--> ABSTRACT: Correlation Coefficient Matrices, a New Graphic Tool for the Evaluation of Biostratigraphic Data, by Robert L. Ravn; #91020 (1995).

Datapages, Inc.Print this page

Correlation Coefficient Matrices, a New Graphic Tool for the Evaluation of Biostratigraphic Data

Robert L. Ravn

Sequence stratigraphy places both new emphasis and new demands on biostratigraphic control in depositional interpretation. Traditional methods of evaluation and display of biostratigraphic data do not always adequately address sequence stratigraphic issues. Micropaleontological populations reflect the history of sediment deposition. Overall populations within a single depositional sequence tend to be more similar to one another than do populations compared across a sequence boundary. Populations of any two samples can be compared mathematically using any of several simple formulae. The resultant figures, or correlation coefficients, represent an objective measurement of relative sample similarity. The comparison of each sample in a single well against each other sample pr duces a matrix of these correlation coefficient values, which often displays vertical variations in similarity that reveal positions of possible boundaries. These events may not always coincide with conspicuous extinction/appearance events that have historically been used as the major interpretive tool for the biostratigrapher.

Correlation coefficient matrices offer several advantages for the evaluation of biostratigraphic (and potentially other) data. They are visually intuitive. They are strongly objective, being produced directly from the data, without intervening subjective interpretation filters. They may be derived from any presence/absence data set of adequate size, independent of taxonomy or other interpretation, permitting direct visual comparison of data sets from different sources, or even of different disciplines. They can be produced on a normal desktop computer having moderate memory capacity, and they require only inexpensive non-specialized software. Older data sets can yield valuable new information using this method.

AAPG Search and Discovery Article #91020©1995 AAPG Annual Convention, Houston, Texas, May 5-8, 1995