The Estimation of Hourly Global Solar Radiation Using a Cloud Cover Model Developed at Blytheville, Arkansas

Abstract

The lack of a comprehensive solar radiation monitoring network throughout the United States has led to extensive modeling. Some of the models use a measured component, usually the global solar radiation, to predict the other components, direct normal and diffuse sky solar radiation, while other models use meteorological measurements of hours of sunshine, cloud cover, atmospheric moisture content, visibility, and other variables to predict all three radiation components. The models may be hourly, daily, or monthly predictors. This paper is concerned with the modeling of hourly global solar radiation using cloud cover data. The model was developed from a meteorological monitoring station maintained at Blytheville, Arkansas from April 1978 to April 1980. The weather station was part of the Mississippi County Community College (MCCC) solar/thermal photovoltaic project located in Blytheville. The weather station measured and recorded direct normal, global, and diffuse solar radiation, as well as wind speed, wind direction, dry bulb temperature, relative humidity and barometric pressure. The cloud cover model was developed to be used in the simulation of the long-term performance of the MCCC solar/thermal photovoltaic system. Cloud cover data were available from nearby Blytheville Air Force Base (AFB). The military base maintains its own weather station, excluding any solar measurements, and makes regular observations of cloud cover and type, which are reported to the National Oceanic and Atmospheric Administration (NOAA). From the observations of cloud cover made at the AFB and the measurements of global solar radiation made at the MCCC weather station, a model was developed which can be used to estimate hourly global solar radiation. The model developed in Blytheville was compared to other cloud cover models, including the NOAA model, and correlated well. The study also determined that some of the existing cloud cover models are over simplified and are limited geographically. The Blytheville model, however, uses extraterrestrial solar radiation as a predictor, and should be more widely applicable.