|The Regional Impacts of Climate Change|
|Other reports in this collection|
B.2. Simulations with Greenhouse Gas and Aerosol Forcing
Two simulations (x,w) were forced with the historical increase in equivalent
CO2, then a 1%/yr increase in equivalent CO2. The patterns of change are qualitatively
similar to those in the experiments above. In Sections 188.8.131.52 and 184.108.40.206 of
the Working Group I contribution to IPCC (1996), we saw that the inclusion of
the direct sulfate aerosol forcing can improve the simulation of the patterns
of temperature change over the last few decades. Here we consider the effect
of combined greenhouse gas and direct sulfate aerosol forcing derived from IS92a
on simulated patterns of temperature change to 2050 and beyond (y,z). Figure
B-2 shows area averages of summer and winter surface temperature, precipitation,
and soil moisture from two models-one set with CO2 increase only, the other
with CO2 increase and the direct effects of sulfate aerosols. The areas considered
include five from Figure B-1 (i.e., Central
North America, South East Asia, Sahel, Southern Europe, and Australia).
In assessing these results, one should bear in mind the possible exaggeration of the sulfate aerosol concentrations under this scenario, the uncertainties in representing the radiative effects of sulfate aerosols, and neglect of other factors including the indirect effect of sulfates. Nevertheless, these experiments suggest that the direct effect of sulfate aerosols could have strong influence on future temperature changes, particularly in northern mid-latitudes.
IPCC (1990) reported some broad scale changes that were evident in most of the equilibrium 2xCO2 experiments that were then available. The detailed regional changes differed from model to model. In the transient experiments reported in IPCC (1992), it was found that the large-scale patterns of response at the time of doubling CO2 were similar to the corresponding equilibrium experiments (IPCC, 1990), except that there was a smaller warming in the vicinity of the northern North Atlantic and the Southern Ocean in transient experiments. Here we summarize the main features in the seasonal (December to February and June to August) patterns of change in temperature, precipitation, and soil moisture in those experiments with a 1%/yr increase in CO2 for which data were available. The changes are assessed at the time data of CO2 doubling (after 70 years). In experiments w-z, we also contrast the continental-scale response under the IS92a scenario with and without aerosol forcing at around 2040.
With increases in CO2, all models produce a maximum annual mean warming in high northern latitudes (Figures 6.6 and 6.7b of IPCC 1996, WG I). The warming is largest in late autumn and winter, largely due to sea ice forming later in the warmer climate. In summer, the warming is small; if the sea ice is removed with increased CO2, then the thermal inertia of the mixed-layer prevents substantial warming during the short summer season, otherwise melting sea ice is present in both control and anomaly simulations, and there is no change in surface temperature (see Ingram et al., 1989). The details of these changes are sensitive to parameterization of sea ice and, in particular, the specification of sea ice albedo (e.g., Meehl and Washington, 1995). In one simulation (k), there is a marked cooling over the northeastern Atlantic throughout the year, which leads to a cooling over part of northWest Europe in winter. There is little seasonal variation of the warming in low latitudes or over the southern circumpolar ocean.
When aerosol effects are included (y,z cf. x,w respectively), the maximum winter warming in high northern latitudes is less extensive. In mid-latitudes, there are some regions of cooling (e.g., over China), and the mean warming in the tropics is greater than in mid-latitudes. In northern summer, there are again regions of cooling in mid-latitudes and the greatest warming now occurs over Antarctica. Again, including the direct forcing by sulfate aerosols has a strong effect on simulated regional temperature changes, though the reader should bear in mind the limitations of these experiments as noted earlier.
Other reports in this collection