Prediction of water quality and LULC analysis using machine learning and geospatial techniques

ABSTRACT The present study explores the use of machine learning and geographical methods to forecast and assess the quality of water and examine the use and coverage of land. The study focuses on the Karamana River in India and looks at the significant changes in LULC models and water quality parameters between 2001 to 2020. The study reveals material from a dynamic environment manifested in the river through detailed numerical analysis and visualization, including urban development, deforestation, and agricultural development. Moreover, it has the ability to identify fluctuations in water quality metrics, which includes temperature, dissolved oxygen levels, pH, conductivity, and biochemical oxygen demand (BOD), providing important insights into pollutants, ecological stresses, and human impacts. The Water Quality Index (WQI) evaluates the overall water quality. Correlation analysis elucidates the interactions between water quality parameters while also providing insight into pollution sources and ecosystem dynamics. The findings illustrate the efficacy of the random forest method in soil classification, as it effectively covers an even distribution of soil types. Overall, the study emphasizes the importance of informed strategies for land management and techniques for managing water quality in sustaining the Karamana River ecosystem.