Application of geographic information systems with artificial neural networks technique in prediction of wind energy potential; the case of Isparta Province, Türkiye

Today, energy is one of the basic needs that must be met in every process of life. Although the energy production in the world increases by 4–5% every year, a large part of the energy produced is met from fossil energy sources such as coal, oil, and natural gas, which are rapidly depleted. Therefore, the need for renewable energy sources has gained importance daily and studies in this area have gained momentum. Wind energy, which is a renewable, clean, and endless energy source from the sun, is one of the most preferred energy types among renewable energy sources. Reliable wind data is the first requirement to determine the wind energy potential in a region. For this, it is necessary to measure the wind speed and predict the wind speeds in the coming years. In this study, the wind speed values for the next years were estimated using the artificial neural network method, using the data obtained from eight measurement stations in Isparta (Aksu, Eğirdir, Isparta, Süleyman Demirel Airport, Senirkent, Karaağaç, Uluborlu, Yalvaç). Then, using the forecast values obtained, wind speed distribution maps were created with the methodology of geographic information systems. It has been concluded that the most suitable region to establish a wind power plant in Isparta is the Keçiborlu district. The artificial neural network method obtained a high value of 0.994639 for the multiple determination coefficient (R²) while making these estimations.