Time Series Analysis Model for Annual Rainfall Data in Lower Kaduna Catchment Kaduna, Nigeria

Abstract

Time series analysis and forecasting has become a major tool in many applications in water resources engineering and environmental management fields. The effects of climate change and variability on water demand in the 21^st century makes the time series analysis of rainfall, a major replenishing source of water, more imperative than ever before. The major challenge of water demand management is the ability to effectively estimate the contribution of rainfall to the water budget of any given basin. Among the most effective approaches for time series analysis is the Box-Jenkins’ Auto regressive Integrated Moving Average (ARIMA) model. In this study, the Box-Jenkins methodology was used to build an Auto regressive Moving Average (ARMA) model for the annual rainfall data taken from Kaduna South meteorological station within the Lower Kaduna catchment for a period of 47 years (1960 – 2006). From the analysis, the mean annual rainfall was 1385.2mm with a standard deviation of 313.8mm and coefficient of variation of 0.23 (low variation). The range for the period of study is 1407mm.The ARMA model identified is ARMA (1,1) which has Pearson Correlation Coefficient (R²) of 0.969 and residual ACF and PACF that indicated no pattern. The model is therefore adequate and appropriate for the forecast of future annual rainfall values in the catchment which can help decision makers establish priorities in terms of water demand management.