Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

(Imbrie and Imbrie, 1979). One of the findings of the CLIMAP

ocean reconstructions was that ice age tropical SSTs did not fall

greatly relative to the present. However, since then it has been

shown that tropical SSTs were considerably cooler than at present.

Continental reconstructions, however, lagged behind because of

the fragmental nature of the record, often of insufficiently high

quality. Indeed, land-sea correlations remain as one of the greatest

challenges.

Some of the earliest land-sea correlations were between

marine oxygen isotope stratigraphy and the loess sequences

of Central Europe and China. Ice ages were indicated by loess

deposition in arid environments and interglacials were indi-

cated by soil development. The long loess sequences in China,

as well as those offshore in the North Pacific, show that atmo-

spheric dust was an important component of the changing cli-

mate system.

Better and longer pollen sequences on the continents

questioned earlier precepts: for example, it was shown that

temperate forest did not exist during the last ice age in the

Mediterranean but instead steppic plant communities were

extensive, with temperate forest survival in Balkan and Greek

plant refugia. More significantly, much of the tropics and

sub-tropical regions were shown to be arid rather than, as pre-

viously thought, wetter places. This compelled a review of the

climate system and the role of the Hadley-Walker cells, the

trade winds and Intertropical Convergence Zone (ITCZ) in a

drier and cooler ice age world. The pollen record from some

long lake sequences, for example in Greece, Colombia and

Lake Baikal in Asia, matched the deep sea record and con-

firmed that Pleistocene vegetation patterns pulsed primarily to

the orbital beat as the result of changing seasonality and insola-

tion. The 41,000-year obliquity forcing was dominant at mid to

high latitudes, especially in influencing ice sheet growth and

decay, whereas precession (23,000 and 19,000 years) was

dominant at lower latitudes.

Although much is now known about the last glacial cycle,

aided to a large extent by the availability of

C ages for the

past 40,000 years, relatively little is known in detail before

the last interglacial (oxygen isotope sub-stage 5e or event 5.5

of the marine record). Currently some emphasis is placed on

oxygen isotope stage 11, the interglacial some 400,000 years

ago, because of its orbitally similar configuration to the current

(Holocene) interglacial (Droxler et al., 2003). However, rela-

tively little is known about major continental areas during other

ice ages and interglacials.

Millennial cycles

Although the evidence for sub-orbital pacings had been long

available from the Camp Century drill hole in Greenland

(Dansgaard et al., 1969), it was not until additional ice cores

were recovered that their millennial time scale records were

appreciated, especially the 1,450 or 1,500 year pacing (Alley,

2000; Bond et al., 2001). These Dansgaard-Oeschger (D-O)

cycles, with the warmer events called interstadials and colder

ones stadials, are grouped into saw-toothed packages of

increasing cold that are terminated by rapid warming events.

Rapid changes in temperatures at the surface of the ice sheets

occurred in less than a decade. The pattern of square-wave cli-

mate changes shown by the ice core records is matched by evi-

dence from high sedimentation cores in the North Atlantic of

SSTs inferred from the polar-dwelling foraminifer Neoglobo-

quadrina pachyderma (s.) and oxygen isotope variability

(Bond and Lotti, 1995). The evolution of these cycles has been

Figure P73 Four records showing millennial time scale variability over the last glacial cycle. The Greenland record of oxygen isotope variability

shows the last 10,000 years of the present (Holocene) interglacial with which the warmer climates of the last glacial cycle (interstadials) alternate

with colder ones (stadials). The offshore Cariaco Basin records influxes of minerals (high reflectance) from South America during stadials. In Italy

and Florida, the changing vegetation, inferred from the fossil pollen record, shows similar variability, including cooler climates during Heinrich

events (Overpeck et al., 2003).

PLEISTOCENE CLIMATES 801

attributed to oscillations in the strength of heat delivery by the

Atlantic thermohaline circulation that caused changing tem-

peratures at the ice sheet surface (Alley, 2000). Others have

pointed to a solar connection because oxygen isotope variabil-

ity in ice cores corresponds with similar variability in the con-

centration of the cosmogenic isotope

Be, which indicates

changes in solar irradiance (Rind, 2002). Corresponding with

the rapid warming at the end of Bond Cycles (Bond and Lotti,

1995), layers of ice-rafted debris (IRD), representing “Heinrich

events” occur in marine cores. Deposited from “armadas” of

icebergs discharged into the North Atlantic (Broecker, 1994;

Bond and Lotti, 1995), IRDs indicate collapse and surging of

ice sheets into the tidewater zone. Because they do not corre-

spond with every D-O package of events, it is likely they were

partly forced by climate, but only occurred when appropriate

glaciological conditions for collapse and surging existed.

Greenland and Antarctic ice cores contain bubbles of air

from which the near-contemporary concentrations of atmo-

spheric greenhouse gases (CO

and CH

) can be measured

(Alley, 2000). Spikes of CH

provide a means of correlation

of the oxygen isotope temperature record between Greenland

and Antarctica and allow phase relationships between the

Northern and Southern Hemisphere polar regions to be

explored. The pattern of greenhouse gas variability pulsed on

both orbital and millennial time scales. When correlated

with other evidence, they confirm previous inferences of

a cooler and more arid Earth during the ice ages, with relatively

lower atmospheric concentrations (180 ppm) of carbon diox-

ide, and warmer interglacial conditions with relatively higher

( 280 ppm) concentrations.

Further evidence that all the Earth’s systems pulsed at mil-

lennial timescales is available from a wide range of environ-

ments that show that atmospheric patterns changed rapidly:

for example, in the lake systems of western North America,

and monsoonal climates in India, Asia and South America

(Clark et al., 2001). Figure P73 (Overpeck et al., 2003) shows

the GISP2 (Greenland Ice Sheet Project 2) d

O temperature

signal with packages of D-O pacings that end with a rapid

warming. These pacings at 1,450 years continue, with reduced

amplitude, into the present (Holocene) interglacial (Bond et al.,

2001). The reflectance characteristics of sediments from the

Cariaco Basin, offshore of Venezuela, record major changes

in the hydrologic systems of the tropical Atlantic. Increased

precipitation and river discharge from northern South America

Figure P74 Greenland and Antarctic ice core records and Loess in China show millennial pacing for variability in temperature,

carbon dioxide, methane, nitrous oxide and the East Asian monsoonal over the past 50,000 years (from Raynaud et al., 2003).

802 PLEISTOCENE CLIMATES

correspond with interstadial (warmer) events, which suggests

higher moisture fluxes from the Atlantic to the Pacific, inter-

spersed with lower precipitation (light colored sediments) as

the Intertropical Convergence Zone migrates south during

colder stadials. Note the rapid changes in vegetational stability

in Italy and Florida that are correlated with D-O events.

Figure P74 shows millennial scale changes from different

sources (Raynaud et al., 2003). From top to bottom: tempera-

ture variability as shown by the Greenland (GRIP) oxygen iso-

topic record; record of the Asian summer monsoon from

magnetic susceptibility of Chinese loess, peaks indicate times

of strong monsoons; temperature variability as shown by the

Antarctic (Byrd station) oxygen isotope record; the methane

record from Greenland; the N

O record from Greenland; and

the CO

record from Antarctica (Byrd station). Note that the

Greenland and Antarctic records have been correlated by global

methane spikes. Numbers 1–13 indicate Dansgaard-Oeschger

(D-O) events in the Greenland record, while H1–H5 indicate

Heinrich events.

Clearly, different sub-systems that were widely separated

pulsed at the 1,450 pacing. Millennial variability at a global

scale shows that climate signals are transmitted rapidly

throughout the entire system. However, is this compatible with

heat distribution by ocean thermohaline circulation? To what

extent does the atmosphere play a role in the advection of heat

and moisture from low latitude tropical oceans, notably the

Pacific Ocean (Cane and Clement, 1999)?

Sub-millennial changes

Some of the answers to the questions raised above may

come from records of sub-millennial changes in climate on

inter-decadal or inter-annual timescales of physical annular

modes of variability. These include records of the transfer of

heat and moisture by Indian, African and Asian monsoons, El

Niño (ENSO), North Atlantic Oscillations (NAO), Arctic

Oscillation and Antarctic Oscillation. These are all intercon-

nected in transporting heat from tropical oceans to higher lati-

tudes. Their teleconnections are complex and may be driven

in part by variability in solar irradiance as well as eccentri-

city-modulated precession. Cane and Clement (1999) have sug-

gested that scaling El Niño (ENSO) events may well provide

explanations for variability in large-scale Pleistocene climates.

This may also account for the “stage 11 problem,” the appar-

ently disproportionate response of the climate system to weak

eccentricity forcing. Perhaps low eccentricity may dampen pre-

cessional variability and concentrates heat in the low latitude

tropical belt from where it is advected polewards.

Although there seems little doubt that discharges of fresh-

water from the Laurentide Ice Sheet influenced thermohaline

activity in the North Atlantic Ocean during ice ages (Clark

et al., 2001), it is otherwise undeniable that the concept of the

North Atlantic, Greenland and Norwegian Seas’ thermohaline

influence in redistributing heat is under challenge from the tro-

pical Pacific as the main source of heat and moisture. Thus,



N, the “critical latitude” of Milutin Milankovitch, and the

ocean thermohaline circulation climate signal carrier is chal-

lenged by data from the tropical Pacific Ocean.

This debate has rejuvenated the climatologic-paleoclimate

link in the search for mechanisms of abrupt and rapid change

for Pleistocene climates. Increasing interest in solar irradiance

variability is the result of ongoing physical measurements as

well as the record of variability inferred from cosmogenic iso-

tope abundances in ice cores. The not-unrelated role of cosmic

particles may also have a bearing on the elusive role of cloud

and water vapor in climate variability on different timescales.

Concepts of Pleistocene climates have been revolutionized in

less than half a century and, while the Pleistocene past may

hold a key to the present and future, its reciprocal, of the pre-

sent being the key to the past, remains undiminished.

David Q. Bowen

Bibliography

Alley, R.B., 2000. The Two-Mile Time Machine: ice cores, abrupt climate

change, and our future. Princeton: Princeton University Press, 229pp.

Bond, G.C., and Lotti, R., 1995. Iceberg discharges into the North Atlantic

on millennial timescales during the last glaciation. Science, 267,

1005–1010.

Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W.,

Hoffmann, S., Lotti-Bond, R., Hajdas, I., and Bonani, G., 2001. Persis-

tent solar influence on North Atlantic climate during the Holocene.

Sience, 294, 2130–2136.

Broecker, W.S., 1994. Massive iceberg discharges as triggers for global cli-

mate change. Nature, 372, 421–424.

Broecker, W.S., and Denton, G.H., 1989. The role of ocean-atmosphere

reorganizations in glacial cycles. Geochim. Cosmochim. Acta, 53,

2465–2501.

Cane, M., and Clement, A.C., 1999. A role for the Tropical Pacific Coupled

Ocean-Atmosphere System on Milankovitch and Millennial Timescales:

Part II: Global Impacts. In Clark, P.U., Webb, R.S., and Keigwin, L.D.

(eds.), Mechanims of Global Climate Change at Millennial Time Scales.

Washington, DC: American Geophysical Union, pp. 373–384.

Clark, P.U., Marshall, S.J., Clarke, G.K.C., Hostetler, S.W., Licciardi, J.M.,

and Teller, J.T., 2001. Reshwater forcing of abrupt climate change dur-

ing the last glaciation. Science, 293, 283–287.

Cronin, T. M., 1999. Principles of Paleoclimatology, New York: Columbia

University Press, 560pp.

Dansgaard, W.S., Johnsen, S.J., Møller, J., and Langway, C.C., 1969.

One thousand centuries of climatic record from Camp century on the

Greenland Ice Sheet. Science, 166, 377–380.

deMenocal, P., 1995. Plio-Pleistocene African climate. Science, 270,

53–59.

Droxler, A., Poore, R.Z., and Burkle, L.H., 2003. Earth’s Climate and

Orbital Eccentricity: The Marine Isotope Stage 11. Washington, DC:

Geophysical Monograph, American Geophysical Union.

Imbrie, J., and Imbrie, K.P., 1979. Ice Ages: Solving the Mystery. London:

Macmillan, 224pp.

Overpeck, J.T., Whitlock, C., and Huntley, B., 2003. Terrestrial biosphere

dynamics in the climate system: Past and future. In Alverson, K.D.,

Bradley, R.S., and Pederson, T.F. (eds.), Paleoclimate, Global Change

and the Future. Berlin: Springer, pp. 82–103.

Rind, D., 2002. The Sun’s Role in Climate Change. Science, 296, 673

–677.

Raynaud, D., Blunier, T., Ono, Y., Delmas, R., Barnola J.M., Joos, F.,

Petit, J.R., and Spahni, R., 2003. The late Quaternary history of atmo-

spheric trace gases and aerosols: Interactions between climate and bio-

chemical cycles. In Alverson, K.D., Bradley, R.S., and Petersen, T.F.

(eds.), Paleoclimate, Global Change and the Future. Berlin: Springer,

13–32.

Zahn, R., 2002. Milankovitch and Climate: the Orbital Code of Climate

Change. JOIDES J. (Special Issue), 28,17–22.

Cross-references

Aerosol (mineral)

Albedo Feedbacks

Alkenones

Arid Climates and Indicators

Astronomical Theory of Climate Change

Atmospheric Circulation During the Last Glacial Maximum

CLIMAP

Climate Change, Causes

Cosmogenic Radionuclides

Dansgaard-Oeschger Cycles

Eccentricity

PLEISTOCENE CLIMATES 803

Foraminifera

Glaciations, Quaternary

Heat Transport, Oceanic and Atmospheric

Heinrich Events

Human Evolution and Climate Change

Ice Cores, Antarctica and Greenland

Ice-Rafted Debris (IRD)

Interstadials

Lake Level Variations

Last Glacial Maximum

Laurentide Ice Sheet

Little Ice Age

Loess Deposits

Millennial Climate Variability

Moraines

Monsoons, Quaternary

North Atlantic Deep Water and Climate Change

North Atlantic Oscillation (NAO) Records

Obliquity

Ocean Paleotemperatures

Oxygen Isotopes

Paleo-El Niño-Southern Oscillation (ENSO) Records

Paleoceanography

Paleotemperatures and Proxy Reconstructions

Pliocene Climates

Pollen Analysis

Precession, Climatic

Quaternary Climate Transitions and Cycles

Quaternary Vegetation Distribution

Sea Level Change, Quaternary

SPECMAP

Stable Isotope Analysis

Sun-Climate Connections

Thermohaline Circulation

Time-Series Analysis of Paleoclimate Data

Transfer Functions

Uranium-Series Dating

Wisconsinan (Weichselian, Würm) Glaciation

PLIOCENE CLIMATES

Introduction

The Pliocene (Plio More – Cene Recent) epoch represents the

uppermost sub-division of the Tertiary period. It spans a time-

frame from 5.3 to 1.8 million years before present, according

to the geological timescale of Berggren et al. (1995). The epoch

incorporates the time interval in which Earth experienced a

transition from relatively warm climates to the prevailing

cooler climates of the Quaternary period.

Climatically, the Pliocene can be crudely divided into three

phases, (a) an Early Pliocene warm period, (b) a relatively

short-lived ‘warm blip’ during the middle part of the Pliocene

(centered around 3 million years before present), and (c) a cli-

matic deterioration during the late Pliocene. It is important to

realize that available higher resolution climate records for this

time (e.g., Leroy and Dupont, 1994) suggest that even rela-

tively short subdivisions of the epoch were characterized by

almost continuous changes in precipitation and temperature,

occurring on the timescale of several thousand years. This cli-

matic variability is poorly resolved in terms of magnitude

(Haywood et al., 2002a) but often occurred with detectable per-

iodicities of 41,000 and 19,000, 21,000 and 23,000 years,

linked to obliquity and precessional Milankovitch cycles (cyc-

lic variations of the Earth’s orbital position; obliquity is the

angle of Earth’s axial tilt as measured against the plane of the

ecliptic; precession is the degree of wobble of the Earth on its

axis, which is caused by the gravitational pull of the Sun and

the Moon). It is likely that many abrupt, short timescale climate

events currently remain undetected due to the lack of high reso-

lution and long paleoclimatic records, thus making the process

of generalization regarding the climate of the Pliocene extre-

mely hazardous.

Even though a progressive cooling was occurring during the

Tertiary, the Pliocene world appears to have been, for the most

part, warmer than at present. The ancient distribution of plank-

tonic (free-floating marine organisms) and benthic (bottom-

dwellers) foraminifer (minute aquatic organisms that consist

of a single cell or a colonial aggregate of cells without differen-

tiation of function) along with terrestrial fossil fauna and flora,

indicates that mean annual temperatures in the mid-latitudes

were often several degrees higher (3



Cto5



C) than at present

(e.g., Thompson, 1991; Dowsett et al., 1996). The greatest

warming appears to have been in the high latitudes and polar

regions, where temperatures were often warm enough to allow

species of animals and plants to exist at higher latitudes than

their nearest modern relatives (Adam, 1994).

The recognition that significant warming took place at high

latitudes and at the poles has interesting implications for the

behavior and extent of sea ice, the major terrestrial ice sheets

of Greenland and Antarctica, and for Pliocene sea levels.

Sea ice

Diatom (see Diatoms) data from the Southern Ocean (Barron,

1996), and foraminifer, ostracod (see Ostracodes) and mollusk

(invertebrate of the phylum Mollusca) data (Cronin et al.,

1993) suggest that sea ice extents were greatly reduced at

approximately 3 million years before present, with modern

summer sea-ice limits similar to Pliocene winter sea-ice cover

and the possibility that the Arctic Ocean was seasonally sea-

ice free (Cronin et al., 1993).

Ice sheets

Significant debate has centered on the nature of the terrestrial

ice sheets of Greenland and Antarctica during the Pliocene.

The consensus view, through examination of ice-rafted detritus

(IRD) in marine sediment cores (e.g., Kleiven et al., 2002), is

that the initiation of significant Northern Hemisphere glaciation

began during the middle to late Pliocene. However, there may

have been many false starts to this glaciation, with evidence

existing for glacial activity on Greenland as far back as the late

Miocene. Total ice volume on Greenland may have remained

significantly less than at present even during the latter parts

of the Pliocene epoch.

With regard to Antarctica, two contrasting viewpoints exist.

The first, based on evidence derived from Pliocene marine

diatoms (Harwood and Webb, 1986) and the remains of

Nothofagus (Southern Beech) in the Sirius Group glacial

sediments along the Transantarctic Mountain Range (Francis

and Hill, 1996), suggests that the East Antarctic Ice Sheet

(EAIS) may have been significantly reduced in size (by a max-

imum of one-third). The second, based on geomorphic investi-

gations of the landscape in the Dry Valleys region, suggests

that stable polar conditions have persisted for at least the last

14 million years (Denton et al., 1993). The dating of a variety

of landforms including superficial till sheets, old desert pave-

ments, avalanche deposits and volcanic ash layers that span

the Miocene and Pliocene apparently testify to remarkably slow

804 PLIOCENE CLIMATES

rates of weathering and erosion in the Dry Valleys (Marchant

et al., 1993). If the EAIS did behave in a dynamic manner, it

is unlikely that such geomorphic features, if truly Miocene or

Pliocene in age, would have survived.

Sea level

If terrestrial ice cover was reduced during the Pliocene, then glo-

bal sea levels should also have been higher than at present. Evi-

dence exists which indicates that sea levels were higher,

although the potential errors in the data make it difficult to assess

the true magnitude of sea-level change. Based on sequence stra-

tigraphic techniques, sea-level highstands of up to 60 m were

reported by Haq et al. (1987). Dowsett and Cronin (1990) used

the presence of the Orangeburg Scarp (an extensive paleoshore-

line from the southeastern U.S. Atlantic Coastal Plain) com-

bined with micropaleontological evidence to suggest that a

sea-level highstand of approximately 35 18 m higher than pre-

sent existed between 3.5 and 3.0 million years before present.

The stratigraphic distribution of disconformities (type of uncon-

formity where the buried erosion surface lies between two series

of strata that are parallel on a large scale) at Enewetak Atoll sug-

gests a Pliocene sea level 36 m higher than today’s (Wardlaw

and Quinn, 1991). Through the analysis and correlation of

Atlantic Coastal Plain sediments with d

O (see Oxygen iso-

topes) isotopic records from deep-sea cores, sea levels between

4.0 to 3.5 million years ago and 3.0 million years ago have been

estimated as 30–35 and 25 m higher than modern, respectively

(Krantz, 1991). These sea-level changes appear to have been dri-

ven by glacio-eustatic events (see Glacial eustasy), but their pre-

cise correlation to changes in the polar regions remains unclear

(Dowsett et al., 1996).

Climate dynamics and climate modeling

The use of numerical climate models, or General Circulation

Models (GCMs), is one of the most powerful tools at our

disposal for the investigation of past climate dynamics. Numer-

ous climate modeling studies have been conducted for the Plio-

cene using a host of different models of varying sophistication

(see Mid-Pliocene warming this volume; Chandler et al., 1994;

Sloan et al., 1996; Haywood et al., 2000a,b; Kim and Crowley,

2000; Haywood, 2001; Haywood et al., 2001; Haywood et al.,

2002a,b,c; Haywood and Valdes, 2004). The focus of these

modeling studies has been on the period of warming which

occurred around 3 million years before present. This is primar-

ily due to the wealth of geological and boundary condition data

sets (necessary for the initiation of the climate model experi-

ments) that exist for this period (see Dowsett et al., 1999).

Climate modeling – surface temperat ure (



Atmosphere-only climate model studies (experiments that do

not include an interactive ocean model in any form) suggest

that the climate of 3 million years before present was, as a

global annual average, 2 to 5



C warmer than at present (Sloan

et al., 1996; Haywood et al., 2000a; see Figure P75). Three

zones of temperature change were predicted in the climate

modeling study of Haywood et al. (2000a). Between 10



and



North and South of the equator, the predicted warming

was a modest 1 to 4



C. Between 30



and 50



in the Northern

Hemisphere, temperatures warmed by 4–8



C. This pattern

was not duplicated in the Southern Hemisphere, where the

temperature rise was 0–4



C. By far the greatest relative

warming was indicated at latitudes above 70



in both hemi-

spheres (Figure P76). Temperature changes relative to present

in high latitude areas of the Northern Hemisphere during the

winter period (December, January and February; DJF) ranged

from 16–20



C in Greenland to 20–28



C in the Arctic. During

DJF (Southern Hemisphere summer), dramatic warming in

parts of west and east Antarctica (~32–36



C) was predicted.

During June, July and August (JJA), Antarctica warmed in a

Figure P75 Climate model prediction: difference between simulated Pliocene (3 million years before present) and present-day surface temperature

(



C) as an annual mean (redrawn from Haywood et al., 2000a; Haywood, 2001). Model used was the Hadley Center Atmospheric Climate

Model Version 3 (HadAM3), boundary conditions specified from the PRISM2 data set (Dowsett et al., 1999).

PLIOCENE CLIMATES 805

comparable fashion to the DJF period (20–32



C). However,

the warming occurred over a significantly wider area of the

continent.

The pattern of temperature change was not spatially uniform.

During DJF, a zone of discontinuous cooling (maximum of 4



was predicted from 10



North to 10



South of the equator; this

extended over parts of northern Australia and South America.

During JJA, the equatorial/low latitude belt of cooling became

more continuous, especially over parts of Africa and southern

Asia. Temperatures in parts of Africa at 10



North of the equator

cooled by 4 and 8



C as compared to the present.

The predicted high latitude warming (>70



North and South)

can be attributed to a reduction of ice cover in both oceanic and

terrestrial settings during the Pliocene and to the prescribed war-

mer high latitude sea surface temperatures (SSTs). Furthermore,

the significance of a higher Pliocene sea level should not be

underestimated. Oceans cool less than land when exposed to

cold conditions, so higher sea levels act as a positive feedback

mechanism for temperature, especially in the high latitudes

(Sloan et al., 1996). The equatorial and low latitude cooling is

due to more than one contributing factor. SST values for the

oceans in equatorial and low latitudes are likely to have an

effect, as are variations in cloud cover and vegetation in these

areas. Increased precipitation promoting greater evaporation in

this region acts as a cooling feedback. This is probably the main

cause of the tropical cooling observed over Africa.

The GCM indicated that the greatest degree of warming

occurred in the Northern and Southern Hemisphere during their

respective winter seasons. This is attributable to the warmer

SSTs in the mid to high latitudes, with associated changes in

snow and ice cover.

Climate modeling – total precipitation rate (mm day

–1

)

The distribution of total precipitation during the Pliocene was

simulated in Haywood et al. (2000a) and Haywood (2001)to

be broadly similar to present (Figure P77). A general but modest

increase, of between 0 and 2 mm day

–1

, was predicted in both

hemispheres during DJF and JJA. Although sporadic pockets

of less rainfall (–2 mm day

–1

) were common, especially during

JJA, they were frequently not statistically significant. Annual

mean precipitation during the Pliocene was 34.65 cm year

–1

2 cm higher than at present (32.84 cm year

–1

), representing a

6% increase relative to modern values (Figure P77). During

DJF in the Northern Hemisphere (10



–70



North), areas of

enhanced precipitation were noted (Figure P78). These areas

include the Indian Ocean, the Caribbean, the North Atlantic

and northwestern North America, where precipitation was

enhanced by 2–8 mm day

–1

. The increase in mid-latitude preci-

pitation in the Northern Hemisphere was associated with the

warmer surface temperatures.

Increases in precipitation during JJA were seen to occur in a

narrow belt at equatorial and low latitudes. India, Arabia,

Central and West Africa, the Caribbean and the West Coast

of Central and Latin America all showed an increase in preci-

pitation of between 2 and 10 mm day

–1

(Figure P78). The model

indicated that the southern limit of the Hadley Cell (atmospheric

cell which transports heat and momentum in the atmosphere at

low-latitudes) was in approximately the same position as at pre-

sent. However, the northern limit was shifted further to the North

by 5



of latitude. The result was a net ‘broadening’ of the Hadley

Cell in the Pliocene, which allowed more precipitation to fall

over a larger geographical area. This change in the Hadley Cell

was caused by the simulated reduction in the equator-to-pole

temperature gradient (by 6



C). Broadening of the Hadley Cell

explained several of the model’s responses observed at equatorial

and low latitudes, such as increased precipitation, evaporation

and subsequent reduction of surface temperatures.

Over Greenland, a statistically significant increase in preci-

pitation was predicted, except over the southern tip of Green-

land where precipitation decreased. This decrease was related

to a shift in the position of the Atlantic storm track.

Climate modeling – wind (m s

–1

) and surface

pressure (hPa)

As a global annual average, surface wind strength decreased

in the Pliocene experiment compared to the present. This res-

ponse was consistent with the model-predicted reduction in

the equator-to-pole temperature gradient of 6



C, which dimin-

ished the general circulation of the atmosphere. Most variations

in wind dynamics were observed in the westerly component.

In the simulation, Pliocene wind directions were broadly

similar to the present. Increases in wind speed were observed

Figure P76 Model-predicted Pliocene (dashed line) and present-day

(solid line) mean annual surface temperatures (



C) plotted as a function

of latitude. This plot indicates a surface temperature regime with

reduced seasonality and greater equability during the Pliocene

compared to present-day (Haywood, 2001).

Figure P77 Annual mean total precipitation rate as a function of

latitude for present-day and the Pliocene in mm day

–1

(redrawn from

Haywood et al., 2000a; Haywood, 2001).

806 PLIOCENE CLIMATES

over areas of North America, Siberia and the Tibetan Plateau.

This was a function of the altered land cover and surface heat-

ing characteristics of the Pliocene simulation as compared to

the modern one. Wind strengths over the oceans exhibited a

more dramatic change in the Pliocene scenario. Wind speeds

increased by a maximum of 6 m s

–1

during DJF in areas

affected by the Icelandic and Aleutian low-pressure systems.

These pressure systems deepened in the Pliocene simulation

as compared to the present (Figure P79).

Analysis of wind strength fields at 200 mbs indicated that the

strength of the mid-latitude jet streams was enhanced in the Plio-

cene experiment. Increases in wind speeds in other locations

Figure P78 Model-predicted difference in mean total precipitation rate (mm day

–1

) between the Pliocene and present-day for (a) DJF and

(b) JJA. Note that the contour interval is not equally spaced and regions where precipitation change is less than 0.5 mm day

–1

are unshaded. The

grey shading in accompanying mini-boxes shows the geographical areas in which the difference in precipitation rate is significant to a 95%

statistical significance level (redrawn from Haywood et al., 2000a; Haywood, 2001).

PLIOCENE CLIMATES 807

were accounted for by the enhancement of high-pressure zones

over subtropical ocean regions. A strong high-pressure anomaly

was observed in the Southern Ocean west of Australia during

DJF and JJA (Figure P79). This was related to alterations in

Rossby wave (planetary wave) propagation caused by changes

in the Antarctic orography in the Pliocene simulation. The

strength of JJA monsoon circulation in the Indian Ocean was

weakened in the Pliocene experiment.

Climate modeling summary

The results of the modeling study by Haywood et al. (2000a)

and Haywood (2001) are closely linked to, and dependent

Figure P79 Model-predicted difference in surface pressure (hPa) between the Pliocene and present-day for (a) DJF and (b) JJA: Vector

surface mean winds are also shown. The grey shading in accompanying mini-boxes shows the geographical areas in which the difference in

precipitation rate is significant to a 95% statistical significance level (redrawn from Haywood et al., 2000a; Haywood, 2001).

808 PLIOCENE CLIMATES

on, the specified model boundary conditions. From analysis of

the model simulation and from sensitivity experiments

(Haywood, 2001), it is clear that alterations in SSTs combined

with sea and land ice volumes within the model boundary con-

ditions are primarily responsible for the observed variations

between Pliocene and present-day climates. The variations in

climate are caused by the alteration of processes, including

the release of sensible and latent heat (SST driven) and by

variations in heat exchange between the ocean and atmosphere

that are also caused by differences in sea ice.

The warmer SSTs are responsible for the majority of model-

predicted annual mean increase in surface temperatures and pre-

cipitation for the Pliocene. However, changes in high latitude

surface temperatures and precipitation rate are also closely linked

to the altered configuration of terrestrial ice sheets and sea-ice

extents specified within the model boundary conditions. The

enhancements of the mid-latitude jet streams and surface wes-

terly winds predicted in the Pliocene experiment are primarily

linked with changes in atmospheric pressure systems, which

are themselves generated by reduced high latitude terrestrial

and sea-ice cover. The role of altered jet stream flow, in response

to the altered orography of the western cordillera of North Amer-

ica, may also be important in generating the enhancements in

westerly wind strength observed in the Pliocene simulation.

Causes of pliocene climate change

Climatic warmth

The causes of Pliocene climate change have been widely dis-

cussed (e.g., Raymo et al., 1996 ). The applicability and scien-

tific usefulness of the question “what was the cause of

Pliocene warmth?” is arguable. This is because of (a) the multi-

ple interactions of different components of the Earth’s climate

system, (b) equifinality (different climate forcing mechanisms

leading to the production of a similar response in climate),

and (c) climatic memory effects (the existence of a global cli-

mate regime at any one time depends not only on forcing

mechanisms in operation during the said regime but also those

which occurred previously). A more scientifically useful ques-

tion to pose might be “what were the processes that helped pro-

mote, sustain and ultimately terminate the climates of the

Pliocene?”. Such an approach has the advantage of being pro-

cess rather than causation driven, which increases the useful-

ness of studying past warm periods in our quest to better

understand future climate change.

Atmospheric CO

Estimates of atmospheric CO

levels have been derived from

analyses of the stomatal density of fossil leaves (stomata are

small pores on a leaf surface that allow carbon dioxide to enter

the leaf and oxygen to escape, thereby facilitating photosynth-

esis; Kürschner et al., 1996), through analyses of d

C ratios of

marine organic carbon (Raymo and Rau, 1992) and through

measurement of the differences between the carbon isotope

composition of surface and deep waters (Shackleton et al.,

1992). Use of all of these proxy methods suggests that absolute

levels during the epoch ranged from 360 to 400 ppmv,

compared to mid-nineteenth century levels of approximately

280 ppmv and modern concentrations of 375 ppmv. The cause

for this higher Pliocene concentration is not well understood

and may be attributable to the long-term trend of CO

change

throughout the Tertiary (Berner, 1990). Alternatively, feed-

backs between climate and the carbon cycle (the cycling of

carbon in the form of carbon dioxide, carbonates, organic com-

pounds, etc. between various reservoirs, e.g., the atmosphere,

the oceans, land and marine biota and, on geological time

scales, sediments and rocks) may also have been important.

The modest CO

increase reconstructed for the Pliocene

would produce a radiative forcing of approximately 2 W m

–2

This may be sufficient to explain the warmth of the Pliocene

globally, depending on the chosen climate sensitivity para-

meter, and whether there are any significant changes in other

aspects of the radiative forcing. It cannot explain the regional

changes in reconstructed SSTs, which apparently display no

warming trend relative to present day in the tropics (Dowsett

et al., 1996). Therefore, an additional mechanism (or mechan-

isms) working independently or combined with variations in

concentrations in the atmosphere would have been

required to drive and maintain the warmer climates recon-

structed for the Pliocene.

Ocean heat transport

A likely additional factor is a change in the heat transport of

the oceans. This could be achieved through an increase in

thermohaline or surface gyre circulation. This possibility was

highlighted in early modeling studies (e.g., Rind and Chandler,

1991), but many difficulties and unexplained issues remain.

Paradoxically, a reduced latitudinal SST gradient implies wea-

ker atmospheric forcing of surface oceanic circulation, and

hence weaker oceanic heat transport from equator to higher

latitudes. This is a common problem associated with past

warm climate dynamics. Problems exist in any explanation

of Pliocene warmth that is based solely upon strengthening of

the thermohaline circulation as it is, (a) difficult to ascribe a

thermohaline coupling argument to all ocean basins and,

(b) difficult to generate the correct reconstructed hemispheric

temperature distribution (Crowley, 1996).

The arguments above are largely based on a consideration of

global scale energetics. The importance of specific regional

aspects, such as changes in surface wind strength and wind

stress over different ocean basins, has only recently been con-

sidered in detail. Climate modeling results presented in

Figure P79 indicate an increase in westerly component winds

at mid to high latitudes in both hemispheres. This is a direct

response to less ice cover and warmer surface temperatures

over Greenland, the Arctic, Southern Ocean and Antarctica

and to warmer sea-surface temperatures in the North and South

Atlantic Oceans during the Pliocene. Conversely, sub-tropical

high-pressure systems in both hemispheres are also enhanced.

Therefore, regional pressure gradients in the mid to high latitudes

of both hemispheres may have been intensified during the Plio-

cene. This has the potential, when combined with warmer SSTs,

to increase the atmospheric transport of heat and moisture to the

high latitudes (Figure P80). Increasing westerly wind strength

applies greater wind stress (maximum 20 N m

–2

on the basis

of modeling experiments) to the surface of the oceans, thus

potentially strengthening the flow of heat-transporting surface

ocean currents such as the Gulf Stream (Figure P81).

These interpretations, based upon modeling studies (e.g.,

Haywood et al., 2000b), support paleoceanographical evidence

for enhanced oceanic heat transports (e.g., Dowsett et al.,

1996). Enhanced carbonate accumulation and upwelling from

the California margin during the Pliocene relates to increased

wind stress (Ravelo et al., 1997). Results from a study of mar-

ine ostracodes (Cronin, 1991) suggest that: (a) North Atlantic

circulation may have been intensified, and (b) reconstructed

PLIOCENE CLIMATES 809

cooler ocean temperatures off the southeastern and mid-

Atlantic coast of the USA during the Pliocene were a result

of enhanced upwelling driven by the Gulf Stream. Billups et al.

(1998) speculated that positive feedbacks involving Southern

Hemisphere atmospheric circulation and Northern Hemisphere

poleward heat transport, together with sea-ice feedbacks, might

have played an important role in maintaining the climatic

warmth of the Pliocene.

Ocean heat transport – difficulties

The mechanism of enhanced ocean heat transport may have

been important in helping to maintain greater mid to high lati-

tude warmth during the Pliocene. However, the oceans act, to a

first approximation, as a feedback mechanism rather than a

direct forcing on climate. In other words, if we invoke greater

oceanic transport of heat to help explain the warmth of the

Pliocene then we must deal with the following question “what

were the factors responsible for forcing a change in oceanic cir-

culation and heat transports?”. Rind and Chandler (1991) pos-

tulated that changes in poleward ocean heat transport might

be related to changes in orography, which would affect the

planetary wind-driven circulation. However, it is unknown

whether changes in orography are capable of affecting wind-

driven circulation on anything more than a regional scale

(Crowley, 1991). Climate modeling sensitivity experiments

conducted by Haywood (2001) indicate that lowering the ele-

vation of the western cordillera of North America, for example,

does have an impact on the surface circulation across North

America and into the Atlantic sector. However, alterations in

high latitude low-pressure and sub-tropical high-pressure cells,

as a result of reduced ice cover and altered SST gradients, are

required to sustain the changes in circulation over the European

and other sectors. Hence, whilst changes in orography may

have been an important contributing factor in forcing changes

in the wind-driven circulation, other mechanisms must have

been in operation to drive larger-scale alterations in oceanic

circulation, for the Pliocene at least.

The potential influence of changing terrestrial ice cover on

oceanic circulation should not be forgotten. Although the direct

effect of reduced high latitude ice cover on surface tempera-

tures may only be observed over a couple of thousand

kilometers away from the ice edge, it can have a significant

impact on atmospheric pressure systems in the mid to high

Figure P80 Model prediction of the difference in total poleward heat

transport (in PW), by atmosphere and oceans between the Pliocene

climate modeling experiment of Haywood et al. (2000a) and the

present-day (xaiud = mid-Pliocene simulation; xajpa = present-day

simulation).

Figure P81 Coupled ocean-atmosphere model prediction of the difference in ocean current strength (mm s

–1

) at 5 m between the Pliocene

and present-day. Note the increase in the strength of the Gulf Stream and North Atlantic Drift currents in the North Atlantic (redrawn from

Haywood and Valdes, 2004).

810 PLIOCENE CLIMATES