American ice sheet. They can therefore safely be used to
reconstruct continental climate changes associated with
Heinrich events (Sánchez-Goñi et al. 2002; CombourieuNebout et al. 2002).
Dansgaard-Oeschger Events
When the cores are studied at century-scale resolution,
records of calcite d
18 O from foraminifera from the ocean
surface indicate that abrupt changes also occurred during
Dansgaard-Oeschger events. The variations recorded by
marine indicators are very similar to those obtained from
Greenland ice data. It is therefore tempting to directly link an
observed variation in one set of records to one observed in
the other and to assume that they are responses to a common
cause. While this is certainly partially true (the d
18 O of polar
ice is above all a response to temperature and the d
18 O of the
foraminiferal calcite also contains a temperature signal),
especially in the case of marine records close to the ice
sheets, it should be borne in mind that it is paleoclimate
indicators and not climate variables, such as temperature,
that are being measured. Directly correlating every small
variation in the two records is therefore hasty and inordinate.
Despite these difficulties, marine records reveal that during
Dansgaard-Oeschger events, the surface ocean and the deep
ocean undergo changes of great amplitude in the surface
temperature and/or the d
18 O of the sea water (see Rasmussen
et al. 1996; Shackleton et al. 2000). The reconstruction of
ocean surface temperatures from independent indicators
shows that the temperature signals in Greenland cores are
matched in the neighboring North Atlantic.
Associated with these changes in temperature and
hydrological conditions are signals of relatively small
amplitude in oceanic d
13 C records (Elliot 2002). Since d
13 C
is an indicator of the ventilation of ocean water bodies, it
would appear that at least for the North Atlantic and the
Arctic, transitions between stadial and interstadial periods
are not associated with large anomalies in ventilation, and
therefore with any drastic change in thermohaline circulation
in the Atlantic. This observation shows the different mechanisms operating during the Dansgaard-Oeschger events and
Heinrich events, the latter having very marked anomalies in
ocean d
13 C records.
Neighbouring Continents
The evolution of continental paleo-conditions is recorded in
numerous environments, such as in caves (e.g. in concretions
or speleothems), lakes (lake sediments), peat bogs, loess or
marine cores. For continental regions adjacent to the Ruddiman belt, these records show a clear correlation between
the evolution of the ocean and continent during periods of
strong glacial variability. Pollen data, for example, show that
during this period, rapid climate changes, whether D/O
oscillations or Heinrich events, had repercussions on
vegetation.
In Western Europe, D/O events resulted in periodic changes
in forest cover (Fig. 29.2), with warm events corresponding to
the expansion of oak forests, also implying wet conditions (e.g.
Sánchez-Goñi et al. 2002; Combourieu-Nebout et al. 2002).
On the other side of the Atlantic, in Florida, over the same
periods, the oak is associated with large quantities of herbaceous plants (Ambrosia, Poaceae), indicating drier phases
(Grimm et al. 2006). During the Heinrich events, the vegetation cover of western Europe became steppe-like, indicating a
cold and very dry climate, especially in the Mediterranean
region (Fig. 29.2). These cold conditions prevailing over
Europe are associated with high d
13 C values recorded in several stalagmites in the south of France. They show up in a
slowing down, or even a curtailment of growth as the cold
conditions prevent the infiltration necessary for the formation
of stalagmites (Genty et al. 2005).
At the same time, vegetation in Florida was characterized
by an abundance of pine trees and the regression of herbaceous plants and oak trees which is interpreted as a consequence of a warmer and more humid climate (Grimm et al.
2006). The strong contrast between the responses by climate
and vegetation to abrupt climate events on the two sides are
examined with the cases of Europe and Florida. In Florida,
this response may seem to be counter-intuitive, since the
climate becomes warmer and more humid during the Heinrich events, although they are responsible for a major cooling in the North Atlantic. However, as records obtained for
Florida have a relatively poor temporal resolution, it is
possible that this contrasting response with the European one
is due to problems of synchronization in the reconstructions.
Another possibility is that the climates of Europe and Florida
do not have the same sensitivity to freshwater incursions into
the North Atlantic and the associated cooling. Europe,
indeed, has a regime of prevailing westerly winds and is
therefore under the direct influence of the cooling of the
North Atlantic, unlike Florida on the other side of the
Atlantic. It is thus crucial to be able to explain the complexity of these different signals. Modeling, as will be shown
later in this chapter, can provide a coherent framework for
reconstructions of climate changes in regions geographically
distant from each other.
Millennial-Scale Variability in Other Regions
of the World
After the discovery of abrupt climate changes in Greenland,
the North Atlantic and adjacent regions, rapid changes in the
characteristics of the climate system were discovered in
408
M. Kageyama et al.
reconstruct continental climate changes associated with
Heinrich events (Sánchez-Goñi et al. 2002; CombourieuNebout et al. 2002).
Dansgaard-Oeschger Events
When the cores are studied at century-scale resolution,
records of calcite d
18 O from foraminifera from the ocean
surface indicate that abrupt changes also occurred during
Dansgaard-Oeschger events. The variations recorded by
marine indicators are very similar to those obtained from
Greenland ice data. It is therefore tempting to directly link an
observed variation in one set of records to one observed in
the other and to assume that they are responses to a common
cause. While this is certainly partially true (the d
18 O of polar
ice is above all a response to temperature and the d
18 O of the
foraminiferal calcite also contains a temperature signal),
especially in the case of marine records close to the ice
sheets, it should be borne in mind that it is paleoclimate
indicators and not climate variables, such as temperature,
that are being measured. Directly correlating every small
variation in the two records is therefore hasty and inordinate.
Despite these difficulties, marine records reveal that during
Dansgaard-Oeschger events, the surface ocean and the deep
ocean undergo changes of great amplitude in the surface
temperature and/or the d
18 O of the sea water (see Rasmussen
et al. 1996; Shackleton et al. 2000). The reconstruction of
ocean surface temperatures from independent indicators
shows that the temperature signals in Greenland cores are
matched in the neighboring North Atlantic.
Associated with these changes in temperature and
hydrological conditions are signals of relatively small
amplitude in oceanic d
13 C records (Elliot 2002). Since d
13 C
is an indicator of the ventilation of ocean water bodies, it
would appear that at least for the North Atlantic and the
Arctic, transitions between stadial and interstadial periods
are not associated with large anomalies in ventilation, and
therefore with any drastic change in thermohaline circulation
in the Atlantic. This observation shows the different mechanisms operating during the Dansgaard-Oeschger events and
Heinrich events, the latter having very marked anomalies in
ocean d
13 C records.
Neighbouring Continents
The evolution of continental paleo-conditions is recorded in
numerous environments, such as in caves (e.g. in concretions
or speleothems), lakes (lake sediments), peat bogs, loess or
marine cores. For continental regions adjacent to the Ruddiman belt, these records show a clear correlation between
the evolution of the ocean and continent during periods of
strong glacial variability. Pollen data, for example, show that
during this period, rapid climate changes, whether D/O
oscillations or Heinrich events, had repercussions on
vegetation.
In Western Europe, D/O events resulted in periodic changes
in forest cover (Fig. 29.2), with warm events corresponding to
the expansion of oak forests, also implying wet conditions (e.g.
Sánchez-Goñi et al. 2002; Combourieu-Nebout et al. 2002).
On the other side of the Atlantic, in Florida, over the same
periods, the oak is associated with large quantities of herbaceous plants (Ambrosia, Poaceae), indicating drier phases
(Grimm et al. 2006). During the Heinrich events, the vegetation cover of western Europe became steppe-like, indicating a
cold and very dry climate, especially in the Mediterranean
region (Fig. 29.2). These cold conditions prevailing over
Europe are associated with high d
13 C values recorded in several stalagmites in the south of France. They show up in a
slowing down, or even a curtailment of growth as the cold
conditions prevent the infiltration necessary for the formation
of stalagmites (Genty et al. 2005).
At the same time, vegetation in Florida was characterized
by an abundance of pine trees and the regression of herbaceous plants and oak trees which is interpreted as a consequence of a warmer and more humid climate (Grimm et al.
2006). The strong contrast between the responses by climate
and vegetation to abrupt climate events on the two sides are
examined with the cases of Europe and Florida. In Florida,
this response may seem to be counter-intuitive, since the
climate becomes warmer and more humid during the Heinrich events, although they are responsible for a major cooling in the North Atlantic. However, as records obtained for
Florida have a relatively poor temporal resolution, it is
possible that this contrasting response with the European one
is due to problems of synchronization in the reconstructions.
Another possibility is that the climates of Europe and Florida
do not have the same sensitivity to freshwater incursions into
the North Atlantic and the associated cooling. Europe,
indeed, has a regime of prevailing westerly winds and is
therefore under the direct influence of the cooling of the
North Atlantic, unlike Florida on the other side of the
Atlantic. It is thus crucial to be able to explain the complexity of these different signals. Modeling, as will be shown
later in this chapter, can provide a coherent framework for
reconstructions of climate changes in regions geographically
distant from each other.
Millennial-Scale Variability in Other Regions
of the World
After the discovery of abrupt climate changes in Greenland,
the North Atlantic and adjacent regions, rapid changes in the
characteristics of the climate system were discovered in
408
M. Kageyama et al.