Climate change, which results in rising global temperatures, poses a significant threat to Turkey, particularly regarding drought. Increasing temperatures not only jeopardize human health but also facilitate the spread of infectious diseases, disrupt ecological cycles, create irregular precipitation patterns, diminish agricultural productivity, and worsen resource scarcity. Consequently, monitoring temperature trends is essential for enhancing agricultural lands, conserving water resources, implementing sustainable energy initiatives, and formulating effective climate action plans. In this context, the present study focuses on temperature forecasting for Afyonkarahisar, a region of strategic importance for agriculture and renewable energy. Hourly temperature data from 2018 to 2022, obtained from the Afyonkarahisar Meteorological Service, were utilized to implement ARIMA and SARIMA models based on Box-Jenkins methods. The Seasonal Naive Forecast model was used as a basic benchmark to demonstrate the predictive capabilities of these models. Their performance was comparatively analyzed by using performance metrics evaluated over quarterly periods for the last year. The developed ARIMA(2,1,1) model outperformed the SARIMA(2,1,1)(1,1,2)₁₂ model, achieving improvements of 11.06% in RMSE, 10.80% in MAE, and 10.92% in R²; additionally, it surpassed the Seasonal Naive Forecast model with improvements of 60.59% in RMSE and 61.89% in MAE. The experimental results demonstrate that the ARIMA model effectively captures seasonal temperature trends and variations, providing accurate and cost-effective long-term forecasts.
No takes yet. Share an insight, caveat, or question.
Yeşil et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: