Modeling of Sokoto Daily Average Temperature: A Fractional Integration Approach

Abstract

Autoregressive fractional integrated moving average modeling strategy was used to model the daily average temperature (DAT) series of Sokoto metropolis for the period of 01/01/2003 to 03/04/2007. The time plot suggests that there is persistence dependence in the series. The order of fractional integration was found to be 0.6238841. The correct model for the daily average temperature data (DAT) of Sokoto metropolis was built. Two models were found to be more adequate for describing,
explaining and forecasting the temperature, ARFIMA (3, 0.6238841, 1) and ARFIMA (1, 0.6238841, 3). But by checking the forecastability, ARFIMA (3, 0.6238841, 1) model was found to be the best optimal model that will best forecast Sokoto metropolis temperature. The fitted model should be used for future forecast of temperature of Sokoto metropolis. Forecasting temperature is important to Agriculturist, Geographers and Hydrologist. Air temperature determines the rate of evapotranspiration.