Levin et al. (1996) observed desert dust particles coated with sulphate. Such particles may originate from in-cloud scavenging of interstitial dust particles followed by evaporation of the cloud droplets, condensation of SO2 onto dust followed by oxidation, or even coagulation of dust and sulphate particles. The presence of soluble material (which may be of anthropogenic origin) on the desert dust particles converts them into large and effective CCN which may affect the cloud microphysics. Whether this effect results in a significant climate forcing has not been investigated and cannot presently be quantified.
Kulmala et al. (1993, 1995, 1998) argued that enhanced concentrations of condensable vapours (such as HNO3 and HCl) in the atmosphere could affect cloud properties by facilitating the activation of cloud condensation nuclei. The impact of such an effect on the planetary cloud albedo has not been assessed.
Hansen and colleagues have used two alternative approaches to characterise and quantify any "missing" climate forcing besides that due to greenhouse gases, solar constant, O3, and aerosol direct effect. Hansen et al. (1995) used a simplified GCM to investigate the impacts of various climate forcings on the diurnal cycle of surface air temperature and compared them with observations. They found that, although the aerosol direct effect or an increase in continental cloud albedo could contribute to damp the surface temperature diurnal cycle, only an increase in continental cloud cover would be consistent with observations (Karl et al., 1993). The required cloud increase depends on cloud height and would be of the order of 1% global coverage for low clouds (i.e., 2 to 5% over land). We cannot rule out that such a change is an unidentified cloud feedback rather than a forcing. Hansen et al. (1997b) also argued that agreement between observed and computed temperature trends requires the presence of another forcing of at least -1 Wm-2 which is inferred as being due to the indirect effect. In their calculations, the direct tropo-spheric aerosol effect does not play a large net role, because the moderately absorbing aerosol assumption leads to an offset between its sunlight reflecting and absorbing properties insofar as the top of the atmosphere irradiance change is concerned. However, this method assumes that the observed change in temperature since pre-industrial times is primarily a response to anthropogenic forcings, that all the other anthropogenic forcings are well quantified, and that the climate sensitivity parameter (Section 6.1) predicted by the GCM is correct (Rodhe et al., 2000). Therefore it may simply be a coincidence that the estimate of Hansen et al. (1997b) is consistent with the GCM estimates discussed above.
Han et al. (1994) analysed AVHRR satellite radiances to retrieve the cloud droplet size of low-level clouds. They reported significant inter-hemispheric differences for both maritime and continental clouds. Boucher (1995) showed that, if this difference is to be attributed to anthropogenic aerosols, it implies a differential forcing of about -1 Wm-2 between the two hemispheres. Assuming a Northern Hemisphere to Southern Hemisphere ratio of 2:1 for the aerosol indirect effect, this would imply a globally-averaged forcing of -1.5 Wm-2. It is not clear, however, to what extent changes in cloud droplet size are related to change in aerosol concentrations. For instance, Han et al. (1998) showed that cloud albedo decreases with decreasing droplet size for the optically thinner clouds over the oceans. While this does not invalidate the aerosol indirect effect at all, it underlines the limitations in using satellite observed changes in droplet size to compute the aerosol indirect forcing. Therefore it seems difficult at present to use satellite observations to estimate the first aerosol indirect forcing unless some changes in cloud albedo could be tied to changes in aerosol concentrations under the assumption of constant liquid water content. Satellite observations do play, however, a key role for evaluating models of the indirect aerosol radiative effect (Ghan et al., 2001b).
Other reports in this collection