Geoscience Reference
In-Depth Information
NES Model Temperature
0.4
0
0.2
-
100
0
-
0.2
-
200
-
0.4
0.6
-
-
300
-
0.8
-
1
-
400
-
1.2
-
1.4
-
500
1.6
-
242
244
246
248
250
Day of 2005
Fig. 8.23
NES model temperature contours in the last half of the simulation. The model becomes
thermobarically unstable on day 245
mixes over a wide depth range. This has large impact on heat flux at the ice/water
interface (Fig. 8.24), resulting in ice melt. During the initial breakthrough into
the WDW layer, the dynamic boundary layer deepens rapidly (again shown as the
white dashedcurvein Fig. 8.23),exceptfora shorttime earlyon 246whensurface
stress falls almost to zero, but then as the melting begins on day 246, the “thermal
barrier”effectkicks in and limits the dynamicboundarylayer to the upper75m or
so. Despite the attenuation of surface stress by positive buoyancy at the interface,
mixing continues at depth because of instabilities triggered by downward mixing
of cold water from the upper layer, as demonstrated by Fig. 8.25. On day 246, the
combinationofmeltingandlowsurfacestressbeginsformationofanew,shallower
upper layer, yet a new, uniform property layer begins to form between about 100
and 200m, and proceeds to grow both downward and upward. Layers like this are
not uncommon in the upper ocean observed around Maud Rise, and this simple
modelillustrates howtheymightform.
The start and end temperature/salinity states for the NES model (Fig. 8.26)
furnish one further observation. Despite a continuous loss of upper-ocean buoy-
ancy over20 days of heat loss to the atmosphereand about4cm of net ice growth,
the thermocline has risen by about 25m, and there is a reservoir of warm water
nearer the surface than when the simulation started. Much of this re-arranging of
the upper ocean properties, reaching depths more than 500m, results from nonlin-
earities in the equation of state. Obviously, a simple one-dimensional model that
