Frequency Analysis of Outflow from Kufa Barrage by using Probability ‎Distributions for Flood Estimation

Abstract

This research aims to evaluate the suitability of several probability distributions to represent the discharge data from Kufa barrage, and to estimate the maximum hydrological values using the frequency factor method. In this research, three probability distributions were used to characterize the discharge data: the normal distribution, the lognormal distribution, and the gamma distribution. To estimate the maximum hydrological discharge values over different recurrence intervals (25, 50, 75, and 100 years), the frequency factor method was adopted. The relationship between the frequency factor and the recurrence interval was found to be a direct relationship, meaning that the frequency factor increases with the increase in the recurrence interval. Four distributions were used to calculate frequency factor, the normal, the lognormal, the extreme values (Gumbel), and the log Pearson type III distributions. The chi-square and Kolmogorov-Smirnov indices were applied, in addition to graphical consistency tests. According to the chi-square test, the lognormal distribution (using the maximum likelihood method) was found to be the best fit to the data, and according to the k-s index test, the lognormal distribution (using the method of moments) was found to be the best fit to the data.