Forcasting of Temperature in Sokoto Metropolis Using Seasonal Modelling

Abstract

In this paper, the results of seasonal modeling and subsequent forecasting of Sokoto monthly average temperature have been obtained using seasonal autoregressive integrated moving average modeling approach. Based on this seasonal modeling analysis, we conclude that the SARIMA(l,0,0)(0,1,1)12 ,SARIMA (3,0,1)(4,1,0)12 and SARlMA (4,0,2)(5,1,1)12 models are adequate for a good description of temperature in Sokoto. We further asses their forecastibility from the out-of-sample forecast statistic, Results show that for the short forecast statistics SARlMA (3,0,1)(4, I ,0) 12 model minimizes the mean squares error of the forecast, while the middle forecast and long forecast statistics results have shown that SARlMA (4,0,2)(5, I, I) 12 model has optimal forecast to the Sokoto temperature, hence this models have the advantage of capturing and describing and forecasting the seasonal dynamics of Sokoto city temperature.