“…ese, however, still face some practical difficulties such as lack of reliable aerosol information, which affects the accuracy of the estimations, especially under clear sky conditions [11,23]. Artificial neural networks (ANNs) have also been used to compute solar irradiance at the Earth's surface, for example, Sahan and Yakut [24] used a suit of meteorological and geographical parameters as inputs to an ANN model to estimate monthly average global solar radiation on a horizontal surface at five locations in the Mediterranean region, Ozgoren et al [25] developed an ANN model for estimating the monthly mean daily sum of global solar radiation using meteorological data of 31 stations across Turkey, Kumar and Kaur [26] used five parameters, i.e., temperature, pressure, relative humidity, precipitation, and clearness index as input parameters for a feed-forward neural network to estimate solar radiation in the Hamirpur region in India, while Hasni et al [27] were able to estimate the global solar radiation for Bechar city in the south western region of Algeria using only air temperature and relative humidity in addition to hour, day, and month of the year. As highlighted, these artificial neural networks also require meteorological data as input parameters, in addition to temporal and geographical parameters.…”