Seasonal sea surface temperature of the Omura Bay, Japan estimated from multi-date Landsat-5 TM thermal infrared band

Seasonal sea surface (SST) of the Omura Bay was estimated using 13 digital data of Landsat-5 TM thermal infrared (TIR) band. These data consisted of 5 data collected in autumn, 3 in winter, 3 in spring and 2 in summer. The simple regression model between TIR and SST truth data of each season showed that each model is good for estimating the SST with high correlation coefficient and low root mean square (RMS) error. The RMS error are 0.35°C for autumn´s SST ranges of 18-25°C, 0.33°C for winter´s SST ranges of 8-14°C, 0.55°C for spring´s SST ranges of 12-23°C and 0.37°C for summer´s SST ranges of 25-31°C. The potential utility of TIR band in Landsat TM for mapping SST is discussed