Cotton W.R., Pielke R.A. Human Impacts on Weather and Climate
Подождите немного. Документ загружается.
Atmospheric radiation 169
10
100
1000
–4 –3 –2 –1 0
–3 –2.5 –2 –1.5 –1 –0.5 0 0.5 1
Pressure (hPa)
10
100
1000
Pressure (hPa)
Longwave heating (K day
–1
)
[CO
2
] = 0
[CO
2
] = 280 ppmv
[CO
2
] = 360 ppmv
[CO
2
] = 560 ppmv
Change in longwave heating (K day
–1
)
Δ 280
– 0
Δ 360
– 280
Δ 560
– 280
Figure 8.10 Longwave heating rates (upper) and change in longwave heating
rates (lower) for tropical atmosphere with perturbed carbon dioxide. Figure
courtesy of Norman Wood, Colorado State University.
and subarctic summer profiles. For the subarctic winter profile, this doubling
of carbon dioxide concentration produces an increase in downwelling longwave
flux similar in magnitude to that for a 10% increase in water vapor mixing
ratio.
A number of factors potentially contribute to these differences in the effects
of water vapor and carbon dioxide on heating rates and fluxes. Firstly, water
vapor is significantly more prevalent in the lower troposphere than in the upper
troposphere and stratosphere. Consequently, the scaling approach used here to
perturb the trace gas amounts tend to produce stronger perturbations of water
vapor mass mixing ratio in the lower troposphere than at higher altitudes. This
effect is less significant for carbon dioxide since this gas is more uniformly
mixed.
170 Overview of global climate forcings and feedbacks
10
100
1000
10
100
1000
–5 –4.5 –4 –3.5 –3 –2.5 –2 –1.5 –1 –0.5 0
–3 –2.5 –2 –1.5 –1 –0.5 0 0.5 1
Longwave heating (K day
–1
)
Change in longwave heating (K day
–1
)
Pressure (hP
a)
Pressure (hP
a)
[CO
2
] = 0
[CO
2
] = 280 ppmv
[CO
2
] = 360 ppmv
[CO
2
] = 560 ppmv
280 – 0
360 – 280
560 – 280
Figure 8.11 Longwave heating rates (upper) and change in longwave heating
rates (lower) for subarctic summer atmosphere with perturbed carbon dioxide.
Figure courtesy of Norman Wood, Colorado State University.
In terms of radiative factors, the heating rate in a layer of the atmosphere is a
function of the spectrally varying absorption–emission characteristics of the layer,
the spectral fluxes incident on the layer, and the layer temperature. The absorption
by a layer is a function of the abundances of absorbing gases in the layer. In the
longwave spectral region in which carbon dioxide is a significant absorber (for
wavelengths of about 125 m and longer) water vapor is also radiatively active.
For a wavelength at which water vapor is already significantly absorbing, the
addition of an amount of carbon dioxide to the layer will cause relatively little
increase in the flux absorbed by the layer and thus cause relatively little increase
in the radiative heating of the layer.
The spectral fluxes incident on the layer are also a function of the tempera-
tures and emission characteristics of the layer’s surroundings. For a layer with
given absorption characteristics, a stronger incident flux will cause more flux to
Atmospheric radiation 171
10
100
1000
–3 –2.5 –2 –1.5 –1 –0.5 0 0.5 1
–3 –2.5 –2 –1.5 –1 0.5 0
Pressure (hPa)
10
100
1000
Pressure (hPa)
Longwave heating (K day
–1
)
[CO
2
] = 0
[CO
2
] = 280 ppmv
[CO
2
] = 360 ppmv
[CO
2
] = 560 ppmv
Change in longwave heating (K day
–1
)
Δ 280
– 0
Δ 360
– 280
Δ 560
– 280
Figure 8.12 Longwave heating rates (upper) and change in longwave heating
rates (lower) for subarctic winter atmosphere with perturbed carbon dioxide.
Figure courtesy of Norman Wood, Colorado State University.
be absorbed by the layer and contribute to heating. In particular in the lower
troposphere, the proximity of the warm surface of the Earth contributes to this
effect. A warmer surface temperature (as in, for example, the tropical or sub-
arctic summer profiles used here) will contribute to enhanced heating in the
lower troposphere as opposed to a cooler surface (as in the subarctic winter
profile).
Finally, the temperature of the layer itself influences the amount of flux emitted
by the layer. A warmer layer will emit flux more strongly, and thus have a greater
tendency for cooling, than will a cooler layer. Profiles with warmer temperatures
in the lower troposphere, such as the tropical profile, have stronger cooling in the
lower troposphere than will a profile, such as the subarctic winter, that has cooler
temperatures in the lower troposphere.
172 Overview of global climate forcings and feedbacks
10
100
1000
–4 –3 –2 –1 0
–3 –2 –1 0 1
Pressure (hPa)
10
100
1000
Pressure (hPa)
Longwave heating (K day
–1
)
No H
2
O
Base H
2
O
Base H
2
O × 1.05
Base H
2
O × 1.1
Δ Base
– No H
2
O
Δ Base
× 1.05 – Base
Δ Base
× 1.1 – Base
Change in longwave heating (K day
–1
)
CO
2
= 280 ppmv for all cases
Figure 8.13 Longwave heating rates (upper) and change in longwave heating
rates (lower) for tropical atmosphere with perturbed water vapor. Figure courtesy
of Norman Wood, Colorado State University.
The downwelling flux at the surface is a function of the emission characteris-
tics of the atmosphere, in particular the profile of the derivative of transmission
with respect to height, known as the weighting function, and the temperature
profile of the atmosphere. The height at which the weighting function peaks
generally indicates the level of the atmosphere from which emission from the
atmosphere most effectively reaches the surface, and this peak can be either
broad (indicating the flux reaching the surface is strongly blended from differ-
ent levels of the atmosphere) or narrow (indicating the flux reaching the surface
is strongly selected from that particular level of the atmosphere). As trace gas
concentrations in the lower troposphere increase, the general tendency is for
the weighting function to shift lower in the atmosphere. For temperature pro-
files which decrease with height, this shift leads to increased emission to the
surface.
Atmospheric radiation 173
10
100
1000
–4 –3 –2 –1 0
–3 –2 –1 0 1
Pressure (hPa)
10
100
1000
Pressure (hPa)
Longwave heating (K day
–1
)
No H
2
O
Base H
2
O
Base H
2
O x 1.05
Base H
2
O x 1.1
Change in longwave heating (K day
–1
)
Base
– No H
2
O
Base
× 1.05–Base
Base
× 1.1–Base
CO
2
= 280 ppmv for all cases
Figure 8.14 Longwave heating rates (upper) and change in longwave heating
rates (lower) for subarctic summer atmosphere with perturbed water vapor.
Figure courtesy of Norman Wood, Colorado State University.
The downwelling fluxes at the surface for the subarctic profile appear less
sensitive to changes in carbon dioxide and water vapor concentrations than do
the fluxes for the tropical and subarctic summer profiles. The subarctic winter
profile has a relatively weak lapse rate in the lowest part of the troposphere,
so changes in the position of the weighting function may have had little effect
on the downwelling fluxes. In addition, the water vapor amounts in the sub-
arctic winter profile are considerably smaller than those in the two other pro-
files. Since a scaling factor was used to perturb water vapor amounts for this
study, the change in water vapor mixing ratio in the lower troposphere would
be considerably smaller for the subarctic winter profile than for the two other
profiles. This approach probably contributed to cause a less significant change in
the weighting function for the subarctic winter case than that for the other two
cases.
174 Overview of global climate forcings and feedbacks
10
100
1000
–4 –3 –2 –1 0
–3 –2 –1 0 1
Pressure (hPa)
10
100
1000
Pressure (hPa)
Longwave heating (K day
–1
)
No H
2
O
Base H
2
O
Base H
2
O × 1.05
Base H
2
O × 1.1
Δ, (Base
– No H
2
O)
Δ, (Base
× 1.05 – Base)
Δ, (Base
× 1.1 – Base)
Change in longwave heating (K day
–1
)
CO
2
= 280 ppmv for all cases
Figure 8.15 Longwave heating rates (upper) and change in longwave heating
rates (lower) for subarctic winter atmosphere with perturbed water vapor. Figure
courtesy of Norman Wood, Colorado State University.
8.3 Climate feedbacks
8.3.1 Water vapor feedbacks
Water vapor is the principle greenhouse gas, which is quite clearly shown in
Tables 8.1 and 8.2. Therefore, any changes in water vapor concentration in
response to other greenhouse gases would substantially alter the net greenhouse
heating. As the atmosphere and the ocean warm in response to enhanced green-
house warming, more water vapor evaporates from the ocean and land surfaces.
The higher water vapor content of the atmosphere causes further greenhouse
warming, which causes more evaporation. There are other possible feedbacks
associated with higher moisture contents, however, which complicate this feed-
back process. Clouds can form, and as we will see in the following discussion, this
can lead to both warming and cooling of the atmosphere through their interactions
Climate feedbacks 175
Table 8.1 Downwelling longwave flux at the surface for carbon dioxide
scenarios
Flux, W m
−2
CO
2
concentration, ppmv Tropical Subarctic summer Subarctic winter
0 407.84 306.36 161.16
280 408.19 309.05 174.74
360 408.25 309.30 175.59
560 408.34 309.77 176.68
Source: Prepared by Norman Wood, Colorado State University.
Table 8.2 Downwelling longwave flux at the surface for water vapor scenarios
Flux, W m
−2
H
2
O mixing ratio scale factor Tropical Subarctic summer Subarctic winter
0 104.35 89.47 58.36
1.0 408.19 309.05 174.82
1.05 412.07 311.78 175.52
1.10 415.71 314.48 176.19
Source: Prepared by Norman Wood, Colorado State University.
with radiation. Simulation of the moisture feedback in GCMs requires realis-
tic models of the hydrological budget (rainfall and physical evaporation and
transpiration of water vapor from land and ocean surfaces) and of the response
of the ocean to heating of the atmosphere. Unfortunately, the responses of the
various GCMs to simulated surface energy budgets are diverse (e.g., see Wild,
2005).
The fluxes of moisture at the ocean surface, for example, are a function of the
ocean surface temperature as well as the surface wind stresses (Pielke, 1981).
Thus, ocean surface temperatures in tropical regions can warm appreciably in
response to tropospheric warming, but in regions of weak winds, such as over
the intertropical convergence zone, only weak vertical moisture fluxes will be
experienced. Moreover, crucial to water vapor feedback is the vertical distribution
of moisture in the middle and upper troposphere. This, in turn, is related not only
to local convective transports, but also to the strength of regional circulations such
as the Hadley and Walker circulations (Philander, 1990) both of which transport
moisture over long distances horizontally and vertically. Thus, while the moisture
feedback can amplify globally averaged surface air warming by as much as a
factor of three (Ramanathan, 1981), consideration of realistic cloud feedbacks
176 Overview of global climate forcings and feedbacks
and realistic ocean–atmospheric circulation responses requires more skill in the
simulation of climate responses than is possible in current models.
8.3.2 Cloud feedbacks
We have seen that one of the major positive feedbacks to carbon dioxide or other
greenhouse gas warming is that as the ocean surface and ground temperatures rise,
an increase in flux of water vapor into the atmosphere is expected. In a cloud-
free atmosphere, enhanced water vapor content provides a strong enhancement
of greenhouse warming. In response to the enhanced water vapor content of the
atmosphere, however, one would also expect that cloud coverage would increase
and some clouds would become optically thicker. Moreover, clouds alter the
stability of the atmosphere in response to their release of latent heat, and their
associated rising and sinking motions transfer heat, moisture, momentum, and
various particles and trace gases vertically and horizontally in the atmosphere.
In this section, we will examine the radiative feedback of clouds as well as the
feedback associated with changes of atmospheric stability and transport.
The global average radiative effects of high clouds warm the atmosphere while
low clouds cool the atmosphere and middle-level clouds are in near balance
between cooling by reflection of solar radiation and warming by absorption of
longwave radiation. Averaged over the entire Earth, the enhanced albedo of clouds
is slightly greater than their greenhouse warming, so clouds radiatively cool the
atmosphere compared to a cloud-free Earth (Ramanathan et al., 1989a,b; Randall
et al., 1989). In the case of an atmosphere–ocean system warmed by added
greenhouse gases, what is the feedback of clouds to that response? Unfortunately,
clouds are forced by relatively small-scale atmospheric motions, thus climate
models must parameterize clouds in rather crude ways, with the cloud cover being
specified based on climatology or parameterized as a function of relative humidity
and as a by-product of convective parameterization schemes. Even in the more
sophisticated GCMs which have a prognostic equation for cloud liquid water or
ice water amounts, the horizontal and vertical resolution of those models is not
sufficient to simulate the major ascending motions leading to cloud formation.
As a result, the simulated feedbacks of clouds on the greenhouse theory vary
from model to model, depending on the nature of the cloud/radiative scheme and
resolutions of the models.
Randall et al. (1989) suggest that cumulus anvil clouds exert a very powerful
influence on tropical convection by radiatively destabilizing the upper tropo-
sphere, and by trapping terrestrial radiation. They argue that the skill in predicting
the effects of those clouds in GCMs, while being qualitatively correct, are not
Climate feedbacks 177
sufficiently accurate to be relied upon quantitatively in considering their strong
influence on simulated climates. Their conclusion is particularly valid when one
considers the important role that organized mesoscale systems play in produc-
ing long-lasting upper tropospheric anvil clouds in both the tropics and middle
latitudes. These systems produce large areas of optically thick, stratiform–anvil
clouds in response to detrainment of rising moist air from deep convective tow-
ers as well as cloud formation in slowly ascending mesoscale motion. Current
cloud parameterization schemes in GCMs do not generally consider the organized
mesoscale motions in their estimates of anvil cloud cover or optical thicknesses.
The proposal to embed cloud-resolving models within GCMs (Randall et al.,
2003) is computationally very expensive, and moreover, up to the present, this
method used two-dimensional (not more realistic three-dimensional) cloud field
models.
3
Some cloud systems such as mesoscale convective complexes and tropical
cloud clusters exhibit a well-defined nocturnal maximum in their frequency of
occurrence (see Cotton and Anthes, 1989). Thus, if a warmed ocean creates more
tropical cloud clusters, then their net impact would be to cause a positive feedback
since they would have little effect on solar radiation. It is important, therefore,
to consider the diurnal variations of cloudiness when examining the impacts of
clouds on climate.
Lindzen (1990) argues that deep convective clouds should produce a net drying
of the upper troposphere due to the drying influence of compensating subsi-
dence. He bases his argument on the behavior of simple cumulus parameterization
schemes that he has developed. While drying effects of compensating subsidence
do indeed take place in the middle troposphere, it appears that the moistening
effects of detrainment of water substance from cumulus towers and stratiform–
anvil clouds of mesoscale convective systems override the drying effects. As a
result, a general moistening in the upper troposphere associated with deep con-
vection is observed (Rind et al., 1991). This issue of an atmospheric “iris” effect
is still being debated in the literature.
Another interesting feedback from clouds has been hypothesized by Mitchell
et al. (1989). They found a strong negative feedback of clouds related to a
reduction in the depletion of cloud water content in a carbon dioxide warmed
atmosphere due to a reduction in ice phase precipitation formation in middle
latitudes. As a result, cloud cover increased and the stronger albedo caused a
reduced rate of greenhouse-induced warming than in the version of the model
using relative humidity as the primary parameter determining cloud amounts.
3
An alternate approach where parameterizations are replaced by much more computationally efficient look-up
tables has been proposed in Pielke et al. (2006b).
178 Overview of global climate forcings and feedbacks
Several researchers have attempted to evaluate cloud feedbacks to a warming
atmospheric–ocean system by examining the observed relationship between sea
surface temperature (SST) anomalies and changes in cloudiness and net radiative
budgets. Ramanathan and Collins (1991), for example, diagnosed changes in
cloudiness and net radiation associated with warm SST anomalies during the
1987 El Niño. They inferred that in warm ocean regions where SSTs were less
than 300 K, the net effect of enhanced water vapor and cloudiness resulted in a
positive greenhouse effect. However, when SSTs exceeded ∼300 K, they found
that cirrus clouds formed by the flux of water substance into the upper troposphere
by cumulonimbus clouds, became optically thick and, as a result, they reflect
more solar radiation than is absorbed and reradiated downward by terrestrial
radiation. They argue that the highly reflective, optically thick cirrus clouds acts
as a thermostat, which prevents further warming of the oceans. They suggest that
the implications to greenhouse warming is that “it would take more than an order-
of-magnitude increase in atmospheric carbon dioxide to increase the maximum
SSTs by a few degrees in spite of a significant warming outside the equatorial
regions.”
In another study, Peterson (1991) examined the relationship between SST
anomalies and anomalies of high, middle, and low-level cloudiness using satellite
data. He found that over much of the tropics and the south Pacific convergence
zone, high clouds increased with warm SST anomalies, while in subtropical
stratocumulus regions, low cloud coverage decreased with positive SST anomalies.
Averaged over the entire region he sampled, total cloudiness increased with
positive SST anomalies. Because the coverage of optically thick low clouds
decreased while optically thin high clouds increased over regions of warm SSTs,
he calculated that the average, net radiative flux to space decreased in response
to warm SST anomalies. He concluded that his results provide observational
evidence that clouds provide a positive feedback loop to global warming scenarios.
There are other effects of clouds that can create positive or negative feedbacks
to greenhouse-gas-induced warming. Some of these are associated with changes
in the stability of the troposphere including the height of the tropopause. In
response to a warming climate, for example, deep convective overturning will
raise the height of the tropopause in the tropics, yielding a colder tropopause.
The temperature of the tropical tropopause has a strong influence on the water
vapor content of the stratosphere. The colder the tropopause, the lower the water
vapor content of the stratosphere, since the tropopause acts as an effective trap
of moisture as more water will be condensed and precipitated out of the air. As
a result, with a colder tropopause, there will be little absorption of terrestrial
radiation in the stratosphere causing a cooling effect on surface temperatures.
Overall, simulation of the feedback of clouds on climate change can be considered