|Working Group I: The Scientific Basis|
|Other reports in this collection|
The radiatively important properties of atmospheric aerosols (both direct and indirect) are determined at the most fundamental level by the aerosol composition and size distribution. However, for purposes of the direct radiative forcing calculation and for assessment of uncertainties, these properties can be subsumed into a small set of parameters. Knowledge of a set of four quantities as a function of wavelength is necessary to translate aerosol burdens into first aerosol optical depths, and then a radiative perturbation: the mass light-scattering efficiency asp, the functional dependence of light-scattering on relative humidity f(RH), the single-scattering albedo wo, and the asymmetry parameter g (cf., Charlson et al., 1992; Penner et al., 1994a).
Light scattering by aerosols is measurable as well as calculable from measured
aerosol size and composition. This permits comparisons, called closure studies,
of the different measurements for consistency. An example is the comparison
of the derived optical depth with directly measured or inferred optical depths
from sunphotometers or satellite radiometers. Indeed, various sorts of closure
studies have been successfully conducted and lend added credibility to the measurements
of the individual quantities (e.g., McMurry et al., 1996; Clarke et al., 1996;
Hegg et al., 1997; Quinn and Coffman, 1998; Wenny et al., 1998; Raes et al.,
2000). Closure studies can also provide objective estimates of the uncertainty
in calculating radiative quantities such as optical depth.
Aerosols in the accumulation mode, i.e., those with dry diameters between 0.1
and 1 mm (Schwartz, 1996) are of most importance. These aerosols can hydrate
to diameters between 0.1 and 2 mm where their mass extinction efficiency is
largest (see Figure 5.1). Accumulation mode aerosols not
only have high scattering efficiency, they also have the longest atmospheric
lifetime: smaller particles coagulate more quickly while nucleation to cloud
drops or impaction onto the surface removes larger particles efficiently. Accumulation
mode aerosols form the majority of cloud condensation nuclei (CCN). Hence, anthropogenic
aerosol perturbations such as sulphur emissions have the greatest climate impact
when, as is often the case, they produce or affect accumulation mode aerosols
(Jones et al., 1994).
The direct radiative effect of aerosols is also very sensitive to the single scattering albedo wo. For example, a change in wo from 0.9 to 0.8 can often change the sign of the direct effect, depending on the albedo of the underlying surface and the altitude of the aerosols (Hansen et al., 1997). Unfortunately, it is difficult to measure wo accurately. The mass of black carbon on a filter can be converted to light absorption, but the conversion depends on the size and mixing state of the black carbon with the rest of the aerosols. The mass measurements are themselves difficult, as discussed in Section 126.96.36.199. Aerosol light absorption can also be measured as the difference in light extinction and scattering. Very careful calibrations are required because the absorption is often a difference between two large numbers. As discussed in Section 5.2.4, it is difficult to retrieve wo from satellite data. Well-calibrated sunphotometers can derive wo by comparing light scattering measured away from the Sun with direct Sun extinction measurements (Dubovik et al., 1998).
Some encouraging comparisons have been made between different techniques for measurements of wo and related quantities. Direct measurements of light absorption near Denver, Colorado using photo-acoustic spectroscopy were highly correlated with a filter technique (Moosmüller et al., 1998). However, these results also pointed to a possible strong wavelength dependence in the light absorption. An airborne comparison of six techniques (extinction cell, three filter techniques, irradiance measurements, and black carbon mass by thermal evolution) in biomass burning plumes and hazes was reported by Reid et al. (1998a). Regional averages of wo derived from all techniques except thermal evolution agreed within about 0.02 (wo is dimensionless), but individual data points were only moderately correlated (regression coefficient values of about 0.6).
Continues on next page
Other reports in this collection