Odekon M. Encyclopedia of paleoclimatology and ancient environments
Подождите немного. Документ загружается.
the depth of deposition relative to sea level is determinable.
Terrestrial deposits within a sequence, such as peats and flood-
plain sediments, may usefully indicate levels not reached by the
sea. Various methods are used to establish the ages. For depos-
its less than 40,000 years old, radiocarbon dating is used
widely, although uranium-series dating is preferable in the case
of coral formations and has a time range that extends to 0.5
Myr. Thermoluminescence (TL) and optically-stimulated lumi-
nescence (OSL) methods are increasingly used for dating
Upper Quaternary shoreline deposits, and amino acid racemiza-
tion has proven to be useful despite its low precision. Sea level
indicator deposits also are correlated to marine oxygen isotope
records, described later, that have a chronology based on slow
variations in the Earth’ s orbit, which affect the seasonal
receipts of solar radiation and acted as an ice-age pacemaker.
Orbital chronology rests on astronomical observations and has
been extrapolated several million years into the past. Wherever
possible, the dating of older Quaternary and pre-Quaternary
deposits is tied to geomagnetic reversal chronology, which in
turn is secured by potassium-argon dating methods.
Separat ion of eust atic, isostati c and tectonic
compon ents
Raw separat ion of uniform uplift or subsid ence
A first step towards separation of the tectonic, isostatic and
eustatic contributions is subtraction of any obvious vertical
crustal movements from a relative sea level curve. More often
than not, this is done by assuming that the sea level for some
reference deposit in the record is known and that the rate of uplift
(or subsidence) was constant throughout the duration of the
record. With these assumptions, the local sea level S represented
by a deposit of age t t hat ac cu mu la t ed a t d ep th d be lo w s ea le vel
and now is height H above present sea level, is given by
S ¼ H þ d Ut ; with U ¼ðS
r
H
r
þ d
r
Þ= t
r
ð1 Þ
where H
r
, d
r
and t
r
are height, depositional depth and age of
the reference deposit, and S
r
is the sea level at its time of for-
mation. For Upper Quaternary studies, the local height of the
Last Interglacial shoreline is widely used to determine uplift
rate U, as evidence at tectonically stable sites indicates that
the Last Interglacial sea level was 5 2 m above present level,
120,000–127,000 years ago (Chappell et al., 1996). However,
for every study, each variable in Equation 1 has an error term
arising from dating errors and uncertainties in field relation-
ships. Moreover, in assuming a constant rate of uplift, this
approach neglects reversing vertical movements that arise from
the global glacio-hydro-isostatic response to advancing and
retreating ice sheets. Figure S13b shows the “uplift-free” sea
level curve derived by this method from the coral terraces
illustrated in Figure S15.
Similar principles are used to derive sea level changes
from sedimentary basin sequences, although here the vertical
Figure S13 Sea level variation at two time scales. (a) 10
8
years, inferred from seismic sequence stratigraphy (Haq et al., 1987; Hallam, 1992). The
higher frequency changes reflect both global and local effects; the large, slow changes reflect continental breakup and changes of ocean ridge
systems. (b) 10
5
years of relative sea level at Huon Peninsula (HP), Papua New Guinea, driven by changes in northern continental ice sheets. Bars
show marine oxygen isotope stages (OIS) discussed in text (from Lambeck and Chappell, 2001).
894 SEA LEVEL CHANGE, QUATERNARY
Figure S14 Observed variability of sea level change. Graphs a–c show sea level over the last 10,000 years relative to A
˚
ngerman, Gulf of Bothnia,
Sweden; Richmond Gulf, Hudson Bay, Canada, and southwest England. Sites A and B were formerly covered by ice sheets and the fall of relative sea
level largely reflects isostatic uplift of the land since the ice melted, whereas southwest England lay outside the ice-covered region and rising sea
level in this area largely reflects the influx of ice meltwater to the oceans. The counteracting effects of isostatic uplift and ocean increase are
reflected in complex sea level changes over the last 20,000 years at Andøya in Norway (d), whereas the ocean increase dominates at sites remote
from the former ice sheets, such as Barbados (e) (data sources given in Lambeck and Chappell, 2001).
Figure S15 Typical expression of Late Quaternary sea level changes superimposed on tectonically rising terrain: coral terraces Huon Peninsula,
Papua New Guinea. The top terrace at right of the inset photo is 350 m a.s.l. and formed at the beginning of the Last Interglacial, 127,000 years
ago. Each of the smaller reef structures within the downstepping stratigraphic sequence was formed during a sea level oscillation within the last
glacial cycle. Subtracting the effect of uplift yields the sea level curve shown in Figure S13b.
SEA LEVEL CHANGE, QUATERNARY 895
movement is downwards. In terms of sedimentary facies, indi-
vidual cyclothems often are very similar from bottom to top of
a thick cyclothemic sequence ( Figure S16), implying fairly uni-
form subsidence of the basin.
Glacio -hydro- isosta sy
When ice sheets melt, to a first approximation the sea level
rises by an amount D z
e
( t) related to the land-based ice volume
V
i
(using the notation of K. Lambeck: see Lambeck and
Chappell, 2001),
Dz
e
ðt Þ¼ðr
i
=r
o
Þ
ð
ð1 =A
o
ð t Þd V
i
=d t Þdt ð2 Þ
where A
o
( t) is the ocean surface area (which changes as sea
level rises or falls) and r
i
, r
o
are the average densities of ice
and ocean water, respectively. D z
e
( t ) is the ice-volume equiva-
lent sea level change (or equivalent sea level change), which
equals eustatic change if no other factors contribute to changes
in ocean volume. The relative sea level change Dz
rsl
( j , t )at
position and time t, ignoring tectonic displacements, is
Dz
rsl
ð j; t Þ¼Dz
e
ðt Þþ Dz
i
ð j; t ÞþD z
w
ð j; t Þð3 Þ
where D z
i
and Dz
w
are the glacio- and hydro-isostatic contribu-
tions. Both are functions of position and time. The water depth
or terrain height, expressed relative to coeval sea level, is
hð j; t Þ¼hð j; t
0
ÞD z
rsl
ðj; t Þð4 Þ
where h( t
0
) is the present-day ( t
0
) bathymetry or topography
at j . Both isostatic terms in Equation 3 are functions of Earth
rheology as well as of fluctuations in the ice sheets over time.
In formerly gla c iat e d a reas , the glac io-isostatic term
Dz
i
( j, t)
i
dominates during and after deglaciation, and leads
to uplift at a rate that can exceed the global eustatic rise, so that
sea level locally falls (the Ångerman result: Figure S14).
Rebound is smaller near the ice margin and although it may
dominate initially, the global sea level rise becomes important
later. When all melting has ceased, the residual rebound leads
to falling local sea level (the Andøya result: Figure S14).
During ice sheet growth, mantle material beneath the loaded
area is displaced outward and broad bulges develop around
the perimeter, which subside when the ice melts, leading to
local sea level that slowly rises after melting has ceased. Much
further from the ice, the water load becomes the dominant
cause of planetary deformation (the hydro-isostatic contribution
Dz
w
( j, t)), producing subsidence of the sea floor and adjacent
margins. The effect is most pronounced at continental margins
far from the ice sheets, such as the Australian coast, and once
melting has ceased, sea levels continue to fall at a slow but
perceptible rate. The amplitude of this postglacial “highstand ”
effect can vary by several meters from site to site.
Isostatic corrections and a global eustatic curve can be
derived from local sea level curves, using a rheologically appro-
priate earth model to predict surface deformation in response to
changing ice and water loads, with the ocean surface remaining
a gravitational equipotential surface at all times. The sea level
signal at sites far from the former ice margins approximates
the equivalent sea level function to about 10–15% and the iso-
static contribution is mainly from water loading, which is
insensitive to the details of the ice sheets, provided that the
total ice volumes are correct to within about 10–20%. Hence,
through an iterative procedure, it becomes possible to estimate
changes in ice volumes V
i
from observed sea level changes
Dz
rsl (obs)
, using Equations 2 and 5:
Dz
rsl
¼ Dz
rslðobsÞ
ðDz
i
þ Dz
w
Þð5Þ
The use of local sea level curves from widely separated
places allows earth rheology parameters and models of ice
distribution to be evaluated. Recent models include deforma-
tion of the basins over time, movement of grounded ice across
the shelves, modification of sea level by the time-dependent
gravitational attraction between the solid earth, ocean, and
ice, and the effect of glacially induced changes in earth rotation
on sea level.
Sea level through the Last Glacial Cycle
Sea level data for the last glacial cycle are more plentiful
than for earlier periods and at Huon Peninsula, Papua New
Guinea provide a near-complete relative sea level curve
(Figure S13b), which has been used for reconstructing ice-
equivalent sea level for the past 140,000 years (Lambeck and
Chappell, 2001). Results indicate that ice melting has varied
since the Last Glacial Maximum, with two periods of rapid
sea level rise from 16,000 to 12,500 and from 11,500 to
8,000 years ago, separated by the Younger Dryas cold episode
when sea level seems to have risen less rapidly. By 7,000 years
ago, the northern ice sheets except for Greenland had gone and
ocean volume approached its present level, but Antarctic ice
melting may have since contributed a few meters of equivalent
sea level.
The Last Interglacial, when sea level was similar to the pre-
sent, ended about 118,000 years ago with rapidly falling sea
level, associated with the growth of northern continental ice.
Cyclic sea level changes from 118,000 to 60,000 years ago
reflect the effect of the 20,000-year orbital precession cycle
on the ice sheets, although other fluctuations also appear. These
occur repeatedly about every 6,000 years from 60,000 to
30,000 years ago, with amplitudes of 10–15 m, and each rise
apparently coincides with a major episode of ice-rafted sedi-
ment deposition recorded in the North Atlantic, suggesting that
the rise was caused by a large, rapid discharge of continental
ice (Chappell, 2002).
Oxygen isotopes and long sea level records
Long, continuous records of oxygen isotopes in calcareous
foraminifera preserved in deep-sea sediments have become
standard records of Quaternary sea level and temperature
changes. The oxygen isotope ratio in foraminifera, conven-
tionally expressed as d
18
O (th e % difference of
18
O/
16
Oina
sample from an international standard), depends on the tem-
perature and isotope ratio of the seawater in which they lived.
Furthermore, the seawater isotopic ratio is related to the size
of polar ice sheets, because
18
O is preferentially removed from
atmospheric water vapor as it makes its way poleward, caus-
ing the icecaps to be depleted in
18
O rela tive to seawater
by 25–55%, varying with atmosp heric temperature. Thus,
foraminiferal isotopes Dd
18
O
f
respond to ice volume changes
DV
i
according to
Dd
18
O
f
d
18
O
i
DV
i
=V
w
þ C
T
DT þ e ð6Þ
where DV
i
is relative to present day V
i
, V
w
is the present volume
of seawater (mean d
18
Oofmodernseawater=0%
SMOW
), C
T
896 SEA LEVEL CHANGE, QUATERNARY
Figure S16 Late Pliocene sea level changes inferred from shallow marine cyclothems in New Zealand. Centre columns show 12 sedimentary
cycles; left column shows cyclic variations of water depth at the site of sedimentation, inferred from fossil marine invertebrates.
Correlations to global marine oxygen isotope cycles (Shackleton et al., 1995) are shown to the right, tied to magnetic reversal chronology
(from Pillans et al., 1998).
SEA LEVEL CHANGE, QUATERNARY 897
is the coefficient of temperature-dependent isotope fractionation
for calcite (– 0.23%
C
– 1
), D T is temperature change, and e
represents any local change of seawater d
18
O not related to DV
i
.
(Equation 6 is approximate because the mean isotopic com-
position of the ice sheets d
18
O
i
is assumed constant, but the
uncertainty here probably is smaller than the effects of ice
volume and temperature). Finally, the first term on the right hand
side of Equation 6 can be expressed in terms of equivalent
sea level Dz
e
: comparison of isotopes and sea levels for the last
glacial cycle indicates that ▵d
18
O
w
/ D z
e
–0.009 % m
– 1
.
The numerical value of d
18
O
f
becomes increasingly positive
under a fall of both sea level and temperature. Thus, the isoto-
pic affect of temperature and ice-volume changes are additive
and the two factors can be decoupled if one or the other is
known. After removing the temperature effect, isotopic sea
levels for the last glacial cycle derived by Shackleton ( 1987)
co mp ar e f av or ab ly w it h Figure S13b. Comparisons further back
in time are also encouraging: for example, Figure S14 il lu str ate s
typically close correspondence between sea level and isotopic
cycles in Late Pliocene times. Thus, oxygen isotopes in continu-
ous deep-sea cores provide proxy records of equivalent sea level
(ice volume) changes through and beyond the Quaternary that
reveal both progressive global cooling and the onset of ice-
driven sea level cycles in the Pliocene (Figure S17). As the his-
tory of ocean temperature becomes increasingly well determined,
through trace-element analysis of microfossils and other techni-
ques, the sea level contribution in long isotope records is now
being isolated (see Zachos et al., 2001).
Shore line reconstr uction s and geomorp holog ic effect s
Once a global eustatic curve z
e
( t) is established, the course of
shoreline changes through time can be predicted. Provided that
the present-day shallow water bathymetry is known with high
resolution, water depths for any region at any time within
the range of the eustatic curve follow from Equation 4,
and the paleo-shorelines at time t correspond to the contours
h(j, t) = 0. Thus, it becomes possible to examine the migrations
of shorelines through time for intervals for which sufficient
observational data exist to constrain the isostatic variables.
Predictions of shorelines since the time of the Last Glacial
Maximum have been published for both global and regional
reconstructions, which can provide useful insights into the inter-
pretation of prehistoric sites. As Lambeck (1996) has shown, for
example, the interpretation of post-Palaeolithic archaeology of
the Aegean is intimately linked to reconstructions of shoreline
changes, which were a powerful factor in trade and changing
human activities in the region.
Rising sea level since the Last Glacial Maximum led to
widespread changes in coastal regions. Under the influence of
ice-age low sea levels, today’s coastal valleys tended to
become incised well below present sea level and then became
traps for floodplain aggradation when sea level rose, as ice
sheets retreated. The alternation of incision and aggradation
doubtless was repeated in each glacial cycle throughout the
Quaternary, leading to development of broad floodplains on
thick sediment valley-fills, which in tectonically stable regions
often extend hundreds of kilometers inland. Near-coastal lime-
stone regions under the influence of low sea levels developed
vadose karst systems below present sea level, including stream
passages and speleothem formations. At coastlines, a range of
landforms from drowned valleys through sand-barrier basins
to estuarine and deltaic plains were the result of the post-glacial
sea level rise, the particular form depending on sediment sup-
ply, tidal range and wave energy. In tropical seas, coral reefs
were re-established after being reduced by subaerial erosion
during glacial-age lowstands.
In tectonically stable terrain, the geomorphic expression of
any given Quaternary sea level cycle tends to overprint the
record of earlier cycles. In contrast, tectonic uplift results in
flights of coastal terraces built at times of high relative sea level
(e.g., Figure S15) that often pass inland into flights of river
terraces. However, even though a terraced river-valley may
grade to present sea level, not all its terraces necessarily relate
to sea level highstands. Where continental shelves are broad, a
river carrying sufficient fluvial sediment, when sea level was
low and the coast far seawards of its present position, may have
aggraded to a level higher than the recent floodplain, which
thus becomes inset below a Pleistocene lowstand terrace. Given
the tendency for all such features to become degraded and
overprinted, the geomorphic legacy of past sea levels – while
very sharp for recent events – becomes increasingly blurred
with time. From this fact springs the irony that, beyond the last
glacial cycle, Quaternary sea level changes have become better
determined from isotope records in deep sea sediments than
from ancient coastal features themselves.
John M. Chappell
Bibliography
Chappell, J., 2002. “Sea level changes forced ice breakouts in the last
glacial cycle: New results from coral terraces.” Quaternary Sci. Rev.,
21, 1229–1240.
Chappell, J., Omura, A., Esat, T., McCulloch, M., Pandolfi, J., Ota, Y., and
Pillans, B., 1996. Reconciliation of late Quaternary sea levels derived
Figure S17 Marine oxygen isotope variations for last 4 Myr indicate sea level changes associated with glacial cycles. The first cycle at the left
represents the general curve for the last glacial cycle, shown in detail in Figure S13b. The increase of amplitude in d
18
O cycles around 2.6 Myr
represents the onset of large fluctuations of ice sheets and sea level, which become even more pronounced around 1 Myr, when they adopt a
characteristic period of 100,000 years. The upper curve shows the changing amplitude of the 100,000 year cycle (after Zachos et al., 2001).
898 SEA LEVEL CHANGE, QUATERNARY
from coral terraces at Huon Peninsula with deep-sea oxygen isotope
records. Earth Planet. Sci. Lett., 141, 227–236.
Hallam, A.J., 1992. Phanerozoic Sea-level Changes. New York: Columbia
University Press.
Haq, B.U., Hardenbol, J., and Vail, P.R., 1987. “Chronology of fluctuating
sea levels since the Triassic.” Science, 235, 1156–1167.
Lambeck, K., 1996. “Sea-level change and shore-line evolution in Aegean
Greece since Upper Palaeolithic time.” Antiquity, 70, 588–611.
Lambeck, K., and Chappell, J., 2001. “Sea level change through the last
glacial cycle.” Science, 292, 679–686.
Pillans, B., Chappell, J., and Naish, T., 1998. “A review of the
Milankovitch beat: Template for Plio-Pleistocene sea-level changes
and sequence stratigraphy.” Sediment. Geol., 122,5–21.
Shackleton, N.J., 1987. “Oxygen isotopes, ice volume and sea level.”
Quaternary Sci. Rev., 6, 183–190.
Shackleton, N.J., Hall, M.A., and Pate, D., 1995. “Pliocene stable isotope
stratigraphy of ODP site 846,” in Proceedings of ODP, Science Results
138. College Station, TX: Ocean Drilling Program, pp. 337–353.
Zachos, J., Pagani, M., Sloan, L., Thomas, E., and Billups, K., 2001.
“Trends, rhythms, and aberrations in Global Climate 65 Ma to Present.”
Science, 292, 686–693.
Cross-references
Cyclic Sedimentation (Cyclothems)
Dating, Amino Acid Geochronology
Dating, Luminescence Techniques
Dating, Magnetostratigraphy
Glacial Eustasy
Glacial Isostasy
Oxygen Isotopes
Sea Level Change, Last 250 Million Years
Sea Level Change, Post-Glacial
SEA LEVEL INDICATORS
Indicators for sea level change encompass direct instrumental
evidence of recent changes in sea level at local and global
scales to proxy indicators for former sea level variations over
the past millennia. The evidence for sea level changes since
the late nineteenth century has been mostly derived from tide
gauges. However, our knowledge of the sea level record prior
to this has been drawn from sources as disparate as former
shoreline positions to geochemical and paleoecological data
reflecting past changes in local mean tide levels. Considering
the variety of indicators, differing widely in timescale, spatial
coverage, and accuracy, it is not surprising that they pose con-
siderable problems for the researcher attempting to reconstruct
and understand the sea level record.
Instrumental records
Instrumental records for sea level change first appeared in the
late seventeenth century in Western Europe. Until the advent
of satellites capable of measuring global changes in the ocean
surface, instrumental indicators for sea level change – the
precision and reliability of the instruments changing with
improvements in technology – actually recorded changes in
tidal levels over time; variations in mean sea level for an area
could be deduced from these data. Because such records are
site-specific, there has been considerable discussion concerning
how such an inherently local nature for sea level change relates
to global and even regional sea level trends. In some cases,
depending on how the instrument was deployed and monitored,
the tidal records derived – and the changes in mean sea level
inferred from them – may have very little to do with the
regional sea level history of an area. Historically, many of the
instruments for delineating changes in tidal levels were located
in ports to aid navigation. Later changes in harbor tidal hydro-
graphy from port construction and modification, and even the
sinking of docks upon which the instruments were placed
into soft, unconsolidated sediments, often introduced consider-
able errors. Methods (largely statistical) exist for highlighting
potential errors in instrument records. These can be augmented
by archival evidence for port construction and modification. In
particular, comparable instrumental records for sites elsewhere
in the region can be very useful in identifying discrepancies.
However, such methods may not always be able to pinpoint
smaller, but still cumulatively significant, errors due to human
activities. Generally, the more recent the instrumental records,
the more they are likely to be reliable and free of error – older
records should always be approached with caution.
The first instrumental records were derived from a very
simple instrument, the tide staff, or graduated staff or rod
anchored along a shoreline or dock. Unless monitored through-
out the course of the tidal cycle, most early tide staff records
were often records of changes in the mean high water mark,
the evidence being the highest wetted mark left on the staff.
Depending on whether the staff was protected from wave
activity, reading the mean high water level accurately as waves
rose and fell could be challenging, and making judgment of
actual water levels was difficult. Nonetheless, tide staff records
comprise the oldest instrumental records of sea level change
available; and in the case of the Amsterdam record (Mörner,
1973), extend back to the late seventeenth century, one of the
coldest periods of the Little Ice Age.
Since the late nineteenth century, modern tide gauges
(mareographs) have provided detailed evidence on tidal varia-
tions as indicators of sea level variation. Although problems
reflecting such effects as port construction and harbor dredging
on tidal hydrography (and the sea level signal) plague many
early tide gauge records, modern tide gauges – protected from
the “noise” of waves on the tidal signal by stilling wells – yield
high precision measurements of daily, monthly and annual tidal
variations at a locality. Increasingly automated and capable of
downloading tidal data by telemetry, tide gauges are the source
for much of what is known about twentieth century sea level
rise. They have also revealed considerable information on sea-
sonal (intra-annual), yearly (inter-annual) and longer-term (dec-
adal) fluctuations in regional and global sea level trends
(Douglas, 2001). Little progress would have been made in the
last two decades toward understanding the linkages between
global climatic fluctuations and sea level changes without tide
gauge data, though the worldwide rate of sea level rise docu-
mented by these records has proved remarkably amenable to
interpretation (see Douglas, 1991).
Perhaps the greatest limitations confronting the use of tide
gauge records as sea level indicators are the often-short lengths
of record available for many areas (especially for developing
countries) and the concentration of some of the longest and
most continuous archives for the Northern Hemisphere (espe-
cially, Northern Europe and Eastern North America). With
respect to length of record, 60 years is now recognized as the
minimum timescale for which reliable determination of average
rates of sea level rise may be made from tide gauge records
(Douglas, 2001). The latter problem concerns the lingering
effects of glacio-isostasy associated with the Laurentide (North
America) and Fenno-Scandinavian (Northern Europe) Ice sheets,
SEA LEVEL INDICATORS 899
which complicate the delineation of purely eustatic trends in
tidal records. Rheological models (e.g., Peltier, 2001) allow
for some correction (“detrending”; Douglas, 1991) of northern
hemisphere tide gauge records for the phenomena of postgla-
cial rebound and forebulge collapse (for the Atlantic Coast
of North America, the hingeline between the two is approxi-
mately in the region of the Hudson River estuary). Ultimately,
satellite altimetry (e.g., Topex/Poseidon), showing the ocean
surface model, should address the spatial limitations of tide
gauge data. A recent attempt has been made to reconcile
tide gauge data with the ocean surface changes observed from
the Topex/Poseidon satellite (Cabanes et al., 2001).
Shoreline features
Ancient shorelines above mean sea level have long been
viewed as one of the more persuasive indicators of sea level
change. In the last several decades, technological advances
like high-resolution seismic stratigraphy (in particular, bottom
profilers) have extended the search for evidence of former
shorelines to the inner continental shelves of many areas.
Nonetheless, though the potential suite of ancient shorelines
to be investigated is no longer limited to subaerial features,
controversies surrounding whether some fossil shorelines have
anything really to do with global sea level trends have not
abated. No one argues that former shorelines do not represent
a change in mean tide level locally for a sustained period (in
some cases, even this argument has been hard to make); rather,
the argument centers on whether the change in the elevation of
tidal frame was driven by global (i.e., climatic) fluctuations,
or was the result of tectonics or possibly even coastal storms.
The answer generally lies in the type of feature observed.
Ancient reefs
In the warm tropical waters where ambient temperatures aver-
age about 26
C, coral polyps are able to construct reefs, which
over time can build large reef complexes exemplified by the
Great Barrier Reef (about 18 million years old). Some of
the strongest evidence for global sea level (and climatic with
respect to Milankovitch cycles) changes during the Quatern-
ary comes from fossil reefs in Barbados (Mesolella et al.,
1969) and New Guinea (Bloom et al., 1974; Chappell and
Shackleton, 1986). The interpretation of fossil reefs as evidence
of sea level change must confront the issue of what the rela-
tionship of the reef is to modern sea level. For Pleistocene
age reefs, the impacts of tectonics are important, as many
of the reported reef complexes (e.g., New Guinea) have
undergone considerable uplift. Fortunately, in most instances,
the rate of long-term uplift appears to have been constant
throughout the Quaternary, and adjustment of the elevation of
the reef to former sea level sea (and modern sea level) can
be accommodated.
In reefs of late Holocene age (ca. 5,000
BP to present), the
species of coral becomes critical to determining the relationship
of the reef to former sea level. In areas where the rate of tec-
tonic uplift is very slow, a fossil reef dating from within the last
few millennia may be only slightly above present mean sea
level. In other instances where tectonism or other crustal move-
ments have been absent, only a few meters may submerge
a recent, subfossil reef that is no longer active. Coral species
differ significantly in their occurrence with depth of water.
Some species (e.g., Acropora palmata, or elkhorn coral) grow
in very shallow waters, often becoming exposed at low tide.
Others, like several brain and head corals (e.g., Montastrea),
thrive best in water depths of 5–6 m (Blatt et al., 1972). With
the magnitude of sea level changes during the late Holocene
amounting to <10 m even along the subsiding US middle
Atlantic Coast (Kearney, 2001), the error imparted to a sea
level reconstruction by not accounting for the dominant coral
species – and relation to former mean sea level – composing
a fossil reef could be significant.
“Raised beaches”: berms, strandlines and wave-cut
scarps
Apparent beach features occurring above the present limits of
the tides (supertidal) have been widely reported as evidence
of former sea level high stands. Some of the features undeni-
ably document past sea level highs – whether in response to
regional tectonics or possible eustatic events – while others
are more ambiguous in the interpretation of their relationship
to actual sea level changes.
Along many sandy coasts, like the Florida Panhandle
(Stapor, 1973), beach-like features a few meters above present
mean sea level have been reported as marking significant
transgressive episodes during the late Holocene. These features
can be either erosional (scarps) or depositional (berms), but in
most cases are relatively small-scale forms (a few meters high
at most in height). A common criticism, given their small
size, is that these features originated from the penetration
of large waves above former tidal limits – perhaps abetted
by storm surges – and not from eustatic sea level changes.
Certainly, scarps and berms can be observed to have formed
after large coastal storms that produced significant storm surges
and wave setup. Moreover, some of the most widely cited
examples of wave-cut scarps and berms touted as associated
with past high stands occur in regions of considerable tropical
storm (i.e., hurricane) activity, like the Gulf of Mexico, a
coincidence not lost on those who argue that they are largely
storm phenomena.
On the other hand, the similar elevation above mean sea
level of equivalent-age supertidal scarps and berms over rela-
tively large coastal areas is hard to reconcile with an origin
other than a rise in sea level. Wave approach, setup, and similar
characteristics from any storm differ across a region, with con-
comitant changes in the landward penetration of storm waves
and the elevation of supertidal forms they create. Assessment,
then, of “raised beach” features as sea level indicators is likely
to remain contentious.
Wave-cut notches
Wave-cut notches, a phenomenon of rocky cliff coasts, are
erosional features produced from abrasion by wave-carried
debris at the base of cliffs. Although rock hardness varies,
especially comparing relatively soft carbonates to volcanic
materials like basalt, wave-produced abrasion is generally a
slow process in most rocks. Hence, a significant wave-cut, par-
ticularly in hard rocks, indicates a long period of wave impacts
against the same spot on the cliff. Most reported wave-cut
notches are above present mean sea level, but submerged
wave-cut notches marking sea level high stands immediately
following deglaciation occur at 60 m depth in areas of Hawaii
(cf. Fletcher and Sherman, 1995).
Marine terraces
Marine terraces are former wave-cut platforms that have under-
gone uplift. Wave-cut platforms typically slope gently seaward,
900 SEA LEVEL INDICATORS
marking the mean low water level (seaward) and high water
level (landward). These features typically characterize rocky
cliff coasts. A spectacular flight of marine terraces (>20)
fringes parts of San Clemente Island off California. Along
the California coast, some of the youngest subaerial terraces
date from 100,000
BP, or approximately isotopic Stage 5c.
The oldest marine terrace in the Palos Verde Hills of Los
Angeles, California rises 394 m above present mean sea level,
and is 600,000 years old.
Because they originate as wave-cut platforms, requiring
decades or more of corrasion to create, depending on rock
hardness, marine terraces clearly represent a considerable per-
iod of slow transgression or stable sea level. Thus, if the
long-term rate of uplift was known, the original elevation of
the terrace – as a proxy for the approximate former sea level
position – could be determined. However, coseismic activity
is not uniform along long reaches of coast, and deformation
of Wisconsinan-age terraces by synclinal and anticlinal folding
has been reported for the Santa Barbara area in California
(Gurrola and Keller, 1997). Moreover, surface erosion and
mass wasting of older Pleistocene terraces can affect their abso-
lute elevation relative to past sea levels, even if corrected for
uplift. The determination of sea level position from Pleistocene
marine terraces can thus be subject to some error, especially in
low terraces of late Sangamonian age (ca. 80–90 ka).
Biological indicators
Biological indicators encompass a great variety of fossil and
subfossil animal and plant materials that have been used to
reconstruct past sea level changes. Not all are equally useful,
and often the accuracy of interpretations of former sea level
position derived from many of these indicators is open to ques-
tion. The keys to using biological indicators are: (a) reliable
and precise knowledge of where the organism lives in relation
to modern mean sea level, and (b) some certainty that the
occurrence of the remains of the organism in the sedimento-
logical record are in situ, and have not been subject to post-
mortem transport and deposition.
Shells and sponge spicules
Shells of pelecypods, like clams (e.g., Mercenaria mercenaria)
and oysters (e.g., Crassostera virginica), commonly can be
found along most coasts. Many of these animals clearly inhabit
relatively shallow waters, generally within 5 m or less of mean
tide level. Thus, the finding of these shells (if they can be
assumed to have buried where the organism lived) can be
indicative of relatively shallow water conditions, a particularly
useful indicator of former sea level position in long cores of
littoral sediments spanning millennia. Nonetheless, because
many species of pelecypods can inhabit a range of water depths
(e.g., between 3–10 m for Mytilus edulis), the usefulness
of their shells in littoral sediments of late Holocene age
(when the total variation in sea levels was generally <10 m
in many areas of the US Atlantic Coast; Kearney, 2001)is
substantially limited.
Similar criticisms can be made of employing the occurrence
of spicules of shallow water sponges in sediment records as
sea level indicators. An additional problem is that these small
generally siliceous needles can be widely scattered across
the ocean floor after the death of the organism (without break-
age and indication of post-mortem transport, necessarily) and,
therefore, are often not found in situ.
Tree stumps
The finding of rooted fossil tree stumps now submerged in
the lower shoreface has been interpreted as marking past sea
levels (e.g., Kearney, 1996). However, reconstruction of former
sea level position employing submerged tree stumps can be
fraught with considerable error, as many riparian or fastland
tree species can often occur in strictly upland forests, rendering
questionable what the elevation of the tree was to mean tide
level at the time of death. Again, this is less of a problem in
early Holocene or Pleistocene sea level reconstructions, where
the range of total elevational change encompassed in these
records is tens of meters. In these instances, they clearly func-
tion as dramatic evidence of a substantial change in sea level,
despite their limitations in precision.
Intertidal deposits: coastal marshes
The influence of sea level rise on the upward growth (vertical
accretion) and overall development of coastal marshes is well-
recognized (Stevenson et al., 1986). Theoretically, since the
elevation of the marsh surface adjusts to changes in mean tide
level, older marsh sediments below the substrate must record
similar relationships to ancient tidal levels, and thereby to sea
level. Unfortunately, this reasonable assumption was under-
mined by findings some time ago that older marsh sediments
can undergo considerable loss in volume upon burial, initially
from dewatering (Kearney and Ward, 1986), and later, by auto-
compaction (the collapse of old buried peats under their own
weight and that of younger sediments; Kaye and Barghoorn,
1964). These phenomena result in peat sediments sinking
below the level at which they were originally deposited in
relation to mean tide level, making reliable reconstruction of
age-depth relations in long marsh sedimentary records proble-
matic. They also affect the use of geochemical (Varekamp et al.,
1992) and faunal indicators of changes in sea level archived
in marsh sediments, though these data are typically used for
more relative interpretations of the magnitude of past transgres-
sions and regressions. Scott and Medioli (1978) nonetheless
have shown that accurate reconstructions of former sea levels
are still possible by delineating vertical zonations of marsh
foraminifera in marsh cores.
The present use of marsh sediment records as sea level indi-
cators generally is in collection of radiocarbon-dated marsh
basal peats. Basal peats are assumed to record the initial
conversion of the former upland surfaces to marsh by rising
sea (a phenomenon termed marsh lateral accretion). Since the
pre-existing surfaces of marshes are typically compacted, dense
muds, dates collected from basal peats in contact with these
sediments are more likely to be at their original elevation and
not to have undergone any autocompaction. Thus, a more reli-
able reconstruction of time-depth relationships for sea level
change is possible. The only limitation of such indicators is that
they are most suited to documenting past rises (transgressions)
in sea level, as new marsh covered existing fastlands, and are
not suited to documenting significant falls in sea level (regres-
sions), when the marsh may have disappeared with the seaward
drop in the elevation of the tides.
Archaeological evidence
Evidence for sea level change drawn from the world archaeolo-
gical record spans shell middens from Woodland and earlier
fishing communities in various sites in North America (see
Fairbridge, 1976) to Medieval through Neolithic-age sites
SEA LEVEL INDICATORS 901
across Mediterranean and Europe (Kearney, 2001). The wealth
of archaeological data that has been applied as sea level indica-
tors is so large that it defies easy description. However, archae-
ological material that is truly useful in sea level reconstruction
must fulfill the same criteria as other sea level indicators:
(a) the relationship to mean sea level at the time of construction
and/or burial can be determined, (b) the material must be
in situ, and (c) in terms of structures, later modifications of
either the structure or the surrounding site must not have chan-
ged the initial relationship to mean sea level. The last criterion
is especially critical regarding ancient harbor and port facilities
that were used for long periods of time. For example, subse-
quent harbor construction – often many centuries later – to meet
demands of increased shipping or better defense, often involved
extensive modifications (if not their replacement) of original
structures like docks that changed initial relationships to mean
sea level in ways that may be not readily discernible. For exam-
ple, the Port of Alexandria, Egypt, underwent substantial modi-
fication over the centuries since Hellenistic times, as waves of
conquerors from the Romans to the Mamelukes came and went.
Even when acknowledging the potential problems inherent in
the use of archaeological materials, it is not surprising that such
evidence is still regarded as highly suspect by many sea level
researchers. Nonetheless, the evidence often has the undeniable
advantage of being able to be dated to within a decade, even a
particular year, especially in the case of Roman (empire) materi-
als. Moreover, in some instances, the relationship to former
mean sea level can be deduced with a high degree of precision
that similar-age biological materials would never allow.
Michael S. Kearney
Bibliography
Blatt, H., Middleton, G., and Murray, R., 1972. Origin of Sedimentary
Rocks. Englewood Cliffs, NJ: Prentice-Hall, 634pp.
Bloom, A.L., Broecker, W.S., Chappell, J., Mathews, R.K.,and Mesolella, K.J.,
1974. Quaternary sea level fluctuations on a tectonic coast: new
230Th
/
234
U dates from the Huon Peninsula. Quaternary Res., 4,185–205.
Cabanes C., Cazenave A., and Le Provost C., 2001. Sea level rise during
past 40 years from satellite and in situ observations. Science, 294,
840–842.
Chappell, J.M., and Shackleton, N.J., 1986. Oxygen isotopes and sea level.
Nature, 324, 137–138.
Douglas, B.C., 1991. Global sea level rise. J. Geophys. Res., 96 (C4),
6981–6992.
Douglas, B.C., 2001. Sea level change in the era of the recording tide
gauge. In Douglas, B., Kearney, M.S., and Leatherman S.P. (eds.),
Sea Level Rise: History and Consequences. New York: Academic Press,
pp. 37–62.
Fairbridge, R.W., 1976. Shellfish-eating preceramic Indians in coastal
Brazil. Science, 191, 353–359.
Fletcher, C.H., and Sherman, C., 1995. Submerged shorelines on Oahu,
Hawaii: Archive of episodic transgression during deglaciation?
J. Coast. Res., 17, Special Issue: Holocene cycles: Climate, sea level,
and sediment, pp. 141–152.
Gurrola, L.D., and Keller, E.A., 1997. Tectonic geomorphology of the
Santa Barbara Fold Belt, Western Transverse Ranges, California. Geol.
Soc. Am. Abs., 29(6), 344.
Kaye, C.A., and Barghoorn, E.S., 1964. Late Quaternary sea-level change
and crustal rise at Boston, Massachusetts, with notes on the autocom-
paction of peat. Geol. Soc. Am. Bull., 75(2), 63–80.
Kearney, M.S., 1996. Sea-level change during the last thousand years in
Chesapeake Bay. J. Coast. Res., 12, 977–983.
Kearney, M.S., 2001. Late Holocene sea level variation. In Douglas, B.,
Kearney, M.S., and Leatherman S.P. (eds.), Sea Level Rise: History
and Consequences. New York: Academic Press, pp. 13–36.
Kearney, M.S., and Ward, L.G., 1986. Accretion rates in brackish marshes
of a Chesapeake Bay estuarine tributary. Geo-Mar. Lett., 6,41–49.
Mesolella, K.J., Mathews, R.K., Broecker, W.S., and Thurber, D.L., 1969.
The astronomical theory of climatic change: Barbados data. J. Geol.,
77, 250
–274.
Mörner, N.-A., 1973. Eustatic changes during the last 300 years.
Palaeogeogr. Palaeoclim. Palaeoceanogr., 13,1–14.
Peltier, W.R., 2001. Global glacial isostatic adjustment and modern instrumen-
tal records of relative sea level history. In Douglas, B., Kearney, M.S.,
and Leatherman S.P. (eds.), Sea Level Rise: History and Consequences.
New York: Academic Press, pp. 65–86
Scott, D.B., and Medioli, F.S., 1978. Vertical zonations of marsh foraminifera
as accurate indicators of former sea levels. Nature, 272,528–531.
Stapor, F.W. Jr., 1973. Coastal Sand Budgets and Holocene Beach
Ridge Plain Development, Northwest Florida. Ph.D. thesis, Florida
State University, Tallahassee.
Stevenson, J.C., Ward, L.G., and Kearney, M.S., 1986. Vertical Accretion
in marshes with varying rates of sea level rise. In Wolf D. (ed.),
Estuarine Variability. New York: Academic Press, pp. 241–259.
Varekamp, J.C., Thomas, E., and van de Plassche, O., 1992. Relative sea
level rise and climate change over the last 1,500 years. Terra Nova, 4,
293–304.
Cross-references
Ancient Cultures and Climate Change
Beachrock
Coastal Environments
Coral and Coral Reefs
Glacial Isostasy
Radiocarbon Dating
Sea Level Change, Post-Glacial
Sea Level Change, Quaternary
SEDIMENTARY INDICATORS OF CLIMATE CHANGE
Introduction
Innumerable publications deal with the sedimentary record of past
climates (e.g., Nairn, 1961, 1963; Schwarzbach, 1963;Frakes,
1979; Parrish, 1998; and Cecil and Edgar, 2003). Earlier paleo-
climatic studies concentrated on identifying paleolatitudes and
their shifts in response to plate tectonic movements on timescales
of millions of years. Spurred by advances in the understanding of
Pleistocene and Holocene climate records, more recent studies
have focused on climate changes in response to orbital parameters
on the timescale of tens of thousands of years and even shorter
time frame variability. The sedimentary indicators of climate
change are inextricably linked to biological and chemical criteria,
which are typically recorded in sediments. The sedimentary
indicators of climate change consist of two broad groupings:
(a) sedimentary features that directly relate to climatic parameters
such as rainfall or temperature, and (b) successions of sedimentary
strata that indirectly reflect relative changes in these parameters.
In both cases, the link to climate is the regional distribution and
stratigraphic coherency of indicators. Many studies now assume
a climate driver for variability unless other mechanisms can be
identified.
Sedimentary features directly reflecting climatic
conditions
Some sedimentary features reflect conditions of temperature
or rainfall, and can be used to reconstruct past climate condi-
tions. However, even the best indicators need to be used in
902 SEDIMENTARY INDICATORS OF CLIMATE CHANGE
context of their regional setting and associations. Furthermore,
some of the features may be confused with other types of
deposits that do not have the same climatic constraints.
Extended periods of cold weather are necessary to produce gla-
ciers and their deposits. Striated boulders and tills are excellent
indicators of past glaciers. These, however, may be confused with
other boulder deposits such as alluvial fans, talus cones, or debris
flows in the geologic record. The mechanical grinding of bedrock
to clay-sized sediment (rock flour) by glaciers can be recognized
in till deposits and other environments well removed from the
glaciers. The re-sedimentation of glacial deposits long after the
glaciers have retreated may be misconstrued as evidence of colder
conditions. Associated glacial features such as eskers, outwash,
and dropstones (Figure S18a) are even less easily distinguished
from non-glacial deposits in the rock record. Features formed by
ground ice such as solifluction lobes, pingos, and ice wedges
are good indicators of cold climates (i.e., Washburn, 1980). Some
mineral phases are more stable at low temperatures such as ikaite
(CaCO
3
–6H
2
O) (Figure S18f), mirabilite (Na
2
SO
4
–10H
2
O),
natron (Na
2
CO
3
–10H
2
O) and hydrohalite (NaCl–2H
2
O). Within
the proper context, these are strong indicators of cold conditions.
Indications of warm or hot climate are less clear. In marine
settings, warm tropical waters favor growth of coral reefs and
precipitation of carbonate minerals. The biological component
of carbonate buildups in the geological record is less certain,
particularly with documented ‘‘cool-water ’’ carbonate deposits
forming at higher latitudes (see James and Clarke, 1997). There
is no clear correlation between accumulations of lacustrine
carbonates and temperature.
Continuous growth of vegetation to produce thick accu-
mulations of peat is best developed in rainy climates, as are
coal and lignite deposits (Figure S18b). Even in areas of mod-
erate rainfall, well-developed root structures are common in
soils. The presence of coal or lignite is not unequivocal evi-
dence of rainy conditions, however, because plants will grow
in abundance wherever there is enough water, even in an arid
climate. Similarly, perennial rivers, lakes, and swamps may
all exist in areas of low rainfall, if the source of water is suffi-
cient. Chemical weathering is best developed in wet, rainy, and
warm climates, whereas mechanical weathering is best devel-
oped in dry and/or cold climates. The production of high-
aluminum or lateritic clays and the removal of labile minerals
from sands are good indications of prolonged chemical weath-
ering. Even marine units that are dominated by these sediments
may be correlated with rainy paleoclimates if regional and
stratigraphic constraints are considered. The absence of these
features, however, does not eliminate rainy climates since other
factors such as relief, sediment transport, and diagenesis are
also important.
Indicators of aridity include saline minerals, eolian sand
deposits, ephemeral stream deposits, closed-basin lakes, and
dry mudflats. Both marine and non-marine syndepositional sal-
ine minerals (Figure S18e) require evaporation to exceed water
supply in order to concentrate solutes. The environment of for-
mation, however, may reflect only local dryness or the evapor-
ites may be superimposed on materials originally deposited in a
wetter climate. Extensive eolian sand seas (Figure S18c) form
where vegetation is sparse in dry climates. Local eolian sand
deposits may form in a variety of climatic conditions. River
and stream channels may remain completely dry for prolonged
periods in areas of low rainfall. Ephemeral streams deposits,
however, may be very similar to the deposits of perennial
streams. Dry channels may also reflect local phenomena such
as channel switching or small drainage areas. Closed-basin
lakes are maintained by evaporation exceeding inflow. The cri-
teria for recognizing closed-basin lakes, which include frequent
shifts in depth indicators and accumulation of soluble minerals,
are not unique or easily recognized in the rock record. Further-
more, closed-basin lakes may form at a small scale in a range
of climates under local conditions. Dry mudflats, which are
dominated by desiccation cracks and vesicles, require a de-
pressed groundwater table. The criteria for their recognition
are still being developed. The predominance of mechanical
weathering in both cold and hot arid settings suggests a miner-
alogical indication of paleoclimate. The abundance of imma-
ture sands such as arkoses or lithic arenites has been used as
climate indicators. Red sedimentary deposits, stained by iron
oxides, have also been cited as indicators of aridity. The mount-
ing evidence suggests that these are unreliable indicators of
paleoclimate. Immature sands appear to reflect relief and rate
of transport more than climate, and redbeds are diagenetic
overprints that are independent of the depositional climate.
The most promising sedimentary features indicative of cli-
mate change are soils (for instance, Retallack, 2001). Modern
soils show a strong relationship with both local and regional
climate as should their fossil counterparts. Some of the features
already mentioned are incorporated into the soil classification,
such as aridisol (saline minerals or mudcracks and vesicles),
histosol (coal or peat), or gelisol (ground ice features). The
climatic divisions, however, can be much finer based upon
abundance and depth of mineral horizons (such as carbonate,
aluminum, or iron), eluviation features, parting textures (peds),
and vegetation overprint. Bauxitic and lateritic soils are good
indicators of high rainfall. Vertisols (soils with abundant sticky
or swelling clay) and calcisols require at least seasonal desicca-
tion or dryness (Figure S18d). Thick eluviated clay B horizons
with abundant roots and no soluble minerals, such as in spodo-
sols (soils with a subsurface horizon containing organic matter,
iron, and/or aluminum) or ultisols (soils with a well-developed
argillic (clay) horizon and low level of exchangeable base metal
cations) (Figure S18b), reflect rainier climates that support vege-
tation and remove soluble minerals. The drawbacks to using
paleosols include the difficulty in recognizing fossil counterparts
due to diagenetic changes in color and composition, similar fea-
tures in different soils, and restriction to subaerial conditions.
Sedimentary features that indirectly reflect climate
change
Relative changes in the world climate may be indicated by the
changing character of sedimentary environments. Shifts in net pre-
cipitation are reflected in certain continental or nearshore settings
by relative indicators of wetness or dryness. Relative shifts in
temperature are usually more difficult to assess from sedimentary
criteria. The problem with indirect climatic indicators is that not all
environments react the same way to similar climatic changes and
the degree of preservation of these changes is highly variable. In
the best cases, relative climatic change is recorded at the millennial
or centennial scale of resolution or even finer (for instance Fischer,
1986). More often, the patterns of climate change are coarsely pre-
served, with numerous gaps or uncertainties. Laminated sediment
from lakes or deep marine settings often shows variations in
mineralogy, sediment type, or thickness that reflect changes in
temperature or precipitation (Figure S19). In some circumstances,
these can be documented as annual layers that provide incredible
SEDIMENTARY INDICATORS OF CLIMATE CHANGE 903