Technical note Monitoring the Oil & Natural Gas Corporation (ONGC) oil well fire at Pasarlapudi,Andhra Pradesh,India

Methods used to infer sea surface temperatures (SSTs) from satellite have traditionally been based on regression-tuned split-window fixed-coefficient algorithms. These can give inaccurate SST results when local atmospheric conditions are significantly different from those encapsulated by the regression averaging. The new generation of SST algorithms attempts to correct for atmospheric variability. These approaches include the R₅₄ transmittance-ratio methods of other workers, and the dynamic water vapour (DWV) correction method of the authors. The relative performances of the various methods are compared by applying each to an ocean and satellite dataset obtained off the west coast of Tasmania, Australia in 1987. We also investigate the performance of the NESDIS operational multi-channel, cross-product, and nonlinear formulas for NOAA-9, -11, -12, and-14 when applied to the same dataset. We find the DWV method gives SST retrievals which have significantly smaller bias errors than those returned by the three transmittance-ratio methods. The best overall performance was returned by the NESDIS multichannel (MCSST) formula for NOAA-9, indicating that in low water vapour loading situations, the standard regression-based algorithms work well.