explanations today relies on an internal oscillation of oceanic
circulation under given boundary conditions (Paillard 2004),
without involving any external forcing. The big advantage of
this hypothesis is that it accounts both for the frequency of
cycles, compatible with what we know of the ocean, and for
their global character. A coupled ocean-atmosphere internal
oscillation has more recently been proposed as the reason for
the occurence of Dansgaard-Oeschger events as well (Peltier
and Vettoretti 2014). This study has the advantage of
showing such an internal oscillation of the climate system in
a more realistic General Circulation Model.
Nevertheless, opposing theories coexist nowadays when
trying to explain the triggering mechanisms of DansgaardOeschger events. Banderas et al. (2015) also relate their
occurence to an internal climatic oscillation, but in their
case, it is controlled by the effects of Southern-Ocean wind
stress and CO 2 on the oceanic circulation. Similarly, Dima
et al. (2018) proposed that the meridional overturning circulation changes are ultimately controlled by the Southern
Ocean with links to gradual global climate changes. Besides,
it has recently been proposed that Dansgaard-Oeschger
events owe their existence to the interactions between the
North Atlantic ice shelves, sea ice and oceanic circulation
(Boers et al. 2018).
To conclude, there is no consensus today on the cause of
the Dansgaard-Oeschger events, despite an increasing
understanding of the effects that these events have had on
global climate.
Sudden Events During Interglacial Periods
The Discovery
Although the climate oscillation called ‘the 8200-year event’
left traces in the Greenland cores, the first publications
presenting the isotopic records from the ice in Greenland
from GRIP cores (Dansgaard et al. 1993) described the climate of the Holocene as stable, as the weak oscillation
occurring around 8200 years before today was considered to
be background noise. Even the discovery of a second record
with exactly the same oscillation did not allow the correct
recognition of this climate event. It was not until four years
later that the notion of ‘the 8200-year event’ was introduced
in an article by a team led by R. Alley. Efforts to place this
climate evolution event on a hemispheric scale have indicated that this event had in fact been identified since 1973
from other records.
The Observations
The 8200-year event has been recorded in numerous climate
archives, including the Greenland polar ice cap, continental
archives (lakes, vegetation, speleothems) and the ocean.
In Greenland cores, this event shows up as a decrease in
d
18 O of about two permil over a period of about 100 years.
This decrease is interpreted as a drop in local temperature in
Greenland of 6 ± 2 °C (see Thomas et al. 2007). It was
estimated at 7.4 °C using an independent method based on
oxygen isotopes (d
15 N) (Kobashi et al. 2007). This was a
fairly dramatic event, at least over Greenland. This oscillation is found in all the cores of adequate resolution obtained
from the Greenland ice sheet: GRIP, GISP2, NGRIP, Dye-3.
This is therefore an event with a regional impact, at least, on
the scale of Greenland.
The cold period equating to the 8200-year event in
Greenland is also recorded in other paleoclimate indicators.
Although this event was not recognized as such in polar ice
records, it was well identified in records from Norwegian
lakes and in the extended ice sheets reconstructed from the
end moraines. Indeed, the cold and dry period ‘Finse
1
’ was
identified (Dahl and Nesje 1994), and then connected to the
8200-year event in the ice cores. Simultaneously, this event
was associated with a cold spell reconstructed from variations in planktonic foraminiferal fauna in a marine core from
the south of Norway. This record proves that it is indeed a
significant climate event on the scale of the North Atlantic,
but also that the ocean was affected almost simultaneously.
This interpretation is reinforced by other more distant
records of the event. One of the first studies to show this was
the study of the d
18 O of the ostracod shells from a Bavarian
lake (Ammersee, Germany), which recorded changes in the
d
18 O of the precipitations associated with cooling (Chap. 15,
von Grafenstein 1998).
In the oceans, records marking the 8200-year event are
rare, as identification requires a temporal resolution of much
less than a century. Of these records, those from the northern
seas show a decrease of about 3 °C in ocean surface temperature. This study proves that this event was the largest
during the entire Holocene in this region. Finally, these
authors show that the oceanic change affected not only the
surface but also the deep ocean. These last changes have
been confirmed recently in a core from southern Greenland,
which shows changes in the formation of deep water bodies
in the North Atlantic.
On the continent, the 8200-year event is expressed as a
cooling especially evident in Europe. Studies of the changes
1
Named after the Finsevatn, the Norwegian Hardanger Lake where the
record originated.
416
M. Kageyama et al.
circulation under given boundary conditions (Paillard 2004),
without involving any external forcing. The big advantage of
this hypothesis is that it accounts both for the frequency of
cycles, compatible with what we know of the ocean, and for
their global character. A coupled ocean-atmosphere internal
oscillation has more recently been proposed as the reason for
the occurence of Dansgaard-Oeschger events as well (Peltier
and Vettoretti 2014). This study has the advantage of
showing such an internal oscillation of the climate system in
a more realistic General Circulation Model.
Nevertheless, opposing theories coexist nowadays when
trying to explain the triggering mechanisms of DansgaardOeschger events. Banderas et al. (2015) also relate their
occurence to an internal climatic oscillation, but in their
case, it is controlled by the effects of Southern-Ocean wind
stress and CO 2 on the oceanic circulation. Similarly, Dima
et al. (2018) proposed that the meridional overturning circulation changes are ultimately controlled by the Southern
Ocean with links to gradual global climate changes. Besides,
it has recently been proposed that Dansgaard-Oeschger
events owe their existence to the interactions between the
North Atlantic ice shelves, sea ice and oceanic circulation
(Boers et al. 2018).
To conclude, there is no consensus today on the cause of
the Dansgaard-Oeschger events, despite an increasing
understanding of the effects that these events have had on
global climate.
Sudden Events During Interglacial Periods
The Discovery
Although the climate oscillation called ‘the 8200-year event’
left traces in the Greenland cores, the first publications
presenting the isotopic records from the ice in Greenland
from GRIP cores (Dansgaard et al. 1993) described the climate of the Holocene as stable, as the weak oscillation
occurring around 8200 years before today was considered to
be background noise. Even the discovery of a second record
with exactly the same oscillation did not allow the correct
recognition of this climate event. It was not until four years
later that the notion of ‘the 8200-year event’ was introduced
in an article by a team led by R. Alley. Efforts to place this
climate evolution event on a hemispheric scale have indicated that this event had in fact been identified since 1973
from other records.
The Observations
The 8200-year event has been recorded in numerous climate
archives, including the Greenland polar ice cap, continental
archives (lakes, vegetation, speleothems) and the ocean.
In Greenland cores, this event shows up as a decrease in
d
18 O of about two permil over a period of about 100 years.
This decrease is interpreted as a drop in local temperature in
Greenland of 6 ± 2 °C (see Thomas et al. 2007). It was
estimated at 7.4 °C using an independent method based on
oxygen isotopes (d
15 N) (Kobashi et al. 2007). This was a
fairly dramatic event, at least over Greenland. This oscillation is found in all the cores of adequate resolution obtained
from the Greenland ice sheet: GRIP, GISP2, NGRIP, Dye-3.
This is therefore an event with a regional impact, at least, on
the scale of Greenland.
The cold period equating to the 8200-year event in
Greenland is also recorded in other paleoclimate indicators.
Although this event was not recognized as such in polar ice
records, it was well identified in records from Norwegian
lakes and in the extended ice sheets reconstructed from the
end moraines. Indeed, the cold and dry period ‘Finse
1
’ was
identified (Dahl and Nesje 1994), and then connected to the
8200-year event in the ice cores. Simultaneously, this event
was associated with a cold spell reconstructed from variations in planktonic foraminiferal fauna in a marine core from
the south of Norway. This record proves that it is indeed a
significant climate event on the scale of the North Atlantic,
but also that the ocean was affected almost simultaneously.
This interpretation is reinforced by other more distant
records of the event. One of the first studies to show this was
the study of the d
18 O of the ostracod shells from a Bavarian
lake (Ammersee, Germany), which recorded changes in the
d
18 O of the precipitations associated with cooling (Chap. 15,
von Grafenstein 1998).
In the oceans, records marking the 8200-year event are
rare, as identification requires a temporal resolution of much
less than a century. Of these records, those from the northern
seas show a decrease of about 3 °C in ocean surface temperature. This study proves that this event was the largest
during the entire Holocene in this region. Finally, these
authors show that the oceanic change affected not only the
surface but also the deep ocean. These last changes have
been confirmed recently in a core from southern Greenland,
which shows changes in the formation of deep water bodies
in the North Atlantic.
On the continent, the 8200-year event is expressed as a
cooling especially evident in Europe. Studies of the changes
1
Named after the Finsevatn, the Norwegian Hardanger Lake where the
record originated.
416
M. Kageyama et al.