UPDATE AND STATUS OF THE MARS CLIMATE MODELING CENTER
AT NASA AMES RESEARCH CENTER
M. A. Kahre (melinda.a.kahre@nasa.gov), R. J. Wilson, A. S. Brecht, R. M. Haberle*, C.E. Harman, NASA
Ames Research Center, Moffett Field, CA, USA, R. A. Urata, A. Kling, K. E. Steakley, C. M. Batterson, Bay
Area Environmental Research Institute, Moffett Field, CA, USA, V. Hartwick, L. Gkouvelis, NASA Postdoctoral
Program, NASA Ames Research Center, Moffett Field, CA, USA, T. Bertrand, LESIA Observatoire de Paris,
MEUDON, France. *retired Jan 4, 2022.
Introduction: NASA Ames Research Center has
a long-established history in the study and numerical
modeling of Mars’ atmosphere and climate. Such
studies began with the late Jim Pollack and have continued under the leaderships of Bob Haberle, Jeff Hollingsworth, and now Melinda Kahre. Our group has
grown to more than 10 members, including civil servants, research scientists, postdocs, and students.
While we still prioritize science and the model development that supports it, we have recently enhanced
efforts to make our codes and output publicly available and boost community engagement through the
hosting of modeling tutorials, etc. Here our goal is to
present the status of the Ames Mars GCM, our ongoing science projects, the tools we have recently made
publicly available, and our plans for releases and continued community engagement.
Status of the New Ames Mars GCM: We are
retiring our older, latitude-longitude dynamical core
(our Legacy Mars GCM) and are completing the transition to the NOAA/GFDL cubed-sphere finite volume (FV3; Figure 1). This new dynamical core is
highly parallelizable and scalable, which allows for
higher resolution simulations and vastly improved
throughput, has improved conservation properties,

Figure 2: Current Mars processes diagram.

and does not have the “pole problem” that arises due
to the singularities that are produced by the converging meridians at the poles in latitude/longitude grids.
Additionally, this dynamical core has non-hydrostatic
and nesting/stretching grid options.

Figure 1: The Cubed-sphere grid.

We have completed the porting of Ames-developed Mars physics routines [1] into the new GCM to
complement the Mars physics routines that previously
existed in the NOAA/GFDL FV3-based Mars GCM,
and we have completed initial testing of this new capability by comparing the new GCM results to both
the Legacy results and to available observations. We
are now continuing model development efforts with

the new GCM that are motivated by our science, mission support, and community engagement goals.
Physics Development: We continue to make progress developing physics routines to improve our simulation of the current and past Martian climates (Figures 2 and 3). Unless specifically referred to as the
Legacy GCM, all future references to the GCM in this
abstract are to the new Ames FV3-based GCM.
Improved Aerosol Treatment. The GCM currently
has multiple options for including both the radiative
effects of aerosols and the physical processes that
govern their spatial and temporal distribution. We use
both particle bins and moments to represent dust and
water ice cloud particle size distributions.
Dust is critical to the Martian climate, so one area
of continued development has been on improving the
representation of dust in the model. Brownian and
gravitational coagulation has been included [2] and a
bi-modal distribution for dust has been implemented
(Urata et al., this meeting [3]). These improvements
have been used in investigations of global dust storms
(see below). In the lower atmosphere, collisions between gas molecules and dust grains keep them thermally equilibrated at nearly the same temperature, but
this equilibrium breaks down as low as 40 km above
the surface and their temperatures diverge [4]. We
have developed a 1-D radiative-convective model that
includes the coupling physics of dust and gas, and we
have implemented these modifications into the GCM
(Haberle et al., this meeting [5]). The model conserves total dust energy while apportioning the available radiative energy streams into gas and dust heating.
Water ice clouds also critically influence the climate system, and we have improved the representation clouds and their radiative effects in the model.

Figure 3: Early Mars processes diagram.

We have included the full, moment-based microphysics package that is described in [1] that includes water
ice cloud nucleation growth and sedimentation with
time splitting. In addition, we have the simple microphysics scheme [6, 7, 8], which allows for highly controlled cloud radiative effects.
Extended Vertical Domain. The GCM currently
contains the physics packages needed to realistically
simulate the atmosphere up to ~120 km, including
Non-LTE near-IR heating and IR cooling, orographic
and non-orographic gravity wave drag, and photochemistry. We are continuing to improve the representation of both orographic and non-orographic
gravity wave forcing by comparing the effects of
gravity waves in simulations at refined spatial resolution to the effects produced by the parameterizations
in the GCM. We are extending the model top up
higher than ~120 km by including variable specific
heat and implementing and testing parameterizations
for UV heating and thermal conduction. We are
switching from our current photochemistry scheme
(from [9]) to the Kinetic Pre-Processor (KPP) scheme
to be flexible for the science question being addressed
(e.g., upper atmosphere, Exo-Mars, and Early Mars
applications).
Early Mars Physics. Although the early Martian
climate is challenging to simulate with GCMs, new
capabilities of the FV3-based GCM will enable improved representations of the ancient hydrological cycle due to better polar region resolution with a cubed
sphere grid and parallel architecture, which allows for
longer runtimes. Our goal is to develop a comprehensive FV3-based GCM that can be used to understand
the early Martian climate. We are porting the appropriate physics packages from the Legacy GCM to the
FV3-based GCM, including an improved radiation

code that includes the effects of collision-induced absorption (for CO2, H2, and CH4), bulk H2O and CO2
cloud microphysics, and moist convection. Once we
have thoroughly tested each package in the new
GCM, we will focus on equilibrating the water cycle
(including the locations of surface water ice) and testing the sensitivity of couplings between the CO2,
H2O, and dust cycles.
Science Overview: We have many ongoing science projects related to current and past Mars, and
Mars-like Exoplanets. We give an overview here,
with a focus on projects that are presented at this
meeting.
Climate Cycles. We are currently working on several projects that focus on the dust and water cycles
on current-day Mars. We briefly describe these projects here.
Recent modeling efforts of the 2018 GDS highlight that climate models do not simultaneously capture the evolution of surface temperatures, semi-diurnal tide amplitude, and the decay rate of global column dust opacities, which suggests that significant
changes in dust particle sizes may occur during the
dust storm (e.g., [10, 11]). We show that using a selfconsistent bimodal dust lifting scheme with a minor
fraction of a small mode leads to an improvement in
areas such as the diurnal surface temperature cycle
and the semi-diurnal tide amplitude during the global
dust storm (Urata et al., this meeting [3]).
The B regional dust storm is observed to occur at
high southern latitudes near southern summer solstice. We model the B storm with the GCM at ~1°x1°
horizontal resolution (Batterson et al., this meeting
[12]). In our simulations, the B storm is made up of
a series of dust plumes that are driven by strong radiative-dynamic feedbacks between airborne dust and
incoming shortwave radiation.
Our ongoing water cycle investigations focus on
the microphysical and radiative role of clouds on current day Mars. While in [1] we found that time-splitting the microphysical processes significantly reduced the cloudiness over the cap and yielded more
realistic simulations of the water cycle, we continue
to find that cloud formation over the NPRC is extremely sensitive to small changes in the environment.
We are exploring the effects of high resolution on
cloud formation over the NPRC and elsewhere, and
we are utilizing our simple microphysics scheme in
addition to our full microphysics scheme to explore
the radiative effects of clouds in the aphelion cloud
belt and in the polar hoods.
Circulation and Dynamics. The evolving distribution of radiatively active dust and water ice clouds
plays a major role in modulating the seasonal and interannual variation in the thermal forcing of the Martian atmosphere, and thus the resulting intensity of the
circulation. Although we are interested in multiple
aspects of these feedbacks, we focus here on tides.
The global tide includes westward propagating
(sun-synchronous) waves driven in response to solar

heating, as well as nonmigrating waves that result
from zonal variations in the thermotidal forcing that
are caused by variations in the surface and the distribution of aerosols. The migrating tides are of particular interest since they tend to be directly responsive
to the aerosol distribution. There now exists an extended record of surface pressure observations at four
locations in the Martian tropics, which can be used to
put valuable constraints on the distribution and evolution of aerosols in the Martian atmosphere. We show
that changes to the strength of the cloud radiative forcing yields better agreement to various tidal components as observed in the pressure records from the
Martian surface (Wilson et al., this meeting [13]).
Middle and Upper Atmosphere Processes. The
middle and upper atmosphere are influenced by the
lower atmosphere and solar environment. Observations from the recent Mars missions are increasing our
understanding of these influences. With the newly extended MGCM, we can now examine drivers that influence the coupling between the lower and upper atmosphere in comparison to observations.
Polar warming and O2 IR nightglow emissions are
a product of the meridional circulation and are observed in similar latitude and altitude regions. However, their seasonal trends are observed to be different. We are examining the correlation between the
polar warming and O2 IR nightglow with respect to
thermal and mechanical forcing mechanisms to
achieve a deeper understanding of the meridional atmospheric transport (Brecht et al., this meeting [14]).
We are examining the behavior of the middle atmosphere nighttime thermal structure with respect to
different lower atmosphere forcings. The recent discovery of the nightside warm layer [15] has stimulated curiosity of what could be driving this unexpected heating. We are utilizing our newly extended
MGCM in conjunction with the most abundant temperature observations of the middle atmosphere (7
years of MAVEN/IUVS observations) to gain physical intuition on the nightside thermal structure variability (Gkouvelis et al., this meeting [16]).
One mechanism for energy and momentum
transport through the whole atmosphere is gravity
waves (GW). We use gravity-wave resolving GCM
simulations at high horizontal (1/4°x1/4°) resolution
with ~100 vertical layers and simulations at low horizontal (~4°) resolution with the orographic gravity
wave drag parameterization from [17] in the new
FV3-based GCM. The objective is to use the resolved
wave-mean flow forcings at high resolution to refine
the characteristics for the GWs implemented in the
subgrid-scale parameterization (at lower resolution).
The end goal is to better represent the unresolved dynamical effects of the GWs on the atmosphere in the
new FV3-based GCM (Kling et al., this meeting [18]).
Early Mars. The early Martian climate has been
the subject of scientific debate for decades. A range
of observed fluvial features [19, 20, 21] imply warm
and wet conditions occurred at least intermittently

~3.5-4 Gya but reproducing these conditions with climate models is difficult [22]. We aim to improve our
understanding of early Mars climate conditions and
active processes during the Noachian. We are currently investigating the role of CO2 clouds on early
Mars and how they affect the thermal structure of the
atmosphere through condensation and through their
radiative effects as they scatter in both the visible and
the infrared (Steakley et al., this meeting [23]). Additionally, we are exploring reducing greenhouse environments with the GCM that result from impact degassing [24]. We find that large impactors (>100 km
in diameter) that produce H2 in 2 bar CO2 atmospheres
can induce long lived above-freezing surface conditions (Steakley et al., submitted [25]). Individual impacts could induce 10s-100s of m of surface degradation over 105 years, which would account for a fair
portion of the total Noachian crater degradation estimated by [26] (~200-1000 m).
Mars-like Exoplanets. Arid land planets like
Mars may be common in planetary systems outside of
our Solar System. It is important to understand their
climate, potential habitability and defining characteristics to aid in their detection and characterization.
Our results show that dust radiative heating, particularly in the high atmosphere, fundamentally modifies
the climate of a dry planet at Earth insolation (Hartwick et al, this meeting [27]). When dust is lofted into
the middle atmosphere, heating rapidly strengthens
the zonal mean circulation and amplifies the diurnal
and higher harmonic components of the thermal tide,
which together positively feeds back on the process of
dust lifting. The atmosphere becomes very dusty and
strong greenhouse warming of the surface is predicted
to occur. We plan to continue investigating this topic
with interactive and coupled dust and water cycles to
study how cloud formation and the radiative effects of
clouds affects the dust cycle and climate. We also
plan to simulate land planet climates around M-stars,
the most likely targets for future observation and characterization of terrestrial exoplanets.
Public Releases and Community Tutorial: We
have focused recently on publicly releasing GCM
source code, GCM output, and an analysis software
package for processing model output. These efforts
are summarized in the following paragraphs.
Legacy Mars GCM Source Code and Output. We
have released source code and output from the Legacy
GCM. The source code is available on the NASA
GitHub (https://github.com/nasa/legacy-mars-globalclimate-model), and output from the simulations described in [1] are available on the NAS Data Portal
(https://data.nas.nasa.gov/mcmc/data_legacygcm.php).
Community Analysis Pipeline (CAP). Analyzing
and visualizing complex multi-dimensional GCM
output is challenging. We are therefore developing an
analysis software pipeline for community and internal
use that is accessible, comprehensive, and versatile.
We are using Python for its open-source and crossplatform utility and the self-descriptive netCDF data

standard. We have developed an initial version of the
analysis pipeline that provides a streamlined workflow for users to access, analyze, and visualize the
Legacy and FV3-based GCM output. It is available
on GitHub (https://github.com/alex-kling/amesgcm).
We are adding advanced diagnostics and interfaces to
observational data sets and other publicly available
models/types of output (e.g., Mars Climate Database,
OpenMars, EMARS) for comparison with our GCMs.
Legacy GCM and CAP Tutorial. We hosted a virtual modeling tutorial in the Fall of 2021 with the goal
of teaching new users how to use and analyze output
from the Legacy Mars GCM. The tutorial included
lectures and hands-on sessions to teach participants
about the basic physics in the model, its subroutines
and flow diagrams, how to make changes to the code,
how to compile and run the model, and how to analyze
model output. Students, teachers, and researchers
with a range of numerical modeling experiences participated. Tutorial materials are available at:
https://www.nasa.gov/mars-climate-modeling-centerames/MarsGlobalClimateModelTutorial.
New Ames Mars GCM Code and Output Release.
We plan to release the new FV3-based Ames Mars
GCM in the Fall of 2022. This will include both a
source code release on GitHub and a model output release on the NAS Data Portal. Our plan going forward
is to release a new version of the code and output
yearly. The initial release will include basic current
Mars physics only (CO2 and dust cycles; model top at
~90 km), but future releases will include increased capabilities.
Conclusions: In addition to continuing to make
progress on our science and model development
goals, we are committed to becoming as valuable of a
community resource as possible. This will involve future model and tool releases, tutorials, and workshops.
We are open to suggestions, so please feel free to get
in touch.
References: [1] Haberle et al. (2019), Icarus, 333; [2]
Bertrand et al. (2021), EPSC Meeting; [3] Urata et al. (2022),
7th MAMO; [4] Goldensen et al. (2008), GRL, 35(8); [5]
Haberle et al., (2022), 7th MAMO; [6] Montmessin et al. (2004),
JGR, 109(E10); [7] Hinson and Wilson (2004), JGR, 109(E1);
[8] Wilson and Guzewich (2014), GRL, 31(10); [9] Levèfre et
al. (2004), JGR, 109(E7); [10] Bertrand et al. (2020), 125(7);
[11] Montabone et al. (2020), JGR, 125(8); [12] Batterson et
al., (2022), 7th MAMO; [13] Wilson et al., (2022), 7th MAMO;
[14] Brecht et al., (2022), 7th MAMO; [15] Nakagawa et al.
(2020), JGR, 125(9); [16] Gkouvelis et al., (2022), 7th MAMO;
[17] Palmer et al. (1986), QJRMS 112(474); [18] Kling et al.,
(2022), 7th MAMO; [19] Craddock and Howard (2002)
JGRwilson a, 107; [20] Kite (2019) SSR, 215(1); [21] Kite and
Daswani (2019) EPSL, 524; [22] Wordsworth (2016) AREPS,
44; [23] Steakley et al., (2022), 7th MAMO; [24] Haberle et al.
(2019) GRL, 46(22); [25] Steakley et al. (2022), Icarus, submitted; [26] Golombek et al. (2006), JGR, 111(E12); [27] Hartwick et al., (2022), 7th MAMO.