Assessing the potential capability of reconstructing glacial Atlantic water masses and AMOC using multiple proxies in CESM

Abstract

Reconstructing the Atlantic Meridional Overturning Circulation (AMOC) during the Last Glacial Maximum (LGM) is essential for understanding glacial-interglacial climate change and the carbon cycle. However, despite many previous studies, uncertainties remain regarding the glacial water mass distributions in the Atlantic and the AMOC intensity. Here we use an isotope enabled ocean model with multiple geotracers (δ13C, εNd, 231Pa/230Th, δ18O and Δ14C) and idealized water tracers to study the potential constraints on LGM ocean circulation from multiple proxies. Our model suggests that the glacial Atlantic water mass distribution can be accurately constrained by the air-sea gas exchange signature of water masses (δ13CAS), but εNd might overestimate the North Atlantic Deep Water (NADW) percentage in the deep Atlantic probably because of the boundary source of Nd. A sensitivity experiment with an AMOC of similar geometry but much weaker strength suggests that the correct AMOC geometry is more important than the AMOC strength for simulating the observed glacial δ13C and εNd distributions. The kinematic tracer 231Pa/230Th is sensitive to AMOC intensity, but the interpretation might be complicated by the AMOC geometry and AABW transport changes during the LGM. δ18O in the benthic foraminifera (δ18Oc) from the Florida Straits provides a consistent measure of the upper ocean boundary current in the model, which potentially provides an unambiguous method to reconstruct glacial AMOC intensity. Finally, we propose that the moderate difference between AMOC intensity at LGM and PD, if any, is caused by the competition of the responses to CO2 forcing and continental ice sheet forcing.