RETRIEVALS OF ATMOSPHERIC AEROSOLS IN HIGH AEROSOL
CONDITIONS FROM MARS CLIMATE SOUNDER MID- AND FAR
INFRARED LIMB MEASUREMENTS
A. Kleinböhl, D. M. Kass, S. Piqueux, S. Suzuki, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA (armin.kleinboehl@jpl.nasa.gov).

Introduction:
Limb sounding of thermal emission in the infrared wavelength range is a powerful technique for
measuring dust and water ice aerosols in the martian
atmosphere. It provides vertical profile information
and, due to the long optical path, typically has a
higher sensitivity compared with nadir viewing
measurements. However, the limb path might become opaque in high aerosol conditions, preventing
a limb sounding measurement from penetrating the
atmosphere. Atmospheric opacities of aerosols in the
far IR are typically lower than in the mid-IR, allowing atmospheric retrievals in conditions of higher
aerosol opacity.
Here we present atmospheric profile retrievals of
dust and water ice from Mars Climate Sounder
(MCS) limb measurements that include radiances in
the far IR. We use empirically derived extinction
efficiencies for dust and water ice to convert extinctions obtained in the far IR to the mid-IR. In dust
profile retrievals the use of the 32 μm channel of
MCS (channel B1) extends the retrievable altitude
range in dust storm conditions typically by about a
scale height. In water ice retrievals the use of the
MCS 42 μm channel (channel B2) extends the available vertical range in conditions of thick water ice
clouds typically by at least a scale height as well. In
cloudy conditions with low levels of dust, such as in
the aphelion cloud belt at nighttime, the 32 μm chan-

nel can additionally be used for water ice, allowing
retrievals down to the lowest scale height in most
conditions. The increased vertical range of the retrieved profiles, which often extend to the lowest
scale height, allows us to report aerosol column
quantities derived from these profiles by extrapolation to the surface.

MCS Instrument and Retrievals:
The Mars Climate Sounder [1] is a passive infrared radiometer onboard Mars Reconnaissance Orbiter (MRO), which views the martian atmosphere in
limb, nadir, and off-nadir geometries. It has 5 midIR, 3 far IR, and one broadband visible/near-IR
channels. Each spectral channel uses a linear detector array consisting of 21 elements, which provides 10 to 90 km altitude coverage with 5 km vertical
sampling when pointed at the Mars limb.
Profile retrievals from MCS radiance measurements use a modified Chahine method together with
a Curtis-Godson approximation in the radiative
transfer [2] and employ a single-scattering approximation to account for scattering in the limb radiative
transfer [3]. The retrieval algorithm employs twodimensional radiative transfer for both temperature
and aerosol retrievals in limb-viewing geometry in
order to correct for lateral inhomogeneities along the
line-of-sight of limb measurements [4]. Retrievals of
temperature profiles use the MCS mid-infrared
channels A1, A2, and
A3 that cover frequencies within the 15
μm gaseous absorption band of CO2.
Water ice extinction
retrievals are based on
limb measurements of
an absorption feature
in channel A4 with a
center frequency of
843 cm-1. Water ice
extinction is represented in the MCS
radiative
transfer
Figure 1: Ratio between extinction efficiencies for dust in the MCS channels B1
through particles that
and A5 (left, solid line), and for water ice in the MCS channels B2 and A4 (right,
follow a modified
solid line) and B1 and A4 (right, dashed line) as calculated by Mie theory for varigamma-distribution
ous particle sizes following a geometric series (dashes on x-axis). Dotted horizonwith an effective radital lines indicate empirical extinction efficiency ratios derived for dust (left, B1/A5)
us of 1.41 μm [3] and
and water ice (right, B2/A4). Arrows indicate the theoretical extinction efficiency
spectroscopic parameratios adopted for water ice in channel B1 in case this channel is used for water ice
retrievals.

ters by Warren [5]. Dust extinction is retrieved from
limb measurements using an absorption feature in
channel A5 with a center frequency of 463 cm-1.
Dust extinction is represented through particles that
follow a modified gamma-distribution with an effective radius of 1.06 μm [3]. Dust spectroscopic parameters are based on Wolff et al. [6] in the midinfrared.
A challenge for the use of far IR radiances in the
retrieval process is the conversion of retrieved extinctions to the nominal MCS wavelengths in the
mid-IR. We use extinction efficiency ratios derived
empirically from MCS limb measurements to accomplish this task [7]. Figure 1 (left) shows a graphical representation of the empirically derived dust
extinction efficiency ratio between channels B1 at 32
μm and channel A5 at 22 μm in comparison with
theoretical extinction efficiency ratios calculated by
Mie theory for various particle sizes following a
geometric series. The empirical extinction efficiency
ratio agrees very well with the ratio for the nominal
particle size of 1.06 μm used in MCS retrievals. We
note this result does not strongly depend on particle
size such that the agreement is similarly good for
slightly smaller (0.75 μm) and larger (1.5 μm) particle sizes. Only for particle sizes above ~2 μm the
ratios start to differ significantly.

In the right panel of Figure 1 a graphical representation of the empirically derived extinction efficiency ratio with theoretical extinction efficiency
ratios calculated by Mie theory between channels A4
at 12 μm and B2 at 42 μm is shown. Due to significant variability in observed ratios in terms of latitude
and local time, three different values were derived
for equatorial latitudes on the day and night side,
respectively, as well as north polar latitudes. Extinction efficiency ratios between A4 and B2 suggest
particle sizes of order 3 μm on the dayside and 6 μm
on the nightside. We note that this analysis is
weighted towards the altitude in which water ice
tends to become opaque 12 μm in a limb view. The
average altitude where a cloud becomes too opaque
for a on the nightside is about 13 km while on the
dayside in the equatorial belt this altitude is around
28 km. So the smaller particles on the dayside are
associated with higher clouds than the larger ones on
the nightside. This behavior is consistent with results
from limb observations by the CRISM instrument,
which find decreasing particle sizes with altitude and
a typical particle size of 2-3 μm at 20-30 km in the
daytime aphelion cloud belt [8].
We use these derived extinction efficiency ratios
separately in retrievals for the north polar region
(applied north of 50°N) and the equatorial region

Figure 2: Temperature (top), dust extinction (center) and water ice extinction profiles (bottom) retrieved
from MCS measurements during the dayside part of one MRO orbit around northern fall equinox in the
standard retrieval version (v5.2) using only mid-IR channels (left) and in the new retrieval version using a
combination of mid- and far IR channels (right). Dashed lines indicate the locations of the individual profiles.

(applied south of 40°N). In addition, daytime and
nighttime measurements are considered separately in
the equatorial region. Results of this analysis were
not robust enough to derive empirical extinction
efficiency ratios between channels A4 and B1 for
water ice. For the use of B1 in water ice retrievals,
which is only considered within the equatorial cloud
belt, extinction efficiency ratios are based on theoretical values calculated by Mie theory (dashed line in
Figure 1) at particle sizes that are part of a geometric
series (2.83 μm and 5.66 μm, arrows in Figure 1),
which are closest to the empirically derived sizes.
Results:
Aerosol profiles: Figure 2 shows a comparison
between the original retrieval and the new retrieval using far IR channels along a transect of
the daytime part of an
orbit at Ls=185°. Although after what is considered the aphelion season, the atmosphere in the
equatorial region is still
dominated by substantial
cloud occurrence. Clouds
are present between about
20°S and 30°N and reach
as high as 50-60 km. The
opacity of the clouds is
sufficient to prevent limb
retrievals that use the
mid-IR channels only
(left panel in Figure 2).
Because aerosol is not
retrieved up to middleatmospheric
altitudes,
temperature is not reported either over this altitude
range.
The right panel in
Figure 2 shows retrievals
of the same measurements with the new algorithm, using a combination of mid-IR and far IR
channels for aerosol retrievals. Retrievals are
now possible from the
middle atmosphere down
to about 10 km altitude
for nearly all measurements. This extension in
vertical coverage is largely enabled by the use of
channel B2 for water ice
retrievals. The resulting
aerosol profiles show that

the equatorial region is dominated by thick water ice
clouds between 20 and 50 km altitude that overlay a
dust layer. The equatorial cloud structure is roughly
bell-shaped, with the clouds reaching highest at the
equator. They fall off in altitude towards higher
latitudes, forming tendrils that reach down to 10-20
km altitude around 20°S and 35°N. Due to the successful aerosol retrieval also temperature is now
reported down to altitudes near the surface based on
retrievals from combined limb and on-planet observations in the 15 μm CO2-band.
Aerosol columns: The extended vertical range of
aerosol profile retrievals using combinations of mid-

Figure 3, top: Dust column optical depths from MCS daytime measurements in
January 2009 (Ls=184°-201°), derived from retrieved dust profiles, extrapolated
homogeneously mixed from the lowest retrievable altitude to the surface. Center:
Daytime water ice column optical depths derived from retrieved water ice profiles
for the same time period. Invalid column derivations are marked by black and red
circles in both panels. Bottom: Lowest retrievable altitudes for aerosol profile
retrievals used to derive column quantities.

and far IR limb views allows the derivation of aerosol column quantities. For dust, extinction profiles
are extended homogeneously mixed from the lowest
retrievable altitude to the surface and vertically integrated in order to derive a column. For water ice the
retrieved extinction profile is integrated directly to
derive a column. Column quantities are typically
reported if aerosol profiles reach down to 15 km
altitude or below.
The top and center panels of Figure 3 show dust
and water ice columns, respectively, derived from
profile retrievals on the dayside in January 2009.
Slight enhancements in dust are visible in the equatorial region as well as in the Hellas and Argyre
basins due to their low surface elevations. Water ice
clouds are present in the equatorial region, centered
at three longitudes around 110°W, 30°W, and 100°E.
In addition, the emerging polar hood is visible north
of 50°-60°N. The south polar region exhibits a partial annulus of clouds around 70°S that reaches from
about 120°E to 120°W.
The bottom panel of Figure 3 shows the minimum altitudes for daytime aerosol profile retrievals
that were achieved in January 2009 using combined
mid- and far IR limb measurements. Typically, retrievals reach altitudes of 10-15 km. Only in the
equatorial region a few retrievals still cut off at significantly higher altitudes due to the large ice opacity.
Summary and Outlook:
We presented new atmospheric profile retrievals
of dust and water ice from Mars Climate Sounder
(MCS) limb measurements that include radiances in
the far IR. Empirically derived extinction efficiencies for dust and water ice are used to convert extinctions obtained in the far IR to the mid-IR. Dust profile retrievals use the 32 μm channel of MCS while
water ice profile retrievals use the 42 μm channel
and occasionally also the 32 μm channel in conditions of the aphelion cloud belt. The use of the far IR
channels extends the retrievable vertical range in
high aerosol conditions typically by at least a scale
height. The extended vertical coverage of the profile
retrievals enables the derivation of dust and water ice
columns through vertical integration. In addition, the
improved representation of aerosols allows a more
accurate estimate of atmospherically corrected surface temperatures from on-planet views. It is planned
to re-retrieve the complete MCS dataset, spanning
the time period from September 2006 to the present,
with this new retrieval version (v6). The updated
retrievals will be re-delivered as a revised version of
MCS Level 2 data to NASA’s Planetary Data System, where it will be made publicly available. This
new version of Level 2 data will also include dust
and water ice column quantities.

Acknowledgements:
This work was performed at the Jet Propulsion
Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space
Administration. © 2022, California Institute of
Technology. Government sponsorship acknowledged.
References:
[1] McCleese, D. J. et al. (2007) JGR, 112, E05S06.
[2] Kleinböhl, A. et al. (2009) JGR, 114, E10006.
[3] Kleinböhl, A. et al. (2011) JQSRT, 112, 15681580.
[4] Kleinböhl, A. et al. (2017) JQSRT, 112, 511-522.
[5] Warren, S. G. (1984) Appl. Opt., 23, 1206–1225.
[6] Wolff, M. J. et al. (2006) JGR 111, E12S17.
[7] Kleinböhl, A. et al. (2017) Sixth Workshop on
Mars Atmosphere Modeling and Observations.
[8] Guzewich, S. D. et al. (2019) JGR 124,
636-643.