Impact of Ocean Eddies on the Response of the Global Coupled Climate System to Strong Greenhouse Gas Forcing

Dian Putrasahan1, Jin-Song von Storch1, Katja Lohmann2, Johann Jungclaus2, Helmuth Haak2 and Oliver Gutjahr2, (1)Max Planck Institute for Meteorology, Climate Variability, Hamburg, Germany, (2)Max Planck Institute for Meteorology, Hamburg, Germany
Abstract:
The efficiency of the ocean in uptaking heat and carbon can notably affect the transient and equilibrium climate response, and these uptakes can be strongly affected by oceanic mesoscale eddies. However, current generation climate models generally rely on eddy parametrizations as they do not resolve ocean mesoscale eddies. The effect of parameterized eddies can deviate from that of resolved eddies and may therefore induce differences in the climate response projected by a climate model with eddy-resolving vs non-eddy-resolving ocean. Thus far, this possibility has not been systematically studied. Here, we quantify whether and to what extent climate response is affected by oceanic mesoscale eddies using a set of abrupt 4xCO2-runs carried out with the Max Planck Institute - Earth Systems Model (MPI-ESM) with T127 atmospheric resolution (~1o) and varying ocean resolution (1o, 0.4o and 0.1o). We aim at identifying the effect of mesoscale eddies on the fast and slow climate responses measured by global mean surface temperature (GMST) and global ocean heat budget in each depth layer. We find that in the first 3-4 decades, the rise in GMST in eddy-resolved runs is less than that seen in eddy-parameterized runs. However, in later decades, they are indistinguishable. 4xCO2 leads to a weaker vertical eddy flux convergence in the upper 50m, but the strongest weakening occurs in the eddy-resolving run, which may help to explain the weaker rise in GMST for eddy-resolving runs. The heat gain in 4xCO2 simulations generally peaks around 700-1200m, and it is between 500-1500m depths that the rate of heat gain increases with time, coinciding with an increase in total vertical flux convergence and indicating a build up and storage of heat at those depths. However, differences in resolution show that build up is larger between 500-1000m and smaller between 1000-1500m in eddy-resolving runs compared to eddy-parameterized runs, and coincident with differences in total vertical flux convergence. Between 500-1000m depths, this difference in total vertical flux convergence is namely dominated by the eddy component, while between 1000-1500m depths, both mean and eddy components provide similar contributions.