Evaluation of the ability of meteorological parameters in the prediction of the average solar radiation in a coastal city in Iran

Solar radiation data play an important role in designing solar systems. These data measured are not easy for some locations; Therefore, solar radiation data have to be predicted by using solar radiation estimation models. In this paper was to investigate the effect of meteorological parameters sunshine, temperature and relative humidity on solar radiation based on Stepwise Multidimensional regression analysis according to the research literature and empirical studies (eight models including models suggested by Angström and Prescott, Ogelman, El-Sebaii, Karakoti and Abdalla) and we used daily solar radiation data of Bandar Abbas station during 19-year statistical period (1987-2005). The accuracy of the models was evaluated using the statistical indicators of Coefficient of Determination (R2), Root Mean Square Error (RMSE), Mean Percentage Error (MPE) and Mean Bias Error (MBE). The results of this analysis show that the variables of “sunshine duration (square) ” and “temperature” are significantly associated with solar radiation with a coefficient of determination of 0. 7566 (%99 confidence level). The results of this research necessarily cannot be generalized to other stations and in each Region, according to the specific characteristics of that Regions، we can find other variables effects on the solar radiation and its changes.