Observational Requirements for Modeling of Global...
43
as expected from the baroclinic instability argument (but not consistently in the southern
midlatitudes, because snow/ice-albedo feedback and thus polar amplification are less in the
Antarctic than in the Arctic). In the tropical east Pacific, cloud cover increases as convection
shifts eastward, but it decreases over South America and Australia/Indonesia, which experience
anomalous subsidence. A high pressure anomaly forms over the west and central Pacific, and
a possible teleconnection pattern emanates poleward from there, with a low pressure anomaly
(and increased cloud) along the west coast of the United States, high pressure (and decreased cloud) in the North Pacific south of Alaska, and low pressure again over eastern Canada
(but without as obvious a cloud change). A similar pattern may propagate into the Southern
Hemisphere.
c) Regional climate forcing.
Although greenhouse gas concentration changes are fairly uniform globally (with the possible
exception of ozone), the competing negative climate forcing associated with aerosols is quite inhomogeneous spatially. Thus, a third approach to understanding regional climate change might
be tied to an assessment of the magnitude of regional aerosol impacts. Aerosol climate forcing
consists of three components (Charlson et aI., 1992): (1) A direct radiative effect associated
with reflection of sunlight by aerosols in otherwise clear air; (2) An indirect radiative effect due
to the fact that higher aerosol concentration implies more cloud condensation nuclei and thus
smaller and more reflective cloud droplets for a given cloud water content; (3) A related indirect
microphysical effect because smaller droplets precipitate less efficiently than larger ones, thus
implying higher cloud water content.
Even if there were no climate forcing due to increasing aerosol concentrations, an indirect aerosol effect on climate sensitivity might still exist because of the climatological difference between
continental and marine air. Continental air has several orders of magnitude greater concentration of cloud condensation nuclei (CCN) than does marine air, and consequently cloud droplets
are smaller over land and precipitate less easily for a given water content (Twomey, 1977; Han
et ai., 1994). As climate changes, the different microphysical settings of land and ocean clouds
might cause them to react differently to changes in the hydrologic cycle. In the GISS GCM,
which has higher effective precipitation thresholds over land than ocean, midlatitude cloud cover generally tends to decrease over ocean but increase over land for a doubling of CO2 (Fig.
2.7). As climate warms, there is less opportunity for cloud water content to increase over ocean
before precipitation sets in than is the case over land. This might contribute to longer cloud
lifetimes over land in the warmer climate, since clouds with larger cloud water content take
longer to dissipate by evaporation.
But in reality, tropospheric aerosol concentrations appear to be increasing at a fairly rapid
rate due to anthropogenic activity, including combustion of fossil fuels and biomass burning
(Charlson et ai., 1991; Penner et aI., 1992). There is significant uncertainty in the total aerosol
direct climate forcing, because ground-based data exist over only a small fraction of the Earth,
satellite monitoring of the relatively small aerosol optical depth is difficult, and the radiative
properties of different aerosols vary with particle size, shape and composition. Best estimates of
the direct forcing suggest that it is of order -1 to -2 W m -2 globally, with the signal much larger
in the Northern Hemisphere and especially near and downwind of industrial regions. If correct,
this is a substantial fraction of the current radiative imbalance due to increased greenhouse gas
concentrations, and it may explain why climate has warmed to a lesser extent thus far than
many GCM predictions (Charlson et aI., 1992). But unlike greenhouse forcing, which is fairly
uniform globally, aerosol forcing is quite inhomogeneous, so net climate forcing to date may be
near zero or negative in some regions while being strongly positive in others.
The indirect radiative effect is even more uncertain. One complication is that cloud droplet
number and size do not simply respond linearly to an increase in CCN concentration. Continen-
43
as expected from the baroclinic instability argument (but not consistently in the southern
midlatitudes, because snow/ice-albedo feedback and thus polar amplification are less in the
Antarctic than in the Arctic). In the tropical east Pacific, cloud cover increases as convection
shifts eastward, but it decreases over South America and Australia/Indonesia, which experience
anomalous subsidence. A high pressure anomaly forms over the west and central Pacific, and
a possible teleconnection pattern emanates poleward from there, with a low pressure anomaly
(and increased cloud) along the west coast of the United States, high pressure (and decreased cloud) in the North Pacific south of Alaska, and low pressure again over eastern Canada
(but without as obvious a cloud change). A similar pattern may propagate into the Southern
Hemisphere.
c) Regional climate forcing.
Although greenhouse gas concentration changes are fairly uniform globally (with the possible
exception of ozone), the competing negative climate forcing associated with aerosols is quite inhomogeneous spatially. Thus, a third approach to understanding regional climate change might
be tied to an assessment of the magnitude of regional aerosol impacts. Aerosol climate forcing
consists of three components (Charlson et aI., 1992): (1) A direct radiative effect associated
with reflection of sunlight by aerosols in otherwise clear air; (2) An indirect radiative effect due
to the fact that higher aerosol concentration implies more cloud condensation nuclei and thus
smaller and more reflective cloud droplets for a given cloud water content; (3) A related indirect
microphysical effect because smaller droplets precipitate less efficiently than larger ones, thus
implying higher cloud water content.
Even if there were no climate forcing due to increasing aerosol concentrations, an indirect aerosol effect on climate sensitivity might still exist because of the climatological difference between
continental and marine air. Continental air has several orders of magnitude greater concentration of cloud condensation nuclei (CCN) than does marine air, and consequently cloud droplets
are smaller over land and precipitate less easily for a given water content (Twomey, 1977; Han
et ai., 1994). As climate changes, the different microphysical settings of land and ocean clouds
might cause them to react differently to changes in the hydrologic cycle. In the GISS GCM,
which has higher effective precipitation thresholds over land than ocean, midlatitude cloud cover generally tends to decrease over ocean but increase over land for a doubling of CO2 (Fig.
2.7). As climate warms, there is less opportunity for cloud water content to increase over ocean
before precipitation sets in than is the case over land. This might contribute to longer cloud
lifetimes over land in the warmer climate, since clouds with larger cloud water content take
longer to dissipate by evaporation.
But in reality, tropospheric aerosol concentrations appear to be increasing at a fairly rapid
rate due to anthropogenic activity, including combustion of fossil fuels and biomass burning
(Charlson et ai., 1991; Penner et aI., 1992). There is significant uncertainty in the total aerosol
direct climate forcing, because ground-based data exist over only a small fraction of the Earth,
satellite monitoring of the relatively small aerosol optical depth is difficult, and the radiative
properties of different aerosols vary with particle size, shape and composition. Best estimates of
the direct forcing suggest that it is of order -1 to -2 W m -2 globally, with the signal much larger
in the Northern Hemisphere and especially near and downwind of industrial regions. If correct,
this is a substantial fraction of the current radiative imbalance due to increased greenhouse gas
concentrations, and it may explain why climate has warmed to a lesser extent thus far than
many GCM predictions (Charlson et aI., 1992). But unlike greenhouse forcing, which is fairly
uniform globally, aerosol forcing is quite inhomogeneous, so net climate forcing to date may be
near zero or negative in some regions while being strongly positive in others.
The indirect radiative effect is even more uncertain. One complication is that cloud droplet
number and size do not simply respond linearly to an increase in CCN concentration. Continen-