DIURNAL AND SEASONAL VARIATIONS OF AEROSOL OPTICAL
DEPTH AT JEZERO CRATER, MARS
Michael D. Smith, NASA Goddard Space Flight Center, Greenbelt, MD, USA (Michael.D.Smith@nasa.gov),
G. M. Martínez, LPI/USRA, Houston, TX, USA, E. Sebastián, V. Apéstigue, I. Arruego, D. Toledo Carrasco, D. Viúdez-Moreiras, J. A. Rodriguez-Manfredi, Centro de Astrobiología (INTA-CSIC), Madrid, Spain,
M. T. Lemmon, Space Science Institute, Boulder, CO, USA, M. de la Torre Juarez, NASA Jet Propulsion Laboratory, Pasadena, CA, USA.

Introduction: The two upward-looking TIRS
sensors from the MEDA instrument suite on-board
the Perseverance rover enable the retrieval of total
aerosol optical depth (dust plus water ice cloud)
above the rover for all observations when TIRS is
taken. Because TIRS observes in the thermal-IR, the
retrievals are possible during both the day and night
and thus, they provide an excellent way to monitor
both the diurnal and seasonal variations of aerosols
above Jezero Crater.

TIRS Data: The Thermal InfraRed Sensor
(TIRS) package on the Perseverance rover consists
of five sensors used to characterize the upward and
downward fluxes of visible and infrared radiation at
the rover site (Rodriguez-Manfredi et al., 2021). Of
interest here are the sensors TIRS IR1, which covers
a broad portion of the thermal-infrared spectrum over
the range 6–35 µm and TIRS IR2, which covers the
CO2 band between 14.5 and 15.5 µm. Both sensors
view upward at an elevation angle centered at 35°
above the rover deck. The field of view of the TIRS
IR1 and IR2 sensors is 20° in the vertical direction
by 40° in the horizontal direction. As part of the
MEDA suite of atmospheric sensors, TIRS observations are taken systematically throughout the sol at a
frequency of 1 Hz. Observations in one-hour blocks
are generally taken so that odd-numbered hours are
covered on odd-numbered sols, while evennumbered hours are covered on even-numbered sols.
In that way the entire 24-hour diurnal cycle is fully
covered over a span of 2 sols. Occasionally, additional one-hour blocks are added to this baseline
providing additional coverage, and observations are
also routinely taken during the first five minutes of
each hour. These observations are a part of the background baseline set for MEDA that runs essentially
every day providing excellent diurnal and seasonal
coverage.

Aerosol Optical Depth Retrieval: We use a
radiative transfer model to compute the expected
TIRS IR1 and IR2 signal for a given aerosol optical
depth and temperature profile. We can then perform
the retrieval by varying the atmospheric temperatures

and aerosol optical depth to match the values observed by TIRS. The radiative transfer model includes aerosol scattering using a 2-stream approximation (e.g., Smith et al., 2006) and treats absorption
by CO2 gas using the correlated-k approximation
(Lacis & Oinas, 1991). The elevation angle above
the horizon for the TIRS is important and is computed for each observation using the known rover pitch,
tilt, and yaw.
Determination of atmospheric temperatures. In
general, the observed signal from TIRS IR1 and IR2
is sensitive to a combination of the atmospheric
opacity (from aerosols and gases) and the atmospheric temperature profile. In practice, because of the
very high opacity of CO2 in the middle of the 15-µm
band compared to aerosols, the IR2 signal is only
sensitive to atmospheric temperatures in the lowest 1
km of the atmosphere and is nearly insensitive to
aerosols. As a baseline, we use the temperature profile returned from the Mars Climate Database (MCD)
(Forget et al., 1999; Millour et al., 2018) for the given latitude, longitude, local time, and Ls of each observation. We use our radiative transfer model to
compute the expected IR2 signal and we offset the
temperature profile in the lowest 1 km until the computed IR2 signal matches the observed signal. Outside of the case of large dust storms, the offset typically no larger than a few K.
The temperature profile above an altitude of 10
km is estimated using concurrent observations taken
by the EMIRS thermal-IR spectrometer on-board the
Emirates Mars Mission (Edwards et al., 2021, Smith
et al., 2022), with output from the MCD used to help
link the two regions where EMIRS and TIRS IR2 are
sensitive.
Once the temperature profile has been estimated
we are left with a single observation (TIRS IR1) and
a single unknown (total aerosol optical depth), which
results in the straightforward task of varying aerosol
optical depth in our radiative transfer model until the
computed IR1 value matches the observed value.
The relative contributions from dust and water
ice cloud cannot be separated using TIRS data alone.
To perform the retrieval, we assume a seasonally
dependent dust vs. ice fraction based on climatology
(e.g., Smith, 2004). In practice, the retrieved total
aerosol optical depth is not very sensitive to this frac-

tion for any reasonable values.
Results: Figure 1 shows the retrieved aerosol
optical depth for TIRS observations taken between
sol 15 (5 March 2021, MY 36, Ls=13°) and sol 364
(28 February 2022, MY 36, Ls=182°) as a function
of season (Ls) and local true solar time (LTST). Retrievals are performed on 5-minute blocks of data,
which are the averages of 300 individual observations. Shown here is the total aerosol extinction optical depth referenced to 1075 cm-1 (9 µm), which includes scattering. Periods of time when the Sun was
within the TIRS field-of-view are excluded since
those results are often unreliable.
Apparent in Figure 1 between Ls=30° and 130° is
the aphelion cloud belt. During this period there is a
clear and consistent diurnal variation of aerosol optical depth with highest values between 6:00 and 9:00
LTST. This diurnal variation is almost certainly
caused by a variation in water ice clouds, which is
superimposed on top of a low baseline of dust optical
depth. While there is a clear seasonal trend, there is
also significant variation from one sol to the next,
which is most likely caused by water ice clouds as
well.

The red band in Figure 1 at Ls=153° shows an
early-season regional dust storm. Figure 2 shows the
complex time history of aerosol optical depth during
this January 2022 event. Here, the sol numbers label
midnight LTST. During the active period of this dust
event between sols 313 and 317 (5–9 January 2022,
Ls=153°–156°) there were numerous spikes in aerosol optical depth with several exceeding unity (at 9
µm). These spikes in aerosol (dust) optical depth
occurred preferentially during the day but appear
equally in both the morning and the afternoon.
Returning to Figure 1, the latest retrievals shown
are from the season where it is expected that dust is
becoming the dominant aerosol over water ice
clouds. These observations taken after the regional
dust storm show a higher overall baseline of aerosol
optical depth (likely caused by dust), but with some
water ice clouds still present. The local time variation also appears to shift with the diurnal maximum
moving later toward the middle of the day, and there
appears to be more sol-to-sol and diurnal variability
in general.

Figure 2. The detailed time history of aerosol
(dust) optical depth retrieved from TIRS observations during the January 2022 regional dust
storm. The red points indicate times when the Sun
was within the field of view of TIRS and are
therefore less reliable.

Figure 1. The seasonal and diurnal variation of
aerosol optical depth retrieved from TIRS observations.

Summary: Observations taken by MEDA/TIRS
onboard the Perseverance rover enable total aerosol
optical depth to be retrieved systematically as a function of local time and season, including both during
the day and the night. The retrieved values show
clear seasonal and diurnal variations in aerosol optical depth. For water ice clouds, the two dominant
patterns involve the seasonal growth and decay of the
aphelion cloud belt and a consistent diurnal pattern
with greatest cloud optical depth between 6:00 and
9:00 LTST. A large regional dust storm in January
2022 highlights the utility of nearly constant retrievals of aerosol optical depth for better understanding
the complex evolution of dust opacity during the
active portion of a storm. Systematic TIRS observations continue as part of the baseline set of MEDA
observations on the Perseverance rover promising
exciting new information to come about the diurnal

variation of dust during the upcoming perihelion
season.

References:
Edwards, C.S. et al., 2021. Space Sci. Reviews,
217:77, doi:10.1007/s11214-021-00848-1.
Forget, F. et al., 1999. J. Geophys. Res., 104, E10,
24155–24176.
Lacis, A.A. & Oinas, V., 1991. J. Geophys. Res., 96,
9027–9063.
Millour, E. et al., 2018. “From Mars Express to
ExoMars” workshop. ESAC Madrid, Spain.
https://www.cosmos.esa.int/documents/1499429/
1583871/Millour_E.pdf
Rodriguez-Manfredi, J.A. et al., 2021. Space Sci.
Reviews,
217:48,
doi:10.1007/s11214-02100816-9.
Smith, M.D., 2004. Icarus, 167, 148–165.
Smith, M.D. et al., 2006. J. Geophys. Res., 111,
E12S13, doi:10.1029/2006JE002770.
Smith, M.D. et al., 2022. An overview of the Mars
atmospheric state as observed by EMIRS. 7th
Mars Atmosphere Modeling and Observations
conference, Paris.