GRAPHIC CORRELATION OF OXYGEN ISOTOPE STRATIGRAPHY APPLICATION TO THE LATE QUATERNARY

A geological time scale is only as useful as the stratigraphy to which it is tied. The stratigraphy of deep-sea sediments of the Brunhes chron is largely based on the 19 oxygen isotope stages defined by Emiliani (1955) and Shackleton and Opdyke (1973). To improve the reliability and precision of this isotope stratigraphy, we have applied the technique of graphic correlation (Shaw, 1964) to isotopic events that can be consistently recognized on a global scale. Accordingly, we have devised a numerical taxonomy of isotope stratigraphy to include not only the 19 stage boundaries but also 56 isotopic events that are recorded as delta O-18 maxima or minima within these stages. Because samples are taken at discrete intervals, each event is recorded as a depth range which depends both on the sampling density and the structure of the isotopic record. Graphic correlation proceeds by selecting a reference section (V28-238) and then graphing the depth range of isotopic events that are common to both sections. The overlapping ranges define correlation boxes, within which the true event must lie. The line of correlation must pass through all correlation boxes. Surprisingly. empirical tests have shown that correlation between stratigraphic sections can be accomplished as a series of straight line segments. The number of segments required and their slopes and offsets identify changes in accumulation rate. stratigraphic gaps, and zones of deformation. The line of correlation relates any level in a given core to the standard section and enables all isotopic records to be recast on the basis of a common depth scale. In this form, isotopic records can be stacked( averaged) to construct a global average record that can be used to differentiate between global and local isotopic variations. Preliminary results with 13 deep-sea cores suggest that correlation may be precise to within a few centimeters and will provide an accurate and reliable method for the application of time scales in the Brunhes chron.