Figure 6-9: Zonal and annual mean radiative
Figure 6-10: Radiative forcing (dotted line), global
In Sections 6.3 and 6.4 we have reported radiative forcing calculations for various aspects of aircraft-induced perturbations to radiatively active substances. One of the basic ideas behind this calculation was the validity of the concept of RF as a quantitative predictor for climate change (see Section 6.2.1). We implicitly assume that contributions from individual perturbations to the change in global mean surface temperature are additive, at least as a first-order approximation. The radiative forcing concept requires a quasi-constant climate sensitivity parameter l within the same model for different types of RF and different magnitudes of RF. This assumption holds for changes in most of the well-mixed greenhouse gases and for variations in solar irradiance (e.g., Hansen et al., 1984b; Shine et al., 1995; Hansen et al., 1997b; Roeckner et al., 1999). As noted in previous IPCC reports (e.g., IPCC, 1996), there are limitations to the applicability of the radiative forcing concept, especially for RFs that are highly variable geographically or seasonally (i.e., typical of aircraft perturbations).
As a clear example of large spatial differences in radiative forcing, the annual mean radiative imbalance as a function of latitude is shown in Figure 6-9 for the big four RFs for subsonic aviation in 1992. Whereas the radiative imbalances from CO2 (positive) and CH4 (negative) are nearly uniform from south pole to north pole, those from O3 (peaking at 0.065 W m-2 at 30�N) and contrails (peaking at 0.10 W m-2 at 40�N) are highly concentrated in northern mid-latitudes. The global means of all four contributions are about equal (see Table 6-1). The NOx-driven perturbations to O3 and CH4 produce RFs (+0.23 and -0.14 W m-2, respectively) that are of similar magnitude and in part cancel. However, the latitudinal distribution of the radiative imbalance from these two perturbations does not cancel: The combined O3+CH4 forcing is positive in the Northern Hemisphere and negative in the Southern Hemisphere. The response of the climate system to such geographically non-homogeneous forcing is unknown. At least regional differences in climate response can be expected, and there may even be differences in the global mean response (see Section 6.5.2).
The limitations and advantages of the radiative forcing concept have recently been demonstrated by Hansen et al. (1997b). A classic test case considers radiative forcing and associated climate change for ozone perturbations occurring at different altitudes. In a series of numerical experiments with a low-resolution GCM (Hansen et al., 1997a), the impact of 100 DU ozone added to the various model layers was investigated. As in the calculations of Lacis et al. (1990), Hauglustaine and Granier (1995), Forster and Shine (1997), and Brasseur et al. (1998), strong variation of radiative forcing with the altitude of ozone perturbation is found: negative forcing for perturbation in the middle stratosphere, and positive forcing for perturbation in the troposphere and lower stratosphere. RF is a maximum for a perturbation close to the tropopause (Figure 6-10, dotted curve).
Without cloud and water vapor feedbacks, the change in global-mean surface air temperature exhibits a similar dependency on the altitude of the ozone perturbation as radiative forcing does (with a maximum close to the tropopause, not shown in figure). However, when all model feedbacks are included, the maximum global-mean surface temperature change in the GISS model occurs for an ozone perturbation in the middle troposphere (Figure 6-10, dashed line). The climate sensitivity parameter l varies widely as a function of the altitude of the ozone perturbation varies widely from the middle stratosphere to the surface layer (Figure 6-10, solid line). Peak sensitivity in this model occurs in the lower troposphere because of feedbacks on clouds. However, for the altitude range at which aircraft O3 perturbations contribute significantly to the total RF (e.g., 2-14 km), the value of l is within a factor of 2 of the climate sensitivity parameter for doubling of CO2 in the same model, 0.92 K/(W m-2) (Hansen et al., 1997b).
In GCM experiments with a similar model that focused on aircraft-like perturbations, the specific feedbacks that minimized the response for aircraft releases were high-level cloud cover (Rind and Lonergan, 1995-considering ozone impacts) and sea ice (Rind et al., 1996-considering water vapor releases). Sensitivity of l to layered perturbations is also found by Ponater et al. (1998) but with opposite sense: The climate sensitivity parameter for aircraft-induced ozone is higher than that for well-mixed greenhouse gases. The difference in these results likely depends on the formulation of clouds within the two models.With our present understanding of climate modeling and critical feedback processes in the troposphere, we can do no better than adopt the tropopause value of RF after stratospheric adjustment. However, for aircraft-induced climate perturbations, the additivity of RFs across all perturbations cannot be taken for granted and adds further uncertainty.
Other reports in this collection