Optimizing Numerical Weather Prediction for Clean Energy

Modern weather forecasts are made by computers that solve the complicated equations for air motion, heat, and moisture. Different computer codes, called weather models, use different atmospheric approximations, creating slightly different forecasts. This forecast diversity is good, because the average of all forecasts is often the most accurate, and the spread between forecasts measures uncertainty. Utility companies such as BC Hydro need accurate weather forecasts to manage their hydroelectric reservoirs, anticipate power arriving from wind and run-of-river sources, efficiently maintain their facilities, and optimize energy trading. At BCHydro’s request, we will enhance the forecast model diversity, and will further improve the output from each model using statistics to remove other errors. BC Hydro is keen for us to merge these day-to-week forecasts with seasonal forecasts, and to enhance the shorter-range “nowcasts” with local observations. The benefit is improved safety, reliability and economic productivity of BC Hydro operations.

Intern: 
Simone Sperati
David Siuta
Pedro Odon
Anthony Di Stefano
Nadya Moisseeva
Timothy Chui
Dominique Bourdin
Faculty Supervisor: 
Roland Stull
Province: 
British Columbia
Partner: 
Sector: 
Partner University: 
Program: