Nature Versus Nurture in Shallow Convection

David Romps Lawrence Berkeley National Laboratory
Zhiming Kuang Harvard University

Category: Modeling

Working Group: Cloud Life Cycle

We use tracers in a large-eddy simulation of shallow convection to show that stochastic entrainment, not cloud-base properties, determine the fate of convecting parcels. The tracers are used to diagnose the correlations between a parcel’s state above the cloud base and both the parcel’s state at the cloud base and its entrainment history. We found that the correlation with the cloud-base state goes to zero a few hundred meters above the cloud base. On the other hand, correlations between a parcel’s state and its net entrainment are large. Evidence is found that the entrainment events may be described as a stochastic Poisson process. We construct a parcel model with stochastic entrainment that is able to replicate flux profiles and, more importantly, the observed variability. Turning off cloud-base variability has little effect on the results, which suggests that stochastic mass-flux models may be initialized with a single set of properties. The success of the stochastic parcel model suggests that it holds promise as the framework for a convective parameterization.

This poster will be displayed at ASR Science Team Meeting.

POSTER in PDF: for proper viewing, it should be viewed with Adobe Acrobat Reader. Download the latest version from the Adobe Reader website.