TOWARD EVALUATING SHORT-TERM PREDICTIONS OF SOLAR IRRADIANCE AT THE SURFACE: PERSISTENCE, SATELLITE-BASED TRAJECTORY AND NUMERICAL WEATHER PREDICTION MODELS

摘要

Herein we describe our efforts and initial results of collaborations between the Cooperative Institute for Research in the Atmosphere (CIRA) and National Renewable Energy Laboratory (NREL) to develop and evaluate short-term forecasting methods of solar irradiance at the surface by way of persistence, satellite-based cloud trajectory, and Numerical Weather Prediction (NWP) model cloud forecasting. We enlist an assortment of well-calibrated surface radiation sites (including the National Oceanic and Atmospheric Administration (NOAA) SURFace RADiation (SURFRAD) network and NREL-affiliated stations) as validation for models of persistence, satellite analyses from Geostationary Operational Environmental Satellites (GOES; i.e., retrievals of cloud mask, type, and optical properties that feed into a surface irradiance model), and NWP (predictions from the NOAA High Resolution Rapid Refresh, or HRRR). Also discussed is a method in development to leverage satellite-derived cloud cover statistics, conditioned on NWP-predicted regional-scale meteorological patterns, to forecast mid- to long-term surface irradiance trends.