Evaluation of optical multi-spectral satellite data for crop type and land cover identification in Marathwada, India: a disaster management perspective

This study evaluates the use of optical multi-spectral satellite data for crop type and land cover identification in Marathwada, India, with a specific focus on disaster management. The region is highly susceptible to various disasters including droughts and other climate-related events that significantly impact agricultural productivity. The study involves analyzing both single-date and multi-temporal satellite imagery to develop composite images using different band combinations, aiming to identify the most accurate combination for crop and land cover identification. A multi-class classification approach based on random forest is employed for feature extraction and the significance of different bands in the imagery is assessed. The results demonstrate that a composite image composed of Red, Green, Blue, Near Infrared and Shortwave Infrared bands yields the highest accuracy with an overall accuracy (OA) of up to 93.69% for all land cover classes and 91.18% for crop classes alone, using six-date multi-temporal imagery. The findings highlight the potential of optical multi-spectral satellite data as an effective tool for crop type and land cover identification in Marathwada, India, particularly in the context of disaster i.e. agricultural draught management. The methodologies and results presented in this study can serve as a valuable reference for similar research endeavors in other agricultural draught prone regions of India and beyond.