Computational linear model for predicting solar energy in Botswana

Energy is key for the future. It plays a pivotal role in the socio and economic development of a country. With the depleting fossil fuels, it is crucial that the developing countries utilize other renewable sources and tap the potential for its development. Botswana being a landlocked country has an abundance of resources and receives the sunshine throughout the year. This paper discusses two different models for the development of linear regression model and polynomial model for solar radiation in Botswana. By using data of global solar radiation, extra-terrestrial solar radiation and sunshine hours data recorded for six locations, Gaborone, Francistown, Kasane, Maun, Shakawe, Ghanzi the solar constants were obtained. Regression analysis computed between measured and calculated solar radiation was applied to generate reliable Angstrom constants. The results of both the linear model and the polynomial model show significant agreements for each of the six locations.