生态与农村环境学报 ›› 2014, Vol. 30 ›› Issue (6): 717-723.doi:

• 自然保护与生态 • 上一篇    下一篇

复合指纹识别泥沙来源:潜在泥沙源地的选择

常维娜, 周慧平, 高燕   

  1. 环保部南京环境科学研究所
  • 收稿日期:2014-07-03 修回日期:2014-12-03 出版日期:2014-11-25 发布日期:2014-12-05
  • 通讯作者: 周慧平 环保部南京环境科学研究所 E-mail:zhp@nies.org
  • 作者简介:常维娜(1987—),女,江苏徐州人,博士生,主要研究方向为水资源与水环境。E-mail:changweina2006@126.com
  • 基金资助:

    国家自然科学基金(41101496);江苏省自然科学基金(BK2011080)

Identification of Sediment Sources Using Composite Fingerprinting:Selection of Potential Sediment Sources

CHANG  Wei-Na, ZHOU  Hui-Ping, GAO  Yan   

  1. Nanjing Institute of Environmental Sciences,Ministry of Environmental Protection
  • Received:2014-07-03 Revised:2014-12-03 Online:2014-11-25 Published:2014-12-05
  • Contact: ZHOU Hui-Ping Nanjing Institute of Environmental Sciences,Ministry of Environmental Protection E-mail:zhp@nies.org

摘要:

复合指纹识别技术应用于泥沙源解析的前提是假设研究者能够确定流域内泥沙的潜在来源类型,并且各泥沙源地所占比例之和为1,对潜在源地预判的不确定性将会对结果产生一定的影响。以南京市九乡河上游流域为例,对不同源地各指纹因子做均值显著性检验,检验各个预判源地能否单独作为潜在泥沙源地,同时对比分析不同沙源地对复合指纹因子判断泥沙来源正确率的影响,定量分析沙源地预判过程中的不足,探讨如何减少沙源地选择对泥沙来源识别的影响。通过对比研究发现,预判的4 种泥沙源地被重新调整为农田、林地和道路3 种类型。调整后各源地指纹因子差异性明显提高,不同泥沙来源的正确判别率最高达到89. 2%,泥沙贡献率从高到低依次为农田(39. 9%~87. 8%)、林地(<0. 1%~47. 7%)和道路用地(0~25. 6%)。

关键词: 复合指纹, 小流域, 泥沙源地

Abstract:

The application of the technique of composite fingerprinting to the identification of sediment sources is based on the assumption that the type of potential sediment sources could be confirmed and the sum of the contributions of the sediment sources equals to 1. However, the uncertainty in prediction of potential sediment sources would have some impact on accuracy of the results. The upstream of the Jiuxiang River in Nanjing was investigated for case study. Mean significance tests were carried out of fingerprint factors of the sediment sources in order to test whether each predicted source could be used individually as a potential sediment source. Meanwhile, effects of different sediment sources on the accuracy of the determination based on composite fingerprinting were analyzed and compared, in an attempt to quantitatively analyze shortages in the process of sediment source prediction and explore for ways to mitigate the impact of sediment source selection on identification of sediment sources. Results show that the four predicted sediment sources were readjusted as farmland, woodland and road. The readjustment significantly enhanced  the differences in fingerprint factors between the sediment sources. The highest identification rate of 89.2% was obtained in discriminating sources by using optimum composite fingerprinting. In terms of relative contribution of sediment, the sediment sources displayed an order of farmland (39.9%~87.8%) > wood land (<0.1%~47.7%) > and road land (0~25.6%).

Key words: composite fingerprinting, small catchment, sediment source

中图分类号: