Land surface temperature retrieval from AVHRR: influence of surface emissivity and atmospheric water vapor

The retrieval of land surface temperatures (LSTs) from satellite infrared radiometry is a more challenging problem than the corresponding surface temperature retrieval over sea. This because of problems associated with defining the surface, large spatial variability exhibited by LSTs, unknown surface emissivity, and variable atmospheric water vapor column. The authors present a method which retrieves LSTs from AVHRR radiances sensed through atmospheres with large and strongly varying water vapor content. The method recognizes that the emissivities are unknown, but likely to lie within a narrow range which is close to, but less than, unity. They have applied this method to the satellite and in-situ data obtained from the 1989 FIFE (First International Satellite Land Surface Climatology Project (ISLSCP) Field Experiment) in Kansas. Results show their method outperforms standard fixed coefficient algorithms which over-simplify atmospheric effects