الفهرس 
Introduction
Scope of studyOnly 14 pages are availabe for public view
 |  | from | 178 |  |  |
| | | from | 178 | | |
Abstract
The critical design discharge value to be used in the economical design of pivotal hydraulic structures was found by using a statistical method called frequency distribution analysis. The frequency analysis is an important mathematical modeling technique. The estimation of design peak discharge of a design return period is a standard in many water resources projects. Frequency analysis is an efficient tool in design which reduces the cost of the projects due to employment of the forecasting techniques. This Thesis is focused mainly on the modeling of ’at-site method’of frequency analysis, involving the method of L-moments on four distributions functions; namely: Gumbel’s Extreme value type 1 distribution (EV1), Logistic distribution (LOG), Generalized Pareto distribution (GPA), and Generalized Extreme Value distribution (GEV). Three plotting position formulas of Hazen, Weibull, and Gringorten are selected for plotting order-ranked data series to achieve this work. L-moments method is applied for fittings these distributions to annual maximum discharges of four selected stations in the Nile River tributaries, and outcome estimates thus obtained are compared. This thesis presents a comparative study made to determine which distribution represented the statistical characteristics of observed flow discharge data in order to be utilized for predicting design discharge of different return period T-years design flow discharge (T = 2, 5, 10, 20, 50 and 100 or any desired year) for the Nile River tributaries. Comparisons of distributions were based on correlation coefficient (R2) in addition to the Root Mean Square Error (RMSE) between observed and estimated data in all stations. from the results and analysis of each individual station, GEV distribution based on L-moments best represented the statistical characteristics of observed data. Therefore, it could be concluded that for the study region, GEV distribution is the best distribution with Gringorten plotting position formula. Whereas, the quantile estimates by Gumbel distribution gives less fit to peak flow results.