GFDL - Geophysical Fluid Dynamics Laboratory

Sensitivity of the North Atlantic Ocean Circulation to an Abrupt Change in the Nordic Sea Overflow in a High Resolution Global
Coupled Climate Model

Zhang, R., T. L. Delworth, A. Rosati, W. G. Anderson, K. W. Dixon, H.-C. Lee, and F. Zhang

J. Geophys. Res., doi:10.1029/2011JC007240, in press. (accepted 28 September 2011)


Abstract

The sensitivity of the North Atlantic Ocean Circulation to an abrupt change in the Nordic Sea overflow is investigated for the first time using a high resolution eddy-permitting global coupled ocean-atmosphere model (GFDL CM2.5). The Nordic Sea overflow is perturbed through the change of the bathymetry in GFDL CM2.5. We analyze the Atlantic Meridional Overturning Circulation (AMOC) adjustment process and the downstream oceanic response to the perturbation.

The results suggest that north of 34?N, AMOC changes induced by changes in the Nordic Sea overflow propagate on the slow tracer advection time scale, instead of the fast Kelvin wave time scale, resulting in a time lead of several years between subpolar and subtropical AMOC changes. The results also show that a stronger and deeper-penetrating Nordic Sea overflow leads to stronger and deeper AMOC, stronger northward ocean heat transport, reduced Labrador Sea deep convection, stronger cyclonic Northern Recirculation Gyre (NRG), westward shift of the North Atlantic Current (NAC) and southward shift of the Gulf Stream, warmer sea surface temperature (SST) east of Newfoundland and colder SST south of the Grand Banks, stronger and deeper NAC and Gulf Stream, and stronger oceanic eddy activities along the NAC and the Gulf Stream paths. A stronger/weaker Nordic Sea overflow also leads to a contracted/expanded subpolar gyre (SPG).

This sensitivity study points to the important role of the Nordic Sea overflow in the large scale North Atlantic ocean circulation, and it is crucial for climate models to have a correct representation of the Nordic Sea overflow.


[CM2.5 model bathymetry - Control and Perturbed]

About the bathymetry differences in the two experiments

The bathymetry in the Control experiment (lower panel to right) is based
upon observational data sets. To study the sensitivity of the North
Atlantic Ocean circulation to an abrupt change of the Nordic Sea
overflow, the bathymetry in a region south of the Denmark Strait is
deepened by 300m whenever the background ocean depth in this area in the
Control experiment is greater than 300m (area indicated by black
shading in upper panel to right.) With this modification of the
bathymetry, the downstream pathways of the Nordic Sea overflow (in
particular the Denmark Strait overflow) in the Perturbed Bathymetry
experiment are deepened and widened, resulting in an instantaneous
strengthening and deeper penetration of the Nordic Sea overflow in the
deep ocean south of the Denmark Strait.
This paper reports on the the
AMOC adjustment process to the abrupt change in the Nordic Sea overflow
using the first 10 years of the GFDL CM2.5 Control and Perturbed
Bathymetry experiments.

About the passive dye tracer

In these GFDL CM2.5 experiments, a passive dye tracer is continuously introduced over the
depth range from 572 to 926m along a 67km line extending across the
Denmark Strait at 66.29N – an area that encompasses 3 vertical levels by
6 points along the model’s X-axis. Initialized everywhere to a value of
0, waters passing through these 18 sub-surface model grid cells in the
Denmark Strait are tagged with a value of 1. The subsequent
model-simulated transport of the passive dye tracer is handled the same
way as are the transports of potential temperature and salinity.

The
non-linear color coding used in the maps shown below can be interpreted as
depicting the percentage of water at each model grid cell at 3000m depth
that has passed through that sector of the Denmark Strait (the
aforementioned 18 grid points) within the past “N” months, where “N” is
the number of months since initialization as listed at the top center of
the images. As constructed, the maximum value of 100% is found at
the 18 grid points where water is tagged with this tracer (not seen here
because they are at shallower depths than the 3000m level shown.) Being substantially deeper and downstream of where the dye
is introduced, the maximum concentration at 3000m in these maps
approaches 30%.  The mapped dye values are monthly
means.


 

webapp/imagecrossfade

img1
Denmark Strait dye distribution base maps

In this base map and the maps to follow land is light gray. Ocean depths less than 3000m are dark gray. Ocean depths greater than 3000m are black. Greenland is located at top center, Iceland (top right), and the Mid-Altantic Ridge (center). The eastern US and Canada appear to the left, including Labrador and Cape Hatteras. The maps that follow show paths taken by passive dye introduced in the Denmark Strait.

img2
Denmark Strait dye distribution at 3000m after 12 months

After 1 year, we see that some passive dye introduced in the Denmark Strait at depths between 572 and 926m has appeared at 3000m depth around the perimeter of the Labrador Sea. In the Perturbed Bathymetry experiment (right), one sees higher dye concentrations extending father downstream – a sign of a deeper and stronger deep western boundary current.

? [larger month 12 image]

img3
Denmark Strait dye distribution at 3000m after 48 months

After 4 years, in the Perturbed Bathymetry experiment, the passive dye at 3000m shows a much higher
concentration near the western boundary, and passes the southern tip of
the Grand Banks due to the stronger deep current. However, in
the Control experiment, the passive dye does not penetrate to south of
the Grand Banks by the end of year 4.

? [larger month 48 image]

img4
Denmark Strait dye distribution at 3000m after 84 months

By year 7, in the Perturbed Bathymetry experiment, the front of the passive
dye has already reached Cape Hatteras due to the stronger deep flow. In contrast, at this depth the Control experiment’s passive dye is still
confined to the subpolar region and spreads through gaps in the Mid-Atlantic Ridge into the eastern basin more so than in the Perturbed Bathymetry experiment.

? [larger month 84 image]

 

Animations

animation showing a 3-D view of Denmark Strait passive dye

(months 1 to 120)

[1920×1080 mov; 21 MB]

[animation with 3D view of GFDL CM2.5 Denmark Str dye]

animation of the Denmark Strait dye at 3000m

(months 1 to 120)

[1280×720 mpg; 2.4MB]

[icon for 2-D animation of GFDL CM2.5 Denmark Strait dye]