Collaborative Research: Understanding Changes in the Atlantic Meridional Overturning Circulation (AMOC) During the 20th Century Using IPCC AR5 Model Ensembles

This proposal is in response to the solicitation: NOAA Climate Program Office’s Earth System Science (ESS) Program − Atlantic Meridional Overturning Circulation (AMOC) Mechanisms and Decadal Predictability. The AMOC strength is expected to decrease in response to global warming because warming-induced changes in surface heat and freshwater fluxes in high latitudes can lead to a buoyancy increase at the ocean surface, reducing deep convection in subpolar seas. Indeed, an analysis of 20th century simulations by coupled climate models used in the upcoming IPCC fifth assessment report (AR5) shows a decreasing trend in the AMOC strength over the past century. However, an ensemble of ocean/sea-ice model simulations forced by the observationally based wind stresses and surface buoyancy fluxes under a common framework of Coordinated Ocean-Ice Reference Experiments (COREs) show an increasing trend in AMOC strength. The upward trend in AMOC strength is also found in nearly all the existing ocean reanalysis products, including ECCO, SODA, … etc. The inconsistency between the fully coupled climate model and ocean/sea-ice model simulations over the 20th century raises a series of important questions concerning long-term AMOC changes: What are the dominant ocean/atmosphere processes controlling the long-term AMOC change over the 20th century? How well are these processes represented in IPCC models? Can the different AMOC trends simulated by the coupled climate models and ocean/sea-ice models be attributed to systematic model biases? If so, how do these model biases affect our ability to predict and project future AMOC changes under global warming?

To address these important issues, we propose to conduct a comprehensive inter-model comparison analysis of IPCC AR5 coupled climate model simulations and ocean/sea-ice model simulations of the 20th century climate. The analysis will focus on ocean/atmosphere processes governing long-term AMOC changes in the coupled and uncoupled model simulations, including atmosphere/ocean processes at a wide range of time scales, ranging from synoptic scale atmospheric variability associated with the North Atlantic storm track to long-term changes in ocean circulations, sea-ice conditions and upper ocean density field in response to global warming. To complement the data-analysis, we also propose to carry out two sets of ocean/seaice model simulations to assess the sensitivity of long-term AMOC changes to a variety of alternative atmospheric state choices. The first set of experiments is designed to examine the effect of observed atmospheric variability on the AMOC and will include the usage of modified forcing derived from the 20th century reanalysis by [Compo et al., 2011]. The second set of ocean/sea-ice model experiments is designed to examine how atmospheric variability in coupled climate models, in contrast to that in reality, affect the AMOC, and will include the usage of surface forcing and its deviates derived from coupled climate model simulations.

This research is highly relevant to the objectives of the NOAA’s Next Generation Strategic Plan (NGSP) and this year’s ESS program. In particular, it directly addresses the first objective of the NGSP − an improved scientific understanding of the changing climate system and its impacts, through understanding and modeling fundamental climate processes. Improving our understanding of AMOC variability and predictability is vital to climate prediction and projection at decadal and longer time scales, and it forms the primary objective of this year’s ESS program. The proposed research focuses on the understanding of the mechanisms governing long-term AMOC variability and aims at improving our ability to predict and project future AMOC changes under global warming. As such, it has a direct bearing on the goals of the ESS program this year.