Martian ice cloud formation in a GCMdriven detailed microphysical model

1
Martian ice cloud formation in a GCM-driven
detailed microphysical model

• Frank Daerden (_at_aeronomie.be)
• C. Verhoeven, D. Moreau
• N. Mateshvili, D. Fussen
• Belgian Institute for Space Aeronomy BIRA-IASB
  (Belgium)
• A. Akingunola, J.C. McConnell, J.W. Kaminski
• York University (Canada)
• N. Larsen
• Danish Meteorological Institute DMI (Denmark)

2
1. Introduction

• Model heritage PSCBox
• detailed microphysical box model for terrestrial
  PSC and cirrus Larsen, 2000
• Size-bin resolved, Eulerian
• Explicit description of homogeneous and
  heterogeneous nucleation condensation/sublimation
  sedimentation optical calculations gas-phase
  mass balances heterogeneous chemistry
• Fully online coupled in 3D-CTM Daerden et al,
  2007 driven by external fields (ECMWF, GCM,
  ...)

3
2. Model description

• Adaptation to Mars MARSBOX
• 2 particle types Dust and Ice
• Dust particles serve as cloud condensation nuclei
• Dust generic, homogeneous composition
  (?2.5.103kg/m3)
• Ice particle hexagonal cilinders with aspect
  ratio 2
• 60 size bins, rmin 10-3µm, rmax103µm

4
2. Model description (ctd)

• Dust initialization
• Vertical eddy diffusion Krasnopolsky, 1993
• No dust loss at boundaries
• Steady-state assumption size-dependent Conrath
  formula
• where
• q(z,r)part/kgair at altitude z for particle
  size r
• ?0/40 terminal fall velocity at altitude 0/40km
• H atmospheric scale height (H10km)
• To be solved sequentially from the surface upwards

5
2. Model description (ctd)

• q(z0,r) is assumed lognormal with parameters
• ? 0.4 (various sources)
• rm 0.2µm (gives ?0(r)0.03 - moderate dust
  load)
• Nt 30/cm3 (gives ?0.2 - consistent with
  GCM)
• Steady state verification 20 sol simulation, no
  ice
• Consistency with SPICAM (rm Rannou et al 2006,
  ext Montmessin et al 2006)

6
3. Simulations and data

• GCM driving GM3
• Global Mars Multiscale Model Moudden
  McConnell, 2005
• Box model stacked on GM3 vertical levels
  (?z1.2km)
• ?t1/24 sol

H2Odriving H2OGM3,g H2OGM3,i - H2Obox,i
3D information into box model
7
3. Simulations and data (ctd)

• Observations SPICAM/Mars Express, UV nadir mode
  Mateshvili et al, 2007

8
3. Simulations and data (ctd)

• Simulations set-up
• Focus on Tropical Cloud Belt season during MY27
• 1 year of offline GM3 simulations with low dust
  load (?0.2)
• In GM3 grid cells with high density of SPICAM
  data
• Average diurnal cycle in bins of 10 Ls (20
  sols)
• Sensitivity w.r.t. T and water budget

9
4. Results

• Differences GM3/bulk ? MARSBOX/microphysics
• Nucleation introduces delay in cloud formation
• Cloud formation is limited by
• nr. of available dust condensation nuclei
• Supersaturation ratio
• ? bulk scheme upper limit for cloud formation

10
4. Results (ctd)

• Daily cycles some selected cases (Ls110-120)
• 70E (Syrtis Major)
• 62W (Xanthe)
• 169E (Elysium)

11
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

12
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

13
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

14
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

15
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

16
4. Results (ctd)

• Case 1 Syrtis Major 67.5E

17
4. Results (ctd)

• Case 2 Xanthe

18
4. Results (ctd)

• Case 2 Xanthe

19
4. Results (ctd)

• Case 2 Xanthe

20
4. Results (ctd)

• Case 3 Elysium

21
4. Results (ctd)

• Case 3 Elysium

22
4. Results (ctd)
23
5. Conclusions

• Cloud formation is highly sensitive to
  temperature and water vapor
• Within small variations of GCM temperature and
  water budget, MARSBOX driven by GM3 is in many
  cases able to reproduce the SPICAM cloud optical
  thickness in the TCB
• Bulk scheme overestimates the water content in
  clouds