A Climate-Adaptive Digital Twin for the Floating Solar-Hydropower-Pumped Storage Nexus

Abstract

Climate change is intensifying hydrological volatility, and the resulting droughts and flash floods increasingly threaten the baseload reliability of hydropower. We propose a real-time, AI-driven digital twin that co-optimizes existing hydropower reservoirs retrofitted with floating photovoltaic (FPV) arrays and pumped hydro energy storage (PHES), coupling a hydro-climatic predictive model with a non-convex power dispatch algorithm to shift the system dynamically between hydro-generation, solar-generation, and water-conservation modes. Using a coupled thermo-evaporative and hydraulic model evaluated against a 10-year climatic record, the digital-twin-coordinated nexus reduces reservoir evaporation by up to 34.0%, sustains firm power commitments through severe drought years, and increases system revenue by 22.4% relative to standard rule-curve operation.