temperature change at the surface and within the whole
atmosphere. Its value is approximately − 3.2 W/m
2 K. There
is very little change in this value from one model to another,
and the sign convention used corresponds to a decrease in
the Earth’s energy balance when the surface temperature
increases. In response to a radiative forcing, DQ, the surface
temperature response can be calculated if the Planck
parameter k P is the only non-zero feedback parameter:
T P ¼ ÀDQ=k P :
This so-called Planck response causes an increase of 1.2 °
C for a forcing of 3.7 W/m
2 , resulting from a doubling of the
CO 2 concentration. It is the response of an idealized system
in which only the atmospheric and surface temperatures can
change uniformly and where only the radiation emission law
is affected (see Sect. “Simple Radiative Climate Models and
Their Limitations” of this chapter). It can be said that this is
the response with no feedback from the climate system. By
combining the above equations, the increase in temperature
at equilibrium DT can be written as a function of the Planck
response:
DT ¼ DT P = 1 À g
ð
Þ
where g is the feedback gain of the system:
g ¼ À k L þ k c þ k w þ k a
ð
Þ =k P :
If the gain g is positive and less than 1, the feedbacks will
amplify the temperature increase DT, relative to DT P . Conversely, if the gain is negative, the DT increase will be
attenuated. In the preparation of the Fifth IPCC Assessment
Report (IPCC 2013), climate change simulations were conducted within the CMIP5 project by around forty climate
models. In particular, for a doubling of the CO 2 concentration, the models simulate a global warming of 3 °C on
average (between 2.0 and 4.6 °C depending on the model),
until a new energy equilibrium is found. We have seen that
in the absence of feedbacks, this warming would be 1.2 °C.
This means that climate feedbacks amplify this warming by
a factor of 2 to 4 depending on the model.
Other developments make it possible to use the feedback
parameters to estimate the temperature increase due to the
Planck response and to the various feedbacks (Dufresne and
Bony 2008). They were applied to twelve CMIP3 (Coupled
Models Intercomparison Project) models, and the results are
shown in Fig. 31.10 illustrating both the average contribution
of the models and the inter-model dispersion. These results
were confirmed for the CMIP5 models (Vial et al. 2013).
Several of the mechanisms governing the feedback
parameter values, and hence the gain, are now well identified
(Bony 2006). For example, an increase in the temperature of
the atmosphere increases the saturation vapor pressure of the
water vapor. If there is little variation in the relative
humidity, this results in an increase in the concentration of
water vapor in the atmosphere, and therefore in the greenhouse effect, thus constituting a very powerful amplification
mechanism of the warming: 1.7 °C on average for the
models considered here.
For thermodynamic reasons (variation of the adiabatic
temperature gradient as a function of humidity), it is also
expected that, in the case of a humid tropical atmosphere,
variations in water vapor will be accompanied by greater
warming at high altitudes than close to the surface (except at
high latitudes). This increases the emission of infrared
radiation from the upper atmosphere and is the only negative
feedback: it decreases warming (− 0.8 °C on average). As
these two feedbacks are very strongly correlated for physical
reasons, their combined effect is generally considered to
contribute about 1 °C to the increase in mean temperature
(Fig. 31.10a).
The mechanisms behind the surface albedo feedback are
also well identified: an increase in temperature increases the
melting of snow and ice. This decreases the area covered by
surfaces that reflect solar radiation, and therefore increases
the amount of energy absorbed by the Earth. This effect
contributes about 0.2 °C to the temperature increase
(Fig. 31.10a).
Finally, the increase in temperature is likely to impact
cloud cover. As seen before, clouds exert two opposing
effects on the terrestrial radiative balance: on the one hand,
by reflecting solar radiation, they contribute to a cooling of
the Earth, and on the other, by absorbing infrared radiation,
they contribute to increase the greenhouse effect. The relative importance of these two effects depends on multiple
factors. The average contribution of clouds estimated by the
models considered here is equivalent to a warming of 0.7 °C
(Fig. 31.10a). However, the dispersion between models is
high (Fig. 31.10b): while some models predict a relatively
neutral cloud response, most predict a decrease in cloud
cover as temperature increases and an increase in global
warming of up to 2 °C.
Recent studies indicate that this uncertainty stems mainly
from the way different climate models predict the response
of low clouds (stratus, stratocumulus or small cumulus
clouds) to global warming. The way other clouds (including
large cumulonimbus storm clouds) respond to climate
change is also uncertain. This response contributes only
poorly to the uncertainty on the magnitude of global
warming, but contributes strongly to uncertainties in regional precipitation changes associated with global warming.
Other feedbacks exist in the climate system, for example,
those potentially causing changes in the atmospheric carbon
storage capabilities by the ocean and the biosphere. They
will be discussed below (Sect. “The CO 2 Cycle”).
31 From the Climates of the Past to the Climates …
465
atmosphere. Its value is approximately − 3.2 W/m
2 K. There
is very little change in this value from one model to another,
and the sign convention used corresponds to a decrease in
the Earth’s energy balance when the surface temperature
increases. In response to a radiative forcing, DQ, the surface
temperature response can be calculated if the Planck
parameter k P is the only non-zero feedback parameter:
T P ¼ ÀDQ=k P :
This so-called Planck response causes an increase of 1.2 °
C for a forcing of 3.7 W/m
2 , resulting from a doubling of the
CO 2 concentration. It is the response of an idealized system
in which only the atmospheric and surface temperatures can
change uniformly and where only the radiation emission law
is affected (see Sect. “Simple Radiative Climate Models and
Their Limitations” of this chapter). It can be said that this is
the response with no feedback from the climate system. By
combining the above equations, the increase in temperature
at equilibrium DT can be written as a function of the Planck
response:
DT ¼ DT P = 1 À g
ð
Þ
where g is the feedback gain of the system:
g ¼ À k L þ k c þ k w þ k a
ð
Þ =k P :
If the gain g is positive and less than 1, the feedbacks will
amplify the temperature increase DT, relative to DT P . Conversely, if the gain is negative, the DT increase will be
attenuated. In the preparation of the Fifth IPCC Assessment
Report (IPCC 2013), climate change simulations were conducted within the CMIP5 project by around forty climate
models. In particular, for a doubling of the CO 2 concentration, the models simulate a global warming of 3 °C on
average (between 2.0 and 4.6 °C depending on the model),
until a new energy equilibrium is found. We have seen that
in the absence of feedbacks, this warming would be 1.2 °C.
This means that climate feedbacks amplify this warming by
a factor of 2 to 4 depending on the model.
Other developments make it possible to use the feedback
parameters to estimate the temperature increase due to the
Planck response and to the various feedbacks (Dufresne and
Bony 2008). They were applied to twelve CMIP3 (Coupled
Models Intercomparison Project) models, and the results are
shown in Fig. 31.10 illustrating both the average contribution
of the models and the inter-model dispersion. These results
were confirmed for the CMIP5 models (Vial et al. 2013).
Several of the mechanisms governing the feedback
parameter values, and hence the gain, are now well identified
(Bony 2006). For example, an increase in the temperature of
the atmosphere increases the saturation vapor pressure of the
water vapor. If there is little variation in the relative
humidity, this results in an increase in the concentration of
water vapor in the atmosphere, and therefore in the greenhouse effect, thus constituting a very powerful amplification
mechanism of the warming: 1.7 °C on average for the
models considered here.
For thermodynamic reasons (variation of the adiabatic
temperature gradient as a function of humidity), it is also
expected that, in the case of a humid tropical atmosphere,
variations in water vapor will be accompanied by greater
warming at high altitudes than close to the surface (except at
high latitudes). This increases the emission of infrared
radiation from the upper atmosphere and is the only negative
feedback: it decreases warming (− 0.8 °C on average). As
these two feedbacks are very strongly correlated for physical
reasons, their combined effect is generally considered to
contribute about 1 °C to the increase in mean temperature
(Fig. 31.10a).
The mechanisms behind the surface albedo feedback are
also well identified: an increase in temperature increases the
melting of snow and ice. This decreases the area covered by
surfaces that reflect solar radiation, and therefore increases
the amount of energy absorbed by the Earth. This effect
contributes about 0.2 °C to the temperature increase
(Fig. 31.10a).
Finally, the increase in temperature is likely to impact
cloud cover. As seen before, clouds exert two opposing
effects on the terrestrial radiative balance: on the one hand,
by reflecting solar radiation, they contribute to a cooling of
the Earth, and on the other, by absorbing infrared radiation,
they contribute to increase the greenhouse effect. The relative importance of these two effects depends on multiple
factors. The average contribution of clouds estimated by the
models considered here is equivalent to a warming of 0.7 °C
(Fig. 31.10a). However, the dispersion between models is
high (Fig. 31.10b): while some models predict a relatively
neutral cloud response, most predict a decrease in cloud
cover as temperature increases and an increase in global
warming of up to 2 °C.
Recent studies indicate that this uncertainty stems mainly
from the way different climate models predict the response
of low clouds (stratus, stratocumulus or small cumulus
clouds) to global warming. The way other clouds (including
large cumulonimbus storm clouds) respond to climate
change is also uncertain. This response contributes only
poorly to the uncertainty on the magnitude of global
warming, but contributes strongly to uncertainties in regional precipitation changes associated with global warming.
Other feedbacks exist in the climate system, for example,
those potentially causing changes in the atmospheric carbon
storage capabilities by the ocean and the biosphere. They
will be discussed below (Sect. “The CO 2 Cycle”).
31 From the Climates of the Past to the Climates …
465