74
The seasonal cycle of the forcings is retained in the third simulation - hereafter III. Of
course, it is this numerical experiment that is likely to perfonn best.
All model results presented below are yearly means, evaluated when the model has reached a
well established regime, i.e., when the transients associated with the initial conditions have
become negligible.
Table 5. Flow (in Sverdrups) through the Drake passage (eastward) and the Indonesian passage
(from Pacific to Indian); North Atlantic overturning (in Sverdrup) and North Atlantic Deep
Water (NADW) flow toward the Southern Ocean at 20 0 S (in Sverdrups). The observed
estimates are due to Nowlin and Klinck (1986) for the Drake passage, Godfrey (1989) for the
Indonesian passage, Gordon (1986) and Hall and Bryden (1982) for the North Atlantic
overturning, Gordon (1986) and Broecker (1991) for the southward NADW flow at 20 0 S. See
also Figs. 13-17.
Drake passage:
Indonesian passage:
N. Atlantic overturning:
NADW southward (20 0 S) :
!O
" 9C
'.0
300
590
2300
512.5
-80
observed
120-150
13.
0(12)
8.4
15- 20
meaningless
13-20
meaningless
-40
o
40
80
Latilude
II
III
134.
168.
15.1
15.6
12.9
17.6
6.3
9.0
Figure 15. Meridional streamfunction, P, in the Atlantic and Arctic for simulation II (all
forcings without seasonal cycle). The contour labels are in Sverdrups. The arrows indicate the
direction of the circulation.