The surface is what the satellite can see

A satellite looking downward may see a roof, bare soil, snow or the upper leaves of a forest. Land-surface temperature refers to those visible surfaces, rather than automatically to the air around them. NASA’s global-map description makes this distinction explicit. A hot patch on a thermal map can therefore be a real feature even when a nearby weather observation reports a lower temperature. The two measurements concern different physical things. Over a forest, the mapped surface can be the canopy rather than the shaded ground beneath it.

Sources: lst-definition

The radiation must be translated

Thermal-infrared observations are inputs to a retrieval, not ready-made surface thermometer readings. The Landsat processing documented by USGS combines thermal bands and brightness temperature with surface-emissivity information, vegetation information and atmospheric profiles. Emissivity describes how efficiently a material emits thermal radiation at the wavelengths being used. Atmospheric information helps account for what happens between the surface and sensor. The estimated surface temperature therefore depends on more than the brightness of a pixel: interpreting the signal requires both the emitting material and the intervening atmosphere.

Sources: lst-inputs

Two thermal bands can help separate effects

NASA’s MODIS documentation describes a split-window approach that uses thermal bands 31 and 32, with emissivity estimates informed by land-cover type. The retrieval also accounts for atmospheric water vapor and temperature conditions. These additional inputs help distinguish changes in the surface from changes affecting the outgoing radiation. A separate day/night method uses paired observations in several thermal bands to retrieve temperatures and emissivities together. These are named processing approaches, not interchangeable labels for every thermal sensor or every surface-temperature dataset.

Sources: lst-modis

Time and viewing angle travel with the temperature

The MODIS product description includes observation time, viewing angle, emissivity and quality assessment alongside temperature. Those fields help make a colored image interpretable. A daytime measurement and a nighttime measurement belong to different parts of a heating-and-cooling cycle, while the viewing angle describes the direction from which the surface was observed. A comparison that retains only the temperature layer throws away useful context already supplied with the product. The accompanying fields let readers distinguish a physical comparison from a comparison of unlike observations.

Sources: lst-modis

Clouds change which surfaces enter the record

ESA identifies cloud cover and incomplete sampling of the day–night cycle as important gaps in satellite surface-temperature records. A thermal-infrared product that needs a clear view of the ground cannot simply report the hidden surface through an opaque cloud. Over a longer interval, the available observations may therefore represent particular clear-sky conditions and observation times. A monthly map’s averaging period alone does not explain its coverage. The contributing observations and gap treatment determine which parts of the surface’s temperature history are represented.

Sources: lst-esa

A hot road and cool grass can share the same air

ESA uses the contrast between warm tarmac and cooler grass on a summer day to explain why surface temperature cannot be converted to air temperature by a universal offset. The surface is also involved in evaporation, convection and exchanges of heat with the atmosphere. That makes thermal maps useful for investigating how landscapes behave, including differences associated with vegetation and water availability. Their strongest interpretation starts with the observed surface and its energy exchanges, before attempting to infer an air-temperature quantity through an additional model.

Sources: lst-esa