however initiated much research on these major deposits of
coarse detrital material, which were named Heinrich events
(HE) by Broecker et al. in 1992. Heinrich events are defined
by these authors in terms of both their characteristic low
concentration of planktonic foraminifera and high concentration of detrital material. They further demonstrate that
these events are related to armadas of icebergs that broke off
from the ice sheets rather than to the melting of these ice
sheets. The layers of detrital material may be several meters
thick in the Labrador Sea. This thickness is less, although
still significant, close to the European coasts. Subsequent
studies of the composition and properties of this
ice-rafted detrital material show that its origin is mostly from
the Canadian shield and has therefore been transported by
icebergs which detached from the North American ice sheet.
During the same period, the analysis of the d
18 O isotopic
signal from ice cores taken from the summit of the Greenland ice cap (Dansgaard et al. 1993) shows the contrast
between interglacial periods, and particularly between the
remarkably stable Holocene, and the last glacial period,
characterized by high amplitude oscillations. This amplitude
can be as much as half of the glacial-interglacial difference in
Greenland.
The
warming
events
were
named
‘Dansgaard-Oeschger events’ (D/O), because these two
authors had already detected rapid d
18 O fluctuations in the
ice in a core from Camp Century. At the time, the record
appeared dubious due to large simultaneous variations in the
CO 2 content in the air bubbles trapped in the ice. We now
know that the latter are artefacts caused by the presence of
carbonate dust attacked by the sulfuric acid present in the
ice. However, the rapid fluctuations of d
18 O are indeed
significant.
D/O warming appears to have occurred particularly
quickly, over a few decades, solely when ice sheets developed on the continents of the northern hemisphere. The
relatively warm period following the D/O event is called the
‘interstadial period’ and is characterized by gradual cooling.
It ends with a rapid return to the coldest levels recorded,
known as ‘stadial periods’. This return completes a D/O
cycle, which lasts a total of about 1500 years. There are
therefore more D/O events than Heinrich events.
The existence of abrupt variations in both the surface
climate in Greenland and in the ocean conditions over time
scales far shorter than the Milankovich cycles contributed to
the emerging idea that the climate system could have multiple equilibrium points and that it could be abruptly reorganized as it transfers from one equilibrium to another. This
idea was reinforced by the correlations between marine and
glacial records, first demonstrated by Bond et al. (1993).
These authors suggest a perspective of glacial variability that
integrates the Heinrich and Dansgaard-Oeschger events. The
glacial millennial variability is organized in cycles, later
called “Bond cycles” (Fig. 29.1). Each cycle begins at the
end of a Heinrich event, by the first Dansgaard-Oeschger
event, which has a large amplitude. This is followed by a
few other Dansgaard-Oeschger cycles, of diminishing
amplitude. The last cycle ends with a massive discharge of
icebergs from the Laurentide sheet, in other words, by a
Heinrich event. Each cycle lasts approximately 7000–
10,000 years.
Since these discoveries in the early 1990s, millennial
variability has become a subject of intense research.
Numerous records were analyzed with as fine a resolution as
possible and they showed that variability at shorter time
scales than those of Milankovitch was not limited to the
North Atlantic and adjacent regions. The challenge is then to
be able to synchronize these different records to better
characterize this type of variability, as well as to better
understand the connections between climate signals recorded
all over the globe. In the following sections, we focus on the
signature of the Heinrich and D/O events, on their impact on
climate and on the efforts to understand these climate
instabilities through modeling.
The Regional Impacts of the Heinrich
and Dansgaard-Oeschger Events: North Atlantic
and Adjacent Regions
Atlantic Ocean
The Heinrich Events
The melting of massive armadas of icebergs profoundly
modified the ocean surface conditions in the North Atlantic.
The changes are recorded in ocean sediment cores, particularly in d
18 O signals from planktonic foraminifera. Indeed,
excursions towards lighter d
18 O are measured in the calcite
of these foraminifera, indicating either less saline surface
waters or warmer temperatures. The fauna of fossil foraminifera in sediments reveal very cold conditions during
Heinrich events (see, for example, Hemming 2004; Cortijo
et al. 2005). The melting of icebergs is responsible for a
huge inflow of freshwater, desalinating the surface of the
ocean and introducing a highly negative d
18 O signal, typical
of ice contained in the ice sheets.
Other paleoceanographic records reveal significant reorganizations during the Heinrich events. In particular, the
study of d
13 C in sediment cores proves that the bottom
waters of the Atlantic Ocean were less well ventilated (Elliot
2002), demonstrating a reorganization of ocean circulation.
These studies were corroborated by analyses of the magnetic
properties of sediment cores collected in the North Atlantic
(see Kissel 2005 for a compilation), which show that the
melting of icebergs was accompanied by a slowing of deep
currents and of the thermohaline circulation.
406
M. Kageyama et al.
coarse detrital material, which were named Heinrich events
(HE) by Broecker et al. in 1992. Heinrich events are defined
by these authors in terms of both their characteristic low
concentration of planktonic foraminifera and high concentration of detrital material. They further demonstrate that
these events are related to armadas of icebergs that broke off
from the ice sheets rather than to the melting of these ice
sheets. The layers of detrital material may be several meters
thick in the Labrador Sea. This thickness is less, although
still significant, close to the European coasts. Subsequent
studies of the composition and properties of this
ice-rafted detrital material show that its origin is mostly from
the Canadian shield and has therefore been transported by
icebergs which detached from the North American ice sheet.
During the same period, the analysis of the d
18 O isotopic
signal from ice cores taken from the summit of the Greenland ice cap (Dansgaard et al. 1993) shows the contrast
between interglacial periods, and particularly between the
remarkably stable Holocene, and the last glacial period,
characterized by high amplitude oscillations. This amplitude
can be as much as half of the glacial-interglacial difference in
Greenland.
The
warming
events
were
named
‘Dansgaard-Oeschger events’ (D/O), because these two
authors had already detected rapid d
18 O fluctuations in the
ice in a core from Camp Century. At the time, the record
appeared dubious due to large simultaneous variations in the
CO 2 content in the air bubbles trapped in the ice. We now
know that the latter are artefacts caused by the presence of
carbonate dust attacked by the sulfuric acid present in the
ice. However, the rapid fluctuations of d
18 O are indeed
significant.
D/O warming appears to have occurred particularly
quickly, over a few decades, solely when ice sheets developed on the continents of the northern hemisphere. The
relatively warm period following the D/O event is called the
‘interstadial period’ and is characterized by gradual cooling.
It ends with a rapid return to the coldest levels recorded,
known as ‘stadial periods’. This return completes a D/O
cycle, which lasts a total of about 1500 years. There are
therefore more D/O events than Heinrich events.
The existence of abrupt variations in both the surface
climate in Greenland and in the ocean conditions over time
scales far shorter than the Milankovich cycles contributed to
the emerging idea that the climate system could have multiple equilibrium points and that it could be abruptly reorganized as it transfers from one equilibrium to another. This
idea was reinforced by the correlations between marine and
glacial records, first demonstrated by Bond et al. (1993).
These authors suggest a perspective of glacial variability that
integrates the Heinrich and Dansgaard-Oeschger events. The
glacial millennial variability is organized in cycles, later
called “Bond cycles” (Fig. 29.1). Each cycle begins at the
end of a Heinrich event, by the first Dansgaard-Oeschger
event, which has a large amplitude. This is followed by a
few other Dansgaard-Oeschger cycles, of diminishing
amplitude. The last cycle ends with a massive discharge of
icebergs from the Laurentide sheet, in other words, by a
Heinrich event. Each cycle lasts approximately 7000–
10,000 years.
Since these discoveries in the early 1990s, millennial
variability has become a subject of intense research.
Numerous records were analyzed with as fine a resolution as
possible and they showed that variability at shorter time
scales than those of Milankovitch was not limited to the
North Atlantic and adjacent regions. The challenge is then to
be able to synchronize these different records to better
characterize this type of variability, as well as to better
understand the connections between climate signals recorded
all over the globe. In the following sections, we focus on the
signature of the Heinrich and D/O events, on their impact on
climate and on the efforts to understand these climate
instabilities through modeling.
The Regional Impacts of the Heinrich
and Dansgaard-Oeschger Events: North Atlantic
and Adjacent Regions
Atlantic Ocean
The Heinrich Events
The melting of massive armadas of icebergs profoundly
modified the ocean surface conditions in the North Atlantic.
The changes are recorded in ocean sediment cores, particularly in d
18 O signals from planktonic foraminifera. Indeed,
excursions towards lighter d
18 O are measured in the calcite
of these foraminifera, indicating either less saline surface
waters or warmer temperatures. The fauna of fossil foraminifera in sediments reveal very cold conditions during
Heinrich events (see, for example, Hemming 2004; Cortijo
et al. 2005). The melting of icebergs is responsible for a
huge inflow of freshwater, desalinating the surface of the
ocean and introducing a highly negative d
18 O signal, typical
of ice contained in the ice sheets.
Other paleoceanographic records reveal significant reorganizations during the Heinrich events. In particular, the
study of d
13 C in sediment cores proves that the bottom
waters of the Atlantic Ocean were less well ventilated (Elliot
2002), demonstrating a reorganization of ocean circulation.
These studies were corroborated by analyses of the magnetic
properties of sediment cores collected in the North Atlantic
(see Kissel 2005 for a compilation), which show that the
melting of icebergs was accompanied by a slowing of deep
currents and of the thermohaline circulation.
406
M. Kageyama et al.