摘要：In order to demonstrate that silicate (SiO3-Si) can be used as an indicator to study upwelling in the northern South China Sea, hierarchical cluster analysis (CA) and principle component analysis (PCA) were applied to analyse the metrics of the data consisting of 14 physical-chemical-biological parameters at 32 stations. CA categorized the 32 stations into two groups (low and high nutrient groups). PCA was applied to identify five Principal Components (PCs) explaining 78.65% of the total variance of the original data. PCA found important factors that can describe nutrient sources in estuarine, upwelling, and non-upwelling areas. PC4, representing the upwelling source, is strongly correlated to SiO3-Si. The spatial distribution of silicate from the surface to 200 m depth clearly showed the upwelling regions, which is also supported by satellite observations of sea surface temperature.
关键词：Multivariate ; statistical analysis ; Remote sensing ; Upwelling ; Silicate ; South China Sea
