Determination of the best fit distributions with FRANMOD model of daily maximum precipitation data for annual period of DSI 7. district provinces

The main input of water resources is rainfall with a high random variability. In order to plan and design water structures safely and economically, future yearly daily maximum rainfall quantities must be reliably estimated. The ability to make reliable and realistic estimates with frequency analysis depends on the correct determination of the best fitting optimal probability distribution for the data sequences. This study has been applied with the 87 probability distribution association, FRANMOD model, which is formed by 11 parameters estimation method of 32 continuous probability distributions to the yearly daily maximum rainfall series measured at 37 rainfall observation stations for at least 30 years in the field of activity of DSİ 7th Region Directorate. To determine the optimal distribution, the fitting test statistics of Kolmogorov-Smirnov, mean deviation from 1-1 line, Anderson-Darling, modified Anderson-Darling, average deviation, and adequacy coefficient were used. It was found that the most suitable distributions determined differ according to rainfall observation station. As a result of the weighted evaluation of the fitting test statistics, the first distribution which best represents the yearly daily maximum precipitation series is the Five Parameter Wakeby (WA5), the second distribution is the Three Parameter Beta-P (BP3), and the third most suitable distribution is the Three Parameter Generalized Logistics (GL3) distribution. These there distributions have achieved distinguish good results comparing to the other distributions. This study of thesis is expected to be a guide in the hydrological works and in the course of designing all hydraulic structures within the Yeşilırmak basin basically Amasya, Tokat, Samsun, Sinop and Ordu provinces and to contribute sustainable development to the area and entire country.