Evaluation of River Water Quality Index Using Remote Sensing and Artificial Intelligence Models

To restrict the entry of polluting components into water bodies, particularly rivers, it is critical to undertake timely monitoring and make rapid choices. Traditional techniques assessing quality are typically costly time-consuming. With advent remote sensing technologies availability high-resolution satellite images in recent years, a significant opportunity for has arisen. In this study, index (WQI) Hudson River been estimated using Landsat 8 OLI-TIRS four Artificial Intelligence (AI) models, such as M5 Model Tree (MT), Multivariate Adaptive Regression Spline (MARS), Gene Expression Programming (GEP), Evolutionary Polynomial (EPR). way, 13 parameters (WQPs) (i.e., Turbidity, Sulfate, Sodium, Potassium, Hardness, Fluoride, Dissolved Oxygen, Chloride, Arsenic, Alkalinity, pH, Nitrate, Magnesium) were measured between 14 March 2021 16 June at site near Poughkeepsie, New York. First, Multiple Linear (MLR) models created these WQPs spectral indices images, then, most correlated selected input variables AI models. reference values WQPs, WQI was determined according Canadian Council Ministers Environment (CCME) guidelines. After that, developed through training testing stages, then compared actual values. The results models’ performance showed that MARS model had best among other WQI. demonstrated high effectiveness power estimating utilizing combination artificial intelligence