Since this will be the first intercomparison study on shallow cumulus convection, our objective in selecting a case was to keep it as simple as possible but yet realistic enough. In view of this we have selected BOMEX. This is a trade wind cumulus case with vertical profiles which are typical for a large part of the trade wind region. It is a simple case since there are: i) no mesoscale complications, ii) no transitions from/to or remains of stratocumulus and iii) from the observational point of view it is in a satisfying quasi steady-state, i.e. the small scale turbulent cumulus response is in balance with the large scale forcing. One obvious weak point of BOMEX is that there are little cloud data available for this case. The main intercomparison with observations therefore will be limited to average profiles and the various turbulent flux divergences.

During phase 3 of the Barbados Oceanographic and Meteorological Experiment (BOMEX) from 22 June to 30 June 1969 a detailed observational budget study in a 500x500 km^2 square near Barbados has been performed. Data were obtained from rawinsondes launched every 1.5 hr from four ships located at the corners of the BOMEX array. From these data large-scale heat and moisture budgets have been deduced (Holland and Rasmusson 1973; Nitta and Esbensen 1974). For the purpose of the intercomparison, we will concentrate on the undisturbed BOMEX period of phase 3 from 22 to 26 June during which non-precipitating cumuli were the only type of cumulus convection that was observed. Ideally we would like to initialize the LES model with the average observed profiles of u, v, theta_l and qt of this 5-day period and run the model using the diagnosed large-scale forcing. This is however not feasible since the temporal and spatial variations of the fields are such that an actual inversion, such as appears on most individual soundings, is not found in the mean soundings (Nitta and Esbensen 1974). Therefore, instead, we have selected from the BOMEX Rawinsonde atlas (1975) a mean profile of one individual ship over a shorter period.

We use the mean profiles of the Oceanographer, the most northern ship of the BOMEX square, averaged over 22 and 23 June during which a well defined steady state with a strong inversion was present. See section 3.2 for an explicit description of the initial profiles. From the BOMEX Low Level Atlas (1975) the oceanic surface values can be found. See section 3.3 for the explicit values. Our choice of the prescribed forcing is dictated by two conditions. Firstly, we want to choose the forcing as simple as possible in order to keep the case transparent. Secondly the forcing has to be realistic, i.e. in agreement with the diagnosed forcing. We prescribe three important forcings for the run:

1. the large-scale subsidence w

2. the radiative cooling

3. large-scale advection of q_t

See section 3.4 of the case description for the prescribed profiles of the forcings which are kept constant during the run.