Figure 6.7: Examples of the geographical distribution of present-day annual-average radiative forcing (1750 to 2000) due to (a) well-mixed greenhouse gases including CO2 ,CH4 ,N2 O, CFC-11 and CFC-12 (Shine and Forster, 1999); (b) stratospheric ozone depletion over the period 1979 to 1994 given by WMO, 1995 (Shine and Forster, 1999); (c) increases in tropospheric O3 (Berntsen et al., 1997; Shine and Forster, 1999); (d) the direct effect of sulphate aerosol (Haywood et al., 1997a); (e) the direct effect of organic carbon and black carbon from biomass burning (Penner et al., 1998b; Grant et al., 1999); (f) the direct effect of organic carbon and black carbon from fossil fuel burning (Penner et al., 1998b; Grant et al., 1999), (g) the direct effect of anthropogenic emissions of mineral dust (Tegen et al., 1996); (h) the "first" indirect effect of sulphate aerosol calculated diagnostically in a similar way to Jones and Slingo (1997), but based on a more recent version of the Hadley Centre model (HadAM3; Pope et al., 2000), using sulphur emission scenarios for year 2000 from the SRES scenario (Johns et al., 2001) and including a simple parametrization of sea salt aerosol (Jones et al., 1999); (i) contrails (Minnis et al., 1999); (j) surface albedo change due to changes in land use (Hansen et al., 1998), (k) solar variability (Haigh, 1996). Note that the scale differs for the various panels. Different modelling studies may show considerably different spatial patterns as described in the text. (Units: Wm-2)
An example of the radiative forcing due to the combined effects of present day concentrations of CO2, CH4, N2O, CFC-11 and CFC-12 is shown in Figure 6.7a (see Shine and Forster, 1999, for further details). The zonal nature of the radiative forcing is apparent and is similar to the radiative forcing due to well-mixed greenhouse gases from Kiehl and Briegleb (1993) presented in IPCC (1994). The radiative forcing ranges from approximately +1 Wm-2 at the polar regions to +3 Wm-2 in the sub-tropics. The pattern of the radiative forcing is governed mainly by variations of surface temperature and water vapour and the occurrence of high level cloud (Section 6.3).
An example of the radiative forcing due to stratospheric ozone depletion is shown in Figure 6.7b which was calculated using zonal mean stratospheric ozone depletions from 1979 to 1994 (WMO, 1995) by Shine and Forster (1999). The zonal nature of the radiative forcing is apparent with strongest radiative forcings occurring in polar regions which are areas of maximum ozone depletion (Section 6.4). The gradient of the radiative forcing tends to enhance the zonal gradient of the radiative forcing due to gaseous species shown in Figure 6.7a.
An example of the radiative forcing due to modelled increases in tropospheric O3 is shown in Figure 6.7c (Berntsen et al., 1997; Shine and Forster, 1999). While the exact spatial distribution of the radiative forcing may differ in other studies, the general pattern showing a maximum radiative forcing over North Africa and the Middle East is common to many other studies (Section 6.5). However, observational evidence presented by Kiehl et al. (1999) suggests that this might be an artefact introduced by the chemical transport models. The radiative forcing is much less homogeneous than for well-mixed greenhouse gases, the maximum radiative forcing being due to the coincidence of a relatively large O3 change, warm surface temperatures, high surface reflectance, and cloud-free conditions.
An example of the direct radiative forcing due to sulphate aerosol is shown in Figure 6.7d (Haywood et al., 1997a). In common with many other studies (see Section 6.7), the direct radiative forcing is negative everywhere, and there are three main areas where the radiative forcing is strongest in the Northern Hemisphere corresponding to the main industrialised regions of North America, Europe, and Southeast Asia. In the Southern Hemisphere, two less strong regions are seen. The ratio of the radiative forcing in the Northern Hemisphere to the Southern Hemisphere has been reported by many studies and varies from 2 (Graf et al., 1997) to approximately 7 (Myhre et al., 1998c). Generally, the strongest sulphate direct radiative forcing occurs over land areas although the low surface reflectance means that areas of water close to heavily industrialised regions such as the Mediterranean Sea, the Black Sea and the Baltic Sea result in strong local radiative forcings. Due to the large areal extent of ocean regions, the contribution to the total annual mean radiative forcing from ocean regions is significant. The ratio of the annual mean radiative forcing over land to that over oceans varies from approximately 1.3 (Kiehl et al., 2000) to 3.4 (Boucher and Anderson, 1995) (see Table 6.4).
An example of the direct radiative forcing due to organic carbon and black carbon from biomass burning is shown in Figure 6.7e (Penner et al., 1998b; Grant et al., 1999). While the radiative forcing is generally negative, positive forcing occurs in areas with a very high surface reflectance such as desert regions in North Africa, and the snow fields of the Himalayas. This is because biomass burning aerosols contain black carbon and are partially absorbing. The dependency of the sign of the radiative forcing from partially absorbing aerosols upon the surface reflectance has been investigated by a number of recent studies (e.g., Chylek and Wong, 1995; Chylek et al., 1995; Haywood and Shine, 1995; Hansen et al., 1997a). The strongest negative radiative forcing is associated with regions of intense biomass burning activity namely, South America, Africa, and Southern Asia and Indonesia and differ from the regions where the sulphate radiative forcing is strongest (Figure 6.7d), being confined to approximately 30oN to 30oS.
An example of the direct radiative forcing due to organic and black carbon from fossil fuel burning is shown in Figure 6.7f (Penner et al., 1998b; Grant et al., 1999). In contrast to the direct radiative forcing from biomass burning (Figure 6.7e), the modelled direct radiative forcing is generally positive except over some oceanic regions near industrialised regions such as the Mediterranean Sea and Black Sea. This is because, on average, aerosols emitted from fossil fuels contain a higher black/organic carbon ratio than biomass aerosols (Penner et al., 1998b; Grant et al., 1999) and are thus more absorbing. Comparison of the radiative forcing due to sulphate aerosols reveals that the areas of strongest sulphate direct radiative forcing are offset to some degree by the radiative forcing due to fossil fuel emissions of black carbon as shown in calculations by Haywood et al. (1997a) and Myhre et al. (1998c). Additional regions of moderate positive radiative forcing are present over areas of high surface reflectance such as northern polar regions and the North African deserts.
An example of the direct radiative forcing due to anthropogenic emissions of mineral dust is shown in Figure 6.7g (Tegen et al., 1996). Areas of strong positive forcing are shown over regions with high surface reflectance such as desert regions in Africa and over the snow surfaces of the Himalayas and areas of strong negative forcing are apparent over ocean areas close to mineral dust sources such as off the coasts of Arabia and North Africa. The exact switchover between areas of positive and negative radiative forcing are not well established owing to uncertainties in the modelled mineral aerosol optical properties and depends upon the assumed single scattering albedo (Miller and Tegen, 1998), the long-wave properties and altitude of the aerosol (Section 6.7.6).
An example of the "first" indirect radiative effect (i.e., changes in the cloud reflectivity only) due to anthropogenic industrial aerosols is shown in Figure 6.7h. The forcing is calculated diagnostically in a similar way to Jones and Slingo (1997), but is based on a more recent version of the Hadley Centre model (HadAM3; Pope et al., 2000), uses updated sulphur emission scenarios from the SRES scenario for the year 2000 (Johns et al., 2001) and also includes a simple parametrization of sea salt aerosol (Jones et al., 1999). The spatial distribution of the indirect radiative forcing is quite different from the direct radiative forcing with strong areas of forcing off the coasts of industrialised regions (note the change in scale of Figure 6.7h). There is a significant radiative forcing over land regions such as Europe and the Eastern coast of North America, and Southeast Asia. The spatial distribution of the indirect radiative forcing will depend critically upon the assumed spatial distribution of the background aerosol field and the applied anthropogenic perturbation and differs substantially between studies (see Section 6.8.5). It would have a very different spatial distribution if the effect of biomass burning aerosols were included. The "second" indirect effect whereby inclusion of aerosols influences the lifetime of clouds is not considered here due to the complications of necessarily including some cloud feedback processes in the estimates (Section 6.8.5), but may well resemble the spatial distribution of the "first" indirect effect.
The spatial distribution of three other radiative forcing mechanisms are considered in this section: the radiative forcing due to contrails, land-use change, and solar variability. The radiative forcing due to other constituents such as nitrate aerosol and aviation-induced cirrus that are very difficult to quantify at present are not presented as geographic distributions of the radiative forcing are currently considered to be speculative.
An example of the present day radiative forcing due to the effect of contrails is shown in Figure 6.7i (Minnis et al., 1999). The radiative forcing is very inhomogeneous, being confined to air-traffic corridors (IPCC, 1999). Future scenarios for aircraft emissions may shift the current geographical pattern of the radiative forcing as discussed in IPCC (1999).
An example of an estimate of the radiative forcing due to changes in land use is shown in Figure 6.7j (Hansen et al., 1998). The areas of strongest negative forcing occur at northern latitudes of the Northern Hemisphere due to the felling of forests which have a lower albedo when snow is present (see Section 6.13). Additional effects are due to the change in albedo between crop lands and naturally occurring vegetation. Examples where the radiative forcing is positive include areas where irrigation has enabled crop-growing on previously barren land.
An example of the present day radiative forcing due to solar variability is shown in Figure 6.7k. The solar radiative forcing was calculated by scaling the top of the atmosphere net solar radiation such that the global average is +0.3 Wm-2 (as deduced for global average radiative forcing since 1750, see Section 6.13.1). Thus it assumes a 0.125% increase in solar constant and no change in any other parameter (e.g., O3, cloud). The cloud and radiation fields were calculated within a run of the UGAMP GCM (Haigh, 1996). The strongest radiative forcings exist where the surface reflectance is low (i.e., oceanic regions) and the insolation is highest (i.e., equatorial regions). The solar radiative forcing is also modulated by cloud amount, areas with low cloud amount showing the strongest radiative forcing. The solar radiative forcing is more inhomogeneous than the radiative forcing due to gaseous species (Section 6.14.1), but more homogeneous than the radiative forcing due to aerosol species (Section 6.14.2).
While the preceding sections have shown that the radiative forcing due to the different forcing mechanisms have very different spatial distributions, it is essential to note that the forcing/response relationship given in Section 6.2 relates global mean radiative forcings to global mean temperature response. Thus, it is not possible to simply map the geographical radiative forcing mechanisms by assuming a globally invariant climate sensitivity parameter to predict a geographic temperature response, due to the complex nature of the atmosphere-ocean system. Rather, the effects of spatial inhomogeneity in the distribution of the radiative forcing may lead to locally different responses in surface temperature (Section 6.2) indicating that the spatial distributions of the radiative forcing need to be accurately represented to improve regional estimates of surface temperature response and other physical parameters.
Other reports in this collection