基于信息筛选和拉依达准则识别地下水主要组分水化学异常的方法研究

地下水水化学组分的异常值识别是获取水化学背景值的重要的一环,以往提出的基于水化学各组分内在联系的计算背景值的水化学图法及Durov图法在柳江盆地的研究中均取得了良好的效果,但是水化学图法的计算步骤繁多,过程复杂,对采样精度要求严格.因此,为简化计算过程,快速识别异常,本文借鉴英国地调局识别异常的方法,以沙颍河流域为研究区,针对该区浅层地下水水质资料,利用信息筛选法替代水化学图法,对研究区内水化学主要组分进行异常识别,并将不同识别方法的识别效果进行对比分析.结果表明,尽管存在两种方法单独识别出的异常数据较多,但是水化学图法+拉依达准则和信息筛选法+拉依达准则对地下水主要组分的异常识别效果均较好,异常识别后的剩余数据的阈值范围较为一致,信息筛选法能够在大尺度区域内精度较低条件下,有效地替代水化学图法快速识别出异常值.

Research on the method to identify the outliers of the main components of groundwater based on the information screening coupled with PauTa criterion

Identification of outliers of hydrochemical components of groundwater is an important part of obtaining hydrochemical background values. As proposed by the study of the Liujiang River Basin for calculating the background values, the hydrochemistry diagram method and the Durov diagram method based on the water chemical groups with tight internal relations have achieved good results. Those methods are strict to the sampling precision. In order to simplify the various and complex calculation processes. In this study, instead of the hydrochemistry diagram method, the information screening method suggested by the British Geological Survey (BGS) was applied to identifying the outliers of the main regular indices of shallow groundwater in the Shayinghe River Basin. Then the two identifying methods were compared. The result shows that both of the hydrochemistry diagram method coupled with PauTa criterion and the information screening method coupled with PauTa criterion are satisfactory, although there are more outliers identified by each method alone. The threshold values of the remaining data after identifying outliers for both methods are close. The information screening method proved to be an effective method for replacing the hydrochemistry diagram method. Besides, it can be used quickly and efficiently in large scale region with low sampling precision.