Using in situ measurements of ice residuals to evaluate ice nucleation and convective transport in CAM5

Analysis of in situ measurements of ice residuals from several NASA aircraft campaigns by Cziczo et al. [2013] suggests that cirrus formation in the upper troposphere is predominately caused by heterogeneous nucleation of mineral dust and metallic aerosols, while a similar analysis by Froyd et al. [2010] suggests that homogeneous freezing of sulfate aerosols is most important for the formation of subvisible cirrus in the tropics. Global simulations by Gettelman et al. [2012] using the Community Atmosphere Model (CAM5) predict that homogeneous freezing is the most important formation mechanism for cirrus in the upper troposphere and appear to be at odds with the results of Czizco et al. [2013]. For a more direct evaluation of these results, we use the specified dynamics version of CAM5 sampled along the same aircraft flight paths to simulate these cloud and aerosol measurements. Initial results suggest that there is too much mixing of convectively transported aerosol to the large scale in the upper troposphere affecting the availability of ice nuclei. We explore the sensitivity of these simulations to changes in the convective transport and cloud microphysics parameterizations.