Projecting impacts of climate change on water quality and quantity using an AI-based model as an innovative alternative to physically-based models

This study explores an innovative approach to assess the impact of climate change on water quality and quantity by employing an Artificial Intelligence (AI) model, as an alternative to conventional physically-based models. Physically-based models, like SWAT (Soil and Water Assessment Tool), are widely used but face limitations in climate change scenario analysis due to the extensive input data requirements, complex model parameterization and prolonged simulation times. To overcome these challenges, we investigate the application of AI-based models, specifically the Random Forest (RF) method, for climate change scenario analysis in the Simiyu catchment of the Lake Victoria basin. Using meteorological data as input data, the RF model is trained with SWAT results for river flow and sediment load. Subsequently, the RF model is validated and utilized to predict river flow and sediment load under different climate change scenarios using new meteorological data. The RF model is compared to SWAT, demonstrating advantages, such as reduced input data requirements and computational costs. This research provides a streamlined and efficient approach for projecting climate change impacts on water quality and quantity, emphasizing the potential benefits of AI-based models over traditional physically-based counterparts.