|Working Group I: The Scientific Basis|
|Other reports in this collection|
6.13 Global Mean Radiative Forcings
The global, annual mean radiative forcing estimates from 1750 to the present (late 1990s; about 2000) for the different agents are plotted in Figure 6.6, based on the discussions in the foregoing sections. As in the SAR, the height of the rectangular bar denotes a central or best estimate of the forcing, while the vertical line about the bar is an estimate of the uncertainty range, guided by the spread in the published results and physical understanding, and with no statistical connotation. The uncertainty range, as employed in this chapter, is not the product of systematic quantitative analyses of the various factors associated with the forcing, and thus lacks a rigorous statistical basis. The usage here is different from the manner “uncertainty range” is defined and addressed elsewhere in this document. The SAR had also stated a “confidence level” which represented a subjective judgement that the actual forcing would lie within the specified uncertainty range. In order to avoid the confusion over the use of the term “confidence level”, we introduce in this assessment a “level of scientific understanding” (LOSU) that represents, again, a subjective judgement and expresses somewhat similar notions as in the SAR (refer also to IPCC, 1999). The LOSU index for each forcing agent is based on an assessment of the nature of assumptions involved, the uncertainties prevailing about the processes that govern the forcing, and the resulting confidence in the numerical value of the estimate. The subjectivity reflected in the LOSU index is unavoidable and is necessitated by the lack of sufficient quantitative information on the uncertainties, especially for the non-well-mixed greenhouse gas forcing mechanisms. In the case of some forcings, this is in part due to a lack of enough investigations. Thus, the application of rigorous statistical methods to quantify the uncertainties of all of the forcing agents in a uniform manner is not possible at present.
The discussions below relate to the changes with respect to the SAR estimates. In many respects, there is a similarity between the estimates, range and understanding levels listed here, and those stated in the recent studies of Hansen et al. (1998) and Shine and Forster (1999). Table 6.11 compares the numerical values with the estimates in the SAR. Also, the Northern to Southern Hemisphere ratio is shown for the present estimates (see also Section 6.14). Table 6.12 summarises the principal aspects known regarding the forcings, along with a brief listing of the key uncertainties in the processes which, in turn, lead to uncertainties in and affect the reliability of the quantitative estimates.
The total forcing estimate for well-mixed greenhouse gases is slightly less now, by about 1% (see Section 6.3) compared to the estimate given in the SAR. The uncertainty range remains quite small and these estimates retain a “high” LOSU. This forcing continues to enjoy the highest confidence amongst the different natural and anthropogenic forcings.
The estimate for stratospheric O3 has increased in magnitude, owing mainly to the inclusion of observed ozone depletions through mid-1995 and beyond. It is an encouraging feature that several different model calculations yield similar estimates for the forcing. The uncertainty range remains similar to that given in the SAR. These arguments suggest an elevation of the confidence in the forcing estimate relative to the SAR. Accordingly, a “medium” LOSU is assigned here. A still higher elevation of the rank is precluded because the O3 loss profile near tropopause continues to be an uncertainty that is significant and that has not been adequately resolved. Also, the global strato-spheric temperature change calculations involved in the forcing determination are not quantitatively identical to the observed changes, which in turn, affects, the precision of the forcing estimate.
The estimate for tropospheric O3 (0.35 ± 0.15 Wm-2) is on firmer grounds now than in the SAR. Since that assessment, many different models have been employed to compute the forcing, including one analysis constrained by observations. These have resulted in a narrowing of the uncertainty range and increased the confidence with regards to the central estimate. The preceding argument strongly suggests an advancement of the confidence in this forcing estimate. Hence, a “medium” LOSU is accorded for tropospheric O3 forcing. Key uncertainties remain concerning the pre-industrial distributions, the effects of stratospheric-tropospheric exchange and the manner of its evolution over time, as well as the seasonal cycle in some regions of the globe.
As the LOSU rankings are subjective and reflect qualitative considerations, the fact that tropospheric and stratospheric O3 have the same ranks does not imply that the degree of confidence in their respective estimates is identical. In fact, from the observational standpoint, stratospheric O3 forcing, which has occurred only since 1970s and is better documented, is on relatively firmer ground. Nevertheless, both O3 components are less certain relative to the well-mixed greenhouse gases, but more so compared with the agents discussed below.
The estimate for the direct sulphate aerosol forcing has also seen multiple model investigations since the SAR, resulting in more estimates being available for this assessment. It is striking that consideration of all of the estimates available since 1996 lead to the same best estimate (–0.4 Wm-2) and uncertainty (–0.2 to –0.8 Wm-2) range as in the previous assessment. As in the case of O3, that could be a motivation for elevating the status of knowledge of this forcing to a higher confidence level. However, there remain critical areas of uncertainty concerning the modelling of the geographical distribution of sulphate aerosols, spatial cloud distributions, effects due to relative humidity etc. Hence, we retain a “low” LOSU for this forcing.
The SAR stated a radiative forcing of +0.1 Wm-2 for fossil fuel (FF) black carbon aerosols with a range +0.03 to +0.3 Wm-2, and a “very low” level of confidence. For biomass burning (BB) aerosols, the SAR stated a radiative forcing of -0.2 Wm-2 with a range -0.07 to -0.6 Wm-2, and a “very low” level of confidence. In the present assessment, the radiative forcing of the black carbon component from FF is estimated to be +0.2 Wm-2 with a range from +0.1 to +0.4 Wm-2 based on studies since the SAR. A “very low” LOSU is accorded in view of the differences in the estimates from the various models. The organic carbon component from FF is estimated to yield a forcing of –0.1 Wm-2 with a range from –0.03 to –0.30 Wm-2; this has a “very low” LOSU. Note that extreme caution must be exercised in adding the uncertainties of the organic and black carbon components to get the uncertainty for FF as a whole. For BB aerosols, no attempt is made to separate into black and organic carbon components, in view of considerable uncertainties. The central estimate and range for BB aerosols remains the same as in the SAR; this has a “very low” LOSU in view of the several uncertainties in the calculations (Section 6.7).
Mineral dust is a new component in the current assessment. The studies on the “disturbed” soils suggest an anthropogenic influence, with a range from +0.4 to -0.6 Wm-2. In general, the evaluation for dust aerosol is complicated by the fact that the short-wave consists of a significant reflection and absorption component, and the long-wave also exerts a substantial contribution by way of a trapping of the infrared radiation. Thus, the net radiative energy gained or lost by the system is the difference between non-negligible positive and negative radiative flux changes operating simultaneously. Because of this complexity, we refrain from giving a best estimate and accord this component a “very low” LOSU.
As explained in Section 6.8, the “indirect” forcing due to all tropospheric aerosols can be thought of as comprising two effects. Only the first type of effect as applicable in the context of liquid clouds is considered here. As in the SAR, no best estimate is given in view of the large uncertainties prevailing in this problem (Section 6.8). The range (0 to -2 Wm-2) is based on published estimates and subjective assessment of the uncertainties. Although several model studies suggest a non-zero, negative value as the upper bound (about -0.3 Wm-2), substantial gaps in the knowledge remain which affect the confidence in the model simulations of this forcing (e.g., uncertainties in aerosol and cloud processes and their representations in GCMs, the potentially incomplete knowledge of the radiative effect of black carbon in clouds, and the possibility that the forcings for individual aerosol types may not be additive), such that the possibility of a very small negative value cannot be excluded; thus zero is retained as an upper bound as in the SAR. In view of the large uncertainties in the processes and the quantification, a “very low” LOSU is assigned to this forcing. Inclusion of the second indirect effect (Chapter 5) is fraught with even more uncertainties and, despite being conceptually valid as an anthropogenic perturbation, raises the question of whether the model estimates to-date can be unambiguously characterised as an aerosol radiative forcing.
Aviation introduces two distinct types of perturbation (Section 6.8). Contrails produced by aircraft constitute an anthropogenic perturbation. This is estimated to contribute 0.02 Wm-2 with an uncertainty of a factor of 3 or 4 (IPCC, 1999); the uncertainty factor is assumed to be 3.5 in Figure 6.6. This has an extremely low level of confidence associated with it. Additionally, aviation-produced cirrus is estimated by IPCC (1999) to yield a forcing of 0 to 0.04 Wm-2, but no central estimate or uncertainty range was estimated in that report. Both components have a “very low” LOSU.
Volcanic aerosols that represent a transient forcing of the climate system following an eruption are not plotted since they are episodic events and cannot be categorised as a century-scale secular forcing, unlike the others. However, they can have substantial impacts on interannual to decadal scale temperature changes and hence are important factors in the time evolution of the forcing (see Section 6.15). Some studies (Hansen et al., 1998; Shine and Forster, 1999) have attempted to scale the volcanic forcings in a particular decade with respect to that in a quiescent decade.
Land-use change was dealt with in IPCC (1990) but was not considered in the SAR. However, recent studies (e.g., Hansen et al., 1998) have raised the possibility of a negative forcing due to deforestation and the ensuing effects of snow-covered land albedo changes in mid-latitudes. There are not many studies on this subject and rigorous investigations are lacking such that this forcing has a “very low” LOSU, with the range in the estimate being 0 to –0.4 Wm-2 (central estimate: -0.2 Wm-2). Note that the land-use forcing here is restricted to that due to albedo change.
Solar forcing remains the same as in the SAR, in terms of best estimate, the uncertainty range and the confidence level. Thus, the range is 0.1 to 0.5 Wm-2 with a best estimate of 0.3 Wm-2, and with a “very low” LOSU.
Other reports in this collection