slope to at least the southern coast of Africa
(Gordon, 1986; Beal et al., 2000), whereas offshore of the Agulhas axis, within the anticyclonic
zone, the intermediate depths are dominated by
lower-salinity AAIW coming from the subtropical
Indian Ocean. Beal et al. (2000) conclude that all
of the Red Sea overflow into the Indian Ocean is
eventually exported by the western boundary current. As Red Sea water is not observed within the
Agulhas Return Current, either its signature is
removed by mixing or it fully contributes to the
Agulhas leakage. Possibly the inshore component of
the Agulhas preferentially contributes to the interocean exchange, and to the Atlantic’s high salinity.
4.7.5 Discussion
Interocean exchanges have been subject to much
attention in recent years (Schmitz, 1995, 1996a,b).
The ACC, Bering Strait, Indonesian Seas and Agulhas Retroflection (and perhaps the westward flow
immediately south of Australia) offer pathways for
interocean exchange. Interocean fluxes of mass,
heat and fresh water are expected to vary across
the full range of temporal scales in concert with
climate variability. As they vary, heat and freshwater budgets of neighbouring oceans change; temperature and salinity anomalies from long-term
means can develop. These anomalies may play a
role in the climate phenomena (oscillations) associated with each ocean basin. Peterson and White
(1998) find that heat and freshwater anomalies
forming in the western subtropical South Pacific
Ocean spread eastward within the ACC entering
the subtropical gyres of the Southern Hemisphere.
The migration of these anomalies is linked to the
3- to 5-year Antarctic Circumpolar Wave (White
and Peterson, 1996).
Transport of Pacific waters entering the Indian
Ocean via the Indonesian Seas displays strong
dependence of the phase of ENSO. During El Niño
the Indonesian Throughflow is greatly reduced,
delivering less heat and fresh water into the Indian
Ocean. Might these variations introduce heat and
freshwater anomalies into the Indian Ocean thermocline, linking ENSO and monsoon climate phenomena? The transfer of subtropical Indian Ocean
water into the South Atlantic also displays variability. The transport of the Benguela Current may
be modulated by Agulhas leakage (Garzoli et al.,
1996). The pathway of Agulhas eddies across the
South Atlantic varies interannually, along with pulsations in the form of the subtropical gyre (Witter
and Gordon, 1999), influencing the distribution of
heat and fresh water within the South Atlantic, and
perhaps altering the access of the Agulhas leakage
to the pathways along the Brazilian coast leading
into the northern hemisphere (Stramma and Schott,
1996; Schott et al., 1998).
Roach et al. (1995) find that there can be interannual variations of up to 1 ppt in salinity of the
Bering Strait throughflow. Niebauer (1998) discusses the variability of wind over the Bering Sea
to ENSO and how this relationship changed in the
‘regime shift’ of the later 1970s. Changing wind
fields and sea ice distribution in the Bering Strait
region may have a downstream effect on the Arctic
freshwater budget and the flux of fresh water
through Fram Strait, affecting NADW formation
within the Greenland, Norwegian and Labrador
Seas. As the Bering Strait did not exist during the
lowered sea level of the glacial epochs, a vital
freshwater source for the Arctic was severed.
While other parts of the Arctic hydrological cycle
would be greatly altered by the glacial condition, reduced inflow of Pacific fresh water into the
Arctic pycnocline may be expected to reduce its
stability and allow greater vertical heat flux from
the warmer deep water to the sea ice cover.
The Atlantic Ocean receives low-salinity water
from the Arctic and saline water from the Indian
Ocean in association with Agulhas leakage (Gordon
et al., 1992; De Ruijter et al., 1999). The former
acts to attenuate convection in the northern North
Atlantic (Zaucker et al., 1994), while the latter has
the opposite effect. On entering the South Atlantic
a density anomaly of Indian Ocean water is only
small, due to the counteracting effects of the heat
and salt anomalies on the density. However, the
heat is quickly lost to the atmosphere while the salt
remains in the water column (Gordon et al., 1992).
According to modelling studies the effect of the
warm and salty Indian Ocean source is to strengthen
and stabilize the northern meridional overturning
of the Atlantic (Weijer et al., 1999; Weijer, 2000;
Weijer et al., 2001a,b), while the effect of the
Bering Strait (Arctic source) freshwater flux is to
weaken the northern overturning. In the presentday climate state the Indian Ocean source effect
dominates over the northern freshwater fluxes.
Shutting down the source of Indian Ocean water
brings the ocean circulation close to a state where
4.7 Interocean Exchange
313
Gordon
(Gordon, 1986; Beal et al., 2000), whereas offshore of the Agulhas axis, within the anticyclonic
zone, the intermediate depths are dominated by
lower-salinity AAIW coming from the subtropical
Indian Ocean. Beal et al. (2000) conclude that all
of the Red Sea overflow into the Indian Ocean is
eventually exported by the western boundary current. As Red Sea water is not observed within the
Agulhas Return Current, either its signature is
removed by mixing or it fully contributes to the
Agulhas leakage. Possibly the inshore component of
the Agulhas preferentially contributes to the interocean exchange, and to the Atlantic’s high salinity.
4.7.5 Discussion
Interocean exchanges have been subject to much
attention in recent years (Schmitz, 1995, 1996a,b).
The ACC, Bering Strait, Indonesian Seas and Agulhas Retroflection (and perhaps the westward flow
immediately south of Australia) offer pathways for
interocean exchange. Interocean fluxes of mass,
heat and fresh water are expected to vary across
the full range of temporal scales in concert with
climate variability. As they vary, heat and freshwater budgets of neighbouring oceans change; temperature and salinity anomalies from long-term
means can develop. These anomalies may play a
role in the climate phenomena (oscillations) associated with each ocean basin. Peterson and White
(1998) find that heat and freshwater anomalies
forming in the western subtropical South Pacific
Ocean spread eastward within the ACC entering
the subtropical gyres of the Southern Hemisphere.
The migration of these anomalies is linked to the
3- to 5-year Antarctic Circumpolar Wave (White
and Peterson, 1996).
Transport of Pacific waters entering the Indian
Ocean via the Indonesian Seas displays strong
dependence of the phase of ENSO. During El Niño
the Indonesian Throughflow is greatly reduced,
delivering less heat and fresh water into the Indian
Ocean. Might these variations introduce heat and
freshwater anomalies into the Indian Ocean thermocline, linking ENSO and monsoon climate phenomena? The transfer of subtropical Indian Ocean
water into the South Atlantic also displays variability. The transport of the Benguela Current may
be modulated by Agulhas leakage (Garzoli et al.,
1996). The pathway of Agulhas eddies across the
South Atlantic varies interannually, along with pulsations in the form of the subtropical gyre (Witter
and Gordon, 1999), influencing the distribution of
heat and fresh water within the South Atlantic, and
perhaps altering the access of the Agulhas leakage
to the pathways along the Brazilian coast leading
into the northern hemisphere (Stramma and Schott,
1996; Schott et al., 1998).
Roach et al. (1995) find that there can be interannual variations of up to 1 ppt in salinity of the
Bering Strait throughflow. Niebauer (1998) discusses the variability of wind over the Bering Sea
to ENSO and how this relationship changed in the
‘regime shift’ of the later 1970s. Changing wind
fields and sea ice distribution in the Bering Strait
region may have a downstream effect on the Arctic
freshwater budget and the flux of fresh water
through Fram Strait, affecting NADW formation
within the Greenland, Norwegian and Labrador
Seas. As the Bering Strait did not exist during the
lowered sea level of the glacial epochs, a vital
freshwater source for the Arctic was severed.
While other parts of the Arctic hydrological cycle
would be greatly altered by the glacial condition, reduced inflow of Pacific fresh water into the
Arctic pycnocline may be expected to reduce its
stability and allow greater vertical heat flux from
the warmer deep water to the sea ice cover.
The Atlantic Ocean receives low-salinity water
from the Arctic and saline water from the Indian
Ocean in association with Agulhas leakage (Gordon
et al., 1992; De Ruijter et al., 1999). The former
acts to attenuate convection in the northern North
Atlantic (Zaucker et al., 1994), while the latter has
the opposite effect. On entering the South Atlantic
a density anomaly of Indian Ocean water is only
small, due to the counteracting effects of the heat
and salt anomalies on the density. However, the
heat is quickly lost to the atmosphere while the salt
remains in the water column (Gordon et al., 1992).
According to modelling studies the effect of the
warm and salty Indian Ocean source is to strengthen
and stabilize the northern meridional overturning
of the Atlantic (Weijer et al., 1999; Weijer, 2000;
Weijer et al., 2001a,b), while the effect of the
Bering Strait (Arctic source) freshwater flux is to
weaken the northern overturning. In the presentday climate state the Indian Ocean source effect
dominates over the northern freshwater fluxes.
Shutting down the source of Indian Ocean water
brings the ocean circulation close to a state where
4.7 Interocean Exchange
313
Gordon