Climate and Climate-System Modelling - Environmental Modelling: Finding Simplicity in Complexity - page 152

the stratosphere and troposphere. These processes can be
adequately modelled only with three-dimensional atmo-
spheric models (in the case of the troposphere) or with
two-dimensional (latitude-height) models (in the case of
the stratosphere). Atmospheric chemistry is also critical to
the removal of CH 4 from the atmosphere and, to a lesser
extent, all other greenhouse gases except H 2 OandCO 2 .
InthecaseofCH 4 , a change in its concentration affects
its own removal rate and, hence, subsequent concentra-
tion changes. An accurate simulation of changes in the
removal rate of CH 4 requires specification of the concur-
rent concentrations of other reactive species, in particular
NO x (nitrogen oxides), CO (carbon monoxide) and the
VOCs (volatile organic compounds); and use of a model
with latitudinal and vertical resolution. However, simple
globally averaged models of chemistry-climate interac-
tions have been developed. These models treat the global
CH 4 -CO-OH cycle in a manner that takes into account
the effects of the heterogeneity of the chemical and trans-
port processes. They provide estimates of future global
or hemispheric mean changes in the chemistry of the
Earth's atmosphere. An even simpler approach, adopted
by Osborn and Wigley (1994), is to treat the atmosphere
as a single well-mixed box but to account for the effects
of atmospheric chemistry by making the CH 4 lifetime
depend on CH 4 concentration in a way that roughly
mimics the behaviour of globally averaged models or of
models with explicit spatial resolution.
Atmospheric chemistry is also central to the distribu-
tion and radiative properties of small suspended particles
in the atmosphere referred to as aerosols, although
chemistry is only part of what is required in order to
simulate the effects of aerosols on climate. The primary
aerosols that are affected by atmospheric chemistry are
sulphate (SO 4 3 − ) aerosols (produced from the emission
of SO2 and other S-containing gases), nitrate aerosols
(produced from emission of nitrogen oxides), and
organic carbon aerosols (produced from the emission
of a variety of organic compounds from plants and
gasoline). The key processes that need to be represented
are the source emissions of aerosols or aerosol precursors;
atmospheric transport, mixing, and chemical and phys-
ical transformation; and removal processes (primarily
deposition in rainwater and direct dry deposition onto
the Earth's surface). Since part of the effect of aerosols on
climate arises because they serve as cloud condensation
nuclei, it is also important to be able to represent the
relationship between changes in the aerosol mass input to
the atmosphere and, ultimately, the radiative properties
of clouds. Establishing the link between aerosol emissions

and cloud properties, however, involves several poorly
understood steps and is highly uncertain.
To simulate the increase in the amount of a given
aerosol in the atmosphere in response to the increase in
emissions of the precursors to that aerosol requires the
simultaneous simulation of all the major aerosols in the
atmosphere due to coupling between the different aerosols
(Stier et al ., 2006). For example, a reduction in sulphur
emissions, while reducing the sulphur aerosol loading,
leads to an increase in the lifetime and loading of other
aerosol species, especially at high latitudes and especially
for black carbon and particulate organic matter. The
geographical distribution of emissions is also important.
Between 1985 and 2000, global emissions fell by 12%
but the atmospheric sulphate aerosol loading is estimated
to have fallen only 3% because the locus of emissions
shifted southward to latitudes where in-cloud processing
of sulphur oxides (SO, SO 2 ) into sulphate ismore effective
(Manktelow et al ., 2007).

9.3 The research frontier

A long-term goal of the climate-research community is
the development of increasingly sophisticatedmodels that
couple more and more components of the climate sys-
tem. A large number of modelling groups have created
three-dimensional models that couple the atmospheric
and oceanic components of the climate system, and that
include increasingly realistic representations of sea ice
and land surface processes (in particular, the buildup
and melting of snow cover, runoff generation, and the
coupled fluxes of water, CO 2 and between vegetation and
the atmosphere). Vegetation-atmosphere fluxes of CO 2
and water are coupled through the partial dependence of
both on plant stomata. The size of the stomatal open-
ings depends on the photosynthetic demand for carbon,
as well as on the simulated leaf temperature and soil-
moisture content. The most recent land-surface models
that have been incorporated intoAOGCMs compute pho-
tosynthetic and respiration fluxes every 20 to 60 minutes
of simulated time. This is referred to as a biogeochemical
land-surface model. The annual net carbon gain or loss
is used to update the amounts of carbon in a three- or
five-box model of the terrestrial biosphere at each grid
point. From the carbon mass, a leaf-area index might be
computed that, in turn, would be used in the calculation
of the amount of solar radiation absorbed by the plant
canopy, which in turn affects the calculated annual photo-
synthesis (see, for example, Arora et al ., 2009). The PFTs