PROVISIONAL TIMESCALE

A provisional timescale based on the astrochronology of the P/E boundary at ODP Hole 690 (Röhl et al., 2000) was developed for sediments from Holes 1220B and 1221C. Three common points, recognizable on the ¹³C curves from Site 690, and Holes 1220B and 1221C were chosen for correlation. These points are the onset of the excursion at 55.234 Ma, the minimum ¹³C value at 55.1815 Ma, and the inflection point where carbon isotope ratios recover from the excursion at 55.0011 Ma. The locations of the datum points are indicated on Figures F1 and F2. A timescale for Holes 1220B and 1221C was created by interpolation between these three datum points and can be found in Table T1 and in Figure F3. Sedimentation rates were assumed to be constant above the highest datum point and below the lowest datum point.

The accuracy of this timescale depends on appropriate selection of datum points. Gaps in the isotope record of ~30 cm in Hole 1220B and of ~20 cm in Hole 1221C restrict our ability to identify the exact onset of the carbon isotope excursion. The start of the carbon isotope excursion is unlikely to have been incorrectly located by more than the duration of these gaps. We have assumed that carbonate burndown during the PETM is minimal because errors as large as these would require extensive dissolution of seafloor sediments deposited before the start of the carbon isotope excursion. Nonetheless, it is possible that the isotope record from Sites 1220 and 1221 does not capture the most negative value of the ¹³C excursion either due to dissolution or a hiatus. Despite this potential source of error, we believe the minimum ¹³C value is a reasonable datum point because it is more readily identified in each of the records, whereas attempting to correlate a different point in the isotope curve, particularly a point where the curve changes rapidly, is likely to be more ambiguous. Although the use of biostratigraphic datums would strengthen the accuracy of this timescale, there are intervals in the sedimentary record where this information is not available because the calcareous material is absent. At the moment, there is no well-defined set of biostratigraphic events that can be used to subdivide the interval within the P/E carbon isotope excursion, and cyclostratigraphies have been constructed for only a few sites. For this reason we believe the approach used here provides a good chronological estimate for these two sites. At the very least, we think that our chronology allows us to accurately compare parts of the isotope record that fall before, during, and after the PETM.