Large-eddy simulation of the response of shallow cumuli to land surface heterogeneity

Poster PDF

Description

In this study, we examine the response of shallow cumuli over a heterogeneous land surface, highlighting the combined influence of land surface heterogeneity length scale and the background wind speed. Our LES experiments use the new composite case of summertime, fair-weather, shallow cumulus clouds over the ARM SGP site (Zhang et al., 2017). The land surface is prescribed to be a chessboard pattern with alternating warm/dry and cool/wet patches. While only the patch size and the imposed background wind speed change in all cases, the domain averaged surface heat fluxes, surface available energy, and the heterogeneity amplitude remain the same. To quantify the strength of land surface heterogeneity in terms of heterogeneity length scale and atmospheric wind speed, we adopt the non-dimensional parameter Lhetero that is the ratio of the time for the flow to cross over the patch to the large-eddy turnover time. Then, we analyze the shallow-to-deep transition cases using Lhetero as a predictor. In general, continental shallow cumuli transition either into precipitating shallow cumuli or deep convective clouds when Lhetero becomes larger than 20 and the underlying heterogeneity has a length scale greater than 2.4km. The transition cases are characterized with the well-defined mesoscale circulation that attributes to the enhanced buoyancy and moisture convergence over warm/dry patch and the subsequent development of deep cumuli. The shallow cumuli in non-transition cases are slightly larger over wet patches as the PBL becomes wetter from the higher surface evaporation. The PBL turbulent profiles in the transition cases deviate greatly from the non-transition cases due to the large contribution from the mesoscale component. It should be noted that the turbulent profiles in non-transition cases resemble the profiles over the homogeneous land surface. Overall, the non-dimensional parameter serves as a good predictor for the occurrence of precipitation; however, it does not scale linearly with the precipitation intensity, timing of precipitation maximum, or cloud size. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-716532.