A use of network diagrams in depicting stratigraphic time-correlation

Abstract

Time-correlation among palaeontological samples or other items of geological information located in measured sections can be depicted graphically as a network, in which the samples, represented by points, are linked by arrows, each of which represents a decision as to which of the samples linked is the older and which the younger. The idea of a correlation network stems naturally from the concept of correlation as a process of ordering, not equating, items of geological information in time. Examples of correlation networks are presented for a palynological reconnaissance of Triassic-Jurassic sections in eastern Svalbard, and for correlation of Lower Carboniferous sections in North Yorkshire based on calcareous microfossils.

Revision of the Dinantian stratigraphy of the Craven Basin, N England

“…The more successful use of goniatites in time correlation is restricted to the higher parts of the sequence, where Bisat (1928) founded the locally useful Cracoean (B) and Bollandian (P) stages, each subdivided into zones. The distribution of the foraminifera and other stratigraphically important calcareous microfossils seen in thin section was described for the rock units of the eastern area by Fewtrell & Smith (1978), while time-correlation based on this information, but avoiding the establishment of zones, was given elsewhere (Smith & Fewtrell, 1979). The availability and diversity of both foraminifera and conodonts in the limestones of the Craven Basin has already been demonstrated by Fewtrell & Smith (1978) and Metcalfe (1978).…”

Techniques for quantitative stratigraphic correlation: a review and annotated bibliography

“…Is stratigraphic correlation about equating samples from section to section, for instance, or is it about ordering them (Hughes & Moody-Stuart, 1969)? Is it ever adequate to indicate contemporaneity, or is that itself an admission of defeat (Smith & Fewtrell, 1979)? Can massaging the fossil record, in whatever way, be expected to reveal an underlying time signal, or must hypotheses about patterns of faunal replacement always be used to drive the interpretation process?…”

“…The contrast between the deterministic and probabilistic attitudes (7 A and 7B respectively) can be demonstrated in most areas of quantitative stratigraphic correlation, both lithostratigraphic and biostratigraphic. Segmentation of time-series, for instance, can be carried out on either basis (Kulinkovich, Sokhranov & Churinova, 1966, in contrast to Gill, 1970, as can sequencing (Smith & Fewtrell, 1979, in contrast to Blank, 1979), or pairwise comparison of time-series (Christopher, 1978, in contrast to Shuey et al 1978. Correlations can also be expressed both in deterministic and probabilistic terms, for instance the conventional correlation line between two sections as opposed to a correlation bracket (Hughes & Moody-Stuart, 1969).…”