Changes in geographical configuration through continental drift, or more radically, through the opening or
closing of ocean basins, are therefore likely to cause profound changes in ocean circulation. Either shallow or deep,
ocean gateways play a crucial role in water masses circulation and associated heat and salt fluxes (Ferreira et al. 2018).
During the Cenozoic, the position of the continents has not
drastically changed but some ocean gateways have been
opened or closed thereby impacting ocean circulation. The
Tasman seaway opened during the Late Eocene, the Drake
passage during the Late Eocene/Oligocene, the Fram Strait
during the Oligocene. Conversely the East Tethys seaway
closed during the Early/Middle Miocene, the Central
American Seaway during the Late Miocene, the Indonesian
throughflow during the Late Miocene/Pliocene. The timeline
of these opening/closures are therefore crucial for paleoclimate reconstruction purposes, and can be constrained
through the use of paleoceanographic markers and sea level
records.
Sea level changes profoundly change the face of the
Earth. Periods of high sea level, such as during the Cenomanian (*95 Ma) studied by Eduard Suess at the end of the
nineteenth century, caused the flooding of large continental
areas. Vast shelf seas, generally shallow (<200 m), covering
up to 30% of the land area, were thus observed during the
Phanerozoic period. In some cases the combination of a high
sea level and a flexure of the lithosphere can result in the
formation of shelf seas several hundred meters deep. In other
cases, flooding or drying is controlled by dynamic
topography.
The dynamics of these epicontinental seas are still not
well understood because there is no analog of such large
shallow water basins today. These epicontinental seas would
have favored climates with little thermal contrast due to the
high calorific capacity of the water and their thermal inertia.
Conversely, periods with low sea levels would have favored
more contrasted climates. Analysis of the sedimentary facies
makes it possible to locate the coastlines. These have
changed with eustatic variations (change of sea level), but
also with continental uplift and subsidence, the flexure of the
lithosphere in response to the build up of ice sheet or its
melting or variations in the sedimentary fluxes. Changes in
coastlines inferred from the sedimentary facies do not
therefore constitute a direct marker of sea level.
In 1977, Peter Vail and his team, geologists with the
American oil company Exxon, produced a curve of sea level
changes. Vail showed that the geometry and position of the
sedimentary units deposited on the continental shelves and
in the basins had varied according to sea level, the subsidence of the area of deposition, the sedimentary flow and the
carbonate production. The geometry and position of the
sedimentary units are the main influencers of variations in
sea level, as subsidence, deposition and carbonate
production are considered to be less variable. The determination of sedimentary facies from core samples taken by
drilling and the location of seismic reflectors on the continental margins and adjacent basins make it possible to
reconstruct the arrangement of sedimentary deposits and to
infer eustatic variations. Several curves have since been
produced (Haq et al. 1987; Haq and Schutter 2008). The
curves obtained (Fig. 2.4) show nested eustatic cycles. The
largest eustatic variations, about 200 m, with a timescale of
tens of million years, are related to life cycle of supercontinents (Wilson cycle), plate reorganization, dynamic
topography and crustal production variations. Secondly,
significant variations (several tens of meters) but with a
lower timescale (from *2 to *10 Ma) are likely due to
regional tectonic forcing. Thirdly, the sea level fluctuations
on the timescale of 0.5 to *2 Ma are not well understood, it
could be related to the climate changes or ice volume variations. Even more rapid variations on the timescale of tens to
hundred years are clearly related to glacioeustatism whereas
postglacial rebound acts on timescale of thousands to hundreds of thousands of years.
However, this interpretation of the high frequency variability found on an eustatic graph at the scale of geological
time has been called into question. Moreover, the covering
of coasts observed in sedimentary systems may not necessarily translate into eustatic variations.
A different method from that of Haq et al. (1987) was
proposed by Miller et al. (2005). It determines eustatic
variations using an inverse model by calculating the effects
of the sediment load, compaction and the oceanic load
necessary to simulate the deformation of a basin (subsidence
or uplift) located on a passive margin. The variations in sea
level obtained by this method do not exceed 100 m, which is
half that proposed by Haq et al. (1987) for the Phanerozoic.
Moreover, the eustatic variations obtained by these two
methods may be out of phase with each other. Unlike the
curve by Haq et al. (1987), which is based on numerous
records, Miller’s sea level variation curve is based on a small
number of sites on the eastern margin of North America. The
deformation undergone by this basin caused by internal
dynamic processes originating in the mantle is not corrected.
This basin may have recorded a “dynamic” topography with
long wavelength radiation (as opposed to the topography
linked to isostasy of a lithospheric column) partially skewing
the eustatic signal. Debates still exist regarding the exact
evolution of sea-level throughout the Phanerozoic, as a more
recent study, based on a full geodynamical model, provided
a large range of sea-level fluctuations for the last 500 Ma
(Vérard et al. 2015) (Fig. 2.4).
To understand these discrepancies, one must analyze the
mechanisms of sea level change. Variations in the position
of coastlines can come from several different sources: variation in the volume of water (the content) or variation in the
32
F. Fluteau and P. Sepulchre
closing of ocean basins, are therefore likely to cause profound changes in ocean circulation. Either shallow or deep,
ocean gateways play a crucial role in water masses circulation and associated heat and salt fluxes (Ferreira et al. 2018).
During the Cenozoic, the position of the continents has not
drastically changed but some ocean gateways have been
opened or closed thereby impacting ocean circulation. The
Tasman seaway opened during the Late Eocene, the Drake
passage during the Late Eocene/Oligocene, the Fram Strait
during the Oligocene. Conversely the East Tethys seaway
closed during the Early/Middle Miocene, the Central
American Seaway during the Late Miocene, the Indonesian
throughflow during the Late Miocene/Pliocene. The timeline
of these opening/closures are therefore crucial for paleoclimate reconstruction purposes, and can be constrained
through the use of paleoceanographic markers and sea level
records.
Sea level changes profoundly change the face of the
Earth. Periods of high sea level, such as during the Cenomanian (*95 Ma) studied by Eduard Suess at the end of the
nineteenth century, caused the flooding of large continental
areas. Vast shelf seas, generally shallow (<200 m), covering
up to 30% of the land area, were thus observed during the
Phanerozoic period. In some cases the combination of a high
sea level and a flexure of the lithosphere can result in the
formation of shelf seas several hundred meters deep. In other
cases, flooding or drying is controlled by dynamic
topography.
The dynamics of these epicontinental seas are still not
well understood because there is no analog of such large
shallow water basins today. These epicontinental seas would
have favored climates with little thermal contrast due to the
high calorific capacity of the water and their thermal inertia.
Conversely, periods with low sea levels would have favored
more contrasted climates. Analysis of the sedimentary facies
makes it possible to locate the coastlines. These have
changed with eustatic variations (change of sea level), but
also with continental uplift and subsidence, the flexure of the
lithosphere in response to the build up of ice sheet or its
melting or variations in the sedimentary fluxes. Changes in
coastlines inferred from the sedimentary facies do not
therefore constitute a direct marker of sea level.
In 1977, Peter Vail and his team, geologists with the
American oil company Exxon, produced a curve of sea level
changes. Vail showed that the geometry and position of the
sedimentary units deposited on the continental shelves and
in the basins had varied according to sea level, the subsidence of the area of deposition, the sedimentary flow and the
carbonate production. The geometry and position of the
sedimentary units are the main influencers of variations in
sea level, as subsidence, deposition and carbonate
production are considered to be less variable. The determination of sedimentary facies from core samples taken by
drilling and the location of seismic reflectors on the continental margins and adjacent basins make it possible to
reconstruct the arrangement of sedimentary deposits and to
infer eustatic variations. Several curves have since been
produced (Haq et al. 1987; Haq and Schutter 2008). The
curves obtained (Fig. 2.4) show nested eustatic cycles. The
largest eustatic variations, about 200 m, with a timescale of
tens of million years, are related to life cycle of supercontinents (Wilson cycle), plate reorganization, dynamic
topography and crustal production variations. Secondly,
significant variations (several tens of meters) but with a
lower timescale (from *2 to *10 Ma) are likely due to
regional tectonic forcing. Thirdly, the sea level fluctuations
on the timescale of 0.5 to *2 Ma are not well understood, it
could be related to the climate changes or ice volume variations. Even more rapid variations on the timescale of tens to
hundred years are clearly related to glacioeustatism whereas
postglacial rebound acts on timescale of thousands to hundreds of thousands of years.
However, this interpretation of the high frequency variability found on an eustatic graph at the scale of geological
time has been called into question. Moreover, the covering
of coasts observed in sedimentary systems may not necessarily translate into eustatic variations.
A different method from that of Haq et al. (1987) was
proposed by Miller et al. (2005). It determines eustatic
variations using an inverse model by calculating the effects
of the sediment load, compaction and the oceanic load
necessary to simulate the deformation of a basin (subsidence
or uplift) located on a passive margin. The variations in sea
level obtained by this method do not exceed 100 m, which is
half that proposed by Haq et al. (1987) for the Phanerozoic.
Moreover, the eustatic variations obtained by these two
methods may be out of phase with each other. Unlike the
curve by Haq et al. (1987), which is based on numerous
records, Miller’s sea level variation curve is based on a small
number of sites on the eastern margin of North America. The
deformation undergone by this basin caused by internal
dynamic processes originating in the mantle is not corrected.
This basin may have recorded a “dynamic” topography with
long wavelength radiation (as opposed to the topography
linked to isostasy of a lithospheric column) partially skewing
the eustatic signal. Debates still exist regarding the exact
evolution of sea-level throughout the Phanerozoic, as a more
recent study, based on a full geodynamical model, provided
a large range of sea-level fluctuations for the last 500 Ma
(Vérard et al. 2015) (Fig. 2.4).
To understand these discrepancies, one must analyze the
mechanisms of sea level change. Variations in the position
of coastlines can come from several different sources: variation in the volume of water (the content) or variation in the
32
F. Fluteau and P. Sepulchre