Evaluation of the Heliosat-4 and FLASHFlux models for solar global daily irradiation estimate in Uruguay

Abstract

Solar resource assessment is a fundamental part of solar energy projects risk evaluation. In absence of in-situ measurement or well-calibrated satellite-based estimates, stakeholders tend to rely in freely available solar irradiation data sets, which usually have not been validated against local ground measurements. In this work we provide a performance assessment of Heliosat-4 (CAMS) and FLASHFlux (NASA) solar irradiation models for Uruguay on a daily time scale. It is found that both models present low bias and a rRMSD of 10.0% and 11.6%, respectively, making the former a better option from the end-user perspective. Evidence that the FLASHFlux model overestimate irradiation for coastal oceanic stations is given. The uncertainty obtained for these models is higher than previously locally adjusted satellite-based models, like the modified JPT statistical model, which remain as the best option for the region.