SPATIAL ANALYSIS OF UNDERGROUND WATER QUALITY USING MULTIVARIATE STATISTICAL TECHNIQUES IN SHUNYI DISTRICT, BEIJING

Multivariate statistics techniques, such as hierarchical clustering analysis and principal basis analysis, were applied for the evaluation of spatial variations of water quality data set generated during 2012 monitoring of 8 parameters at 28 different sites of Shunyi District, Beijing. Hierarchical clustering analysis grouped 28 sampling sites into 4 clusters. Four (or more) typical sampling points from the four clusters could be selected to monitor instead of all the 28 monitoring points. Principal basis analysis gave the results for spatial analysis. It provided an important feature selection as it selected only 4, 3, 4,3 parameters respectively in 4 clusters. Therefore, principal basis analysis allowed a reduction in the dimensionality of the large data set. Thus, this study illustrates the usefulness of multivariate statistical methods in the analysis of reducing the number of monitoring stations and chemical parameter.