MODELLING AND FORECASTING RAINFALL AND TEMPERATURE FOR UYO, AKWA IBOM STATE, NIGERIA

Akpan, Eno-obong Edem, Udo-Inyang, Uduak C. and Iren, Otonbong B.

ABSTRACT

The monthly air temperature and rainfall time series recorded between January, 1990 -December, 2022 from Nimet station were modeled and forecasted. In our forecasting, we used the methods of the Box-Jenkins Autoregressive Integrated Moving Average (ARIMA) model to analyze annual rainfall and maximum temperature of Uyo, Akwa Ibom State and the forecast. Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) were used to identify the models by aid of visual inspection. From graphical analysis on time plot and Autocorrelation Function (ACF), the series seems not to have a seasonal component and was also non- stationary. Stationary tests were conducted using the Augmented Dickey-Fuller (ADF). In order to attain stationary, first-order and second-order differencing (d = 1, 2) was carried out on rainfall and temperature data respectively. The chosen models were evaluated and validated using Akaike Information Criterion Corrected (AICC) and Schwartz Bayesian Criteria (SBC). Moreover, diagnostic check tests on the residuals of the models on mean rainfall and temperature satisfies the normality, constant variances (homoscedascity), P-P and Q-Q plots assumptions respectively. The best ARIMA model for rainfall for Uyo was (1, 1, 1) with AICC value of 275.15. That of maximum temperature for Uyo was (1, 2, 2) and the corresponding AICC value of 36.92. The model’s efficiency was checked using Sum of square error (SSE), Mean square error (MSE), Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) respectively. In order to assess the relationship between the observed values and the predicted values, Pearson Correlation analysis was employed. The result showed a correlation coefficient (r) of 0.464 and 0.385 between the observed and the predicted rainfall and temperature data respectively indicating a positive correlation between the two variables. The mean, standard deviation, coefficient of variation, skewness and kurtosis of monthly and annual rainfall and temperature was calculated to check the rainfall and temperature variability. Finally, ARIMA (1, 1, 1) and ARIMA (1, 2, 2) were used to forecast mean of monthly rainfall and temperature from the period 2024 – 2028 for Uyo Area, Akwa Ibom State.

Keywords: ARIMA Model, Modelling, Forecasting, Rainfall, Temperature.