restriction is not a detail but a condition of correctness: FIRMS already delivers every other satellite with a full attribute set. A second, attribute-less copy of the same detections under a different identifier would double every count in the situation picture. The gain is an additional observation window from a second, independent operator chain.
What is missing, and how samoEKO handles it
The GWIS hotspots carry only an identifier, a timestamp and a class. There is no fire radiative power, no confidence, no brightness temperature and no day/night flag.
samoEKO leaves these fields empty rather than inventing them. Confidence is recorded as “nominal”, the same house convention applied to unreadable VIIRS values. The day/night flag is derived from the sun's elevation by the interface, which states that derivation explicitly in the response header so that nobody mistakes it for a sensor value.
The distinction between Sentinel-3A and -3B remains reconstructable only through the GWIS identifier; in the database both are recorded as Sentinel-3.
Limits
Without fire radiative power, Sentinel-3 points are always the smallest on the map: circle size is computed from intensity, and that is missing here. Such a detection therefore looks weak regardless of the actual fire. For situation assessment this means: Sentinel-3 points say WHERE, not HOW STRONG. The difference becomes visible in the detail panel, which names the sensor. For prioritisation by fire intensity, FIRMS remains indispensable.
Then there is the time of day: SLSTR fire detections are in practice pure night passes, measured between 19:00 and 21:00 UTC. They therefore close a window that VIIRS already serves, and not the afternoon gap.
An archive trap belongs here too: the identically named EFFIS layer is a frozen archive covering September 2019 to October 2021 and must never enter live operation. samoEKO therefore reads a rolling weekly window.
The daily image
Alongside the points, samoEKO carries the OLCI daily mosaic from Sentinel-3A and -3B as an image layer, 0.06 degrees, roughly six kilometres. It is the coarser European counterpart to the VIIRS daily image, but from a different operator chain: if two independent chains show the same smoke plume, it is not a processing artefact.
As a raster layer, however, samoEKO deliberately does not carry SLSTR, although the service offers it. These are individual strips one hour and forty minutes apart: over the working area the layer is almost always empty, and when it is populated it shows the same detections that already appear as points on the map. A second, coarser depiction of the same thing would have doubled the counter without adding a statement.