摘要：In this study, spatial and seasonal variations of water quality in the Morava River system in Serbia were evaluated using multivariate statistical techniques, such as cluster analysis, principal component analysis, factor analysis and discriminant analysis. Water quality data set was collected during the period of 8 years by monitoring 18 parameters at 14 sampling stations. The hierarchical cluster analysis classified monitoring stations into 3 clusters, i.e. relatively less polluted (LP), medium polluted (MP), and high polluted (HP) areas, reflecting thus different chemical properties and pollution levels. FA/PCA applied to the data sets of the 3 groups obtained from cluster analysis, resulted in 6, 4 and 6 latent factors explaining water quality data sets of LP, MP and HP areas, respectively. The varifactors obtained from factor analysis indicate that the parameters responsible for water quality variations are mainly related to mineral components and temperature (natural), nutrients, and organic matter pollution (anthropogenic) out of point and non-point sources. The discriminant analysis offered an important data reduction as it only used 2 parameters (water temperature, calcium) for the temporal analysis, and 3 parameters (pH value, electrical conductivity, potassium) for the spatial analysis.
关键词：The Morava River system; surface water quality;
cluster analysis; factor analysis; principal component analysis; discriminant analysis
