Geoscience Reference
In-Depth Information
and Carmack,
1989
). Density is largely a function of salinity at low temperatures.
Consequently, only a moderate freshening, such as from a change in the Fram
Strait outflow, could cause freshwater capping, leading to a reduction or cessation
of deepwater formation, an idea first proposed by P. Weyl (
1968
). Because of the
differential compressibility of seawater, with cold water being more compressible
(the thermobaric effect) it is likely that within some critical range, small amount of
salinity stratification actually serves to promote convection (Aagaard and Carmack,
1989
). However, too much freshwater caps the convection.
The study of M. Holland et al. (
2001
), based on a global coupled ice-ocean-atmo-
sphere model, suggests that realistic variations in the Fram Strait ice export can lead
to decadal-scale variability in the THC. The basis of this interaction is that reduced
ice export leads to less ice melting in the North Atlantic. This destabilizes the ocean
column, causing more deepwater formation. This results in a strengthening of the
THC (that is, stronger ocean heat transport), which then reduces ice growth in the
GIN seas. This warms and freshens the surface, reducing the original anomalous
high ocean density, acting as a negative feedback on the system. The sensitivity of
the THC to freshwater export to the North Atlantic finds further support from simu-
lations of coupled ice-ocean models (e.g., Häkkinen,
1993
,
1999
).
An interesting facet of the THC debate is that the Arctic “back door” can be influ-
enced by changes within the Arctic itself. The Fram Strait outflow is sensitive to the
regional atmospheric circulation. Freshwater inputs to the Arctic Ocean from river
runoff are larger than the Bering Strait inflow. Changes in the atmospheric circula-
tion and river discharge, as well as in net precipitation over the Arctic Ocean itself,
will also ultimately be seen in the Fram Strait outflow. The Holland et al. (
2001
)
study also indicates that variability in the THC affects the atmospheric state, alter-
ing air temperature, precipitation and runoff.
7.5.4
The Great Salinity Anomaly
As also noted by Holland et al. (
2001
), model simulations of the variability and
sensitivity of the THC appear to depend strongly on model design and complexity.
Whether the THC is as sensitive to disruption as some models have indicated is
still in debate. However, there is observational evidence, the best example being the
“Great Salinity Anomaly” (GSA) of 1968-1982. During the late 1960s and early
1970s, the upper 100 m of the waters in the GIN seas underwent reductions in salin-
ity from 0.1 to 0.5 psu, with water temperature anomalies of −1 to −2 K. This was
associated with positive anomalies in sea ice extent in the Greenland Sea, peaking
during 1968-1969 and propagating with the salinity anomalies into the Labrador
Sea in 1971-1972. The salinity anomaly circulated around the Atlantic sub-Arctic
gyre, returning to the Greenland Sea in 1981-1982 (Dickson et al.,
1988
).
Observational studies (Walsh and Chapman,
1990
; Wohlleben and Weaver,
1995
) as well as model evidence (Häkkinen,
1993
) suggest that strong northerly
winds caused an increase in the sea ice transport through the Fram Strait and into
the Greenland Sea. The large freshwater flux anomaly associated with this transport
