Identification of weather events from INSAT-3D RGB scheme using RAPID tool

Abstract

Real-time analysis of products and information dissemination (RAPID), a web-based quick visualization and analysis tool for INSAT satellite data has been presented for identification of weather events. The combination of channels using red-green-blue (RGB) composites of INSAT-3D satellite and its physical significant value content is presented. The solar reflectance and brightness temperatures are the major components of this scheme. The shortwave thermal infrared (1.6 μm), visible (0.5 μm) and thermal IR channels (10.8 μm) representing cloud microstructure is known as Day Microphysics (DMP) and the brightness temperature (BT) differences between 10.8, 12.0 and 3.9 μm is referred to as Night Microphysics (NMP). The thresholds technique have been developed separately for both the RGB products of two years (2015-17 of December to February) of data for the identification of fog, snow and low clouds. The validation of these thresholds has been carried out against in situ visibility data from IMD observatories. The RGBs, i.e. DMP and NMP have a reasonable good agreement with ground-based observations and Moderate Resolution Imaging Spectroradiometer (MODIS) data. This threshold technique yields a very good probability of fog detection more than 94% and 85% with acceptable false alarm conditions less than 8% and 10% for DMP and NMP respectively. The technique has significantly minimized the misclassification between low clouds, snow, and fog and found useful for day-to-day weather forecast.