Geostrophic Wind Dependent Probabilistic Irradiance Forecasts for Coastal California

Abstract

Coastal California has enormous potential for rooftop solar photovoltaic (PV) energy production. To reduce grid-integration costs, accurate and certain solar energy forecasts are required. However, frequent marine layer fog and stratus conditions limit the accuracy of numerical weather prediction models, especially during the summer. Thus, irradiance uncertainty is large and probabilistic forecast intervals are generally wide. To produce narrow and meaningful forecast intervals, the correlation of uncertainty to local meteorological conditions describing synoptic-scale atmospheric flow was considered. Specifically, the direction and magnitude of geostrophic flow were used as an indicator of coastal cloud cover probability to produce regime-dependent forecast intervals. The method was tested in summer 2011 for coastal California. The forecast interval was smaller than that provided by previous methods for all except clear conditions and contained 80% of irradiance measurements.