Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

Leaf physiognomy predictions are relatively accurate, but

several assumptions do exist. The first assumption is that

the relationship between leaf morphology and climate has

remained constant through time. The second is that taphonomic

biases are small, though they are known to exist (Roth and

Dilcher, 1978; Greenwood, 1992). The third is that regional

differences due to climatic or taxonomic history are not great.

Several authors have suggested that it is important to use pre-

dictor data sets that are similar geographically, taxonomically

or climatically to those likely to be encountered in the fossil

assemblage (Stranks and England, 1997; Gregory-Wodzicki,

2000; Jacobs, 2002; Kowalski, 2002). This restriction should

reduce the problems that might be encountered by the third

assumption above. Leaf physiognomy methods also require a

relatively high number of fossil taxa (~20) to reduce paleocli-

mate estimate errors. Only the leaf characters of woody angios-

perms have been shown to correlate with climate, restricting

the use of the method to the Cretaceous and Cenozoic.

Both stomatal density (number of stomata per area) and sto-

matal index (the ratio of the number of epidermal leaf cells to

the number of stomata in a defined area, multiplied by 100)

can be used to determine atmospheric CO

concentrations, tem-

perature or moisture availability. Stomata pores are surrounded

by guard cells that control the size of the stomatal opening and

thus the gas exchange and water loss from the leaf to the atmo-

sphere. The number of stomata that differentiate is determined

at the time of leaf unfolding (Kurschner et al., 1998) when

the leaf acts as a biosensor of atmospheric CO

. Stomatal index

is used more often than stomatal density, as it eliminates the

effect of leaf size on density counts (large leaves with large

cells might have a decreased stomatal density, but not a

decreased stomatal index). Quercus, Ginkgo and Metasequoia

have been used, among other taxa, to calibrate CO

-stomatal

index curves from which to estimate paleoCO

levels (e.g.,

Kurschner et al., 1998; Royer, 2001; Beerling and Royer,

2002). All species have a stomatal density and stomatal index

with a negative relationship to atmospheric carbon concentra-

tion. These taxa are broad-ranging temporally (both Ginkgo

and Metasequoia are termed “living fossils”) and can track

atmospheric carbon changes over long periods of time.

Like all methods, stomatal index and stomatal density have

advantages, but must be used with caution. As CO

rises, stomatal

index responds non-linearly. Thus, problems arise when estimat-

ing paleoCO

during periods of higher than modern atmospheric

concentrations (Kurschner et al., 1998; Beerling and Royer,

2002). Stomatal index decreases as temperature increases, even

under the same ambient CO

(Beerling and Chaloner, 1993). This

can influence atmospheric CO

estimates if temperature is not

constant over the time period being analyzed. However, tempera-

ture can be estimated in combination with other leaf physiognomy

methods, which can minimize potential interpretation problems.

Stomatal density also appears to correlate, at least qualitatively,

with precipitation amount, especially in shade leaves (Sun et al.,

2003). Stomatal index does not seem to reflect this correlation

and as such is a better measure for atmospheric CO

concentration.

A major benefit to using stomatal index and stomatal density to

estimate CO

is that this method can be used from the Paleozoic

to the Recent (Chaloner and McElwain, 1997).

Leaf venation density is a method that is more recently

being refined as a way to estimate paleomoisture parameters.

Leaf venation density changes in response to the amount of leaf

transpiration, which is due to water availability or temperature

(Uhl and Mosbrugger, 1999; Uhl et al., 2002). Leaf venation

density is influenced by climate factors during leaf production

in one growing season; thus, this method can record small-

scale changes that may vary on the order of seasons. However,

several problems still exist with this method, including density

differences with leaf size dependent on species, how to define

density, and differences between sun and shade leaves. How-

ever, if these problems are addressed, this method could poten-

tially be used from the Paleozoic to the Recent.

Wood

Several aspects of wood can be used to determine paleoclimate,

including wood anatomy, tree rings and charcoal. Wood is

often found in fossil environments that are not suited for leaf

preservation. Thus, using wood as a climate proxy can increase

the number of environments in which paleoclimatic reconstruc-

tion is possible from plants.

Paleoclimate reconstruction from wood anatomy is based on

cellular level proxies, such as the presence, organization or

size of vessel elements, ray cells, axial parenchyma and fiber

cells (Wiemann et al., 1999; Woodcock and Ignas, 1994). The

relationship between wood anatomy and climate has been

demonstrated qualitatively based upon relationships observed

between vessel density and diameter with moisture, vessel dia-

meter as a function of temperature, and the presence of ring

porosity as an indicator of seasonality. Quantitative equations

have been derived, most notably by Wiemann et al. (1998,

2001), that result in numerical climate estimates from these

same features. These equations ideally predict temperature to

within 5



C, although equations using differing predictive char-

acters from the same site can differ by almost 6



C. Tempera-

ture correlates fairly well with wood anatomical characters,

but precipitation correlates poorly (Wiemann et al., 1998,

2001). These equations need a relatively large number of taxa

(~25) to estimate climate in order to reduce error. However,

Terral and Mengal (1999) have shown that it is possible to

use one species through time to estimate climate, if that species

shows variation in anatomical characters with respect to cli-

mate. In their study, olive wood was used to estimate Holocene

paleoclimate.

Figure P4 Percentage of entire margined leaves in a flora versus mean

annual temperature for 177 sampled floras worldwide. Data from Wolfe

(1993); Gregory-Wodzicki (2000); and Kowalski, (personal observation).

PALEOBOTANY 681

Both qualitative and quantitative climate estimates, based

upon wood anatomy, work relatively well to distinguish tropi-

cal from more temperate or seasonal sites. The cellular anatomy

of wood grown in temperate or seasonal climates is more influ-

enced by climate than wood grown in tropical environments.

As such, climate reconstruction using wood allows for broad

categorical statements about paleoclimate, and can distinguish

well between worldwide sites. However, in smaller areas

(e.g., eastern North America), taxonomic overlap between sites

can result in temperature differences that are not reflected in the

wood assemblage (Kowalski, personal observation). In some

areas, anatomical characters used in this type of climate recon-

struction, such as the presence or absence of spiral thickening,

do not vary significantly within species.

Cautions with using wood anatomy to determine paleocli-

mate are similar to those using leaf methods, such as the

assumption that the relationship between climate and anatomy

has been constant through time. Ideally, wood anatomy techni-

ques could be used back into the Paleozoic. However, some

wood characters used to estimate climate are derived (e.g., ves-

sel characters such as vessel density and ring porosity). These

wood characters developed with the evolution of angiosperms

in the late Mesozoic (Francis and Poole, 2002), and thus these

wood characters cannot be used to estimate the paleoclimate of

the Paleozoic and early and middle Mesozoic.

Dendrochronology or tree ring analysis can be used to deter-

mine climate over several seasons, as each ring records aspects

of moisture and temperature of one growing season. The result-

ing information details a long series of annual climate and cli-

mate changes, with a precision that few other records approach.

Most information regarding the response of tree ring character-

istics to climate has come from studies of conifers (Francis

and Poole, 2002), although angiosperms are occasionally stu-

died as well. A continuous record of annual tree rings has been

established in the southwestern USA, extending back nearly

9,000 years. Changes in moisture and temperature of local cli-

mates often can be correlated to ancient human habitations

and migrations. The presence/absence of tree rings is an indi-

cation of seasonality, often related to temperature (Francis and

Poole, 2002). Ring presence, as an indicator of seasonality of

temperature, is best determined from more than one taxa, as

some tropical tree species possess rings and could be confused

with temperate species. Tropical tree rings are most likely due

to seasonality of precipitation or seasonal leaf loss (Woodcock

and Ignas, 1994). Ring width is useful in determining growing

conditions, with wide rings the product of warmer and wetter

conditions or increased CO

(Beerling, 1999), and narrow rings

produced under cooler or drier conditions (Francis and Poole,

2002). Occasionally, false rings occur, which can indicate a halt

during the growing season due to fire, drought, insect damage

or frost. Ring widths are best used for broad floral compari-

sons, rather than estimating the absolute temperature or humid-

ity of one flora. Mean sensitivity measures the variability of

tree rings from year to year, and gives an indication of the con-

stancy of climate regime (Francis and Poole, 2002). Tree ring

analyses can be performed on floras from the Recent to the

Paleozoic but most applications have documented climate dur-

ing the last 10,000 years.

Finally, the presence of charcoal in an assemblage can deter-

mine whether paleoO

levels are within the relatively narrow

range associated with charcoal formation. Charcoal can only form

within a relatively narrow O

range (Chaloner and McElwain,

1997). This area needs more research because plant parts are

frequently charcoalified and found in sediments from the

Devonian to the Recent. Such fragments are known as mesofossils

and can be used in floristic studies to assign nearest-living-relative

climatic relationships.

Chemical signatures

Plants take up carbon, oxygen and hydrogen isotopes from

atmospheric CO

and water during growth. These isotopes

are incorporated into the plant structure. They preserve a record

of the atmosphere and climate in which the plant grew. These

records can show an individual season’s values by looking at

leaves or individual tree rings, or can show long-term fine-scale

variations by comparing tree ring values over the life of the

tree. d

C measurements are used to estimate paleoclimate,

but have also been used to match CO

records of tree rings

with ice core data to determine the exact ages of the trees being

measured.

C is the most commonly measured isotopic ratio

used to determine paleoclimate from plants, and is taken into

the plant from CO

in the atmosphere during growth. In wood,

C ratios are measured in the cellulose and show a negative

relationship with relative humidity and temperature (Edwards

et al., 2000).

C measurements from the leaf cuticle also show

a negative correlation between

C ratios and precipitation or

humidity (Sun et al., 2003), (Figure P5).

In addition to

C, both d

O and d

H have been measured

using wood cellulose and wood nitrite to determine variability

in meteoric water, seasonal moisture (relative humidity and pre-

cipitation) and precipitation amount (Buhay and Edwards,

1995; Edwards et al., 2000). In addition, d

H shows a relation-

ship with mean annual and seasonal temperature, though this

relationship is best measured in areas where the water supply

has a constant isotopic value (Buhay and Edwards, 1995).

Because of the complex interaction between precipitation and

temperature, climate estimates using measurements from only

one isotope may be misleading. Using a combination of d

Figure P5

C isotopic composition of modern Ginkgo leaves

versus annual precipitation for three sites in China. Data

from Sun et al. (2003).

682 PALEOBOTANY

H and d

O measurements seems to offer the most paleocli-

matic information from plants (Edwards et al., 2000).

Elizabeth A. Kowalski and David L. Dilcher

Bibliography

Beerling, D.J., 1999. Atmospheric carbon dioxide, past climates and the

plant fossil record. Bot. J. Scotl., 51,49–68.

Beerling, D.J., and Chaloner, W.G., 1993. The impact of atmospheric CO

and temperature change on stomatal density: Observations from Quer-

cus robur Lammas leaves. Ann. Bot., 71, 231–235.

Beerling, D.J., and Royer, D.L., 2002. Reading a CO

signal from fossil

stomata. New Phytol., 153, 387–397.

Buhay, W.M., and Edwards, T.W.D., 1995. Climate in southwestern

Ontario, Canada between AD 1610 and 1885 inferred from oxygen

and hydrogen isotopic measurements of wood cellulose from trees in

different hydrologic settings. Quaternary Res., 44, 438–446.

Chaloner, W.G., and Creber, G.T., 1990. Do fossil plants give a climatic

signal? J. Geol. Soc. London, 147, 343–350.

Chaloner, W.G., and McElwain, J., 1997. The fossil plant record and global

climatic change. Rev. Palaeobot. Palynol., 95,73–82.

Davies-Vollum, K.S., 1997. Early Palaeocene palaeoclimatic inferences

from fossil floras of the western interior, USA. Palaeogeogr. Palaeocli-

matol. Palaeoecol., 136, 145–164.

Dilcher, D.L., 1974. Approaches to the identification of fossil leaf remains.

Bot. Rev., 40,1–157.

Dilcher, D.S., 1973. A Paleoclimatic interpretation of the Eocene Floras of

Southeastern North America. In Graham, A. (eds.), Vegetation and

Vegetational History of Northern Latin America. Amsterdam: Elsevier.

Chapter 2, pp. 39–59.

Dilcher, D.L., 2000. Toward a new synthesis: Major evolutionary trends in

the angiosperm fossil record. PNAS, 97, 7030–7036.

Edwards, T.W.D., Graf, W., Trimborn, P., Stichler, W., Lipp, J., and Payer,

H.D., 2000. d13C response surface resolves humidity and temperature

signals in trees. Geochim. Cosmochim. Acta , 64, 161–167.

Francis, J.E., and Poole, I., 2002. Cretaceous and early Tertiary climates of

Antarctica: Evidence from fossil wood. Palaeogeogr. Palaeoclimatol.

Palaeoecol., 182,47– 64.

Givnish, T.J., 1987. Comparative studies of leaf form: Assessing the rela-

tive roles of selective pressures and phylogenetic constraints. New Phy-

tol., 106, 131–160.

Greenwood, D.R., 1992. Taphonomic constraints on foliar physiognomic

interpretations of Late Cretaceous and Tertiary paleoclimates. Rev.

Palaeobot. Palynol., 71

, 149–190.

Greenwood, D.R., and Wing, S., 1995. Eocene continental climates and

latitudinal temperature gradients. Geology, 23, 1044–1048.

Gregory-Wodzicki, K.M., 2000. Relationships between leaf morphology

and climate, Bolivia: Implications for estimating paleoclimate from fos-

sil floras. Paleobiology, 26, 668–688.

Halloy, S.R.P., and Mark, A.F., 1996. Comparative leaf morphology spectra

of plant communities in New Zealand, the Andes and the European

Alps. J. R. Soc. N. Z., 26,41–78.

Jacobs, B.F., 2002. Estimation of low-latitude paleoclimates using fossil

angiosperm leaves: Examples from the Miocene Tugen Hills, Kenya.

Paleobiology, 28, 399–421.

Kowalski, E.A., 2002. Mean annual temperature estimation based on leaf

morphology: A test from tropical South America. Palaeogeogr. Palaeo-

climatol. Palaeoecol., 188, 141–165.

Kowalski, E.A., and Dilcher, D.L., 2003. Warmer paleotemperatures for

terrestrial ecosystems. PNAS, 100, 167–170.

Kurschner, W.M., Stulen, I., Wagner, F., and Kuiper, P.J.C., 1998. Compar-

ison of paleobotanical observations with experimental data on the leaf

anatomy of Durmast Oak [Quercus petraea (Fagaceae)] in response to

environmental change. Ann. Bot., 81, 657–664.

Martin, H.A., 1997. The use of ecological tolerances for the reconstruction

of Tertiary palaeoclimates. Aust. J. Bot., 45, 475–492.

Mosbrugger, V., and Utescher, T., 1997. The coexistence approach – a

method for quantitative reconstructions of Tertiary terrestrial palaeocli-

mate data using plant fossils. Palaeogeogr. Palaeoclimatol. Palaeoe-

col., 134,61–86.

Parkhurst, D.F., and Loucks, O.L., 1972. Optimal leaf size in relation to

environment. J. Ecol., 60, 505–537.

Roth, J.L., and Dilcher, D.L., 1978. Some considerations in leaf size and

leaf margin analysis of fossil leaves. Cour. Forsch. – Inst. Senckenberg,

30, 165

–171.

Royer, D.L., 2001. Stomatal density and stomatal index as indicators of

atmospheric CO

concentration. Rev. Palaeobot. Palynol., 114,1–28.

Sinka, K.J., and Atkinson, T.C., 1999. A mutual climatic range method for

reconstructing palaeoclimate from plant remains. J. Geol. Soc. London,

156, 381–396.

Spicer, R.A., 1989. The formation and interpretation of plant fossil assem-

blages. Adv. Bot. Res., 16,95–191.

Stranks, L., and England, P., 1997. The use of a resemblance function in the

measurement of climatic parameters from the physiognomy of woody

dicotyledons. Palaeogeogr. Palaeoclimatol. Palaeoecol., 131,15–28.

Sun, B., Dilcher, D.L., Beerling, D.J., Zhang, C., Yan, D., and Kowalski, E.,

2003. Variation in Ginkgo biloba L. leaf characters across a climatic gra-

dient in China. PNAS, 100, 7141–7146.

Terral, J.-F., and Mengal, X., 1999. Reconstruction of Holocene climate in

southern France and eastern Spain using quantitative anatomy of olive

wood and archeological charcoal. Palaeogeogr. Palaeoclimatol.

Palaeoecol., 153,71– 92.

Uhl, D., and Mosbrugger, V., 1999. Leaf venation density as a climate and

environmental proxy: A critical review and new data. Palaeogeogr.

Palaeoclimatol. Palaeoecol., 149,15–26.

Uhl, D., Walther, H., and Mosbrugger, V., 2002. Leaf venation density in

angiosperm leaves as related to climatic and environmental conditions-

problems and potential for palaeoclimatology. N. Jb. Geol. Paläont.

Abh., 224,49–95.

Wiemann, M.C., Wheeler, E., Manchester, S.R., and Portier, K.M., 1998.

Dicotyledonous wood anatomical characters as predictors of climate.

Palaeogeogr. Palaeoclimatol. Palaeoecol., 139,83–100.

Wiemann, M.C., Manchester, S.R., and Wheeler, E., 1999. Paleotempera-

ture estimation from dicotyledonous wood anatomical characters.

Palaios, 14, 459–474.

Wiemann, M.C., Dilcher, D.L., and Manchester, S.R., 2001. Estimation of

mean annual temperature from leaf and wood physiognomy. Forest Sci.,

47, 141–149.

Wilf, P., 1997. When are leaves good thermometers? A new case for leaf

margin analysis. Paleobiology, 23

, 373–390.

Wilf, P., Wing, S., Greenwood, D.R., and Greenwood, C.L., 1998. Using

fossil leaves as paleoprecipitation indicators: An Eocene example.

Geology, 26, 203–206.

Wing, S., and Greenwood, D.R., 1993. Fossils and fossil climate: The case

for equable continental interiors in the Eocene. Phil. Trans. R. Soc.

Lond. B, 341, 243–252.

Wolfe, J.A., 1993. A method of obtaining climatic parameters from leaf

assemblages. USGS Bull., 2040, 71.

Woodcock, D.W., and Ignas, C.M., 1994. Prevalence of wood characters in

eastern North America: What characters are most promising for inter-

preting climates from fossil wood? Am. J. Bot., 81, 1243–1251.

Cross-references

Carbon Isotopes, Stable

CLAMP

Dating, Dendrochronology

Dendroclimatology

Deuterium, Deuterium Excess

Nearest-Living-Relative Method

Oxygen Isotopes

Paleotemperatures and Proxy Reconstructions

Palynology

PALEOCEAN MODELING

Introduction

Paleoceanographic (paleocean) modeling is the application

of analytic or numerical techniques to predict and understand

the ocean’s past circulation, biogeochemical interactions, and

climate. Paleocean modeling is primarily a tool for exploring

PALEOCEAN MODELING 683

the mechanistic functioning of the ocean and for making con-

crete predictions testable against paleoenvironmental proxies.

Paleocean modeling may be used to investigate such specific

questions as: (a) What was the current structure and tracer distri-

bution of a given past time interval? (b) How sensitive are major

oceanographic features, e.g., the Gulf Stream or the meridional

overturning circulation (MOC) to boundary condition changes?

dal, and millennial-scale variability change under different con-

ditions? It may also answer some more general questions, such

as: (d) What is the role of the ocean in governing the Earth sys-

tem’s response to greenhouse gas (GHG) forcing, and what is

the ocean’s role in GHG storage and release through Earth’shis-

tory? (e) To what degree are ocean heat transport (OHT) changes

important for understanding climate change, and to what is OHT

sensitive: GHGs, gateways, multiple equilibria? Due to recent

advances in modeling, answers to these questions are now in

the offing, once suitable data and a framework with which to

interpret them can be developed.

Conditions are sufficiently close to the present for modeling

intervals within the past several million years, such that modern

numerical ocean modeling techniques are likely to work as well

for the past as they do for modern times. Analytic, “box,” and

simplified physical models are general enough that they can be

applied as readily to paleoceanographic problems. Biogeo-

chemical models are crucial but difficult to extend into the deep

past because many ecological and biogeochemical processes

that only are moderately understood today are poorly known

in the deep past.

This article will focus largely, therefore, on the use of

sophisticated numerical models in time periods deep in the

past (>3 Ma), before continents and climate had reached their

near-modern states – first, on the unique challenges that arise

in these periods; and next, on the techniques used to overcome

these challenges. These special challenges are balanced by

unique opportunities to study the evolution of climate and the

Earth system over its full dynamic range. If chosen carefully,

i.e., only when the environmental “signal” provided by proxies

is much greater than the “noise” inherent in the uncertain phy-

sics of models, paleocean modeling answers questions that are

impossible to answer within near-modern settings. We con-

clude with the main results of this young field and discuss

some important future directions.

Challenges of paleocean model ing

The ocean is coupled to climate through its surface: via its state,

i.e., sea surface temperatures (SSTs), and via fluxes of sunlight,

heat, water, and momentum. It is also biogeochemically coupled

to the biosphere, through fluxes of sunlight, carbon dioxide, and

dust, and traces of organic and inorganic chemicals.

Thus, in order to understand the ocean’s dynamics and phe-

nomenology and to predict its role in climate and biogeochem-

ical cycles it is necessary to constrain the state of – and fluxes

through – the interface between ocean and atmosphere (Power

and Kleeman, 1994; Seager et al., 1995).

Because the ocean is driven, to first order, from its top boun-

dary – the conditions of which are poorly known in the past –

paleocean modeling requires that either these surface conditions

be quantitatively estimated, or in the best case, objectively prog-

nosed within a coupled ocean-atmosphere model. There are,

furthermore, only weak constraints on paleobathymetry, an

important boundary condition for paleo-oceanographic modeling

(Crowley, 1998), and vertical and horizontal ocean current

velocities that might be used as a test of model predictions are

poorly known. Thus, there are two sets of problems – even given

a perfect ocean model. How does one: (a) set up the model experi-

ment, given significant sources of uncertainty, and (b) test model

predictions if significant mismatches occur between what models

predict and what proxies record? We begin with (a).

Setting up deep paleocean model experiments

Beyond the normal challenges associated with ocean model-

ing (e.g., Gent et al., 1998), deep past paleoceanographic

simulations face three challenges: uncertainty introduced by

choice of initial conditions, boundary conditions, and model

spin up.

Initial conditions. Proxies that might be used as initial con-

ditions for modeling experiments are sparsely distributed in

time and space, have inherent uncertainties associated with

them, and may not constrain important quantities (e.g., sali-

nity). This challenge is usually overcome by specifying an

initial sea surface temperature (SST) distribution derived from

proxies or using output from an atmospheric general circulation

model (AGCM), which in turn has predicted SSTs, assuming

some distribution of oceanic mixed layer depth and heat trans-

port. Surface salinities are normally based on output from such

AGCM simulations after assuming an empirical relationship

between precipitation minus evaporation (P–E) distributions

and surface salinity, or alternatively, a global constant value

can be specified as an initial condition. In paleoceanographic

simulations carried out with a coupled ocean-atmosphere

model, these initial conditions values may strongly deviate as

a consequence of the model’s evolution. In simulations that

are not coupled to an atmospheric model, these distributions

are more or less fixed (discussed below). Deep ocean tempera-

tures are normally taken from benthic foraminifera-derived

temperature estimates (where available) and salinity is fre-

quently set at the modern global mean value or sometimes set

to a different global mean value designed to include past varia-

tions in global mean salinity.

Boundary conditions. These include the way in which

heat, moisture, and momentum fluxes are passed to the ocean

model throughout the simulation as well as static features such

as bathymetry. Fluxes have traditionally been handled through

the use of so-called Haney restoring (Haney, 1971) of tempera-

ture and the specification of net virtual salt fluxes, i.e., mixed

boundary conditions. In general, the “weaker” the restoring

the more closely this boundary condition mimics atmospheric

interaction, but mixed boundary conditions have been shown

to be susceptible to spurious multiple equilibria (Weaver and

Sarachik, 1991; Zhang et al., 1993; Power and Kleeman,

1993, 1994; Saravanan and McWilliams, 1995). The use of

so-called bulk forcing (Brady et al., 1998; Doney et al.,

1998), or preferably inclusion of even a simple representation

of atmospheric fluxes as in Nong et al. (2000) and Najjar et al.

(2002), is best in those cases in which a coupled model is not

used. Surface wind stress patterns are normally specified based

on modern conditions or a theoretical distribution, or produced

by an AGCM.

Since a non-interactive atmosphere – i.e., one in which

winds, temperature, and freshwater fluxes cannot adapt to chan-

ging ocean conditions – is inherently unrealistic, in most cases a

coupled model is preferable to an uncoupled model, even when

the focus is strictly oceanographic (Doney et al., 1998). This also

allows for a more realistic pattern of surface state and flux varia-

tions, and for OHT to be a well-posed prognostic variable, which

684 PALEOCEAN MODELING

it is not if surface temperature patterns are specified (Seager

et al., 1995;Nongetal.,2000; Huber et al., 2003).

Ocean bathymetry is a boundary condition that has fre-

quently been a subject of concern in the construction of paleo-

cean modeling experiments (Bice et al., 1998). Some studies

have employed a flat bottom configuration, eschewing bathy-

metric variation altogether (Barron and Peterson, 1991; Bush

and Philander, 1997) and dynamical arguments suggest that

most ocean current patterns and transports are not sensitive to

bathymetric variations (summarized in Bush, 1997). Neverthe-

less, the details of some currents certainly involve bathymetry

(Bice et al., 1998). Perhaps more importantly, the degree of

upwelling from the abyss is governed to a large extent by the pre-

sence of abyssal bathymetric rises (e.g., geostrophically balanced

meridional flows along ocean ridges) and thus some representa-

tion is probably important even for the gross features of the ocean

circulation (e.g., Toggweiler and Samuels, 1995;Vallis,2000).

The freshwater (salt) forcing of the ocean is another poten-

tially important variable, with potential issues being the place-

ment of continental runoff into marginal seas and determination

of the overall meridional P–E. The former issue is important

for determining the exact location of deep water formation (Bice

et al., 1997) and the latter is a factor in determining strength of

the MOC. However, neither of these issues proves fatal to the

study of paleoclimates with ocean models (see below).

Deep ocean spin up procedure. Another difficulty with per-

forming paleocean modeling experiments has been in integrating

the entire system, including the deep ocean, to a quasi-steady

state. Nevertheless, ocean models – at least the non-eddy resol-

ving models employed in climate and paleoclimate studies – are

computationally inexpensive relative to their atmospheric cou-

sins. In ocean-only ocean general circulation model (OGCM)

studies, it is now feasible to integrate the model to equilibrium

without so-called deep ocean acceleration, although accelerating

the deep ocean tracer fields during spin up has been shown to

speed equilibration without unduly affecting the final climate

state (Danabasoglu et al., 1996, 2004).

However, as described above, a coupled model is preferred

for most climate purposes, so that surface temperature, salinity

fields, ocean heat, and freshwater transport are fully prognostic.

Many coupled models historically display a “drift” of climate

away from modern values even under modern conditions, which

have precluded them for use in paleoclimatology, but this model-

ing challenge has been overcome in the recent generation of

models. It has not been feasible with existing computational

resources to attempt the ideal solution with a coupled general cir-

culation model (CGCM), which is to integrate several multi-mil-

lennial simulations. Two options exist: (a) using a simplified

coupled model that is computationally less expensive, or (b)

using a technique to accelerate convergence of a full CGCM.

Many simplified CGCMs or “intermediate” complexity

models now exist, and in general, they are a useful tool in the

modeling toolbox (Saravanan and McWilliams, 1995; Poussart

et al., 1999; Weaver et al., 2001; Claussen et al., 2002). When

these tools are not sufficient, iterative accelerated spin up tech-

niques have been developed to allow several thousand years of

deep ocean simulation to be carried out in only several hundred

years of surface ocean computation using fully coupled GCMs.

There is a variety of techniques – most draw on two existing

procedures: iteration of coupled and uncoupled modeling steps,

and deep ocean acceleration. The iteration technique has been

used and validated in somewhat different forms by Sausen

and Voss (1996), Kutzbach and Liu (1997), Liu et al. (1999),

Vavrus et al. (2000), Huber and Sloan (2001), Otto-Bliesner

et al. (2002), and Liu et al (2004). Shin et al. (2003) describe

a different technique that employs deep ocean acceleration

directly in the fully coupled integration and this has been

shown to produce convergent climates with the iterative tech-

nique in the Eocene, thus either family of techniques is a

potential candidate for use. There are no guarantees that a

given acceleration technique will apply equally well to all time

intervals or that it can be generalized for different models or

model configurations. Therefore, initial validation of the con-

vergence of a given modeling framework to a reference condi-

tion (presumably the present) to some appropriate level of

similarity is a primary requirement of any technique.

Minimal tests of model suitability. Although the models

used in paleoceanography are usually well-validated for modern

day conditions, even a “perfect” model for the present may not

be perfect for the correct physical reasons (instead, tuning may

be responsible), nor nearly as accurate with less accurate

boundary conditions. Our incomplete knowledge of boundary

and initial conditions for paleocean modeling, simplification

of model physics, and/or use of acceleration techniques intro-

duces errors into the simulation. Therefore, paleocean model-

ing investigations should be tailored to answering questions

that can be answered with the level of detail available for the

specified time interval. One method to assess the suitability

of a model is a process called “degradation” (Huber and Sloan,

2001). This involves simplifying boundary conditions in the

control case (i.e., changing bathymetry) and appropriate

“detuning” of the model. A degraded control solution provides

a benchmark to determine what model-predicted variables

might potentially be robustly determined in the past. Degrada-

tion has been shown to provide one (minimum) estimate of

the errors introduced into model predictions by misrepresenting

multiple aspects of the system (Huber et al., 2003).

Modeling choices. No model is perfect, and many parame-

terizations and models in existence are better for some purposes

than others. Nevertheless, there are some general guidelines, at

least with complex, GCM caliber models, that should be con-

sidered. Oceanic vertical diffusion is one of the most important

and most contentious issues in modern oceanography. It is

likely to have been even more important in the past. The abso-

lute “background” value of the diffusivity and the physical pro-

cesses that increase it above this background value are much

debated. While some spread exists for an estimate of this back-

ground value, a reasonable observed number is 0.1 cm

1

, and

values in this range should be the default in paleoceangraphic

applications. Values 10–100 times greater than this are com-

monly used even in modern simulations, which highlights the

deficiencies in those models. The inclusion of state-of-the-art

treatments of diapycnal and isopycnal eddy mixing is impor-

tant, and currently it appears that a combination of the KPP

and Gent-McWilliams treatments of these processes is best

from the point of view of the tracer circulation and tracer prop-

erties (Gent et al., 1998). There is substantial room for

improvement in these parameterizations and ideally several

vertical diffusion methods should be tested because the result-

ing ocean circulations can be fundamentally different (e.g.,

Nilsson et al., 2003).

Predictions of paleocean modeling

Horizontal ocean currents. These are prognosed by OGCMs

as well as a host of simplified models, and are the quantities

most directly tied to the well-understood and well-represented

PALEOCEAN MODELING 685

dynamics of the ocean. In general, an OGCM is not necessary

to predict velocities. The surface and upper ocean flows largely

reflect the driving wind fields, and consequently “shallow

water” ocean models, or even analytic calculations such as

the “island rule” or the Sverdrup transport allow estimates of

the strength, shape, and disposition of gyre circulations as well

as the western boundary currents in equilibrium with them for a

given wind field (Huber and Nof, 2006; Nof, 2000; Nof and

Van Gorder, 2002; see Figure P6). In the deep ocean, OGCM

results have mainly reproduced features that were adducible

from Stommel-Arons theory (see results in Huber et al.,

2003). Such predictions may be verified with records of

paleocirculation intensity and direction.

Unfortunately, current directions are weakly constrained and

current magnitudes are among the most poorly known features

in “deep-time” paleoceanography, since these features are not

strongly constrained by direct proxies, and the indirect proxies

are qualitative (Carter et al., 1996; McGowran et al., 1997;

Watkins and Kennett, 1972). Nevertheless, a substantial

amount of work has been done in predicting paleocean current

directions and testing these against proxy interpretations. Some

notable studies include those in the Permian (Winguth et al.,

2002), the Cretaceous Tethyan current regime (Barron and

Peterson, 1991; Bush and Philander, 1997), and in the global

ocean (Poulsen et al., 1998, 2001), and on a regional scale, in

the Turonian seaway (Slingerland et al., 1996). For more recent

periods, attempts have been made to understand changes in

flow direction between the Atlantic and Pacific, i.e., during

the Miocene (Nof and Van Gorder, 2002; Nisancioglu et al.,

2003; Omta and Dijkstra, 2003), although explicit model-data

comparison has been lacking in these studies. The more wide-

spread use of sediment and biological transport and dispersal

models in such investigations would substantially assist in the

comparison to proxy records. Some first steps in that direction

have been taken in Huber et al. (2004).

Upwelling and productivity. Upwelling is a feature more

closely tied to the proxy record than ocean current direc-

tion because it is more directly linked to an observable quan-

tity (productivity and burial). Understanding how upwelling

patterns may have changed during the past can provide us

with clues to the nature of circulation changes (Toggweiler and

Samuels, 1995;Vallis,2000), primary productivity shifts (Moore

et al., 2004), and alterations in global cycling of chemical

species that may have occurred. Upwelling predictions may be

especially useful for placing constraints on organic carbon burial

rates (Handoh et al., 1999, 2003) over these time intervals.

For the purely wind-driven component of upwelling, an

OGCM is not particularly necessary; a “shallow water” model

is adequate to the task (Handoh et al., 1999, 2003). Calculation

of Ekman pumping from the wind fields produces nearly iden-

tical results.

Quantitatively tying the upwelling velocity to a sediment

accumulation rate involves a deep understanding of many eco-

logical and biogeochemical processes. Currently, our ability to

make quantitative predictions on productivity, burial, and the

carbon cycle in the past is more limited by our understanding

of past biogeochemical cycles and paleoecology than by

insights into past physical oceanography (more below).

Deep water formation and MOC

Transformation of surface waters into shallow mode and inter-

mediate and deep waters, and subsequent recycling of these

waters into the surface, is an important and complex process.

It is fundamental to the redistribution of heat, nutrients, and

other tracers in the ocean and is consequently of great interest

in paleoceanography.

Figure P6 Results from Huber et al. (2004) for the theoretical wind-driven and general circulation model derived ocean circulations. The estimated

wind-driven ocean gyre transport in equilibrium with the AGCM-derived winds calculated from the Sverdrup transport driven by the wind stress

curl. Theory suggests to first order that barotropic velocities should correspond closely to the Sverdrup transport. Right, this is confirmed by

comparison to the barotropic stream function produced by the ocean model. Clockwise gyres are indicated in blue, counterclockwise gyres in red.

The flow is contoured at 10 Sverdrup intervals. The correspondence between the upper ocean velocities shows clearly that – if wind fields are well-

known, or simply prescribed – the major ocean current systems will be entirely governed by this distribution.

686 PALEOCEAN MODELING

Just as there are weak data-based constraints on ocean velo-

city, there are weak constraints on its integral properties, such

as the meridional overturning streamfunction. One commonly

applied constraint is the calculation of vertical and spatial

gradients of d

C, which is an isotopic tracer related to nutrient

distribution and the carbon cycle (e.g., Corfield and Norris,

1996). Where sufficient d

C gradients exist, the location of

regions of deep water formation can be ascertained and bulk

flow patterns may be estimated. The use of other water mass

tracers, such as neodymium (e.g., Thomas, 2004), is a recent

innovation that may add significant detail to the picture that

has developed from d

C tracer distributions.

Rates of deep water formation and deep water flow are not

currently predictable from these methods, which is unfortunate

because these are the quantities that bear directly on the critical

paleoceanographic questions, whereas the location of deep

water formation reveals little, by itself, about the ability of

the circulation to transport heat or nutrients.

The rate at which heat is transported by circulation is a func-

tion of the velocities, diffusion, and temperature gradients, and

analogous balances obtained for the transport of all tracers.

Thus the main importance of the location of deep water forma-

tion, i.e., whether it occurs at high latitudes or at low latitudes,

is indirect, through its relationship with the resulting transport

rates and tracer gradients. Some debate exists as to whether true

deep water formation at low latitudes was physically possible

in the Earth’s past – results and interpretations vary from

model to model (compare Saravanan and McWilliams, 1995;

Bice and Marotzke, 2002; Zhang et al., 2002). In general, even

in simulations in which low latitude deep water formation

occurs, it is either weak or transient, and is associated with a

decrease in OHT (see below).

Proxy records of benthic oxygenation have been employed as

constraints on overturning circulation rates in the oceans. These

records are difficult to interpret because they involve deconvolu-

tion of the competing influences of the supply of organic matter

and oxygen to the deep. Both factors may increase as the over-

turning rate increases – increased oxygen supply occurs as the

overturning strengthens, but increased upwelling of nutrient rich

waters may enhance export of organics out of the surface and

into the deep. For a revealing view of the divergences of inter-

pretation that can occur , compare Zhang et al. (2001)and

Hotinski et al. (2001).

Without improvement of our understanding of past biogeo-

chemical cycles and the development of multiple and indepen-

dent proxies for upwelling, productivity, oxygenation, and deep

water formation, it is unlikely that it will be possible to discri-

minate between different models of the ocean’s overturning cir-

culation in the deep past.

Ocean gateways and heat transport

Past climate changes have been frequently attributed to ocean

gateway changes – such as the opening of Drake Passage and

closing of the Panamanian Gateway (Kennett, 1977 ; Crowley

1998; Lawver and Gahagan, 1998). This belief finds its support

in the approximate correlation between major gateway changes

and climatic change (Zachos et al., 2001). A number of model-

ing studies have demonstrated that the ocean’s overturning

circulation, heat transport, and resulting temperature distribu-

tion are somewhat sensitive to changes in ocean gateways

(Mikolajewicz et al., 1993, 1997). There is, furthermore, a

strong line of theoretical and modeling evidence that backs

up the theory that wind-driven upwelling of deep water in the

Southern Ocean plays a dominant role in modern day ocean

circulation (Toggweiler and Samuels, 1995; Nof, 2000) and

that past changes in this upwelling may have been especially

important at both high latitudes (Toggweiler and Bjornsson,

2000) and low latitudes (Hotinski and Toggweiler, 2003). The

most frequently cited example of a direct impact of gateway

changes on OHT is due to the creation of the Antarctic Circum-

polar Current (ACC). As noted in Toggweiler and Bjornsson

(2000), creation of the ACC leads to a cooling of the Southern

Hemisphere, and a warming of the Northern Hemisphere,

because the flows associated with the ACC “steal” heat across

the equator.

While all of the above studies yield similar predictions with

respect to the physical oceanographic responses to gateway

changes, the surface temperature response, i.e., the climatic

impact of this change shows a wide spread. In studies in which

a simple representation of coupling of ocean-atmosphere

exchanges has been included, opening and closing of Southern

Ocean gateways result in Southern Ocean surface temperature

changes from a low value of 0.8



C (Mikolajewicz et al.,

1993), to a middle range of 1.5



C (Nong et al., 2000), and a

high value of 3.5



C (Toggweiler and Bjornsson, 2000; Najjar

et al., 2002). When a CGCM is used, the resulting change

within the Southern Ocean is in the middle of the range of pre-

vious estimates (Huber and Sloan, 2001; Huber et al., 2003).

The spread of values represents differences in the treatment

of the “top” boundary condition, including sea ice feedbacks

and damping by the atmosphere, rather than representing an

important difference in predicted ocean dynamics behavior.

The mechanism that links gateways to climate change is

usually presumed to be ocean heat modifications induced by

shifts in ocean currents. The impacts that such ocean current

changes have had on ocean productivity and burial and hence

global inventories of atmospheric GHGs have received less

attention in comparison (Huber et al., 2003).

Some lessons learned

Paleocean modeling has been applied to a suite of paleoclimate

problems in Earth’s history – mostly, it has focused on the

leading role played by changes in greenhouse gas concentra-

tions. Results from this work indicate that, within a sophisti-

cated coupled modeling framework, ocean temperatures in

line with proxies for both past “greenhouse” and “icehouse”

intervals are generally achievable without major changes in

ocean circulation, but are driven by changes in deep water tem-

peratures (Huber and Sloan, 2001; Otto-Bliesner et al., 2002;

Poulsen et al., 2002; Shin et al., 2002; Peltier and Solheim,

2004). See Figure P7 for a Cretaceous example.

The main current limitation of physical paleocean modeling

in paleoclimatology is the availability and accuracy of data for

inclusion in initial and boundary conditions and the need for a

suite of proxies sufficient to verify and validate the model pre-

dictions. Fully coupled modeling offers the advantage of

increasing the range of model predictions testable against

proxies (Huber and Caballero, 2003). Modeling of biological,

isotopic, ecological, and geochemical processes for the deep

past lags behind physical modeling by a significant margin

and the current sophistication of these efforts may be consid-

ered rudimentary, despite significant efforts to improve the

situation (e.g., Heinze and Crowley, 1997). The next main chal-

lenge of paleoceanographic modeling is to close the gap

PALEOCEAN MODELING 687

between the physics of the ocean and the biology, sedimentol-

ogy, and chemistry that provides key constraints on the paleo-

climatic record as well as being of fundamental importance to

paleoclimate itself.

Matthew Huber

Bibliography

Barron, E.J., and Peterson, W.H., 1991. The Cenozoic ocean circulation

based on ocean general circulation model results. Palaeogeogr. Palaeo-

climatol. Palaeoecol., 83,1–28.

Bice, K.L., Barron, E.J., and Peterson, W.H., 1997. Continental runoff and

early Cenozoic bottom water sources. Geology, 25, 951–954.

Bice, K.L., Barron, E.J., Peterson, W.H., 1998. Reconstruction of realistic

Early Eocene paleobathymetry and ocean GCM sensitivity to specified

ocean basin configuration. In Crowley, T.J., and Burke, K.C. (eds.),

Tectonic Boundary Conditions for Climate Reconstructions. New York:

Oxford University Press, pp. 212–226.

Bice, K.L., and Marotzke, J., 2002. Numerical evidence against reversed

thermohaline circulation in the warm Paleocene /Eocene ocean.

J. Geophys. Res., 106, 11529–11542.

Brady, E.C., DeConto, R.M., and Thompson, S.L., 1998. Deep water for-

mation and poleward ocean heat transport in the warm climate extreme

of the Cretaceous (80 Ma). Geophys. Res. Lett., 25, 4205–4208.

Bush, A.B.G., and Philander, S.G.H., 1997. The Late Cretaceous: Simula-

tion with a coupled ocean – atmosphere general circulation model.

Paleoceanography, 12, 495–516.

Bush, A.B.G., 1997. Numerical simulation of the Cretaceous Tethys

Circumglobal Current. Science, 275, 807–8.

Carter, R.M., Carter, L., and McCave, I.N., 1996. Current controlled sedi-

ment deposition from the shelf to the deep ocean: The Cenozoic evolu-

tion of circulation through the SW Pacific gateway. Geol. Rundsch.,

438–451.

Claussen, M., Mysak, L.A., Weaver, A.J., Crucifix, M., Fichefet, T.,

Loutre, M.F., Weber, S.L., Alcamo, J., Alexeev, V.A., Berger, A.,

Calov, R., Ganopolsky, A., Goosse, H., Lohman, G., Lunkeit, F.,

Mohkov, I.I., Petoukhov, V., Stone, P., and Wang, Z., 2002. Earth

System Models of Intermediate Complexity: Closing the gap in the

spectrum of climate system models. Clim. Dynam., 18, 579–586.

Corfield, R., and Norris, R.D., 1996. Deep water circulation in the

Paleocene Ocean. In Knox, R.W.O.B., Corfield, R.M., and Dunay,

R.E. (eds.), Correlation of the Early Paleogene in Northwest Europe.

Geological Society Special Publication, pp. 443–456.

Figure P7 Cretaceous paleotemperature estimates published from ODP and DSDP cores based on deep-dwelling fauna are sparse and exhibit a

wide range of values, but point to a much warmer deep ocean. For a transect in the Atlantic Ocean, the NCAR Climate System Model

predicts ocean temperatures below 2.5 km depth of approximately 10



C(bottom values) compared with proxy estimates of 6–15



C(top

values). The model simulation indicates that the source of the warmer deep ocean temperatures is sinking of warmer southern polar waters.

See Otto-Bliesner et al. (2002) for more details.

688 PALEOCEAN MODELING

Crowley, T.J., 1998. Significance of tectonic boundary conditions for

paleoclimate simulations. In Crowley, T.J., and Burke, K.C. (eds.),

Tectonic Boundary Conditions for Climate Reconstructions. New York:

Oxford University Press, pp. 212–226.

Danabasoglu, G., 2004. A comparison of global ocean circulation model

solutions obtained with synchronous and accelerated integration

methods. Ocean Model., 7, 323–341.

Danabasoglu, G., McWilliams, J.C., and Large, W.G., 1996. Approach

to equilibrium in accelerated global oceanic models. J. Climate, 9,

1092–1110.

Doney, S.C., Large, W.G., and Bryan, F.O., 1998. Surface ocean fluxes and

water-mass transformation rates in the coupled Climate System Model.

J. Climate, 9, 1420–1441.

Gent, P.R., Bryan, F.O., Danabasoglu, G., Doney, S.C., Holland, W.R.,

Large, W.G., and McWilliams, J.C., 1998. The NCAR Climate System

Model global ocean component. J. Climate, 11, 1287–1306.

Haney, R.L., 1971. Surface thermal boundary condition for ocean circula-

tion models. J. Phys. Oceanogr., 1, 241–248.

Handoh, I.C., Bigg, G.R., Jones, E.J.W., and Inoue, M., 1999. An ocean

modeling study of the Cenomanian Atlantic: Equatorial paleo-upwelling,

organic-rich sediments and the consequences for a connection between

the proto-North and South Atlantic. Geophys. Res. Lett., 26,223–226.

Handoh, I.C., Bigg, G.R., and Jones, E.J.W., 2003. Evolution of upwelling

in the Atlantic Ocean Basin. Palaeogeogr. Palaeoclimatol. Palaeoecol.,

202,31–58.

Heinze, C., and Crowley, T.J., 1997. Sedimentary response to gateway cir-

culation changes. Paleoceanography, 12, 742–754.

Hotinski, R.M., Bice, K.L., Kump, L.R., Najjar, R.G., and Arthur, M.A.,

2001. Ocean stagnation and end-Permian anoxia. Geology, 29,7–10.

Hotinski, R.M., and Toggweiler, J.R., 2003. Impact of a Tethyan circum-

global passage on ocean heat transport and “equable” climates.

Paleoceanography, 18, doi: 10.10.1029/2001PA000730.

Huber, M., Brinkhuis, H., Stickley, C.E., Doos, K., Sluijs, A., Warnaar, J.,

Schellenberg, S.A., and Williams, G.L., 2004. Eocene circulation of the

Southern Ocean: Was Antarctica kept warm by subtropical waters.

Paleoceanography, 19, PA4026.

Huber, M., and Caballero, R., 2003. Eocene El Niño: Evidence for robust

tropical dynamics in the “hothouse”. Science, 299, 877–881.

Huber, M., Sloan, L.C., and Shellito, C., 2003. Early Paleogene oceans and

climate: A fully coupled modelling approach using NCAR’s CSM. In

Wing, S.L., Gingerich, P.D., Schmitz, B., and Thomas, E. (eds.), Causes

and Consequences of Globally Warm Climates in the Early Paleogene.

Geolog. Soc. Amer. Special Paper, 369,25–47.

Huber, M., and Sloan, L.C., 2001. Heat transport, deep waters, and thermal

gradients. Geophys. Res. Lett., 28, 3481–3484.

Huber, M., and Nof, D., 2006. The ocean circulation in the Southern Hemi-

sphere and its climatic impacts in the Eocene. Palaeogeogr. Palaeoclim.

Palaecol., 231,9–28.

Kennett, J.P., 1977. Cenozoic evolution of Antarctic glaciations, the

circum-Antarctic ocean and their impact on global paleoceanography.

J. Geophys. Res., 82, 3843–3860.

Kutzbach, J.E., and Liu, Z., 1997. Response of the African monsoon to

orbital forcing and ocean feedbacks in the middle Holocene. Science,

278, 440–443.

Lawver, L.A., and Gahagan, L.M., 1998. The Initiation of the Antarctic

Circumpolar Current and its impact on Cenozoic climate. In

Crowley, T., and Burke, K. (eds.), Tectonic Boundary Conditions for

Climate Model Simulations. New York: Oxford University Press,

213–226.

Liu, Z.R., Gallimore, G., Kutzbach, J.E., Xu, W., Golubev, Y., Behling, P.,

and Selin, R., 1999. Coupled modeling of long-term climate change

with the equilibrium asynchronous coupling scheme. Clim. Dyn., 15,

325–240.

Liu, Z., Lewis, W., and Ganopolski, A., 2004. A coordinated acceleration

scheme for the simulation of long term climate evolution. Clim. Dyn.,

22, 771–781.

McGowran, B., Li, Q., Cann, J., Padley, D., McKirdy, D.M., and Shafik, S.,

1997. Biogeographic impact of the Leeuwin Current in southern

Australia since the late middle Eocene. Palaeogeog. Palaeoclimatol.

Palaeoecol., 136,19– 40.

Mikolajewicz, U., Maierreimer, E., Crowley, T.J., and Kim, K.Y., 1993.

Effect of Drake and Panamanian Gateways on the circulation of an

ocean model. Paleoceanography, 8, 409

–426.

Mikolajewicz, U., and Crowley, T.J., 1997. Response of a coupled ocean/

energy balance model to restricted flow through the Central American

isthmus. Paleoceanography, 12, 429–441.

Moore, T. C. Jr., Backman, J., Raffi, I., Nigrini, C., Sanfilippo, A.,

Palike, H., and Lyle, M., 2004. Paleogene tropical Pacific: Clues to

circulation, productivity, and plate motion. Paleoceanography, 19(3),

PA3013.

Najjar, R.G., Nong, G.T., Seidov, D., and Peterson, W.H., 2002. Modeling

geographic impacts on early Eocene ocean temperature. Geophys. Res.

Lett., 29, doi: 10.1029 /2001GL014–438.

Nilsson, J., Broström, G., and Walin, G., 2003. The thermohaline circula-

tion and vertical mixing: does weaker density stratification give stron-

ger overturning? J. Phys. Oceanogr., 33, 2781–2795.

Nisancioglu, K.H., Raymo, M.E., and Stone, P.H., 2003. Reorganization

of Miocene deep water circulation in response to shoaling of

the Central American Seaway. Paleoceanography, 18, doi:10.1029/

2002PA000767.

Nof, D., 2000. Does the wind control the import and export of the South

Atlantic? J. Phys. Oceanogr., 30, 2650–2667.

Nof, D., and Van Gorder, S., 2002. Did an open Panama Isthmus corre-

spond to an invasion of Pacific water into the Atlantic? J. Phys. Ocea-

nogr., 33, 1324–1336.

Nong, G.T., Najjar, R.G., Seidov, D., and Peterson, W.H., 2000. Simulation

of ocean temperature change due to the opening of the Drake Passage.

Geophys. Res. Lett., 27, 2689–2692.

Omta, A.W., and Dijkstra, H.A., 2003. A physical mechanism for the

Atlantic-Pacific flow reversal in the early Miocene. Glob. Planet.

Change, 788,1–12.

Otto-Bliesner, B.L., Brady, E.C., and Shields, C., 2002. Late Cretaceous

Ocean: Coupled simulations with the NCAR CSM. J. Geophys. Res.,

107, doi: 10.1029/ 2001JD000821.

Peltier, W.R., and Solheim, L.P., 2004. The climate of the Earth at Last Gla-

cial Maximum: statistical equilibrium state and a mode of internal varia-

bility. Quaternary Sci. Rev., 23, 335–357.

Poulsen, C.J., Seidov, D., Barron, E.J., and Peterson, W.H., 1998. The

impact of paleogeographic evolution on the surface oceanic simulation

and the marine environment within the mid-Cretaceous Tethys. Paleo-

ceanography, 13, 546

–559.

Poulsen, C.J., Barron, E.J., Arthur, M.A., and Peterson, W.H., 2001.

Response of the mid-Cretaceous global oceanic circulation to tectonic

and CO2 forcings. Paleoceanography, 16, 576–592.

Poulsen, C.J., Jacob, R.L., Pierrehumbert, R.T., and Huynh, T.T., 2002.

Testing paleogeographic controls on a Neoproterozoic snowball Earth.

Geophys. Res. Lett., 29, doi: 10.10129/ 2001GL014352.

Poussart, P.F., Weaver, A.J., and Barnes, C.R., 1999. Late Ordovician gla-

ciation under high atmospheric CO2: A coupled model analysis. Paleo-

ceanography, 14, 542–558.

Power, S.B., and Kleeman, R., 1993. Multiple equilibria in a global ocean

general circulation model. J. Phys. Oceanogr., 23, 1670–1681.

Power, S.B., and Kleeman, R., 1994. Surface flux parameterisation and the

response of OGCMs to high latitude freshening. Tellus, 46,86– 95.

Saravanan, R., and McWilliams, J.C., 1995. Multiple equilibria, natural

variability, and climate transitions in an idealized ocean-atmosphere

model. J. Climate, 10, 2296–2323.

Sausen, R., and Voss, R., 1996. Techniques for asynchronous and periodi-

cally synchonous coupling of atmosphere and ocean models. I. General

strategy and application to the cyclo-stationary case. Clim. Dyn., 12,

595–604.

Seager, R., Kushnir, Y., and Cane, M.A., 1995. On heat flux boundary con-

ditions for ocean models. J. Phys. Oceanogr., 25, 3219–3230.

Shin, S.I., Liu, Z., Otto-Bliesner, B., Brady, E.C., Kutzbach, J.E.,

Harrison, S.P., 2003. A simulation of the last Glacial Maximum Climate

using the NCAR-CCSM. Clim. Dyn., 20, 127–151.

Slingerland, R., Kump, L.R., Arthur, M.A., Fawcettt, P.J., Sageman, B.B.,

and Barron, E.J., 1996. Estuarine circulation in the Turonian

Western Interior seaway of North America. Geol. Soc. Am. Bull., 108,

941–952.

Thomas, D.J., 2004. Evidence for production of North Pacific deep waters

during the early Cenozoic Greenhouse. Nature, 430,65–68.

Toggweiler, J.R., and Samuels, B., 1995. Effect of Drake Passage on the

Global Thermohaline Circulation. Deep Sea Res., 42, 477–500.

Toggweiler, J.R., and Bjornsson, H., 2000. Drake Passage and paleocli-

mate. J. Quatern. Sci., 15, 319–238.

PALEOCEAN MODELING 689

Watkins, N.D., and Kennett, J.P., 1972. Regional sedimentary disconformi-

ties and upper Cenozoic changes in bottom water velocities between

Australia and Antarctica. Antarct. Res. Ser., 19, 317–334.

Weaver, A.J., and Sarachik, E.S., 1991. The role of mixed boundary condi-

tion in numerical models of ocean’s climate. J. Phys. Oceanogr., 21,

1470–1493.

Weaver, A.J., Eby, M., Wiebe, E.C., Bitz, C.M., Duffy, P.B., Ewen, T.L.,

Fanning, A.F., Holland, M.M., MacFadyen, A., Matthews, H.D.,

Meissner, K.J., Saenko, O., Schmittner, A., Wang, H., and

Yoshimori, M., 2001. The UVic Earth System Climate Model: Model

description, climatology and application to past, present and future

climates. Atmosphere Ocean, 39, 361–428.

Winguth, A.M.E., Heinze, C., Kutzbach, J.E., Maier-Reimer, E.,

Mikolajewicz, U., Rowley, D., Rees, A., and Zieger, A.M., 2002. Simu-

lated warm polar currents during the middle Permian. Paleoceano-

graphy, 17,4.

Vallis, G.K., 2000. Large-scale circulation and production of stratification:

Effects of wind, geometry, and diffusion. J. Phys. Ocean., 30, 933–954.

Vavrus, S., Gallimore, R., and Liu, Z., 2000. A mixed-flux equilibrium

asynchronous coupling scheme for accelerating convergence in ocean-

atmosphere models. Clim. Dyn., 16, 821–831.

Zachos, J., Pagani, M., Sloan, L.C., Thomas, E., and Billups, K., 2001.

Trends, rhythms, and aberrations in global climate 65 Ma to Present.

Science, 292, 686–693.

Zhang, S., Greatbach, R.J., and Lin, C.A., 1993. A reexamination of the polar

halocline catastrophe and implications for coupled ocean-atmosphere

modeling. J. Phys. Oceanogr., 23, 287–299.

Zhang, R., Follows, M., Grotzinger, J.P., and Marshall, J., 2001. Could

the late Permian deep ocean have been anoxic? Paleoceanography,

16, 317–329.

Zhang, R., Follows, M., and Marshall, J., 2002. Mechanisms of thermo-

haline mode switching with application to warm equable climates.

J. Clim., 15, 2056–2072.

Cross-references

Carbon Isotopes, Stable

Heat Transport, Oceanic and Atmospheric

Ocean Paleocirculation

Ocean Paleotemperatures

Paleoclimate Modeling, Pre-Quaternary

Paleoclimate Modeling, Quaternary

Paleoceanography

Plate Tectonics and Climate Change

Thermohaline Circulation

PALEOCEANOGRAPHY

Introduction

Paleoceanography is the study of the history of the oceans. It

encompasses aspects of oceanography, climatology, biology,

chemistry and geology. The main sources of information are

biogenic and inorganic marine sediments, as well as corals.

Biogenic sediment includes planktonic and benthic fossils

whereas inorganic sediment includes ice-rafted debris and dust.

On land, paleo-shorelines and erosional features as well as out-

croppings of paleomarine sediments are the principal sources of

proxy data. Glaciological records can also give indirect infor-

mation about paleoceans. The ocean’s high heat capacity and

its ability to transport energy and to sequester and release

greenhouse gases give it an important role in helping to deter-

mine the state of the planet’s climate. Thus, paleoceanographic

research is also intimately linked to paleoclimatology.

Methods

The reconstruction of paleocean characteristics and dynamics

requires climatic detective work. It involves the dating and

interpretation of paleoclimatic records as well as the definition

of physical and dynamical constraints which specify possible

circulation patterns and characteristics.

Reconstructions (Proxy Data)

Direct measurements of the quantities of interest to oceanogra-

phers extend only into the relatively recent past and in most

cases do not go further back than the mid-nineteenth century.

To study the ocean during periods for which there are no direct

measurements one must rely on indirect evidence. Historical

documents can be used as sources of data. Ship logbooks and

sailing times across frequently traveled routes have provided

estimates of the directions and strengths of past prevailing

winds (Brazdil et al., 2005). The frequency and intensity of

El Niño events since the 1500s have been reconstructed based

partially on historical accounts of large floods and crop losses

(Quinn and Neal, 1987). This type of analysis furnishes quali-

tative descriptions of the past.

Quantitative reconstructions are possible by proxy, where a

quantity which is preserved in a natural archive and can be

measured, stands as a surrogate of the parameter of interest.

A basic requirement is that the relationship between the proxy

parameter and the quantity of interest has to be known. When

this is the case, the history of the proxy variable can be con-

verted into the history of the variable of interest by the use of

mathematical expressions of the type:

Int

¼ f ðPrx

Þð1Þ

which state that the parameter of interest, Int, is a function

of the proxy quantity, Prx.Thet inde x refers to time.

Equations of this kind are commonly called transfer functions.

The confidence with which Int can be estimated will depend on

a series of factors, starting with the quality of the proxy mea-

surements. Also relevant, and a common source of uncertainty,

is how well f represents the relationship between Prx and Int.

In most cases, transfer functions are obtained empirically by

comparing directly measured values of the quantity of interest

to a pertinent set of proxy data. A potential source of error is

that the function obtained by this procedure might not be gen-

eral, but in fact could represent a relationship between Prx and

Int that is peculiar to the data sets used to generate f. This pro-

blem can be minimized by expanding the spatial and temporal

coverage of the data used to establish the transfer function.

Still, even assuming that f is a perfect representation of how

the proxy and the quantity of interest are connected to each

other in the present, there is no guarantee that the relationship

between them was the same in the past.

Another source of error can be easily understood by re-

writing Equation (1) so that it expresses the proxy quantity as

a function of the variable of interest. It is reasonable to assume

(and in many cases it has been demonstrated) that Int is not the

only factor controlling Prx, so that in fact we end up with:

Prx

¼ f

1

ðInt

; E

; :::; E

Þð2Þ

where E

, E

,..., E

represent environmental parameters that

also influence the proxy variable but are independent of the

quantity of interest. An immediate conclusion is that recon-

structions of Int based on Prx will be “contaminated” by other

690 PALEOCEANOGRAPHY