Cycles were visually identifiable in L* reflectance data from all five of the sites in this study. For Sites 1209, 1210, and 1211, we worked with spliced composite records rather than data from a single hole. Reflectance data were initially placed on the magnetostratigraphic age model based on the polarity timescale of Cande and Kent (1995). Power spectra for untuned sections of reflectance data placed on this age model consistently show a concentration of power at orbital frequencies, particularly around the 41-k.y. obliquity cycle (Fig. F9).
The reflectance records were then tuned to the astronomical solution for obliquity from Laskar et al. (1993), as this was the most visually identifiable cycle in the reflectance data and the power spectra for different time intervals in all the sites showed a concentration of power at the obliquity frequency (Fig. F9). In constructing the astrochronological age model, we assume that there was no phase lag between the orbital forcing and the response. For convenience, the reflectance data were broken up into 1-m.y. intervals when compared to the astronomical solution and each site was tuned independently. Cycles were readily apparent in the reflectance data for all sites, and tuning of the record required a minimum of adjustment of peaks in the reflectance data to the astronomical solution (Figs. F10, F11, F12, F13). Astronomically tuned ages were calculated for polarity reversals in the 1- to 8-Ma interval at Site 1207 (Table T3). At Site 1209, tuning was performed in the 1- to 7-Ma interval and at Sites 1210 and 1211 in the 1- to 5-Ma interval. Site 1208 has also provided an astrochronological age model for the 1- to 6-Ma interval (Fig. F11) and is included in Table T3. The tuned age models are compared to each other (Table T3) and are compared with other recently published astrochronologies for this time period (Table T4). The output of a bandpass filter centered on 41 k.y. is shown below the astronomical solution for obliquity and the raw reflectance data in Figures F10, F11, F12, F13, and F14.
To test the validity of the timescale we used cross-spectral analysis performed using the Blackman-Tukey method and Analyseries software (Paillard et al., 1996). Coherence between the reflectance data and the astronomical solution for obliquity was significant at all the sites, although the coherence values depend on which time interval is being examined. At Site 1207 coherence was ~0.8 for the 1.2- to 1.8-Ma and 6.2- to 6.8-Ma intervals (Fig. F15A). The coherence values at Site 1208 were >0.8 for the entire 1-to 6-Ma interval. Sites 1209, 1210, and 1211 also showed coherence values between 0.8 and 1 (Fig. F15C–F15E).