Impact of Autoconversion Schemes On Prediction of the Aerosol Indirect Effect

Online GCM (NASA GISS II') simulations of aerosol and cloud droplet number are used to assess the sensitivity of predicted climate and aerosol indirect effects to the treatment of autoconversion for warm clouds. Four autoconversion parameterizations and direct integration of the kinetic collection equation (KCE) with kernels that account for both gravitational and turbulent effects on collection are used to represent autoconversion. Predictions of the combined cloud albedo and lifetime effects on aerosol indirect forcing are found to be quite sensitive to the autoconversion scheme used, with differences between present-day and preindustrial aerosol indirect forcing ranging from -0.89 W/m² to -2.05 W/m². Autoconversion parameterizations are evaluated by the extent to which the predicted timescale for conversion of cloudwater into drizzle is consistent with observations. Of the schemes considered, the MC, P6, and KK parameterizations tend to predict values within the observed range of autoconversion timescale when clouds are close to the precipitation threshold. Analysis of the link between global annual average changes in LWP, autoconversion, and aerosol indirect forcing shows that the contribution of cloud albedo effects to aerosol indirect forcing is about -0.9 W/m², with the remainder being the cloud lifetime contribution. Reduction of autoconversion uncertainty to a factor of 3-5 would result in a 30-50% aerosol indirect forcing uncertainty. Quantifying autoconversion prediction uncertainty using the observational approach outlined here offers a robust method for determining the aerosol indirect forcing uncertainty associated with the combined cloud albedo and lifetime effects.