CERFACS in Activity 2

1
CERFACS in Activity 2

• Ronan Paugam
• CERFACS Team AE
• Toulouse Feb. 6-9 2006

2
Plan

• Introduction
• Meso-scale simulations of aircraft plumes
• Small scale simulations of ship plumes
• Integration of EEI in large scale models
• Conclusion

3
INTRODUCTION
Details on the engagements of CERFACS in Activity
2

• Meso-scale simulations of aircraft plumes - WP
  2.2.2
• - Include results from small scale simulation
  (ONERA)
• - Model microphysics of aerosols
• - Run meso-scal simulations with heterogeneous
  processes
• - Results in terms of EEI
• Small scale simulations of ship plumes - WP 2.2.2
• - develop dynamical plume model
• - provide tracks particles to IVL

4
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
Evolution of an aircraft contrails
Description of contrail evolution in 4 regimes1
Jet 1-20 s.
Vortex 20 - 100 s.
Dissipation 100 1000 s.
Diffusion up to some hours
Robins (98)
Paoli (05)
Holzapfel (00)
Gerz (99)
Durbeck (98)
Typical meso scale simulation
Buoyancy effect controls dynamics
Vortices propagates downward by mutual induction
Exhaust material is entrained in counter
rotating vortices
1 Contour plot of axial vorticity, except for
dissipation regime where it a passive scalar.
5
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
Evolution of an aircraft contrails
What we want to simulate ?
6
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
Evolution of an aircraft contrails
To show you where we work
7
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
The MesoNH code
Quick details on MesoNH
MesoNH is a non-hydrostatic mesoscale atmospheric
model, developped by the Laboratoire dAerologie
(LA) and the Centre Nationale de Recherche en
Meteorologie (CNRM) - horizontal
resolution mesh ranging from 1 m. to 1 km.
(possibity of 8 nested models) - basic
prognostic variables the velocity field (u,v,w)
and the potential temperature ? -
different turbulent model tuned for L.E.S. or
meso-scale simulation (TKE) -
atmospheric microphysical module (warm and cold
microphysics) - homogeneous chemical
scheme
ri, mixing ration of all water phases
Concentration of chemical species cj
8
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
Coupling between small and meso scale simulations
Coupling processes based on the LES version of
MesoNH
Meso Scale Simulations
LES Simulations
ONERA

1 km
Small scale simulations
400 m
Start meso scale simulation after dx gt 1km
1 km
410 m
dx 1m
To be continued N times
256 m
dx 10 m
Interpolation filtering
Interpolation filtering
5 km
distributions (vapor, ice, soot
particules) velocity field (counter rotating
vortices) energy spectrum (jet turbulence) chemica
l species
prognostic variable (u,v,w,?) mixing ratio of ?
water phases chemical species turbulent kinetic
energy
1000 s
100 s
20 s
200 s
1h
Vortex
Diffusion
Dissipation
Jet
9
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
First simulations
2D simulations of the vortex regime
Physical Domain is quasi 2D, - grid point 5
256 465 - (flight direction span wise
vertical) - resolution 8 1 1 m. model is
purely dynamical - no microphysics - no
chemistry Initial condition case of a
B747 2 counter rotating vortices (Lamb - Oseen)
with - a circulation of 600 m2s - an
initial distance between vortices of b0 47
m. - no jet turbulence ambient atmosphere
corresponding of the flight altitude of a B747
with - a brunt Vaisala frequency of N 0.0108
s-1 - no shear S 0. s-1 - no atmospheric
turbulence - initial field of potential
temperature is not perturbed (acts as tracer)
10
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
First simulations
2D simulations of the vortex regime (continued)
Plot of the axial vorticity form -0.12 s-1 to
0.12 s-1 with a spacing contour of 0.015 s-1
11
MESO-SCALE SIMULATIONS OF AIRCRAFT PLUMES
First simulations
2D simulations of the vortex regime (continued)
Plot of potential temperature from 331.55 K to
333.3 K with a contour spacing of 0.06 K
12
SMALL-SCALE SIMULATIONS OF SHIP PLUMES
Provide particles tracks to IVL
Using MesoNH and results from the experimental
campaign BOMEX
Marine boundary layer
Develop a lagrangian particle tracer tuned for
ship exhaust (based on balloon tracer)
Potential temperature and Vapor mixing ratio
along the tracks
potential temperature
13
PLUME INTEGRATION IN LARGE SCALE MODELS
Petry Meijer what we suggest
Petry
Inject an initilal quantity which will be the
same after a time tref in GCM (ID) and in meso
scale simulation
GCM simulation Meso scale simulation
concentraton
EEI
EI
tref
Plume is now resolved by the GCM
Meijer
Inject a quantity which correspond to end of the
plume life
concentraton
EI
ECF
twake
End of the plume life
14
PLUME INTEGRATION IN LARGE SCALE MODELS
Petry Meijer what we suggest
What we suggest a model for a subgrid plume
Inject in a reservoir an initial quantity which
is the integration of all the quantity produced
during the life plume, and let it diffuse in GCM
variables with a suitable characteristic time.
SGP
EI
twake
End of the plume life
- No function to minimize (Petry) - A GCM cell
keeps an history of the plume - Conservative
equations can be derived for emitted and non
emitted species
15
PLUME INTEGRATION IN LARGE SCALE MODELS
sub-grid plume methods
About emitted species x (example Nox)
(1)
First approximation
meso scale simulation
Second order
rf stands for the quantity of fuel burned at
time t (parameterization of plume expansion)
To model the age of the plume
SGPx is deduced from
Note
16
PLUME INTEGRATION IN LARGE SCALE MODELS
sub-grid plume methods
About emitted species x (example Nox)
(continued)
We define
(1)
by definition as all the chemistry from the wake
is in the EEI coefficient
Then using equation of rf, this yields a closed
system for rx
()
17
PLUME INTEGRATION IN LARGE SCALE MODELS
sub-grid plume methods
About non emitted species x (example O3)
By analogy with () for emitted species, we write
rx created at time t
where
To close the system we need to estimate the last
term. We only consider here the reaction where
one of the products is emitted, for example j is
emitted and i belong to the background
We introduced Keff
Keff is deduced from
Quantity produced in the plume at meso scale
18
PLUME INTEGRATION IN LARGE SCALE MODELS
sub-grid plume methods
About non emitted species x (example O3)
(continued)
Then Keff is defined by
Finally we get a closed system
19
CONCLUSION
Things are undergoing

• - The Main achievement for this first year is the
  installation of MesoNH on our local machine
• - We just have start 3D simulations of contrails
• Particles tracks of exhaust material from ship
  engines should be soon available