Abstract: Scheme for Multivariate Analysis in Carbonate Petrology with Example from Silurian Tonoloway Limestone

Richard Smosna, Steven M. Warshauer

Two methods of exploratory data analysis--a numerical classification system applied in conjunction with an ordination technique--were used to group, classify, and decipher 61 Tonoloway samples from Pinto, Maryland. Q-mode cluster analysis served as an excellent aid in facilitating paleoenvironmental interpretation, as samples of each cluster were assumed to have formed under similar conditions. In the Tonoloway Limestone seven microfacies (clusters) were identified: micrite-pellet of the lower intertidal mud flat; micrite, micrite-dolomite, and dolomite from the upper-intertidal zone; intraclast and ooid-cement representing a shallow-subtidal environment; and pellet-cement from the deeper-subtidal environment. Multidimensional scaling illustrated relations among the clusters and the gradational nature of samples not seen by cluster analysis alone. The gradational nature of samples led, then, to interpretations of environmental gradients that were perpendicular to the ancient shoreline. Hydrodyna ic energy was low on both the tidal flat and offshore near wave base, systematically increasing from both directions toward the shallow-subtidal environment. Petrographic-fossil diversity was highest near wave base with a second, lesser peak in the lower-intertidal mud flat; this variable decreased into the shallow-subtidal high-energy belt and also onto the upper-intertidal mud flat. Dolomitization was most pronounced in rocks of the upper-intertidal zone. We suggest that the scheme of multivariate analysis presented here, because it involves nonparametric methods, is more applicable to typical geologic data than more commonly used strategies.

AAPG Search and Discovery Article #90961©1978 AAPG Annual Convention and Exhibition, Oklahoma City, Oklahoma