MODELING IMPACT OF GRAVITY WAVES ON THE THERMOSPHERE
OF MARS WITH A NON-OROGRAPHIC WHOLE ATMOSPHERE
GRAVITY
WAVE
SCHEME:
M-GITM
APPLICATION
FOR
INTERPRETING MAVEN MEASUREMENTS
S. W. Bougher, K. J. Roeten, CLASP Department, U. of Michigan, Ann Arbor, MI. USA,
(bougher@umich.edu), E. Yiğit, Department of Physics and Astronomy, George Mason University, Fairfax, VA
USA, A. S. Medvedev, Max Planck Institute., Germany, M. K. Elrod, M. Benna, NASA Goddard Space Flight
Center, Greenbelt, MD USA.

Introduction:
Gravity waves are ubiquitous throughout the
Mars atmosphere. They play an important role in
producing the general circulation pattern and temperature structure in the middle and upper atmosphere of Mars (e.g., Medvedev and Yiğit, 2012) and
facilitate cloud formation in the upper mesosphere
and lower thermosphere (Yiğit et al., 2018). Expected variations of this thermosphere structure and
winds owing to changing solar insolation are significantly modified by upward propagating gravity
waves. Indeed, the effects of GWs in the Martian
upper atmosphere have been observed by the
MAVEN (Mars Atmosphere Volatile EvolutioN)
mission up to at least 250 km (Yiğit et al. 2015).
Nevertheless, a comprehensive characterization of
GWs in the Martian upper atmosphere is still lacking. In this regard, improvements in numerical modeling can serve to address the impacts of GWs on the
upper atmosphere. For this, we implement a state-ofthe-art GW scheme into a Martian General Circulation Model and study GW effects in the thermosphere. We provide a possible interpretation of the
thermospheric temperatures retrieved from MAVEN.
MAVEN Neutral Gas and Ion Mass Spectrometer
(NGIMS) measurements include both Science Orbits
(SO) and a set of 9-Deep Dip (DD) campaign orbits
throughout the mission. Neutral argon densities are
typically
used
to
extract
corresponding
thermospheric temperatures. SO sampling and temperature extraction typically occurs above ~150 km,
while DD campaign sampling can reach to ~125 km,
providing lower thermospheric measurements. For
the latter, Deep Dip 2 (DD2) near noon at the equator is selected for detailed study of its extracted temperature profile (Stone et al. 2018). General circulation model (GCM) simulations to date (using modern
EUV heating efficiencies, but without accounting for

subgrid-scale GWs) have found it difficult to reproduce the details of this DD2 profile. Similarly, recent
NGIMS measurements of global thermospheric wind
components can be interpreted using general circulation models (Benna et al., 2019; Roeten et al. 2019).
GCM simulations with solar forcing alone cannot
capture the full range of variability or the campaign
averaged behavior seen in these measured winds over
roughly 10-orbit monthly campaigns spanning ~150220 km. This missing GW physics is likely an important underlying factor.
M-GITM Model and GW Scheme Setup:
To examine the thermospheric impacts of using a
modern GW scheme, new Mars Global IonosphereThermosphere Model (M-GITM) (Bougher et al.
2015) simulations are conducted and outputs compared to corresponding DD2 temperatures and global
winds from selected NGIMS campaigns. Modern
M-GITM model simulations are driven by daily solar
EUV-UV fluxes (from the MAVEN/EUVM instrument). Specifically, EUVM daily fluxes from the
FISM-M empirical model (Thiemann et al., 2017)
are used to supply inputs to the M-GITM code for
calculating solar EUV-UV heating, photodissociation, and photo-ionization rates (e.g., Roeten
et al. 2019). In addition, a fast and modern formulation for non-LTE CO2 15-µm cooling is now used
within the M-GITM code from Gonzalez-Galindo et
al. (2013). This scheme serves to accurately capture
the CO2 cooling rates, especially near the mesopause.
Most importantly, a spectral nonlinear whole atmosphere GW scheme (Yiğit et al. 2008, Medvedev and
Yiğit, 2012) is newly incorporated into M-GITM,
using a robust set of standard GW parameters, and its
impact on thermospheric temperatures and winds is
investigated. Specifically, the global distribution of
gravity wave momentum and energy deposition terms
are now computed interactively for the M-GITM

code (Roeten et al., 2022). This GW scheme is tailored for the treatment of non-orographic gravity
waves that reach the thermosphere from their launching point at the top of the planetary boundary layer.
This scheme is one of the most recent parameterizations appropriate for gravity waves that propagate to
thermospheric altitudes. It accounts for wave dissipation due to molecular viscosity in the upper atmosphere, molecular thermal conduction, radiative
damping, and nonlinear breaking-saturation (Yiğit et
al., 2008). Standard gravity model parameters utilized in M-GITM code are as follows: (a) horizontal
wavelength (300 km), (b) GW spectrum source
height (~top of the planetary boundary layer, ~9 km),
and (c) the momentum source flux (0.0025 m2/s2).
Some of these key parameters are based upon existing GW measurements obtained by MAVEN and
previous Mars missions, e.g. horizontal wavelength
(Medvedev et al. 2011); other parameters are unconstrained thusfar (e.g. momentum source flux).

Figure 1: Zonal momentum deposition rate in units
of m/s/sol. 15-day zonal average.

Figure 2: Zonal energy deposition rate in units of
K/sol. 15-day zonal average.

General Modeling Results:
For initial investigation, we have selected Ls ~
270-280° seasonal conditions (only), and solar moderate EUV-UV fluxes. Key fields (temperatures plus
zonal and meridional winds) are compared with and
without GW effects included. These fields are specifically plotted for zonally and temporally (15° ΔLs)
averaged values, for ease of comparison with previous MGCM simulations incorporating the whole
atmosphere GW scheme (e.g., Medvedev et al.,
2011; Yiğit et al., 2018).
Figures 1 and 2 illustrate that the chosen spectrum of
GWs primarily deposits its momentum and energy
over a range of altitudes from ~90-160 km.

Figure 3. Temperature differences for two cases: GW
case temperatures minus no GW case values (K).

The net impacts on simulated M-GITM temperatures
(see Figure 3) are twofold: (a) significant cooling in
the thermosphere at all latitudes, but especially at
high-latitudes (up to 45°K cooling at S. polar region), and (b) S. hemisphere warming (up to ~20 K)
at mesosphere altitudes (~60-90 km). The latter can
be compared with the MRO/MCS temperature climatology up to about ~90 km.
The corresponding net impacts on zonal and
meridional winds are best represented by raw wind
plots, for which “no GW” and “with GW” case wind
components are compared visually (see Figure 4).
Generally, zonal wind magnitudes are reduced by

about a factor of ~2 with the incorporation of the
GW momentum term. In addition, the single westerly “deep” zonal jet (spanning 0-250 km, in the summer hemisphere ) is now nearly closed off over
~100-130 km altitude, giving rise to separate westerly jets in the lower atmosphere (peaking at ~50±25
km) and in the thermosphere (above ~150 km).

Figure 5: DD2 campaign averaged temperature
profiles: (a) NGIMS derived (from Stone et al.
2018), (b) M-GITM “no GW” case, and (c) MGITM “with GW” case. Simulated cross-hatched
lines corresponds to the 1-σ variability of the orbitor-orbit temperatures around the computed mean.

Figure 4: Zonal and meridional wind plots for: (a)
left column (“no GW” case), and (b) right column
(“with GW” case). Wind magnitudes are in units m/s.

Specific Modeling Results: Data Comparisons
Specific simulations using M-GITM were conducted
(with and without GWs) corresponding to the DD2
campaign in April 2015. This was the portion of the
MAVEN mission when solar fluxes were near the
peak of solar cycle #24 conditions. Corresponding
seasonal conditions were near Ls ~ 328 (after perihelion). Finally, extracted NGIMS DD2 temperatures
sampled at low SZA (~9.0°) are compared to newly
simulated M-GITM profiles.

Figure 5 shows the simulated temperature profile
including GWs is cooler by ~15-20°K (above 180
km) and ~10-15°K at lower altitudes. These computed temperatures are much closer to NGIMS extracted
values. The reason for this improvement is due to the
slowed circulation, which reduces the transport of
atomic O from the dayside to the nightside of Mars.
This yields increased dayside O/CO2 ratios near
~150 km where the CO2 15-µm cooling layer peaks.
These O abundances now match NGIMS measured
values in the lower thermosphere. Recall that collision of O and CO2 enhances 15-µm emission under
non-LTE conditions (e.g. Bougher et al. 2017). Direct cooling of the thermosphere by GWs has also
been modeled on Earth using the whole atmosphere
scheme (Yiğit and Medvedev, 2009), suggesting that
GWs should be included in the energy balance of the
thermosphere.
New M-GITM simulations were also run for the January 2017 NGIMS neutral wind observational campaign. Figure 6 shows both the averaged NGIMS
velocities and corresponding M-GITM simulations,
with and without GWs. This campaign occurred
over Jan 11-13, 2017, from Ls=297-299 (in the
southern summer season), near midnight local time,
and over southern mid-latitudes.

Bougher, S. W., D. A. Brain, J. L. Fox, F. GonzalezGalindo, C. Simon-Wedlund, and P. G. Withers
(2017). Chapter 14 in The Atmosphere and Climate
of Mars, ed. B. Haberle, M. Smith, T. Clancy, F.
Forget, R. Zurek, Cambridge University Press,
doi:10.1017/9781107016187.
González-Galindo, F., Chaufray, J.-Y., LópezValverde, M., Gilli, G., Forget, F., et al. (2013).
JGR: Planets, 118.doi.org/10.1002/jgre.20150
Medvedev, A. S., Yiğit, E., Hartogh, P., & Becker,
E. (2011). JGR, 116, doi: 10.1029/2011JE003848
Medvedev, A. S., and Yiğit, E. (2012), GRL, 39,
L05201, doi:10.1029/2012GL050852.

Figure 6. Data-model comparison plots for the
January 2017 wind campaign illustrate the improvement of the computed mean wind magnitude, but no
improvement in the direction with the “with GW”
case.
With the addition of the GW scheme, the M-GITM
wind speed profile slows by over 100 m/s, bringing it
much closer to the averaged values observed by
NGIMS. The simulated direction also shifts with the
addition of GW effects, becoming less easterly and
more westerly, while retaining the southerly component. While it cannot be said that this aspect of the
data-model comparison improves, comparing the
mean direction is difficult in this particular campaign
due to extreme orbit-to-orbit variability in wind direction observed by NGIMS at the time.
Overall, these upgraded M-GITM simulations
demonstrate that gravity waves can have a significant
influence in shaping thermospheric winds and temperatures at low and high latitudes.
References:
Benna, M., S. Bougher, Y. Lee, K. Roeten, E. Yiğit,
P. Mahaffy, B. Jakosky (2019), Science,
doi:10.1126/science.aax1553
Bougher, S.W., Pawlowski, D., Bell, J. M., Nelli, S.
McDunn, T., Murphy, J. R., et al. (2015). JGR:
Planets, 120. doi.org/10.1002/2014JE004715

Roeten, K J., S. W. Bougher, M. Benna, P. R.
Mahaffy, Y. Lee, D. Pawlowski, F. Gonzalez Galindo, M. A. Lopez-Valverde (2019), JGR, 124,
1-21. doi:10.1029/2019JE005957
Roeten, K.J., Bougher, S. W., Yiğit, E., Medvedev,
A. S., Benna, M. and Elrod, M. K., (2022). Impacts
of Gravity Waves in the Martian Thermosphere using
M-GITM Coupled with a Whole Atmosphere Gravity Wave Scheme. JGR: Planets, submitted.
Stone, S.W., Yelle, R. V., Benna,M., Elrod, M. K.,&
Mahaffy, P. R. (2018). JGR, 123, 2842–2867.
doi.org/10.1029/2018JE005559
Thiemann, E. M. B., Chamberlin, P. C., Eparvier, F.
G., Templeman, B., Woods, T. N., Bougher, S.W., &
Jakosky, B.M. (2017). JGR: Space Physics, 122,
2748–2767. doi.org/10.1002/2016JA023512
Yiğit and Medvedev (2009), GRL, 36,
doi :10.1029/2009GL038507.
Yiğit, E., A. D. Aylward, and A. S. Medvedev
(2008), JGR, 113, doi:10.1029/2008JD010135
Yiğit,, E., S. L. England, G. Liu, A. S. Medvedev, P.
R. Mahaffy, T.Kuroda, B..M. Jakosky (2015), GRL,
42, doi:10.1002/2015GL065307.
Yiğit, E., A. S. Medvedev, P. Hartogh, (2018), Ann.
Geophys, 36, 1631-1646, doi: 10.5194/angeo-361631-2018.