A new method for evaluating the impact of vertical distribution on aerosol radiative forcing in general circulation models
Abstract. The quantification and understanding of direct aerosol forcing is essential in the study of climate. One of the main issues that makes its quantification difficult is the lack of a complete comprehension of the role of the aerosol and clouds vertical distribution. This work aims at reducing the incertitude of aerosol forcing due to the vertical superposition of several short-lived atmospheric components, in particular different aerosols species and clouds. We propose a method to quantify the contribution of different parts of the atmospheric column to the forcing, and to evaluate model differences by isolating the effect of radiative interactions only. Any microphysical or thermo-dynamical interactions between aerosols and clouds are deactivated in the model, to isolate the effects of radiative flux coupling. We investigate the contribution of aerosol above, below and in clouds, by using added diagnostics in the aerosol-climate model LMDz. We also compute the difference between the forcing of the ensemble of the aerosols and the sum of the forcings from individual species, in clear-sky. This difference is found to be moderate on global average (14%) but can reach high values regionally (up to 100%). The non-additivity of forcing already for clear-sky conditions shows, that in addition to represent well the amount of individual aerosol species, it is critical to capture the vertical distribution of all aerosols. Nonlinear effects are even more important when superposing aerosols and clouds. Four forcing computations are performed, one where the full aerosol 3-D distribution is used, and then three where aerosols are confined to regions above, inside and below clouds respectively. We find that the forcing of aerosols depends crucially on the presence of clouds and on their position relative to that of the aerosol, in particular for black carbon (BC). We observe a strong enhancement of the forcing of BC above clouds, attenuation for BC below clouds, and a moderate enhancement when BC is found within clouds. BC forcing efficiency amounts to 44, 171, 333 and 178 W m-2 per unit optical depth for BC below, within, above clouds and for the 3-D BC distribution, respectively. The different behaviour of forcing nonlinearities for these three components of the atmospheric column suggests that, an important reason for differences between cloudy-sky aerosol forcings from different models may come from different aerosol and clouds vertical distributions. Our method allows to evaluate the contribution to model differences due to aerosol and clouds radiative interactions only, by reading 3-D aerosol and cloud fields from different GCMs, into the same model. This method avoids differences in calculating optical aerosol properties and forcing to enter into the discussion of inter-model differences. It appears that the above and in-cloud amount of BC is larger for SPRINTARS (190 compared to 179), increasing its cloudy-sky forcing efficiency with respect to LMDz, being thus potentially an important factor for inter-model differences.