A GLOBAL STUDY OF MARTIAN ATMOSPHERIC DUST WITH THE
NEAR-INFRARED
IMAGING
SPECTROMETER
OMEGA/MARS
EXPRESS.
Y. Leseigneur and M. Vincendon, Institut d’Astrophysique Spatiale, Université Paris-Saclay, CNRS, Orsay,
France (yann.leseigneur@ias.u-psud.fr.

Introduction:
Dust, composed of micrometer-sized particles, is
an important component of both the atmosphere and
the surface of Mars. This high albedo material modifies heat balance while being highly mobile: it is thus
a major contributor to current Mars atmosphere and
surface dynamics. Numerous studies realized from
the Martian orbit or from the ground already provided
information on dust spatial and time distribution (e.g.,
[1], [2], [3]). However, some characteristics of the

dust cycle remain uncertain. This is notably related to
the fact that dust movements occur over a large range
of spatial scale, from Recurring Slope Lineae (RSL)
to Global Dust Storm (GDS), and with important interannual variability ([4], [5]). Further global monitoring of dust movements with high spatial sampling
and over several years may help improving our
knowledge about Martian dust. Here we present a new
method used to derived atmospheric dust optical
depth using OMEGA observations acquired between
2004 and 2010, i.e. over three Martian Years (MY).

Figure 1: Six seasonal atmospheric VIS-NIR dust optical depth maps of Mars during MY 29, derived from the OMEGA dataset
(see text for details).

Data and method:
The near-infrared imaging spectrometer OMEGA
[6] onboard Mars Express collected about 8300 observations over that time range, with a spatial sampling
typically close to 1 km.
First, we analyze the prominent 2 µm atmospheric
CO2 absorption band to detect dust, as airborne dust
reduces atmospheric optical paths. We have developed a model to predict the CO2 optical depth 𝜏!"!
under clear atmospheric conditions: observed CO2 optical depths are compared to these predictions to derive the difference “Δ𝜏!"! ”, an index sensitive to dust.
More details about this first part of the method can be
found in [7].
Then, we calibrate this Δ𝜏!"! using the visible
(VIS) dust optical depth (from 880 nm images) measured by the Mars Exploration Rovers (MER: Spirit
and Opportunity) [2], which operated during the same
period as OMEGA. Concomitant MER/OMEGA observations are used to characterize the relation between Δ𝜏!"! (OMEGA) and the ground optical depth
(MER), which notably depends on solar incidence angle and albedo. This relation is then extrapolated to
the whole OMEGA dataset to calculate a VIS-NIR
dust optical depth (𝜏#$%& ).
Results and discussion:
The OMEGA dataset allows to produce global
seasonal maps of dust optical depth with a spatial coverage typically greater than 50% for most 60° wide
intervals in solar longitude (𝐿% ). We provide the six
seasonal maps of the MY 29 in Figure 1. Firstly, we
observe the well-known seasonality of dust: overall
higher values of dust optical depth during the “dusty
season” (𝐿% = 180 − 300°) than during the “clear atmosphere season” (𝐿% = 0 − 120°). Some localized

higher values are also observed in all maps, such as in
the 𝐿% = 0 − 60° map (top left panel of Figure 1) near
315°E and 20°S, that can be associated to dust storms.
Then, some main dust storm travel routes [8] (from
the northern hemisphere to the southern one) can be
identified, in particular in the 𝐿% = 300 − 360° map,
the Acidalia-Chryse-Margaritifer route which correspond to moderate and high values of dust optical
depth. This area is also known by the observation of
Recurring Slope Lineae (RSL), which are active in
this 𝐿% = 300 − 360° period [9]. This indicates that
large-scale atmospheric dust events may be connected
to local surface dust movements occurring at RSL location.
We are also able to plot latitude/solar longitude diagrams with the OMEGA dataset, such as for MY 27
in Figure 2, that provides information on the time evolution of the dust optical depth. We observed characteristics of the well-known dust cycle, in particular the
late storms called “C-storms” between 𝐿% = 310° and
330° (as defined in [10]). Other main dust optical
depth investigations were made with thermal infrared
data [3], where results are preferential represented in
this type of diagrams. We can notice an overall good
agreement between our study in VIS-NIR and in thermal studies (see Fig. 16 of [3]).
But there are still differences that can be studied
by computing the ratio between our VIS-NIR
OMEGA values derived at 2 µm (and converted to
880 nm using MER) and the thermal ones at 9 µm.
We show the histogram of this ratio in Figure 3. The
maximum is reached for a ratio of 2.4, which is in
agreement with the value of 2.6 considered in [3] and
also the values found locally with the MER (see Fig.
10 of [2]). We can also notice that the distribution is
very large, with a 2.0 equivalent standard deviation,
which may be related to variations of the mean dust
particle size according to [2].

Figure 2: OMEGA VIS-NIR dust optical depth as a function of solar longitude (𝐿! ) and latitude for MY 27. For each 𝐿! and
latitude, we average 𝜏"#!$ obtained at various longitudes. Each pixel of the diagram is 5° wide in 𝐿! and in latitude. The
sampling in longitude is variable (see examples of maps for another MY in Figure 1).

Figure 3: Global histogram of the ratio between OMEGA VIS-NIR dust optical depth (computed at 2 µm and converted to
match the 880 nm MER reference) and thermal-infrared (TIR) optical depth (9 µm) presented in [3].

Conclusion:
We have developed a new method to detect atmospheric dust using near-infrared spectrometer data. The
method relies on CO2 absorption band at 2 µm and is
applied to OMEGA data. We also implemented a calibration procedure based on MER ground measurements to compute a VIS-NIR dust optical depth. We
constructed seasonal global maps (Figure 1) and latitude/solar longitude diagrams (Figure 2). Overall, results are in good agreements with previous studies
based on thermal-infrared observations, but with
some differences noticed by a particular distribution
of the dust optical depth ratio between VIS-NIR and
thermal-infrared (Figure 3), with a maximum occurrence value of 2.4. This OMEGA dataset may also be
used to connect dust activity over several spatial
scales. For example, preliminary results indicate that
large scale atmospheric dust events may be related to
local RSL activity occurring in the Chryse-Acidalia
area.
Acknowledgments:
The OMEGA/Mex data are freely available on the
ESA PSA at https://archives.esac.esa.int/psa/#!Table%20View/OMEGA=instrument.

References:
[1] Smith M. D. (2004) Icarus, 167, 148-165. [2]
Lemmon M. T. et al. (2015) Icarus, 251, 96-111. [3]
Montabone L. et al. (2015) Icarus, 251, 65-95. [4]
Szwast M. A. et al. (2006) JGR, 111, E11008. [5] Vincendon M. et al. (2019) Icarus, 325, 115-127. [6] Bibring J-P. et al. (2004) ESA Publication Division,
1240, 37-49. [7] Leseigneur Y. et al. (2021), 52nd
LPSC, LPI contribution No. 2548, id. 1741. [8] Wang
H. et al. (2015) Icarus, 251, 112-127. [9] Stillman D.
E. and Grimm R. E. (2018) Icarus, 302, 126-133. [10]
Kass D. M. et al. (2016) Geophys. Res. Lett., 43,
6111-6118.