232
Mackensen and Bickert
from the Atlantic, Indic, and Pacific Ocean deeper
than 2000 m and wheighed by the volume of each
ocean, Duplessy et al. (1988) try to consider an
additional variability ofo 13 C in intermediate water
masses. Therefore the value given by Curry et al.
(1988) seems to be a maximum estimation for this
global shift. As one possible reason for such a shift
Shackleton (1977) proposes that rebuilding ofthe
continental biomass during deglaciation retrieves
carbon from the atmosphere and raises therefore
the 013C in the ocean. Broecker (1982) favours the
deposition of organic material on the shelves
during the rising sea level which leads to a higher
013C. Also, a redistribution of nutrients between the
intermediate and the deep ocean, as modeled by
Boyle (1988), could have lead to a shift in carbon
isotopes of the average deep ocean compensated
by an equal and opposite shift in the upper ocean.
The debate, however, about origin and amplitude
of the global shift is still open (Raymo et al. 1997).
New Results from the Southern Ocean
"Thermodynamic" Effect
Based on a new data set, which extended the
GEOSECS 013l:C02 measurements of the SouthAtlantic and Southern Ocean sections into the
Weddell Sea, Mackensen et al. (1996) suggested
a method to differentiate between two specific
modes of bottom water formation. On two sections,
between the tip of the Antarctic Peninsula and
Kapp Norvegia at the eastern Weddell Sea continental margin and along the Greenwich meridian
between 45°S and Kapp Norvegia, respectively,
full-depth vertical profiles of temperature and conductivity were measured with a conductivitytemperature-depth (CTD) probe. Water samples
were taken at up to 24 depths at each station for
salinity, dissolved oxygen, nutrients, and stable isotope determinations (Fig. 1).
Mackensen et al. (1996) used the 1.1 "Redfield"
slope to calculate the excess O!3C values and to
quantify the decoupling between P0 4 and
013C LC02 ' This slope is calculated for data from
ocean basins where only one water mass contributes to deep water, assuring the isolation ofbiological processes only. In addition, the 1.1 slope coincides with the latest theoretically predicted
"Redfield" slope. The calculation ofAo13C Lco2 depicts significantly high values between 3000 and
5000 m water depth in the north-western abyssal
Weddell Sea (Fig. 2).
Moreover, the distribution pattern in the two
sections clearly indicate that these high deep and
bottom water values are derived from the surface
at sites of bottom water formation, i.e. advected
from the south-western continental slope off the
Filchner-Ronne Ice Shelf and, in addition, directly
cascading down the Antarctic Peninsula slope off
Larsen Ice Shelf.
These data suggest a thermodynamic imprint
(A013CLC02) of >0.6%0 in the surface water of the
central Weddell Sea. If an anthropogenic lowering
of the ocean-atmosphere Ol3C because of fossil
fuel combustion is considered, a higher imprint, by
up to 0.4 %0, would be expected in Weddell Sea surface water during pre-industrial times (MaierReimer, pers. communication, 1994). In the
subpolar surface water masses driven by the Antarctic Circumpolar Current, excess Ol3C values
of >0.8 %0 were calculated. These values would
match the 1.3 %0 positive deviation from the
Redfield line in the Antarctic ocean as simulated
by the Hamburg Ocean GCM (Broecker and
Maier-Reimer 1992). A doubling of the global
air-sea exchange rate of CO 2 in the Hamburg
model resulted in an increase of Ol3C values for
cold surface waters and a decrease of those for
warm surface waters. Because the maximum of
sub-Antarctic A0l3CLC02 today lies in the Polar
Frontal Zone, a region of high wind speeds, this
model result corroborates the view that enhanced
gas exchange rates caused by larger than average
wind speeds and cold surface water lead to high
Aol3C. This process may be important also during
bottom water formation when strong catabatic
winds consistently keep the surface water free of
ice and simultaneously water cools down in contact with the atmosphere. Consequently, polynyas
are the prime candidate for producing surface
water with high A0 13 C values. After becoming
dense enough by the continuous process of repeated freezing, brine release, and subsequent
Mackensen and Bickert
from the Atlantic, Indic, and Pacific Ocean deeper
than 2000 m and wheighed by the volume of each
ocean, Duplessy et al. (1988) try to consider an
additional variability ofo 13 C in intermediate water
masses. Therefore the value given by Curry et al.
(1988) seems to be a maximum estimation for this
global shift. As one possible reason for such a shift
Shackleton (1977) proposes that rebuilding ofthe
continental biomass during deglaciation retrieves
carbon from the atmosphere and raises therefore
the 013C in the ocean. Broecker (1982) favours the
deposition of organic material on the shelves
during the rising sea level which leads to a higher
013C. Also, a redistribution of nutrients between the
intermediate and the deep ocean, as modeled by
Boyle (1988), could have lead to a shift in carbon
isotopes of the average deep ocean compensated
by an equal and opposite shift in the upper ocean.
The debate, however, about origin and amplitude
of the global shift is still open (Raymo et al. 1997).
New Results from the Southern Ocean
"Thermodynamic" Effect
Based on a new data set, which extended the
GEOSECS 013l:C02 measurements of the SouthAtlantic and Southern Ocean sections into the
Weddell Sea, Mackensen et al. (1996) suggested
a method to differentiate between two specific
modes of bottom water formation. On two sections,
between the tip of the Antarctic Peninsula and
Kapp Norvegia at the eastern Weddell Sea continental margin and along the Greenwich meridian
between 45°S and Kapp Norvegia, respectively,
full-depth vertical profiles of temperature and conductivity were measured with a conductivitytemperature-depth (CTD) probe. Water samples
were taken at up to 24 depths at each station for
salinity, dissolved oxygen, nutrients, and stable isotope determinations (Fig. 1).
Mackensen et al. (1996) used the 1.1 "Redfield"
slope to calculate the excess O!3C values and to
quantify the decoupling between P0 4 and
013C LC02 ' This slope is calculated for data from
ocean basins where only one water mass contributes to deep water, assuring the isolation ofbiological processes only. In addition, the 1.1 slope coincides with the latest theoretically predicted
"Redfield" slope. The calculation ofAo13C Lco2 depicts significantly high values between 3000 and
5000 m water depth in the north-western abyssal
Weddell Sea (Fig. 2).
Moreover, the distribution pattern in the two
sections clearly indicate that these high deep and
bottom water values are derived from the surface
at sites of bottom water formation, i.e. advected
from the south-western continental slope off the
Filchner-Ronne Ice Shelf and, in addition, directly
cascading down the Antarctic Peninsula slope off
Larsen Ice Shelf.
These data suggest a thermodynamic imprint
(A013CLC02) of >0.6%0 in the surface water of the
central Weddell Sea. If an anthropogenic lowering
of the ocean-atmosphere Ol3C because of fossil
fuel combustion is considered, a higher imprint, by
up to 0.4 %0, would be expected in Weddell Sea surface water during pre-industrial times (MaierReimer, pers. communication, 1994). In the
subpolar surface water masses driven by the Antarctic Circumpolar Current, excess Ol3C values
of >0.8 %0 were calculated. These values would
match the 1.3 %0 positive deviation from the
Redfield line in the Antarctic ocean as simulated
by the Hamburg Ocean GCM (Broecker and
Maier-Reimer 1992). A doubling of the global
air-sea exchange rate of CO 2 in the Hamburg
model resulted in an increase of Ol3C values for
cold surface waters and a decrease of those for
warm surface waters. Because the maximum of
sub-Antarctic A0l3CLC02 today lies in the Polar
Frontal Zone, a region of high wind speeds, this
model result corroborates the view that enhanced
gas exchange rates caused by larger than average
wind speeds and cold surface water lead to high
Aol3C. This process may be important also during
bottom water formation when strong catabatic
winds consistently keep the surface water free of
ice and simultaneously water cools down in contact with the atmosphere. Consequently, polynyas
are the prime candidate for producing surface
water with high A0 13 C values. After becoming
dense enough by the continuous process of repeated freezing, brine release, and subsequent