253
eastern boundary. If so, the implied trans- Atlantic spreading rate would
be very much [x 3] faster than that assumed by Read and Gr'~ld [1992] or
Ellett [1993], as Sy et al point out.
(iii) As Labrador Sea Water characteristics have changed, the southward spread of these anomalous conditions into the western Atlantic has also been recognised in the hydrographic record. McCartney,
Worthington and Raymer [1980] compare meridional sections along 55W
to highlight the arrival, in July 1977, of small cells of pure LSW, 800 km
from their nearest possible source. Localised salinity anomalies of up to
-0.106 were encountered in the Slope Water, under the Gulf Stream and
beneath the centre of the recirculating gyre to its south, and were clearly
new features compared with the KNORR 60 section of the previous October. Though McCartney et al were evidently unaware that convection in
the Labrador Sea had resumed since winter 1971-2, we would now probably
assign salinities as low as these [S=34.86 at t=3.52 0C] to the vintage of
LSW formed in about 1973 during that first intense period of renewal [ef
source function, Figure 24], implying a spreading rate of 0.56 cm/s. This
is very similar to the estimated spreading rate to the eastern boundary,
described above.
More recently, Koltermann and Sy [1994] give evidence of the continued
increase in LSW production during the 1980's in showing a deepening of
the LSW east of Newfoundland from 1400 to 2100 m in July 1993, and a
more-than-doubling of LSW volume between 1982 and 1993 at 48 ON in
the West Atlantic.As we have seen, this dramatic increase of southward
spreading has coincided with the succession of exceptionally-harsh and
worsening winters in the Labrador Sea during the 1980's and 1990's, with
postwar maxima developing in windstress, northwesterly wind component
and windstress curl (Figure 22 a,b,c).
5 Summary and Conclusions
This review suggests the following conclusions:
1. The convective activity of the North Atlantic is subject to radical
interannual and interdecadal changes.
2. In the case of the Greenland Sea convection-centre, intensifying deep
convection from the 1950's to the early 1970's, when overturn reached
eastern boundary. If so, the implied trans- Atlantic spreading rate would
be very much [x 3] faster than that assumed by Read and Gr'~ld [1992] or
Ellett [1993], as Sy et al point out.
(iii) As Labrador Sea Water characteristics have changed, the southward spread of these anomalous conditions into the western Atlantic has also been recognised in the hydrographic record. McCartney,
Worthington and Raymer [1980] compare meridional sections along 55W
to highlight the arrival, in July 1977, of small cells of pure LSW, 800 km
from their nearest possible source. Localised salinity anomalies of up to
-0.106 were encountered in the Slope Water, under the Gulf Stream and
beneath the centre of the recirculating gyre to its south, and were clearly
new features compared with the KNORR 60 section of the previous October. Though McCartney et al were evidently unaware that convection in
the Labrador Sea had resumed since winter 1971-2, we would now probably
assign salinities as low as these [S=34.86 at t=3.52 0C] to the vintage of
LSW formed in about 1973 during that first intense period of renewal [ef
source function, Figure 24], implying a spreading rate of 0.56 cm/s. This
is very similar to the estimated spreading rate to the eastern boundary,
described above.
More recently, Koltermann and Sy [1994] give evidence of the continued
increase in LSW production during the 1980's in showing a deepening of
the LSW east of Newfoundland from 1400 to 2100 m in July 1993, and a
more-than-doubling of LSW volume between 1982 and 1993 at 48 ON in
the West Atlantic.As we have seen, this dramatic increase of southward
spreading has coincided with the succession of exceptionally-harsh and
worsening winters in the Labrador Sea during the 1980's and 1990's, with
postwar maxima developing in windstress, northwesterly wind component
and windstress curl (Figure 22 a,b,c).
5 Summary and Conclusions
This review suggests the following conclusions:
1. The convective activity of the North Atlantic is subject to radical
interannual and interdecadal changes.
2. In the case of the Greenland Sea convection-centre, intensifying deep
convection from the 1950's to the early 1970's, when overturn reached