Empirical models for estimating the global solar radiation based on air temperature and sunshine duration

Abstract

This work aims to develop and validate empirical models for estimating global solar radiation at site of Alhoceima based on the meteorological data measured during the period 2010-2017. The empirical models studied in this paper are broken down into two categories: the first category consist of linear and nonlinear regression which relates the monthly average daily global solar radiation to the sunshine duration, such as: the model of Angstrom-Prescott, the cubic, quadratic model (Ogleman et al.), exponential (Almorox), logarithmic and power model. The second category consists of mathematical models calculating the monthly average daily global solar radiation based on the air temperature. And finally this study proposed model wish is based on sunshine duration and gives the best results for the city of Alhoceima. Statistical evaluation of the performance and accuracy of the models on the basis of The Root Mean Square Error (RMSE), Mean Bias Error MBE and R2 coefficient of determination, showed that all calibrated models estimates very well the monthly average daily global radiation. However the proposed model is more accurate with smaller errors and the best coefficient of determination.