Bridging weather realism with future climate insights for energy system modeling - Livestream

Energy system models often depend on hourly weather timeseries to simulate supply and demand dynamics. Yet, most global climate models offer only daily projection data, limiting their direct utility for forward-looking energy system analysis. This talk presents a method used at EPRI that combines the advantages of historical weather records with the best available climate model simulations. Using a monthly quantile anomaly mapping technique, we transform historical hourly profiles to reflect future climatological shifts while preserving key real-world characteristics - such as local extremes, natural variability, diurnal ramps and patterns, and the physical interdependence of variables like wind, solar, and temperature. This hybrid approach enables the creation of hundreds of realistic synthetic hourly timeseries for future climates, supporting robust resource adequacy assessments and stress-testing as well as representative profiles for capacity expansion and load forecasting. By combining the fidelity of historical data with scenario-based climate modeling, we create new inputs for resilient and forward-looking system planning.

Erik Smith and Delavane Diaz, both from the Electric Power Research Institute, will present the methodology and its applications, highlighting how this hybrid approach is being used to inform planning and decision-making.