Development of an aerosol-robust algorithm for remotely sensed sea surface temperatures from the NOAA AVHRR

Since 1981, remotely sensed sea surface temperature (SST) measurements have been produced operationally by the National Oceanic and Atmospheric Administration (NOAA) using data obtained from the Advanced Very High Resolution Radiometer (AVHRR) on board NOAA polar orbiting satellites. This long-term continuous time series of global data has been invaluable for various climatological, meteorological, oceanographic, and civil applications. The conventional algorithms are derived from regression analyses of AVHRR window channel radiances against in situ buoy measurements under clear-sky conditions which provide a correction for water vapor absorption. However for atmospheric conditions with high aerosol content (viz., persistent stratospheric aerosol layers caused by volcanic eruptions, and/or tropospheric dust and smoke outflows from continents), such algorithms lead to significant negative biases in SST due to the IR attenuation arising from aerosol absorption and scattering.

This paper describes the empirical derivation of a first-phase aerosol-robust daytime SST retrieval algorithm by utilizing a unique merged AVHRR Pathfinder data set. The data set consists of Pathfinder Oceans AVHRR-buoy matchup data that have been merged with the Pathfinder Atmosphere (PATMOS) daily gridded, clear-sky radiances (11 and 12 micron) and aerosol optical depths (0.63 micron) for the years 1990-1992. The time period includes measurements obtained during the residence time of the Mt. Pinatubo (mid 1991 to 1993) stratospheric aerosol layer. Daytime coefficients are derived separately for stratospheric, tropospheric and "combined" aerosol layers. Aerosol correction equations are derived for use with the NOAA operational NLSST and for retrospective PATMOS data. Use of these equations for the elimination of cold biases in the AVHRR SST should greatly improve its utility in monitoring climate change.