General presentation
- Fig 1. Satellite image of the Northeast Pacific illustrating the stratocumulus to cumulus transition
The latest GCSS-CFMIP meetings emphasized the importance of further evaluating the ability of both SCM and LES to reproduce the transition between marine stratocumulus and shallow cumulus (Fig 1). In this context two Lagrangian transitional cases are proposed for an intercomparison study of both SCM and LES within the EUCLIPSE-GCSS framework. The first case is the "revisited" ASTEX case, based on the Atlantic Stratocumulus to Cumulus Transition Experiment field campaign (Albrecht et al., 1995), which already served in the past for an intercomparison of SCMs (Bretherton et al., 1995). The increase in computational power allows us now to perform 3-D Large Eddy Simulations of this case as well. The second case gathers in fact a set of three "composite" transitions, based on the observational study of the transitions in boundary layer cloudiness described in Sandu et al. (2010). While ASTEX offers the opportunity to evaluate models against in situ data, this set of "composite" transitions represents a more idealized framework for model evaluation, which offers the possibility of comparing the models for a variety of transition cases, which differ for example in terms of amplitude or timescale of the transition.
The "composite" transitions
The "composite" reference case, and two of its variations corresponding to a faster, and respectively a slower transition in cloud fraction are described in the attached 
pdf. Figure 2 illustrates the time evolution of the cloud fraction obtained for the three cases with UCLA-LES and ECHAM6.
The intercomparison of the different models for the reference case will mainly try to assess whether they agree on:
- the decrease in cloud albedo and cloud cover during the 3 days;
- the time evolution of the cloud fraction (the time scale of the transition);
- the growth rate of the boundary layer height.
Furthermore, by simulating the two proposed variations of the reference case, we hope to understand whether the models are able to reproduce the differences in the amplitude and the timescale of the transitions induced by changes in the environmental context.
- Fig 2. Time evolution of the cloud fraction for the slow, reference and fast transitions, obtained with UCLA-LES (top) and ECHAM6 (bottom).
Initialization
Netcdf files containing the initial profiles and the large scale forcing necessary for the LES and SCM models can be downloaded for the reference, fast and slow transitions. The program used to compute the non-equidistant vertical grid (see cell edges) can be found here zgrid composite.f.
Requested output
For LES, the same netcdf files containing the time series of scalars and hourly averaged profiles as for ASTEX are required (see Stephan de Roode webpage). For the first round, only LES of the reference case are requested, while simulations of the fast and slow cases are optional.
The SCM may simulate the 3 cases from the beginning. For SCMs we recommend to follow the format of the requested data output by Stephan de Roode for the ASTEX case .
For LES results should be sent to irina.sandu@zmaw.de, and for SCM to suvarchal.kumar@zmaw.de
Updates
The initial date should be 15 of July. The initial time, i.e. 10 local time, corresponds to 18 UTC (standard UTC time).
Everybody should relax thetal and qt in the sponge layer towards the initial profiles.
For the surface albedo, eveybody should use the time varying albedo provided by Marat.
For SCM's, the wind speeds should be relaxed towards the initial wind profiles in order to avoid internal oscillations, and subsequent errors in the computation of the surface fluxes.
New: Roel Neggers is going to put all the SCM results for the composite cases on a website where is possible to interactively compare your results with those of other SCMs or LES models. If you haven't submitted your SCM results yet please send them to neggers@knmi.nl and cc irina.sandu@zmaw.de.