Geologic time series from deep-sea sediments: Time scales and distortion by bioturbation

Abstract

In order to evaluate the time- and frequency-domain characteristics of geologic time series obtained from deep-sea sediments, two important problems must be considered. First, an accurate time scale must be established for the sediment records to be studied. An accurate time scale is a basic requirement for such techniques as spectral and cross-spectral analysis. Secondly, the distortion of a geologic time series by sedimentation processes such as bioturbation must be considered if inferences are to be made about the mechanisms controlling climatic and oceanographic processes reflected in the geologic record. Recently, it has been suggested that the periodic components in paleoclimatic records which can be related to variations in the Earth's orbital parameters, can be used as a means for providing very detailed control for late Pleistocene sediment sections. A time scale is estimated by tuning an individual frequency component in a climate record to an orbital parameter. The tuning procedure commonly uses the technique of bandpass filtering which removes all frequency components except the one equivalent to an orbital parameter and then adjusting the geologic time scale so that the filtered record and orbital record align. Since bandpass filtering will produce a periodic record from any data set, a question is raised as to the circularity of this tuning approach. A series of numerical experiments using synthetic time series are used to determine if the tuning of a geologic record artificially induces improved correlation and increased periodic character in a climatic record. These experiments indicate that the tuning does not artificially induce these effects. Thus, it is suggested that the careful use of orbital tuning provides a powerful means of defining a geologic time scale. The degree to which bioturbation distorts a paleoclimatic record is studied by comparing the spectra of oxygen-isotopic records obtained from cores with sedimentation rates ranging from 1 to 12 cm 1000 yr −1. It is shown that the systematic change in the variance spectra of these records that is predicted by models of bioturbation cannot be observed for sedimentation rates greater than 2 cm 1000 yr −1. This can be explained in part if bioturbation is a function of sedimentation rates. If this is the case, then bioturbation must proceed to depth of tens to hundreds of centimeters. However, it seems likely that the degree to which bioturbation alters the climate signal preserved in deep-sea sediment cores has been overestimated.