242
In detail, the instantaneous heat budget in the central Labrador Sea is
not easy to reconcile. The air-sea heat flux and the heat storage below
(assuming a one-dimensional heat balance over short times) show the right
"events" yet disagree in magnitude [Smith and Dobson, 1984] ; yet after
averaging over a year, the anomalies in surface flux and storage were in
better agreement. Monthly mean heat flux can, according to Smith and
Dobson, exceed 300 Wjm 2 in severe winters. Their annual average heat
loss for the central Labrador Sea is about 85 W j m 2 in the severe winter of
1972, yet over the more-placid period 1964-73, averages 28 Wjm 2 . This is
far less than the values obtained using Bunker's flux formulas, which yield
98 Wjm 2 . As Clarke and Gascard [1983] emphasise, the air temperature
is much lower close to the rim of the Labrador Sea and the oceanic heat
loss is likely to be greater there.
These small annual heat flux numbers highlight the delicate balance
of the high latitude ocean: while we assume and believe some connection
between heat loss and convection, the net cooling is difficult to establish in
detail, and this leads to widely-varying estimates of the rate of production
of Labrador Sea Water.
More recent modelling results by Wallace and Lazier [1988] suggest
that cooling may not be the sole factor at work. In their experiments,
most of the vertical profiles were not prone to convection below 1000 m
under normal conditions, and deep convection to 1400 m could only be
achieved patchily or with unrealistic heat fluxes. They conclude that some
form of pre- conditioning of the watercolumn,- a cyclonic circulation for
example,- may also be required for deep convection to occur, but as already described, (Figure 22), the observed changes in air temperature and
windstress curl have tended to act in the same sense during the postwar
period.
4.4 Changes in LSW density
It is plain enough from Figure 24 that deep () - S changes occur in the core
of the Labrador Sea Water layer. Figure 25 shows equally clearly that in
most years for which we have records, the changes in () and S at BRAVO
have been mutually-compensating in density, so that annual changes in the
() - S characteristics of LSW have tended to occur along isopycnals. In
fact, Clarke and Gascard [1993] proposed that the density of Labrador Sea
Water should remain constant with time because of this effect.
In detail, the instantaneous heat budget in the central Labrador Sea is
not easy to reconcile. The air-sea heat flux and the heat storage below
(assuming a one-dimensional heat balance over short times) show the right
"events" yet disagree in magnitude [Smith and Dobson, 1984] ; yet after
averaging over a year, the anomalies in surface flux and storage were in
better agreement. Monthly mean heat flux can, according to Smith and
Dobson, exceed 300 Wjm 2 in severe winters. Their annual average heat
loss for the central Labrador Sea is about 85 W j m 2 in the severe winter of
1972, yet over the more-placid period 1964-73, averages 28 Wjm 2 . This is
far less than the values obtained using Bunker's flux formulas, which yield
98 Wjm 2 . As Clarke and Gascard [1983] emphasise, the air temperature
is much lower close to the rim of the Labrador Sea and the oceanic heat
loss is likely to be greater there.
These small annual heat flux numbers highlight the delicate balance
of the high latitude ocean: while we assume and believe some connection
between heat loss and convection, the net cooling is difficult to establish in
detail, and this leads to widely-varying estimates of the rate of production
of Labrador Sea Water.
More recent modelling results by Wallace and Lazier [1988] suggest
that cooling may not be the sole factor at work. In their experiments,
most of the vertical profiles were not prone to convection below 1000 m
under normal conditions, and deep convection to 1400 m could only be
achieved patchily or with unrealistic heat fluxes. They conclude that some
form of pre- conditioning of the watercolumn,- a cyclonic circulation for
example,- may also be required for deep convection to occur, but as already described, (Figure 22), the observed changes in air temperature and
windstress curl have tended to act in the same sense during the postwar
period.
4.4 Changes in LSW density
It is plain enough from Figure 24 that deep () - S changes occur in the core
of the Labrador Sea Water layer. Figure 25 shows equally clearly that in
most years for which we have records, the changes in () and S at BRAVO
have been mutually-compensating in density, so that annual changes in the
() - S characteristics of LSW have tended to occur along isopycnals. In
fact, Clarke and Gascard [1993] proposed that the density of Labrador Sea
Water should remain constant with time because of this effect.