A TIME SERIES ANALYSIS OF THE MONTHLY DISTRIBUTION OF RAINFALL
ABSTRACT
Water supply is widely acknowledged to be a pressing issue in the twenty-first century. As a result, research into the role of adequate rainfall in the development of human and natural resources is worthwhile. The monthly amount of rainfall in Enugu city during the period of was used in this project work (2000- 2012). A preliminary examination of the data revealed that there is no trend and that the data consists of multiplicative seasonal movements. Furthermore, the Augmented Dickey Fuller test of unit root and the Autocorrelation test for serial correlation of the error term revealed that the monthly data was stationary and serially uncorrelated.
The exponential smoothing procedures were used to build the best fit model for future prediction. Enugu’s rainfall pattern. This was accomplished through the use of algorithms aimed at smoothing out all irregular components inherent in the series. To forecast the monthly rainfall distribution for 2013, the best fit model parameters were used. The outcome predicted heavy rainfall in general for the year in question, with the amplitude peaking in October.