Estimating the solar energy potential based on radiation values for different climate conditions

The rate of energy use in the world is increasing day by day. Energy is widely used in many fields such as industry, health, residential heating and lighting. Countries actively need energy in different areas such as health, education, transportation and production. Countries, which cannot meet their energy needs, cannot make positive progress in many areas. In today's world, many countries have turned to the energy field to increase the level of welfare. There is an energy deficit in the world due to the increase in population and industrialization day by day. Because of depletion of fossil-based energy resources, the new energy production methods are investigated. Hence, the trend towards renewable energies, which are inexhaustible and environmentally friendly, has increased. The main resource of renewable energies is the sun. The fact that solar energy is free, and our country has geography advantage that can receive so much sunlight has led us to work on this problem. In this study, research on solar energy radiation potential estimation, which is one of the renewable energy resources, have been carried out. Renewable energy types and their potentials are mentioned and the capacity of renewable energy types in Turkey is explained. In the next section, detailed information about the sun is given. In the literature part, the previous studies which include many techniques and models related to the potential estimation of solar energy are presented. It is seen that the most significant factor in electricity production from solar energy is the received radiation values in that region. For this reason, monthly estimates for solar energy potential for 2021 and 2022 were calculated by ARIMA and artificial neural network based on the radiation values of the cities of Istanbul, Ottawa, and Beijing, which are in different climate zones, for the years 2015-2020. Istanbul was the most compatible city in ARIMA and artificial neural networks methods. Among the ARIMA and artificial neural networks methods, the artificial neural network method shows the most similarity between the real value and the prediction model in all these three cities.