摘要
在一个农业和采矿活动较多的小流域,分析了不同用地类型,包括农田、林地、采矿用地等潜在泥沙来源的土壤及泥沙的地球化学性质特征,利用复合指纹和多元混合模型研究了流域输出泥沙的主要来源及相对贡献。研究表明：该方法对不同泥沙来源的正确判别率最高达到87.5%,模型拟合优度达0.93。各泥沙样品中来源于农田的相对贡献为25.3%～65.2%,来源于矿山用地的相对贡献为34.8%～74.7%,来源于林地的泥沙总体上不到0.1%。指纹识别技术在地表物质均一的小流域也能有效识别泥沙的来源。
In a small catchment with frequent farming and mining,the properties of potential sediment sources and catchment outlet sediment were analyzed.Composite fingerprints and multivariate mixing models were used to estimate the relative contribution of the potential sediment sources.The results showed that：A discriminatory efficiency of 87.5% was provided for discriminating the different sources by the optimum composite fingerprint and a mean model GOF（goodness of fit）was up to 0.93.The relative contributions from agricultural land,mine land and wood land ranged between 25.3%～65.2%,34.8%～74.7% and less than0.1% respectively.It concluded that sediment fingerprinting could also provide reliable results in small catchments with relative uniform geological or geochemical properties.
出处
《水土保持学报》
CSCD
北大核心
2014年第6期106-110,154,共6页
Journal of Soil and Water Conservation
基金
国家自然科学基金项目(41101496)
江苏省自然科学基金项目(BK2011080)
关键词
泥沙来源
复合指纹
小流域
sediment sources
composite fingerprinting
small catchment
