Patrick Mathiesen
Mechanical and Aerospace Engineering PhD Student
University of California, San Diego

Operational Solar Forecasting at UCSD

North American Model (NAM) WRF-CLDDA

Power Forecast (kW):

Power Forecast (kW)

Recent Performance

Recent Performance

Typically, the NAM model over-predicts GHI at UCSD. To correct for this, historical accuracy was related to to the solar zenith angle (SZA) and forecast clear sky index (kt*). This correction function was then applied to future forecasts to produce an improved irradiance forecast. For the NAM, this has reduced RMSE by 20 W/m2 and effectively eliminated mean bias error.



Evaluation of numerical weather prediction for intra-day solar forecasting in the continental United States
Patrick Mathiesen and Jan Kleissl, 2011. Solar Energy. 85(5), pp. 967-977

Summarized Abstract:
Numerical weather prediction (NWP) is generally the most accurate tool for forecasting solar irradiation several hours in advance. This study validates the North American Model (NAM), Global Forecast System (GFS), and European Centre for Medium-Range Weather Forecasts (ECMWF) global horizontal irradiance (GHI) forecasts for the continental United States (CONUS) using SURFRAD ground measurement data.