Validation of Alternative Methods of Clear Sky Irradiance Modeling for Solar Energy Resource Assessment

While solar geometry and cloudiness are generally considered to be the first order determinants of solar irradiance, it is increasingly recognized that aerosols and associated variability of “clear sky irradiance” are an important factor contributing to the variability of solar irradiance, especially at typical candidate sites for solar development, where cloudiness may be uncommon but aerosol variability is large.

Two categories of models can be used to predict direct normal irradiance (DNI) for solar concentration or tracking photovoltaic applications. The first type uses methods to extract DNI from global horizontal irradiance (GHI). The second type is based on a broadband radiative model that predicts DNI under clear skies from atmospheric data. While many validation studies have been done on the former, the latter have only been evaluated in localized studies of a dozen or fewer sites and those mostly in the USA. In this study, we perform a global validation of the latter approach, specifically an implementation of the REST2 direct irradiance model, at 100+ locations worldwide against both GHI and DNI. Further, we compare at the same sites GHI and/or DNI following the former approach, specifically, the SUNY Perez clear sky methodology, which derives DNI from clear sky GHI estimates.  The two methods make use of a variety of global data sources, including geostationary visible imagery from multiple satellite platforms, snow cover data derived from the National Ice Center, MODIS Aerosol Optical Depth data, and aerosol and meteorological data from the ECMWF-MACC and MERRA2 global reanalysis datasets.

The results of this validation study will allow us to see how different methods of deriving GHI and DNI fare against independent ground station measurements. Results will indicate whether model performance varies regionally, or rather, that one model is superior to the other across all regions included in the global validation study. This information will provide the industry with guidance on improving the accuracy of solar energy resource assessments, and therefore decrease the risk and uncertainty associated with large project developments.