Prediction of probable extreme monthly average temperature by using polynomial regression

Increase in global warming since the end of the 19. century has increased the duration of extreme temperature events and caused the frequency and intensity of heat waves to increase. These changes in temperature have caused changes in climate systems and the hydrological cycle. These changes cause disasters and economic losses. Therefore, being able to predict the behavior of temperature has become an important issue. In this study, the missing parts of the observed air temperature data at meteorological stations in Turkey were completed using the SPSS program with the help of expectation maximization. Then, with PolReg, a software using polynomial regression, the extreme monthly average air temperatures that are likely to occur in the future were predicted for each month of the year within the limits of the 95% confidence interval. Thus, it is aimed to create a prediction about which values the air temperatures may rise to in which region throughout Turkey. After predicting the expected possible maximum values for 12 months of the year using the average air temperature values of 82 meteorological observation stations in various parts of Turkey, these predictions were created in the ArcMap program using the inverse distance weighted method for each month of Turkey's extreme monthly average air temperature maps.