Assessment of some homogeneous methods for the regional analysis of suspended sediment yield in the south and southeast of the Caspian Sea

Journal of Earth system science

Mehdi Vafakhah1¹  Hossein Kheirfam1¹ 

[1] Department of Watershed Management Engineering, Faculty of Natural Resources, Tarbiat Modares University, P.O. Box 46417-76489, Noor, Mazandaran Province, Iran.$$

关键词: Suspended sediment;    regional analysis;    input selection techniques;    watershed classification;    Caspian Sea coastal watersheds.;   

DOI  :  

学科分类：天文学(综合)

来源: Indian Academy of Sciences

【 摘 要 】

Regional analysis of suspended sediment yield (SSY) is commonly used to estimate sediment at a particular site where little or no information is available on sediment yield. In this research, the efficiency of three input selection and homogenization methods were evaluated in the estimation of SSY. Therefore, 42 sediment measurement stations and their upstream watersheds were selected and sediment rating curve was estimated by using regression models for each station. Mean annual SSY was estimated by using sediment rating curve and daily discharge. In the present study, in order to determine the independent variables in sediment yield, 11 physiographical, one climatic and two hydrologic variables of whole study watersheds were selected. Then the most effective independent variables were selected by using principal component analysis (PCA), Gamma test (GT) and stepwise regression (SR) techniques. After reducing 14 input variables to five (using PCA and GT) and two (using SR techniques), they are divided into homogeneous areas by Andrew curve (AC), cluster analysis (CA) and canonical discriminate function (CDFs) techniques. The watersheds were divided into two (using PCA-AC), three (using PCA-CA, PCA-CDFs and GT-CDFs), four (using GT-CA, GT-AC and SR-CA) and five (using SR-AC) homogenous regions. Multiple regression models to estimate mean annual SSY as a function of five (using PCA and GT) and two (using SR techniques) watershed characteristics were built in each homogeneous region, and compared to actual mean annual SSY in each station using relative error (RE), efficiency coefficients (CE) and relative root mean square error (RRMSE). The results showed that preprocessing the input variables by means of PCA and GT techniques has improved the homogeneous stations determination and the development models. According to the results, the best technique for determining homogeneous watersheds was AC technique with RE=49.24%, RRMSE=43.75% and CE=71.04%.

【 授权许可】

Unknown