Climate Change 2001:
Working Group I: The Scientific Basis
Other reports in this collection

6.8 The Indirect Radiative Forcing of Tropospheric Aerosols

6.8.1 Introduction

Aerosols serve as cloud condensation and ice nuclei, thereby modifying the microphysics, the radiative properties, and the lifetime of clouds. The physics and chemistry of the indirect effect of aerosols is discussed in detail in Chapter 5. Only aspects directly relevant to quantifying the indirect radiative forcing by aerosols are presented here. The aerosol indirect effect is usually split into two effects: the first indirect effect, whereby an increase in aerosols causes an increase in droplet concentration and a decrease in droplet size for fixed liquid water content (Twomey, 1974), and the second indirect effect, whereby the reduction in cloud droplet size affects the precipitation efficiency, tending to increase the liquid water content, the cloud lifetime (Albrecht, 1989), and the cloud thickness (Pincus and Baker, 1994). Until recently, the first indirect effect has received much more attention than the second. IPCC (1994) and the SAR only considered the first indirect effect. Shine et al. (1996) retained a range of radiative forcing from 0 to -1.5 Wm-2 with no best estimate, although a value of -0.8 Wm-2 was used for the year 1990 in the IS92a scenario (Kattenberg et al., 1996). Here we review and discuss the various estimates for the globally averaged aerosol indirect forcing available in the literature. Because of the inherent complexity of the aerosol indirect effect, GCM studies dealing with its quantification necessarily include an important level of simplification. While this represents a legitimate approach, it should be clear that the GCM estimates of the aerosol indirect effect are very uncertain. Section 6.8.2 investigates the indirect radiative forcing due to sulphate aerosols, on which most efforts have concentrated, while other aerosol types are treated in Section 6.8.3. Section 6.8.4 is devoted to alternative approaches, while Section 6.8.6 describes the aerosol indirect effects on ice clouds.

6.8.2 Indirect Radiative Forcing by Sulphate Aerosols

6.8.2.1 Estimates of the first indirect effect

The studies reported in Table 6.6 use different GCMs and methods for computing the droplet number concentration (i.e., empirical relationships between the sulphate mass and the cloud droplet number concentration, empirical relationships between the sulphate aerosol number concentration and the cloud droplet number concentration, or parametrization of cloud nucleation processes). The forcing estimates for the first indirect effect from sulphate aerosols range from -0.3 to -1.8 Wm-2, which is close to the range of 0 to -1.5 Wm-2 given in the SAR when only a few estimates were available.

Table 6.6: The global mean annual average aerosol indirect radiative forcing from different global studies. Letters P (prescribed) and C (computed) refer to off-line and on-line sulphate aerosol calculations, respectively. CCN and CDN stand for cloud condensation nuclei and cloud droplet number, respectively. In studies indicated by an asterisk, the estimate in flux change due to the indirect effect of aerosols was computed as the difference in top of atmosphere fluxes between two distinct simulations and therefore does not represent a forcing in the strict sense (see text). When several simulations are performed in the same study, "base" indicates the baseline calculation, while the range of estimates is given in parenthesis.
Reference
 
Aeroso
Type
Forcing estimate (Wm-2)
   
Remarks
             
   
First effect
Second effect
Both effects
         
Boucher and Rodhe (1994)*   Sulphate     -0.65 to -1.35 P   Uses 3 relationships between sulphate mass and CCN/CDN concentrations.
Chuang et al. (1994)   Sulphate -0.47     C   Includes a parametrization of cloud nucleation processes.
Jones et al. (1994)   Sulphate -1.3     P   Uses a relationship between aerosol and droplet number concentrations.
Boucher and Lohmann (1995)   Sulphate -0.5 to -1.4     P   LMD GCM Uses 4 different relationships between sulphate mass and CCN/CDN concentrations (A, B, C, and D).
Boucher and Lohmann (1995)   Sulphate -0.45 to -1.5     P   ECHAM
Jones and Slingo (1996)   Sulphate -0.3 to -1.5     P   Uses 2 different sulphate distributions. Follows Jones et al. (1994), Hegg (1994), Boucher and Lohmann (1995) ‘D'.
Uses a cloud climatology rather than GCM-simulated clouds.
Chuang et al. (1997)   Sulphate -0.4 to -1.6     C   Includes a parametrization of cloud nucleation processes.
Uses a mixture of pre-existing aerosols.
Feichter et al. (1997)   Sulphate -0.76     C   Uses Boucher and Lohmann (1995) ‘A' relationship.
Jones and Slingo (1997)   Sulphate -0.55 to -1.50     P   Uses 2 different versions of the Hadley Centre model.
Lohmann and Feichter (1997)*   Sulphate -1   -1.4 to -4.8 C   Uses Boucher and Lohmann (1995) ‘A' relationship.
Rotstayn (1999)*   Sulphate base -1.2 (-1.1 to -1.7) base -1.0 (-0.4 to -1.0) base -2.1 (-1.6 to -3.2) P   Includes a (small) long-wave radiative forcing.
Jones et al. (1999)*a   Sulphate -0.91 base -0.66 -1.18 C   Includes a (small) long-wave radiative forcing.
The two effects add non-linearly.
Kiehl et al. (2000)   Sulphate -0.40 to -1.78     C        
Ghan et al. (2001a)*   Sulphate ~50% for base ~50% for base base -1.7 (-1.6 to -3.2) C   Includes a parametrization of cloud nucleation.
Predicted aerosol size distribution.
Lohmann et al. (2000)* Sulphate     base -0.4 (0 to -0.4) C Includes a parametrization of cloud nucleation processes.
Carb.     base -0.9 (-0.9 to -1.3) C
Sulphate and Carb. -40% for base -60% for base base -1.1 (-1.1 to -1.9) C
Chuang et al. (2000b) Sulphate -0.30     C Includes a parametrization of cloud nucleation processes. Includes the effect of BC absorption in clouds.
Carb. base -1.51 (-1.27 to -1.67)     C
Sulphate and Carb. -1.85     C
aThis model predicts too low sulphate concentrations on average.

There is a tendency for more and more studies to use interactive (on-line) rather than prescribed (monthly or annual mean) sulphate concentrations. Feichter et al. (1997) pointed out that the first indirect effect calculated from monthly mean sulphate concentrations is 20% larger than calculated from interactive sulphate concentrations. Jones et al. (1999) found that the total indirect effect was overestimated by about 60% when they used seasonal or annual mean sulphate concentrations.

The various GCM studies show some disagreement on the spatial distribution of the forcing, an example of which is shown in Figure 6.7h. The Northern to Southern Hemisphere ratio varies from 1.4 to 4 depending on the models. It is generally smaller than the Northern to Southern Hemisphere ratio of anthropogenic sulphate aerosol concentrations because of the higher susceptibility of the clouds in the Southern Hemisphere (Platnick and Twomey, 1994; Taylor and McHaffie, 1994). The ocean to land ratio depends very much on the method used to relate the concentration of sulphate mass to the cloud droplet number concentration and on the natural background aerosol concentrations. It was generally found to be smaller than unity (Boucher and Lohmann, 1995; Jones and Slingo, 1997; Kiehl et al., 2000). Larger ratios, such as 1.6 (Chuang et al., 1997) and 5 (Jones and Slingo, 1997), are reported in some of the sensitivity experiments. Using a detailed inventory of ship sulphur emissions and a simple calculation of the aerosol indirect effect, Capaldo et al. (1999) suggested that a significant fraction of the effect over the oceans (-0.11 Wm-2, averaged globally) could be due to ship-emitted particulate matter (sulphate plus organic material). So far this source of aerosols has not been included in the GCM studies.

Kogan et al. (1996, 1997) used the Warren et al. (1988) cloud climatology over the oceans rather than a GCM to predict the indirect effect by sulphate aerosols on cloud albedo. The cloud albedo susceptibility was evaluated from a large eddy simulation model applied to stratocumulus clouds. They found an indirect short-wave forcing of -1.1 Wm-2 over the oceans with a small hemispheric difference of 0.4 Wm-2 (i.e., a Northern to Southern Hemisphere ratio of about 1.4). In their study, the forcing had a strong seasonal cycle, with the Southern Hemisphere forcing prevailing in some seasons.



Other reports in this collection