General models for estimating daily and monthly mean daily diffuse solar radiation in China's subtropical monsoon climatic zone

Solar energy is the most prominent kind of renewable energy resource. Accessibility to accurate information on diffuse solar radiation is critical to solar energy applications. This study primarily aims to present a general approach for estimating diffuse solar radiation in China's subtropical monsoon climatic zone. For this purpose, 46 existing empirical models in eight sub-categories were investigated for estimating diffuse solar radiation. Long-term, continuously measured solar radiation and other meteorological data were used for calibration and validation models. General models on both daily and monthly mean daily bases were established from the best-performing models. Among the daily basis models, FCS3 exhibited superiority as compared to the other models with the lowest average MAE, RMSE, U95, MAPE, and the largest average R2 of 0.854 MJ/m2, 1.376 MJ/m2, 3.658, 10.924%, and 0.888, respectively. The best overall performance on a monthly mean daily basis was obtained by CCS3 with corresponding statistical indicators of 0.569 MJ/m2, 0.723 MJ/m2, 1.866, 8.870%, and 0.938. Moreover, the percentage gap between the general and best site-dependent model performances was lower than 20% for most of the statistical indicators. Overall, the general models could be used to estimate the diffuse solar radiation in areas without diffuse solar radiation records.