Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

uplift of the Tibetan Plateau (Quade et al., 1989; Harrison et al.,

1992; Molnar et al., 1993).

The growth of the Tibetan Plateau represents a clear case in

which tectonically-driven surface uplift (i.e., an increase in mean

elevation) has caused fundamental changes in regional climates.

Moreover, the 5 km high topography of Tibet presents a physical

barrier that impacts global patterns of atmospheric circulation

and affects regional climates throughout the Northern Hemisphere

(Kutzbach et al., 1993).

Active mountain building and climate change

The CO

-weathering hypothesis suggests that weathering of

silicate rocks can extract CO

from the atmosphere, thereby

reducing the amount of greenhouse gases and causing global

cooling (Raymo and Ruddiman, 1992). Such a reaction can

be chemically expressed as:

CaSiO

þ CO

gðÞþ2H

O ¼ CaCO

þ H

SiO

in which calcium silicate combines with atmospheric CO

and

creates calcium carbonate which is then precipitated and

sequestered in oceans. Although drawdown of atmospheric car-

bon dioxide should happen all the time due to such a weather-

ing process, the uplift of mountains is proposed to accelerate

chemical weathering rates because rapid mechanical denuda-

tion in mountains will provide abundant fresh mineral surfaces

for chemical attack. This theory, therefore, links orogenesis

with changes in mineral weathering rates, greenhouse gases,

and ultimately global temperature.

The growth of the Himalaya and secondarily of the Andes dur-

ing the late Cenozoic (Raymo et al., 1988) has been cited as a

potential trigger for enhanced rates of CO

drawdown and global

cooling. These ranges are characterized by mechanical erosion

rates averaging >1mmyr

1

: rates far exceeding those at which

rocks are converted to soils through chemical and mechanical

weathering. The primary way in which these ranges sustain

long-term erosion rates of this magnitude is via landslides invol-

ving bedrock, because soil-related processes are too slow

(Burbank et al., 1996). Such landsliding persistently exposes fresh

rock surfaces that are ideal sites for chemical weathering. Chemi-

cal weathering is enhanced by two additional features of the Hima-

laya and Andes. First, the tropical temperatures promote more

rapid rates of chemical weathering than do cooler conditions at

northern latitudes. Second, these ranges are sufficiently high to

be glaciated. The “rock flour” that is produced by glacial grinding

is perfect grist for chemical weathering because it commonly

comprises tiny, unweathered rock particles with large surface

area/mass ratios.

The drawdown of atmospheric CO

is counterbalanced by

three feedbacks. First, more than 95% of the Earth’s surface

is not experiencing mountain building. Erosion rates are low

(commonly <0.01 mm yr

1

) in such areas, and chemical

weathering also proceeds slowly even in the tropics, except in

areas of higher relief. These areally extensive, low rates of

atmospheric drawdown may balance the high rates of draw-

down predicted for areas of rapid mountain building and ero-

sion. Second, because erosion isostatically drives rock uplift,

formerly buried carbon reservoirs can be exposed to erosion

and returned to the atmosphere. Third, mountain building asso-

ciated with subduction can lead to metamorphic degassing of

carbon-rich rocks. These mechanisms serve to increase atmo-

spheric CO

concentrations.

Impacts of climate and climate change on

mountain uplift

Defining uplift

The term “uplift” should be defined in an appropriate context (e.g.,

Figure M46). The change of elevation at any point in a landscape

(surface uplift at a point) is a function of the magnitude of erosion,

deposition, and vertical movement of the underlying rock. Some

coupling exists among these factors: erosion or deposition should

induce a compensating isostatic response, such that 5/6 (the

ratio of crustal to mantle densities) of an instantaneous change in

altitude due to erosion or deposition will be recovered over time.

The vertical movement of rocks, whether in response to isostasy

or tectonism, can be defined as “rock uplift” and should be mea-

sured in absolute terms with respect to a reference surface,

such as the geoid or sea level. Whereas such measurements may

be feasible at the coast, this formal reference frame is commonly

lost inland. Nonetheless, relative amounts of rock uplift can be

defined by comparing offset points that were once contiguous at

the same altitude. The average change of all points on a surface

serves to define “surface uplift” or the change in the mean eleva-

tion of an area (Figure M46). Although some geophysicists argue

that the term surface uplift should be restricted to areas of a size

(10

) capable of inducing a response to crustal loading

(England and Molnar, 1990), changes of mean surface elevation

of smaller areas can provide useful geomorphic and geologic

insights (Medwedeff, 1992; Talling and Sowter, 1999).

Absolute field measurement of any form of uplift at geologic

time scales (>10

yr) is commonly difficult. The widespread

availability of digital elevation models allows ready determination

of modern surface elevations; a comparable reconstruction of

ancient elevations could require many hundreds of points.

Although the science of paleoaltimetry is improving (Sahagian

and Maus, 1994; Forest et al., 1999), there are few reliable or pre-

cise measures of the previous altitude of the surface or rock

beneath it. Even when the amount of erosion or deposition can

be specified, the vertical movement of rocks in response to iso-

stasy or to tectonic loading is hard to measure. Whereas coastal ter-

races provide a means to define rock-uplift (Lajoie, 1986), for

most inland sites, relative displacements of surfaces or rocks are

commonly studied with respect to a reference frame defined

by an assumed earlier form, such as a river gradient (Thompson

et al., 2002) or the surface of a former sedimentary basin (Burbank

et al., 1999; Bullen et al., 2003).

What drives surface uplift of mountain ranges?

Four phenomena can commonly cause positive changes in mean

elevation. Tectonic thickening of the crust will induce a change

in the mean height equal to 1/6 of the magnitude of thickening

after the crust is compensated isostatically. The commonly

observed low-density root beneath contractional ranges is at least

partly responsible for elevated topography above it. Replacement

of high-density mantle with lower density asthenosphere willdrive

a similar isostatic response, but in proportion to the density con-

trasts between them and the thickness of the replaced mantle.

Dynamic support of topography can occur when the strength of

the flexed crust induces surface uplift over a flexure. For example,

underthrusting of old, thick,rigid crust can generatedynamic topo-

graphy. In contrast to the previous two mechanisms, which cause

surface uplift, dynamic support will create a range out of isostatic

equilibrium. Finally, volcanism can cause surface uplift by

two mechanisms. Heating of the crustlowers its density and makes

it more buoyant, whereas addition of new volcanic material to the

MOUNTAIN UPLIFT AND CLIMATE CHANGE 599

surface or intrusion of plutons adds to the crustal mass and causes

an isostatic response. Because crustal additions occur from

point sources (vents) in volcanic systems, they generate more irre-

gular, punctuated surface uplift than non-volcanic systems,

because they create isolated high peaks or volcanic cones that

may be widely separated.

Surface uplift of mountain peaks (positive elevation changes)

will typically result from all of the processes described above.

Because volcanoes add material to the Earth’ssurface,rock

beneath will subside in response to new volcanic loads, such that

only 1/6 of any crustal thickening will be expressed as a long-

term elevation change.

Erosion can also induce peak uplift. Consider the situation

in which erosion is focused in only part of the landscape, such

as the valley bottoms. The removal of mass by erosion will

cause a net thinning of the crust, and although the mean eleva-

tion will decrease, the crustal thinning induces isostatic uplift

(rock uplift). If uneroded, the intervening peaks will rise by

an amount equal to 5/6 of the average amount of erosion

(Molnar and England, 1990). Consequently, 1 km of erosion

averaged across a region could cause the mean elevation to

decrease by 150 m, whereas uneroded peaks could uplift iso-

statically >800 m (Figure M47).

The erosional system

The primary means by which climate can affect tectonics is

through the influence of climate on erosion. Isostasy requires

that wherever erosion removes significantly more mass from

an active orogen than elsewhere, differential rock uplift will

occur preferentially beneath the more highly eroded area.

Figure M47 Isostatic responses to erosion. Left: An uneroded “crustal” block in isostatic balance with surrounding denser mantle; its upper surface

floats above the surface of the denser material. Center: Square “valleys” 1.5 km deep remove mass across 2 /3 of the top of the crustal block

(average erosion equals 1 km). Rebound is artificially restrained. Right: Following isostatic rebound, the mean elevation decreases by ~165 m, whereas

the base of the block and the remaining part of its uneroded upper surface rebound ~830 m. The amount of rebound or “rock uplift” equals the

average amount of erosion (1 km) times the ratio of the density of the crustal block (r = 2.75 g cm

3

) to that of the mantle (r = 3.3 g cm

3

Figure M46 Definitions for aspects of “uplift.” (a) Initial configuration depicting mean surface height. (b) Rock uplift (vertical arrows) represents the

vertical motion of rock with respect to sea level or the geoid. “Regional surface uplift” is defined by changes in mean surface height (difference

from “time 1” to “time 2”). The difference between the magnitude of rock uplift and the change in the mean surface height is the mean

denudation. (c) “Surface uplift at a point” can depart from the mean surface uplift due to variations in erosion or deposition at specific sites in a

landscape. (Modified after Burbank and Anderson, 2001.)

600 MOUNTAIN UPLIFT AND CLIMATE CHANGE

Because differential rock uplift usually requires faulting, differ-

ential erosion can focus tectonic activity in regions where ero-

sion is most rapid.

A linked climate-geomorphic system underlies this feedback.

From a geomorphic framework, most mountain belts consist

of networks of rivers and adjoining hillslopes (Figure M48). The

rivers control the rate of lowering of base level within the moun-

tains, whereas the hillslopes respond to base-level lowering and

provide a flux of sediment to the rivers. Two classes of hillslope

processes can be defined based on whether they involve primarily

soil or bedrock. Soil erosion processes on hillslopes include rain

splash, creep, bioturbation, and shallow (soil) landsliding.

Because these occur primarily within soils, they cannot proceed

at rates faster than soils are produced by weathering of bedrock.

The best documented studies at present indicate that the conver-

sion of rock to regolith or soil occurs at rates of <0.5 mm yr

1

(Heimsath et al., 1997). Thus, if hillslopes are eroding at

rates slower than this, processes that move soils downslope

may be sufficient to account for all hillslope sediment fluxes. As

soon as rates climb above 1 mm yr

1

, bedrock landslides must

begin to play an important role. For bedrock landslides to occur,

a hillslope must be sufficiently steep to overcome the stabilizing

forces resulting from cohesion (rock strength) and gravitational

normal stresses on a potential failure plane. At equilibrium, slope

angles should hover around that angle required for failure: the

threshold slope angle. In many rapidly eroding terrains, hillsides

are long and straight, with slope angles that are commonly clus-

tered around 30



–34



, independently of variations in erosion rates

>1mmyr

1

. These characteristics are indicative of threshold

slopes poised for failure (Figure M48). Any processes that

decrease cohesion or normal stresses, and increase pore pressures

or shear stresses, will tend to lead to bedrock landsliding. Thus,

rock weathering, heavy rainfall, or seismic shaking will all tend

to trigger landslides.

For ranges whose peaks rise above the regional snowline,

glacial erosion becomes an important component of the geo-

morphic system (Figure M49). Glaciers that slide on their beds

are very effective erosive agents, capable of eroding at rates

exceeding 10 mm yr

1

(Hallet et al., 1996). Glacial erosion is

commonly defined as a function of ice velocity (Humphrey and

Raymond, 1994). Because velocity typically peaks near the glacial

snowline, vertical erosion is considered to be maximized in the

vicinity of the snowline (MacGregor et al., 2000). When coupled

with enhanced vertical erosion, rapid headward erosion by glaciers

causes a flattening of the slope of the long profile of glaciers

(Brocklehurst and Whipple, 2002). Although glaciers are capable

of overdeepening (scouring below the level of their toes), the

rate of lowering of the entire glacial profile is ultimately regulated

by the rate at which the rivers erode at the glacial terminus.

Mountains of sufficient height may have their upper reaches in

a frozen zone that almost never rises above freezing. Here, glaciers

are cold-based and frozen to their beds, freeze-thaw activity is

minimal, and the primary means of erosion is through bedrock

landslides (Figure M49). In such a setting, warm-based glaciers

commonly occur at lower altitudes and cause both downward

and headward erosion. Meanwhile, peaks within the frozen zone

become higher and narrower. These steeple-like summit regions

episodically collapse via large-scale bedrock landslides.

Climatic impacts on the geomorphic system and

mountain topography

During the repeated, large-scale Quaternary oscillations

(Shackleton and Opdyke, 1976), changing temperature and

precipitation regimes caused ice sheets and mountain glaciers

to wax and wane. Snowlines generally were lowered 800–

1,000 m in most temperate regions during maximum glacial

times (Porter, 1989). This snowline shift displaced the pre-

dicted zone of maximum glacial erosion farther down the

Figure M48 Hillslope and erosion processes in a steady-state, non-glaciated landscape. Left panel: Geomorphic process zones. River incision

determines changes in local base level. Rates of rock-to-soil conversion limit the rate at which the landscape can erode (rain splash, creep, shallow

landslides) in the absence of removal of unweathered bedrock. Right panel: Rates of relative rock-uplift increase from left to right. Assuming

topographic steady-state, hillslope erosion and river incision match rock uplift rates. For slow uplift, soil processes (creep, rain splash) balance uplift

rate. At higher uplift rates, shallow landslides rapidly strip soil from hillslopes. When relative rock-uplift rates exceed the rock-to-soil conversion

rate, soil erosion balances uplift only if bedrock erosion, usually via landslides, also occurs. In most such landscapes, many hillslopes hover near the

threshold angle for failure by landsliding. (Modified after Burbank, 2002.)

MOUNTAIN UPLIFT AND CLIMATE CHANGE 601

mountain flanks and was responsible for the low-gradient gla-

cial troughs that exist downstream of modern glaciers. The ver-

tical and lateral excavation of these troughs lowers base level

and steepens adjacent valley walls, thereby promoting acceler-

ated lowering of the ridges between glacial valleys. During gla-

cial times, therefore, the average height of the glaciated parts of

mountains should be lowered due to enhanced erosion of both

valleys and ridges.

Changes in hydrology and sediment loads also affect fluvial

erosion during climate changes. For any given climatic state, a

river tends toward a graded condition, whereby its channel

slope and width are adjusted to transport the sediment supplied

to it from hillslopes. For a given discharge, steeper and nar-

rower channels allow more sediment to be transported down-

stream. The longitudinal profiles of rivers typically show

systematic downstream decreases in channel slope with

increasing drainage area. The concavity (y) of many river sys-

tems is defined by the relationship of channel slope (S) to drai-

nage area (A): S =kA

y

, where k is a constant (Whipple et al.,

1999). For most rivers, channel concavity hovers around 0.4–

0.5. The constant k can be conceptualized as a ratio between

rock uplift and climate or hydrology. If rock uplift rates

increase, channel steepness should also increase, whereas if cli-

matic conditions enhance erosiveness, channel steepness

decreases (Whipple et al., 1999). In general, climatic conditions

that lead to higher discharges will promote erosion and cause

the entire steam profile to become less steep (Figure M50).

Sediments transported by rivers abrade the channel bed through

repeated impacts. In the absence of sediment, rivers cannot erode

efficiently, whereas too much sediment will cover the bed of the

river and protect it from further erosion (Sklar and Dietrich,

2001). For underloaded rivers, i.e., where the sediment supply is

much less than the transport capacity, a climatically-induced

increase in the sediment supply may promote more rapid channel

erosion. Irrespective of the cause (changing sediment or water dis-

charge) of a lower channel slope, unless there is a compensating

change in drainage spacing or rock strength, hillslopes and ridge

crests will be lowered as well, and the mean elevation of a range

should decrease (Figure M50). Because climate oscillates between

glacial and interglacial conditions, the erosivity of fluvial systems

should oscillate as well, such that the mean elevation, mean

relief, and height of the ridge crests should vary with climate.

The rapidity of climate change, however, often outpaces the

response time of river systems (Whipple, 2001), such that the

hillslopes and rivers will always tend to be somewhat out of

equilibrium.

Because glaciers are very effective erosional agents, their

presence will create a different mountain topography than

would occur on the same range in the absence of glaciers.

Two attributes influence glacial effects on topography. First,

the focusing of bedrock erosion by glaciers at the altitude of

the glacial snow line causes a lowering of mean elevation at

that level. Second, as climate changes, the glacial snowline

shifts altitudinally up or down, causing a synchronous shift in

the zone of maximum glacial erosion. The well known elongate

glacial troughs of alpine terrains are excavated by this mechan-

ism. In comparison to the topographic profiles of rivers, gla-

ciers produce high-altitude low-gradient profiles that extend

Figure M49 Erosion processes in steady-state, glaciated landscapes experiencing rapid relative rock uplift and erosion. Below the glacial limit,

erosion (white downward arrows and dark gray region) via bedrock landsliding and river incision balance rock uplift (dark arrows along base). Left

panel: Warm-based glaciers slide on their beds and erode bedrock, leading to a concentration of higher altitude topography near the glacial

snowline. Right panel: Cold-based glaciers are frozen to their beds and, the absence of freeze-thaw cycles prevents erosion in the frozen zone.

As warm-based glaciers continue to erode at lower altitudes, rock uplift steepens the summit regions, making them more unstable, and causing

massive landslides. (Modified after Burbank, 2002.)

Figure M50 Predicted changes in river-channel and ridge-crest profiles

that result from a sustained change from a less to more erosive climate

state or from a regime of more rapid to slower rock uplift. More erosive

climates increase discharge to each channel, thereby enhancing their

ability to erode and lower river profiles. Because only a small fraction of

a landscape typically lies between the channel head and ridge crest, a

decrease in the channel gradient causes ridge lines to lower as well.

(Modified after Whipple et al., 1999.)

602 MOUNTAIN UPLIFT AND CLIMATE CHANGE

upward to the base of the cirque headwalls. Because the head-

walls occupy only a small fraction of the landscape, the glacial

troughs cause a concentration of landscape area in the altitude

range defined by climatically-driven variations in snowline

(Figure M51) (Brozovic et al., 1997). The net results of clima-

tically-driven snowline variations are a lowering of the mean

elevation, a focusing of land area near the snowline, and devel-

opment of isolated horns or summits formed by coalescence of

cirques. These summits may act as topographic “lightning

rods” where storms gather along their flanks (Brozovic et al.,

1997). The additional snow fed into the nearby glaciers would

be expected to cause them to erode even more effectively.

Although reconstruction of the paleoaltitude of past moun-

tain belts is difficult and typically contains large uncertainties

(100s of m), changes in the flux of sediments from a mountain

belt can be assessed through examination of the sedimentary

record in basins adjacent to ranges. An enhanced sediment

flux can occur either because erosion is occurring with greater

efficienc y compared to the past or because the rates of tectonic

fluxes into an eroding mountain belt have increased. Compilation

of sedimentation-accumulation rates from basins around the

world indicate that both the volume and grain size of sediments

has significantly increased since 2–4Ma(Zhangetal.,2001).

The global synchrony of this change indicates that tectonics can-

not be the cause: only climate change could cause such a perva-

sive and synchronous response. Whereas one might expect

expanded glaciations to drive increases in erosion rates (Hallet

et al., 1996), many of the studied basins are not bounded by gla-

ciated ranges. Instead, it is proposed that strongly oscillating

Late Cenozoic climates have imposed such abrupt and frequent

changes in precipitation, vegetation, and temperature that the

landscape was always out of equilibrium and susceptible to

rapid erosion (Zhang et al., 2001).

Steady-state orogens, topography, and climate

Conceptually, in a convergent mountain range, the flux of rock

into an orogen could be balanced by the efflux through erosion,

such that a steady-state mass would be maintained (Willett and

Brandon, 2002). In a topographic steady state, the mean height

and relief of the range would also remain steady through time.

In such a condition, erosion by rivers, hillslopes, and glaciers

must be adjusted so that they collectively erode at the same rate

as rock moves into the orogen through tectonism. Steady-state

topography requires a tectonic influx, because by itself the iso-

static response to erosion will always result in a reduction in

the mean height of a range. Given the rate of climatic change

and the response time of geomorphic systems to approach equi-

librium, orogenic topography should oscillate between higher

and lower mean elevation and relief. Thus, at short time scales

(<10

yr), steady state would never be strictly attained. How-

ever, if the average topography remained steady at time scales

exceeding the primary glacial cycles (10

yr), the topography

could be considered to be in a steady state.

Recent numerical models suggest that climate can play a major

role in shaping orogens and determining where strain is focused

within an orogenic belt (Beaumont et al., 1992; Willett, 1999).

The models suggest that, in the absence of erosion, a convergent

orogen will simply widen and thicken over time. If moisture is

advected into the orogen from one side, the resultant asymmetric

erosion causes profound changes in the shape of the orogen, preci-

pitation patterns, and distribution of thickening and strain. Both

the topographic divide and the zone of maximum crustal thicken-

ing shift away from the moisture source, whereas the region

of maximum strain shifts toward the moisture (Figure M52).These

changes all result from the preferential focusing of erosion on only

one side of the orogen.

At a large scale, the predicted impacts of climate on oro-

genic growth may be well displayed by the Himalayan-Tibetan

Orogen. Strong climatic contrasts exist on the northern and

southern margins of the Tibetan Plateau. In the south, the Asian

monsoon delivers 3–5 m of rain over the summer months to the

southern flank of the Himalaya (Barros et al., 2000). In the

north, generally arid to semi-arid conditions prevail along

the margins of the Tarim Basin and the Hexi Corridor. After

50 Myr of the ongoing Indo-Asian collision, 1,000s of km

of convergence have been absorbed within the orogen. Accom-

modation of the converging crust can occur by thickening and

widening, removal of mass through erosion, or lateral escape of

mass out of the pathway of collision. The monsoon’s highly

effective removal of mass from the Himalayan side of the Tibe-

tan Plateau suggests that a near steady state may prevail there

(Beaumont et al., 2001). Many of the major Himalayan struc-

tures that were active 10 million years ago are still active today,

and the zone of active deformation has not significantly

widened. In contrast, on the arid northern side of the Tibetan

Plateau, extensive northward expansion of the Plateau has

occurred during this same interval. Both northward propagation

of active thrusting and the incorporation, shortening, and thick-

ening of the former northern foreland has enabled large scale

Figure M51 Hypsometry (altitude versus frequency) and mean hillslope

angles of a glaciated Himalayan region. In glaciated regions above

3 km, much of the topography is concentrated between 4 and 5 km – a

range coinciding with the altitudinal range of glacier snow lines

between interglacial and glacial times. Glacial erosion is most efficient

near the snow line, where ice flows fastest. Such altitudinally-

dependent erosion focuses more topography near that altitude.

Hillslope angles are steep in the zone just below the glaciers (<3.5 km)

where bedrock landsliding occurs, but become more gentle in the

altitude range where most glaciers occur, before steepening again in

the headwalls and higher altitudes where cold-based glaciers prevail.

(Modified after Brozovic et al., 1997.)

MOUNTAIN UPLIFT AND CLIMATE CHANGE 603

accretion of new mass into the northern edge of the Tibetan

Plateau. The apparent steady state on one flank of the Himala-

yan-Tibetan Orogen and its obvious absence on the other flank

support the prediction that large scale climatic contrasts exert a

controlling influence on patterns of orogenic deformation

(Willett, 1999).

Summary

Mountain ranges impact climate and climate change through

multiple mechanisms, including the control of the pathways

moisture follows across the landscape, patterns of orographic

precipitation as storms impinge on the flanks of mountains,

the enhancement of summer monsoons, and the magnitude of

continentality. The chemical composition of the atmosphere

and, in particular, the concentrations of CO

may be affected

by rates of mountain building, whereby rapid mountain growth

causes high rates of mechanical weathering and enhances the

extent of glaciation – both of which serve to provide fresh rock

surfaces for chemical weathering.

Climate can also exert some surprising controls on mountain

topography and orogenic growth. In nonglaciated mountains, a

coupling exists between river erosion, bounding hillslopes, and

climate. Rates of river downcutting control both local base-level

lowering and incipient changes in the angle of hillslopes.

Increased discharge during wetter climates will promote more

rapid fluvial erosion and lower stream gradients. The flux of sedi-

ment off slopes represents the primary means of denudation

in most orogens. As climate changes, precipitation variations alter

hillslope pore pressures, change stable hillslope angles, and ulti-

mately cause variability in hillslope sediment fluxes. When

climates are wetter or more erosive, the mean altitude and relief

of ranges should be lowered.

In glaciated ranges, glacial erosion commonly limits

the mean height of mountain ranges by removing much of

the topography above the snow line. Climatically-induced

variations in the glacial snow line cause changes in the ele-

vation of maximum glacial erosion and tend to define an altitu-

dinal range within which most topography is concentrated.

Overall, glaciated ranges will have lower mean elevations than

would nonglaciated ranges under the same tectonic conditions.

In convergent orogens, gradients in erosion can cause gradi-

ents in crustal strain. Increased moisture on one flank of a

Figure M52 Contrasting magnitudes of geometries of erosion, strain, and crustal thickening resulting from the direction in which storms

approach an orogen. In both cases, the tectonic framework is identical, with two blocks of continental lithosphere colliding and with the mantle

lithosphere detaching on one side of the orogen. In each case, erosion and crustal strain are concentrated on the side from which moisture

approaches the orogen, whereas the topographic divide and zone of maximum crustal thickening move away from the moisture source.

The erosion resulting from greater rainfall on one side of the orogen creates fundamental contrasts in deformation patterns, despite the

identical tectonic framework. (a) Moisture enters the orogen from above the retro-wedge side (the non-subducting side. (b) Moisture

approaches from the pro-wedge side. (Modified after Willett, 1999.)

604 MOUNTAIN UPLIFT AND CLIMATE CHANGE

range will tend to increase erosion on that side and induce more

rapid deformation in the underlying rock. Whereas a simple

isostatic response can restore 80% of the eroded mass, if rock

is entering a convergent orogen from the side, then a steady

state topography can be maintained at time scales (>10

yr)

exceeding those of major climatic cycles.

Douglas W. Burbank

Bibliography

Barros, A.P., Joshi, M., Putkonen, J., and Burbank, D.W., 2000. A study of

the 1999 monsoon rainfall in a mountainous region in central Nepal

using TRMM products and rain gauge observations. Geophys. Res.

Lett., 27, 3683–3686.

Beaumont, C., Fullsack, P., and Hamilton, J., 1992. Erosional control of

active compressional orogens. In McClay, K.R. (ed.), Thrust Tectonics.

London: Chapman and Hall, pp. 1–18.

Beaumont, C., Jamieson, R.A., Nguyen, M.H., and Lee, B., 2001. Himala-

yan tectonics explained by extrusion of a low-viscosity crustal channel

coupled to focused surface denudation. Nature, 414, 738–742.

Brocklehurst, S.H., and Whipple, K.X., 2002. Glacial erosion and relief

production in the eastern Sierra Nevada, California. Geomorphology,

42,1–24.

Brozovic, N., Burbank, D.W., and Meigs, A.J., 1997. Climatic limits

on landscape development in the northwestern Himalaya. Science,

276, 571–574.

Bullen, M.E., Burbank, D.W., and Garver, J.I., 2003. Building the northern

Tien Shan: Integrated thermal, structural, and topographic constraints.

J. Geol., 111, 149–165.

Burbank, D.W., 1991. Late Quaternary snowline reconstructions for

the southern and central Sierra Nevada, California: Re-assessment of

the Recess Peak Glaciation. Quaternary Res., 36, 294–307.

Burbank, D.W., 2002. Rates of erosion and their implications for exhuma-

tion. Mineral. Mag., 66,25–52.

Burbank, D.W., and Anderson, R.S., 2001. Tectonic Geomorphology, 1st

edn. Malden, MA: Blackwell Science, 273pp.

Burbank, D.W., Leland, J., Fielding, E., Anderson, R.S., Brozovic, N.,

Reid, M.R., and Duncan, C., 1996. Bedrock incision, rock uplift and

threshold hillslopes in the northwestern Himalayas. Nature, 379,

505–510.

Burbank, D.W., McLean, J.K., Bullen, M., Abdrakhmatov, K.Y., and

Miller, M.G., 1999. Partitioning of intermontane basins by thrust-

related folding, Tien Shan, Kyrgyzstan. Basin Res., 11,75– 92.

Burbank, D.W., Blythe, A.E., Putkonen, J., Pratt-Sitaula, B., Gabet, E.,

Oskin, M., Barros, A., and Ojha, T.P., 2003. Decoupling of erosion

and precipitation in the Himalayas. Nature, 426, 652–655.

England, P., and Molnar, P., 1990. Surface uplift, uplift of rocks, and

exhumation of rocks. Geology, 18, 1173–

1177.

Fielding, E.J., Isacks, B.L., Barazangi, M., and Duncan, C., 1994. How

flat is Tibet? Geology, 22, 163–167.

Forest, C.E., Wolfe, J.A., Molnar, P., and Emanuel, K.A., 1999. Paleoalti-

metry incorporating atmospheric physics and botanical estimates of

paleoclimate. Geol. Soc. Am. Bull., 111, 497–511.

Guo, Z.T., Ruddiman, W.F., Hao, Q.Z., Wu, H.B., Qiao, Y.S., Zhu, R.X.,

Peng, S.Z., Wei, J.J., Yuan, B.Y., and Liu, T.S., 2002. Onset of Asian

desertification by 22 Myr ago inferred from loess deposits in China.

Nature, 416, 159–163.

Hallet, B., Hunter, L., and Bogen, J., 1996. Rates of erosion and sediment

evacuation by glaciers: A review of field data and their implications.

Glob. Planet. Change, 12, 213–235.

Harper, J., and Humphrey, N.F., 2003. High altitude Himalayan

climate inferred from glacial ice flux. Geophys. Res. Lett., 30,

10.1029/ 2003GL017329 (1764).

Harrison, T.M., Copeland, P., Kidd, W.S.F., and Yin, A., 1992. Raising

Tibet. Science, 255, 1663–1670.

Heimsath, A.M., Dietrich, W.E., Nishiizumi, K., and Finkel, R.C., 1997.

The soil production function and landscape equilibrium. Nature, 388,

358–361.

Humphrey, N.F., and Raymond, C.F., 1994. Hydrology, erosion, and

sediment production in a surging glacier: The Variegated Glacier surge,

1982–1983. J. Glaciol., 40, 539–552.

Kutzbach, J.E., Prell, W.L., and Ruddiman, W.F., 1993. Sensitivity of

Eurasian climate to surface uplift of the Tibetan Plateau. J. Geol.,

100, 177–190.

Lajoie, K.R., 1986. Coastal Tectonics. In Active Tectonics. Washington,

DC: National Academy Press, pp. 95–124.

MacGregor, K.R., Anderson, R.S., Anderson, S.P., and Waddington, E.D.,

2000. Numerical simulations of glacial-valley longitudinal profile

evolution. Geology, 28, 1031–1034.

Medwedeff, D.A., 1992. Geometry and kinematics of an active, laterally

propagating wedge thrust, Wheeler Ridge, California. In Mitra, S.,

and Fisher, G.W. (eds.), Structural Geology of Fold and Thrust Belts.

Baltimore, MD: Johns Hopkins University Press, pp. 3–28.

Molnar, P., and England, P., 1990. Late Cenozoic uplift of mountain ranges

and global climatic change: Chicken or egg? Nature, 346,29–34.

Molnar, P., England, P., and Martinod, J., 1993. Mantle dynamics, uplift of

the Tibetan Plateau, and the Indian monsoon. Rev. Geophys., 31, 357–396.

Porter, S.C., 1977. Present and past glaciation threshold in the Cascade

Range, Washington, U.S.A.: Topographic and climatic controls, and

paleoclimatic implications. J. Glaciol., 18, 101–116.

Porter, S.C., 1989. Some geological implications of average Quaternary

glacial conditions. Quaternary Res., 32, 245–261.

Pratt, B., Burbank, D.W., Putkonen, J., and Ojha, T., 2002. Climatic

implications of modern and paleo ELAs in the central Nepal Himalaya.

EOS (Transactions of the American Geophysical Union), 83, F319.

Quade, J., Cerling, T.E., and Bowman, J.R., 1989. Development of Asian

monsoon revealed by marked ecological shift during latest Miocene

in northern Pakistan. Nature, 342, 163–166.

Raymo, M.E., and Ruddiman, W.F., 1992. Tectonic forcing of late

Cenozoic climate change. Nature, 359,117–122.

Raymo, M.E., Ruddiman, W.F., and Froelich, P.N., 1988. Influence of

late Cenozoic mountain building on ocean geochemical cycles.

Geology, 16, 649–653.

Roe, G.H., Montgomery, D.R., and Hallet, B., 2002. Effects of orographic

precipitation variations on the concavity of steady-state river profiles.

Geology, 30, 143–146.

Sahagian, D.L., and Maus, J.E., 1994. Basalt vesicularity as a measure

of atmospheric pressure and palaeoelevation. Nature (London), 372,

449–451.

Shackleton, N.J., and Opdyke, N.D., 1976. Oxygen-isotope and paleo-

magnetic stratigraphy of Pacific core V28–239 Late Pliocene to Latest

Pleistocene. Geol. Soc. Am. Mem., 145, 449–464.

Sklar, L., and Dietrich, W.E., 2001. Sediment supply, grain size and

rock strength controls on rates of river incision into bedrock. Geology,

29, 1087–1090.

Talling, P.J., and Sowter, M.J., 1999. Drainage density on progressively

tilted surfaces with different gradients, Wheeler Ridge, California.

Earth Surf. Processes and Landforms, 24 (9), 809–824.

Thompson, S.C., Weldon, R., III., Rubin, C.M., Abdrakhmatov, K.E.,

Molnar, P., and Berger, G.W., 2002. Late Quaternary slip rates across

the central Tien Shan, Kyrgyzstan, central Asia. J. Geophys. Res.,

107, 10.1029/2001JB000596.

Whipple, K.X., 2001. Fluvial landscape response time; how plausible

is steady-state denudation? The steady-state orogen; concepts, field

observations, and models. Am J. Sci., 301, 313–325.

Whipple, K.E., Kirby, E., and Brocklehurst, S.H., 1999. Geomorphic limits

to climate-induced increases in topographic relief. Nature, 401,39–43.

Willett, S.D., 1999. Orogeny and orography: The effects of erosion on the

structure of mountain belts. J. Geophys. Res., 104, 28957–28982.

Willett, S.D., and Brandon, M.T., 2002. On steady states in mountain belts.

Geology, 30, 175–178.

Williams, V.S., 1983. Present and former equilibrium-line altitudes near

Mount Everest, Nepal and Tibet. Arct. Alp. Res., 15, 201–211.

Zhang, P., Molnar, P., and Downs, W.R., 2001. Increased sedimentation

rates and grain sizes 2–4 Myr ago due to the influence of climate

change on erosion rates. Nature, 410, 891–897.

Cross-references

Glacial Geomorphology

Monsoons, Pre-Quaternary

Monsoons, Quaternary

Paleo-Precipitation Indicators

Plate Tectonics and Climate Change

Weathering and Climate

MOUNTAIN UPLIFT AND CLIMATE CHANGE 605

NEAREST-LIVING-RELATIVE METHOD

The nearest-living-relative (NLR) method represents a standard

and powerful tool in paleoclimatology and paleoecology. It uses

the climatic or ecologic characteristics of the NLRs of fossil taxa

or assemblages to estimate the paleoclimatic or paleoecologic

conditions under which the fossil taxa or assemblages lived.

Since the NLR method is based on the assumption that the cli-

matic or ecologic characteristics of a fossil taxon or assemblage

are similar to those of its nearest living relative, it is largely

restricted to the Cenozoic. The paleobotanist Oswald Heer

(1809–1883), who is also considered to be one of the founders

of paleoclimatology, was probably the first to use this technique

in order to estimate climate parameters for the Tertiary based on

fossil floras (see History of Paleoclimatology). Since that time,

numerous qualitative and quantitative variations of the NLR

method have been developed, some of them based on fossil taxa,

others based on fossil assemblages, all of them, however, have

their specific advantages and limitations.

Qualitative techniques

The qualitative indicator taxon approach may be viewed as the

classic variation of the NLR method: individual indicator taxa

are used to derive information about the paleoenvironment or

about paleoenvironmental trends. For instance, the modern

bivalve species Mytilus edulis tolerates hyposaline conditions;

hence, mass occurrence of Mytilus in a Cenozoic shell-bed

may indicate brackish or varying salinities. Similarly, an increase

in abundance of Lauraceae (laurel family) or Arecaceae (palm

family) leaves in a sequence of fossil floras is typically inter-

preted as reflecting a warming trend since today both families

show a predominantly tropical to subtropical distribution.

Although very simple and rapid to apply – at least if the

ecology and climatology of the NLRs of the fossil taxa are

known – this qualitative indicator taxon approach has serious

limitations. First, it provides only qualitative and not quantita-

tive data. Second, the accuracy and reliability of the paleoenvir-

onmental information depends on the taxonomic level to

which the NLR of a fossil taxon can be determined and on

the stratigraphic or evolutionary age of the fossil taxon. Typi-

cally, more detailed information can be extracted from fossil

taxa if their NLRs can be identified not only on the family

but also on the genus or even species level. The accuracy, how-

ever, to which the NLR of a given fossil can be determined, is a

function of the evolutionary and stratigraphic age of the fossil.

Angiosperms have their major radiations in the Cretaceous and

Paleogene. Correspondingly, the NLR of Neogene angiosperms

can mostly be identified on the genus level, Pleistocene angios-

perms on the species level. The situation is completely different

for mammals, which show important radiations in the Neogene

and Pleistocene. Hence, the NLRs of many early Pleistocene

mammals can only be identified on the family or genus level.

A third shortcoming of the qualitative indicator taxon approach

is its sensitivity to taphonomic effects since the presence or

absence of certain indicator taxa may not only be controlled

by paleoclimatic or paleoenvironmental change but also by

taphonomy. For instance, Lauraceae are quite common in

European Paleogene leaf floras but are rare in pollen floras

due to differential degradation processes. For the same reason,

angiosperms are rare in Cenozoic wood floras since their trunks

and branches decay more rapidly than those of conifers.

Finally, it should be emphasized that the qualitative indicator

taxon approach is not a well-defined, objective and reproduci-

ble technique. The same fossil flora or fauna may hence lead

to different interpretations if analyzed by different scientists.

Instead of indicator taxa, it is possible to use fossil indicator

assemblages to obtain qualitative paleoenvironmental informa-

tion. This qualitative indicator assemblage approach is particu-

larly common in Quaternary research: pollen diagrams of

sedimentary sequences are interpreted as a succession of char-

acteristic vegetation types; from the vegetation succession, a

qualitative or semi-quantitative climate trend or climate curve

may then be derived. The technique can be applied to all kinds

of fossil assemblages but it makes more rigorous assumptions

than the qualitative indicator taxon approach. In the qualitative

indicator assemblage approach, it is required that the NLRs of

fossil taxa and of fossil assemblages are identified correctly

and that both the fossil taxa and the fossil assemblages have

climatic or environmental requirements similar to those of their

modern analogues. Hence, the qualitative indicator assemblage

technique is more restricted in its applicability than the qualita-

tive indicator taxon approach. In general, it works quite well in

the Quaternary but becomes increasingly less reliable in the

Neogene. In addition, it is even more sensitive to taphonomy

than the indicator taxon approach because transportation and

fossilization processes may heavily influence the composition

of a fossil assemblage.

Quantitative techniques

Both the indicator taxon and indicator assemblage techniques

may also be used in a more systematic and objective way in

order to obtain quantitative paleoclimate or paleoenvironmental

data. Grichuk (1969) was the first to develop Heer’s indicator

taxon approach into a well-defined methodology that allows deter-

mination of absolute climate estimates from fossil floras; his tech-

nique was later modified into the so-called “coexistence approach”

and adapted to modern computer technology (Mosbrugger

and Utescher, 1997). A similar methodology was developed by

Atkinson et al. (1987) for beetles and applied to Quaternary beetle

faunas. All these techniques are quite similar (Figure N1). Given a

fossil fauna or flora with taxa A to Z for which the NLRs A’ to Z’

are known, if the climatic (or ecologic) tolerances of the modern

taxa A’ to Z’ (e.g., concerning the various climatic parameters

such as mean annual temperature, mean temperature of the warm-

est month, etc.) are known, then by overlapping all these toler-

ances the “coexistence interval” canbedeterminedforany

environmental or climatic parameter. This coexistence interval is

interpreted to bracket the real climatic or environmental parameter

value under which the fossil fauna or flora lived. The climatic reso-

lution of the methodology can be improved if the various climatic

parameters such as mean annual temperature, mean annual preci-

pitation, etc. are not considered independently but if the multidi-

mensional climatic spheres, which describe the climatic

tolerances of the modern taxa A’ to Z’, are calculated and over-

lapped (e.g., Kühl et al., 2002).

All these quantitative techniques have found wide applica-

tion in paleoclimatology and proved very useful and reliable

in Cenozoic paleoclimate reconstructions (e.g., Utescher et al.,

2000; Mosbrugger et al., 2005). Unfortunately, they suffer

from similar shortcomings as the qualitative indicator taxon

approach. However, the method becomes virtually independent

from taphonomy if many taxa are included in the analysis. For

instance, if some taxa are missing in the fossil fauna or flora

for taphonomic reasons, this will only influence the width of

the resulting coexistence interval (cf. Figure N1) but the new

interval will still bracket the real value. Similarly, the simulta-

neous consideration of many taxa will allow statistical tests to

determine if some modern taxa have changed their climatic tol-

erances as compared to their fossil relatives. For instance, in

Figure N1, the NLR taxon C’ forms a climatic outlier with

respect to parameter I, i.e., the climatic tolerance of C’ does

not overlap with the climatic tolerances of the other NLRs of

the fossil flora/fauna. This clearly indicates that the climatic

tolerance of modern taxon C’ does not reflect the climatic tol-

erance of fossil taxon C. From many such analyses it turns

out that all modern relict taxa, such as the conifer genus

Sequoia, that today have a very restricted distribution, as com-

pared to the distribution area of the fossil taxon, are not very

useful in paleoclimate reconstructions.

A variety of assemblage-based methods that allow quantita-

tive estimates of paleoclimate or paleoenvironment parameters

also exists. The modern analogue technique may serve as an

example. In a first step, a calibration dataset has to be estab-

lished. For this purpose, many modern assemblages covering

a wide range of climatic (or ecologic) conditions are character-

ized with respect to their climatic (or ecologic) requirements.

Then for a given fossil assemblage, a certain number (e.g.,

1–5) of the best “modern analogues” are picked from the cali-

bration dataset; the climatic characteristics of these modern

analogues are averaged and the mean value is used as a paleo-

climate estimate for the fossil assemblage. The classic transfer

functions may be viewed as just another variety of the quanti-

tative assemblage-based methods.

The limitations of these techniques are similar to those

described for the qualitative indicator assemblage approaches:

the NLRs of the fossil taxa and of the fossil assemblages

have to be identified and it is assumed that both fossil assem-

blage and its modern analogue have similar climatic or envir-

onmental requirements. Hence, quantitative assemblage-based

paleoclimate reconstructions are again largely restricted to the

Quaternary, with rapidly increasing uncertainties further back

in time. Calibration of the modern data set and taphonomy play

a crucial role in these techniques. It has to be kept in mind that

there are fossil assemblages, such as the Pleistocene mammoth

steppe, which have no close modern analogues. In addition,

taphonomic filter processes may strongly influence the relative

abundance of taxa in a fossil assemblage and thus hamper the

identification of the nearest living relative assemblage.

Summary

The nearest-living-relative method exists in numerous taxon-

and assemblage-based variations and is still one of the most

powerful approaches to obtain qualitative and quantitative

paleoclimate (or paleoenvironment) information. Taxon-based

approaches relying only on the presence-absence of taxa are

less sensitive to taphonomy than assemblage-based approaches,

which typically also rely on the relative abundance of taxa.

Stratigraphically, taxon-based methods are limited to periods

after the time when the major evolutionary radiation of the

respective taxon group occurred; typically, they cannot be

applied to pre-Cenozoic times. In contrast, assemblage-based

techniques are mostly restricted in their application to Quatern-

ary sequences.

Volker Mosbrugger

Figure N1 General concept of quantitative indicator taxon approaches.

Taxon C’ forms a climatic outlier (further explanation in the text).

608 NEAREST-LIVING-RELATIVE METHOD

Bibliography

Atkinson, T.C., Briffa, K.R., and Coope, G.R., 1987. Seasonal temperatures

in Britain during the past 22,000 years, reconstructed using beetle

remains. Nature, 325, 587–592.

Grichuk, V.P., 1969. An attempt to reconstruct certain elements of the cli-

mate of the northern hemisphere in the Atlantic period of the Holocene.

In Neishtadt, M.I. (ed.), Golotsen. Moscow: Izd-vo Nauka, pp. 41–57.

Kühl, N., Gebhardt, G., Litt, T., and Hense, A., 2002. Probability density

functions as botanical-climatological transfer functions for climate

reconstruction. Quaternary Res., 58, 381–392.

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

for quantitative reconstructions of T e rtiary terrestrial palaeoclimate data

using plant fossils. Palaeogeogr. Palaeoclimatol. Palaeoecol., 134,61–86.

Mosbrugger, V., Utescher, T., and Dilcher, D., 2005. Cenozoic continental

climatic evolution of Central Europe. Proc. Natl. Acad. Sci., 102,

14964–14969.

Utescher, T., Mosbrugger, V., and Ashraf, R.A., 2000. Terrestrial climate

evolution in Northwest Germany over the last 25 million years.

PALAIOS, 15, 430–449.

Cross-references

Animal Proxies, Invertebrates

Animal Proxies, Vertebrates

History of Paleoclimatology

Paleobotany

Paleoclimate Proxies, an Introduction

Pollen Analysis

Transfer Functions

NEOGENE CLIMATES

Introduction

The Neogene includes the Miocene and Pliocene epochs, which

span 23.03–5.33 and 5.33–1.81 million years ago, respectively

Loureus et al., 2004. The Neogene witnessed several major

paleoclimatological changes. In general terms, the Neogene went

from a warm climatic optimum in the late early Miocene through

rapid cooling during the middle Miocene, to the onset of North-

ern Hemisphere Glaciation (NHG) in the Pliocene.

The Miocene

During the Miocene, especially the middle Miocene, the Earth

went through several major changes – both climatic and tectonic.

The climatic transition involved gradual cooling, starting about

15 million years ago, and establishment of major ice sheets on

Antarctica by 10 million years ago (Zachos et al., 2001), as well

as significant changes in global carbon cycling and deep ocean cir-

culation. Initially, researchers argued for minimal continental ice

from sometime in the Cretaceous to the Miocene, with a transition

to so-called “icehouse” conditions in the middle Miocene. How-

ever, evidence now exists in support of the development of glacial

conditions in the early Oligocene, with waxing and waning of glo-

bal ice volume occurring through the Oligocene and early Mio-

cene. Trends in microfossil stable oxygen isotope composition

provide evidence for several glacial intervals during the Miocene

(e.g., Miller et al., 1991). However, a period of global warmth fol-

lowed a major glacial episode at the Oligocene/Miocene bound-

ary, resulting in a reduced latitudinal temperature gradient and

higher average surface water and deep-water temperatures

(Zachos et al., 1997, 2001). At the height of the climatic optimum

(17–15 million years ago) deep-water and high-latitude surface

water temperatures were as much as 6



Cwarmerthantoday

(Shackleton and Kennett, 1975; Savin et al., 1975). The most

dramatic climatic change that occurred during the Miocene took

place around 15 million years ago, during the middle Miocene,

at which time there was evidence for rapid deep-water cooling

and east Antarctic ice-sheet expansion (Figure N2). Evidence from

ice-rafted debris indicates that glaciation in the Nordic seas and

Arctic regions began in the late Miocene. During the late Miocene,

glaciers reached sea level in both the North Atlantic and North

Pacific, resulting in the deposition of ice-rafted debris across both

these oceans.

There are several different hypotheses concerning the cause

of the Miocene climatic optimum and the following cooling.

One of these hypotheses involves changes in atmospheric

content, with an elevated pCO

possibly responsible for

the global warmth in the late early Miocene optimum. Evidence

for Neogene pCO

variability comes from different sources.

Evidence regarding a possible organic carbon control on

pCO

can be found in the stable carbon isotope composition

of microfossils that live on the ocean floor. In the middle

Miocene, there is a positive shift in the stable carbon isotope com-

position of marine carbonates, which is associated with organic

carbon-rich deposits around the Pacific, such as the Monterey

Formation (Vincent and Berger, 1985). Increased upwelling is

thought to be the major cause for this increased organic carbon

burial and subsequent CO

removal. The increase is followed

by a positive shift in the stable oxygen isotope composition of

marine carbonates, indicating a cooling associated with decreased

levels of pCO

. However, Raymo (1994) suggested that the pri-

mary cause of decreasing pCO

during the Miocene was due to

changes in rates of continental chemical weathering. Based on a

strontium isotope record, Raymo (1994)arguedforincreasing

silicate chemical weathering rates as the primary mechanism of

removal. The enhanced weathering rate has been attributed

to a major phase of deformation of the Himalayan orogen that

occurred about 21–17 million years ago. On the other hand,

reconstructions of the evolution of atmospheric pCO

during the

Miocene that are based on the carbon analyses of diunsaturated

alkenones and planktonic foraminifers from deep-sea sediment

cores show no link between atmospheric pCO

and climate

change during the late early and middle Miocene (Pagani et al.,

1999). Another cause of the late early Miocene climate optimum

and East Antarctic ice expansion in the middle Miocene could be

the opening of the Drake Passage between South America and the

Antarctic Peninsula. The opening of the Drake Passage had a

great effect on Southern Ocean circulation.

Several different hypotheses tie Miocene climate change to

deep-water circulation patterns (Wright et al., 1992 and refer-

ences therein). Based on stable carbon and oxygen isotope

records from the Atlantic, Pacific, Indian and Southern oceans,

middle Miocene growth of the East Antarctic Ice Sheet was

controlled by the interplay between Northern Component

Water, Southern Component Water and a water mass originat-

ing in the eastern Tethys (Wright et al., 1992; Flower and

Kennett, 1995). According to Wright et al. (1992) a combina-

tion of the relatively warm Northern Component Water and

Tethyan water masses about 20–16 million years ago produced

a significant meridional heat transport that warmed the deep

ocean by several degrees. At around 16 million years ago,

deep-water temperatures cooled after the fluxes of Northern Com-

ponent Water and Tethyan outflow water were reduced, allowing

the deep oceans to be ventilated by only Southern Component

NEOGENE CLIMATES 609