Although the development of a permanent ice cap on
Antarctica started early, during the early Neogene, the
development of perennial continental ice caps in the high
latitudes of the northern hemisphere did not occur until the
end of the Neogene. The first deposits of IRD in the Norwegian Sea, proof of the early development of an ice cap
(although perhaps not a permanent one) on Greenland are
not observed before 5.5 Ma (Jansen and Sjoholm 1991). The
rapid increase in d
18 O from *3.2 Ma onwards may be
interpreted as the beginning of permanent glaciation in high
latitudes of the northern hemisphere. This glaciation intensified rapidly around 2.1–2.6 Ma, as evidenced by the
massive IRD deposits in the Norwegian Sea.
Many studies have focused on the hypothesis of the
‘closure of the Panama isthmus’ as a potential trigger for the
development of northern hemisphere ice sheets. The mechanism would involve warm intertropical waters no longer
being able to cross from the Atlantic to the Pacific. They
would therefore deviate into the North Atlantic, increasing
oceanic evaporation in this basin and thus snow accumulation in the high latitudes. However, recent studies suggest
that the closure of the Panama isthmus could have occurred
during the Miocene, well before the intensification of
glaciations (Montes et al. 2015). The recent work of Rohling
et al. (2014) also observes a large temporal offset during the
onset of the Plio-Pleistocene ice ages, between a marked
cooling step at 2.73 My ago and the first major glaciation
starting 2.15 My ago. Other theories indicate that a decrease
in atmospheric CO 2 may have been responsible for a cooling, an increase of deep water formation in the North
Atlantic and a change of circulation that together induced the
start of the glaciations.
The ‘Middle Pleistocene Transition’
and the Establishment of 100-Ka Cycles
The trend towards the climatic decline (seen as an increase in
benthic d
18 O) discussed in Sect. “From ‘The Greenhouse
Effect Era’ to the ‘Ice Ages’” continued over the last two
million years, as is shown in detail in Fig. 21.2. Superimposed on this trend are quasi-periodic oscillations. They
reflect the alternating glacial periods—corresponding to a
cooling of deep waters and an increase in ice volume at high
latitudes—and interglacial periods, with warming and relative melting of the ice caps. It should be emphasized here
that the use of the terms ‘interglacial’ and ‘glacial’ does not
imply a total melting of ice sheets. During interglacial
periods, ice sheets do not disappear, even if they are greatly
reduced in the northern hemisphere. For example, during the
LGM, the ice sheets in the northern hemisphere covered a
large portion of North America and Europe. During the
Holocene, the interglacial period we currently live in, these
caps were largely diminished; the meltwater derived from
them has caused sea level to rise by 120 m since the LGM.
However, an ice cap of 2.8 million km
3 continues to exist on
Greenland that would cause a further sea level rise of about 7
meters, if it were to completely melt.
The amplitude of the glacial-interglacial oscillations
increased sharply between 1.2 and 0.6 Ma (Fig. 21.2).
During this period, called ‘the Middle Pleistocene Transition’ (Clark et al. 2006; McClymont et al. 2013), a threshold
response to longer-term atmospheric CO 2 decline has been
proposed (Raymo et al. 1997). However, recent atmospheric
partial pressure CO 2 reconstructions have failed to show this
long-term decrease during the Pleistocene (Hönisch et al.
2009). The gradual increase in glacial-interglacial amplitude
is mainly due to increasingly high values of d
18
O during
glacial periods. The few available reconstructions of deepwater temperature during this period indicate near-freezing
temperatures at every glacial maximum instead of a gradual
cooling (Elderfield et al. 2012), which suggests that an
increase in Antarctic ice volume would be responsible for
the rapid and steep increase in seawater d
18
O at 0.9 Ma.
This change in amplitude of glacial-interglacial oscillations is accompanied by a disruption in the frequency content of the global d
18
O signal. While benthic foraminifera
d
18 O oscillations show mainly a cycle of *41 ka over most
of the Neogene and early Quaternary, the last 600,000 years
are dominated by oscillations with a cyclicity of *100 ka
(Fig. 21.2). Some authors have agreed on the progressive
nature of the ‘Middle Pleistocene Transition’, with the
amplification of the 100 ka cycles occurring over hundreds
of thousands of years. However, in some ocean regions, the
records fail to demonstrate this progressive nature. This is
the case, for example, in the equatorial Atlantic, where the
dynamics of the thermocline, reconstructed from micropaleontological tracers, suddenly change its variability
around *930 ka. The mechanisms responsible for this
transition are still unclear, although it appears that an
important role can be attributed to the enormous Laurentide
ice sheet, which may have favored the frequency of 100 ka
through its inertia (Clark and Pollard 1998) (see Chap. 28).
The Last Glacial Maximum (LGM)
The LGM has long been, and still remains, a major area of
interest in paleoclimatology, in particular because it presents
another extreme on the climate spectrum on which Earth
System models can be validated (Kageyama et al. 2018).
Early studies defined this period as the time encompassing
21 Climate and the Evolution of the Ocean: The Paleoceanographic …
243
Antarctica started early, during the early Neogene, the
development of perennial continental ice caps in the high
latitudes of the northern hemisphere did not occur until the
end of the Neogene. The first deposits of IRD in the Norwegian Sea, proof of the early development of an ice cap
(although perhaps not a permanent one) on Greenland are
not observed before 5.5 Ma (Jansen and Sjoholm 1991). The
rapid increase in d
18 O from *3.2 Ma onwards may be
interpreted as the beginning of permanent glaciation in high
latitudes of the northern hemisphere. This glaciation intensified rapidly around 2.1–2.6 Ma, as evidenced by the
massive IRD deposits in the Norwegian Sea.
Many studies have focused on the hypothesis of the
‘closure of the Panama isthmus’ as a potential trigger for the
development of northern hemisphere ice sheets. The mechanism would involve warm intertropical waters no longer
being able to cross from the Atlantic to the Pacific. They
would therefore deviate into the North Atlantic, increasing
oceanic evaporation in this basin and thus snow accumulation in the high latitudes. However, recent studies suggest
that the closure of the Panama isthmus could have occurred
during the Miocene, well before the intensification of
glaciations (Montes et al. 2015). The recent work of Rohling
et al. (2014) also observes a large temporal offset during the
onset of the Plio-Pleistocene ice ages, between a marked
cooling step at 2.73 My ago and the first major glaciation
starting 2.15 My ago. Other theories indicate that a decrease
in atmospheric CO 2 may have been responsible for a cooling, an increase of deep water formation in the North
Atlantic and a change of circulation that together induced the
start of the glaciations.
The ‘Middle Pleistocene Transition’
and the Establishment of 100-Ka Cycles
The trend towards the climatic decline (seen as an increase in
benthic d
18 O) discussed in Sect. “From ‘The Greenhouse
Effect Era’ to the ‘Ice Ages’” continued over the last two
million years, as is shown in detail in Fig. 21.2. Superimposed on this trend are quasi-periodic oscillations. They
reflect the alternating glacial periods—corresponding to a
cooling of deep waters and an increase in ice volume at high
latitudes—and interglacial periods, with warming and relative melting of the ice caps. It should be emphasized here
that the use of the terms ‘interglacial’ and ‘glacial’ does not
imply a total melting of ice sheets. During interglacial
periods, ice sheets do not disappear, even if they are greatly
reduced in the northern hemisphere. For example, during the
LGM, the ice sheets in the northern hemisphere covered a
large portion of North America and Europe. During the
Holocene, the interglacial period we currently live in, these
caps were largely diminished; the meltwater derived from
them has caused sea level to rise by 120 m since the LGM.
However, an ice cap of 2.8 million km
3 continues to exist on
Greenland that would cause a further sea level rise of about 7
meters, if it were to completely melt.
The amplitude of the glacial-interglacial oscillations
increased sharply between 1.2 and 0.6 Ma (Fig. 21.2).
During this period, called ‘the Middle Pleistocene Transition’ (Clark et al. 2006; McClymont et al. 2013), a threshold
response to longer-term atmospheric CO 2 decline has been
proposed (Raymo et al. 1997). However, recent atmospheric
partial pressure CO 2 reconstructions have failed to show this
long-term decrease during the Pleistocene (Hönisch et al.
2009). The gradual increase in glacial-interglacial amplitude
is mainly due to increasingly high values of d
18
O during
glacial periods. The few available reconstructions of deepwater temperature during this period indicate near-freezing
temperatures at every glacial maximum instead of a gradual
cooling (Elderfield et al. 2012), which suggests that an
increase in Antarctic ice volume would be responsible for
the rapid and steep increase in seawater d
18
O at 0.9 Ma.
This change in amplitude of glacial-interglacial oscillations is accompanied by a disruption in the frequency content of the global d
18
O signal. While benthic foraminifera
d
18 O oscillations show mainly a cycle of *41 ka over most
of the Neogene and early Quaternary, the last 600,000 years
are dominated by oscillations with a cyclicity of *100 ka
(Fig. 21.2). Some authors have agreed on the progressive
nature of the ‘Middle Pleistocene Transition’, with the
amplification of the 100 ka cycles occurring over hundreds
of thousands of years. However, in some ocean regions, the
records fail to demonstrate this progressive nature. This is
the case, for example, in the equatorial Atlantic, where the
dynamics of the thermocline, reconstructed from micropaleontological tracers, suddenly change its variability
around *930 ka. The mechanisms responsible for this
transition are still unclear, although it appears that an
important role can be attributed to the enormous Laurentide
ice sheet, which may have favored the frequency of 100 ka
through its inertia (Clark and Pollard 1998) (see Chap. 28).
The Last Glacial Maximum (LGM)
The LGM has long been, and still remains, a major area of
interest in paleoclimatology, in particular because it presents
another extreme on the climate spectrum on which Earth
System models can be validated (Kageyama et al. 2018).
Early studies defined this period as the time encompassing
21 Climate and the Evolution of the Ocean: The Paleoceanographic …
243