188.8.131.52 Direct RF for Combined Total Aerosol
The TAR reported RF values associated with several aerosol components but did not provide an estimate of the overall aerosol RF. Improved and intensified in situ observations and remote sensing of aerosols suggest that the range of combined aerosol RF is now better constrained. For model results, extensive validation now exists for combined aerosol properties, representing the whole vertical column of the atmosphere, such as τaer. Using a combined estimate implicitly provides an alternative procedure to estimating the RF uncertainty. This approach may be more robust than propagating uncertainties from all individual aerosol components. Furthermore, a combined RF estimate accounts for nonlinear processes due to aerosol dynamics and interactions between radiation field and aerosols. The role of nonlinear processes of aerosol dynamics in RF has been recently studied in global aerosol models that account for the internally mixed nature of aerosol particles (Jacobson, 2001a; Kirkevåg and Iversen, 2002; Liao and Seinfeld, 2005; Takemura et al., 2005; Stier et al., 2006b). Mixing of aerosol particle populations influences the radiative properties of the combined aerosol, because mixing changes size, chemical composition, state and shape, and this feed backs to the aerosol removal and formation processes itself. Chung and Seinfeld (2002), in reviewing studies where BC is mixed either externally or internally with various other components, showed that BC exerts a stronger positive direct RF when mixed internally. Although the source-related processes for anthropogenic aerosols favour their submicron nature, natural aerosols enter the picture by providing a condensation surface for aerosol precursor gases. Heterogeneous reactions on sea salt and dust can reduce the sub-micron sulphate load by 28% (H. Liao et al., 2004) thereby reducing the direct and indirect RFs. Bauer and Koch (2005) estimated the sulphate RF to weaken from –0.25 to –0.18 W m–2 when dust is allowed to interact with the sulphur cycle. It would be useful to identify the RF contribution attributable to different source categories (Section 2.9.3 investigates this). However, few models have separated out the RF from specific emission source categories. Estimating the combined aerosol RF is a first step to quantify the anthropogenic perturbation to the aerosol and climate system caused by individual source categories.
A central model-derived estimate for the aerosol direct RF is based here on a compilation of recent simulation results using multi-component global aerosol models (see Table 2.6). This is a robust method for several reasons. The complexity of multi-component aerosol simulations captures nonlinear effects. Combining model results removes part of the errors in individual model formulations. As shown by Textor et al. (2006), the model-specific treatment of transport and removal processes is partly responsible for the correlated dispersion of the different aerosol components. A less dispersive model with smaller burdens necessarily has fewer scattering and absorbing aerosols interacting with the radiation field. An error in accounting for cloud cover would affect the all-sky RF from all aerosol components. Such errors result in correlated RF efficiencies for major aerosol components within a given model. Directly combining aerosol RF results gives a more realistic aerosol RF uncertainty estimate. The AeroCom compilation suggests significant differences in the modelled local and regional composition of the aerosol (see also Figure 2.12), but an overall reproduction of the total τaer variability can be performed (Kinne et al., 2006). The scatter in model performance suggests that currently no preference or weighting of individual model results can be used (Kinne et al., 2006). The aerosol RF taken together from several models is more robust than an analysis per component or by just one model. The mean estimate from Table 2.6 of the total aerosol direct RF is –0.2 W m–2, with a standard deviation of ±0.2 W m–2. This is a low-end estimate for both the aerosol RF and uncertainty because nitrate (estimated as –0.1 W m–2, see Section 184.108.40.206) and anthropogenic mineral dust (estimated as –0.1 W m–2, see Section 220.127.116.11) are missing in most of the model simulations. Adding their contribution yields an overall model-derived aerosol direct RF of –0.4 W m–2 (90% confidence interval: 0 to –0.8 W m–2).
Three satellite-based measurement estimates of the aerosol direct RF have become available, which all suggest a more negative aerosol RF than the model studies (see Section 18.104.22.168.3). Bellouin et al. (2005) computed a TOA aerosol RF of –0.8 ± 0.1 W m–2. Chung et al. (2005), based upon similarly extensive calculations, estimated the value to be –0.35 ± 0.25 W m–2, and Yu et al. (2006) estimated it to be –0.5 ± 0.33 W m–2. A central measurement-based estimate would suggest an aerosol direct RF of –0.55 W m–2. Figure 2.13 shows the observationally based aerosol direct RF estimates together with the model estimates published since the TAR.
Figure 2.13. Estimates of the direct aerosol RF from observationally based studies, independent modelling studies, and AeroCom results with identical aerosol and aerosol precursor emissions. GISS_1 refers to a study employing an internal mixture of aerosol, and GISS_2 to a study employing an external mixture. See Table 2.4, Note (a) for descriptions of models.
The discrepancy between measurements and models is also apparent in oceanic clear-sky conditions where the measurement-based estimate of the combined aerosol DRE including natural aerosols is considered unbiased. In these areas, models underestimate the negative aerosol DRE by 20 to 40% (Yu et al., 2006). The anthropogenic fraction of τaer is similar between model and measurement based studies. Kaufman et al. (2005a) used satellite-observed fine-mode τaer to estimate the anthropogenic τaer. Correcting for fine-mode τaer contributions from dust and sea salt, they found 21% of the total τaer to be anthropogenic, while Table 2.6 suggests that 29% of τaer is anthropogenic. Finally, cloud contamination of satellite products, aerosol absorption above clouds, not accounted for in some of the measurement-based estimates, and the complex assumptions about aerosol properties in both methods can contribute to the present discrepancy and increase uncertainty in aerosol RF.
A large source of uncertainty in the aerosol RF estimates is associated with aerosol absorption. Sato et al. (2003) determined the absorption τaer from AERONET measurements and suggested that aerosol absorption simulated by global aerosol models is underestimated by a factor of two to four. Schuster et al. (2005) estimated the BC loading over continental-scale regions. The results suggest that the model concentrations and absorption τaer of BC from models are lower than those derived from AERONET. Some of this difference in concentrations could be explained by the assumption that all aerosol absorption is due to BC (Schuster et al., 2005), while a significant fraction may be due to absorption by organic aerosol and mineral dust (see Sections 22.214.171.124, and 126.96.36.199). Furthermore, Reddy et al. (2005a) show that comparison of the aerosol absorption τaer from models against those from AERONET must be performed very carefully, reducing the discrepancy between their model and AERONET derived aerosol absorption τaer from a factor of 4 to a factor of 1.2 by careful co-sampling of AERONET and model data. As mentioned above, uncertainty in the vertical position of absorbing aerosol relative to clouds can lead to large uncertainty in the TOA aerosol RF.
The partly absorbing nature of the aerosol is responsible for a heating of the lower-tropospheric column and also results in the surface forcing being considerably more negative than TOA RF, results that have been confirmed through several experimental and observational studies as discussed in earlier sections. Table 2.6 summarises the surface forcing obtained in the different models. Figure 2.12 depicts the regional distribution of several important parameters for assessing the regional impact of aerosol RF. The results are based on a mean model constructed from AeroCom simulation results B and PRE. Anthropogenic τaer (Figure 2.12a) is shown to have local maxima in industrialised regions and in areas dominated by biomass burning. The difference between simulated and observed τaer shows that regionally τaer can be up to 0.1 (Figure 2.12b). Figure 2.12c suggests that there are regions off Southern Africa where the biomass burning aerosol above clouds leads to a local positive RF. Figure 2.12d shows the local variability as the standard deviation from nine models of the overall RF. The largest uncertainties of ±3 W m–2 are found in East Asia and in the African biomass burning regions. Figure 2.12e reveals that an average of 0.9 W m–2 heating can be expected in the atmospheric column as a consequence of absorption by anthropogenic aerosols. Regionally, this can reach annually averaged values exceeding 5 W m–2. These regional effects and the negative surface forcing in the shortwave (Figure 2.12f) is expected to exert an important effect on climate through alteration of the hydrological cycle.
An uncertainty estimate for the model-derived aerosol direct RF can be based upon two alternative error analyses:
1) An error propagation analysis using the errors given in the sections on sulphate, fossil fuel BC and organic carbon, biomass burning aerosol, nitrate and anthropogenic mineral dust. Assuming linear additivity of the errors, this results in an overall 90% confidence level uncertainty of 0.4 W m–2.
2) The standard deviation of the aerosol direct RF results in Table 2.6, multiplied by 1.645, suggests a 90% confidence level uncertainty of 0.3 W m–2, or 0.4 W m–2 when mineral dust and nitrate aerosol are accounted for.
Therefore, the results summarised in Table 2.6 and Figure 2.13, together with the estimates of nitrate and mineral dust RF combined with the measurement-based estimates, provide an estimate for the combined aerosol direct RF of –0.50 ± 0.40 W m–2. The progress in both global modelling and measurements of the direct RF of aerosol leads to a medium-low level of scientific understanding (see Section 2.9, Table 2.11).
Table 2.6. Quantities related to estimates of the aerosol direct RF. Recent estimates of anthropogenic aerosol load (LOAD), anthropogenic aerosol optical depth (τaer), its fraction of the present-day total aerosol optical depth (τaer ant), cloud cover in aerosol model, total aerosol direct radiative forcing (RF) for clear sky and all sky conditions, surface forcing and atmospheric all-sky forcing.
|No ||Modela ||LOAD ||τaer (0.55 µm) ||τaer ant (0.55 µm) ||Cloud Cover ||RF top clear sky ||RF top all sky ||Surface Forcing all sky ||Atmospheric Forcing all sky ||Reference |
| || ||(mg m–2) || ||(%) ||(%) ||(W m–2) ||(W m–2) ||(W m–2) ||(W m–2) || |
|Published since IPCC, 2001 || || || || || || |
|A ||GISS ||5.0 || || ||79% || ||–0.39b +0.01c ||–1.98b –2.42c ||1.59b 2.43c ||(Liao and Seinfeld, 2005) |
|B ||LOA ||6.0 ||0.049 ||34% ||70% ||–0.53 ||–0.09 || || ||(Reddy and Boucher, 2004) |
|C ||SPRINTARS ||4.8 ||0.044 ||50% ||63% ||–0.77 ||–0.06 ||–1.92 ||1.86 ||(Takemura et al., 2005) |
|D ||UIO-GCM ||2.7 || || ||57% || ||–0.11 || || ||(Kirkevag and Iversen, 2002) |
|E ||GATORG ||6.4d || || ||62% ||–0.89 ||–0.12 ||–2.5 ||2.38 ||(Jacobson, 2001a) |
|F ||GISS ||6.7 ||0.049 || || || ||–0.23 || || ||(Hansen et al., 2005) |
|G ||GISS ||5.6 ||0.040 || || || ||–0.63 || || ||(Koch, 2001) |
|AeroCom: identical emissions used for year 1750 and 2000 (Schulz et al., 2006) || || || || |
|H ||UMI ||4.0 ||0.028 ||25% ||63% ||–0.80 ||–0.41 ||–1.24 ||0.84 ||(Liu and Penner, 2002) |
|I ||UIO_CTM ||3.0 ||0.026 ||19% ||70% ||–0.85 ||–0.34 ||–0.95 ||0.61 ||(Myhre et al., 2003) |
|J ||LOA ||5.3 ||0.046 ||28% ||70% ||–0.80 ||–0.35 ||–1.49 ||1.14 ||(Reddy and Boucher, 2004) |
|K ||LSCE ||4.8 ||0.033 ||40% ||62% ||–0.94 ||–0.28 ||–0.93 ||0.66 ||(Schulz et al., 2006) |
|L ||ECHAM5 ||4.3 ||0.032 ||30% ||62% ||–0.64 ||–0.27 ||–0.98 ||0.71 ||(Stier et al., 2005) |
|M ||GISS ||2.8 ||0.014 ||11% ||57% ||–0.29 ||–0.11 ||–0.81 ||0.79 ||(Koch, 2001) |
|N ||UIO_GCM ||2.8 ||0.017 ||11% ||57% || ||–0.01 ||–0.84 ||0.84 ||(Kirkevag and Iversen, 2002) |
|O ||SPRINTARS ||3.2 ||0.036 ||44% ||62% ||–0.35 ||+0.04 ||–0.91 ||0.96 ||(Takemura et al., 2005) |
|P ||ULAQ ||3.7 ||0.030 ||23% || ||–0.79 ||–0.24 || || ||(Pitari et al., 2002) |
| Average A–G ||5.1 ||0.046 ||42% ||67% ||–0.73 ||–0.23 ||–2.21 ||2.07 || |
|Average H–P ||3.8 ||0.029 ||26% ||63% ||–0.68 ||–0.22 ||–1.02 ||0.82 || |
|Stddev A–G ||1.4 ||0.004 || || ||0.18 ||0.21 || || || |
|Stddev H–P ||0.9 ||0.010 ||11% ||5% ||0.24 ||0.16 ||0.23 ||0.17 || |
|Average A–P ||4.3 ||0.035 ||29% ||64% ||–0.70 ||–0.22 ||–1.21 ||1.24 || |
|Stddev A–P ||1.3 ||0.012 ||13% ||7% ||0.26 ||0.18 ||0.44 ||0.65 || |
|Minimum A–P ||2.7 ||0.014 ||11% ||57% ||–0.94 ||–0.63 ||–1.98 ||0.61 || |
|Maximum A–P ||6.7 ||0.049 ||50% ||79% ||–0.29 ||0.04 ||–0.81 ||2.43 || |
External mixture. c
Internal mixture. d
The load excludes that of mineral dust, some of which was considered anthropogenic in Jacobson (2001a).