DIURNAL VARIATIONS IN THE MARTIAN ATMOSPHERE FROM
ENHANCED MAVEN/IUVS STELLAR OCCULTATION DATASET
S. Gupta, N. M. Schneider, S. K. Jain, J. Deighan, Laboratory for Atmospheric and Space Physics, University
of Colorado, Boulder, CO, USA, (sumedha.gupta@lasp.colorado.edu), R. V. Yelle, F. Jiang, Lunar and Planetary Laboratory Science, University of Arizona, Tucson, AZ, USA, L. Verdier, A. S. Braude, F. Montmessin,
LATMOS, CNRS/UVSQ Université Paris-Saclay/UPMC, Guyancourt, France

Introduction: Over the last five decades, diverse
observations and theoretical models have provided a
comprehensive account of thermal structure and composition for different regions of the Martian atmosphere. Nonetheless, there are still gaps in our interpretation when it comes to the study of day/night differences for the Martian upper mesosphere to lower thermosphere, a region that is critical in understanding atmospheric loss, coupling processes, global circulation
patterns, energy balances, along with the influences of
H2O, CO2, and dust cycles prevalent on Mars. Part of
this region is explored by a few Martian missions,
e.g., SPICAM onboard Mars Express observed the
nighttime temperature and density profiles in the altitude range 70-130 km (Montmessin et al., 2017),
NGIMS on MAVEN during its deep dip campaigns
can retrieve in-situ density and temperatures for different local solar times, but altitudes from ~125 km
(Stone et al., 2018), while ACS onboard TGO probes
the lower mesosphere, resolving the diurnal cycle
over a 54-sol period (Guerlet et al., 2022).
Stellar occultations, one of the observation modes
of the Imaging Ultraviolet Spectrograph (IUVS)
aboard the Mars Atmosphere and Volatile EvolutioN
(MAVEN) spacecraft, are being designed and executed in dedicated bimonthly campaigns for the last
seven years to fill this gap (Jakosky et al., 2015;
McClintock et al., 2014). This study presents a methodology for enhancing and expanding the usable mission-wide MAVEN/IUVS stellar occultation dataset
to provide the first complete diurnal coverage of Martian upper mesosphere to lower thermosphere (~80160 km) for a range of latitudes, longitudes, and Martian seasons. We here present the preliminary results
of this study.
MAVEN/IUVS Stellar Occultations: As the
stellar signal of UV bright star traverses the Martian
middle atmosphere, it gets attenuated due to absorption by the atmospheric constituents. Information on
the column abundances of CO2, O2, O3, together with
aerosol optical depth, can be obtained from the transmission spectra due to such absorption features, from
which are retrieved the local densities and profiles of
temperature and pressure by vertical inversion in the
20 – 160 km altitude range (Gröller et al., 2018).
There have been forty-two stellar occultation campaigns from 24 March 2015 to 29 January 2022

(covering MY 32 to 36) that are executed every 2 to 3
months and last between 1 and 2 days.
The limitation of the current dataset is that it contains only the nighttime events. Of the total 3003 stellar occultation events to date, ~53% of the observations correspond to the nightside while the remaining
47% are during dayside or bright terminator such that
the sunlight is scattered to the MUV channel of IUVS
and saturates it while part of this first-order stray light
contaminates the second-order FUV channel. In the
current stellar occultation data pipeline, dayside
events (LT 6–18) with stray light contamination are
not processed, therefore resulting in a data gap and
significant loss of important information and science,
especially on a diurnal scale.
The objective of this study is to expand and enhance the usable stellar occultation dataset by processing the daytime events. FUV (110 – 190 nm) is
crucial for the retrieval of CO2 and O2 above 90 km
and for examining the diurnal and seasonal variability
of the O2 mixing ratio. The majority of these unprocessed dayside cases have first-order stray light in the
FUV confined to the altitude range where FUV starlight is fully attenuated. An example is shown in Figure 1. Ignoring the contaminated MUV spectrum and
performing the retrievals on only the FUV spectrum
for such dayside cases can rescue the daytime events
and considerably expand the usable dataset.

Figure 1. An example of a dayside FUV spectrum with the
unattenuated stellar spectrum, rescuable finite absorption
region, and stray light contaminated records.

Methodology: We have devised a stray light
avoidance methodology to rescue at least the FUV
spectrum for the dayside occultation measurements
by 1) only considering l=115-170 nm, 2) eliminating
aerosol contribution in the spectral fits, 3) limiting the
altitude range to where the retrievals are credible (restricting transmissions 0.1-0.95), and 4) testing the
validity of these modifications on the nightside events
that are processed by the current pipeline (using 115300 nm). This is to eliminate any bias introduced due
to considering only the FUV wavelengths. The criterion for choosing a retrievable altitude range is shown
in Figure 2.

The stray light avoidance methodology has enabled the retrieval of ~1000 dayside events (LT 6–18).
Based on this methodology, the nightside events are
also reprocessed. To validate the stray light avoidance
approach, these reprocessed nightside events are then
compared to the pipeline retrievals that considers both
FUV and MUV. An example is shown in Figure 3. It
can be seen that the two profiles are consistent with
each other. We have been able to successfully process
2300 events to date using only the FUV wavelengths,
providing unprecedented day/night observations of
retrieved CO2 number density and temperature in the
80-160 km altitude range.
The enhanced and expanded usable dataset covers
a wide range of local times, latitudes, longitudes, and
the Martian seasons. The data coverage for these 2300
events is shown in Figure 4, covering three complete
Martian years.

Figure 2. Limiting retrieved altitude range based on transmission at two different wavelength ranges. The retrievals
in this example are credible between 95-130 km.

Figure 3. CO2 column abundance for the nightside event retrieved from only the FUV spectrum (black), with pipeline
retrieval from both FUV and MUV (red). The credible altitude range for the FUV analysis is given by green lines.

Observations and Results: Based on these dayside and nightside retrievals for the FUV, Figure 5 (a)
shows the diurnal cycle for temperature variations on
a vertical scale. It is to be noted that for each LT and
pressure bin, the temperature values are averaged over
corresponding Martian seasons, longitudes, and latitudes in the range 65°N – 65°S. There are temperature
anomalies at the top side due to the assumption of an
isothermal atmosphere in retrieving the upper boundary temperature from CO2 number density before integrating downward following hydrostatic equilibrium.
Prominent day/night differences are seen with the
dayside observations warmer than the nightside over
all pressure levels. The lowest temperatures are seen
in the midnight mesosphere, while the maximum temperatures reaching ~250 K are observed for LT 14-17
in the lower thermosphere. We also see evening terminator temperatures to be warmer than the dawn, especially in the lower thermosphere.

Figure 4. Data coverage for 2300 processed events considering only FUV with inverse square of heliocentric distance
(R) for MY 33-36. Each data point represents median latitude and local time for all the orbits corresponding to a star in
a campaign.

(a)

(a)

(b)
(b)

Figure 6. (a) Temperature and (b) pressure level at mesopause for different local times as given by the retrievals on
FUV and the corresponding MCD values.

Figure 5. Diurnal variations in temperature (a) based on retrievals for FUV, and (b) corresponding MCD values. These
observations are over all the Martian seasons, longitudes,
and latitudes 65°N – 65°S.

The analysis of the enhanced dataset shows that
the mesopause is significantly warmer in the daytime
than nighttime. The dayside upper mesospheric temperatures (up to 10-4 Pa) do not show much variation
(~140 K) with a warm and weak mesopause. This is
in contrast with the nightside mesospheric temperatures, whereby mesopause is placed at 10-3 – 10-4 Pa
with the minimum temperatures reaching ~100 K.
This agrees well with the nightside mesopause reported by SPICAM (González-Galindo et al., 2009).
Comparison with the MCD reveals an opposite
trend in the mesosphere with the nighttime mesopause
warmer than daytime. We compare our retrievals with
the predicted profiles from Mars Climate Database
(MCD) v5.3 using its climatological dust scenario for
average solar EUV conditions (Forget et al., 1999;
Millour et al., 2018). A quite opposite trend is seen by
MCD in Figure 5 (b) where values are taken corresponding to each of the observation conditions. This
discrepancy between the retrievals and the MCD is
also distinctly seen in Figure 6. While the MCD mesopause is quite stable at 10-3 – 10-4 Pa, it is colder by
~20 K during the dayside, resulting in a difference of
~40 K with the retrievals. On the other hand, it is
warmer and lower in altitude for the nightside. This
overestimation of cooling in the mesosphere during
the daytime could be due to improper parametrization
of CO2 15 µm cooling and radiative balance in LMDMGCM simulations (González-Galindo et al., 2015).

The day/night differences are predicted to be
much lower by the MCD. Quantification of such diurnal differences between the retrievals and the MCD,
for the entire altitude range of 80-160 km (5 km binning), is shown in Figure 7. The day/night differences
become more relevant above 90 km as solar forcing
imposes a strong diurnal cycle in the thermosphere.
Further, the rapidly increasing temperatures seen at
times lower in the atmosphere are not predicted by the
MCD.

Figure 7. Average day and night temperature profiles (65°N
– 65°S) for the retrievals on FUV (solid line) and corresponding MCD values (dashed lines).

The seasonal and latitudinal variations of the diurnal cycle exhibit expected characteristics. The retrieved temperature profiles will also be studied for
varied combinations of seasons and latitude ranges to
examine maximum and minimum day/night differences. Two examples are presented here in Figure 8
to show expected minimum nightside temperatures
for northern low/mid-latitudes around aphelion and
maximum dayside temperature for southern latitudes

during dust storm season around perihelion. We have
also seen nightside southern polar winter warming
that is due to an enhanced interhemispheric circulation around solstice and downwelling at winter poles,
but analogous warming for northern winter is not seen
due to sparse observations for the northern winter
high latitudes.
(a)

(b)

Figure 8. Day and night temperature profiles for low/midlatitudes (a) in the northern hemisphere around aphelion,
and (b) in the southern hemisphere around perihelion.

Conclusions: We have added ~1000 new dayside
events to the stellar occultation dataset that were previously not processed due to the contamination by
stray light. The formulated methodology to avoid
stray light considering only the FUV spectrum gives
consistent results when compared to the nightside test
cases processed by considering both FUV and MUV
spectra. Apart from considerably expanding and enhancing the usable dataset, this effort has enabled the
first complete day/night study of the upper mesosphere/lower thermosphere (80-160 km). It is seen
that the diurnal variations dominate any other variations. The temperature profiles so obtained demonstrate discrepancies with LMD-MGCM simulations,
especially around the mesopause.
Though these are some of the preliminary results,
our future endeavor would be focused on justifying
these observations with the mechanisms responsible –
the role of vertical winds and dynamics, heating rates,
and radiative transfer. These observations have the
potential to provide important inputs and constraints
to the models to improve their predictions.

Data Availability: The L1B fits files for each of
the stellar occultation events used in this study, along
with the higher-level data products, are archived in the
Planetary Atmospheres Node of the Planetary Data
System (PDS).
Bibliography:
Forget, F. et al. (1999). Improved general circulation
models of the Martian atmosphere from the
surface to above 80 km. Journal of
Geophysical Research E: Planets, 104(E10),
24155–24175.
https://doi.org/10.1029/1999JE001025
González-Galindo, F. et al. (2009). A ground-toexosphere Martian general circulation model:
1. Seasonal, diurnal, and solar cycle variation
of thermospheric temperatures. Journal of
Geophysical Research E: Planets, 114(4), 1–
21. https://doi.org/10.1029/2008JE003246
González-Galindo, F. et al. (2015). Variability of the
Martian thermosphere during eight Martian
years as simulated by a ground-to-exosphere
global circulation model. Journal of
Geophysical Research: Planets, 120(11),
2020–2035.
https://doi.org/10.1002/2015JE004925
Gröller, H. et al. (2018). MAVEN/IUVS Stellar
Occultation Measurements of Mars
Atmospheric Structure and Composition.
Journal of Geophysical Research: Planets,
123(6), 1449–1483.
https://doi.org/10.1029/2017JE005466
Guerlet, S. et al. (2022). Thermal structure and
aerosols in Mars’ atmosphere from
TIRVIM/ACS onboard the ExoMars Trace
Gas Orbiter : validation of the retrieval
algorithm. Journal of Geophysical Research:
Planets. https://doi.org/10.1029/2021je007062
Jakosky, B. M. et al. (2015). Initial results from the
MAVEN mission to Mars. Geophysical
Research Letters, 42, 8791–8802.
https://doi.org/10.1002/2015GL065271.Receiv
ed
McClintock, W. E. et al. (2014). The imaging
ultraviolet spectrograph (IUVS) for the
MAVEN mission. Space Science Reviews, 1–
50. https://doi.org/10.1007/s11214-014-0098-7
Millour, E. et al. (2018). The Mars Climate Database
(Version 5 . 3 ). Scientific Workshop: “From
Mars Express to ExoMars,” February, 27–28.
Montmessin, F. et al. (2017). SPICAM on Mars
Express: A 10 year in-depth survey of the
Martian atmosphere. Icarus, 297, 195–216.
https://doi.org/10.1016/j.icarus.2017.06.022
Stone, S. W. et al. (2018). Thermal structure of the
Martian upper atmosphere from MAVEN
NGIMS. Journal of Geophysical Research:
Planets, 123(11), 2842–2867.
https://doi.org/10.1029/2018JE005559