Linear and non-linear response of late Neogene glacial cycles to obliquity forcing and implications for the Milankovitch theory

Constraints are given for the geometry and time lags of the prominent obliquity-paced glacial stages 100, 98 and 96, which mark a major phase in Northern Hemisphere (NH) glaciations during the late Pliocene (2.56–2.4 Ma ago). For this purpose a high-resolution benthic δ18O record was constructed from the astronomically tuned Mediterranean ODP Site 967 and decomposed into an ice volume and an annual NH (40–80° N) temperature component using an inverse modelling approach. Our results indicate that the dominant 41 ky component in δ18O lags obliquity by 6.5 ± 0.6 ky, which approximates late Pleistocene estimates. Maximum (minimum) ice volume growth occurred in phase with obliquity minima (maxima), which invoked low (high) total summer energy conditions that reduced (increased) ice-sheet ablation. Sea level dropped 60–70 m during full glacial conditions. Similar to late Pleistocene ice core and marine δ18O records, our late Pliocene δ18O record reveals significant power at ∼28 ky, which appear to be bound to the major glacial terminations. We argue that this beat most likely reflects the sum frequency of the 41 ky prime and its multiples of 82 and 123 ky, supporting the theory that the late Neogene glacial cycles are primarily determined by the linear and non-linear response mechanisms of the ice sheets to the obliquity forcing. Evidence for such a scenario may come from the alignment between the Devils Hole δ18O chronology and the sum of the filtered linear and non-linear obliquity-related components of late Pleistocene climate records, suggesting that the ∼28 ky beat is intrinsic to the climate system or at least an important constituent of the annual high-latitude NH temperature changes that have set the stage for the geometry of the glacial–interglacial variability throughout the course of the Pliocene and Pleistocene.