다변량 분석을 이용한 도심하천 수질의 시간적 변동특성 분석

This study investigates the applicability of multivariate statistical techniques including cluster analysis(CA), factor analysis(FA) and principal component analysis(PCA) for the assessment of seasonal or temporal variations in the urban stream water quality. The characteristics of water quality for each streams were evaluated based on water quality data in 2005-2013 using a statistical package, SPSS ver.18.0 and R ver.3.2.2. Cluster analysis for the time variation was resulted in 2 groups(summer and from fall to spring) in which rainfall and temperature were shown to contribute to the clustering. Four factors were found as responsible for the data structure explaining 62.4%~75.8% of the total variance in the Nam stream and five factors were explaining 69.4~74.1% in the Changwon stream. Factor analysis showed main factors changed according to the seasonal characteristics with different pattern for each stream.