Cumulus Entrainment: Verifying Retrieval Methods and Interpreting Environmental Sensitivities Using Large-Eddy Simulation

A quantitative understanding of cumulus entrainment remains elusive, in part due to the difficulty of directly observing cumulus entrainment rates. To overcome this challenge, multiple approaches to retrieving bulk fractional entrainment rates within cumuli using ground-based in situ and remote data at the Department of Energy’s Atmospheric Radiation Measurement (ARM) climate research sites have been developed recently. Among these retrievals are the parcel-based method by Jensen and Del Genio (2006), the iterative Entrainment Rate in Cumulus Algorithm (ERICA; Wagner et al. 2013) and a newly developed turbulent-kinetic-energy (TKE) similarity theory based retrieval approach. To provide a first-ever numerical verification of these methods, large-eddy simulations (LES) of a broad range of continental and maritime shallow cumulus convection cases are used as Observing System Simulation Experiments (OSSEs). From these simulations, simulated entrainment retrievals are performed and compared directly to bulk diagnosed entrainment rates. To begin, all quantities used in the retrievals are assumed to be perfectly observable, and these quantities are evaluated over the full LES domain. This evaluation reveals a mean absolute error for the different retrieval methods ranging from 23% (TKE) to 43% (ERICA). The OSSE framework also facilitates physical interpretation of the environmental sensitivities of the simulated cumulus entrainment. While the simulated fractional entrainment rates are largely insensitive to environmental humidity, they are strongly sensitive to continentality: maritime cloud fields exhibit substantially higher entrainment rates than continental cloud fields. Analysis of the TKE retrieval suggests that this difference stems from a larger cloud base mass flux over land, driven by larger surface sensible heat fluxes, which effectively decreases the turbulent dissipation per unit cloud mass. Furthermore, the diurnal cycle of cloud-layer fractional entrainment rate over land is analyzed and physically interpreted based on the corresponding diurnal cycles of relevant environmental parameters.