Statistical Modeling of Rainfall Distribution in Jos, Plateau State, Nigeria

Abstract

Rainfall is a form of precipitation that occurs when water vapor in the atmosphere condenses into droplets which can no longer be suspended in the air. Flood is as a result of rainfall overflowing onto land. Flood is associated with a lot of negative consequences on the activities of man and animals in general and even threatens their existences. In this study, rainfall data obtained on rainfall occurrence in Jos, Plateau State were described. Five theoretical distributions; Weibull, Log-normal, Gamma, Extreme value type 1 and Log-Pearson type III distributions were fitted to the data. Kolmogorov-Smirnov and Anderson-Darling tests of goodness of fit were used to identify the most appropriate distribution. The Maximum likelihood estimator, Bayesian estimator and Principle of maximum entropy were used to estimate the parameters of the identified distribution. Alkaike information criterion (AIC) was used to compare the estimates of the parameters from the different estimators. The probability of the returning periods and the future amount of rainfall were then predicted. The results obtained show that the distribution is skewed to the left, with median annual rainfall of 2383.5mm and inter-quartile range of 432.4mm.The results obtained also show that the Log-Pearson Type III distribution best fitted the rainfall data with the Bayesian method as the best estimator of the parameters of the distribution. Results on the probability of the returning period show that it decreases as the number of years increases while the amount of rainfall increases as the number of year’s increases.