Shine and Forster (1999) proposed a value of the aerosol indirect forcing due to all aerosols of -1 Wm^-2 with at least a factor of two uncertainty. Several observational studies (see Chapter 5) support the existence of the first aerosol indirect effect on low-level clouds and a negative sign for the associated radiative forcing, but these studies do not give indications on what a (negative) upper bound of the forcing would be. GCM studies predict radiative forcing for the first indirect effect of industrial aerosols in the range of –0.3 to –1.8 Wm^-2. However, because of the uncertainties in the estimates discussed above, the limited validation of GCM parametrizations and results, and because in-cloud absorption by black carbon aerosols was not considered in all but one of the GCM studies, we retain 0 Wm^-2 as an upper bound for the first aerosol indirect effect. A lower bound of –2 Wm^-2 is selected on the basis of available GCM studies for the first indirect effect. Not too much emphasis should be given to the exact bounds of this interval because they do not carry any statistical meaning (Section 6.13.1) and because of the very low level of scientific understanding associated to this forcing (see Section 6.13.1 where this concept is defined).

Available GCM studies suggest that the radiative flux perturbations associated with the second effect could be of similar magnitude to that of the first effect. There are no studies yet to confirm unambiguously that the GCM estimates of the radiative impact associated with the second indirect effect can be interpreted in the strict sense of a radiative forcing (see Sections 6.1 and 6.8.2.2), and very few observations exist as yet to support the existence of a significant effect. Therefore we refrain from giving any estimate or range of estimates for the second aerosol indirect effect. However, this does not minimise the potential importance of this effect.

6.8.6.1 Contrails and contrail-induced cloudiness

Using meteorological and air traffic data scaled to regional observations of contrail cover, Sausen et al. (1998) estimated the present day global mean cover by line-shaped contrails to be about 0.1%. This results in a global and annual mean radiative forcing by line-shaped contrails of 0.02 Wm^-2 (Minnis et al., 1999), subject to uncertainties in the contrail cover, optical depth, ice particle size and shape (Meerkötter et al., 1999). We follow Fahey et al. (1999) and retain a range of 0.005 to 0.06 Wm^-2 for the present day forcing, around the best estimate of 0.02 Wm^-2.

Contrails can evolve into extended cirrus clouds. Boucher (1999) and Fahey et al. (1999) have shown evidences that cirrus occurrence and coverage may have increased in regions of high air traffic compared with the rest of the globe. Smith et al. (1998) reported the existence of nearly invisible layers of small ice crystals, which cause absorption of infrared radiation, and could be due to remnant contrail particles. From consideration of the spatial distribution of cirrus trends during the last 25 years, Fahey et al. (1999) gave a range of possible best estimates of 0 to 0.04 Wm^-2 for the radiative forcing due to aviation-induced cirrus. The available information on cirrus clouds was deemed insufficient to determine a single best estimate or an uncertainty range.

6.8.6.2 Impact of anthropogenic aerosols on cirrus cloud microphysics

Measurements by Ström and Ohlsson (1998) in a region of high air traffic revealed higher crystal number concentrations in areas of the cloud affected by soot emissions from aircraft. If the observed enhancement in crystal number density (which is about a factor of 2) is associated with a reduction in the mean crystal size, as confirmed by the measurements of Kristensson et al. (2000), a change in cloud radiative forcing may result. Wyser and Ström (1998) estimated the forcing, although very uncertain, to be in the order of 0.3 Wm^-2 in regions of dense air traffic under the assumption of a 20% decrease of the mean crystal size. No globally averaged radiative forcing is available.

The sedimentation of ice particles from contrails may remove water vapour from the upper troposphere. This effect is expected to be more important in strongly supersaturated air when ice particles can fall without evaporating (Fahey et al., 1999). The impacts of such an effect on cirrus formation, vertical profile of humidity and the subsequent radiative forcing have not been assessed.

Aerosols also serve as ice nuclei although it is well recognised that there are fewer ice nuclei than cloud condensation nuclei. It is conceivable that anthropogenic aerosols emitted at the surface and transported to the upper troposphere affect the formation and properties of ice clouds. Jensen and Toon (1997) suggested that insoluble particles from the surface or soot particles emitted by aircraft, if they serve as effective ice nuclei, can result in an increase in the cirrus cloud coverage. Laaksonen et al. (1997) argued that nitric acid pollution is able to cause an increase in supercooled cirrus cloud droplet concentrations, and thereby influence climate (see Chapter 5, Section 5.3.6). Such effects, if significant at all, are not quantified at present.

IPCC Homepage