Current State of the Art in Solar Forecasting

As solar thermal and photovoltaic generation begin to have a larger role in electrical generation in California, the California Independent System Operators needs to accommodate their variable nature in its forecasting and dispatching. This project reviews and evaluates current knowledge and models for forecasting solar resources and considers options for improving forecasts through RD&D and additional measurements. Satellite and numerical weather prediction (NWP) have been shown to be the best tools for hour ahead and day ahead forecasts at this time. However, NWP solar forecast performance has yet to be evaluated for California, where the coastal microclimate especially may present a significant challenge. To validate and calibrate such forecasts, an aggregated real-time production database for all metered PV systems is deemed to be the most spatially dense and economical set of “measurements.” A research roadmap for improving Direct Normal Irradiance forecasts is provide