Recognition of flood estimation by generalized skewness method in some of Karkheh river sub-basins
2007
Tamahsebipour, N., Former Ph. D. Student University of Tehran, Faculty of Natural Resources (Corresponding Author) | Sharifi, F., Assistant Professor in Soil Conservation and Watershed Management Research Institute | Mahdavi, M., Professor in University of Tehran, Faculty of Natural Resources | Pezeshk, H., Associated Professor in University of Tehran, Faculty of Sciences
Flood estimation as a criterion for designing water structures such as small and large dams is one of the approaches to increase safety against their failure. There are many different methods to estimate flood magnitude and other characteristics. One of the most important and common method is statistical flood frequency of measured data. The study area includes some sub-basins of Karkheh reiver in Lorastan province. The test of outlier was conducted by using statistical parameters of peak flow in the hydrometric stations, and it was determined that there are no outlier points among data. So completing and lengthening for common statistical period performed. Then, statistical distribution of aforementioned data was fitted. Using girding method, the centeriod of high area of each hydrometric station was determined.Generalized skewness coefficients of points then were computed using the unbiased skewness coefficient, the weight coefficient of data and the distance of each hydrometric gauge from centeriod of sub-basin. Spline (Smooth Plate Line) method was applied to generalize the skewness with the mean square error of %34. The results show that the range of percentage of differences between unbiased and generalized skewnesses are from %58 to %137. The observed data have been fitted well with the normal distribution. Using this method results less differences between observed and estimated values of peak discharges as where generalized skewnesses were used, the differences of peaks for return periods of 2, 100 and 100 years, were %12, %77 and %180 respectively. It can be concluded that the fitnes of selected probability distributiion with the data is quite best using the generalized skewness in estimation of peak discharge.
Показать больше [+] Меньше [-]