CHALLENGES IN MARS CLIMATE MODELLING WITH THE LMD
MARS GLOBAL CLIMATE MODEL, NOW CALLED THE MARS
“PLANETARY CLIMATE MODEL” (PCM)
F. Forget, E. Millour, A. Bierjon, A. Delavois, S. Fan, L. Lange, J. Liu, C. Mathe, J. Naar, T. Pierron, R.
Vandemeulebrouck, A. Spiga, L. Montabone, Laboratoire de Météorologie Dynamique, IPSL, Paris, France
(forget@lmd.ipsl.fr), J-Y Chaufray, F. Lefèvre, A. Määttänen, F. Montmessin, L. Rossi, M. Vals Laboratoire Atmosphères, Milieux, Observations Spatiales (LATMOS), IPSL, Paris, France. F. Gonzalez-Galindo,
M.-A. Lopez-Valverde, Instituto de Astrofísica de Andalucía, Granada, Spain, M. J. Wolff Space Science
Institute (USA) & LMD, R. Young, UAE University (Al Ain, UAE), S. R. Lewis, School of Physical Sciences,
The Open University, Milton Keynes, UK, P.-L. Read, Atmospheric, Oceanic & Planetary Physics, University
of Oxford, UK, and the Mars PCM development teams.

Introduction:
The Mars atmosphere Global Climate Model
(GCM) developed at the Laboratoire de Météorologie Dynamique [1] in collaboration with several
teams around the world (LATMOS, the Instituto de
Astrofisica de Andalucia, UAE University, University of Oxford, The Open University), and with the
support of ESA and CNES is currently used for
many kinds of applications. It simulates Mars from
the subsurface to the top of the thermosphere and
includes the cycles of dust, water and CO2 that control the current Martian climate as well as a photochemical/ionospheric module.
The aim of this modeling is high: ultimately to
build a numerical simulator based only on universal
equations, yet able to consistently reproduce available observations. The goal is to create a realistic
virtual planet on which all observed phenomena and
climate-induced geological landforms arise naturally.
Like for the other similar models in the community,
this specific goal is a scientific endeavour by itself.
Such a GCM can also provide useful environmental predictions that can be used to process observations or prepare space missions. For this purpose
our teams have produced the Mars Climate Database (See Millour et al., this issue) which provides
climatologies derived from GCM simulations completed by dedicated tools. The GCM is also used to
perform meteorological data assimilation to create an
optimal description of the Martian environment
obtained by combining observation and model simulations (See e.g. Young et al., Read et al., Holmes et
al., this issue).
A new name: The Mars PCM
The different planetary version of the LMD
GCM (for Mars, Venus, Titan, Pluto, Triton, giant
planets, and a “generic” version for exoplanets and
primitive atmospheres) are more and more used and
co-developed by other teams around the world.
Within that context, we wish to use new model
names and specifically a new naming convention to
designate the LMD GCM. We now (very simply)

call it the “Planetary Climate Model (PCM)”,
preceded by the name of the planet and -if neededfollowed by the name of the dynamical core used to
solve the fluid dynamical equations to compute the
atmospheric circulation and transport. For instance :
the “Mars PCM-Dynamico”
In the case of our Mars PCM, 4 different dynamical cores are available (all using the same kind of
physical parameterizations to model radiative transfers, surface and subsurface physic, clouds, dust,
subgrid-scale dynamical processes, chemistry,
etc…):
− LMDZ: the baseline latitude-longitude grid
“dynamical core”, notably used to produce the
Mars Climate Database version 5 and 6 (See
Millour et al., this issue).
− UK-spectral : a spectral dynamical core (originally developed at the University of Reading [2]
and adapted to Mars at the University of Oxford
and the Open University) combined with a semiLagrangian advection scheme (e.g. Holmes et
al., Read et al. this issue).
− Dynamico: a new, state of the art energyconserving finite differences model on a quasiuniform icosahedral-hexagonal grid (developed
at LMD and IPSL [3]). The hexagonal mesh is
much more uniform than the regular latitudelongitude grids adopted in our previous Global
Climate Models and avoids the pole singularity
problem.
− WRF. The dynamical core of the Weather Research and Forecasting (WRF [4]) Model is used
in combination with the Mars PCM physic to
create a meso-scale and micro-scale/Large Eddy
Simulation (LES) model [5].
Current challenges.
In this talk, we will present the major challenges
and remaining enigmas that we currently face when
modeling the Martian climate. In the lower atmosphere, key challenges are
−

Dust cycle. Even in the version where dust
lifting is driven by the observed climatology,
it is not easy to predict a 4D dust distribution

that matches the available observations, and
in particular the Mars Climate Sounder observations. Creating the detached dust layers
observed throughout the year remains a challenge, in spite of our effort to parametrize
the effect of “Rocket dust sorms” and
“Mountain top flows” (See Bierjon et al.,
this issue). It is notably difficult to create detached dust layers at the observed altitudes.
−

Mean temperatures. When the dust vertical
extension seems to match the MCS observations, the mean atmospheric temperatures
appear to be overestimated when compared
to observations. Even after paying a particular attention to the dust optical properties at
various wavelengths, this discrepancy is not
easy to solve. It is possible that we are missing a physical process, or that we do not understand well the observations. The modeled
variations of mean temperature with latitude,
altitude and season can be significantly improved when taking into account and tuning
the effect of gravity waves (See Liu et al.,
this issue).

−

Thermal tides. The phase of the modeled
thermal tides seem systematically slightly
different than what is observed by MCS,
TGO/ACS/TIRVIM and EMM/EMIRS (see
Fan et al., this issue), probably for a fundamental reason that we have not yet identified. The amplitude and the phases of the
tides is nevertheless improved when taking
into account and tuning the effect of gravity
waves (Liu et al., this issue).

−

Water cycle and clouds. In the past years
we have conducted many studies to improve
our modeling of the water cycle. It is found
to be sensitive to complex couplings between microphysics, radiative transfer (heating and cooling of the ice particles and their
radiative effect), mixing by the convective
and turbulent motions, gravitational sedimentation, and scavenging by CO2 ice condensation in the polar night. These coupling
result in a significant sensitivity to the GCM
physical timestep. By exploring the sensitivity of the modeled water cycle to some of the
poorly known parameters that control this
cycle, we can set the model to simulate a
realistic water/cloud cycle. However it remains very sensitive to model parameters
(positive feedbacks) and in particular it is
difficult to simulate realistic water vapour
columns at low and mid latitudes without
predicting a northern summer pole water vapour column higher that observed (i.e. 70-80
pr-µm instead of ~50 pr-µm.

Modelling Mars Through Time.
The Mars PCM is also used to explore the possible climate on Mars in the past, in particular when
the obliquity was different (see Lange et al., Vos et
al., Aharonson et al. this issue). These developments
are notably achieved with the support of the European Research Council (ERC) within the context of
advanced grant project “Mars Through Time” [5]
For this purpose, we have included several improvements in the PCM, and in particular a new
parametrizations to model micro-climates on subgrid scale slopes (Lange et al. this issue). This allows
us to better understand the formation of glaciers
(H2O and CO2) and subsurface ice deposits on possible slopes anywhere on past or current Mars.
In addition, to simulate where the different volatile reservoirs stabilize and how they evolve over
tens of thousands of years as the orbital and spin
parameters evolve, we are developing a new kind of
model, the Mars Planetary Evolution Model
(PEM) (See Vandemeulebrouck et al., this issue). It
calls the much slower Planetary Climate Model
(PCM) regularly to update the local condensation
and sublimation rates as well as the surface temperatures, and use this information to extrapolate the
evolution of glaciers, sub-surface ice, etc. Ultimately, this will allow us to realistically simulate the
environment on Mars at low and high obliquity, and
the formation of periglacial landforms such as the
Northern Polar Layered deposits.
References
[1] Forget, F., Hourdin, F., Fournier, et al. (1999)
J. Geophys. Res., 104, 24,155-24,176.
[2] Hoskins and Simmons (1975] Q. J. R. Meteorol. Soc., 109, 771–783.
[3] Dubos, T., Dubey, S., Tort, M., Mittal, R.,
Meurdesoif, Y., Hourdin, F. (2015). Geoscientific
Model Development 8, 3131-3150.
[4] Skamarock and Klemp (2008), J. Comput.
Phys., 227, 3465–3485.
[5] Spiga, A. and Forget, F. (2009) , Journal of
Geophysical Research (Planets), vol.114, pp.E02009
Acknowledgements
This project has received funding from the European Space Agency (ESA), the Centre National
d’Etude Spatiale (CNES) and The European Research Council (ERC) under the European Union’s
Horizon 2020 research and innovation programme
(grant agreement No 835275).