Prediction of Monthly Mean Surface Air Temperature Using SARIMA in Jos North, Plateau State, Nigeria

Abstract

Fluctuation in temperature causes an untold hardship to human, animals and plants. In this study, the forecasting model was developed using the Seasonal Auto Regressive Integrated Moving Average (SARIMA) model, which is based on the Box-Jenkins system. The monthly mean temperature data for Jos city was acquired from Meteorology and Climatology unit of the Geography Department University of Jos within the period thirty two years (January 1986 to December 2023), The SARIMA model most appropriate orders were determined using the autocorrelation and partial autocorrelation functions for time series results. The monthly mean Temperature was used to verify the validity of these models. Statistical criteria such as Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and R-square (R²) were used to measure the model's accuracy and compare them. The model SARIMA(4,1,6)(2,1,2)¹² was chosen as the most reliable model and was used in predicting the monthly mean temperature for the study area.