How sensitive are aerosol-cloud-precipitation interactions to warm rain representation?

We present results from an intercomparison of precipitation susceptibility (S0) from widely used warm microphysics schemes, with the 1D kinematic driver model (KiD) as the dynamic basis for the intercomparison. The results show that, in general, single-moment bulk schemes produce the most rapid onset of precipitation for all cloud droplet number concentrations (Nd) and liquid water paths (LWP) tested, and the widest range in response to Nd. Employing double-moment rain results in better agreement with a detailed bin microphysics scheme when considering maximum precipitation rate and, in general, a slower onset of precipitation. Comparison of the S0 from the schemes tested shows that double moment schemes can be up to twice as sensitive to changes in Nd than the bin scheme, but the qualitative behaviour of S0, i.e. the change with increasing LWP, altitude and sub-cloud RH, is consistent with the bin scheme. In contrast, the single-moment schemes fail to capture this consistency in the behaviour of S0 and modifying parametrisation alone does not improve this. Hence, it is argued that even with all else equal, if a single moment scheme is used to simulate aerosol-cloud-precipitation interactions, there will be uncertainty in the response of precipitation to changes in Nd.