Prsentation PowerPoint

1
First high resolution simulations with AROME Yann
Seity (Météo-France CNRM/GMAP)
26th EWGLAM / 11th SRNWP meeting Oslo, 4 - 7
October 2004
2
Introduction (1/1)
3
Plan

• Presentation of AROME
• Squall line cases (academic and real)
• Flood case GARD 09-08-2002
• Conclusion

4
Presentation of AROME (1/6)
Prototype built in 2004 For an operational
use in 2008

Dynamics ALADIN NH 2TL and 3TL SL schemes

P/C scheme
5
The AROME Physics (2/6)
Microphysics ICE3 sophisticated
vapour/cloud/rain/ice/graupel/snow scheme
6
The AROME Physics (3/6)
Microphysics ICE3 sophisticated
vapour/cloud/rain/ice/graupel/snow
scheme Radiation operational ECMWF (SW
Fouquart-Morcrette and LW RRTM) Turbulence at
first step, 1D version of the 3D MesoNH scheme.
Prognostic TKE, Bougeault-Lacarrère (1989)
closure condition.
7
The AROME Physics (4/6)
Surface town TEB (Masson, 2000)
nature improved ISBA (Noilhan and Planton,
1989) sea and water Charnocks
formulation with constant SST
8
The AROME Physics (5/6)
Surface characteristics (TEB, ISBA) from
Ecoclimap classification (Masson, 2003)
242 types of cover cover fraction at 1km
resolution
9
The AROME Physics (6/6)
Microphysics ICE3 sophisticated
vapour/cloud/rain/ice/graupel/snow
scheme Radiation operational ECMWF (SW
Fouquart-Morcrette and LW RRTM) Turbulence at
first step, 1D version of the 3D MesoNH scheme.
Prognostic TKE, Bougeault-Lacarrère (1989)
closure condition. Surface Externalized
schemes (town, nature, sea, water)
Documentation http//www.aero.obs-mip.fr/mesonh
/
10
2D Academic Squall Line (1/3)

• Idealised squall line
• initiated by a cold pool produced by a 0.01 K.s-1
  cooling rate applied for 10 min at low level
• Wind, temperature and humidity profiles derived
  from a COPT81 RS.
• 2.5 km resolution, 8 hours run, 180 points
• Tubulence Microphysics

11
2D Academic Squall Line
(2/3)
Theta-theta0
Vertical velocity
Relative wind
Cloud
humidity
Sum hydro
TKE
12
2D Academic Squall Line
(3/3)
AROME(7.5s) (8h) AROME (30s)
13
Real squall line case (1/4)
Ile de France case (08-04-94)
Simulation parameters Size 144x144 points Full
Physics Radiation called every 15 Begin at 15
TU, end at 18 TU Time step 15s for AROME, 5s for
MésoNH Goal Check that AROME is able to
simulate a squall line as realistic as MesoNH.
14
Real squall line case (2/4)
15
Real squall line case (3/4)
Ile de France case at 16TU
MésoNH
AROME
16
Real squall line case (4/4)
Ile de France case at 16TU radar reflectivity
MésoNH (800m height)
AROME (800m height)
Trappes Radar
17
Gard real case (1/6)
GARD flood 09-08-2002
700 mm in 24h 23 deaths, 1.2 10 9 euros
damage Stationary MCS
18
Gard real case (2/6)
GARD flood 09-08-2002
Simulation parameters Size 192x192 points Full
Physics Radiation called every 15 Begin at 12TU
Sept. 8th, end Sept. 9th 00TU Focus on first
stage of the system 12TU-22TU Goal As good as
referenced mesoNH with lower cost
19
Gard real case (3/6)

20
Gard real case (4/6)
12-22 TU Nîmes radar cumulated rainfall
Arome 15s
Arome 60s
MésoNH 4s
319 mm
304 mm
274 mm
gt 300 mm
BC Aladin 3h Forecasts IC Mesoscale surface
data reanalysis microphysics adjustment
21
Cost
Gard real case (5/6)

• MésoNH
• TSTEP 4s , CPU 24h20
• AROME
• TSTEP15s, CPU 9h
• TSTEP45s, CPU 3h23
• TSTEP60s, CPU 2h30

22
Gard real case (6/6)
Adjustment
Interface
Radiation
Turbulence
Dynamics
Microphysics
Surface
Physics (other)
23
Conclusion (1/2)

• First comparative tests between mesoNH and AROME
  in 2D academic or 3D real cases show equivalent
  quality results

• AROME 10 less expensive than MésoNH,
  3.4 more than ALADIN oper

24
Conclusion (2/2)

• Diagnostics

• Other cases (fog, MAP )

• Code optimisation (Benchmark)

• Work on coupling

• AROME with ALADIN 3D VAR

25
First high resolution simulations with AROME Yann
Seity (Météo-France CNRM/GMAP)
26th EWGLAM / 11th SRNWP meeting Oslo, 4 - 7
October 2004