the last great cold maximum (as recorded by micropaleontology and pollen) as well as the maximum spread of the ice
sheets (marked by the position of moraines on land masses).
Radiocarbon dating has placed the maximum at around
16-20 ka
14 C (equivalent to 18–23 ka in calendar age). This
period was the first to be the subject of a global paleoclimate
study, thanks to the CLIMAP group. The isotopic maximum
in d
18 O of planktonic and benthic foraminifera, interpreted
as reflecting the cumulative effects of the cold and ice volume maxima, was used as a stratigraphic marker, and
summer and winter sea surface temperatures were determined from micropaleontological transfer functions. This
established the CLIMAP maps (CLIMAP 1981) (Fig. 21.3)
that served as boundary conditions for the first comprehensive paleoclimate modeling experiments. The CLIMAP
results have had a profound impact. For the first time, the
magnitude of the temperature change between an ice age
(LGM) and an interglacial period (modern times) could be
quantified: the average global temperature dropped by 6 °C.
However, this cooling was far from uniform: it exceeded
10 °C at high northern latitudes, while it was only a few
degrees in the intertropical region. This intense cooling was
associated with the development of large ice sheets over the
landmasses of the northern hemisphere, which, with about
50 million cubic kilometers of ice, was the most glaciated
hemisphere. In addition, analyses of Antarctic ice showed
that atmospheric CO 2 concentration was about 100 ppmv
below pre-industrial values (Petit et al. 1999).
It quickly became necessary to expand these early studies. Continental tracers in tropical regions, such as pollen
series or concentration of noble gases in aquifers (which are
dependent on the temperature of the rains feeding these
aquifers), indicated a cooling of 3–6 °C during the LGM. In
the nearby ocean, sea surface temperature alkenone reconstructions indicated a cooling of only about 2 °C, and
micropaleontological transfer functions showed little or no
change. Detailed studies were therefore conducted in later
decades to explain the observed differences. It appeared that
many cores taken from the tropical Pacific Ocean and used
for the CLIMAP reconstruction had very low sedimentation
rates, so that bioturbation caused the contrasts in fauna over
time to disappear. In addition, the fauna from warm waters
exhibited variability that did not solely respond to temperature changes, with the result that micropaleontological
transfer functions became insensitive at the temperatures
above 25 °C common in tropical regions. At the same time,
high-resolution studies started to indicate strong climatic
variability between 17 ka and 25 ka. In the North Atlantic,
for example, the LGM does not correspond to the coldest
conditions, which are instead associated with two periods
framing the LGM: Heinrich Stadial (HS) 2 at around 24–
22 ka and HS1 at around 19–17 ka.
It therefore became essential to reconsider the reconstruction of the surface ocean during the LGM. The latest
and most comprehensive synthesis was carried out in the
MARGO program (MARGO Project Members 2009), which
focused on the period 19–23 ka corresponding to the LGM
sensu stricto. This period corresponds to the maximum
expansion of ice sheets, as opposed to the coldest conditions
of Heinrich Stadials. The LGM MARGO reconstructions
(Fig. 21.3) agree relatively well with CLIMAP, but they also
revealed some important differences:
– the northern seas were ice-free during the summer;
– latitudinal and longitudinal thermal gradients were
strong; the mid-latitudes of the North Atlantic Ocean
experienced the strongest cooling (*−10 °C);
– the decrease in temperature was generally larger on the
eastern side than on the western side of the oceans; this
was particularly marked along the African margin,
especially in coastal upwelling zones of Namibia and
South Africa;
– the cooling of tropical waters was close to 2 °C, although
some localized waters of the Pacific and Indian Oceans
experienced moderated warming;
– in the Southern Ocean, a cooling of 2–6 °C marked a
northward displacement of the polar front.
Other studies have focused on the deep ocean during the
LGM. We have mentioned several of them in the description
of the various methodological techniques that have been
developed over the last forty years. Significant differences
between the LGM and the present day include:
– the downwelling of surface waters in the North Atlantic
happened in open ocean, leading to the formation of very
cold deep water that found its density equilibrium at
2000 m depth;
– the very cold and dense bottom waters formed in the
southern hemisphere spread throughout the deep ocean,
occupying a much larger volume than today;
– the boundary between deep and bottom waters was
characterized by a much stronger gradient of physical (T,
S, density) and geochemical (d
18 O, d
13 C) properties than
today;
– the ventilation and renewal rates of deep waters are still
poorly constrained because of conflicting information
from different tracers with a complex geochemical
behavior (
14 C,
231 Pa/
230 Th); this uncertainty is also
reflected in the simulations from general circulation
models of the ocean and coupled ocean-atmosphere
models. Most proxies do indicate, however, lower ventilation of the deep ocean and a resulting large accumulation of carbon dioxide in the deeper waters.
244
T. Caley et al.
sheets (marked by the position of moraines on land masses).
Radiocarbon dating has placed the maximum at around
16-20 ka
14 C (equivalent to 18–23 ka in calendar age). This
period was the first to be the subject of a global paleoclimate
study, thanks to the CLIMAP group. The isotopic maximum
in d
18 O of planktonic and benthic foraminifera, interpreted
as reflecting the cumulative effects of the cold and ice volume maxima, was used as a stratigraphic marker, and
summer and winter sea surface temperatures were determined from micropaleontological transfer functions. This
established the CLIMAP maps (CLIMAP 1981) (Fig. 21.3)
that served as boundary conditions for the first comprehensive paleoclimate modeling experiments. The CLIMAP
results have had a profound impact. For the first time, the
magnitude of the temperature change between an ice age
(LGM) and an interglacial period (modern times) could be
quantified: the average global temperature dropped by 6 °C.
However, this cooling was far from uniform: it exceeded
10 °C at high northern latitudes, while it was only a few
degrees in the intertropical region. This intense cooling was
associated with the development of large ice sheets over the
landmasses of the northern hemisphere, which, with about
50 million cubic kilometers of ice, was the most glaciated
hemisphere. In addition, analyses of Antarctic ice showed
that atmospheric CO 2 concentration was about 100 ppmv
below pre-industrial values (Petit et al. 1999).
It quickly became necessary to expand these early studies. Continental tracers in tropical regions, such as pollen
series or concentration of noble gases in aquifers (which are
dependent on the temperature of the rains feeding these
aquifers), indicated a cooling of 3–6 °C during the LGM. In
the nearby ocean, sea surface temperature alkenone reconstructions indicated a cooling of only about 2 °C, and
micropaleontological transfer functions showed little or no
change. Detailed studies were therefore conducted in later
decades to explain the observed differences. It appeared that
many cores taken from the tropical Pacific Ocean and used
for the CLIMAP reconstruction had very low sedimentation
rates, so that bioturbation caused the contrasts in fauna over
time to disappear. In addition, the fauna from warm waters
exhibited variability that did not solely respond to temperature changes, with the result that micropaleontological
transfer functions became insensitive at the temperatures
above 25 °C common in tropical regions. At the same time,
high-resolution studies started to indicate strong climatic
variability between 17 ka and 25 ka. In the North Atlantic,
for example, the LGM does not correspond to the coldest
conditions, which are instead associated with two periods
framing the LGM: Heinrich Stadial (HS) 2 at around 24–
22 ka and HS1 at around 19–17 ka.
It therefore became essential to reconsider the reconstruction of the surface ocean during the LGM. The latest
and most comprehensive synthesis was carried out in the
MARGO program (MARGO Project Members 2009), which
focused on the period 19–23 ka corresponding to the LGM
sensu stricto. This period corresponds to the maximum
expansion of ice sheets, as opposed to the coldest conditions
of Heinrich Stadials. The LGM MARGO reconstructions
(Fig. 21.3) agree relatively well with CLIMAP, but they also
revealed some important differences:
– the northern seas were ice-free during the summer;
– latitudinal and longitudinal thermal gradients were
strong; the mid-latitudes of the North Atlantic Ocean
experienced the strongest cooling (*−10 °C);
– the decrease in temperature was generally larger on the
eastern side than on the western side of the oceans; this
was particularly marked along the African margin,
especially in coastal upwelling zones of Namibia and
South Africa;
– the cooling of tropical waters was close to 2 °C, although
some localized waters of the Pacific and Indian Oceans
experienced moderated warming;
– in the Southern Ocean, a cooling of 2–6 °C marked a
northward displacement of the polar front.
Other studies have focused on the deep ocean during the
LGM. We have mentioned several of them in the description
of the various methodological techniques that have been
developed over the last forty years. Significant differences
between the LGM and the present day include:
– the downwelling of surface waters in the North Atlantic
happened in open ocean, leading to the formation of very
cold deep water that found its density equilibrium at
2000 m depth;
– the very cold and dense bottom waters formed in the
southern hemisphere spread throughout the deep ocean,
occupying a much larger volume than today;
– the boundary between deep and bottom waters was
characterized by a much stronger gradient of physical (T,
S, density) and geochemical (d
18 O, d
13 C) properties than
today;
– the ventilation and renewal rates of deep waters are still
poorly constrained because of conflicting information
from different tracers with a complex geochemical
behavior (
14 C,
231 Pa/
230 Th); this uncertainty is also
reflected in the simulations from general circulation
models of the ocean and coupled ocean-atmosphere
models. Most proxies do indicate, however, lower ventilation of the deep ocean and a resulting large accumulation of carbon dioxide in the deeper waters.
244
T. Caley et al.