What band 13 sees
This layer is band 13, the clean longwave infrared window, at about 10.3 microns on the GOES Advanced Baseline Imager and about 10.4 microns on Himawari's Advanced Himawari Imager. Both sample at 2 km at the sub-satellite point.
"Clean" means the channel sits where water vapor absorption is comparatively weak, so what reaches the satellite is mostly what the cloud top or the surface emitted rather than what the atmosphere in between did to it. The instrument measures upwelling radiance, which converts through the Planck function to a brightness temperature: the temperature a perfect blackbody would need to emit that much energy.
Why temperature stands in for height
Thick clouds behave nearly as blackbodies at 10.3 microns, so the measured brightness temperature is close to the actual temperature of the cloud top. Temperature falls with height through the troposphere, roughly 6.5 K per km in a standard atmosphere, so a colder top is a higher top. That is the whole basis of conventional infrared enhancement, and why cold tops are drawn bright.
In round numbers, a deep convective top near the tropopause reads around -60 to -80 degrees C, mid-level cloud perhaps -10 to -30, and low cloud or clear ground sits near surface temperature. Converting a brightness temperature into an actual altitude requires a temperature profile, so infrared gives relative top height robustly and absolute altitude only approximately.
The reason to carry this layer at all is that it works at night. A visible channel measures reflected sunlight and goes dark on the night side. A thermal channel measures what the Earth and its clouds emit themselves, so the picture is continuous around the clock. The trade is sensitivity to thin and low cloud, which is covered below.
Parallax, and why the cloud is not where it looks
This is the limitation most worth understanding, because it systematically misplaces exactly the clouds an aviation reader cares about.
A geostationary satellite sits over the equator and views high latitudes at a steep oblique angle. A cloud top at altitude is therefore projected away from its true ground position, displaced toward the limb. The higher the cloud, the further it moves. Documented magnitudes: roughly 26 km for a 10 km cloud top at 60 degrees latitude, and in one Alaska case thunderstorm tops near 30,000 ft were drawn about 32 km away from where they actually were.
These layers are not parallax corrected. At middle and high latitudes, a cold top drawn over an aircraft's position may physically be tens of kilometers poleward of where it appears on this map.
Why Europe, Africa and South Asia are blank
NASA GIBS carries geostationary imagery from GOES-East, GOES-West and Himawari. Meteosat is not in GIBS, and Meteosat is the satellite that covers the 0 degree and Indian Ocean positions. There is consequently no infrared layer here for Europe, Africa, the Middle East, South Asia or the western Indian Ocean. That is a gap in the data service, not a bug on this map, and we would rather state it than let you assume clear skies.
There is a second, self-inflicted edge. Each satellite's tiles have an opaque background, so if the three were drawn overlapping the gray would stack and darken the map. Each is therefore clipped to its own band of longitudes. The visible consequences are thin seams at the clip boundaries, and a satellite's imagery stopping before its true useful field of view ends, so a storm near a boundary can be truncated even though the satellite can see it.
What this layer does not tell you
Every data source has an edge. These are the ones that matter for reading this layer correctly.
- It shows cloud tops, not precipitation. A cold top suggests deep convection but says nothing direct about rain reaching the ground.
- Thin cirrus is semi-transparent at this wavelength, so radiation from below mixes in, the retrieved temperature reads too warm and the cloud appears lower than it is.
- Low cloud and fog at night can be nearly the same temperature as the ground beneath and effectively disappear in a single infrared channel.
- Imagery is typically 30 to 60 minutes old, which is several times staler than the radar layer it sits beside.
- Geostationary parallax displaces high cloud tens of kilometers toward the limb at middle and high latitudes, and these layers are not corrected for it.
- Native tiles stop at zoom 6, so this is a continental-scale picture; anything sharper is upsampling.
- Europe, Africa, the Middle East and South Asia have no coverage because Meteosat is not carried by GIBS.
AeroScope is an situational awareness and research tool. Nothing here is approved for air traffic control, navigation, flight planning or collision avoidance. For operational decisions, use the authoritative source named above.