Working days on the "EFB" closure, Toulouse, 18-25 March 2013
Context
During the workshop on “Parameterization of stable boundary layer in numerical weather prediction models†organized in Finnish Meteorological Institute (FMI), Helsinki, 3-5 December 2012, (http://netfam.fmi.fi/Stable12/), it was decided to organize working days to identify and solve main issues for implementing “Energy- & flux-budget (EFB) closure†in AROME/HARMONIE turbulence schemes.
Participants (with affiliation, participation days)
- Sergej Zilitinkevich (FMI, Nizhni Novgorod State University (NNSU), 18/3 -> 25/3)
- Carl Fortelius (FMI, 18 -> 25)
- Evgeny Kadantsev (FMI, NNSU, 18 -> 22)
- Andrey Glazunov (Institute for Numerical Mathematics Moscow State Univ., NNSU, 18 -> 25)
- Igor Rogachevskii (Ben-Gurion University (BGU) of the Negev Israel, NNSU, 18 -> 25)
- Nathan Kleeorin (BGU, NNSU, 19 -> 25)
- Elena Masciadri (INAF-OAA, 22 & 23 in visio-conference)
- Ivan Bastak Duran (Czech Hydrometeorological Institute CHMI) (22 & 25)
- Jean-François Geleyn (Météo-France/CNRM-GAME) (22 & 25)
- Alexandre Paci (Météo-France/CNRM-GAME) (18, 19, 21, 25)
- Pascal Marquet (Météo-France) (22 & 25)
- Valery Masson (Météo-France/CNRM-GAME) (18 -> 25)
- Yves Bouteloup (Météo-France/CNRM-GAME) (18 -> 25)
- Eric Bazile (Météo-France/CNRM-GAME) (18 -> 25)
- François Bouyssel (Météo-France/CNRM-GAME) (18 -> 25)
Presentations
- Overview of NWP systems used in ALADIN/HIRLAM (F. Bouyssel)
- The physics in HAAA galaxy and 1D model MUSC (E. Bazile)
- Eddy-Diffusivity/Mass Flux parameterization in AROME and ARPEGE (Y. Bouteloup)
- Turbulence closure for stably stratified geophysical flows (S. Zilitinkevich)
- 3-fold decomposition EFB closure for convective turbulence and organized structures (I. Rogachevskii)
- LES with localized mixed dynamic subgrid/subfilter closure. Stably stratified Couette flows and turbulent convection (A. Glazunov)
- Turbulent energies in the CBR scheme in a stable case. Preliminary findings (C. Fortelius)
- Modifications of AROME turbulence scheme to fit the TPE approach (V. Masson)
- EFB in ARPEGE/AROME: some practical aspects (E. Bazile)
- 3-fold decomposition EFB closure for stably stratified turbulence and large-scale inertial waves (N. Kleeorin)
- Turbulent dissipation time and length scales (S. Zilitinkevich)
- Turbulent fluxes of buoyant and passive scalars (I. Rogachevskii)
- Short presentation on the CNRM-GAME water flume (A. Paci)
- Presentation on thermodynamics and TOUCANS turbulence scheme (J.F. Geleyn): (1) COSTWG12, (2) COST2013, (3):
TOUCANS’ digest - On the use of moist entropy in moist turbulence (P. Marquet) : (part 1), (part 2)
Summary of developments and discussions
- Several technical issues have been solved during the working days in the 1D model MUSC to simulate GABLS1 case with cycle CY38T1 and to use new GFL for additional turbulence prognostic variables (turbulent potential energy, etc.).
- A comparison of EFB closure (Zilitinkevich et al., 2013) and Valery’s coding proposition for Arome turbulence scheme has been performed. This analysis has identified some differences, in particular in the parameterization of pressure correlation terms.
- A preliminary implementation of EFB type closure has been performed in AROME turbulence scheme.
- A preliminary implementation of EFB closure “4.2 equations†in ARPEGE turbulence scheme was performed. This code has been tested on GABLS1 case leading to an improvement on simulated wind and temperature profiles. The simulation of GABLS3 was also tested and useful to solve some problems linked with unstable conditions.
- Definition of an extension of GABLS1 case with higher Richardson number. Further collaboration: comparison of LES simulation done by Andrej and 1D simulations with Harmonie turbulence schemes including EFB closure (Eric, Carl, Valery, Ivan) http://webdav.cnrm.meteo.fr/public/proc/bouyssel/WD_EFB_201303/Glazunov_baroclic_PBL.pdf
- Investigate the use of Dome C observations (mast of 40m). Selection of one date for the development of one case study for GABLS4 model intercomparison (1D and LES modeling). Work will be done in collaboration between Eric Bazile, Olivier Traulle, Tihmo Vihma.
- A discussion was conducted on LES simulation (uncertainties, filtering, spatial resolutions, etc.). Stably-stratified plain Couette flow is considered to be one of the best simulation type to validate EFB closure. More computer time is needed to reach higher Richardson number.
- Interesting discussion on turbulent dissipation time and length scales in the free atmosphere. Further collaboration needed to compare the turbulent dissipation time scales (EFB proposition, Bougeault Lacarrere mixing length ; TOUCANS’s options for diagnostic versus prognostic). What happens in neutral stratification? How to treat the cloudy case? (This entire important topic can be treated separately from energetic part of EFB closure.)
- Visit of the water flume has been organized by A. Paci. The large size of the water flume has the potential to respect anisotropy features of stable stratified flows. Possibility to define lab experiment with simultaneously high Richardson and Reynolds numbers, therefore that can reach regimes that can not be performed with LES and DNS experiments. Field experiments can never provide fully controlled conditions and precise observations are more expensive and uncertain than in laboratory experiments. General recommendation: Need to strengthen lab experiments in meteorology. Lab experiments can be done not only in the air, but also in water and others fluids. In comparison, there are much more connections between operational oceanography and lab experiments. Further collaborations: i) carry special simulation with the current experiment configuration to produce a new dataset suitable for validating EFB closure ; ii) combine this new dataset with a previous available dataset (Richardson number up to 2) and perform data analysis for EFB closure validation ; iii) plan a future lab simulation designed in collaboration.
- Write down a proposition on eddy viscosity and eddy conductivity to take into account evaporation/condensation ; a paper could be finalized.
- Interesting discussion on modeling moist turbulence. Moist-air entropy potential temperature is the quantity to be mixed by the moist turbulence, not the Betts’s potential temperature. It was recommended to pursue by addressing sequentially four steps in the long term: 1) logarithmic rather than linear elimination of qt separately for ql and qi cases; 2) recombination of ql and qi cases at the end ; 3) comparison of classical closure with cloudiness and closure with condensation and evaporation dependency on the prognostic variables ; 4) extension of the previous topic to the partial cloudiness situation ; 5) convergence between the ways condensation/evaporation generally speaking interact with turbulence.
- Create a Web page including workshop materials and further collaborative works.