Mathematical Modeling of Rooftop Solar Energy Potential Using Geospatial and AI-Based Approaches

Abstract

Floating solar photovoltaic (FPV) systems provide an effective approach to increase renewable energy production while avoiding competition for land resources. This paper introduces a mathematical modeling framework for the automated identification and preliminary assessment of water bodies suitable for FPV deployment. The methodology combines satellite remote sensing data (Sentinel-2, Landsat-8) with geospatial analysis and mathematical modeling techniques. Water bodies are delineated using the Normalized Difference Water Index (NDWI), after which filtering criteria—such as minimum surface area, seasonal stability, vegetation cover, and proximity to infrastructure—are applied. A multi-criteria mathematical scoring model is developed to evaluate and rank the suitability of candidate sites. Preliminary experiments confirm the ability of the proposed model to detect and prioritize high-potential areas, thereby reducing the need for extensive field surveys. The research advances the integration of mathematical modeling and geospatial methods for renewable energy planning, with ongoing work directed toward refining model accuracy and environmental adaptability.