Modelling the Temperature of South Africa Using Box Jenkins Methodology

The study employed the Box-Jenkins methodology to predict the annual temperature of South Africa. The study utilised time series data spanning from January 2012 to December 2020, comprising a total of 108 observations sourced from The World Bank Group. The best model was chosen using two criteria: the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC). Seasonal Autoregressive Integrated Moving Average (SARIMA)(1,0,1) × (0,1,1)[12] model outperformed SARIMA(0,0,1) × (0,1,1)[12], SARIMA(0,0,2) × (0,1,1)[12], and SARIMA(1,0,2) × (0,1,1)[12] because it has the lowest AIC and BIC value. ACF, quantile-quantile plots (qqplots), and the Ljung-Box test were among the statistical tests performed on the model’s residuals. These indicated that the chosen model is adequate and can be used to forecast temperature in South Africa. The best model was used to forecast South African temperatures for the next three years. The findings revealed seasonal fluctuations that were adjusted for previous years and a narrow confidence interval.