Quantification of Water Sources in a Coastal Gold Mine through an End-Member Mixing Analysis Combining Multivariate Statistical Methods

Guowei Liu,Gang Liu,Hongyu Gu,Xueliang Duan,Jie Guo,Fengshan Ma

doi:10.3390/w12020580

Abstract

Mixing calculations have been widely applied to identify sources of groundwater recharge, but these calculations have assumed that the concentrations of end-members are well known. However, the end-members of water remain unclear and are not easily available in practical applications. To better determine end-members and mixing ratios, an end-member mixing analysis combining multivariate statistical methods was used on a large, complex water chemistry dataset collected from the Shashandao gold mine in China. Multivariate statistical methods, including principal component analysis (PCA) and hierarchical cluster analysis (HCA), were applied to determine the specific end-members (these two methods verified each other). On the basis of the identified end-members, a maximum likelihood method was then used to estimate the mixing ratios of the water sources. The combined method proposed in this study can help to identify more accurate end-members and deal with uncertainty in end-member concentrations, and it can also adjust the concentrations until the optimal mixing ratios for the calculation are obtained. This method can be a powerful tool for groundwater management and in predicting water inrush in mining operations.

Highlights

In coastal gold mines, it is essential to accurately predict a mixture’s water sources and mixing ratios
The variables used in the principal component analysis (PCA) and hierarchical cluster analysis (HCA) were standardized according to the formula according to the formula [9]
The HCA showed that the water samples from the −600-m sublevel were divided into four clusters

Summary

Introduction

It is essential to accurately predict a mixture’s water sources and mixing ratios. In hydrology, mixing calculations have often been used to estimate the mixing ratios of a sample (based on an assumption that a mixture has two or more specific end-members) [6]. Laaksoharju et al have created a multivariate mixing and mass balance (M3) method using complex hydrochemical data to trace the original water sources of groundwater and compute its mixing proportions and mass balances [10]. This method uses a multivariate analysis called PCA to summarize the datasets and uses PCA plot mixing calculations

Methods

Results

Discussion

Conclusion