Probabilistic Subseasonal Weather Forecasts for the Energy Sector

With SBIR funding, CFAN has developed a new probabilistic multi-model subseasonal forecasting system for U.S. temperatures and 100 m winds. Innovations include extreme event probabilities, ensemble clustering, 2-step calibration process, real-time verification, and a dynamic user interface.

Key Takeaways:

1. There is a great need for reliable subseasonal forecasts in the energy and financial sectors, particularly for extreme events.

2. Communication of forecast uncertainty and real-time verification of forecasts helps overcome skepticism of end users about forecast accuracy

3. A "forecast window of opportunity” approach is useful to identify situations with high versus low predictability.