GEOAI-DRIVEN LAND USE/LAND COVER CHARACTERIZATION OF RAJASTHAN USING SCATSAT DATA

Abstract

Land Use/Land Cover (LULC) characterization has an important part to play in environmental monitoring, natural resource management, agricultural planning and how to develop the region sustainably. However, cloud cover, atmospheric disturbances and differences in time can limit conventional optical remote sensing techniques, especially in large-scale regional studies. High revisit frequency and the ability of microwave remote sensing to observe the Earth during all weather conditions are a suitable alternative. The present study aims to suggest a framework of LULC characterization using scatterometer data from satellite SCATSAT in Rajasthan using GeoAI. The research uses K-Means clustering, geospatial Artificial Intelligence (AI) techniques, Normalized Difference Vegetation Index (NDVI) and Leaf Area Index (LAI) to identify and categorize the regional patterns of land cover in the State of Rajasthan. The primary data for the unsupervised clustering are the SCATSAT backscatter coefficient data, and the NDVI and LAI products are used as the sets to validate the characterisation of vegetation. The study will examine the potential to use the SCATSAT data to resolve regional-scale spatial variations in land surface properties, as well as to separate land cover classes. Advanced geospatial data preprocessing, transformation from raster to feature, optimization of clusters, spatial pattern recognition, and comparative accuracy assessment using Kappa statistics have been added to the framework proposed. The utility of the SCATSAT derived backscatter signatures for the identification of the vegetation density, barren lands, agricultural fields and mixed-use categories will be expected to be proven. The study addresses the lack of studies in the field of GeoAI by introducing the machine learning algorithms in the form of clustering methods to microwave data for creating LULC maps at large-scale. The results might be helpful for environmental planning, crop monitoring, drought assessment and sustainable land use management in the semi-arid areas of India.