A radiometric calibration of the SPECMAP timescale

The astronomical theory of climate change asserts that Earth's climate is affected by changes in its orbit, which vary the seasonal and latitudinal distribution of solar radiation. This theory is the basis of the orbitally tuned SPECMAP timescale. A key constraint for this important chronology was the mid-point of the Penultimate Deglaciation, initially dated to 127,000 years ago. Recent work suggests this event may be considerably older, casting doubt on the astronomical theory, the SPECMAP timescale, and the accuracy of orbitally tuned chronologies. Difficulties with U/Th coral dating of sea-level events have impeded progress on this problem, because most corals are not closed systems. Here, we use a new approach to U/Th dating that corrects for open-system behavior and produces a sea-level curve of sufficient resolution to confidently correlate with SPECMAP over the last 240,000 years, permitting a reassessment of both this critical chronology and a central tenet of climate change theory. High-precision ages for 24 oxygen isotope events provide a 240,000-year chronology for marine 6180 records that is independent of orbital tuning assumptions. Although there appear to be significant differences between the radiometric and orbitally tuned timescales near the lastglacial maximum and at the Marine Isotope Stage 7/6 boundary, a comparison of radiometric and SPECMAP ages for identical isotope events suggest that the SPECMAP timescale is quite accurate and that its errors were, in general, overestimated. Despite suborbital complexity, orbital cyclicity is clearly evident in our record. High-amplitude sea-level oscillations at periods greater than similar to 20,000 years are very close in phase to summer insolation in the Northern Hemisphere. Although sea-level changes cannot be uniquely tied to a specific season or latitude of insolation forcing, the simplest explanation is that long-period, high-amplitude sea-level change is linked to Northern Hemisphere insolation forcing. These results validate the principles of orbital tuning and suggest such timescales are generally robust. (c) 2006 Elsevier Ltd. All rights reserved.