Forecasting of water quality parameters of Sandia station in Narmada basin, Central India, using AI techniques

Abstract

Abstract In addition to the influence of climate change on water availability and hydrological risks, the effects on water quality are in the early stages of investigation. This study aims to consolidate the latest interdisciplinary research in the application of artificial intelligence (AI) in the field of assessment of water quality parameters and its prediction. This research paper specifically explores the intricate relationship between climate change and water quality parameters at Sandia station, situated within the Narmada basin in Central India. As global climatic patterns continue to shift, the repercussions on water resources have gained prominence. In this work, electrical conductivity is predicted using the KERAS data processing environment on TensorFlow. The root-mean-square error (RMSE), coefficient of determination (R2), Nash–Sutcliffe efficiency (NSE), etc. are calculated between observed and predicted values to assess the model performance. A total of ten models are developed depending upon the input geometry from past monthly timelines. The results indicate that model 8, with ten inputs, performs the best based on the R2 value of 0.889. These results indicate that AI can be very helpful in analyzing the possible threats in the future for drinking water, livestock feeding, irrigation, and so on.