›› 2019, Vol. 33 ›› Issue (4): 488-491.

• 工程实录 • 上一篇    下一篇

基于聚类分析岛礁地下水水质分类方法

李焜耀1,2,唐健健2,陈拥军3,张晨阳2,朱 旭2   

  1. (1.桂林理工大学土木与建筑工程学院,广西桂林 541004;2.中国科学院武汉岩土力学研究所,武汉 430071;3.中国核工业第二二建设有限公司,湖北宜昌 443000)
  • 收稿日期:2018-11-29 修回日期:2018-12-04 出版日期:2019-08-20 发布日期:2019-08-16
  • 作者简介:李焜耀(1994-),男,硕士研究生,研究方向为岛礁地下水淡化。
  • 基金资助:

    国家自然科学基金(41572304);中科院科技先导专项A中科院科技先导专项类子课题(XDA13010301)

Classification Method of Water Quality of Freshwater Lens in Coral Reefs and Islands

LI Kunyao1,2, TANG Jianjian2, CHEN Yongjun3, ZHANG Chenyang2, ZHU Xu2   

  1. (1.College of Civil Engineering, Guilin University of Technology, Guilin 53004;
    2.State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences, Wuhan 430071;3.China Nuclear Industry 22nd Construction Co. Ltd., Yichang 443000)
  • Received:2018-11-29 Revised:2018-12-04 Online:2019-08-20 Published:2019-08-16

摘要: 地下淡水体是珊瑚岛礁的重要淡水资源,但是长久以来学者们对于岛礁地下淡水体水质分类并没有太多研究,一般就是以氯离子浓度600 mg/L为分类界限,而且由于岛礁地下淡水体的特殊性,常用的地下水质量标准以及海水水质标准并不能给我们提供足够的参考。随着项目的发展,以单一的氯离子浓度分类已经无法满足我们对水质的调查以及评价。采取聚类算法对目标值进行分类,它是通过计算各个数据点到聚类中心的距离来判别每个水样的类别。结果表明,结合电导率、氧化还原电位以及浊度的分类结果是比较满意的,并且为现阶段的水质调查评估提供了分类参考。

关键词: 珊瑚岛礁, 淡水透镜体, 地下水分类, Kmeans

Abstract: The freshwater lens below the coral reefs and island is a very important fresh water source, however, not many researches on this aspect are reported. The conventional classification borderline for the saline water is using the 600 mg/L chloride concentration. Due to the uniqueness of the freshwater lens underneath the coral reefs and islands, the conventional groundwater quality and seawater quality standards could not provide enough references for the theoretical study and practical applications. It is recognized that the single chloride concentration classification is unable to meet the investigation and evaluation of water quality for these freshwater lenses. This paper presents a new classification by clustering algorithm, which classifies each water sample by calculating the distance from each data point to the cluster center. The results show that the classification results using SpCond, ORP and Turbidity are satisfactory.

Key words: coral reefs and islands, freshwater lenses, groundwater classification, Kmeans