Comparison of the tuned ages for polarity reversal boundaries at the five sites in the 1.5- to 2-Ma interval showed that polarity chron ages are in good agreement. For other time intervals there are some significant differences (more than an obliquity cycle) between sites (Table T3). Intervals with enhanced 41-k.y. power in reflectance data are considered more reliable (bold in Table T3). Site 1208 showed the strongest cyclicity, with Site 1207 also showing a clear signal in some intervals, particularly the 2.1- to 2.7-Ma and 4.5- to 5-Ma intervals.
During ODP Leg 138 to the eastern equatorial Pacific, 11 sites were drilled and most of them showed a prominent cyclicity in GRA density. Shackleton et al. (1995) used these cycles in GRA density records to produce an orbitally tuned age model for the 0- to 12.5-Ma interval. They worked entirely in the time domain comparing smoothed GRA density records with the target record of summer insolation at 65°N. In their tuning they assumed that no phase lag existed between insolation and GRA density controlled by proportion of SiO2 and CaCO3 (high density), high carbonate content being associated with high Northern Hemisphere insolation. Age control points were added to the data to align prominent groups of density maxima. The records were broken into 0.8-m.y. intervals for convenient viewing. Each site was tuned independently over the chosen time interval. Shackleton et al. (1995) found that some intervals in these records were more easily tuned than others, similar to results from Leg 198. Shackleton et al. (1995) noted that it was difficult to tune the 0- to 1-Ma interval, which was also the case at four of the Leg 198 Sites (1208, 1209, 1210, and 1211). The 1- to 2-Ma interval for the Leg 138 sites carries a clear 41-k.y. obliquity cycle. For Leg 198 sites, the 1.2- to 1.6-Ma interval also carries a very clear obliquity cycle (Figs. F10, F11, F12, F13, F14). In the 2.4- to 2.6-Ma interval, a very strong obliquity cycle was observed at Site 846 (Leg 138), and this same interval also carries a strong 41-k.y. signal at Sites 1209, 1210, and 1211. Comparison between the Site 1207 age model and ages from Shackleton et al. (1995) indicate consistency for the 1- to 8-Ma time interval (Table T4).
Hilgen et al. (1995) developed an astronomical timescale for the interval from 3 to 9.7 Ma using lithologic cyclicity seen in sedimentary sections in the Mediterranean. These sections comprise open marine sediments that alternate between carbonate-rich and carbonate-poor marls or homogeneous marls and sapropels. The individual sapropels were related to precession minima and the clusters of sapropels to the 400-k.y. eccentricity cycles. In tuning the section, the target curve used was the 65°N summer insolation curve. To obtain an astronomical age for the youngest polarity reversal in the sequence, Hilgen et al. (1995) took the Shackleton et al. (1995) age for the onset of Subchron C3An.2n of 6.576 Ma. They then matched the lithologic cycles in the section to the astronomical solution using the correlation of sapropel clusters to eccentricity. The age of the calibration point (6.576 Ma) had to be adjusted to 100 k.y. older to establish a consistent correlation between sapropel clusters and eccentricity maxima. The ages from Hilgen et al. (1995) differ significantly from those from Leg 198 in the 6- to 8-Ma interval (Table T4). At the top of Subchron C3Bn the difference is >200 k.y. In the interval from 7.2 to 8.1 Ma, the difference is ~100 k.y., which is the amount of adjustment of the 6.576-Ma tie point used by Hilgen et al. (1995) for the age of the youngest polarity reversal in their section.
ODP Site 926 on the Ceara Rise also produced an orbitally tuned timescale from 5 to 14 Ma (Shackleton and Crowhurst, 1997). This timescale cannot be directly compared with the Leg 198 timescale because of a lack of polarity reversals at Site 926. Backman and Raffi (1997) used the cyclostratigraphic age model from Site 926 to calibrate ages of the calcareous nannofossil datums for the late Miocene. These ages were then compared with the biomagnetochronology from Site 853 (ODP Leg 138). The center of the peak in abundance of transitional morphotypes of Triquetrorhabdulus rugosus at Site 853 occurred 120–130 k.y. after the corresponding peak at Site 926. The age estimates of Hilgen et al. (1995) were then applied to the Site 853 data and the peak center was found to coincide at Sites 853 and 926. Therefore, Backman and Raffi (1997) considered that the Hilgen et al. (1995) ages are more reliable in this interval than the ages of Shackleton et al. (1995).
Lourens et al. (in press) recalibrated the Miocene astronomic timescales of Shackleton and Crowhurst (1997) and Hilgen et al. (1995) using the astronomic solution of Laskar et al. (2004). For the last 13 m.y., the retuning resulted in almost negligible changes in the ages of reversal boundaries (Lourens et al. in press). For the 6- to 8-Ma interval, the ATNTS2004 is in close agreement with that of Hilgen et al (1995) and therefore differs significantly from the results of this study.