M2 - Benchmarking and Specification


In M2, benchmark simulations with two established LES models (PALM and UCLA-LES) are performed. In a first step, strongly idealized simulations of a dry and a moist convective boundary layer, each with varying resolutions and domain sizes, are produced to serve as reference for ICON-LES developed within module M1 (Fig. 1).

Currently, as a second step, semi-idealized simulations with both LES models are underway. The simulations cover a continuous period of 12 days of the HOPE measurement campaign (April and May 2013). With this so-called HD(CP)2 Prototype Simulations (HPS), we make a step towards less-idealized simulations by considering processes occurring on larger scales than those typically covered by LES. Large-scale forcing, which is needed for this kind of simulations, is taken from COSMO-DE analysis data. Compared to the planned HD(CP)2 simulations with ICON-LES in a limited-area mode, the HPS is more idealized but has the advantage of higher (horizontal) resolution (50 m), easier model setup and less required computing resources due to a smaller domain. Our aim is to validate the HPS with the observational datasets from the HOPE campaign, to compare it with the final HD(CP)2 simulation with ICON-LES and to provide HPS model data for the development of diagnostics.

One of the processes that are poorly resolved at a (horizontal) resolution of 100 m (as envisioned for the ICON-LES HD(CP)2 simulations) is the stable nocturnal boundary layer. A resolution study of idealized night-to-day time transitions with UCLA-LES indicates that, although the nocturnal boundary layer is poorly presented at 100 m resolution, influence of these biases on daytime convection is very limited.

Fig. 1: Intercomparison of simulations of a moist-convective boundary layer with 50 m resolution for four model configurations (UCLA-LES, PALM with two different advection schemes by Wicker and Skamarock (2002) (WS) and by Piacsek and Williams (1980) (PW), and ICON-LES). Top: Time series of cloud fraction, mean local boundary layer height, and mean latent heat flux. Bottom: Vertical profiles of potential temperature, total water mixing ratio, and liquid water mixing ratio.