OBSERVING FAR-ULTRAVIOLET OXYGEN AURORAE IN THE MARTIAN
NIGHT-SIDE ATMOSPHERE WITH MAVEN-IUVS.
E. Lieb, Laboratory for Atmospheric and Space Physics (emma.lieb@colorado.edu), N. Schneider, Laboratory
for Atmospheric and Space Physics, J.C. Gerard, LPAP, STAR Institute, University of Liege, L. Soret, LPAP, STAR
Institute, University of Liege, S. Jain, Laboratory for Atmospheric and Space Physics.

Introduction
There are at least three types of aurorae on Mars; discrete, diffuse, and proton. Discrete aurorae were first
discovered by the Mars Express SPICAM instrument
and are further supported by MAVEN IUVS observations. Figure 1 panel A shows a typical discrete aurorae
emission. These aurorae are defined by their small,
short-lived patches of emission usually related to the
strong crustal field region in the southern hemisphere
(Bertaux et al. (2005)). Diffuse aurorae were first discovered by MAVEN IUVS and are defined by emission
that spans much of the Martian nightside atmosphere
(Schneider et al. (2015); Jakosky et al. (2015); Schneider et al. (2018)). Figure 1, Panel B shows an example
of a diffuse aurorae. These emissions are caused by
solar energetic particles (SEPs), specifically electrons
and protons accelerated to energies of roughly 100 keV.
Lastly, MAVEN IUVS also discovered proton aurorae
which occur on the Martian day side and are caused by
penetrating protons from the solar wind (Deighan et al.
(2018); Halekas et al. (2017); Ritter et al. (2018)).

Figure 1: Three types of aurorae on Mars. Mars Atmosphere
and Volatile Evolution (MAVEN)/Imaging UltraViolet Spectrograph (IUVS) apoapse images.

The focus of the present study is to showcase new
analysis of MAVEN IUVS night side observations in
the Far-Ultraviolet (hereafter FUV) wavelength range,
specifically emissions at 130.4 nm and 135.6 nm. Until
now, discrete aurorae have been extensively studied in
the Mid-Ultraviolet (hereafter MUV) range and it has
been shown that strong aurorae occur at 190 - 270 nm
which correspond to CO Cameron Band emissions and at
298 & 299 nm which correspond to CO+
2 UVD Doublet
emissions. Ritter et al. (2019), studied oxygen emissions at 130.4 nm and 135.6 nm on the Martian day side
and showed that these features are strongly correlated

and behave similarly to each other there, and can provide an understanding of the variability of the day side
Martian atmosphere. There have also been significant
results from the Emirates Mars Misssion (EMM) EMUS
instrument showing strong discrete aurorae in the FarUltraviolet. Due to the sensitivity and high vantage point
of EMUS observations, their detection of aurorae shows
variations with solar wind, interplanetary magnetic field
conditions, and local time (Lillis et al. (2021)). The
current study focuses on these emissions on the Martian
night side using MAVEN IUVS for the first time and so
far has revealed unique auroral events that have yet to
be understood.

Observations and Analysis
We have analyzed MAVEN IUVS periapse data, now in
the FUV, ranging in time from January 2015 to near the
present in February 2021. We have found 47 examples
of emissions at 130.4 nm or 135.6 nm. Both of these
emissions are due to electron impact on atomic oxygen
and carbon dioxide molecules. The 130.4 nm triplet is
caused by the OI3 S transition to the ground state emitting
at 130.22, 130.49, and 130.60 nm, but we refer to it here
as 130.4 nm, which is an allowed transition subject to
resonance scattering. The 135.6 nm quintet-triplet is
caused by the OI5 S0 -3 P transition which is a forbidden
quintet-triplet transition to the ground state at 135.56
and 135.85 nm, but we refer to it here as 135.6 nm, and
features no resonance scattering (Ritter et al. (2019)).
We define a detection of an aurora via statistical tests
and a by-eye analysis. Emissions at 130.4 nm pass our
statistical tests if the observation consists of more than
10 pixels that have a signal-to-noise ratio greater than 2
sigma. Emissions at 135.6 nm pass our tests if they have
more than 5 pixels with a signal-to-noise ratio greater
than 1.5 sigma. The discrepancy between 130.4 nm and
135.6 nm is due to the fact that the emission at 135.6 nm
is far fainter than that of 130.4 nm. Both emissions must
also have a solar zenith angle greater than 105 degrees
to ensure that no daylight scattering corrupts the observation. The MUV emissions pass our statistical tests
if at least 5 pixels show a signal-to-noise ratios greater
than 4 sigma and 3 sigma for CO Cameron bands and
CO+
2 UVD Doublet, respectively. After a scan passes
our statistical tests, a by-eye analysis is required to further refine the detections versus non-detections. Most

3383/8

3526/5

6414/2

6481/3

6675/8

7406/3

7409/6

7409/2

7711/6

7768/9

8605/2

8662/2

11394/11

13137/8

13227/9

14177/10

2584/5

4357/1

13097/5

3453/12

6483/3

6483/2

2622/9

3522/8

3522/10

4559/4

4560/2

6469/7

6481/4

6484/10

7371/6

8319/4

9154/6

9600/8

2512/5

6485/6

9155/6

9251/4

9591/8

11441/2

3520/4

3520/5

8615/3

8675/3

8687/3

8687/5

11279/3

Figure 2: A “family portrait” of all FUV aurorae detections in snapshots which show all four aurorael emissions. Within each
snapshot there are four scan images, the top left shows the CO Cameron Band, top right shows the CO+
2 UVD Doublet, bottom left
shows the 130.4 nm, and bottom right shows the 135.6 nm. The border color of each scan image corresponds to whether or not a
statistical detection was made at that wavelength. The CO Cameron band and the UVD Doublet scan images are bordered in red
and pink, the 130.4 nm scan image is bordered in blue, and the 135.6 nm scan image is bordered in green if there is a detection at
this wavelength.

observations that pass our statistical tests do not pass the
by-eye analysis, as a variety of factors are required to
make a positive detection. For example, it is not only
necessary for there to be a certain number of bright pixels, but those pixels must be relatively spatially coherent
and they must be observed far away from the terminator such that no scattered dayglow appears in adjacent
scans.
Figure 2 showcases emissions at four wavelengths
(CO Cameron bands, CO+
2 UVD Doublet, O130.4, and
O135.6) for all our FUV auroral events. The border
color of each emission depends on whether or not there
is a positive aurora detection at that wavelength (red for
CO Cameron bands, pink for CO+
2 UVD Doublet, blue
for O130.4, and green for O135.6). When considering
the underlying physics of aurorae, one would assume
that the mechanisms generating the MUV aurorae would
always allow for FUV aurorae, and vice versa. However,
this is not what we have found. 47% of our FUV aurorae
occur with no measurable MUV counterpart (Fig. 2,
first and second row). Roughly 10% of detections have
very bright and spatially coherent FUV aurorae with
no measurable MUV counterpart (Fig. 2, orbit 3383

scan 8, orbit 8605 scan 3, and orbit 3453 scan 12). For
detections of concurrent MUV and FUV aurorae, they
tend to be very bright and spatially coherent (Fig 2,
last row), this amounts to about 15% of the detections.
Additionally, more than 93% of MUV aurorae have no
simultaneous FUV aurorae. The process which causes
emission bright enough to be detected in one wavelength
range but not in the other is not yet understood.
There are unique optical depth processes affecting
the detection of FUV aurorae with MAVEN IUVS. The
Martian atmosphere has unit optical depth when horizontal column densities exceed approximately 1017 cm−2 .
This occurs when CO2 densities at the tangent point are
approximately equal to 1010 cm−3 and this local density
occurs at an altitude of 125 km. We assume that all FUV
light is absorbed by the atmosphere anywhere below 115
km.
In periapse mode, there are three points of interest
along the line of sight: the nearside intercept point, the
tangent point, and the farside intercept point. The instrument only reports the tangent point but the aurorae could
be located at any point along the line of sight through
the limb of the atmosphere. The assumed altitude for

aurora on Mars is at an altitude of 135 km, at which both
the near side intercept point and the far side intercept
point are located. For those detections at a tangent point
altitude below 115 km, the aurorae must occur at the
near side intercept point because the atmosphere is not
only optically thick to FUV emission at the tangent point
but more importantly the precipitating electrons which
cause the aurora cannot penetrate the atmosphere at the
tangent point.
Unlike the MUV aurorae, there is no clear correlation between FUV aurorae and their proximity to the
strong crustal field region, defined to be from 50 to 210
degrees longitude and from -60 to -30 degrees latitude.
That is to say, FUV aurorae can be and have been detected at all latitudes and longitudes. However, the FUV
aurorae with simultaneous MUV counterparts are more
likely to occur near the strong crustal field region, which
is expected because the brightest MUV detections occur within the strong crustal field region. Variations in
detection frequency and brightness with solar longitude,
local time, and solar zenith angle are negligible.
Conclusions
The existence of strong aurorae in the FUV wavelength
range with no simultaneous aurorae in the MUV suggests

that there is an interesting phenomena in the Martian atmosphere to be understood. A possible explanation for
the differences lies in the energy distribution of precipitating electrons. The oxygen mixing ratio increases at
higher altitudes where less energetic electrons are able to
penetrate, rather than lower down where they get scattered or absorbed. In this case, electron impact on O
atoms is favored rather than collisions with CO2 which
could allow for more detections of FUV aurorae at higher
altitudes, but this has not yet been proven. Another avenue of investigation is how these differences vary over
the solar cycle, we are lucky to have data that spans from
solar maximum of solar cycle 24 in 2015 to the rise of
solar cycle 25 in 2021. Studying the effects of solar activity will help to prove if there is a correlation between
the distribution of energetic particles and the detection
frequency of FUV aurorae, as there are more energetic
particles hitting Mars when the sun is more active. Future work on this study will also involve modeling of the
Martian atmosphere, in order to simulate the conditions
that allow for the observations we see to help us understand the physical mechanisms behind these unique
night side FUV aurorae.