Abstract
Multivariate discriminant analysis of sediment grain-size data from a series of sub-Antarctic deep-sea sedimentary cores has yielded equations which differentiate between terrigenous, siliceous, and calcareous components. As a test of the methods, grain-size distribution data for 106 specimens from two sedimentary cores were examined using discriminant-predictive equations. A modified discriminant technique is used. The predicted sediment groups for these samples are in good agreement with independent microscopic observations. It is proposed that multivariate discriminant analysis of grain-size data provides automated data collection and rapid diagnosis of sedimentary type for restricted regions. This, in turn, facilitates genetic interpretations of sedimentary texture and depositional processes.
