This work considered the frequency of monthly rainfall from 1996 to 2011 obtained from National Root Crops Research Institute Umudike in Nigeria. The analysis was based on probability time series modelling approach. The Plot of the original data shows that the time series is stationary and the Augmented Dickey-Fuller test did not suggest otherwise. The graph further displays evidence of seasonality and it was removed by seasonal differencing. The plots of the ACF and PACF show spikes at seasonal lags respectively, suggesting SARIMA (0,0,0) (1,1,1)12. Though the diagnostic check on the model favoured the fitted model, the Auto Regressive parameter was found to be statistically insignificant and this led to a reduced SARIMA (0, 0, 0) (0, 1, 1)12 model that best fit the data and was used to make forecast. Comparison of the actual/observed frequency from July to December 2011 was done with their corresponding forecast values and a t-test of significance showed no significant difference.
[1]
G. Schwarz.
Estimating the Dimension of a Model
,
1978
.
[2]
Gwilym M. Jenkins,et al.
Time series analysis, forecasting and control
,
1972
.
[3]
A. C. Akpanta,et al.
Threshold Excess Analysis of Ikeja Monthly Rainfall in Nigeria
,
2015
.
[4]
H. Akaike.
A new look at the statistical model identification
,
1974
.
[5]
G. Box,et al.
On a measure of lack of fit in time series models
,
1978
.
[6]
D. Cox,et al.
An Analysis of Transformations
,
1964
.
[7]
W. Fuller,et al.
Distribution of the Estimators for Autoregressive Time Series with a Unit Root
,
1979
.
[8]
George E. P. Box,et al.
Time Series Analysis: Forecasting and Control
,
1977
.