Odekon M. Encyclopedia of paleoclimatology and ancient environments

Подождите немного. Документ загружается.

precipitation will fall and where arid conditions will prevail.

Similarly, as the continental layout changes with continental

drift, the ocean currents readjust and produce climate change.

Absence of land in the polar regions to anchor potential ice

sheets is believed to have been instrumental in maintaining

the relatively mild temperatures during the Mesozoic period.

Another type of climate forcing and one that is responsible

for the classical shifting of global climate between cold ice

ages and warm interglacial periods over the past several million

years is known as the astronomical or Milankovitch theory

of climate change. This involves slow variations of the Earth’s

orbital parameters due to the gravitational influence of other

planets. As a result, the tilt of the rotational axis of the Earth

with respect to the orbital plane wobbles by more than a degree

with a period of about 41,000 years. This affects the amplit-

ude of the seasonal temperature change. Furthermore, because

the orbit of the Earth is not fully circular, the Earth moves clo-

ser in and farther away from the Sun during the course of the

year. At present, the closest point of approach, or perihelion,

occurs in January, at which point the Earth receives about 7%

more solar energy as compared with July when the Sun-Earth

distance is at its maximum. The seasonal timing of perihelion

occurrence is associated with the precession of the equinoxes,

which takes about 22,000 years for a full cycle. In addition,

the degree of oblateness of the Earth’s orbit varies on a time-

scale of roughly 100,000 years. These perturbations have little

effect on the annual mean solar energy that strikes the Earth,

but they do alter the geographical and seasonal distribution of

solar illumination by as much as 10–20%, being particularly

effective in the polar regions.

Precisely how these slow variations in geographical and

seasonal insolation act to produce an ice age climate is not

yet fully understood. Ice core analysis of trapped air bubbles

in the Antarctic and Greenland Ice Sheets indicates that atmo-

spheric CO

levels 20,000 years ago were roughly half of

present day concentrations. Climate model simulations of ice

age climate show that the radiative forcing characteristic of

ice age climate was about 6 W m

2

less than for the current

climate, due primarily to the effect of vast continental ice

sheets that account for about 3.5 W m

2

, and to the decreased

amount of greenhouse gases (CO

,CH

, and N

O), which con-

tribute about 2.5 W m

2

. This amount of radiative forcing was

sufficient to maintain a global mean temperature 5



C colder

than the present global mean, consistent with the global mean

temperature that has been inferred from polar ice core d

isotope ratios. This implies a climate sensitivity of nearly 1



global mean temperature change per 1 W m

2

of radiative for-

cing for ice age climate, which is basically the same sensitivity

that is obtained from climate model simulations conducted for

current climate.

Nevertheless, a careful analysis of the ice core record shows

that the observed temperature changes tend to precede the CO

and CH

changes by 500–1,000 years. This implies that the

observed changes in CO

and CH

may be positive feedback

responses by the ocean and the biosphere to the Milankovitch

forcing. Just how this CO

and CH

feedback response might

operate has not been fully demonstrated. The distinction

between what constitutes a climate forcing and a climate feed-

back is not a clearly defined concept and depends on the time-

scales that are involved. On the one hand, there are fast

feedback processes such as the evaporation and condensation

of water vapor and the seasonal melting and freezing of snow

and ice. These feedback processes respond quickly to changes

in radiative forcing and provide a positive feedback magnifi-

cation of the original forcing. This is because water vapor is

a strong greenhouse gas and the addition of more water vapor

to the atmosphere produces more greenhouse warming. The

melting of snow and sea ice lowers the albedo of the Earth

and allows more solar radiation to be absorbed. Clouds also fall

in the category of fast feedback processes, but are more compli-

cated in that an increase in high-altitude cirrus would provide

a positive feedback because of their strong greenhouse effect,

while an increase in low-level clouds would increase the plane-

tary albedo and thus provide a negative feedback. On the other

hand, slow geophysical processes such as the growth and decay

of continental ice sheets, changes in vegetation cover, and the

gradual change in atmospheric CO

and CH

levels, may oper-

ate as feedback processes on geological timescales but act as

radiative forcings on short (decadal) timescales.

The susceptibility of the Earth to ice age climate appears

largest when alignment of the orbital parameters is most favor-

able for cold Northern Hemisphere summers. However, in cli-

mate model simulations, the Milankovitch forcing alone is

not enough to initiate an ice age. It appears that a triggering

mechanism, such as strong global cooling from a series of large

volcanic eruptions, or a temporary decrease in solar luminosity,

may be necessary to ensure that the winter snowfall accumula-

tion does not melt during the following summer and becomes

instead the foundation of a new continental ice sheet. However,

once established, ice age climate is stable in the sense that the

prevailing colder global temperature is in equilibrium with the

combined radiative forcings due to the continental ice sheets

and the depleted greenhouse gases.

A closer look at the geological temperature record based on

isotope ratios found in Antarctic ice cores shows a complex

time series of global mean temperature fluctuations. For most

of the past 4 million years, the time series is characterized by

a cold ice age climate. The warm interglacial periods appear

to be relatively brief and tend to occur at intervals of roughly

100,000 years, and last perhaps 10,000–20,000 years. There

are slow periodic fluctuations where the global temperature

undergoes changes of more than several degrees on timescales

of 10,000 years, on which are superimposed faster oscillations

of less than 1,000 years in duration with amplitudes smaller

than 1



C. This only serves to illustrate the limitations of pre-

sent day climate models, which appear capable of calculating

the direct impact of radiative forcings with reasonable accuracy

and in estimating the magnitude of feedback responses of the

fast feedback processes but do not possess the necessary phy-

sics to model the full ice age cycle of climate change.

There may be additional uncertainties and non-linear effects

involved in these slow-process climate forcings as suggested

by the asymmetric nature of the more or less slow and steady

growth of continental ice sheets followed by a more rapid ice

sheet decay. The extent to which an ice sheet will grow or

decay is determined by a complex balance between snow accu-

mulation over the ice sheet relative to ice loss due to melting

and ablation. A strong hydrological cycle is needed to supply

ample moisture for ice sheet growth, and sufficiently cold sum-

mer temperatures are needed to sustain the accumulation of

snow over the ice sheet. Ice sheet melting, however, is not sim-

ply the reverse process of ice accumulation. During ice sheet

decay, ice dynamics aided by ice-melt lubrication injected via

ice crevasses may undermine the ice sheets and lead to an

accelerated breakup, processes that would not be operating dur-

ing ice sheet buildup.

CLIMATE FORCING 175

Radiative nature of climate forcings

The purely radiative aspects of climate forcing have been more

thoroughly studied and can be described more quantitatively.

As illustrated in Figure C52, the atmospheric temperature

decreases steadily with height, due partly to radiative cooling

by atmospheric gases but also due to the convective and

dynamic transport of energy within the troposphere, which acts

to maintain a global mean temperature gradient of about 5.5



1

. About half of the incident solar radiation is absorbed by

the ground surface, from where the absorbed energy is trans-

ported upward in the form of latent and sensible heat and mixed

throughout the troposphere by convective motions. For an atmo-

sphere that is in radiative-convective equilibrium, convective

heat transport along with longwave cooling by atmospheric

water vapor, CO

, clouds, and other gases, work together to bal-

ance the atmospheric heating due to absorbed solar radiation.

Convective heat transport also acts to bind the tropospheric tem-

perature profile to the ground surface and to the very large heat

capacity of the ocean. The stratosphere on the other hand is in a

state of purely radiative equilibrium where the local heating due

to the absorption of solar UV radiation by ozone is balanced by

the longwave cooling due primarily to CO

.Itisthelocalmax-

imum concentration of ozone above the tropopause that pro-

duces the characteristic temperature profile in the stratosphere.

For the current climate, the global mean surface temperature

is about 288K (15



C). The Earth has a planetary albedo of 0.3,

because of which, the global mean solar energy absorbed is

239 W m

2

. For the Earth in radiative equilibrium, this dictates

an effective Planck radiating temperature to space of 255K.

The difference between the surface temperature and the effec-

tive radiating temperature (33



C) defines the total greenhouse

strength of the Earth’s atmosphere. This means that if the atmo-

sphere of the Earth were totally transparent to all radiation

(and if the Earth still had a global albedo of 0.3), the global

mean surface temperature would be 255K (–18



C) instead of

288K. The point within the atmosphere where the physical

temperature is equal to the effective radiating temperature

is determined by the vertical distribution of the longwave

absorbing gases (H

O, CO

,CH

O) and clouds, and

coincides with the point where the effective longwave opacity

(as viewed from space) is unity. For the current climate, this

point occurs at about a height of 6 km.

As illustrated in Figure C52a, the simplest measure of radia-

tive climate forcing is the instantaneous forcing DF

(W m

2

When a radiative forcing such as doubled CO

is applied, the long-

wave fluxes change throughout the atmosphere. By evaluating the

instantaneous radiative flux change at the tropopause, a reason-

ably good estimate of the eventual change in equilibrium surface

temperature is obtained. This is because the tropospheric temp-

erature profile is strongly bound to the surface temperature via

convective heat transport, so that evaluating the flux imbalance

at the tropopause gives a good estimate of the radiative forcing

that is applied to the troposphere-ground system. A better estimate

of the expected surface temperature change is obtained from the

adjusted forcing, DF

, whereby the stratosphere is allowed to

adjust to its new radiative equilibrium before evaluating the radia-

tive flux difference. This provides a more accurate measure of

the radiative flux imbalance that actually drives the climate sys-

tem. Because the heat capacity of the stratosphere is very small,

the stratosphere will reach radiative equilibrium in a matter of

months, long before the surface temperature has had a chance to

respond appreciably.

Figure C53 shows the equilibrium surface temperature

change, DT

, which is obtained after the ocean, with its large

heat capacity, has fully responded to the applied forcing, DF

in the artificial case that no feedback processes are allowed to

operate. As might be anticipated, DT

and DF

, differ only by

a relative scale factor where DT

(



C)  0.3DF

(W m

2

with the exact value of the constant depending somewhat on

the latitudinal, spectral, and height distribution of the applied

forcing. As indicated, instantaneous doubling of CO

increases

the longwave opacity of the atmosphere. This forces the effec-

tive radiating level (the point where longwave opacity is unity)

to a higher altitude where the temperature is colder; therefore,

less thermal energy is radiated to space, thereby trapping the

amount DF

. Since the solar heat input remains fixed, the

Figure C52 Schematic global mean temperature profile of the Earth. (a) Instantaneous forcing DF

evaluated at the tropopause. (b) Adjusted

forcing DF

after allowing stratosphere to cool.

176 CLIMATE FORCING

trapped thermal energy forces the atmospheric and surface tem-

peratures to increase until radiative energy balance is restored.

When feedbacks are allowed to operate, there is a further increase

in atmospheric temperature, as is shown in Figure C53b.Thisis

because the initial warming of the atmosphere by the amount DT

due to doubled CO

forcing induces more evaporation of water

vapor, causes additional melting of snow and sea-ice, increases

the amount of high-altitude cirrus, and decreases low-level

clouds – all of which act to magnify the original temperature

change of the initial radiative forcing perturbation.

The all-feedback response, DT

, can be thought of as being

the sum of the applied no-feedback forcing, DT

, plus a contri-

bution, DT

, that arises entirely due to feedbacks, such that

= DT

+ DT

. In typical doubled CO

climate experiments,

= 1.2



C, with the global mean temperature change attain-

ing DT

= 4.0



C at equilibrium. The ratio f, by which the

no-feedback forcing is magnified, is called the climate feed-

back factor, such that DT

= fDT

, which for current climate

yields f = 3.3.

The strength of the climate feedback response is given by

the ratio g = DT

/DT

, which is also called the feedback gain

factor, and for current climate is about 0.7 (see Albedo feed-

backs). From this we see that the feedback gain factor is related

to the feedback factor by f =(1– g)

1

. In this formulation, the

relative strengths of individual feedback processes can be

directly compared because the gain factors for water vapor,

snow and ice, and cloud feedbacks combine linearly in the

form g = g

+ g

. Thus, the climate system of the Earth

is seen to operate with a strongly positive overall feedback so

that a relatively small change in the gain efficiency of any

one of the feedback processes can have a disproportionately

large effect in magnifying an applied radiative forcing.

The largest uncertainty in estimating the climate system res-

ponse to an applied radiative climate forcing is associated with

determining the strengths of the various feedback processes,

which depend on complicated physical interactions that are dif-

ficult to model. Radiative forcings, on the other hand, can be

calculated with good accuracy and can be expressed in terms

of either the adjusted forcing, DF

, or the no-feedback forcing

response, DT

. The radiative forcing due to changes in atmo-

spheric CO

is shown in Figure C54a, and in Figure C54b for

and N

Figure C53 Schematic illustration of equilibrium response to doubled CO

.(a) No-feedback forcing response DT

.(b) All-feedback response DT

Figure C54 No-feedback forcing DT

for (a)CO

, and (b)CH

and N

O. Filled circles represent current atmospheric amounts. Open circles show DT

for half of the current concentrations.

CLIMATE FORCING 177

These no-feedback forcings show considerable saturation

with increasing absorber concentration, particularly for CO

The curves are roughly logarithmic in nature; hence, the radia-

tive forcing for a doubling of the gaseous concentration is not

overly sensitive to the precise reference point. Also of note, if

all CO

were removed from the atmosphere, the radiative for-

cing would be DT

 7



C, and if we include also the radiative

forcings for the current atmospheric amounts of CH

the CFCs, and ozone, we obtain an additional DT

 3



Together, the greenhouse forcing due to these non-volatile green-

house gases (i.e., greenhouse gases that will not condense

and precipitate from the atmosphere in response to tempera-

ture change) is approximately DT

 10



C. The total green-

house effect of the Earth’s atmosphere (given by the difference

between global mean surface temperature and the effec-

tive radiating temperature of the Earth) is DT

 33



C, which

includes the contributions from all greenhouse gases, including

water vapor, and also the greenhouse contribution due to clouds.

From this, it follows that the ratio f = DT

/DT

, of the total ter-

restrial greenhouse effect to the total no-feedback forcing

due to the non-volatile greenhouse gases, yields a climate sen-

sitivity feedback factor of f  3.3 for the entire atmosphere

of the Earth, which is fully consistent with the feedback sensi-

tivity that is obtained in climate change studies for relatively

small climate forcings such as doubled CO

Andrew A. Lacis

Bibliography

Fleming, J.R., 1998. Historical Perspectives on Climate Change. NY:

Oxford University Press, 194p.

Hansen, J., Lacis, A., Rind, D., Russell, G., Stone, P., Fung, I., Ruedy, R.,

and Lerner, J., 1984. Climate sensitivity: Analysis of feedback mec-

hanisms. In Hansen, J.E., and Takahashi, T. (eds.), Climate Processes

and Climate Sensitivity, Geophysical Monograph 29, Washington,

DC: American Geophysical Union, pp. 130–163.

Hansen, J., Sato, M., and Ruedy, R., 1997. Radiative forcings and climate

response. J. Geophys. Res., 102, 6831–6864.

Cross-references

Albedo Feedbacks

Astronomical Theory of Climate Change

Bolide Impacts and Climate

Carbon Dioxide and Methane, Quaternary Variations

Faint Young Sun Paradox

Mass Extinctions: Role of Climate

Maunder Minimum

Paleoclimate Modeling, Quaternary

Plate Tectonics and Climate Change

Sun-Climate Connections

Volcanic Eruptions and Climate Change

CLIMATE VARIABILITY AND CHANGE, LAST

1,000 YEARS

Geologists and paleoclimatologists study the past on all time-

scales as a basis for future predictions. The past is therefore

highly relevant to modern climate change. It is essential to

document past large-scale climate changes in order to place

recent climate change in a longer-term context. The primary

boundary conditions on the climate (e.g., Earth’s orbital

geometry and global ice mass) have not changed appreciably

over the past millennium. The variations in climate observed

over this timeframe are hence likely to be representative of

the natural climate variability that might be expected in the

absence of any human influence. Placing modern climate

change, including recent global warming, in a longer-term con-

text can thus help establish the role of anthropogenic forcing

(human greenhouse gas concentration increases and aerosol

production) on past and future climate changes.

Approaches to estimating past changes

Instrumental meteorological records can only document large-

scale climate changes over roughly the past 100–150 years.

Relatively few instrumental records are available back into

the eighteenth century. Several distinct approaches nonetheless

exist for describing climate variations over a timeframe of

roughly the past millennium. One such approach is the use of

theoretical models of the climate system that are driven by esti-

mated changes in external parameters or “forcings,” such as

greenhouse gas concentrations and solar output, which tend to

warm the climate, and the frequency and intensity of explosive

volcanic eruptions, which tend to cool the climate through

injecting reflective sulfate aerosol into the atmosphere (Free

and Robock, 1999; Crowley, 2000; Gerber et al., 2003). This

approach to reconstructing climate history is limited by the

imperfectly known history of these changes in forcing, which

are typically estimated indirectly from trapped gas, radioiso-

topes, and volcanic dust signals left behind in ice cores. More-

over, the reconstruction is limited by any uncertainties in the

model’s representation of actual climate processes and res-

ponses. Finally, the approach only indicates the forced compo-

nent of climate change; it cannot estimate the possible role of

internal dynamics (e.g., natural changes in ocean circulation)

in the actual climate changes in past centuries, which are, in

part, unpredictable from a model simulation.

Human documentary evidence provides another source for

reconstructing climate in past centuries. Records of frost dates,

droughts, famines, the freezing of water bodies, duration of

snowcover, and phenological evidence (e.g., the dates of flow-

ering of plants, and the ranges of various species of plants) can

provide insight into past climate conditions (Wigley et al.,

1981; Bradley, 1999). Furthermore, human accounts of moun-

tain glacier retreats and advances during past centuries, wide-

spread historical documentation of weather conditions, and

even a handful of several centuries-long thermometer measure-

ments are available in Europe for more than 1,000 years. Plen-

tiful human documentary evidence is limited, however, to those

regions (primarily Europe and Eastern Asia) where a written

tradition existed. They are thus not useful for documenting

truly global-scale climate variations. Human documentary

records must furthermore be interpreted with caution, as they

are not equivalent in their reliability to actual instrumental mea-

surements of meteorological variables. Typically, these records

provide indirect and potentially biased snapshots in time of

potentially climate-related phenomena.

We must thus turn to more widespread, quantifiable mea-

sures of climate to document the large-scale climate variations

of past centuries. Geological evidence from terminal glacial

moraines indicating the advance of mountain glaciers in past

centuries can also provide inferences into past climate changes.

However, owing to the complex balance between local changes

in melting and ice accumulation, and the effects of topography,

all of which influence mountain glacier extents, it is difficult to

178 CLIMATE VARIABILITY AND CHANGE, LAST 1,000 YEARS

ascertain the true nature of temperature changes simply from

evidence of retreat of mountain. Both increased winter precipi-

tation and cooler summer temperatures can lead to the growth

of a glacier, and large or even moderate glaciers respond slowly

to any underlying climate changes (Folland et al., 2001). Esti-

mates of long-term ground temperature trends from tempera-

ture profiles retrieved from terrestrial borehole data over the

globe can also provide complementary large-scale information

regarding ground-surface temperature trends in past centuries.

However, impacts of changes in seasonal snow cover, land-

surface changes, and other factors unrelated to surface

air-temperature changes may complicate the interpretation of

these data (Folland et al., 2001 ; Mann and Jones, 2003).

“Proxy” climate reconstruction

The most reliable estimates of climate changes in past centuries

are provided by “proxy” indicators of climate change. These

indicators are natural archives of information that, by their bio-

logical or physical nature, record climate-related phenomena.

Certain proxy indicators, including most sediments cores, ice

cores, and preserved pollen cannot record climate changes at

high temporal resolution, and are thus generally not useful for

reconstructing climate changes on centennial or shorter time-

scales. However, high-resolution, annually and/or seasonally

resolved proxy climate records such as growth and density

measurements from tree rings, laminated sediment cores, annu-

ally resolved ice cores, and isotopic information from corals

can be used to describe year-to-year patterns of climate in past

centuries (Jones et al., 2001; Mann, 2001). These indirect mea-

surements of climate change vary considerably in their reliabil-

ity as indicators of long-term climate. While tree-ring data are

the most widespread source of annual proxy climate informa-

tion, they are limited in a number of ways. They only provide

information on sub-polar terrestrial regions, and generally, only

extratropical species are useful for climate reconstruction.

Thus, tree ring data are poorly representative of the full surface

of the globe. The issue of how best to estimate climate variability

on multi-centennial timescales from tree ring data, moreover,

remains unresolved (Briffa and Osborn, 2002; Esper et al.,

2002; Mann and Hughes, 2002), rendering long-term estimates

basedsolelyontreeringdatasomewhat perilous. Coral informa-

tion (e.g., oxygen isotopes and chemical species ratios that

respond to both temperature and salinity changes near the ocean

surface), in contrast, offer information regarding tropical and sub-

tropical climate changes, represent maritime regions, and continu-

ously sample their environment over the full year. However , very

long, multi-century records are rare, and the nature of the possible

influence of “vital effects” (e.g., non-climatic influences on the

coral’s biochemistry) has not yet been confidently established.

Ice cores typically provide information from polar regions (and,

in some cases, high-elevation tropical and extratropical environ-

ments), so that they are spatially complimentary to tree-rings

and corals, but they sample a very small fraction of the global

surface. An incontrovertible interpretation of ice-core oxygen

isotopes in terms of e.g., atmospheric temperature variability,

moreover, remains elusive, and precise annual dating is diffi-

cult. However, when taken together, and particularly when

combined with historical documented information and the few

long instrumental climate records available, these various

proxy sources can provide a mutually complementary, global

sampling (Figure C55) of climate changes in past centuries.

Such “multiproxy” assemblages of data can be used to recon-

struct past large-scale patterns of surface temperature change,

drought, and atmospheric circulation, by establishing a statisti-

cal relationship between the proxy data network and modern

instrumental climate records (Mann, 2001). Indices of modes

or patterns of climate variability, such as the El Nino/Southern

Oscillation or ENSO and the North Atlantic Oscillation (NAO),

which may play an important role in determining patterns of

past climate variability, can also be reconstructed in a similar

manner (see e.g., Jones et al., 2001 for a summary). This

approach has also been used to reconstruct annual patterns of

surface temperature change in past centuries (Mann et al.,

1998; Briffa et al., 2001). Hemispheric or global mean surface

temperature estimates can be derived by spatial averaging

of such estimates of past large-scale patterns of surface tem-

perature change. It is most meaningful to estimate the average

surface temperature changes for the Northern Hemisphere,

where considerably more widespread past data is available

(e.g., Figure C55).

Climate history of the past 1,000 years

Climate reconstructions from proxy data networks have

allowed for reconstructions of large-scale temperature patterns

Figure C55 Geographic distribution of “proxy” and historical /instrumental indicators used in large-scale annual surface temperature

reconstructions back to

AD 1400 (adapted from Mann et al., 1998).

CLIMATE VARIABILITY AND CHANGE, LAST 1,000 YEARS 179

over roughly the past millennium. The detailed spatial patterns

of these reconstructions provide insights into the underlying

governing factors. For example, substantial patterns of warm-

ing and cooling over the extratropical Northern Hemisphere

that vary from year-to-year appear to relate to the influence

of the North Atlantic Oscillation (NAO) or Arctic Oscillation

(AO) atmospheric circulation pattern. A related pattern of cli-

mate variability that appears to have an origin to internal, nat-

ural variations in the North Atlantic ocean-atmosphere system

(Delworth and Mann, 2000) also appears to be important on

multidecadal and century timescales. Other patterns of surface

temperature variability appear to be associated with the global

impacts of the El Nino/Southern Oscillation phenomenon.

The complexity of the resulting regional variations in tempera-

ture trends underscores the premise that it is perilous to draw

conclusions regarding hemispheric or global temperature trends

from isolated regional information.

Different estimates of Northern Hemisphere mean tempera-

ture changes over the past millennium have been made, using

different sources of proxy data, some of which are representa-

tive of the full Northern Hemisphere, and others which are

more indicative of the extratropical regions of the Northern

Hemisphere. These reconstructions can also be compared with

model estimates of forced Northern Hemisphere mean tempe-

ratures changes over the past millennium (Figure C56). These

various estimates indicate relatively modest variations in

Northern Hemisphere mean temperature over the past 1,000

years, prior to the marked warming of the twentieth century.

Hemispheric mean temperatures appear, for example, to have

been slightly higher (by a couple of tenths of a degree Celsius)

during the period

AD 1000–1400 relative to the later, colder

period

AD 1400–1900 (associated, respectively, with the

often-used labels of “Medieval Warm Period” and “Little Ice

Age”). However, even the early interval of relative warmth

does not approach the magnitude of the hemispheric warmth

of the latter twentieth century. The modest estimated changes

in Northern Hemisphere average temperatures are consistent

with evidence of the amplitude of natural (pre-anthropogenic)

changes in the carbon cycle estimated from long-term carbon

dioxide measurements (Gerber et al., 2003).

Some reconstructions of hemispheric temperature changes

based on extratropical, continental tree-ring data suggest wider

swings in temperature, including greater cooling during the

seventeenth to nineteenth century than is evident in the instru-

mental, model, or other proxy-based estimates (Esper et al.,

2002). The greater variability in this case probably arises from

the emphasis on extratropical continental regions and the exclu-

sive use of summer-temperature-sensitive tree ring chronologies

(Mann, 2002). While European temperatures, for example, indi-

cate a distinct cold phase from the seventeenth to nineteenth cen-

turies, most hemispheric estimates (Figure C56) suggest a

relatively steady, long-term cooling. It is probable that greater

cooling in continental regions such as Eurasia and North America

during the “Little Ice Age” resulted from changes in the NAO

atmospheric circulation pattern. Recent evidence suggests that

the NAO can be altered, for example, by changes in the output

of the Sun (Shindell et al., 2001). Such changes may have led to

enhanced winter cooling in Europe during the eieghteenth century,

when solar output was slightly weaker, and winter atmospheric

circulation patterns in Europe appear to have been significantly

altered, favoring exposure to cold, dry, Siberian air masses. Other

complications arise from the use of summer temperature-sensitive

tree-ring proxy information alone, which may exaggerate the

cooling response to volcanic eruptions, which is greatest both

over the continents and during summer (Mann, 2002).

Ground temperature changes estimated from boreholes in

past studies (Huang et al., 2000) appeared to conflict with the

hemispheric temperature trends evident in the instrumental

and proxy-based estimates shown in Figure C56. Recent work

has shown, however, that once these estimates are adjusted

Figure C56 Comparison between various estimates of Northern Hemisphere temperature changes in past centuries. Annual mean Northern

Hemisphere surface temperatures are reconstructed from proxy data (Crowley and Lowery, 2000; Mann et al., 1999) and theoretical energy balance

models (Crowley, 2000), as well as warm-season, extratropical Northern Hemisphere temperature trends (Esper et al., 2002; Jones et al., 1998), scaled

to match the full Northern Hemisphere annual temperature record from 1856–1980 for direct comparison (Jones et al., 1999). The smoothed

reconstructions highlight variations on timescales greater than 40 years; uncertainties (95% confidence interval) in the smoothed Mann et al.

(1999) reconstruction are also shown, as are the instrumental Northern Hemisphere temperature records through the mid eighteenth century

(Mann and Jones, 2003). Estimates of ground surface temperature changes from terrestrial borehole data based are also shown, based on

(a) a simple (unweighted) arithmetic average of available borehole data (Huang et al., 2000), (b) an areal-weighted average of the borehole

data, and (c) statistically optimal hemispheric surface temperatures from available borehole data (Mann and Jones, 2003).

180 CLIMATE VARIABILITY AND CHANGE, LAST 1,000 YEARS

for spatial sampling biases (Briffa and Osborn, 2002; Mann

and Jones, 2003) and the influence of factors unrelated to

annual mean surface temperature changes, such as changes in

the extent and duration of seasonal snowfall (Mann and Jones,

2003), the borehole estimates come into reasonable agreement

with the other proxy and model-based estimates (Figure C56).

Implications

Recent research seeking to reconstruct and model climate

changes of the past 1,000 years highlights the importance of

taking into account regional variations in estimating climate

changes in past centuries. While reconstructions of Northern

Hemisphere temperature changes in past centuries, for exam-

ple, confirm other (e.g., human documentary) evidence of the

phenomenon of a 1–2



C cooling in the winter during the “Lit-

tle Ice Age” of the seventeenth to nineteenth centuries in

Europe, they do not indicate a continuous period of similar mag-

nitude cooling that is synchronous with cold conditions in Eur-

ope. The best current estimates suggest modest amplitude

changes (considerably less than 1



C) at the hemispheric scale,

prior to the twentieth century warming trend. While internal cli-

mate factors are important in describing regional patterns of

climate change, most of the observed variations in hemispheric

mean temperature are consistent with the response of the climate

to external forcing including volcanic and solar influences and,

during the twentieth century in particular, anthropogenic influ-

ences. The comparisons between estimated and modeled changes

imply a sizeable sensitivity of the climate system to anthropo-

genic influences of roughly 2



C warming for a doubling of car-

bon dioxide concentrations from their pre-industrial levels.

Michael E. Mann

Bibliography

Bradley, R.S., 1999. Paleoclimatology: Reconstructing Climates of the

Quaternary. San Diego: Harcourt, Academic Press, 610pp.

Briffa, K.R., and Osborn, T.J., 2002. Blowing hot and cold. Science, 295,

2227–2228.

Briffa, K.R., Osborn, T.J., Schweingruber, F.H., Harris, I.C., Jones, P.D.,

Shiyatov, S.G., and Vaganov, E.A., 2001. Low-frequency temperature

variations from a northern tree ring density network. J Geophys. Res.,

106, 2929–2941.

Crowley, T.J., 2000. Causes of climate change over the past 1000 years.

Science, 289, 270–277.

Crowley, T.J., and Lowery, T., 2000. How warm was the Medieval Warm

Period? Ambio, 29,51–54.

Delworth, T.L., and Mann, M.E., 2000. Observed and simulated multideca-

dal variability in the Northern Hemisphere. Clim. Dyn., 16, 661–676.

Esper, J., Cook, E.R., and Schweingruber, F.H., 2002. Low frequency sig-

nals in long tree-ring chronologies for reconstructing past temperature

variability. Science, 295, 2250–2253.

Folland, C.K., Karl, T.R., Christy, J.R., Clarke, R.A., Gruza, G.V., Jouzel, J.,

Mann, M.E., Oerlemans, J., Salinger, M.J., and Wang, S.-W., 2001.

Observed climate variability and change. In Houghton, J.T. et al., (eds.)

Climate Change 2001: The Scientific Basis. Cambridge, UK: Cambridge

University Press, pp. 99–181.

Free, M., and Robock, A., 1999. Global warming in the context of the

Little Ice Age. J. Geophys. Res., 104, 19057–19070.

Gerber, S., Joos, F., Bruegger, P.P., Stocker, T.F., Mann, M.E., Sitch, S.,

and Scholze, M., 2003. Constraining temperature variations over the

last millennium by comparing simulated and observed atmospheric

CO2. Clim. Dyn., 20, 281–299.

Huang, S., Pollack, H.N., and Shen, P.-Y., 2000. Temperature trends over

the past five centuries reconstructed from borehole temperature. Nature,

403, 756–758.

Jones, P.D., Briffa, K.R., Barnett, T.P., and Tett, S.F.B., 1998. High-

resolution paleoclimatic records for the last millennium: Interpretation,

integration and comparison with circulation model control-run temp-

eratures. The Holocene, 8, 455–471.

Jones, P.D., New, M., Parker, D.E., Martin, S., and Rigor, J.G., 1999.

Surface air temperature and its changes over the past 150 years. Rev.

Geophys.,

37, 173–199.

Jones, P.D., Osborn, T.J., and Briffa, K.R., 2001. The evolution of climate

over the Last Millennium. Science, 292, 662–667.

Mann, M.E., 2001. Climate during the past millennium. Weather, 56,

91–101.

Mann, M.E., 2002. The value of multiple proxies. Science, 297,

1481–1482.

Mann, M.E., and Hughes, M.K., 2002. Tree-ring chronologies and climate

variability. Science, 296, 848.

Mann, M.E., and Jones, P.D., 2003. Global surface temperatures over

the past two millennia. Geophys. Res. Lett., 30(15), 1820,

Mann, M.E., Bradley, R.S., and Hughes, M.K., 1998. Global scale temp-

erature patterns and climate forcing over the past six centuries. Nature,

392, 779–787.

Mann, M.E., Bradley, R.S., and Hughes, M.K., 1999. Northern Hemisphere

temperatures during the Past Millennium: Inferences, uncertainties,

and limitations. Geophys. Res. Lett., 26, 759–762.

Shindell, D.T., Schmidt, G.A., Mann, M.E., Rind, D., and Waple, A., 2001.

Solar forcing of regional climate change during the Maunder Minimum.

Science, 294, 2149–2152.

Wigley, T.M.L., Ingram, M.J., and Farmer, G., 1981. Past climates and their

impact on man: A review. In Wigley, T.M.L., Ingram, M.J., and Farmer, G.

(eds.), Climate and History. Cambridge, UK: Cambridge University Press,

pp. 3–50.

Cross-references

Climate Forcing

Little Ice Age

Medieval Warm Period

North Atlantic Oscillation (NAO) Records

Paleoclimate Proxies, an Introduction

Sun-Climate Connections

Volcanic Eruptions and Climate Change

COAL BEDS, ORIGIN AND CLIMATE

Introduction

Coal is defined in the Glossary of Geology (1997, p. 122) as

follows:

“A readily combustible (sedimentary) rock containing more

than 50% by weight and more than 70% by volume of carbo-

naceous material, including inherent moisture, formed from

compaction and induration of variously altered plant remains

similar to those in peat. Differences in the kinds of plant mate-

rials (type), in degree of metamorphism (rank), and in the range

of impurity (grade) are characteristic of coal.”

Although coal, by definition, may be composed of nearly

50% mineral matter, most commercial-grade coal in the United

States is mined from coal beds that generally contain less than

20% mineral matter by weight (excluding mineral partings),

and commonly less than 10% by weight. The U.S. Geological

Survey excludes impure coal/coaly material containing more

than 33% ash by weight from resource/reserve estimates

(Wood et al., 1983, p. 8). Coal with these grade characteristics

can develop only under very limited conditions of peat forma-

tion. Grade characteristics indicate that resource-quality coal

beds are derived from peat that formed in environments that

were relatively free of dissolved solids and mineral content.

COAL BEDS, ORIGIN AND CLIMATE 181

Furthermore, in order to form commercial grade coal beds, the

concentration of mineral matter in the original peat must have

been relatively low, because metamorphism following burial

of peat tends to increase mineral content. The heat and pressure

involved in the transformation (metamorphism or “coalifica-

tion”) of peat to coal liberates significant amounts of the origi-

nal organic matter in the form of carbon dioxide, methane, and

water. The liberation and loss of organic matter during this

transformation (metamorphism) of peat to coal may effectively

double the mineral content. For example, 25% mineral matter

in peat (dry basis) may be concentrated to 50% mineral

matter when peat is transformed (coalified) to a medium vola-

tile rank coal.

Climate and peat/ coal formation

Although modern wetlands containing organic-rich sediment

occur in a variety of climate settings around the world, only

those deposits that contain less than approximately 25% (dry

basis) mineral matter should be considered as precursors for

coal. Other organic-rich deposits where mineral matter concen-

trations are greater than 25% will become carbonaceous shale

or marl following burial and metamorphism. Cecil et al.

(1985) pointed out a relation between climate and modern pre-

cursors of most commercial-grade coal. These modern coal pre-

cursors occur predominantly in two types of peat-forming

environments, designated as type A and type B (Table C6

and Table C7), which can be differentiated, respectively, by

selected parameters and conditions of peat-formation.

The primary variables that control the mineral matter con-

tent of peat (i.e., peat quality) are nutrient availability and pH

of peat-forming environments. Thus, the highest quality peat/

coal precursors have the lowest mineral matter content, and

they tend to have both a low pH and a low nutrient status (oli-

gotrophic conditions). Oligotrophic peat deposits occur under

humid to perhumid climates (see Cecil, 2003 for climate defini-

tions), where relatively high precipitation causes intense chemi-

cal weathering (leaching) and removal of soluble mineral

matter from peat. These climates also cause removal of nutri-

ents and alkaline materials from mineral soils (pedogenesis)

that are spatially and temporally associated with peat forma-

tion. Mesotrophic peat deposits tend to occur where the climate

is marginally humid. The quality of mesotrophic peat is mar-

ginal as a precursor to coal because nutrients are available that

promote microbial decay of organic matter. Furthermore, the

nutrients tend to remain sequestered in peat as mineral matter

through ion exchange processes. Eutrophic peat deposits, by

definition, have a relatively high nutrient status and they gener-

ally contain relatively high concentrations of alkaline materials,

which result in a near neutral pH that is conducive to high rates

of microbial degradation of organic matter. These deposits tend

to occur in climates that are moist subhumid or sometimes in

drier climates, where soluble salts are enriched in surficial

waters through evapotranspiration. As a result, eutrophic peat

is invariably enriched in mineral matter to the extent that the

peat would become very impure coal or carbonaceous shale

upon burial and metamorphism.

In the modern world, the distribution of peat deposits (as

precursors of coal beds) empirically illustrates climate controls

on peat formation. Laterally extensive, high quality (low ash)

peat deposits form predominantly in regions where monthly

precipitation approximates or exceeds evapotranspiration (humid

to perhumid climates). These regions include certain global

climatic belts and regions that have maritime climates. The

most widespread modern global climate belt of peat formation

Table C6 Parameters and conditions that control peat formation for the two predominant types of peat-forming environments (modified from

Cecil et al. (1985) with climate terminology from Cecil, 2003)

Parameter Type A condition Type B condition

Climate Perhumid Humid

Water source Ombrogenous (rain) Topogenous (rain water, surface water, and ground water)

Nutrient availability Oligotrophic (impoverished) Oligotrophic to mesotrophic to eutrophic (impoverished to rich)

pH <4 4 to ~7

Eh ? ?

Microbial activity Low (cellulose preserved) High (cellulose degraded)

Mechanism of degradation Primarily chemical Primarily microbial

Table C7 Descriptors and relative characteristics of peat-forming environments (modified from Cecil et al., 1985)

Descriptors of peat Type A characteristics Type B characteristics

environments

Surface morphology Domed Planar

Flora communities Low diversity; Zoned distribution; Xeromorphic physiognomy High diversity; Random distribution; Luxuriant physiognomy

Ash content Uniformly low Low to high

Sulfur content Uniformly low Low to high

Nitrogen content Uniformly low Low to high

Cation exchange

capacity

High High

Specific conductivity Low High

Base saturation Low High

[Ca

] Low High

Fiber content Fibric Hemic to sapric

Bioorganic sulfide Low High

Bioorganic methane Low High

182 COAL BEDS, ORIGIN AND CLIMATE

occurs in the vicinity of 60



North latitude along the edge of

the atmospheric polar front. Polar fronts result from high atmo-

spheric pressure and descending air at the poles that moves out

from the polar region where it is then overridden by warmer

and moister air in the vicinity of 60



latitude. The cooling of

the warmer and moister air as it rises results in condensation

and precipitation at the polar front. Even though precipitation

is not particularly high in the region of polar fronts, the climate

is humid to perhumid because cool summer temperatures

and often freezing winter temperatures inhibit evaporation.

At 60



South latitude, however, the absence of landmasses

precludes widespread peat at this latitude in the Southern

Hemisphere.

Areally extensive peat deposits also occur in certain equator-

ial regions. Thick (up to 13 meters) and extensive (over 50 km

across) modern deposits of domed peat occur in equatorial

Indonesia and Malaysia. Domed peat deposits, whose moisture

source is exclusively precipitation, form in this region in

response to a perhumid climate. The perhumid climate results

from the presence of the nearly permanent low pressure,

induced by the rising air of the intertropical convergence zone

(ITCZ). In addition, a maritime effect occurs as moisture-laden

weak winds cool as they move inland and rise over the island

landmasses, thus causing precipitation that supplements rainfall

associated with the low pressure of the ITCZ.

Further examples of peat formation controlled by maritime

climate can be found in the British Isles where prevailing wes-

terlies from the Atlantic Ocean cool as they rise upon reaching

landmasses. The cooling of the westerly winds results in preci-

pitation that contributes to the formation of peat deposits that

blanket upland slopes. The climate in this cool temperate region

is humid to perhumid because humidity and precipitation (in the

form of fog, mist, and rain) are relatively high, whereas evapo-

transpiration is relatively low. Although it is highly unlikely that

blanket peat deposits in upland areas can be preserved in the

rock record, such deposits nevertheless serve to illustrate the

relation between peat formation and wet climates.

Summary and conclusions

Modern peat deposits that could be precursors of resource-

quality coal beds occur predominantly under humid to perhumid

climatic conditions (Cecil, 2003; Cecil and Dulong, 2003).

By analogy, coal beds containing less than approximately

30% ash must have been derived from peat that formed under

humid to perhumid climatic conditions. Although some coali-

fied terrestrial organic matter may have been deposited in

sedimentary strata that were deposited under a wider range of

climatic conditions, resource-quality coal beds that are discrete

and areally extensive accumulated under humid to perhumid

paleoclimate conditions.

C. Blaine Cecil and Sandra G. Neuzil

Bibliography

Cecil, C.B., 2003. The concept of autocyclic and allocyclic controls on

sedimentation and stratigraphy, emphasizing the climatic variable. In

Cecil, C.B., and Edgar, N.T. (eds.), Climate Controls on Stratigraphy:

SEPM (Society for Sedimentary Geology). Special Publication, 77,

pp. 13–20.

Cecil, C.B., and Dulong, F.T., 2003. Precipitation models for sediment sup-

ply in warm climates. In Cecil, C.B., and Edgar, N.T. (eds.), Climate

Controls on Stratigraphy: SEPM (Society for Sedimentary Geology).

Special Publication, 77, pp. 21–27.

Cecil, C.B., Stanton, R.W., Neuzil, S.G., Dulong, F.T., Ruppert, L.F., and

Pierce, B.S., 1985. Paleoclimate Controls on Late Paleozoic Sedimenta-

tion and Peat Formation in the Central Appalachian Basin (U.S.A.). Int.

J. Coal Geol., 5, pp. 195–230.

Jackson, J.A., (ed.), 1997. Glossary of Geology, 4th edn. American Geol.

Institute, Virginia, 469p.

Wood, G.H., Kehn, T.M., Carter, M.D., and Culberston, W.C., 1983. Coal

Resource Classification of the U.S. Geological Survey: Geological Sur-

vey Circular 891. Reston, VA: U.S. Geological Survey, 65p.

Cross-references

Continental Sediments

Sapropels

Sedimentary Indicators of Climate Change

COASTAL ENVIRONMENTS

Introduction

The coastal zone, extending from ca. 10 m depth on the inner shelf

to the landward limit of tidal action in rivers, occupies a small frac-

tion of the Earth’s surface. Despite this, coastal deposits are an

important component of the stratigraphic record because they con-

tain a large amount of information about the environmental condi-

tions that prevailed during deposition. The geomorphology of the

coastal zone (Figure C57) and the characteristics and stratigraphic

architecture of the corresponding deposits (Figure C58) are influ-

enced by many interdependent factors. The most notable of

these are: (a) the history of sea (or lake) level, which, together with

sediment input, determines whether the coastline is migrating

landward (i.e., transgressing) or seaward (i.e., regressing or pro-

grading); (b) the relative influence of waves, tidal currents, and

river outflow, which determines the nature of the resulting facies;

history and the resulting stratigraphic stacking pattern of coastal

deposits; and (d) the climate, which determines the sea-level

history, the intensity of wave action, the amount of fluvial runoff,

the presence of ice, and the role of plants and animals. The inter-

action of these many factors produces enormous geomorphic

variability and complexity in both modern coastal environments

(Bird, 2000; Woodroffe, 2002) and their ancient counterparts

(Pratt et al., 1992; Reading and Collinson, 1996).

Relative sea-le vel history

The direction of shoreline movement is the single most impor-

tant control on the geomorphology and preserved stratigraphic

expression of coastal environments(Boyd et al., 1992).Regressive

situations are associated with the seaward growth of deltas

(Figure C57a, b), the development of strandplains (also called

beach-ridge plains) by prograding beaches and shorefaces, and

the seaward build-out of open-coast tidal flats (Figure C57c).

On the other hand, barrier islands, lagoons, and estuaries

(Figure C57d) exist only where there is, or very recently has been,

transgression. Thus, such environments are particularly common

along modern coasts because of the post-glacial sea-level rise.

Here, we follow the geological definition of “estuary” proposed

by Dalrymple et al. (1992), in which an estuary is a transgres-

sive coastal embayment (usually a drowned river valley) that

receives sediment from both river and marine sources. This defini-

tion is preferred to the salinity-based formulation of Pritchard

COASTAL ENVIRONMENTS 183

(1967), which considers any semi-enclosed coastal area with

brackish water an estuary, regardless of whether it is regressive

or transgressive.

The distinction between regressive and transgressive coasts

is reflected in the vertical stacking pattern of coastal facies:

shallower-water facies overlay deeper-water deposits in regressive

settings, and the reverse in transgressive successions. Regressive

successions are usually more or less complete and gradational

(Figure C58a) becausethe shorelineis migratinginto deeper water.

Only in situations where regressions occur under conditions of

falling sea level (i.e., a forced regression; Posamentier et al.,

1992) is it common to generate a succession that may contain an

erosion surface, with sandy, shallow-water facies lying abruptly

on muddy, offshore deposits (Figure C58a). By contrast, transg-

ressive successions are almost everywhere punctuated by one or

more discontinuities (Figure C58b,c) at which the deposits of a

more seaward environment lie abruptly on a more landward facies

(e.g., brackish-water lagoonal mud overlying fluvial deposits).

Such discontinuities are called flooding surfaces; if there is signif-

icant erosion, such as occurs where the shoreface migrates land-

ward, they are also termed a ravinement surface (Figure C58a,c).

Physical processes

River currents, waves, and tidal currents are responsible for

the geomorphology of coastal environments (Hayes, 1975;

Wright, 1985; Boyd et al., 1992) and the nature of the asso-

ciated deposits. Consequently, coastal environments and their

ancient counterparts are commonly classified as river-, wave- or

tide-dominated.

River-dominated coastal environments occur at the mouths

of rivers, in settings where wave and tidal influence are not

important, and generally consist of deltas with either a birds-

foot or lobate morphology (Figure C57a; Wright, 1985). Distri-

butary mouth bars occur directly seaward of the river mouth(s)

and consist of a succession that passes upward from prodeltaic

mud to the sand of the mouth bars. Interdistributary bays are

developed between the distributary mouths (Figure C57a) and

are sites where muddy and organic-rich sediments accumulate.

Wave-dominated coasts are generally straight or cuspate, and

are fronted by beaches that pass seaward into the shoreface and

offshore shelf (Figure C57d). In progradational settings, progres-

sive seaward migration of the coast produces a strandplain, on

Figure C57 Typical examples of modern coastal environments. (a) The presently active birdsfoot delta of the Mississippi River, United States. Note

the elongate distributary channels, flanked by levees and separated by shallow bays and marshes. (b) Subaerial plain of the tide-dominated Rajang

River delta, Malaysia. Note the complex network of sinuous tidal channels. (c) Muddy, carbonate tidal flats, cut by meandering tidal channels,

Andros Island, Bermuda. (d) Transgressive coast with barrier islands, tidal inlet, back-barrier lagoon, and drowned river-mouth estuaries (upper

right), Pensacola, Florida, United States. The elongate area of land immediately behind the modern barrier is an older barrier island formed during a

late Pleistocene highstand of sea level (all photos except c courtesy of NASA; c courtesy of N. P. James, Queen’s University).

184 COASTAL ENVIRONMENTS