Contribution of photovoltaic cover to reducing evaporation in the context of Climate Change: Experimentation and modelling

Morocco is one of the countries most vulnerable to the impacts of climate change. Reduced rainfall, rising temperatures and declining availability of water resources are likely to plunge the country into a situation of extreme water stress. In agriculture, water loss through evaporation is a critical issue in Morocco. Although the country currently has approximately 154 large dams, with a total storage capacity of about 20.7 billion cubic meters, significant water losses through evaporation have been reported. For instance, it is estimated that nearly one billion cubic meters of water evaporate annually from approximately 90 major dams, which together store about 13 billion cubic meters. Reducing evaporation from water bodies therefore remains a major challenge, particularly in semi-arid agricultural regions such as the Béni Mellal-Khénifra region (Morocco). Our experimental protocol presents an innovative approach to reducing the amount of water evaporated from small water bodies (irrigation ponds, swimming pools, etc.). In this context, we have adopted the photovoltaic cover approach, which not only reduces evaporation but also generates energy for various uses (pumping, lighting, localized irrigation, etc.). The experimental process used is installed at the climatological station of the Faculty of Science and Technology in Béni Mellal. It is based on two rectangular tanks of the same dimensions, one of which is covered by a solar panel, while the other remains in the open air. Parameters (water level, TDS, air temperature, water temperature, air humidity) are measured daily by hand and instantaneously by sensors.Initial results from this experimental approach show a remarkable difference between the evaporation rates of the two tanks. By way of example: over a 33-day period (from 17/04/2025 to 19/05/2025), measurements show a loss of 112.2 liters of water through evaporation in the open basin, compared with 93.72 liters in the covered basin, i.e. a difference of 18.48 liters, equivalent to a reduction in the evaporation rate of around 16.5% thanks to the photovoltaic cover. Admittedly, this percentage is greatly increased during the summer season (22/05/2025 - 22/07/2025) with a percentage of 20% due to the intense sunshine which causes great warming. In this context, we calculated the correlations between various hydro-climatic parameters to better understand the evaporation phenomenon. Predictive models were then developed using Support Vector Regression (SVR), Random Forest, and Gradient Boosting. Their performance metrics (e.g., RMSE and R2) are presented in the Machine Learning section, demonstrating the models' ability to accurately capture the dynamics of evaporation under both experimental conditions. So, this new approach, which is part of the integrated management of water resources, is a promising solution with a dual role: on the one hand, it will help us generate electricity, and on the other, it will preserve the maximum amount of water stored in water bodies.