A framework for the retrieval of all-weather land surface temperature at a high spatial resolution from polar-orbiting thermal infrared and passive microwave data

Land surface temperature (LST) is an important parameter associated with the land-atmosphere interface. Satellite remote sensing is the most effective method of measuring LST at regional and global scales. Satellite thermal infrared (TIR) measurements are widely used to retrieve LST with high accuracy and high spatial resolution but are limited to cloud-free conditions due to their inability to penetrate clouds. By contrast, satellite passive microwave (PMW) measurements are capable of penetrating clouds and providing data regardless of the cloud conditions. However, PMW measurements have limitations, such as a low spatial resolution and low temperature retrieval accuracy. Furthermore, temperature retrieval from PMW measurements yields the subsurface temperature, which differs from the LST retrieved from TIR measurements (skin temperature). This study proposes a framework for the retrieval of all-weather LST at a high spatial resolution by combining the advantages of TIR and PMW measurements. Compared to the MODIS LST product, the all-weather LST reflects the spatial variations in LST accurately. In situ LST measurements at four sites in an arid area of northwest China were used to evaluate the accuracy of the all-weather LST. The root mean square error of the LST under cloud-free conditions was approximately 2 K, whereas that of the LST under cloudy conditions varied from 3.5 K to 4.4 K. The results indicate that the all-weather LST retrieval algorithm can provide an LST dataset with reasonable accuracy and a high spatial resolution under clear and cloudy conditions.