Issues in Developing and Validating Satellite Land Surface Temperature Product

Information on land surface temperature (LST) is important for understanding climate change, modeling the hydrological and biogeochemical cycles, and is a prime candidate parameter for Numerical Weather Prediction assimilation models. Satellite remote sensing is the only resource for providing regular regional and global LST measurements. The satellite LST production has been conducted over 30 years, through a variety of sensors onboard geostationary- and polar-orbiting satellites; a number of different algorithms have been applied for LST derivation from the sensor data. It is well-known that quality of LST production is significantly lower comparing to the satellite production for sea surface temperature (SST) using the same sensor data. Applications of the satellite LST product have been significantly restricted due to such low-quality status. This is particularly true in promoting the LST data usage in numerical weather predicting model, which is a critical application domain of the satellite products at National Oceanic and Atmospheric Administration (NOAA). Understanding issues in LST development and validation is vital in our efforts to improve the satellite LST production.