In this study Landsat TM data was employed to ascertain the water quality of marine water at Lyari River outfall. Lyari River is a seasonal river and main sewage carrier, it is one of the sources of water pollution at Karachi port area. In the present study multiple regression algorithms were derived through multi-spectral data acquired from Landsat TM sensor and the sea truth data collected at different pre-designated sites. Water quality parameters ascertained at different locations included turbidity, total suspended solids and biochemical oxygen demand. The algorithm used was based on exo-atmospheric reflectance in different spectral bands. A window of 3x3 pixels was selected coinciding with the sea truth locations and the mean of digital numbers from these pixels were utilized to calculate volumetric radiance and reflectance. During the calculation of reflectance solar zenith angle taken from meta-data provided with the satellite image and earth to sun distance based on Julian days were used. The efficiency of models was evaluated using Karl Pearson's coefficient of correlation and maps were generated employing calibrated algorithms of water quality parameters.
