Classification of estuarine particles using automated electron microprobe analysis and multivariate techniques

The suspended particulate matter from a series of stations in an estuary is automatically analyzed for 11 elements on a particle-by-particle basis by using electron microprobe X-ray microanalysis and an automatic image analysis system. The interpretation of the large acquired data set implies numerical multivariate analysis. The developed procedure consists of using cluster analysis on the particles of individual samples, followed by nearest centroid sorting over all samples from the study area. For the Ems estuary, the procedure provided 13 geochemically different particle types of which the relative abundance through the estuary could be followed and interpreted in terms of estuarine mixing processes. The presented methodology could be applied in other studies concerning the characterization of particle populations.