CO2 Does Not Drive Tropical SST Cooling during Glaciation: The Case from the Eastern Tropical Pacific

Abstract:

CO₂ is widely considered the key mechanism for translating Northern Hemisphere Milankovitch forcing into globally synchronous glacial-interglacial cycles. A major line of reasoning in support of CO₂ is its apparent correlation with tropical SSTs. I reassess this argument with two new Mg/Ca reconstructions from the eastern tropical Pacific stretching back ~230,000 years. These records are combined with previous ones to construct a regional SST composite with unrivaled resolution and data density. Comparison with Antarctic temperature underscores a tight thermal coupling between the two regions down to the details of distinct millennial scale events. Comparison with CO₂ on the other hand demonstrates a more complex relationship. I focus here on a large and robust lead of SST over CO₂ during the 5e-d transition, implying that CO₂ is not the main driver of tropical SST during glacial inception. Based on independent evidence that deep-ocean temperatures also cooled in advance of CO₂, I propose an alternative mechanism: the cooling signal originates in the surface North Atlantic from Milankovitch forcing, enters the deep ocean with North Atlantic Deep Water (NADW), and spreads globally with deep ocean circulation. In the Southern Ocean NADW becomes precursor to Antarctic Intermediate Water and Antarctic Bottom Water transmitting the signal to large volumes of the ocean interior. In the eastern tropical Pacific the signal emerges when Antarctic Intermediate Water upwells to the surface, spreads westward with the large-scale equatorial currents and enters the Indian Ocean with the Indonesian Through Flow, synchronizing the Indo-Pacific tropical SST response. In this view ocean circulation does the “heavy lifting” in globalizing the climatic cooling during glaciation, but CO₂ may play a more prominent role during deglaciation, which is dynamically not the reverse process of glaciation.