Abstract
Water quality monitoring is critical for maintaining safe water and conserving ecosystem diversity. However, data and information on riverine water quality are sparse in India’s river systems. Remote sensing analytics have huge potential to enhance the ecological state of water resources by monitoring the evolution of water contamination over time. The principal aim of the study is to use empirical modelling approaches in developing models for estimating water quality parameters (WQPs) such as total suspended solids (TSS), dissolved oxygen (DO), Calcium, Chloride, and pH using Landsat-8 and PlanetScope satellite data and laboratory analysis. Surface reflectance and band ratios are mainly utilized as input data to develop linear regression with measured water quality data. Regression-based results with PlanetScope generated significantly higher R2 for all WQPs (0.65–0.78) except pH (0.41) as compared to Landsat-8. Results also showed that the regression models of TSS, DO, Calcium, Chloride, and pH are highly significant to visible (B, G and R) and near-infrared (NIR) bands of PlanetScope which can be attributed to finer spatial resolution. The water quality is mainly very poor around densely populated areas which crosses the permissible limit. Furthermore, the findings of this study illustrated the considerable capacity of water quality models based on remote sensing for conducting periodic monitoring and assessment. The applied empirical approach demonstrates the potential applicability of remote sensing analytics for the formulation of water management strategies, policies, and decision-making.