黄土高原沟壑区典型草地小流域水沙特征对不同雨型的响应

Response of water and sediment characteristics to different rainfall regime in a typical grassland watershed in the gully region of the Loess Plateau

在黄土高原植被恢复已不断接近可持续利用水资源极限背景下,研究草地小流域水沙特征对不同雨型的响应可为合理进行植被恢复提供参考。文中以典型非治理小流域董庄沟2005-2014年径流、泥沙等实测数据为基础,采用K-均值聚类分析法结合判别分析法对次降雨进行分类,在此基础上,对各降雨要素与径流、输沙量之间进行Pearson相关分析,并建立了径流、输沙与降雨要素间的多元回归模型。结果表明:该地区降雨事件划分为3种类型:即A雨型(长历时大雨或暴雨),频次较低(34次);B雨型(中历时小雨或中雨)频次高(196次),C雨型(短历时大雨),频次极低(5次)。其中,A、C雨型径流深与前3天降雨量相关性最高(p<0.05),而B雨型径流、输沙与降雨要素均无显著相关因子。前期降雨量对径流深的影响有效时间约为7d。加入前期降雨量因子可使降雨-径流多元线性回归模型的精度大幅提高。

Under the background that the revegetation in Loess Plateau is approaching sustainable water resource limits, studying the response of water and sediment characteristics of grassland small watersheds to different rain types can provide a reference for reasonable vegetation restoration. On the basis of runoff, sediment and other measured data of typical non-manage small watershed Dongzhuanggou from 2005 to 2014, K-means clustering analysis combined with discriminant analysis method was used to classify the secondary rainfall. On this basis, Pearson correlation analysis was performed between each rainfall element and runoff and sediment, multiple regression models between runoff, sediment and rainfall elements was established. The results showed that the rainfall events in the region were divided into three types: type A(long-term heavy rain with low frequency: 34 times);type B(medium rain or moderate rain with high frequency:196 times);type C(short duration storm with extremely low frequency:5 times). Correlation analysis showed that runoff depths of type A and C were mostly correlated to 7-day antecedent precipitation(p<0.05). runoff and sediment of Type B wasn’t significantly correlated to precipitation factors. Pre-rainfall had a significant impact on runoff depth and was valid for approximately 7 days. The accuracy of the rainfall-runoff multiple linear regression model can be greatly improved after adding the pre-rainfall factor.