Mesoscale simulations above
Antarctica:
How to choose an optimal configuration of the
meteorological model to characterize turbulence
above a potential astronomical site ?
Lascaux Franck, Masciadri Elena, Hagelin Susanna, Stoesz Jeff
Meso-NH: a mesoscale non-hydrostatic
meteorological model (1)
http://mesonh.aero.obs-mip.fr/mesonh/
-Research model developed jointly by Météo-France and Laboratoire d’Aérologie in
Toulouse, France.
-System of equations based upon the Lipps and Hemler (1982) anelastic system (filtering
of acoustic waves, time discretization entirely explicit).
-Vertical following-terrain coordinates (Gal Chen and Sommerville, 1975).
-Use of a 2-way interactive grid-nesting technique (Stein et al., 2000) with increasing
horizontal mesh-sizes.
-Prognostic dynamic variables of the model :
u, v, w, è, q v , TKE and rc , rr , ri , rs , rg , rh (optional)
Meso-NH: a mesoscale non-hydrostatic
meteorological model (2)
http://mesonh.aero.obs-mip.fr/mesonh/
-Includes different standard physical packages:
Turbulence scheme: 1.5 closure scheme with different mixing lengths (Cuxard
et al. 2000).
Convection scheme: Kain-Fritsch-Bechtold scheme (Bechtold et al., 2001).
Externalized surface model: ISBA (Interactions between Soil Biosphere and
Atmosphere) scheme for vegetation (Noilhan and Planton, 1989), TEB scheme for urban
area.
Radiation scheme: ECMWF radiation scheme (Morcrette, 1991).
Microphysical scheme: different cloud schemes (the most used is a 1-
moment scheme with 5 or 6 prognostic variables - Pinty et al. 1998, Lascaux et al. 2006).
Arcetri):
-Plus (only at the Osservatorio Astrofisico di Arcetri):
Optical Turbulence: prediction of Cn², and associated parameters (seeing ε,
wavefront coherence time τAO, spatial coherence outer scale ℒO, isoplanatic angle ΘAO)
-Goal: running Meso-NH on a local workstation -> we have to test the best configuration.
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
Km DEC05 JAN06 FEB06
20 SUMMER
15
10
5
0 10 20 30 m/s -2
0 10 20 30 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of wind velocity above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
m DEC05 JAN06 FEB06
130
90
50
0 SUMMER
0 4 820 12 30 m/s -2
0 10 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of wind velocity above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
Km DEC05 JAN06 FEB06
SUMMER
20
15
10
5
20 250 K
210 10 230
0 30 -2 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of temperature above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
m DEC05 FEB06
JAN06
130
SUMMER
90
50
0
230 10 240
0 20 250 K -2
30 -1 0 1 2 -2 -1 0 1 2 -1 0 1 2 3 4 5
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of temperature above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
Km JUN06 JUL06 AUG06
20 WINTER
15
10
5
0 10 20 30 m/s -2
0 10 20 30 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of wind velocity above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
m JUN06 JUL06 AUG06
130
WINTER
90
50
0
0 4 820 12 30 m/s -2
0 10 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of wind velocity above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
Km JUN06 JUL06 AUG06
WINTER
20
15
10
5
20 250 K
210 10 230
0 30 -2 0 2 -2 0 2 -2 0 2
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of temperature above Dome C
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
m JUN06 JUL06 AUG06
130 WINTER
90
50
0
20 250 K -2
230 10 240
0 30 0 2 4 -2 0 2 4 -4 0 4
ECMWF and Differences: ECMWF - RADIOSOUNDINGS
RADIOSOUNDINGS
Mean vertical profiles of temperature above Dome C
Mesoscale simulations – Initializations (2)
-Initialization and forcing:
-> Analyses from the ECMWF. Choice of a winter day, 07/22/2005.
20 20
km km
15 15
10 10
5 5
0 0
0 10 20 30 40 m/s 0 20 40 60 80 °C
Wind vertical velocity profile above Temperature profile above Dome C
Dome C at 12:00 UTC on 07/22/2005: at 12:00 UTC on 07/22/2005:
-> from the ECMWF (blue); -> from the ECMWF (red);
-> from radiosounding (black). -> from radiosounding (black).
Mesoscale simulations – Configurations
Test of grid-nesting:
->2 configurations:
*One in grid-nesting 2 models;
*One in 1 model.
Test of surface scheme:
Horizontal resolution: (not shown here – work in progress)
->Different configurations between:
*100x100 km over Antarctica;
*1kmx1km over a chosen site (Dome C,
South Pole, Dome A…).
Vertical resolution:
->2 configurations:
*5 levels in the 1st 100 m, starting at
10 m, highest level at ~20 km;
*15 levels in the 1st 100 m, starting at
2 m, highest level at ~20 km.
Antarctica, as seen by Meso-NH
-Orographic file: gtopo30, resolution ~1 km. Δx = 1 Km
-Polar Stereographic Projection.
DOME C
high and Concordia Station
ANTARCTICA – low resolution low
resolution Summit
ΔH = 20 m
6000 Km ΔL = 60 Km
Δx = 100 Km
400 Km
Δx = 100 Km
Influence of Horizontal Resolution
-Same vertical resolution, different horizontal resolution:
->Dome C1: whole Antarctica, Δx = 100 Km, 6000 x 6000 km domain.
->Dome C2: centered above dome C, Δx = 1 Km, 400 km x 400 km domain
WIND SPEED WIND DIRECTION TEMPERATURE
20 20 20
km ECMWF km km
15 Dome C1 15 15
Dome C2
10 10 10
5 5 5
m/s °C
0 5 10 15 20 25 150° 200° 250° 300° -100 -80 -60 -40 -20
4 4 4
km km km
m/s °C
3 3 160° 3-70
0 5 10 15 20 180° 200° 220° -60 -50 -40 -30
Vertical Profiles at Dome C, To + 12H
Influence of Vertical Resolution
-Same horizontal resolution, different vertical resolution:
->Dome C3: 50 levels, 5 levels in the 1st 100 m, starting at 10 m, highest level
at ~20 km;
->Dome C4: 60 levels, 15 levels in the 1st 100 m, starting at 2 m, highest level
at ~20 km.
WIND SPEED WIND DIRECTION TEMPERATURE Configuration:
20 20 20 High horizontal
km
ECMWF resolution,
15 Dome C3 15 15
Dome C4
Δx = 1 km
10 10 10
5 5 5
m/s °C
0 5 10 15 20 25 150° 200° 250° 300° -100 -80 -60 -40 -20
4 4 4
km km km
150 km
3 m/s 3 3 °C
0 5 10 15 20 160° 180° 200° 220° -70 -60 -50 -40 -30
Vertical Profiles at Dome C, To + 12H
Simulations with or without grid-nesting (1)
324O m
10 km 2.5 km
Simulation “Dome C5”:
- Only Domain B.
Simulation “Dome C6”:
- “father” model : A;
- “son” model : B;
- 2-way interactive
grid-nesting technique.
A B Initialization:
300 km
2940 m - ECMWF analysis
1000 km
interpolated on the
“father” model (A for
DomeC6, B for
Same vertical resolution, 50 levels from 10 meters to DomeC5);
20 km. - 07/22/2005.
Influence on wind ? temperature ? on domain B (common to both simulations).
Simulations with or without grid-nesting (2)
WIND SPEED WIND DIRECTION TEMPERATURE
20 20 20
km km km
15 15 15
ECMWF
10 -6 0 6 Dome C5 10 10
-6 0 6
Dome C6
5 5 5
m/s -20 0 20
°C
0 5 10 15 20 25 150° 200° 250° 300° -100 -80 -60 -40 -20
4 4 4
km km km
-6 0 6 -20 0 20
-6 0 6
3 m/s 3 3 °C
0 5 10 15 20 160° 180° 200° 220° -70 -60 -50 -40 -30
Vertical Profiles at Dome C, To + 12H
Conclusions
-Extended comparison of ECMWF analyses with radiosoundings
-Different numerical simulations were performed above Antarctica, with Meso-NH.
-Test of different configurations:
-> low horizontal resolution (Δx=100 Km) / high horizontal resolution (Δx=1 Km).
-> low vertical resolution (1st level @ 10 m) / high vertical resolution
Next steps:
(1st level @ 2 m)
->improving the surface scheme for Antarctica needs;
-> no grid-nesting / grid-nesting.
-Comparisons made looking at 3 vertical profiles at Dome C: Wind Speed, Wind Direction,
Temperature. ->testing other days, in winter and in summer, with
different characteristics of wind and temperature, to
HVR GN
HHR generalize the conclusions;
T + + =/-
->using Meso-NH with the optical turbulence !
WS + = +
WD +/- = +
High resolution simulation with grid-nesting
12 hours simulation, from 00 UTC to 12 UTC on 6 September 2005.
Numerical setup:
5 km
1 km
25 km
High resolution simulation with grid-nesting
12 hours simulation, from 00 UTC to 12 UTC on 6 September 2005.
Initialization: Results:
Temperature
Wind Speed
Wind
Speed
(m/s)
DOME C
0 25 50
Temperature
(K)
190 240
Mesoscale simulations – Initialization (1)
-Initialization and forcing:
ECMWF (European Center for Medium-range Weather Forecasts) analyses.
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
20 15
SUMMER DJ WINTER JJA
2005/2006 2005
15
10
10
5
5 50 days
47 days
km km
m/s 0 10 20 30 0 10 20 30
Mean vertical profiles of wind velocity above Dome C: in blue, from Radiosoundings; in black,
from the ECMWF; in red the difference.
Mesoscale simulations – Initialization (2)
-Comparisons between Vertical Profiles from ECMWF analyses and Radiosoundings (from
Antartide, www.climantartide.it):
PNRA – Programma Nazionale di Ricerche in Antartide, www.climantartide.it):
15 15
SUMMER DJ WINTER JJA
2005/2006 2005
10 10
5 5
50 days 47 days
km
km
° -40 -20 0 20 0 100 200 300
Validation of ECMWF analyses above Dome C !
15 15
SUMMER DJ WINTER JJA
2005/2006 2005
10 10
5 5
50 days 47 days
km
km
° 0 100 200 300 0 100 200 300
Mean vertical profiles of temperature and wind direction above Dome C: in blue, from
Radiosoundings; in black, from the ECMWF; in red the difference.