DNS and LES for Simulating Stratocumulus: Better Together

Abstract

We argue that combining direct numerical simulation (DNS) with large-eddy simulation (LES) and field studies could accelerate current lines of stratocumulus research. LES allows for a faster and more holistic study of the parameter space, but LES is sensitive to details of its formulation because the energetics are tied to unresolved processes in the cloud top region. One way to assess this sensitivity is through field studies. Another way is through DNS. In particular, DNS can be used to test the hypothesis that LES, even with an inadequate representation of the physics of cloud top entrainment, properly quantifies the sensitivity of cloud-topped boundary layers to changing environmental conditions. We support this argument by contrasting theoretical aspects of both techniques, by presenting first DNS results of a stratocumulus-topped boundary layer and discussing their convergence toward Reynolds number similarity, and by showing the consistency of DNS results with LES results and field measurements.