Development of global solar radiation models: A comprehensive review and statistical analysis for Indian regions

The number of developed Global Solar Radiation models is relatively high, making it challenging to choose the most appropriate solar conversion system model. However, there is no comprehensive review of global solar radiation models available for India. This study's main objective is to develop a comprehensive review of the work done by Indian researchers in the field of solar radiation modelling such that it will provide a current state of the art for further research. The other aim is to carry out a statistical analysis of the developed solar radiation model for 522 locations across India. Which will help in the selection of the most appropriate and accurate model based on: (a) Country Best model, and (b) State Best Model. It is concluded that the sunshine duration is the most prominently used input parameter for the prediction of global solar radiation. For the region where the measured ground data is not available, the satellite-derived data can be used within a mean percentage error between -36.65% and 0.93% for all parts of India. As per the statistical analysis, the GSR model M -78 indicates the country's best GSR model, which is the function of relative humidity and coefficient of determination of 0.7959 and MPE of -2.5771%. Also, the GSR model M-78 has the least error in the following test, i.e., RMSRE (0.1483), MAE (0.5402), U95 (1.3771), and RMSE (0.6936). For State best, the GSR model M-179 indicates most suitable for the prediction of global solar radiation for most parts of India where the statistical error varies between; MBE (-0.429 to 0.449), RMSE (0.337-0.998), t-Stat (0.005-15.491), U95 (0.226-1.956), MARE (0.160-0.531), MAE (0.283-0.846), RMSRE (0.064-0.205) and GPI (0.366-2.010). The statistical error indicated by all the state is within the acceptable value. (c) 2021 Elsevier Ltd. All rights reserved.