Abstract
In this paper a large-eddy “smooth” cloud (SC) model will be presented with smooth implying that the entire model converges under a Newton-based solution procedure or that time scales within the SC model are being resolved. Besides ensuring that time scales within microphysical parameterizations are resolved, convergence of Newton’s method requires that advection schemes near cloud boundaries should not induce fast time scales. For example, flux-corrected transport (FCT) schemes that force cloud variables to stay oscillation free near boundaries are typically not differentiable in time and hence may prevent convergence of Newton’s method. To circumvent the use of a FCT scheme, an alternative approach, a cloud-edge (CE) diffusion-based approach, will be presented in this paper. Since the diffusion produced by the CE approach could conceivably lead to the fictitious evaporation of a real cloud, the first major point of this paper will be to document that the SC model when employing an evaporative limiter is able, like most traditional large-eddy cloud models, to reasonably reproduce nondrizzling stratus clouds observed during flight 1 of the Second Dynamics and Chemistry of Marine Stratocumulus field study (DYCOMS-II). However, the SC model obtains the accuracy offered by higher-order time-stepping approaches, unlike most traditional cloud models. In fact, temporal errors from the SC model are shown to be at least two orders of magnitude smaller than those of a traditional large-eddy cloud model. Hence, the second major point of this paper will be to demonstrate the consequence of these large temporal errors found in traditional large-eddy cloud models, that is, the inability to accurately track an identifiable cloud feature in time.
Corresponding author address: J. M. Reisner, Los Alamos National Laboratory, MS D401, Los Alamos, NM 87545. Email: reisner@lanl.gov
