Estimation of monthly mean air-sea temperature difference from satellite observations using genetic algorithm

Abstract

We present a new method to determine the monthly mean air sea temperature difference (ΔT = SST - Ta) from satellite observations. The satellite observed parameters viz., vertically integrated water vapour (W), sea surface temperature (SST) and wind speed (U) are used to derive ΔT. Genetic Algorithm (GA) is used to find the optimum relations between the input (W, SST, U) and output (ΔT) parameters. The input data consist of 6 years (January 1988-December 1993) of monthly averages of water vapour and wind speed from SSM/I (Special Sensor Microwave Imager), and sea surface temperature data from AVHRR (Advanced Very High Resolution Radiometer). Surface Marine Data (based on COADS observations) of SST and Ta (air temperature) are used to develop and evaluate the non-linear empirical relationship. Largest seasonal dependent differences between ΔT derived from the new method and in-situ data are found over the western boundary currents (such as the Kuroshio and Gulf Stream). After removal of the systematic biases, ΔT can be determined with accuracy of 0.40 ± 0.11°C. Copyright 2005 by the American Geophysical Union.