MODELING CO2 CLOUDS CONVECTION IN THE MARTIAN POLAR NIGHTS
WITH LARGE-EDDY SIMULATIONS.
V. Caillé, LATMOS, Sorbonne Université, UVSQ Paris-Saclay, CNRS, Paris, France (vincent.caille@latmos.ipsl.fr),
A. Spiga, LMD/IPSL, Sorbonne Université, Paris, France - Institut Universitaire de France, France, A. Määttänen,
LATMOS, Sorbonne Université, UVSQ Paris-Saclay, CNRS, Paris, France, C. Mathé, LATMOS, Sorbonne Université,
UVSQ Paris-Saclay, CNRS, Paris, France, L. Falletti, LATMOS, Sorbonne Université, UVSQ Paris-Saclay, CNRS,
Paris, France.

Introduction
CO2 is the main component of the Martian atmosphere.
It condenses both at the surface as the polar caps and
in the atmosphere as clouds. Clouds play a crucial role
in the CO2 cycle at different spatial scales, from nucleation to large clouds structure, but their modeling were
restrained because of the lack of observations. However, around twenty years ago, the Mars Orbiter Laser
Altimeter was the first instrument to observe CO2 ice
clouds and snowfall [Neumann et al., 2003, Ivanov and
Muhleman, 2001] and gave, alongside Mars Climate
Sounder observations, the only properties we have about
polar CO2 ice clouds. [Colaprete and Toon, 2002] were
able to determine CO2 ice clouds formation processes
and required atmospheric conditions but did not study
clouds dynamics. Both the scarcity of condensation nuclei in the mesosphere and the strong coupling between
convective clouds dynamics and latent heat release led
to rethink microphysical theories to match the observations. In particular, CO2 theories should be adapted to
the Martian atmosphere. Modeling convective clouds
also imply that we need to use a very high resolution
model due to the scale at which some phenomenon occur. It is the first time that the impact of latent heat release on polar clouds dynamics through CO2 convection
is studied with a modeling technique where convecting
motions are resolved.

microphysics, developed at LATMOS [Listowski et al.,
2012, Listowski et al., 2013] has been coupled to the
MGCM (Määttänen & al., 2022 ; see abstract Mathé &
al. in this conference). This notably includes nucleation of CO2 on both water ice nuclei and dust particles,
accounting for sedimentation and atmospheric conditions being impacted by condensation/evaporation. We
adapted this coupling for the LES mode by adapting
mass and available CO2 calculation formulas and the
required list of tracers transported by the resolved convective dynamics.
Model is here thought to run idealized simulations.
The horizontal domain is set up as a 60x60 grid with
a 50m step in both directions. Top of the model is
8000m above ground with 60 vertical layers. We run
the model on a typical polar night initial profile obtained
from the MGCM, thus a near-saturation initial temperature profile. Thermal inertia and albedo used for surface
temperature calculations are respectively fixed at 260
J.m−2 .K−1 .s−1/2 and 0.26. A thermal perturbation is
added in the atmosphere, around 2000m, to eventually
trigger CO2 condensation. We refer to our perturbation
with the chosen delta of potential temperature, at the
center of the perturbation. Potential temperature gradually decreases from the center in every directions. First
simulations are made with no environment horizontal
winds.

Simulations
Model
We use a model made of physical parameterizations
from the Mars Global Climate Model (MGCM) developed at the Laboratoire de Météorologie Dynamique
(LMD) coupled to a dynamical core from the Weather
Research and Forecast (WRF) hydrodynamical solver
[Skamarock and Klemp, 2008]. Domain size, spatial
and temporal resolutions are fixed to run WRF in idealized Large-Eddy Simulations (LES) mode [Spiga and
Forget, 2009, Spiga et al., 2017]. Principle of LES is to
solve the Navier-Stokes equation at a high enough resolution filtering the smallest length scales to resolve the
largest turbulent eddies, and thus the latent heat release.
Initial profile is a flat plain with periodic borders along
the two horizontal axis. A new model of CO2 ice clouds

Our reference run uses a perturbation of -2K (fig 1),
what seems to be realistic in agreement with [Hu et al.,
2012] temperature profiles in the polar night and could
potentially be caused by gravity waves. We will show
that this perturbation is enough to form a CO2 ice cloud.
Latent heat released by the phase change end up being
important enough to warm up the atmosphere and trigger
a convective upward movement (fig 3). “Strength ” of
the convection is quantified by analysing the associated
vertical winds (fig 4). Because the condensation is more
important at the center of the perturbation, the cloud is
quickly distorted (fig 2). CO2 precipitation can also be
observed in our simulation.
We then study the impact of certain parameters on
CO2 condensation and cloud convection. A colder ini-

Modeling CO2 Clouds Convection in the Martian Polar Nights with Large-Eddy Simulations
tial perturbation leads to more CO2 condensing into the
cloud. The cloud seems to be heavier and slightly move
downward before the latent heat release is strong enough
to trigger the upward convection. Convection does not
seem to be stronger for a colder perturbation, ie the cloud
does not reach a higher altitude. However, temperature
of the perturbation seems to play an important role on
the timescale, convection starting faster for a colder perturbation. Perturbation temperature is then pushed to
extreme and unrealistic values to reveal eventual model
numerical limits. We also study the impact of modifying
the initial dust profile or activating the horizontal winds.
Acknowledgements
We thank the Agence National de la Recherche for
funding (projet MECCOM, ANR-18-CE31-0013). This
work was performed using HPC computing resources
from GENCI-CINES (Grant 2021-A0100110391), and
resources at the ESPRI mesocentre of the IPSL institute

Figure 2: CO2 ice mixing ratio profile at 500s (pressure
against east-west axis). Cloud moved upward because of the
convection, that also changed its global form by not being as
important at the center and at the borders.

Figures

Figure 3: CO2 ice mixing ratio for a -2K simulation, altitude
against time, CO2 cloud starts forming instantaneously. Values are averaged on the center of the perturbation layers. Convection and precipitation can be observed.

Figure 1: Initial potential temperature profile (pressure against
east-west axis), from a north polar night with a perturbation of
-2K around 2000m

Figure 4: Vertical wind for a -2K simulation, altitude against
time. Positive winds starting around 200s and ending around
500s are responsible for the cloud convection.

REFERENCES

References
[Colaprete and Toon, 2002] Colaprete, A. and Toon,
O. B. (2002). Carbon dioxide snow storms during the
polar night on mars. Journal of Geophysical Research:
Planets, 107(E7):5–1–5–16.
[Hu et al., 2012] Hu, R., Cahoy, K., and Zuber, M. T.
(2012). Mars atmospheric co2 condensation above
the north and south poles as revealed by radio occultation, climate sounder, and laser ranging observations.
Journal of Geophysical Research: Planets, 117(E7).
[Ivanov and Muhleman, 2001] Ivanov, A. B. and Muhleman, D. O. (2001). Cloud reflection observations:
Results from the mars orbiter laser altimeter. Icarus,
154(1):190–206.
[Listowski et al., 2012] Listowski, C., Määttänen, A.,
Montmessin, F., and LefŁvre, F. (2012). Modelling
the microphysics of Martian CO2 ice clouds. EAC
2012, European Aerosol Conference. Poster.
[Listowski et al., 2013] Listowski, C., Määttänen, A.,
Riipinen, I., Montmessin, F., and LefŁvre, F. (2013).
Near-pure vapor condensation in the Martian atmo-

sphere: CO<sub>2</sub> ice crystal growth. Journal
of Geophysical Research. Planets, 118(10):2153–2171.
[Neumann et al., 2003] Neumann, G. A., Smith, D. E.,
and Zuber, M. T. (2003). Two Mars years of clouds
detected by the Mars Orbiter Laser Altimeter. Journal
of Geophysical Research (Planets), 108(E4):5023.
[Skamarock and Klemp, 2008] Skamarock, W. C. and
Klemp, J. B. (2008). A time-split nonhydrostatic
atmospheric model for weather research and forecasting applications. Journal of Computational Physics,
227(7):3465–3485. Predicting weather, climate and
extreme events.
[Spiga and Forget, 2009] Spiga, A. and Forget, F.
(2009). A new model to simulate the martian
mesoscale and microscale atmospheric circulation:
Validation and first results. Journal of Geophysical Research: Planets, 114(E2).
[Spiga et al., 2017] Spiga, A., Hinson, D. P.,
Madeleine, J.-B., Navarro, T., Millour, E., Forget, F.,
and Montmessin, F. (2017). Snow precipitation on
Mars driven by cloud-induced night-time convection.
Nature Geoscience, 10(9):652 – 657.