217
pattern there causes upper-layer divergence through Ekman pumping and
drives a single-gyre 35 Sv cyclonic circulation, [figure 4], resulting in the
doming of isopycnal structures across the Greenland Sea, [e.g.Kiilerich,
1945]. Both the cyclonic circulation and the doming have important implications for convection. It is at the centre of this dome, where stratification
is weakest and where dense water is brought closest to the surface that
there is the greatest potential for convective instability, since we have to
remove less buoyancy there to promote overturn. In turn, the upper-layer
divergence has the effect of sharpening the fronts which surround the dome,
thus isolating the convective centre against the lateral spread of fresh water
from the East Greenland Current.
Figure 5, from Meincke, Jonsson and Swift [1992] certainly confirms a
dramatic reduction in windstress curl over the Greenland Sea between the
late 1970's and the 1980's, implying that conditions favourable to convection would have been replaced in the 1980's by weaker doming, a smaller,
leakier gyre circulation and weak convection.
Meincke and Rudels [1995] provide the first direct evidence of the collapsing dome, pointing out that the Eurasian Basin Deep Water layer [the
warmest, most saline and deepest of the Arctic Ocean Deep Waters which
pass south through Fram strait (sill depth=2600 m)] had recently been
observed to increase its depth in the Greenland Sea Basin by several hundred metres, but without changing its density. They conclude that without a convective input, the dome collapses and spreads towards the rim,
causing a compensating influx of Canadian Basin Deep Water [CBDW]
and Eurasian Basin Deep Water [EBDW] into the centre of the Basin. A
succession of hydrographic transects extending towards the centre of the
Greenland Sea from its eastern boundary in 1991-94 appear to show the
spread of a salinity maximum layer across the basin at 2000-2500 m depth
[not shown; Blindheim, pers comm.]
[Note that the increasing CFC concentrations at depth in the Greenland
Sea, reported in a recent informal note by Schlosser, Wallace, Bullister,
Boenisch and Blindheim (1994) may be just a further reflection of the collapsing dome rather than any sign of renewed convective mixing].
(ii) Aside from the change in wind stress curl, it has been supposed
since the time of Nansen [1906] that changes in deep water temperatures
might be some response to changes in winter air temperature, acting
through the cooling and sinking of surface water. Aagaard [1968] constructs
a winter cooling index to confirm the point.
pattern there causes upper-layer divergence through Ekman pumping and
drives a single-gyre 35 Sv cyclonic circulation, [figure 4], resulting in the
doming of isopycnal structures across the Greenland Sea, [e.g.Kiilerich,
1945]. Both the cyclonic circulation and the doming have important implications for convection. It is at the centre of this dome, where stratification
is weakest and where dense water is brought closest to the surface that
there is the greatest potential for convective instability, since we have to
remove less buoyancy there to promote overturn. In turn, the upper-layer
divergence has the effect of sharpening the fronts which surround the dome,
thus isolating the convective centre against the lateral spread of fresh water
from the East Greenland Current.
Figure 5, from Meincke, Jonsson and Swift [1992] certainly confirms a
dramatic reduction in windstress curl over the Greenland Sea between the
late 1970's and the 1980's, implying that conditions favourable to convection would have been replaced in the 1980's by weaker doming, a smaller,
leakier gyre circulation and weak convection.
Meincke and Rudels [1995] provide the first direct evidence of the collapsing dome, pointing out that the Eurasian Basin Deep Water layer [the
warmest, most saline and deepest of the Arctic Ocean Deep Waters which
pass south through Fram strait (sill depth=2600 m)] had recently been
observed to increase its depth in the Greenland Sea Basin by several hundred metres, but without changing its density. They conclude that without a convective input, the dome collapses and spreads towards the rim,
causing a compensating influx of Canadian Basin Deep Water [CBDW]
and Eurasian Basin Deep Water [EBDW] into the centre of the Basin. A
succession of hydrographic transects extending towards the centre of the
Greenland Sea from its eastern boundary in 1991-94 appear to show the
spread of a salinity maximum layer across the basin at 2000-2500 m depth
[not shown; Blindheim, pers comm.]
[Note that the increasing CFC concentrations at depth in the Greenland
Sea, reported in a recent informal note by Schlosser, Wallace, Bullister,
Boenisch and Blindheim (1994) may be just a further reflection of the collapsing dome rather than any sign of renewed convective mixing].
(ii) Aside from the change in wind stress curl, it has been supposed
since the time of Nansen [1906] that changes in deep water temperatures
might be some response to changes in winter air temperature, acting
through the cooling and sinking of surface water. Aagaard [1968] constructs
a winter cooling index to confirm the point.