Climate and Climate-System Modelling - Environmental Modelling: Finding Simplicity in Complexity - page 153

present in each grid cell are prescribed in some AOGCMs,
while in others the biogeochemical module might be used
to determine which PFTs are present in each grid cell
based on the carbon balance as simulated by the bio-
geochemical module for each potentially present PFT
(see, for example, Foley et al ., 1996; Winter et al ., 2009;
and Notaro et al ., 2007). In addition to the above, the
entire seasonal phenological cycle is internally computed
in the vegetation model of Krinner et al . (2005). A fur-
ther advance involves computing the cycling of nitrogen
between soil carbon and plant biomass, as C-N inter-
actions are important for the response of vegetation to
increasing CO 2 (Plattner et al ., 2008; Gerber et al ., 2010).
Sensitivity studies have shown that the change in vegeta-
tion type as the climate changes can significantly modify
the final climate change (at least locally) compared to the
change that would occur if the vegetation was fixed. The
physiological plant response to higher CO 2 (changes in
stomata) can also be important local compared to the
radiative effect on climate (Sellers et al ., 1996; Bounoua
et al ., 1999) (see also Chapter 12).
Another major area of intensive research is in the
representation of clouds in AGCMs. Differences in the
way in which clouds are modelled are the main rea-
son for the wide uncertainty in the long-run warming
resulting from a doubling atmospheric CO 2 concentra-
tion (with global mean responses ranging from 2.1 ◦ C
to 4.4 ◦ C in recent models, as summarized by Randall
et al ., 2007, Table 8.2). The climatic effects of clouds
depend on the details: exactly when, where, and how
high clouds are; their detailed microphysical proper-
ties, such as cloud droplet radius, ice crystal radius and
shape, and the extent of impurities (such as soot) in
cloud droplets. As the climate changes, the various cloud
properties will change in hard-to-predict ways, with a
range of strong but competing effects on climatic change
(some cloud changes will tend to amplify the change,
thereby serving as a positive feedback, whereas others will
serve to diminish the change, thereby serving as nega-
tive feedback). Increasingly sophisticated cloud schemes
are being developed for and implemented in AGCMs
(e.g. Del Genio et al ., 1996; Kao and Smith, 1999) that
explicitly compute more and more of the cloud processes
of climatic importance. However, greater sophistication
and detail do not guarantee that the predicted changes
in climate will be more reliable because even the most
detailed schemes still require parameterizations to rep-
resent the effects of subgrid scale processes, and the
nature of the parameterizations can strongly influence
the results.

Other areas of ongoing research include improved
treatment of sea ice, the realistic routing of runoff to the
oceans, and the incorporation of atmospheric chemistry
within AOGCMs used for climate-simulation purposes.
With regard to the latter, a typical atmospheric chemistry
model might track the concentration of over 100 chemical
species in three dimensions, including aerosol species in
a dozen or more size categories, and would include over
300 different chemical and phytolysis reactions (Jacobson
and Streets, 2009). Due to the large amount of computer
time required for atmospheric chemistry, simulations
with fully coupled three-dimensional atmospheric chem-
istry models embedded in AOGCMs have not yet been
performed. Rather, a recent approach has been to run the
atmospheric chemistry model for two years with winds,
temperature and other meteorological parameters pre-
scribed from a specific two-year period from an AOGCM
simulation, along with the corresponding anthropogenic
emissions, in order to compute the concentrations of
aerosols and short-lived GHGs and the associated radia-
tive forcing. This radiative forcing is then used, along with
the forcing from long-lived GHGs, to simulate another
two to three decades of transient climatic change with
the AOGCM, at which point the chemistry model is re-
run for another two years with the latest meteorological
variables and emissions (Shindell et al ., 2007). Climatic
change also affects changes in the natural emissions of the
precursors to various aerosols and ozone, although the
effects are rather small (Jacobson and Streets, 2009).

9.4 Online material

Rather than delving into the details of selected climate-
model components, the online material that accompanies
this chapter is designed to illustrate basic principles gov-
erning changes in climate in response to a radiative
forcing, and governing changes in the terrestrial bio-
sphere in response to changes in atmospheric CO 2 and
temperature. The change in temperature in response to
a given radiative forcing, once temperatures everywhere
have had time to fully adjust to the radiative forcing and
so are no longer changing, is referred to as the equilib-
rium temperature change. The variation in temperature
over time, as the climate system approaches the equilib-
rium change, is referred to as the transient temperature
change. Similarly, one can speak of equilibrium and tran-
sient changes in the terrestrial biosphere. Of course, if the
radiative forcing (for climate) or climate and CO 2 con-
centration (for the terrestrial biosphere) are themselves