Anisotropy of observed and simulated turbulence in marine stratocumulus

Anisotropy of turbulence near the top of the stratocumulus-topped boundary layer (STBL) is studied using large-eddy simulation (LES) and measurements from the POST and DYCOMS-II field campaigns. Focusing on turbulence similar to 100 m below the cloud top, we see remarkable similarity between daytime and nocturnal flight data covering different inversion strengths and free-tropospheric conditions. With lambda denoting wavelength and z(t) cloud-top height, we find that turbulence at lambda/z(t) similar or equal to 0.01 is weakly dominated by horizontal fluctuations, while turbulence at lambda/z(t) > 1 becomes strongly dominated by horizontal fluctuations. Between are scales at which vertical fluctuations dominate. Typical-resolution LES of the STBL (based on POST flight 13 and DYCOMS-II flight 1) captures observed characteristics of below-cloud-top turbulence reasonably well. However, using a fixed vertical grid spacing of 5 m, decreasing the horizontal grid spacing and increasing the subgrid-scale mixing length leads to increased dominance of vertical fluctuations, increased entrainment velocity, and decreased liquid water path. Our analysis supports the notion that entrainment parameterizations (e.g., in climate models) could potentially be improved by accounting more accurately for anisotropic deformation of turbulence in the cloud-top region. While LES has the potential to facilitate improved understanding of anisotropic cloud-top turbulence, sensitivity to grid spacing, grid-box aspect ratio, and subgrid-scale model needs to be addressed.