Abstract
Information on the global solar radiation is essential for many solar energy applications. Because of the cost of the measuring equipment, data on global radiation are not always available in most places, especially in developing countries. To overcome this difficulty, several attempts have been made to estimate solar radiation components from easy-to-measure atmospheric and/or geographical variables. In this study, monthly mean global solar radiation data for Qassim City, Saudi Arabia for the period from 1971 to 1998 were modelled using three meteorological variables: relative humidity, air temperature and atmospheric pressure. The predictability of the model was superior to the experimental data, showing a correlation coefficient (R) of 0.988. The mean percentage error (MPE) was less than 1%, the root mean square error (RMSE) was 0.02 and the mean bias error (MBE) was -1.1 x 10(-4). The performance model was validated using an independent data set for the period between 2003 and 2005. The statistical results were optimal and showed the ability of the model to predict the monthly global solar radiation for Qassim City for any period of time with less error. The predictability of some of the previously proposed models, which differed from each other in terms of the variables that they used and the number of variables contained, were tested to estimate monthly global solar radiation. The performances of these models were different in terms of predicting the experimental data.