Forecast of Daily Average Air Temperature by Artificial Neural Networks: Yozgat Province Case

Today, with the rapidly development of technology, artificial neural networks (ANN) usage as an alternative to statistical methods increase in many fields such as finance sector, medical science, defense industry, production, robot systems, image processing and data mining. Thanks to ANN, a computing technology created from the working principle of the human brain, it is possible to solve problems that seem very complicated for traditional methods. ANN applications are generally used for classification, estimation, data association, data interpretation and data filtering. The aim of this study is to forecast the daily mean air temperature using real atmospheric parameters by the ANN method. The Data used in the study consist composed of 3173 actual weather data (Daily Average Humidity, Daily Average Air Pressure, Daily Average Wind Speed, Daily Average Steam Pressure and Daily Average Temperature) measured in the center of Yozgat province between January 2007 and November 2017 obtained from the General Directorate of Meteorology. The ANN model used in the study was developed using the SLR (Simple Left-to-Right) algorithm. Data were normalized to avoid problems such as data size, calculation time loss. According to the findings obtained, the correlation value (R2) between the test data of 30% (951 registrations) and the estimation data was 0,97 and the RMSE (root mean square error) value was 1,34. The results of the study show that the ANN method by using the SLR algorithm can be used effectively in the forecast of air temperature.