Uncertainties in Estimates of Radiative Flux Perturbations

Aerosol forcing, as inferred from representing the geographical distribution of anthropogenic aerosols and their radiative influences in climate models (i.e., the so-called forward estimates of forcing), has been consistently larger than "inverse" estimates, whereby aerosol forcing is inferred from the total forcing required to obtain the observed temperature change over the industrial period for the generally accepted range of climate sensitivity, observed ocean heat uptake, and given anthropogenic greenhouse gas forcing (Anderson et al. 2003). Given the large uncertainties in processes connected to both aerosol forcings and climate sensitivity, it is not obvious that this discrepancy can simply be attributed to an overestimate of the aerosol forcings by forward estimates.

In a multimodel comparison, Denman et al. (2007) analyzed the forcing of anthropogenic aerosols with respect to the aerosol compounds and aerosol-cloud interaction processes considered (i.e., whether or not aerosol interactions with mixed-phase and ice clouds are taken into account). If sulfate or sulfate and black carbon are solely used, they found that global mean forcing is larger than if organics and aerosol interactions with mixed-phase and ice clouds are included as well. However, the variations within a given aerosol compound category are at least as large as the differences between different categories.

Calculations of the total aerosol forcing depend critically on estimates of both present-day and preindustrial aerosol (precursor) emissions. In particular, the limited knowledge of preindustrial aerosol concentrations introduces considerable uncertainty about the radiative forcing by aerosols. This is manifest in our lack of knowledge about the physical processes that determine the unperturbed cloud droplet number concentration (CDNC) for use in global climate model parameterizations. Observations show that CDNC, even in very clean regions, falls rarely below some lower bound (approximately 10 cm-3).

0 0

Post a comment