GIS, remote sensing, and machine learning for identifying optimal locations for seawater intake and brine disposal in coastal desalination systems

This is the authors' abstract. We don't add key points for this paper.

The rising global demand for freshwater has made seawater desalination vital, especially in coastal areas like Bhimunipatnam, Andhra Pradesh, India. Improper placement of intake and brine discharge sites can lead to environmental damage and inefficiencies. This study introduces an integrated framework using Geographic Information Systems (GIS), Remote Sensing (RS), and Machine Learning (ML) to identify optimal desalination sites. Key environmental and technical factors—such as bathymetry, land use, ocean current velocity, slope, proximity to sensitive ecosystems, and geomorphology—were analyzed through geospatial datasets. A Random Forest (RF) model was used for predictive site suitability, while the Analytical Hierarchy Process (AHP) guided factor weighting within a Multi-Criteria Decision Analysis (MCDA). The resulting Suitability Index (SI) categorized areas from Very Low to Very High suitability. Over 50 % of the region fell into High to Very High suitability zones. Field validation showed over 85 % prediction accuracy. The approach offers a scalable, reliable method to balance environmental sustainability with technical feasibility, providing actionable insights for decision-makers. This model can be adapted for other coastal regions facing freshwater challenges, supporting long-term water security and coastal resilience.