Sealevel Changes

W hile the theory of plate tectonics was taking center stage, a quieter revolution was helping us to understand other
processes controlling the evolution of ocean margins. Scientists working on Pleistocene and Holocene sea-level
histories provided new insights into the patterns of glaciations that have characterized the last few million years of earth
history.

Studies of sea-level changes prior to the Pleistocene have focused on longer-term, cyclic events in the stratigraphic record
that have been controlled by both global warming/cooling and large-scale tectonics.

Still-stands of sea level prior to historic data can be inferred by features related to steady wave action over a long period
of time (i.e., wave-cut benches. terraces and "nick points"). Fossil material deposited in nearshore sediments and in coral
reefs may contain organisms of limited depth range, which can serve as relative indicators of water depth at the time of
deposition.

Mean sea level is defined as the average height of the sea surface referenced to a fixed point. Changes in sea level
measured at a particular site may reflect either worldwide or local conditions, and changes have occurred many times in
the past. We will examine the causes of sea-level variations, methods of measuring sea level, and then discuss the
importance of sea-level change to marine deposition.

The Concept of Relative Sea Level

W hether measured directly or inferred from past geologic evidence, sea level is a relative term that results from the
eustatic or tectonic changes. For example, global sea level might be fluctuating due to growth and melting of
continental glaciers, but there are many regional processes that result in rise or fall of relative sea level that affect one
coastline and not the other. Some of these processes are:

• thermal expansion of ocean waters,

• changes in meltwater load,
• crustal rebound from glaciation,
• uplift or subsidence in coastal areas related to
various tectonic processes,
• fluid withdrawal, and
• sediment deposition and compaction.

At the same time, the sea floor may be rising in one

ocean and falling in an adjacent basin. As a result, the
sea-level histories reflected in the two basins will seem
different, while in fact the actual level of the world
ocean was falling uniformly. The marked differences in
the sea-level curves derived from the eastern coast of
Australia and the US Atlantic coast for the past 10,000
years reflect tectonic differences in those two areas.

Local tectonic movements have been commonplace

throughout geologic time. Today, the modern U.S.
Atlantic coast is generally subsiding while the Pacific
coast is emerging. Along both coasts, however, local
variations in subsidence and uplift occur. The Oregon is
bring uplifted at a rate roughly equal to sea level rise,
while many areas on the Mississippi River Delta are subsiding due to sediment loading. As a result, the relative sea level
curves for these sites differ significantly from each other and from the global average.
1
If this is not already confusing enough, we must add to this the fact that even today mean sea level is not the same the
world over. For example, mean sea level at opposite ends of the Panama Canal differs by several meters. When discussing
the concept of sea level change one must, therefore, be very careful to distinguish between eustatic and relative sea level.

Eustatic Sea Level Change

E ustatic sea-level changes occur on an oceanic to worldwide scale. They result from either a change in the volume of
seawater, or a change in the size of the ocean basin that contains it. Even in these large-scale cases, however, mean
sea level can vary from place to place due to local tectonic and hydrographic effects. Several processes that can cause
worldwide changes in sea level are listed below.

Glaciation

F or a number of reasons, the volume of glacial ice near the poles waxes and wanes over time. As a result, water is
alternately taken from or added to the world oceans. This can result in sea-level oscillations of up to 200 meters. For
example, modern continental glaciers are 1.5 to 2.5 km thick and have a total estimated volume of 33 million km3. If we
assume the maximum volume of Pleistocene glaciers to have been 71.3 million km3, then the difference is 38 million km3.
Using the assumption that glacial water volume is 91.7% of the volume of sea water from which it is derived, a sea-level
drop of 106 m can be accounted for by Pleistocene glaciation. Melting of the present Greenland and Antarctic glaciers
would produce a sea-level rise of approximately 60 meters.

Since 18,000 years ago, sea level has been rising in response to the melting of glacial ice, mainly from the Laurentian and
the Scandinavian ice sheets of northern Canada. Rapid changes in the oxygen isotopic composition of skeletal
foraminifera and corals (melt water has a low O18/O16 ratio) indicate that this rise was not uniform; two major and several
minor pulses of melt water were added to the ocean. Based on gaps in the branching-coral (Acropora palmata) record
offshore of Barbados, Fairbanks proposed major meltwater pulses at about 14,000 and 11,000 yBP. Whatever the actual
age or number of meltwater pulses, the main point is that sudden changes in glacial volume have resulted in significant
and rapid transients in sea level.

In addition to sea-level changes caused by the transfer of seawater to and from glaciers, sea water volume can change due
to thermal expansion and contraction – for every 1o C decrease in the mean temperature, world-wide sea level will drop by
2 m. Fauna from Pleistocene ocean sediments suggest a 5 degree lower surface temperature at that time, which would
have resulted in a 10-meter lower sea level due solely to thermal contraction.

2
Despite a growing amount of data, we are still have basic questions about glacially driven sea-level change, including the
amount of rise and fall and the cause of glacial cycles. The astronomic control of glaciation proposed by Milankovitch has
been generally accepted to explain Pleistocene glacial cycles. The concept is that the nature of earth's orbit around the sun
plays a primary role in patterns of glaciation and and, therefore, sea level change. The amount of solar radiation reaching
the earth at any latitude or season is determined by three aspects of the position of the earth in its orbit around to the sun.
These include:

• the eccentricity of the earth's orbit,

• the tilt of the earth's axis relative to the orbital plane, and
• seasonal changes in the time when the earth is most proximal to the sun (precession).

The earth follows an elliptical orbit around the sun. The shape of the elliptical
orbit, which is measured by its eccentricity, varies from between one and five
percent through time. During times of minimum eccentricity, solar radiation is
more uniformly distributed through the year, and seasonality is less pronounced
than during times of high eccentricity. The period between successive
eccentricity maxima and minima is about 96,000 years.

The season when the earth is closest to

the sun (the perihelion) changes over
time. Because of this precession, the vernal equinox occurs at perihelion every
22,000 years with the autumnal equinox occurring midway between (i.e., 11,000).
If the earth is closest to the sun during the
northern hemisphere summer and farthest
during the northern winter, then the northern
hemisphere climate will have hot summers
and cold winters. These conditions are not
conducive to glaciation because cold winters
are usually drier than mild winters (i.e., less
snowfall), and snow will likely melt during
the hot summer. In contrast, when the earth is
closest to the sun during the northern
hemisphere winter and farthest during the summer, milder climates prevail year round.
As a result, the wetter winters and cooler summers encourage glacial accumulation.

Finally, the earth's axis of rotation can have a profound effect on global climate patterns. Presently the earth is tilted at
23.5° relative to the orbital plane of earth around the sun. As a result, latitudes north of 66.5°N are totally dark during the
northern winter solstice while latitudes south of 66.5°S are in total daylight. This tilt varies by about three degrees, from
21.5° to 24.5°. This tilt or obliquity, influences the distribution of solar radiation throughout the year and over the 41,000-
year cycle from one extreme to the other. During minimum obliquity, there is slightly less seasonal variability in solar
radiation that during maxima.

All three elements (orbital eccentricity, its precession and the changing tilt of the earth) influence the distribution of solar
radiation reaching the earth's atmosphere, acting to increase or diminish seasonal differences and, therefore, the tendency
for glaciation. Milankovitch mathematically combined these variations to produce a northern hemisphere climate curve
that predicted a clear periodicity to the amount of incident radiation and, therefore, glacial cycles. He claimed that
astronomic factors could introduce significant cyclic variation into the earth's climate even with no changes in solar
activity or the earth's albedo and no changes in composition of the earth's atmosphere. His proposal did not deny that other
factors could be important, but suggested that variations in earth's orbital elements alone could trigger significant climatic
change.

A "Greenhouse effect" is the behavior of solar radiation when it interacts with gases in the Earth's atmosphere like carbon
dioxide, methane and water vapor. The radiation absorbed by atmospheric gasses is infrared (IR) radiation. These
atmospheric gases are known as "greenhouse gases." Recent studies have highlighted carbon dioxide and other

3
greenhouse gases as central to our understanding of global climate change. Ice cores tells us about the Earth's past climate
by small bubbles of air that were trapped in glaciers, that are actual samples of ancient air.

Deep cores in Antarctic ice show that the amounts of carbon

dioxide and methane in the earth's atmosphere have varied for at
least the last 160,000 years. These changes correlate closely with
changes in global temperature and the elevation of sea level. These
variations may have been caused by natural changes in the
biosphere, circulation of oceanic deep water or volcanic activity.

More recently, global warming has been tied to the dramatic

increase in carbon dioxide and chloroflorocarbons associated with
growing industrial pollution, increased automobile emissions and
the destruction of large tracts of forest, especially in the tropical
regions of the planet. According to the report of the
Intergovernmental Panel on Climate Change (IPCC), 1990 levels of
carbon dioxide and methane were 26 and 115 percent, respectively
above pre-industrial levels less than a century earlier. They
continue to increase at rates of 0.5 to 0.9% annually. Because these
"greenhouse gasses" tend to trap heat within the earth atmosphere,
the expected result is an increase in the rate of sea-level rise. This is
discussed in
greater detail
at the end of
this chapter.

In contrast, the
level of
incident light
and resulting
global
temperatures
can be
dramatically
reduced by
increased
particulate material in the atmosphere during periods of unusually
high volcanic eruptions or other large-scale events. For example,
snowfall was recorded in the Bahamas as a result of the eruption of
Pinatubo in the Philippines.

Tectonics

C rustal movements may be local or global in scale. For example,

tectonic activity along mid-ocean ridges will cause eustatic sea-
level changes that are expressed worldwide. Locally the coast may be
raised or lowered relative to regional sea level by subsidence or
uplift.

The most common tectonic mechanism to impact global sea-level

change is the movement of the earth's lithospheric plates. This
includes the opening and closing of major ocean basins, the addition
of new crust along mid-ocean ridges, and changes in continental or
plate margins with subduction and isostatic adjustments to loading by
new crustal material. At the same time, volcanic activity along the
4
mid-ocean ridges (as well as subaerial vulcanism) releases juvenal water that is released from the rock during extrusion.

As sea-floor spreading proceeds, basin shape and volume changes. Near the ridge, the new and more buoyant sea floor
rises. Near the oceanic margins the older and increasingly dense crust sinks or is subducted beneath adjacent plates.
Whether sea-floor spreading causes a sea-level rise or fall is related to the spreading rate. Under conditions of rapid
spreading, the buoyant crust has less time to cool (and sink) and sea level rises in response to the shallowing basin. In
ocean basins where spreading is slow, crustal sinking dominates and sea level will fall.

While the origin of the Atlantic provided a new ocean basin, ensuing events also affected the bathymetry of existing
basins and further changed the volume of the world ocean. The early Atlantic Ocean was shallow until mid to late
Cretaceous time, gradually replacing the deeper Pacific Ocean. As the Atlantic Ocean continued to grow, changes in the
rate of spreading and the volume of the mid-Atlantic ridge resulted in changes in the total volume of the ocean basins,
thereby triggering changes in sea level. Similar tectonic processes have been present over the span of geologic time and
probably account for most of the cyclicity of sea-level change in pre-Pleistocene marine sediments.

Pleistocene glaciations may have been triggered by the continents approaching their present configuration with the south
pole over a landmass and the north pole situated in a shallow inland sea. The restricted heat transfer encouraged a much
more zoned world climate. Glacial and interstadial episodes were then controlled by the diversion of bottom waters away
from the Arctic Sea by features such as the ridge between Greenland and Scotland.

Paleoclimate models highlight the potential importance of continental and oceanic evolution to the extent that they modify
large-scale oceanic circulation. As our understanding of the relationship between this "oceanic conveyor belt" of Broecker
and global climate grows, so does our appreciation of the complex interplay among tectonic, oceanographic and
atmospheric factors.

Measuring Sea Level

M odern sea level is measured by tide gauges, and changes over more than a decade are used to generate local models.
To determine the present mean sea level, hourly measurements of tidal height are taken at land-based stations using
benchmarks and recording tide gauges. In the United States, the values for all stages of the tide are averaged over a
nineteen-year period and used to publish tidal atlases that predict the daily tidal height at coastal sites throughout North
and South America and the Caribbean Sea. Tide-gauge records are only useful for sites that are within a few hundred
miles of each other; even then, local corrections must be made to allow for modifications to the tidal wave caused by
things like restrictions near tidal inlets and broad salt marshes. This limits the determination to a regional framework, and
does not take into account either local vertical motions or larger-scale differences in other basins. The resultant change in
sea level is therefore a relative direction and rate of change, which may vary from region to region. The effect of these
variations can be highly significant in terms of local patterns of coastal erosion. . In fact, fifty years is the minimum sea
level record length that should
be considered in an analysis of
global sea level rise. More
recently, data from Earth-
orbiting satellites such as
TOPEX/Poseidon (T/P) are
being used to measure global
sea level rise.

Studies of glacial-eustatic
changes in sea level during the
Pleistocene and Holocene are
based on the recovery of in-situ
flora and fauna that have a
restricted depth range over which they occur. Using their vertical position and radiometric age, researchers can reconstruct
the sea-level history of an area if enough in-place samples can be found spanning a broad enough period of time. Shells,
wood and other fossils that have been bored or infilled by chemical precipitates and other contaminants cannot be used for

5
dating. The resulting age will be an average that reflects both the original age of the fossil and the age of any
contaminants.

The limits to modern sea level determinations and use of regional frameworks also applies to geologically past sea level
determinations. No one determination should be given worldwide application.

Finding Suitable Material

M aterials commonly used to obtain radiometric dates are the shells of marine organisms. Crassostrea virginica and
Rangia cuneata are often used for dating because the two animals live in a limited depth range (3-4 meters) within
salt marshes and estuaries. Mangrove peat is deposited in swamps and provides another material for dating if tidal range is
limited and the peats are of the short-rooted variety. Similarly, oysters (Isognomon alatus) attached to the rock lived near
mean sea level and could be
dated to accurately determine
when local sea level was at that
elevation. The coral Porites often
forms a reef flat close to sea
level.

Dating of similar skeletal

material in fossil reefs can give a
fairly reliable measurement of
sea level at the time the reef was
active. In the Atlantic Ocean, the
branching staghorn coral
(Acropora palmata) occurs in
water depths between the 0 and
5-7 meters. A number of
researchers have constructed local sea level curves by assuming that the shallowest occurrence of this coral in a Holocene
reef represents sea level at the time the reef was alive. While this establishes a curve, which is less precise than some
based on other organisms, it is relatively easy in outcrop to determine if these corals are in place. It is generally agreed
that, whatever method is used, a local sea-level curve is essential to understanding the relationship of sea level to coastal
erosion and deposition.

The Dating Method

T he age of a fossil sample is determined using the known decay rate of a radioactive element and its transformation
into a stable counterpart. For example, the ratio of radioactive carbon (C14) to stable carbon (C12) is known for living
organisms today. That ratio remains constant until the organism dies, at which time carbon intake ceases and the C14 starts
to "decay" or alter to more stable C12. Using the known rate of decay, we can accurately measure the ratio of C14 to C12 at
any time after the organism died. A shell, piece of coral or similar material is ground up and the C14/C12 ratio is
determined using a mass spectrometer. This reveals the radiocarbon age of the dated material.

Radiocarbon dates are considered reliable back to 15,000 to 20,000 yBP, which covers the Holocene. Generally, C14 dates
older than about 12-18,000 years are suspect and a radiometric series must be used that has a longer half-life. The most
popular mid-range series tracks the decay of Uranium to Thorium. These dates are reliable past 125,000 yBP and are
thought to be accurate within 1% of the absolute age (slightly better than for radiocarbon dates). Furthermore, Uranium-
Thorium dates are less susceptible to metabolic and local seawater effects as the skeleton is being produced. Their primary
drawback is that these dates are two to three times more expensive than radiocarbon age determination.

6
Indirect Methods of Measurement

P ast glacial events have been studied using marine terraces that are exposed above sea level. These can be caused by
either erosional or depositional processes. Erosional features such as a bench or a nick point can be cut into limestone
or even harder cliffs. This type of feature is victim to a two-fold problem. First, the exposure of terraces above sea level
often reflects significant tectonic uplift. Therefore, local changes in the elevation of the feature must be quantified before a
reliable eustatic sea-level curve can be constructed. Second, the age of any material that is recovered reflects the time
when the strata were deposited and not necessarily when they were eroded and the terrace was formed. It must also be
remembered that the very process that
created the terraces (coastal erosion
during a stillstand of sea level) often
removes the datable material which is
so important in determining the exact
age of the feature.

The amount of data collected and the

effort that has gone into developing a
chronology of the last 25 to 30
thousand years of sea level are
surpassed only by the controversy over
their interpretation. The use of
radiometric dating is reasonably
straightforward and widely accepted.
However, arguments over
interpretations of paleo-depth, whether or not a sample is actually in place, the importance of local or regional tectonic
movement, and the likelihood of sample contamination combine to keep scientific skepticism over one sea-level curve
versus another at a healthy level.

Despite all the problems, useful and reasonably accurate sea-level curves often emerge. However, the farther back in time
we look, the greater the number of problems that must be faced and the more imprecise our numbers become. In this case,
several tools are available that can help us to understand general patterns of sea-level change in the geologic record.
General cycles of glaciation can be determined using paleostratigraphy (fossils of a known age and depth of occurrence
that are deposited in a sequence) and oxygen isotope ratios (O18:O16) from deep-sea cores. Oxygen isotope curves directly
reflect changing polar ice volumes and, therefore, sea level. The analysis of Pleistocene climate using planktonic
foraminifera has led to development of a detailed chronology of the past million years. Paleo-temperature was inferred
from large numbers of cores in which the relative number of warm and cold-water species was tabulated. The parallel
development of a paleomagnetic record allowed the oxygen isotopic and foraminiferal record to be tied to a known time
scale.

Seismic stratigraphy is a methodology used for dating older sealevels. Large numbers of seismic reflection profiles have
been examined to identify depositional packages that are "typical" under conditions of rising, falling or stable sea level.
What results is a suite of seismic "signatures," each related to a specific sea-level scenario. The units cover much broader
time spans than fossil dating

As an example, sediment delivered to any broad shelf during a sea level highstand will reach the ocean at some distance
from the shelf break ( scenario 1 and 3 ). Along the U.S. Atlantic coast, this material either aggrades or is trapped in the
upper portions of broad submarine canyons that traverse the shelf. Sedimentation in the adjacent basin is greatly reduced
and is limited to planktic settling.

As sea level falls below the shelf break (scenario 2), sediment is delivered directly to the continental slope and
sedimentation in the adjacent basin increases dramatically. Additional sediment is provided by erosion of newly exposed
shelf sediments. A situation intermediate between the two extremes may occur if sea level is lowered, but does not reach
the shelf break (scenario 4). In this instance, any major unconformity caused by erosion of the shelf will occur in a
submarine environment, in contrast to subaerial erosion during shelf exposure.

7
Each of these scenarios will produce a
recognizable depositional package across
the shelf-to-basin transition. These
patterns of onlap and offlap, aggradation
and bypassing provide evidence of both
the direction and magnitude of sea-level
change at the time of deposition. If
sediment-stacking patterns in the margins
of different ocean basins are similar, then
the patterns are assumed to result from
global sea-level variation. In areas where
the depositional signatures vary
significantly from the global "average,"
the record can provide valuable
information about local tectonic
movements that are responsible for that
divergence.

Hopefully these interpretations are backed

up by supporting information such as
cores, outcrops or biostratigraphic data.
This is not always possible, and many of
the models that have been developed have
given birth to a whole new arena of
scientific debate. Some scientists argue
that the stratigraphic boundaries identified
in seismic records do not necessarily
represent a single time boundary, as is the
case with biostratigraphic boundaries. In
addition, principles developed in
siliciclastic systems may be inapplicable to carbonate margins since sediment is derived not from rivers but from
biological production by marine organisms.

As a carbonate bank is flooded by rising sea level, production by the "carbonate factory" increases dramatically and waves
and currents sweep this material off the bank and into deeper water. Thus, a rise in sea level will increase sediment
introduction to the adjacent basin, in direct contrast to the model described above for the siliciclastic Atlantic shelf. The
sedimentary record could, therefore, be very different between the U.S. Atlantic coast and the adjacent Bahama Banks.
This pattern is further complicated by the fact that windward and leeward margins in the Bahamas have been shown to
behave differently in response to the same change in sea level.

The lesson here is not that we should abandon seismic stratigraphy. To the contrary, this simply challenges us to
understand the nature of the processes that produce a particular seismic signature in one environment or the other. When
used in conjunction with other traditional methods, seismic stratigraphy significantly increases our ability to recognize
major patterns of sea-level rise and fall in the large-scale geologic record.

Patterns of Sea Level Change

T he effects of sea-level change are generally seen across the continental shelf and upper slope. They are exposed to
subaerial processes as sea level falls. The ensuing marine invasion renews the sedimentary regime that was
abandoned after the previous high stand.

8
Mesozoic and Tertiary Sea Level Change

S eismic-stratigraphic studies at Exxon led to proposal of a global sea-level history for Triassic through present time.
From the Triassic to the present, they identified more than 100 global sea-level changes.

While agreement on the magnitude or timing of the older sea level changes is far from universal, the various generations
of the sea-level curves provide a useful point of reference. These curves group sea-level variations into cycles that are
reflected in the seismic records of the world's shelf
margins. First-order cycles coincide with times of major
continental plate break-up and seem to last 200 to 300
million years. More frequent and irregular transgressive
and regressive events are superimposed on these longer
cycles.

The last time the world ocean reached its present level was
in the lower Triassic (ca. 250 million yBP). Midway
through this period (lower Cretaceous), sea level reached a
highstand roughly 100 meters below today's ocean. At
about this time, the new and relatively shallow Atlantic
Ocean was opening, and gradually replacing a deep Pacific
Ocean. Using spreading-rate data for the mid-Atlantic over
the past 180 m.y. the Cretaceous high sea-level stand (ca.
125 MyBP) and the subsequent lowstand (95 MyBP) could
be explained solely by expansion and contraction of the
mid oceanic ridge system. It should be kept in mind,
however, this was also a time of rapid warming in what
has been described as a "greenhouse climate" similar to
that of today. The relative importance of glacio-eustatic
versus tectonic controls remains a topic of debate.

Since the mid-Cretaceous lowstand, sea level oscillations

have continued with a rough periodicity of 100,000 years
superimposed on a gradual rise . The Oligocene highstand
was accompanied by a roughly continuous equatorial
seaway that spawned a cosmopolitan coral fauna in the
9
Caribbean region that has been in decline ever since. The curve hints at an increase in the frequency of sea-level
oscillations since the mid-Cretaceous, perhaps reflecting an increase in the relative importance of astronomic factors.

Pleistocene Sea Levels

P leistocene glacial and interglacial events were first described from Alpine deposits and a chronology of glaciations
was established in which four glacial events were named for the southern extent of continental glaciers:

• in Europe, the Gunz, Mindel, Riss and Wurm intervals;

• in North America, the Nebraskan, Kansan, Illinoian and Wisconsinian glaciations.

The Wisconsin and Illinoian glacial stages are still accepted and widely used for glacial stratigraphy and correlation in
North America and elsewhere. However, the Kansan and Nebraskan are not longer considered valid glacial stages in light
of much evidence for more complicated pre-Illinoian glaciation of the central United States.

Glacial ice may have began forming as early as the Oligocene but certainly by Pliocene time. The sea level curves of
glacial formation and melting show a downward trend in volume of seawater available as more water became locked in
continental glaciers. The concept of four glacial events has been revised as studies of deep-sea sediments disclosed a long
succession of glaciations during the Pleistocene; the most recent of which was shown by radiocarbon dating to equate
with deposits of the last glaciation on the continents. Long Atlantic cores of deep-sea sediments recorded at least nine
Pleistocene ice ages. This led to a long series of investigations of Pleistocene sea level history. Paleomagnetic dating of
deep-sea cores shows that about 30 glacial events occurred during the Pleistocene.

Three different approaches are used in the investigation of Pleistocene

glaciations:

• studies of deep sea microfossil assemblages;

• measurement of the ratio of oxygen18 to oxygen16 in fossil shells;
and
• study of uplifted coral reef terraces on tropical islands.

The studies by different investigators were merged into a large project,

CLIMAP, which has given us a reasonable reconstruction of the Pleistocene
history, and causes of Pleistocene glaciations. The history of this project
was described in Imbrie and Imbrie which is the basic reference for the
following discussion.

Ericson, et al. used the abundance of selected indicator species that they
considered to be sensitive to changes in climate to analyze the Pleistocene
and to determine the climatic history of the Pleistocene. Geochemical
analyses and the absence of an indicator species, Globorotalia menardii,
established the 11,000 yBP boundary that we now call the Holocene. They
also established what they considered the boundary of the onset of the
Pleistocene, marked by the extinction of discoasters, at about 1.5 MyBP.
However, in the same cores, the differences from the results by Emiliani
were significant. Emiliani used the technology for measuring oxygen
isotope ratios (O18:O16) in foraminifera, which was assumed to be an
indicator of paleotemperatures. Ericson's and Emiliani's methods agreed
over the more recent part of the Pleistocene record, but over the older parts,
the results did not agree.

Broecker and van Donk conducted an isotopic analysis of the same core
which allowed the results of Emiliani, Ericson and Imbrie to be compared.
Some environmental factor other than surface water temperature caused
Globorotalia menardii to appear and disappear cyclically in deep water of
10
the Atlantic Ocean, so Ericson was not measuring water temperature. The results led to the conclusion that much of the
isotopic variation noted by Emiliani must be due to changes in the volume of the ice sheets -- not to changes in
temperature.

When an ice age begins, the continental ice sheets grow, removing water and the ocean water left behind becomes
enriched in oxygen18. The calcium carbonate shells of marine organisms reflect the water's isotopic composition at the
time that they were formed. The higher the ratio of O16 to O18 in shells, the more land ice was present when the sediment
was laid down.

Examination of uplifted coral reef terraces instead of cores during the same period of time was yielding data about the last
200,000 years of Pleistocene history with three dates for high sea levels

• the present starting 5,000 to 7,000 yBP,

• 80,000 yBP and
• 120,000 years ago

from reef terraces in Florida and Hawaii. Accretionary terraces in Barbados represented reef growth at a former sea level.
Two terraces were dated at 80,000 and 125,000 years, with a third between the others dated at 105,000 yBP. The 105,000-
year peak did not fit the Milankovitch 65o insolation curve.

A 45o N insolation curve in which precession is given more weight than tilt shows:

• the warm peak at 50 KyBP (from the 65o N curve) was removed
• a new peak appeared at 106 KyBP

The results clearly indicated that the last four high sea stands (122, 103, 82, and 5 KyBP), correspond closely in time to
the last four prominent warm peaks (127, 106, 82, and 11 KyBP) in the modified curve of summer insolation, not only in
chronology but also in magnitude.

The matching of the coral terraces of Barbados, New Guinea, Australia and Hawaii was one of the convincing arguments
in support of the Milankovitch theory. Marine terraces in Haiti, Puerto Rico, and Jamaica, have elevations and ages that
agree with the deep-sea record.

The major cycle in the isotopic record was 100,000 years and this primary climatic cycle has an asymmetrical, saw-
toothed shape in which glacial expansions were terminated by rapid warming. This pattern suggests that climate
responded to some continual forcing oscillation, with the most obvious being variation in the amount of sun energy
received by the earth coupled with positioning of the continents.

Holocene Sea Level Rise

T he Holocene includes the last 11,000 years, since the end of the last major glaciation. There is some debate over the
low point of the previous Wisconsinian sea level. Some early studies off the New England coast placed the lowstand
near 130 meters. Other investigations that take into account the local effects of glacial loading beneath the glacier and
compensatory uplift further away place that low closer to 85 m.

11
The sea-level curve is generally
characterized by a gradual rise in sea
level, although the exact character and
the continuity of that rise continue to be
debated. One group of researchers
proposes a rapid sea-level rise (8-10
mm/yr = 8-10 m/thousand years) from
12,000 to about 6,000 yBP, followed by
a decrease in the rate of rise to 1-1.5
mm/yr between 6,000 yBP to 3,000
yBP and slower yet thereafter.

Another group suggests that the

Holocene sea level was more ragged
and either stopped, slowed dramatically
or even dropped to form marine terraces
at levels 40, 20, 15 and 9 meters below
present sea level. Logically, the rates of
rise during the time intervals between
successive stillstands had to be much
faster than those calculated for a
continuous rise.

All of these curves had little data on sea

level before about 10,000 yBP.
Fairbanks filled this gap with a curve
based on core data from Acropora
palmata in deeply submerged reef terraces off Barbados. The curve contains significant gaps in the depositional record
from terrace to terrace which he interpreted to reflect sudden rises in sea level due to the rapid addition of meltwater into
the world's oceans. Central to Fairbanks' interpretation is the idea that the terraces associated with the Barbados reefs were
actually formed by the development of Acropora palmata reefs on a smooth slope as rising sea level suddenly stopped;
this was followed by a sudden rise
which left the reefs behind. An
alternative possibility is that the reefs
formed on older terraces and that each
reef represents the accretion of coral as
sea level passed by a pre-existing notch
in the otherwise smooth island slope. In
this second scenario, the gaps in the
record would be the result of reefs not
being able to accrete along steep
intervening slopes.

One alternative supports a smooth sea

level rise, while the other points to
pulses in the addition of meltwater to
the world oceans. While the picture is
far from a clear and universally
accepted, there seems to be wider
acceptance of the constant rate of sea-
level rise. Neither concept supports a
highstand of sea level during the past 3,000 - 6,000 years.

The important point is that sea level has generally risen over the past 18,000 years from a lowstand somewhere 100 m
below today's oceans.

12
Global Warming and Present Sea Level Change

R egional studies of modern sea level are generally consistent in their overall picture, but not in the details. Much of the
disagreement is related to the local factors described above. While scientists might debate the specifics, they are in
universal agreement about one thing - both global temperature and sea level continue to rise today, the latter at a rate of
approximately 1-3 mm/year on a global scale similar to that for the late Holocene .

Even this modest rate of rise has had a significant effect on coastal and nearshore areas when private homes and public
facilities have been built close to the water's edge.

More problematic is the question of how much of this rise is part of a larger, natural cycle and how much is the result of
man-induced changes in the planet. Temperature records over the past century show a warming trend . Furthermore, a
strong correlation has been demonstrated between this pattern and anthropogenic activities such as the burning of fossil
fuels, increasing nitrous oxide and methane emissions from factories and automobiles, the use of chlorofluorocarbon-
based aerosol propellants and other activities related to the progressive industrialization in our modern society.

The September 4, 1990 Time Magazine presented new data and a new view of global warming together with figures from
NOAA research. Since 1980 temperatures have climbed in the Arctic faster than the rest of the earth, and climatic changes

13
are expected to move more rapidly in the polar region. Warming in the Arctic is a factor in the global climate system
because the difference in temperatures between the tropics and the poles drives the system.

In 2006, the National Academy of Sciences gave complete support to the concept that global warming is occurring.

If the poles continue to warm faster than the tropics, the circulatory system may diminish as the pull created by the
sinking NADW decreases. Concurrent with this the movement of warm water in the Gulf Stream could slow or stall
reducing temperatures in North America and Europe. The melting 11,000 years ago sent freshwater from the St. Lawrence
River into the North Atlantic to create the 1,300-year Younger Dryas. It has been suggested that this new Arctic melting
could follow the same pattern. A significant result of such a change is that during the Younger Dryas the monsoon
weakened in Asia and the Sahara expanded - a change in precipitation accompanied the climate event and this could have
dramatic effects.

Large sums of money have been invested in research to model the possible changes in global climate over the coming
decades. The magnitude of the proposed changes in these "Global Climate Models" (GCM'S) varies considerably
depending on the assumptions that are made and the particular model that is used. Regardless of the details, however, the
existing record shows a marked increase in the rate of change over recent decades. And whichever model is used, a less
than optimistic prognosis is associated with a scenario in which "business as usual" continues. As a best guess, global
temperature would rise by 3° C over the next century with a worst-case estimate of 5° C. If the worst-case scenario proves
true, then the world ocean could rise by a meter in the next century, at a rate tenfold that of the past 3,000 years. The
impact on the world's weather patterns would have staggering impacts on business, and especially agriculture. Low-lying
areas would be flooded at a rate too high for even the US Army Corps of Engineers to keep up with. Coral reefs might be
left behind, exposing now-protected port areas to increasing wave energy. Combined with recent estimates that hurricane
wind speeds could increase by 7-15 mph with a 2.2° C increase in global temperature, this portends for dramatic coastal-
engineering problems. Because many important fisheries are related to reefs, implications to that industry are profound.

Summary
14
T he change of sea level leaves a record in the distribution of sedimentary facies, the areal extent of marine deposits,
and the rate of sediment accumulation. While none of these are uniquely related to sea level variations, changes in the
world ocean play a very important role. The pattern of sea-level change at any on site is an amalgam of global eustatic
change and local tectonic movement. The resulting sedimentary record is controlled by the relationship between local sea-
level history, sediment supply and physical-oceanographic processes.

Changes in sea level have occurred many times during the geological history of the earth. Pre-historic sea-level changes
are studied by analyzing the sedimentary patterns of deposition from outcrops, cores, seismic records and a host of
indirect geochemical methods. The more recent and shorter-term Pleistocene and Holocene sea-level changes have been
driven primarily by patterns of glaciation and deglaciation, which are in turn controlled by the eccentricity of the earth's
orbit around the Sun and changes in the tilt of the earth's axis. Longer-term sea-level fluctuations are additionally
controlled by global tectonic patterns that have changed as the planet evolved.

Today's ocean margins have been dramatically affected by the most-recent Holocene rise in sea level. Coastlines, reefs,
shelf deposits and their connection to adjacent basins all reflect the patterns of sea-level rise in conjunction with both
larger- and smaller-scale controls. Today's short-term patterns may be impacted more by man-induced changes in the
earth's atmosphere than by any other single factor mentioned above. The understanding of natural controls over past sea-
level change is paramount in understanding the geologic development of the planet. Using these to appreciate the extent to
which we may be responsible for modifying that environment in the near future could have a profound effect on how well
Homo sapiens fares with respect to the longevity of other faunal groups in the geologic record.

15