Researchers Eloi Lindas, Yannig Goude, and Philippe Ciais from the Laboratoire des Sciences du Climat et de l’Environnement in France have developed a new method to improve wind power forecasts at subseasonal-to-seasonal timescales. Their work, published in the journal Renewable Energy, focuses on enhancing the accuracy and reliability of wind power predictions, which are vital for grid stability, balancing energy supply and demand, and managing market risks.
Currently, short-term weather forecasts are commonly used to predict renewable power output. However, forecasts with longer lead times, ranging from a few days to several weeks, have not been as thoroughly explored. This is despite recent advancements in subseasonal-to-seasonal weather probabilistic forecasting. The challenge lies in achieving reasonable skill levels without excessive temporal and spatial aggregation.
The researchers have created a forecasting pipeline that transforms European Centre for Medium-Range Weather Forecasts (ECMWF) subseasonal-to-seasonal weather forecasts into wind power forecasts. This pipeline can provide daily wind power predictions for lead times from 1 day to 46 days. The framework also includes a post-processing step to correct biases and address the lack of dispersion in the weather forecasts.
The study demonstrates that this new method outperforms a climatological baseline by 50% in terms of both Continuous Ranked Probability Skill Score and Ensemble Mean Squared Error. Moreover, the forecasts are nearly perfectly calibrated for lead times ranging from 15 to 46 days. This improvement in forecast accuracy can significantly benefit the energy sector by enabling better grid management, more effective balancing of supply and demand, and improved market risk management.
The practical applications of this research are substantial. More accurate wind power forecasts can help energy companies optimize their operations, reduce costs, and integrate more renewable energy into the grid. This, in turn, supports the transition to a more sustainable and low-carbon energy system. The researchers’ work highlights the importance of continued innovation in forecasting technologies to support the growing renewable energy sector.
Source: Lindas, E., Goude, Y., & Ciais, P. (2023). Achieving Skilled and Reliable Daily Probabilistic Forecasts of Wind Power at Subseasonal-to-Seasonal Timescales over France. Renewable Energy.
This article is based on research available at arXiv.