Premium
Comparison of the Moist Parcel‐in‐Cell (MPIC) model with large‐eddy simulation for an idealized cloud
Author(s) -
Böing Steven J.,
Dritschel David G.,
Parker Douglas J.,
Blyth Alan M.
Publication year - 2019
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.3532
Subject(s) - grid , computer science , turbulence , scale (ratio) , large eddy simulation , eulerian path , flow (mathematics) , convergence (economics) , mixing (physics) , meteorology , eddy , statistical physics , mechanics , environmental science , lagrangian , mathematics , physics , geometry , quantum mechanics , economics , economic growth
The ascent of a moist thermal is used to test a recently developed essentially Lagrangian model for simulating moist convection. In this Moist‐Parcel‐In‐Cell (MPIC) model, a number of parcels are used to represent the flow in each grid cell. This has the advantage that the parcels provide an efficient and explicit representation of subgrid‐scale flow. The model is compared against Eulerian large‐eddy simulations with a version of the Met Office NERC Cloud model (MONC) which solves the same equations in a more traditional Eulerian scheme. Both models perform the same idealized simulation of the effects of latent heat release and evaporation, rather than a specific atmospheric regime. Dynamical features evolve similarly throughout the development of the thermal using the two approaches. Subgrid‐scale properties of small‐scale eddies captured by the MPIC model can be explicitly reconstructed on a finer grid. MPIC simulations thus resolve smaller features when using the same grid spacing as MONC, which is useful for detailed studies of turbulence in clouds. The convergence of bulk properties is also used to compare the two models. Most of these properties converge rapidly, though the probability distribution function of liquid water converges only slowly with grid resolution in MPIC. This may imply that the current implementation of the parcel mixing mechanism underestimates small‐scale mixing. Finally, it is shown how Lagrangian parcels can be used to study the origin of cloud air in a consistent manner in MPIC.