| Home | Case description | Participants | Model details | LEM data |
Results
GABLS LES Intercomparison
Model Details
A summary of the subgrid models and the advection schemes used by each participant is given in the table below. Where relevant, values of the Smagorinsky parameter, Cs, the ratio of basic mixing length to horizontal grid length, are given. Further discussion is given in the intercomparison paper Beare et al.
| Participants | Subgrid Models | Advection Scheme | |
| scalars | momentum | ||
| MO | Smagorinsky with backscatter, Cs=0.23 high res (6.25m and below), Cs=0.15 low res. | TVD / ULTIMATE | Piacsek-Williams central difference scheme |
| CSU | Deardorff (1980), Cs=0.19 | 3-d monotonic and positive definite (Smolarkiewic-Grabowski) | central difference in flux form with 3rd order Adams-Bashforth in time |
| IMUK | Deardorff (1980), Cs=0.1 | Piacsek-Williams central difference scheme | |
| LLNL | Nonlinear Deardorff, Cs=0.2 | Pseudospectral method (Fox and Orszag, 1973) in the horizontal, second order central differencing in the vertical | |
| NERSC | Dynamic mixed subgrid closure, Vreman et al. (1994) Cs=0.14 (average) | 2nd order central differencing with advection term written in divergence form (non-monotonic) | 2nd order central differencing with advection term written in skew-symmetric form |
| WVU | ARAP second order quasi-equilibrium closure model Sykes and Henn (1989) | piecewise parabolic method (monotone) | 2nd order central differencing and leapfrog in time |
| NCAR | Sullivan, McWilliams and Moeng (1994), Cs=0.18 | 3d monotone scheme Koren (1993) "kappa=1/3 scheme" | Pseudospectral method (Fox and Orszag, 1973) in the horizontal, second order central differencing in the vertical |
| UIB | Two part Meso-NH scheme (Cuxart et al., 2000) | positive definite second order central difference scheme | second order central difference scheme |
| CORA | Dynamic Smagorinsky, Cs=0.18 | Pseudospectral method (Fox and Orszag, 1973) in the horizontal, second order central differencing in the vertical | |
| WU | Deardorff (1980), with changes to dissipation formulation. | Piacsek-Williams central difference scheme | |
| NRL | Smagorinsky, Cs=0.17. | Bott (1989), with second order polynomial | Second order central difference with leapfrog in time |
Subgrid Comparisons
Three participants also provided data for different subgrid model configurations, as summarised in the table below. The labels correspond to the legends used on the plots viewed by following the links on the left.
| Participants | Resolution | Subgrid Models | Label |
| MO | 6.25 m | Smagorinsky with backscatter, Cs=0.15 | BKSCT, Cs=0.15 |
| Smagorinsky with backscatter, Cs=0.23 | BKSCT, Cs=0.23 | ||
| Smagorinsky, Cs=0.15 | SMAG | ||
| 1m | Smagorinsky with backscatter, Cs=0.23 | BKSCT | |
| Smagorinsky, Cs=0.23 | SMAG | ||
| LLNL | 6.25 m and 12.5 m | Nonlinear Deardorff, Cs=0.2 | NLD |
| Nonlinear Smagorinsky, Cs=0.2 | NLSMAG | ||
| Smagorinsky, Cs=0.2 | SMAG | ||
| NERSC | 6.25 m | Dynamic mixed subgrid closure, Cs=0.14 (average) | DMS |
| Smagorinsky, Cs=0.1 | SM, Cs=0.1 | ||
| Smagorinsky, Cs=0.14 | SM, Cs=0.14 | ||
| Smagorinsky, Cs=0.23 | SM, Cs=0.23 |