Assessment and categorization of empirical models for estimating monthly, daily, and hourly diffuse solar radiation: A case study of Iran

Abstract

Knowledge of diffuse radiation is essential to design and monitor passive and active solar systems. Measuring diffuse radiation is not common in most meteorological stations. Therefore, developing empirical models for many regions is inevitable. In the current study, a comprehensive review was conducted on all available monthly, daily, and hourly empirical model formats (MFs) to estimate the horizontal diffuse fraction of solar radiation (DFSR). A total of 153 monthly mean daily MFs, 159 daily MFs, and 30 hourly MFs were cited. Monthly and daily MFs were classified into six sub-categories, while hourly MFs were classified into two sub-categories. Eleven new empirical regression MFs were introduced to estimate DFSR over five cities of Iran. Results showed that errors in all 11 newly developed models are not significantly different even at the 10% probability level. The yearly average of DFSR for Bandarabbas, Kermanshah, Mashhad, Tabriz, and Zahedan were 12.41, 9.04, 7.65, 8.99, and 9.93 MJ m−2, respectively. The generality of the newly developed models was proved by fitting them to data of four meteorological stations in the U.S.A. Comparing the newly developed models and the seven existing models showed that the newly developed models were more accurate in fitting the data.