Hydrological modeling to support water management and hydropower production: a case study

Accurate meteorological forecasts and hydrological modeling are essential for sustainable water resources management, especially under the growing pressure of climate change. The increasing demand for hydropower to support power grid stability, together with the need to preserve storage capacity for flood mitigation, requires decision-support systems able to forecast water availability across multiple timescales. This paper presents the framework adopted by Dolomiti Energia within an integrated weather-hydrological modeling chain based on the MIKE Powered by DHI software, enhanced by a machine learning correction layer and forced by a multi-model ensemble of meteorological forecasts, to support the management of 11 reservoirs characterized by different storage capacities and hydrological regimes. The results demonstrate that short-term inflow forecasts significantly enhance both routine hydropower operations and flood events management. Furthermore, the consistency of results across different meteorological forcing models confirms the robustness of the hydrological modeling chain and supports the adoption of a multi-model meteorological ensemble approach to better characterize forecast uncertainty. For snow-dominated storage reservoirs, where Snow Water Equivalent (SWE) estimates are used as the primary decision-support indicator in place of short-term inflow forecasts, estimate accuracy decreases with the increasing complexity of the hydraulic diversion network