Selection of Variable Gain Settings for Irs- P6: Liss Sensors

IRS-P6 spacecraft provides a unique opportunity of acquiring simultaneously multispectral (MS) data in three different spatial resolutions at 5.8 m, 23.5 m and 56 m, respectively, from LISS-4, LISS-3 and AWIFS sensors. Of these, spectral bands of the AWIFS (excepting the MIR band) have been designed to cover a 100% albedo, and with 10 bits per pixel to provide a good distribution of grey values for almost ground features. The LISS-3 and LISS-4 sensors have been designed with variable electronic gain settings, much similar to the LANDSAT-TM and SPOT HRV sensors, in order to cover the scene reflectance within the linear dynamic range of the sensors. Given the gain settings (equivalently Saturation Radiance, SR) permitted onboard for the LISS sensors, a study was conducted to select appropriate sets of gain values applicable to different terrain features and solar illumination conditions. In this paper, we describe the method followed, based upon experiments with the AWIFS data acquired specifically to cover terrain features and theoretical approaches for computing the top-of-atmosphere radiances for these features. Selection of the SR is to be chosen based on application requirements. The SR value larger than the optimum would result the terrain feature of interest covering only a small portion of the dynamic range of the output. On the other hand, if it is set at a lower value, then many common terrain features in the image would saturate, thereby limiting the data usability. Also, as the at-sensor radiance is dependent nonlinearly on the solar incident angle at the given scene, it is essential to take into account of the influence by the solar incident angle, which varies significantly for two major seasons in a year. The required data collected over cities, deserts, vegetation, lakes etc., across the globe were used here. Radiometric model of IRS MS sensors was described first, and a method of evaluating expected probable maximum spectral radiance for each feature of interest was given. Theoretical and experimental results obtained were compared to arrive operable gain settings appropriate for global data usage of IRS-P6 LISS data.