Modeling climate change impacts on water availability and sectoral water demands in a multipurpose reservoir using machine learning: a case study of the Bagré Dam (West Africa)

Abstract The objective of this study is to assess the potential effects of climate change on water inflows, storage, and withdrawals from the Bagré dam in Burkina Faso. This dam is multipurpose, particularly for agricultural and hydropower production. Climate change, specifically shifts in temperature and precipitation, affects water resources and can therefore disrupt hydropower generation and agricultural production. A better understanding of its effects can help in implementing mitigating measures. To this end, a random forest model was developed to evaluate the impacts of projected climate change on reservoir inflows, storage and water needs. The results show a temperature rise of 1 °C to 5 °C and a general increase in rainfall of up to 50% for the MeanEns model by 2100, compared with the reference period 1995–2014. These changes in climate variables lead in an increase in dam inflows of up to 200% in some months, compared with the reference period, an increase in water storage of up to 60%, and a rise in water withdrawals for agriculture and hydropower of up to 50% and 100% respectively. The performance of the developed model is satisfactory for simulating water inflows and withdrawals for both irrigation and hydropower generation. Given the projected increase in agricultural and hydropower water demands, it is crucial to identify effective strategies for the sustainable and efficient management of water resources at the Bagré Dam.