Paleogeographic controls on the evolution of Late Cretaceous ocean circulation

Abstract

Understanding of the role of ocean circulation on climate during the Late Cretaceous is contingent on the ability to reconstruct its modes and evolution. Geochemical proxies used to infer modes of past circulation provide conflicting interpretations for the reorganization of the ocean circulation through the Late Cretaceous. Here, we present climate model simulations of the Cenomanian (100.5–93.9 Ma) and Maastrichtian (72.1–66.1 Ma) stages of the Cretaceous with the CCSM4 earth system model. We focus on intermediate (500–1500 m) and deep (> 1500 m) ocean circulation and show that while there is continuous deep-water production in the southwestern Pacific, major circulation changes occur between the Cenomanian and Maastrichtian. Opening of the Atlantic and Southern Ocean, in particular, drives a transition from a mostly zonal circulation to enhanced meridional exchange. Using additional experiments to test the effect of deepening of major ocean gateways in the Maastrichtian, we demonstrate that the geometry of these gateways likely had a considerable impact on ocean circulation. We further compare simulated circulation results with compilations of εNd records and show that simulated changes in Late Cretaceous ocean circulation are reasonably consistent with proxy-based inferences. In our simulations, consistency with the geologic history of major ocean gateways and absence of shift in areas of deep-water formation suggest that Late Cretaceous trends in εNd values in the Atlantic and southern Indian oceans were caused by the subsidence of volcanic provinces and opening of the Atlantic and Southern oceans rather than changes in deep-water formation areas and/or reversal of deep-water fluxes. However, the complexity in interpreting Late Cretaceous εNd values underscores the need for new records as well as specific εNd modeling to better discriminate between the various plausible theories of ocean circulation change during this period.