Study On Natural Background Levels Of Conventional Components In Shallow Groundwater Of The Liujiang River Basin In Qinhuangdao

Groundwater background value is a very important decisive rule for groundwater contamination assessment.And reasonably determine the background value of groundwater in the different units,is the foundation of scientific and reasonable to carry out the evaluation of groundwater pollution.As a case study,the Liujiang River Basin of Qinhuangdao,was divided into four hydrogeological units according to the local hydrogeological conditions.Seven new methods are used to eliminate the outliers,which combined of piper diagram,ion ratio diagrams and a series of mathematical statistics methods such as clustering analysis and Grubbs.Comparing to the effects of different units and the whole study area,as well as the various methods,results are presented to show importance of the division of groundwater units,and the necessary of hydrochemical map methods.Then through the comparsion for the number of outliers,the change of index curves,and boxplot respectively,pointed out the optional eliminating scheme.Finally,the present baselines of each conventional component calculated using the cumulative frequency curves.The main conclusions are as followes:(1)Comparing to the results of outliers in different hydrogeological units and the whole study area showed that: a.The regional difference of the whole area is ignored in the recognition process of outliers,leads to the phenomenon of eliminating too much;b.Hydrogeological features of each unit and influence of human activity have been given adequate consideration,so as to make the results more reliable.(2)Comparing mathematical statistics methods and hydrochemical map methods of groundwater,it is found that the latter is sufficient study the relation between the seven major ions,and can be used as a good tool identifying outliers in groundwater chemical constituents and classifications.(3)Seven new methods are used to eliminate the outliers,then compare the number of outliers,analyze the change of index curves except outliers,and evaluate effect by boxplot respectively.The results show that the new method combines hydrocheical map method and Grubbs has a significant advantage in the aspects of identifying groundwater monitoring data outliers.(4)Analyzing the reasons indicates that outliers in the study area are mainly distributed in some special sites,such as the coal mining area,the population areas and some plants.Coal mining may result in concentration of sulfates increased;the same consequences occurred that plants and septic tanks may lead to sodium and chloride abnormity.It is demonstrated that the method proposed in this study is feasible.