Fig. 1. Magnetic component inclination for the C4Ar.1n–C5r.1n interval from ODP Site 1092 [1] compared to the geomagnetic polarity timescale of Cande and Kent [3] and [4]. rmcd = revised meters composite depth [2].

Fig. 2. Oxygen isotope records from the C4An–C5r.1n interval at ODP Site 1092. The top frame shows the output of a gaussian filter centered at a frequency of 0.78 m^− 1 applied to the stacked δ¹⁸O record. The stacked δ¹⁸O record with numbered obliquity cycles is shown superimposed on the same record with a 5-point smoothing. The three δ¹⁸O records from different planktic and benthic foraminiferal species were used to generate the stack.

Fig. 3. a) Power spectrum generated from the oxygen isotope stack in the depth domain (solid line). b) Dashed line is the power spectrum generated from the ETP target [15] and the solid line is the power spectrum generated from the stacked oxygen isotope records after tuning. c) Coherence between the δ¹⁸O stack and the ETP target curve, line indicates 95% confidence limit for coherence peaks.

Fig. 4. Upper plot shows the correlation of the filtered (filter centered at 0.0244 ± 0.0073 kyr^− 1) oxygen isotope stack to the astronomical solution for obliquity [15]. Lower plot shows the correlation of the three oxygen isotope records from Site 1092 to the ETP solution [15]. Crosses mark the tie points between the oxygen isotope stack and the ETP curve. Shaded areas indicate critical intervals in the correlation between the records that facilitate an unambiguous match between the oxygen isotope record and the ETP astronomic solution.

Fig. 5. Interval sedimentation rates for the C4Ar.1n–C5r.1n interval calculated using the new astrochronology. Asterisks indicate the position of polarity excursions within C5n.2n.

Fig. 6. Comparison of the age estimates of polarity chrons at ODP Site 1092 (this paper) to the timescale of Cande and Kent [3] and [4], to the ATNTS2004 timescale [11], and to the timescales of Hilgen et al. [4] and [31], Abdul Aziz et al. [12] and Husing et al. [26].

Fig. 7. The Site 1092 relative paleointensity record for C5n.2n [1] (bottom), the deep-tow magnetic anomaly record from the East Pacific Rise at 19°S (middle) and the revised North Pacific Stack [38]. Numbering on the revised NE Pacific stack is after [38]. Arrows from above indicate the proposed correlation [38] of CK92 cryptochrons to the revised N. Pacific Stack and the EPR 19°S deep-tow record. Arrows from below indicate our preferred correlation of the polarity excursion chrons to the deep-tow record.

Astronomical ages from recent timescales compared with those inferred at ODP Site 1092

Errors on Site 1092 astronomical ages take into account the uncertainty in the exact position of the magnetic reversal boundaries but not the error associated with the astronomical solution (< 20 kyr at 10 Ma) or the error from a possible delay in remanence acquisition. Numbers in parentheses indicate the difference between Site 1092 estimates (this paper) and other timescales. CK95 — Cande and Kent [4], ATNTS2004 [11], A2003 — Abdul Aziz et al. [12], S1995 — Shackleton et al. [28], H1995 — Hilgen et al. [13], H2000 — Hilgen et al. [31], H2003 — Hilgen et al. [14], Hus2007 — Husing et al. [26]