Regional L-Moment-Based Flood Frequency Analysis in the Upper Vistula River Basin, Poland

详细信息    查看全文

• 作者：A. Rutkowska ; M. Żelazny ; S. Kohnová ; M. Łyp ; K. Banasik
• 关键词：Catchment characteristics ; clustering ; regional flood frequency ; L ; moments ; multivariate regression ; diagnostic measures
• 刊名：Pure and Applied Geophysics
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：174
• 期：2
• 页码：701-721
• 全文大小：
• 刊物类别：Earth and Environmental Science
• 刊物主题：Geophysics/Geodesy;
• 出版者：Springer International Publishing
• ISSN：1420-9136
• 卷排序：174

文摘

The Upper Vistula River basin was divided into pooling groups with similar dimensionless frequency distributions of annual maximum river discharge. The cluster analysis and the Hosking and Wallis (HW) L-moment-based method were used to divide the set of 52 mid-sized catchments into disjoint clusters with similar morphometric, land use, and rainfall variables, and to test the homogeneity within clusters. Finally, three and four pooling groups were obtained alternatively. Two methods for identification of the regional distribution function were used, the HW method and the method of Kjeldsen and Prosdocimi based on a bivariate extension of the HW measure. Subsequently, the flood quantile estimates were calculated using the index flood method. The ordinary least squares (OLS) and the generalised least squares (GLS) regression techniques were used to relate the index flood to catchment characteristics. Predictive performance of the regression scheme for the southern part of the Upper Vistula River basin was improved by using GLS instead of OLS. The results of the study can be recommended for the estimation of flood quantiles at ungauged sites, in flood risk mapping applications, and in engineering hydrology to help design flood protection structures.