Evaluation of diffuse fraction and diffusion coefficient using statistical analysis

Abstract In this study, eighty models are proposed in order to estimate the diffuse fraction and diffuse coefficient. For the proposed models, sunshine ratio and clearness index are considered as predictors. Monthly average global and diffuse solar radiation together with sunshine duration data of Tamanrasset station form 1995 to 2017 were analyzed. The different proposed models are compared and statistically analyzed to assess the performance of the best fitted model. Nine statistical indicators and Global Performance Indicator are computed to evaluate different proposed models. It is concluded that the cubic model with sunshine ratio and clearness index is selected as the best accurate model to estimate diffuse solar radiation on a horizontal surface in the study area.