432 GREENHOUSE GASES EFFECTS
TABLE 1
Global mean changes in temperature and precipitation caused by doubling
CO 2 in various models in “Equilibrium”Model T(°C) P(%) RemarksUKMO (1987)* 5.2^15
GDFL (1989) 4.0 8
GISS (1984)* 4.8 13 Very low (8° 10°) resolution
SUNY (1991) 4.2 8
CSIRO (1991) 4.8 10
NCAR (1991) 4.5 5
Models with Computed Cloud
Water/Ice
UKMO (1989) 3.2 8 Fixed radiative properties
1.9 3 Variable radiative properties as function of water/
ice contentGDFL Geophysical Fluid Dynamics Laboratory, Princeton, USA
GISS Goddard Institute of Space Studies
SUNY State University of New York
SCRIO Commonwealth Scientific and Industrial Research Organization, Australia
NCAR National Center for Atmospheric Research, Boulder, USAand run in tandem with two-way feedbacks between ocean and
atmosphere transmitted at five-day intervals. Thus the atmo-
spheric model is run separately for five days with unchanged
sea-surface temperatures and sea-ice extents, accumulating
relevant time-averaged surface fluxes, which are then used to
drive the corresponding time step of the ocean model, follow-
ing which the updated sea-surface temperatures and sea-ice
cover are fed back to the atmosphere for the next iteration.
When an internally consistent balance is obtained between all
four main components of the climate system, the final state
may be taken as the starting point for perturbation experi-
ments such as the doubling of carbon dioxide.MODEL PREDICTIONS OF CLIMATE
CHANGES CAUSED BY DOUBLING PRESENT
CONCENTRATIONS OF CARBON DIOXIDEIntroductionWe recall that atmospheric concentrations of carbon dioxide
are likely to double by the second half of this century and
that simple radiative calculations, allowing only for feed-
back from the accompanying increases in water vapour, indi-
cate that this might cause the globally and annually averaged
surface air temperature to rise by about 1.5C. Because, as
discussed by Mason (1995), many other feedback processes,
both positive and negative, operate within the complex cli-
mate system, and because their effects are likely to vary with
season, latitude and geographical location, firmer estimates
can come only from model experiments in which the climate
simulated by a model perturbed by the doubling of CO 2 is
compared with that from an unperturbed (control) model,
the differences being attributed to the enhanced CO 2.We now compare and discuss the results of two types
of experiments, produced by different models. In one set,
involving a global atmosphere coupled to only a shallow
ocean, the CO 2 concentration is doubled in one step and the
climatic effects are assessed after the system has reached
a new equilibrium. In the second set, in which the atmo-
sphere is coupled to a multi-layered deep ocean, the CO 2
is allowed to increase at 1% p.a. compound and so doubles
after 70 years.Prediction of Global Mean Changes in the
‘Equilibrium’ ExperimentsAll six models cited in Table 1 comprise a global atmosphere
with 9–12 levels in the vertical, coupled to a shallow (50 m
deep) ocean with prescribed heat transport. The input solar
radiation to all models follows a seasonal cycle, but only
those marked with an asterisk include a diurnal cycle. All
the models have a rather low horizontal resolution and all the
experiments were run for 50 years. Furthermore, all of
them prescribe the cloud amount and height by empirical
formulae that relate cloud to relative humidity and are based
on satellite observations of cloud. The radiative properties
of the clouds (classified into low, medium and high-level
categories) are also prescribed and remain fixed during the
model simulation.
The predicted globally and annually-averaged increases
in surface air temperature due to doubling of CO 2 are
remarkably similar, ranging from 4.2C to 5.2C with an
average of 4.6C. This is probably because the sea-surface
temperatures and sea-ice cover are constrained to be near
observed values by adjusting the advective heat fluxes in
the shallow ocean. The predicted increase in precipitation,C007_002_r03.indd 432C007_002_r03.indd 432 11/18/2005 10:28:00 AM11/18/2005 10:28:00 AM