生态与农村环境学报 ›› 2016, Vol. 32 ›› Issue (2): 277-282.doi: 10.11934/j.issn.1673-4831.2016.02.016

• 区域环境与发展 • 上一篇    下一篇

基于Mann-Kendall检验和重标极差分析的水质变化趋势

董旭1, 梅琨2, 商栩3, 黄树辉4, 黄宏2   

  1. 1. 温州市水利局珊溪管理局, 浙江温州 325000;
    2. 温州医科大学浙南水科学研究院, 浙江温州 325035;
    3. 浙江省流域水环境与健康风险研究重点实验室, 浙江温州 325035;
    4. 温州医科大学环境与公共卫生学院环境科学系, 浙江温州 325035
  • 收稿日期:2015-11-11 出版日期:2016-03-25 发布日期:2016-04-01
  • 通讯作者: 黄宏,E-mail:huanghongpanda@163.com E-mail:huanghongpanda@163.com
  • 作者简介:董旭(1987-),男,浙江温州人,助理工程师,学士,主要研究方向为水资源利用与保护。E-mail:276205244@qq.com
  • 基金资助:

    温州市公益性科技计划(S20140014);温州市水体污染控制与治理科技创新项目(S20140039,S20140037);温州医科大学人才科研启动基金(QTJ14045);浙江省自然科学基金(LQ16C030004)

Analysis of Variation Trend of Water Quality Based on Mann-Kendall Test and Rescaled Range Analysis

DONG Xu1, MEI Kun2, SHANG Xu3, HUANG Shu-hui4, HUANG Hong2   

  1. 1. Wenzhou Water Resources Bureau, Shanxi Authority, Wenzhou 325000, China;
    2. Southern Zhejiang Water Research Institute, Wenzhou Medical University, Wenzhou 325035, China;
    3. Key Laboratory of Watershed Environmental Science and Health of Zhejiang Province, Wenzhou 325035, China;
    4. Department of Environmental Science, School of Environmental Science and Public Health, Wenzhou Medical University, Wenzhou 325035, China
  • Received:2015-11-11 Online:2016-03-25 Published:2016-04-01

摘要:

将Mann-Kendall检验和重标极差(R/S)分析相结合,建立了一套综合分析过去和未来水质变化趋势的方法。首先,用季节性Mann-Kendall检验识别水质时间序列的过去变化趋势;然后,用R/S分析方法计算水质时间序列的Hurst指数(IH),0< IH< 0.5、0.5< IH< 1.0和IH=0.5分别指示反持续性、持续性和白噪声特征;最后,结合水环境保护规划和进展综合推断水质的未来变化趋势。基于2010-2014年珊溪水库及其6条入库支流TN浓度和CODMn,对该方法进行实例研究。结果表明,未来大部分支流的水质将会停止恶化或得到持续改善,然而水库TN浓度仍无明显下降趋势。水库对TN的滞留效应较强,且污染来源较多,水质的根本性改善将是一个长期的过程。除了继续控制入库支流污染,还要控制消落带污染和内源污染。

关键词: 水质, 变化趋势, Mann-Kendall检验, 重标极差分析, 长程相关性

Abstract:

A comprehensive set of methods was established based on Mann-Kendall test and rescaled range (R/S) analysis for analysis of variation trends of water quality in the past and future. Firstly, seasonal Mann-Kendall tests were performed to determine variation trend of water quality time series in the past. And then, R/S analysis was done used to calculate Hurst indexes (IH)of the water quality time series(0< IH< 0.5, 0.5< IH< 1.0 and IH=0.5 indicates anti-persistence, persistence and white noise characteristics, respectively). And in the end, variation trend of water quality in future was extrapolated by taking into comprehensive account the water environment protection program and its progress. Furthermore, this approach was demonstrated using the data of TN concentrations and CODMn in the Shanxi Reservoir and its six tributaries during the period of 2010-2014 as an example. Results show that water quality in most of its tributaries will stop deteriorating or go on improving in future, but TN in the reservoir will not display any declining trend, because the reservoir has a high TN retention capacity and numerous pollution sources. The reservoir has a long way to go before its water quality is fundamentally improved. In addition to keeping on controlling incoming pollution from its tributaries, countermeasures should be taken to control pollution of the water-level-fluctuating-zone and its endogenous pollution.

Key words: water quality, variation trend, Mann-Kendall test, rescaled range analysis, long-term correlation

中图分类号: