The Breakdown of Paradigmatic CMIP6 Assumptions for Parameterizing Warm Rain Initiation

Operational global climate models—as represented by the most recent phase of the Coupled Model Intercomparison Project (CMIP6)—parameterize the microphysics of warm rain initiation via a shared conceptual framework. An essential, ubiquitous, and oft-critiqued component of this framework is autoconversion, which abstracts the initial phases of drop growth by coagulation (the collision and subsequent coalescence of drops). Here we investigate the breakdown of paradigmatic assumptions underlying CMIP6 representations of autoconversion with idealized size-resolved modeling of coagulation and observed liquid drop size distributions. We demonstrate that shared mathematical properties of CMIP6 autoconversion functional forms can misrepresent the temporal evolution of coagulation and limit variability in the timing of warm rain initiation. We quantify the microphysical regime where the CMIP6 assumptions break down. Our analysis of in-situ warm cloud observations suggests that this breakdown regime occurs in about half of the cloudy air sampled. This work raises concerns about the physical basis of operational global climate models’ representations of warm rain initiation, which influence modeled rain frequency, rain intensity, and cloud albedo.