179
with less density: in the extreme case, too light to sink.
A final emphasis about boundary conditions: there is no "correct"
boundary condition for an ocean-only model since there is no single boundary condition that can give both the correct fluxes and the correct time
change of fluxes: the only correct and consistent boundary condition is
given by a correct atmosphere consistently coupled to the ocean.
5.2 General Considerations of Variability
Most studies of ocean decadal thermohaline variability were carried out
in idealized coarse resolution OGCMs. These mechanistic explorations
mainly investigate internal ocean variability, where the variability arises
from internal processes in the ocean without direct help from a changing
atmosphere. The coarse resolution is necessitated by practical reasons:
the large amounts of time needed to consistently spin up the thermohaline
circulation, on the order of 1000 model years.
The first exploration in that direction was the seminal work of F. Bryan
(1986). Bryan showed that, in a simplified coarse resolution sector model,
a steady circulation, symmetric about the equator, with sinking at high latitudes and upwelling in the rest of the ocean, existed under both restoring
(both temperature and salinity are restored) and mixed boundary conditions (surface temperature restored and salinity flux prescribed). When
the mixed boundary condition case was perturbed with the addition of
fresh water at high southern latitudes, a halo cline capped the sinking region and a single cell pole-to-pole circulation, with sinking in the north
and rising elsewhere, obtained in only 50 years. This rapid adjustment
implied that relatively rapid variability was possible even though the generally agreed-upon time scale for the thermohaline circulation was of order
1000 years. The rapid time scales turned out to be due to the frictional
boundary current components of the thermohaline circulation.
Perhaps the simplest idea about thermohaline variability is this: a
steady circulation requires the delivery of salt by the circulation itself to
balance any freshwater input at high latitudes in the region of the deep
sinking. In particular the Gulf Stream Extension in the models must deliver enough salt to counter the freshening by surface freshwater fluxes in
the sinking regions of the NADW. Clearly if the circulation is limited in
its capacity to deliver salt, there will be some input of freshwater flux to
the surface beyond which the circulation cannot counter: the circulation
with less density: in the extreme case, too light to sink.
A final emphasis about boundary conditions: there is no "correct"
boundary condition for an ocean-only model since there is no single boundary condition that can give both the correct fluxes and the correct time
change of fluxes: the only correct and consistent boundary condition is
given by a correct atmosphere consistently coupled to the ocean.
5.2 General Considerations of Variability
Most studies of ocean decadal thermohaline variability were carried out
in idealized coarse resolution OGCMs. These mechanistic explorations
mainly investigate internal ocean variability, where the variability arises
from internal processes in the ocean without direct help from a changing
atmosphere. The coarse resolution is necessitated by practical reasons:
the large amounts of time needed to consistently spin up the thermohaline
circulation, on the order of 1000 model years.
The first exploration in that direction was the seminal work of F. Bryan
(1986). Bryan showed that, in a simplified coarse resolution sector model,
a steady circulation, symmetric about the equator, with sinking at high latitudes and upwelling in the rest of the ocean, existed under both restoring
(both temperature and salinity are restored) and mixed boundary conditions (surface temperature restored and salinity flux prescribed). When
the mixed boundary condition case was perturbed with the addition of
fresh water at high southern latitudes, a halo cline capped the sinking region and a single cell pole-to-pole circulation, with sinking in the north
and rising elsewhere, obtained in only 50 years. This rapid adjustment
implied that relatively rapid variability was possible even though the generally agreed-upon time scale for the thermohaline circulation was of order
1000 years. The rapid time scales turned out to be due to the frictional
boundary current components of the thermohaline circulation.
Perhaps the simplest idea about thermohaline variability is this: a
steady circulation requires the delivery of salt by the circulation itself to
balance any freshwater input at high latitudes in the region of the deep
sinking. In particular the Gulf Stream Extension in the models must deliver enough salt to counter the freshening by surface freshwater fluxes in
the sinking regions of the NADW. Clearly if the circulation is limited in
its capacity to deliver salt, there will be some input of freshwater flux to
the surface beyond which the circulation cannot counter: the circulation