Obliquity Drives ~100-ky Glacial Cycles Since the Appearance of Major Northern Hemisphere Ice Sheets

Abstract:

It is now established that variations in Earth's orbital configuration pace the late-Pleistocene ~100 ky glacial cycles. It remains unclear, however, whether orbital variations drive these glacial cycles or are merely phase-locked to an internal climate mode. The only statistically significant evidence linking the amplitude of ~100 ky cycles to the amplitude of orbital variations has ~100 ky cycles appearing when eccentricity is weakest (Lisiecki, 2010), seemingly favoring the internal-mode hypothesis. We conduct a Bayesian analysis of the links between orbital amplitude modulations and ~100 ky amplitude in d18O over the epoch of Northern Hemisphere glaciation, 2.8-0 Ma. The amplitude of ~100 ky variations is estimated in a high-resolution benthic d18O record from the North Atlantic using Empirical Nonlinear Orbital Fitting (Liautaud et al., 2020), and statistically significant ~100 ky cycles are identified in four intervals spanning 2.8 to 2.5 Ma, 2.4 to 2.0 Ma, 1.4 to 0.9 Ma, and 0.8 to 0 Ma. Using a Bayesian logistic regression model, we evaluate the probability that ~100 ky cycles are present given the amplitude of insolation forcing. The relative contributions of obliquity and climatic precession to the forcing, as well as the phase of precession, are jointly inferred with the logistic regression parameters. We find that ~100 ky variability appears when the amplitude of obliquity forcing is largest (p<0.001), aligning with maxima in 1.2-My obliquity amplitude modulations. Earlier findings of an anti-correlation between eccentricity and ~100 ky amplitude are proposed to be an artifact that arises because eccentricity and obliquity contain amplitude modulations are anti-phased with respect to each other. Our results suggest that ~100 ky cycles are not simply internal variations that are phase-locked to orbital variations, but are fundamentally generated by large-amplitude orbital forcing.