生态与农村环境学报 ›› 2019, Vol. 35 ›› Issue (7): 925-932.doi: 10.19741/j.issn.1673-4831.2018.0549

• 研究方法 • 上一篇    下一篇

上海地区河网水质空间分异及对河岸带土地利用的响应

汪昱昆1,2, 程锐辉1,2, 曾鹏1,2, 车越1,2   

  1. 1. 华东师范大学生态与环境科学学院, 上海 200241;
    2. 上海市城市化生态过程和生态恢复重点实验室, 上海 200241
  • 收稿日期:2018-09-06 出版日期:2019-07-25 发布日期:2019-07-22
  • 通讯作者: 车越 E-mail:yche@des.ecnu.edu.cn
  • 作者简介:汪昱昆(1994-),男,江苏泰州人,硕士生,研究方向为环境规划与管理。E-mail:532438139@qq.com
  • 基金资助:

    水体污染控制与治理科技重大专项(2017ZX07207003-01)

Spatial Differentiation of Water Quality in River Networks in Shanghai and Its Response to Land Use in Riparian Zones

WANG Yu-kun1,2, CHENG Rui-hui1,2, ZENG Peng1,2, CHE Yue1,2   

  1. 1. School of Ecological and Environmental Sciences, East China Normal University, Shanghai 200241, China;
    2. Shanghai Key Laboratory for Urban Ecological Processes and Eco-Restoration, Shanghai 200241, China
  • Received:2018-09-06 Online:2019-07-25 Published:2019-07-22

摘要:

河岸带土地利用是影响河流水质的重要因素,基于上海市2013年55个河网水质监测点10项水质指标数据,利用SOM+K-means自组织特征映射(self-organization feature mapping,SOM)神经网络,识别全市水质空间分布格局;运用冗余分析(RDA)和Spearman秩相关在不同空间尺度(100、200、500和1 000 m缓冲区)上探讨水质与河岸带土地利用的关系及尺度效应。结果表明:(1)可将上海市55个水质监测点划分为4个聚类,体现出较为明显的空间异质性,监测点分布于淀山湖、崇明岛等城市远郊地区的聚类Ⅰ水质最优,而监测点分布于苏州河沿线的聚类Ⅱ和城市近郊的聚类Ⅲ的水质较差;(2)在空间尺度上,500 m缓冲区对聚类Ⅰ、Ⅲ和Ⅳ的总解释率最强,1 000 m缓冲区对聚类Ⅱ的总解释率最强;(3)在最优空间尺度上,城镇建设用地对各聚类水质都有较高的解释率,且与大部分水质指标呈正相关。

关键词: 平原河网, 自组织特征映射, K-means算法, 空间异质性, 冗余分析

Abstract:

Land use in riparian zones is an important factor affecting river water quality. Data of 10 water quality indicators for 2013 were collected from 55 river network water quality monitoring stations in Shanghai. Based on the data,the self-organizing map was used to identify the spatial distribution pattern of water quality in the city. Moreover,the redundancy analysis (RDA) and Spearman rank correlation analysis were used to investigate the relationship and scale effect between water quality and riparian land use (100,200,500,1 000 m buffer). The results show that:(1) The 55 water quality monitoring stations in Shanghai could be divided into 4 clusters,which reveals an evident spatial heterogeneity. The cluster I composed of monitoring stations in the outer suburbs,including Dianshan Lake and Chongming Island,represents the best water quality. Comparatively,water qualities of cluster Ⅱ distributed along the Suzhou River and cluster Ⅲ in the suburbs of the city are poor. (2) Among all the spatial scales,the 500 m buffer has the strongest total interpretation of clusters Ⅰ,Ⅲ,and Ⅳ,and the 1 000 m buffer has the strongest total interpretation of cluster Ⅱ. (3) On the optimal spatial scale,urban construction land has a high interpretation rate for water quality of each cluster,and is positively correlated with most of the water quality indicators.

Key words: reticular river network area, self-organization feature mapping(SOM), K-means algorithm, spatial heterogeneity, redundancy analysis(RDA)

中图分类号: