A14D-08
GABLS4, an Intercomparison of Models in Extremely Stable Conditions over Antarctica: LES preliminary results F Couvreux, E Bazile, P LeMoigne, and the participants to the LES intercomparison

Monday, 14 December 2015: 17:46
3008 (Moscone West)
Fleur Couvreux Jr1, Eric Bazile2 and Patrick Le Moigne1, (1)CNRM-GAME, Toulouse Cedex 01, France, (2)Météo-France, CNRM/GMAP/PROC, Toulouse, France
Abstract:
In polar regions and under stable stratification, models present large biases that are dependent upon the parameterizations used for the surface and the boundary layer (Holtslag et al., 2013). However, no intercomparison study focuses on very stable stratification as the first three GABLS intercomparison cases dealt with moderately stable conditions. The goal of the new GABLS-4 intercomparison is to study real-time diurnal cycles over the Antarctic Plateau, focusing on the boundary layer characteristics and the coupling with the surface. The intercomparison uses observations collected at the Dome-Concordia (Dome-C) Research Station in Antarctica during the summer, and in particular the observations acquired on the 45-m tower. This site was chosen for its homogeneous surface with a low conductivity as snow and on a flat topography. Three types of numerical simulations were intercompared : 15-day offline land-snow model (LSM) runs ,30 hours runs of single column model (SCM) and 24 hours of large eddy simulations (LES). A large variability in surface fluxes was highlighted in all the types of simulations with variations around 30 W/m² during daytime and nighttime which is about 100% of the ensemble mean value. After a rapid overview of the main outcomes of the intercomparison of the two first types of simulations, we focus on the LES results. In total, 7 different LES models participated to this intercomparison. First, we show how the results are sensitive to the horizontal and vertical resolution used for this case and that horizontal resolution of 5m and vertical resolution of 2m are not sufficient to correctly reproduce the features occurring at night. Then, we characterize the turbulence separating the daytime regime and the nighttime regime. In particular, we highlight the existence of different relationship between the turbulent kinetic energy and the Richardson Number depending on the time of the day.