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

• Published in: Environmental science & technology Vol. 20; no. 5; pp. 467 - 473
• Main Authors: Bernard, Paul C, Van Grieken, Rene E, Eisma, Doeke
• Format: Journal Article
• Language: English
• Published: United States American Chemical Society 01-05-1986
• ISSN: 0013-936X; 1520-5851
• DOI: 10.1021/es00147a005

A commercially available scanning-electron-microscope-based automatic image analyser, the Particle Recognition and Characterization program, combined with multivariate analysis techniques, was used to classify particulate matter in estuaries. The procedure developed used cluster analysis of the particles of individual samples, followed by nearest centroid sorting over all samples from the study area. The procedure was applied to suspended particulate matter from the Ems estuary; 13 geochemically-different particle types were identified, the relative abundance of which could be followed through the estuary and interpreted in terms of estuarine mixing processes.