SATELLITE-DERIVED SEA-SURFACE TEMPERATURES - A COMPARISON BETWEEN OPERATIONAL, THEORETICAL, AND EXPERIMENTAL ALGORITHMS

The Advanced Very High-Resolution Radiometers (AVHRR 2) on the National Oceanic and Atmospheric Administration (NOAA) operational meteorological satellites (currently NOAA-11) provide radiometric data at wavelengths of 3.7, 10.8, and 11.9-mu-m that enable an estimate of sea surface temperature (SST) using a differential absorption technique to allow for atmospheric effects. The operational multichannel SST (MCSST) algorithms used by NOAA/NESDIS (National Environmental Satellite, Data, and Information Service) are derived from a regression analysis of coincident satellite and buoy data. The SST algorithms can also be derived using theoretical models of infrared absorption in the atmosphere. By comparing the coefficients of operational, theoretical, and experimental algorithms it is possible both to assess the performance of the transmission models and to gain an insight into the physical processes underlying the SST algorithms. Two sets of data are also used in this study. First, a set of coincident ship and satellite data is used to derive simple experimental multichannel SST algorithms. The second set of data, collected using two satellites with overlapping swaths, has been used to produce experimental multipath algorithms for satellite-derived SSTs. The collective results discussed in this paper indicate that computer models of infrared transmission through the atmosphere are not yet capable of reproducing the observed satellite measurements. A better understanding of the water vapor absorption and the infrared transmission through the atmosphere is required if accurate prelaunch SST algorithms for future satellite instruments are to be derived.