Skip to content

Topic: Simulating the AMOC in a warming world

Info associated with a Journal of Climate paper (Delworth et al., 2011, in press) and a poster presented at the October 2011 WCRP Open Science Conference (Dixon et al, 2011).

[thumbnail: AMOC vs time in GFDL CM2.5 and CM2.1 experiments]

Though CM2.5’s ocean is more energetic overall, the Atlantic Meridional Overturning Circulation (AMOC) is more vigorous and variable in the coarser resolution CM2.1 model. The time mean AMOC strength in the CM2.5 Control Experiment is 14.4 Sverdrups (Sv; 1 Sv = 106 m3 s­-1) versus 18.1 Sv in CM2.1.   However, in both CM2.5 and CM2.1 Control Experiments, the poleward heat flux in the Atlantic basin peaks at about 1.0×1015 Watts (1PW) – a value less than the ~1.3PW of recent observational estimates (Johns et al., 2011).

We suspect that insufficiently resolved overflows of dense waters through the Denmark Strait and Faroe Bank Channel may contribute to the somewhat weaker than observed North Atlantic heat transport.  Additional studies are exploring the sensitivity of CM2.5’s Atlantic circulation (time mean and internal variability) to Denmark Strait bathymetry (Zhang et al., 2011) and Labrador Sea stratification.

As shown below, in idealized +1% yr-1 CO2 experiments, the AMOC weakens more in CM2.1 (-25%) than in CM2.5 (-15%) (fig. to left). The resuction in the poleward heat transport is also less in CM2.5 than in CM2.1. Accordingly, surface temperatures in the subpolar North Atlantic warm more quickly in CM2.5, as documented in Delworth et al., (2011). This occurs because, as atmospheric CO2 increases, the sizable reduction in CM2.1’s poleward heat transport (a cooling effect) offsets a large part of the greenhouse gas-induced warming in the subpolar North Atlantic.

[Graph of AMOC strength vs time in GFDL CM2.1 and CM2.5 experiments]

caption: thin lines = annual AMOC values, thick lines = 10 year means. “Control” experiments (solid) are run with 1990 forcing conditions. Starting at year 101, CO2 is increased 1% yr-1 compounded to doubling (70 years later) and then held constant in the “+1%/yr CO2” warming experiments (dashed).

  • Delworth, T.L., et al., 2011: Simulated climate and climate change in the GFDL CM2.5 high resolution coupled climate model. Journal of Climate, (in press).
  • Johns, W.E., et al., 2011: Continuous, array-based 1497 estimates of Atlantic Ocean heat transport at 26.5°N, Journal of Climate, (in press).
  • Zhang, R., et al., 2011: Sensitivity of the North Atlantic Ocean Circulation to an Abrupt Change in the Nordic Sea Overflow in a High Resolution Global Coupled Climate Model. JGR – Oceans (in press). [ LINK ]