Deep-sea hiatus record reveals orbital pacing by 2.4 Myr eccentricity grand cycles

The study investigates whether long-period astronomical “grand cycles,” particularly the ~2.4 Myr eccentricity modulation linked to the precession of perihelia of Earth (g3) and Mars (g4), are recorded in a global, long-duration stratigraphic dataset of deep-sea hiatuses. Building on the foundation that Milankovitch-band cycles (19–23 kyr precession, 41 kyr obliquity, 100 and 400 kyr eccentricity) pace climate, the authors test if million-year-scale eccentricity cycles are imprinted in the frequency of hiatus formation—breaks in sedimentation caused mainly by deep-sea current erosion. The purpose is to assess the presence, stability, and perturbations of ~2.4 Myr cyclicity through the Cenozoic and to evaluate links to orbital solutions and to major climatic/oceanographic events, including potential chaotic transitions in the inner Solar System near the PETM.

Prior work identified the 2.4 Myr eccentricity cycle and related long-period signals in diverse geological archives, including Cenozoic δ18O and δ13C records, geophysical stratigraphic signals, and fossil assemblages (e.g., Boulila 2012; Kocken et al. 2019; Crampton et al. 2018). Most grand-cycle studies rely on relatively short (<10 Myr) high-resolution records that resolve multiple 405 kyr eccentricity cycles but rarely resolve 2.4 Myr modulation. Astronomical solutions (La2004; La2010 variants; La2011) are reliable to ~40–50 Ma but diverge earlier due to Solar System chaos, motivating geological constraints. Previous studies also indicate climate-system nonlinearity and reduced predictability in icehouse intervals (Westerhold et al. 2020) and identified potential chaotic transitions in the early Cenozoic (e.g., Westerhold et al. 2017; Zeebe & Lourens 2019).

Data: A merged global dataset of 370 deep-sea hiatuses spanning 70–0 Ma from 293 DSDP/ODP/IODP drill holes (Dutkiewicz & Müller 2022) with age-depth models sourced from the NSB database and converted to GTS2020. Hiatuses (≥0.18 Myr duration) were identified as zero-slope intervals in age-depth models and sampled every 100 kyr to detect presence/absence. Age uncertainties are typically ±0.37 Ma for the Cenozoic, smaller for orbitally tuned sections. Sites span all major ocean basins, crustal types, and a wide range of paleodepths. Preprocessing: Long-term trends in hiatus-frequency time series were removed using loess weighted averages (10% weighted average) to emphasize higher-frequency variability. Spectral analyses: The multi-taper method (MTM) with three 2π prolate tapers and harmonic F-test identified significant periodic components; results were compared with unsmoothed periodograms and AR(1) red-noise confidence levels (95% and 99%). Band-pass filters (Gaussian; and Taner for some analyses) isolated the ~2.4 Myr band and other frequencies. Evolutive FFT amplitude spectrograms (window = 6 Myr; step = 0.1 Myr) tracked temporal evolution of dominant periodicities and identified intervals of unimodal stability and bifurcations. Astronomical comparisons: Band-pass filtered 2.4 Myr components of the hiatus series were compared with eccentricity signals from La2004, La2010a–d, and La2011 over 50–32 Ma (within nominal solution reliability). Phase relations and coherence were quantified using 2π-MTM cross-spectral analysis (interval 42–32 Ma). Additional low-frequency Taner filtering of eccentricity models and hiatus data (cutoffs 0–1.8 cycles/Myr) and evolutive FFT spectrograms were used to assess potential chaotic transitions in the g4−g3 term (~2.4 Myr) between 70–40 Ma, focusing on perturbations near ~56 Ma. Interpretive framework: Mechanistic linkage invokes amplitude modulation of Milankovitch cycles, whereby eccentricity maxima enhance insolation seasonality, amplify warming (notably at high latitudes), strengthen deep-ocean circulation and mesoscale eddies, and increase bottom-current erosion, thereby raising hiatus frequency.

• The global Cenozoic hiatus-frequency record exhibits dominant cycles with periods of ~2–3 Myr, with four intervals of relatively stable unimodal grand-cycle occurrence: 70–56 Ma, 50–34 Ma, 30–22 Ma, and 15–0 Ma.
• Dominant periodicities per interval: 2.34 Myr (70–57 Ma), 2.31 Myr (50–32 Ma), 3.15 Myr (30–20 Ma), and 2.56 Myr (15–0 Ma). These are within or near the expected 2.4 Myr eccentricity modulation band, with longer periods post-34 Ma likely augmented by non-orbital forcings.
• Three bifurcation intervals with reduced amplitudes and frequency splitting occur at onsets ~56 Ma, ~34 Ma, and ~22 Ma, indicating disturbances to the background 2.4 Myr pacing.
• Cross-spectral coherence between hiatus-frequency and eccentricity models at the ~2.4 Myr band is strong (coherence 0.83–0.93 for 42–32 Ma), confirming common cyclicity.
• Phase relationships: For 40–32 Ma, the hiatus cycles are slightly phase-shifted relative to the 2.4 Myr band across models. Between 50–40 Ma, most models show increased phase shifts, except La2004, which remains in phase with hiatus cycles at 50–46 Ma. La2004 and La2010a are in phase from 44–0 Ma; La2010a diverges from La2010b at ~48 Ma; La2010b and La2010c remain in phase over 50–32 Ma; La2010d and La2011 are nearly in phase over 50–32 Ma.
• Evidence for a chaotic orbital transition at ~56 Ma: Evolutive spectrograms show a perturbation and temporary frequency shift in the g4−g3 eccentricity-related term in both hiatus data and several astronomical solutions (La2004, La2010a–c). The hiatus record indicates a shift from ~2 Myr to ~1 Myr periodicity near 56 Ma, returning to ~2 Myr thereafter, implying a transient chaotic transition lasting ~2.6 Myr (model-dependent durations: ~1.5 Myr in La2010a; ~4.2 Myr in La2004; ~4.8 Myr in La2010b/La2010c).
• Mechanistic interpretation: Hiatus maxima align with eccentricity maxima and enhanced insolation/seasonality, promoting intensified deep-water circulation and erosive bottom-current activity, consistent with modeling and proxy evidence that warmer climates exhibit stronger eddy activity and abyssal flow.
• Post-34 Ma longer cycles and bifurcations are attributed to superimposed tectonic/oceanographic changes, including opening/closure of key gateways (Drake Passage, Tasman Gateway, Fram Strait, Greenland-Scotland Ridge) and Tethys closure, influencing AMOC/ACC strength and sediment redistribution.
• The early Eocene bifurcation (~57–52 Ma) coincides with severe narrowing of the Norwegian-Greenland Seaway and frequent short-lived warming events, including the PETM (~56 Ma), temporally aligning with the inferred chaotic resonance transition.

The results demonstrate that a global, discontinuity-based stratigraphic dataset encodes the ~2.4 Myr eccentricity grand cycle, indicating orbital pacing of deep-sea erosion and hiatus formation through the Cenozoic. This addresses the central question of whether million-year-scale astronomical forcing is traceable beyond traditional continuous proxy records. Mechanistically, eccentricity maxima increase insolation seasonality and global temperatures (notably polar amplification), intensify deep-ocean circulation and mesoscale eddies, and thus enhance bottom-current erosion, elevating hiatus frequency. The observed strong coherence with astronomical solutions supports this linkage. Perturbations to the baseline 2.4 Myr pacing occur during major reorganizations of ocean circulation driven by tectonic gateway changes (e.g., Eocene–Oligocene transition, Miocene reconfigurations) and during a likely chaotic resonance transition near 56 Ma. The alignment of a g4−g3 frequency shift in both geological and astronomical data strengthens the case for using sedimentary records to constrain the timing and nature of Solar System chaos. Additionally, the divergence between hiatus-cycle frequencies and orbital solutions during icehouse intervals suggests a more stochastic climate–ocean response, consistent with prior studies. These findings have dual significance: they provide independent validation and constraints for astronomical solutions beyond 40–50 Ma, and they imply that warmer climates are associated with more vigorous deep-ocean circulation, relevant for anticipating future ocean responses to warming.

A 70 Myr-long compilation of deep-sea hiatuses reveals pervasive ~2.4 Myr eccentricity pacing, intermittently disrupted by tectonic and oceanographic reorganizations. The hiatus maxima likely correspond to eccentricity maxima via increased insolation and seasonality, enhancing abyssal circulation and erosive bottom currents. Robust coherence with multiple eccentricity solutions confirms the orbital imprint, while a transient chaotic transition at ~56 Ma—coincident with the PETM—is evident as a temporary shift in the g4−g3 periodicity. These geological constraints aid refinement of astronomical models in deep time and link climate warmth to stronger deep-ocean dynamics. Future work should: (i) expand high-resolution proxy datasets to refine phase relationships and test causality between PETM dynamics and orbital chaos; (ii) improve spatial coverage in underrepresented basins; (iii) integrate coupled climate–ocean modeling to quantify erosion thresholds under eccentricity modulation; and (iv) further evaluate different astronomical solutions using multi-proxy stratigraphic benchmarks.

• Astronomical solutions diverge and become uncertain beyond ~40–50 Ma due to chaos; correlations prior to ~50 Ma should be interpreted with caution and require geological validation.
• Age model uncertainties (typical ±0.37 Ma; section-dependent) and unknown volumes of removed sediment introduce uncertainty in hiatus timing and duration.
• Detection of carbonate dissolution events is challenging; brief (<10 kyr) events are too short to appear as hiatuses, though the dataset suggests dissolution-related hiatuses are few.
• Spatial sampling is uneven, with relative underrepresentation of the South Pacific, Southern Ocean, and Mediterranean, and temporal increases in hole numbers toward the present.
• Identification of chaotic transitions is more sensitive to onset detection than to midpoints or terminations; duration estimates are model- and method-dependent and thus uncertain.
• Post-34 Ma cycles likely include significant non-orbital forcings, complicating isolation of pure astronomical signals.