Hydrogeochemical analysis and groundwater pollution source identification based on self-organizing map at a contaminated site

被引:22
|
作者
Zhang, Yaobin [1 ,2 ,3 ]
Zhang, Qiulan [1 ,2 ,3 ]
Chen, Wenfang [4 ]
Shi, Weiwei [4 ]
Cui, Yali [1 ,2 ,3 ]
Chen, Leilei [4 ]
Shao, Jingli [1 ,2 ,3 ]
机构
[1] China Univ Geosci Beijing, Key Lab Groundwater Circulat & Environm Evolut, Minist Educ, Beijing 100083, Peoples R China
[2] China Univ Geosci Beijing, MNR Key Lab Shallow Geothermal Energy, Beijing 100083, Peoples R China
[3] China Univ Geosci Beijing, Sch Water Resources & Environm, Beijing 100083, Peoples R China
[4] Henan Prov Bur Geoexplorat & Mineral Dev, Inst Geoenvironm Survey 1, Zhengzhou 450045, Peoples R China
关键词
Self-organizing map; Contaminated sites; Hydrogeochemical groups; K-means clustering; Pollution source identification; RED-RIVER DELTA; STATISTICAL-METHODS; HETAO BASIN; WATER; QUALITY; SOM; CLASSIFICATION; VISUALIZATION; RESOURCES; SEDIMENT;
D O I
10.1016/j.jhydrol.2022.128839
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
Groundwater contamination at the site has become a very serious problem. A clear understanding of the hydrogeochemical characteristics of groundwater is indispensable for pollution remediation. It requires taking a number of samples and continuous monitoring. However, it is challenging to interpret hydrogeochemical datasets with diverse compositions and wide range of concentration by linear method. In this work, combination of self-organizing map (SOM) and K-means clustering was applied to investigate the hydrogeochemical characteristics at a contaminated site. The results showed that shallow groundwater hydrogeochemical characteristics were performed by 42 neurons and were classified into 5 clusters. The NO3- in cluster 1 widely distributed in the site. The application of fertilizers led to high NO3- concentration in groundwater. Cluster 2 was dominated by Ca2+, Mg2+, Cr(VI) and NO2- and cluster 3 was characterized by TDS, Na+, Cl-, HCO3- and SO42-. Pollutants were mainly from the migration of components at the chromium slag heap under the effect of convection and dispersion. Cluster 4 was dominated by pH, As and CO32-. Furthermore, the pH with the minimum of 8.3 and the presence of CO32- in groundwater provided a favorable opportunity for arsenic enrichment. Pollutants in cluster 4 originated from rainfall leaching on the chromium slag. Moreover, the migration of components from cluster 4 to cluster 2 was also observed by SOM and numerical simulation. Cluster 5 was mainly dominated by Mn and Fe. Reduced environment and anthropogenic activities caused Fe and Mn to exceed standards. The deep groundwater characteristics were performed using 20 neurons and were identified into 4 clusters. Its contamination was due to the leakage of shallow groundwater. Finally, the Gibbs diagram and the saturation index method performed the chemistry control mechanisms of different clusters. This study demonstrated that SOM could be used to interpret nonlinear and complex contamination datasets.
引用
收藏
页数:12
相关论文
共 50 条
  • [21] Hydrogeochemical interpretation of South Korean groundwater monitoring data using Self-Organizing Maps
    Choi, Byoung-Young
    Yun, Seong-Taek
    Kim, Kyoung-Ho
    Kim, Ji-Wook
    Kim, Hyang Mi
    Koh, Yong-Kwon
    JOURNAL OF GEOCHEMICAL EXPLORATION, 2014, 137 : 73 - 84
  • [22] Self-organizing maps for the identification of groundwater salinity sources based on hydrochemical data
    Haselbeck, Valentin
    Kordilla, Jannes
    Krause, Florian
    Sauter, Martin
    JOURNAL OF HYDROLOGY, 2019, 576 : 610 - 619
  • [23] Identification of homogeneous regions for regional frequency analysis using the self-organizing map
    Lin, GF
    Chen, LH
    JOURNAL OF HYDROLOGY, 2006, 324 (1-4) : 1 - 9
  • [24] Hydrogeochemical characterization and suitability assessment of groundwater in a typical coal mining subsidence area in China using self-organizing feature map
    Di Zhao
    Yifan Zeng
    Qiang Wu
    Aoshuang Mei
    Shuai Gao
    Xin Du
    Weihong Yang
    Environmental Earth Sciences, 2022, 81
  • [25] Hydrogeochemical characterization and suitability assessment of groundwater in a typical coal mining subsidence area in China using self-organizing feature map
    Zhao, Di
    Zeng, Yifan
    Wu, Qiang
    Mei, Aoshuang
    Gao, Shuai
    Du, Xin
    Yang, Weihong
    ENVIRONMENTAL EARTH SCIENCES, 2022, 81 (21)
  • [26] Constructing analysis systems with the self-organizing map based on WEB technology
    Ikeda, Y
    Tokutaka, H
    Fujimura, K
    Maniwa, Y
    ICONIP'02: PROCEEDINGS OF THE 9TH INTERNATIONAL CONFERENCE ON NEURAL INFORMATION PROCESSING: COMPUTATIONAL INTELLIGENCE FOR THE E-AGE, 2002, : 778 - 781
  • [27] Regional Groundwater Water Quality Assessment and Contamination Source Identification by a Self-Organizing Map and Entropy Method in Pinggu Basin, Northeast Beijing
    Lv, Shaojie
    Zhang, Zongwen
    Sun, Ning
    Shi, Zheming
    Li, Jia
    Qu, Shen
    FRONTIERS IN ENVIRONMENTAL SCIENCE, 2022, 10
  • [28] Automatic analysis of chinese dialect tones based on self-organizing map
    He, Yifan
    Jiang, Minghu
    ADVANCED INTELLIGENT COMPUTING THEORIES AND APPLICATIONS: WITH ASPECTS OF CONTEMPORARY INTELLIGENT COMPUTING TECHNIQUES, 2007, 2 : 732 - +
  • [29] Explorative data analysis based on self-organizing maps and automatic map analysis
    Franzmeier, M
    Witkowski, U
    Rückert, U
    COMPUTATIONAL INTELLIGENCE AND BIOINSPIRED SYSTEMS, PROCEEDINGS, 2005, 3512 : 725 - 733
  • [30] Kohonen self-organizing map for ideotype identification chickpea agronomic
    Ortiz, Vithoria M.
    Filho, Alberto C.
    Loro, Murilo V.
    Andretta, Joao Augusto
    dos Reis, Mikael B.
    Schuller, Bruno Raul
    SIGMAE, 2024, 13 (04): : 52 - 61