Chronological constraints on Pleistocene sapropel depositions from high-resolution geochemical records of ODP sites 967 and 968

We completed and merged an existing high-resolution X-Ray Fluorescence (XRF) data set derived from Ocean Drilling Program (ODP) site 968 with new data from adjacent ODP site 967. An astronomical age model spanning the last 1.05 Myr was constructed for the spliced record using the highly linear relation between the elemental ratio of titanium and aluminum in the sediment (Ti/Al) and insolation. This rendered detailed ages for sapropel deposition in the Eastern Mediterranean. Our results imply major revisions of previous sapropel age models below MIS 11, with changes of up to two precession cycles in the interval ~ 450-850 ka of the ODP 967 sapropel chronology. Based on the Ti/Al age model, we find that color reflectance - used as an indicator for sapropels in this interval of ODP site 967 and 968 - is highly incongruent with the Ti/Al proxy as well as insolation forcing during periods of minima in the 405 kyr eccentricity cycle. Our findings indicate that the use of color reflectance as a proxy for so-called ghost sapropels in these intervals is not reliable. Lastly, we demonstrate the presence of a strong obliquity signal in the Ti/Al record that lags obliquity forcing by 4 ± 0.7 kyr). This time lag is only marginally longer than the adopted 2.7 ± 1.1 kyr for the dominantly precession-tuned age model, and is hence much shorter than the generally accepted ice sheet response time to insolation forcing. This suggests that the obliquity-bound changes in Ti/Al are not glacial controlled and most likely reflect changes in low-latitude climate oscillations, such as the monsoon.