Experiments ...

Grid

We are asking for this case to be integrated for four simulated hours, on a domain whose vertical extent is greater than 1250m. Note that the lapse rate of the specified temperature profile would equal the observed free atmospheric lapse rate of approximately 3.5 K km^-1 at 1770 m, thus extending the domain beyond 1770 m would lead to increasingly large departures in the mean state unless the temperature profile is set to a constant lapse rate above 1770 m, and the radiative flux profiles are adjusted accordingly. Note, however that such departures from the observations are likely only to affect gravity wave propagation characteristics whose net effect on the simulation of the lower boundary layer are thought to be negligible. The base case horizontal extent of the domain is to be 3.36 km with a nominal grid spacing of 35m, corresponding to 96 points in each horizontal direction. Because of the fine-scale thermal and radiative effects just above cloud top we are asking for simulations whose vertical discretization is 5m or less in the vicinity of the inversion. Boussinesq models should use a base state density of 1.13 g m^-3, and anelastic models should be careful to check that our failure to consider density variations with height in the formulation of Eq. 1 may alter the structure of the free-tropospheric temperature profile.

Plan of Attack

Preliminary tests of this case, by a number of participating groups, indicate that it is a challenging case to represent. For this reason we are asking participants to perform simulations with increasing complexity as enumerated below:

1. Simple forcing: A case with only longwave radiative cooling from cloud top active (at a value equal to the net forcing of the PBL, which amounts to setting F₀ = 48 Wm^-2 and F₁ = 0 in (1)), fixed thermodynamic fluxes (15 and 115Wm^-2 respectively) at the surface, fixed surface momentum flux (u_* = 0.25 ms^-1), and no large-scale subsidence.
2. Partial forcing: As above (i.e., with fixed surface thermodynamic and momentum fluxes) but with introduction of cloud base warming term, i.e., F₀ = 70 Wm^-2 and F₁ = 22 Wm^-2 in (1).
3. Inversion forcing: Introduce subsidence and inversion layer cooling term.
4. Full forcing: Introduction of interactive surface fluxes — these should only be specified for the last two hours, so as to avoid excess fluxes from winds which are initially too large.

Before integrating any case, we ask all investigators to double check that their initial cloud base is within 10m of 600m, and that cloud top liquid water contents are near 0.45 gkg^-1. In addition we ask investigators to volunteer sensitivity studies, these would be most helpful if they explored resolution issues, as well as physical issues such as the role of surface fluxes, or moisture jumps across the inversion. Because we realize that the computations can be rather intensive and beyond the capacity of some groups we ask that all groups perform at least the first computation, and preferably the first three itemized above.