The Empirical Angular Function Approach: Testing Sea Surface Temperature Satellite Retrievals

Abstract

Recently, a statistical procedure was proposed to analyze the angular effect in the NOAA Advanced Very High Resolution Radiometer (AVHRR) brightness temperatures. The estimated empirical angular functions (EAF) over the oceans allow one to check the algorithms for the sea surface temperature (SST) and the column water vapor content when the observation geometry is variable, as well as to test angular methods of SST retrieval. The EAF approach has been previously applied to the analysis of the AVHRR brightness temperatures in channels 3 and 4 and dual-window SST over the tropical Atlantic in June 1987 and December 1988 from NOAA-10 and NOAA-11, respectively. Here, it is extended to estimate the accuracy of the split-window sea surface temperature and atmospheric water vapor retrievals from NOAA-9 over the tropical and North Atlantic in July 1986. The authors confirm the previously drawn conclusion that in a general case no angle-independent coefficients in a linear SST retrieval algorithm can provide angle-invariant retrievals. More recent operational NOAA angle-dependent algorithms have been shown to improve retrievals in the Tropics. In high latitudes, they seem to slightly overcorrect the angular effect. Using satellite data of higher spatial resolution with better radiometric accuracy is expected to improve the accuracy of the EAFs and the reliability of the conclusions.