Formation

1
Jean-Marcel Piriou Centre National de Recherches
Météorologiques Groupe de Modélisation pour
lAssimilation et la Prévision
Update on model developments Meteo-France NWP
models
CLOUDNET Workshop / Paris 4-5 April 2005
2

• Summary
• Update on model developments
• Work done Validating models within CLOUDNET
  BLH, surface fluxes
• Ongoing work comparing radar vs SYNOP cloudiness
  scores
• Now available Model output on the new sites
• Perspectives reading the CLOUDNET database in
  Toulouse

3
Update on model developments
4
Update on model developments

• 2004-01 Sea ice masks from SSMI, relax towards
  NESDIS 0.5 SSTs, reduce snow evaporation rates,
  
• 2004-03 Use AQUA radiances in data assimilation,
  interactive mixing length,
• 2004-05 Cloudiness (more cirrus clouds, more
  cloudiness intermediate values), FMR radiation
  scheme (3h ARPEGE predictions, 1h assimilation)
• 2004-10 Use AMSU-B data, Seawind Quickscat,

5
NWP GCM Climate GCM 25-70km operations
Limited area ALADIN
Mesoscale modelling 10km operations

•  Unifying  SGS physical schemes
• Radiation
• Turbulence
• SGS convection

Cloud Resolving Model AROME
Precipitating convective clouds explicitly taken
into account 2.5km operations ? 2008
Global ARPEGE, stretched regular grids
6
Validating models within CLOUDNET
7
Validating models within CLOUDNET Anne Mathieu
Selection of days between April and August
2003 Cabauw 95 days Chilbolton 81days SIRTA 75
days Models ARPEGE IFS Met-Office model
turbulent fluxes are not available RACMO
results are strange more test are
needed Comparisons between models and
observations done on an hourly basis
8
Validating models within CLOUDNET Anne Mathieu

• Slightly better agreement than with the CLBH
  predicted
• Essentially same flaws than the predicted CLBH.

9
Validating models within CLOUDNET Anne Mathieu

• For selected days of cloudy convective boundary
  layer on the CLOUDNET stations
• Boundary layer cloud base height predicted
  within more than 300m
• 40 of the hours for IFS
• 55 of the hours for ARPEGE.
• Same behavior in the different stations.
• ARPEGE
• Under-estimation of the CLBH due to warm and
  humid biases at the surface
• Essential condition to have a good prediction of
  dry and cloudy boundary layer diurnal cycle
  right surface field prediction.
• Soil scheme
• Surface layer scheme
• Precipitations (convection)

10
Comparing radar vs SYNOP cloudiness scores
11
Comparing radar vs SYNOP cloudiness scores

• The ARPEGE (Météo-France global model) cloudiness
  scores against CLOUDNET radars improved, as the
  scores against SYNOP became less good
• The validation team has made a more extensive
  comparison CLOUDNET radars vs SYNOP total
  cloudiness
• How to compute a good model equivalent to the
  SYNOP total, low, medium and high cloudiness?
• Validating cloudiness more confident in
  radar/lidar validations than to SYNOP
  observations

12
Model output on the new sites
13
Model output on the new sites

• Since 1st september 2002 sites Chibolton,
  Cabauw, Palaiseau
• Since 16 March 2005 sites Lindenberg and
  Potenza, plus the 5 ARM sites Darwin, Manaus,
  Nauru, North Slope of Alaska, Southern Great
  Plains (10 sites daily, cron) Work done by
  François Vinit.

14
Perspectives

• Reading in 2005 the CLOUDNET 10 sites database in
  Toulouse (François Vinit).
• AROME (2.5km) model data

15

• Summary
• Update on model developments
• Work done Validating models within CLOUDNET
  BLH, surface fluxes
• Ongoing work comparing radar vs SYNOP cloudiness
  scores
• Now available Model output on the new sites
• Perspectives reading the CLOUDNET database in
  Toulouse

16
(No Transcript)
17
(No Transcript)
18
Global ARPEGE Aquaplanet mode
SCM ARPEGE (EUROCS, GATE, TOGA,BOMEX, ARM, )
Global regular ARPEGE / 4DVAR-ass. / 66 km
PHYSICS
LAM ALADIN / coupled / 10 km
Global stretched ARPEGE / 4DVAR-ass. / 20 to 200
km
19
Present operational schemes / modified in 2003 Under progress / Done in 2003
Radiation Geleyn and Hollingsworth (1979), Ritter and Geleyn (1992) More accurate infra-red exchanges between surface and layers
Cloudiness New scheme after Xu Randall 1996
Grid-scale cloud scheme Diagnostic in ql/i, all supersaturation removed, liquid/ice condensation ? T, melting/ freezing/ evaporation/ Kessler (1979), Clough and Franks (1991) Prognostic ql/i, qr/s
Subgrid-scale cloud scheme (convection) mass-flux scheme, CISK-type closure and triggering, water vapour budget using a Kuo-type closure, downdrafts, momentum flux Modified trigger functions (TKE, CIN) and cloud entrainment rates
Turbulence 1st order closure scheme after Louis (1979), Louis and al. (1981), using a flux-gradient K-theory with Ri dependency, variable roughness lengths over sea (Charnock Reduced turb. in st. cond. PrognosticTKE scheme, mixing  Betts  conservative variables thetal and qt instead of theta and qv
20
Description of the large-scale cloud and
precipitation scheme
21
Cloud scheme

• Developed by P. Lopez (QJRMS, 2002)
• Designed for variational assimilation of cloud
  and RR obs
• Prognostic var Qc (cloud condensates) Qp
  (precip water)
• Semi-lagrangian treatment of the fall of
  precipitation

(Lopez,2002)