EVIDENCE OF OBLIQUITY DRVEN CLIMATE FLUCTUATIONS ON
MARS FROM SMALL CRATER SURVEYS
M. E. Hoffman, H. E. Newsom, University of New Mexico, Albuquerque, NM, USA (mehoffman13@unm.edu),
F. Calef, Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA. J. P. Williams,
University of Los Angeles, Los Angeles, California, USA

Introduction: The obliquity, or axial tilt of
Mars has been theorized to fluctuate quasichaotically over semi-periodic timescales [1]. Modeling the obliquity history of Mars for the most recent 20 Ma is supported by measurements of the
layered deposits at the Martian poles [2]. For timescales beyond 20 Ma, the obliquity fluctuations become increasingly chaotic and impossible to predict
without additional constraints.
Fluctuations in the Martian obliquity result in
changes in the atmosphere as the CO2 ice reservoir at
the poles (largely concentrated at the south pole) is
coupled with the predominantly CO2 Martian atmosphere [2] (Figure 1). At higher angles of obliquity,
>40 degrees, the Martian poles experience higher
amounts of solar insolation. This results in greater
sublimation of the CO2 ice at the poles, which increases the amount of atmospheric CO2 and causes
an increase in the atmospheric pressure and density.
At lower obliquities, <40 degrees, the poles experience less solar insolation and are able to accumulate
CO2 ice at the poles. This scenario results in more
CO2 being removed from the atmosphere, which will
cause a decrease in the overall atmospheric pressure
and density. At more extreme angles >65 degrees,
ice begins to be deposited at the mid latitudes; however, the obliquity is thought to be <45 degrees for
the last 20 Ma, so this scenario will not be considered here.

Figure 1. Obliquity angle (degrees) vs. time (Myr) as determined
by Laskar et al., (2004) for the last 20 Myr (top) and the corresponding pressure (mbar) for obliquity from Williams et al.,
(2018).

The changes in obliquity and resulting changes in
atmospheric pressure can be seen indirectly through
geologic features such as small impact craters. For
the present-day Martian atmosphere, the limit for the
diameter of the smallest crater that can form at the
surface is theorized to be D ≈ 0.25 m [3,4,5]. Understanding the population of small decimeter- to decameter-diameter craters will lead to improved interpretations of the recent Martian atmospheric pressure.
Cratering Mechanics: Incoming projectiles
from space that encounter a planetary atmosphere
endure some degree of ablation, where material from
the projectile is vaporized and removed, and deceleration, where the projectile is slowed by the atmosphere [5]. Additionally, some projectiles experience
fragmentation, where one projectile may break apart
into several separate projectiles, often resulting in
multiple, or clustered, impact craters forming at the
surface [5,6].
The atmospheric pressure of Mars is less than 1
percent of the Earth’s, but it is still able to filter out
and remove small projectiles intersecting the orbital
path of Mars. Smaller projectiles are more susceptible to ablation and deceleration as these projectiles
have a larger surface area to volume ratio [5]. Smaller projectiles are also less likely to experience fragmentation, but possible fragmentation events will
still need to be considered [6,7].
Crater Catalog: The best resolution of HiRISE
orbital imagery is approximately 25cm/pixel. This
resolution is not sufficient to accurately resolve the
smallest craters at the surface. Using imagery taken
from rovers at the surface provides detailed context
of possible small crater candidates (Figure 2). An
extensive survey of the Mars Science Laboratory
(MSL) Curiosity rover traverse through the first
2300 sols of the mission (through the completion of
the Vera Rubin ridge campaign) was conducted to
compile a catalog of small craters [8,9]. Additionally, a survey of the first 500 sols of the Mars Exploration Rover (MER) Spirit rover traverse was conducted to compile a crater catalog of the Gusev Plains
and the start of the Columbia Hills region [10].
The MSL Curiosity survey found 198 craters
throughout the first 2300 sols of the traverse. There
were fewer than expected D < 1.0 m craters, and the
smallest crater found was D = 0.33 m [8,9]. Vasavada et al., (1993) predicted an abundance of small cm-

Figure 2. Nine examples of small craters seen throughout the MSL mission. Images are from the Navcam (black and white) and Mastcam
(color) instruments. Each white bar corresponds to an approximate length of 1.0 meter.

size craters across the Martian surface [11]. Over the
first ~20 km of the Curiosity traverse, there were
only five D < 0.50 m craters discovered. There is an
overall lack of cm-sized craters found at the surface
at Gale crater.
The MER Spirit rover survey found and documented a total of 267 small craters [10]. The smallest
crater measured was D = 0.23 m, making it the
smallest crater observed from the combined MER
and MSL catalogs thus far. Almost 80 percent of the
total small craters in this Spirit survey were found
within the first 155 sols of the traverse [10]. This
portion of the traverse was noticeably flatter and
smoother, allowing the Spirit rover to traverse greater distances in less time. Additionally, the flat terrain
allows for unobstructed crater formation and for such
craters to be identified more easily.
Erosion Rates: The small crater populations
gathered from each of the rover traverses can only
provide insight into the behavior of the recent Martian atmosphere for the lifetime of the craters. Erosion rates on Mars can be orders of magnitude slower than on Earth. Eventually, the small craters will be
removed by aeolian processes eroding down the rims
or infilling the central depression with fine material.
Bridges et al., (2012) estimated rates on the order

of 10-50 m/Myr for saltation processes eroding bedrock from analyzing the active dune field from Nili
Patera [12]. Golombek et al., (2014) determined erosion rates from studying small craters along the Opportunity rover traverse in Merdiani Planum to be
~1m/Myr for recently formed craters <1 Ma and ~
<0.1m/Myr for older craters 10–20 Ma [13]. Grant et
al., (2022) used measurements from the 27 mdiameter Homestead hollow, where the InSight
lander is located, to find degradation rates within the
hollow to be 10−4 m/Myr for regolith-covered lava
plains [14]. The InSight lander is located in Elysuim
Planitia, just 600 km to the north of Gale crater.
Newsom et al., (2015) analyzed small craters and
blocks within the first 360 sols of the Curiosity traverse and estimated aeolian erosion rates on the order
of ~0.01m/Myr from crater counts [15]. Golombek et
al., (2006) estimated average erosion rates for the
Columbia Hills portion of the Spirit traverse to be
∼3*10-5 m/Myr by analyzing small craters [16].
There are a wide range of erosion and degradation rates across Mars that reflect different erosional
processes on regional environments (Table 1). The
slower rates at Gale and Gusev reflect a slow deflation by aeolian processes that have likely persisted in
these locations for over a billion years [14,15,16].
Using these rates and known cratering mechanics

Bridges et al., (2012)
Golombek et al., (2014)
Grant et al., (2022)
Newsom et al., (2015)
Golombek et al., (2006)

Rate of Erosion (m/Myr)
10 - 50
1.0
10-4
0.01
3x10-5 - 10

Location
Nili Patera
Merdiani Planum
Elysium Planitia
Gale crater
Gusev cratered plains

Associated Rover/Lander
Opportunity rover (MER-B)
InSight lander
Curiosity rover (MSL)
Spirit rover (MER-A)

Table 1. Erosion/degradation rates (m/Myr) for various locations across Mars along with the corresponding studies and the
rover/lander associated with each location, if applicable.

for approximate depth to diameter ratios [17], it
would take approximately 20 Myr to fill a 1.0 m
diameter crater along the Bradbury Group at Gale
crater and approximately 600 Myr to fill a 1.0 m
diameter crater along the Gusev plains. The slower
erosion rates at the Gusev plains could be contributing to the increase in the number of small craters
identified along the Spirit traverse as the craters
would have longer lifespans before being infilled and
removed from the crater record (Figure 3).

Atmospheric Modeling: A previous study of
small crater production at the surface of Mars was
conducted by Williams et al., (2018) to find a noticeable shift in cratering for possible obliquity induced
atmospheric pressure changes [7]. Williams et al.,
(2018) used insolation and temperature equations
from Ward et al., (1974) to estimate the atmospheric
pressure increase from sublimating CO2 at the poles
experiencing higher solar insolation at higher past
obliquities [7,18]. For present conditions, Mars has
an average global pressure of ~6mbar, which allows
for greater production and retention of small craters
at the surface [7,11]. For an obliquity adjusted scenario, the atmospheric pressure increases, which
allows for a greater degree of filtering by the atmosphere and reduces the number of small impactors
that make it to the surface with enough speed to form
an impact crater [7].
For generalized Martian conditions and cratering
production rates, Williams et al., (2018) modeled the
cumulative crater frequency distributions for a constant 6 mbar atmosphere and an obliquity adjusted
atmosphere modeled for 20 Ma over 1.0 km2 as seen
in figure 4 [7].

Figure 3. A cumulative crater frequency distribution for the Curiosity and Spirit rover traverse surveys. The Spirit craters (green)
show a higher accumulation of D < 1.0 m craters, likely the result
of slower degradation rates at the Gusev Plains. All cratering plots
were made in CraterStats 2.0 [19].

Under present conditions on Mars, the smallest
crater at Gale crater could survive for at least 5 Myr,
meaning that the decimeter- to decameter-scale
crater catalog from the Curiosity and Spirit traverses
are likely to be reflective of the last 20 Myr of the
atmospheric history or longer. These small crater
populations could provide additional evidence to
support an atmospheric pressure increase corresponding to increased obliquity angles as illustrated
by Laskar et al., (2004).

Figure 4. A cumulative crater frequency distribution from Williams et al., (2018) demonstrating a shift in small crater production for a hypothetical atmospheric scenario where atmospheric
pressure is aloud to vary with obliquity. The obliquity adjusted
craters (blue) show decreased production than the scenario of a
constant 6 mbar Martian atmosphere (red) for 20 Ma.

The obliquity adjusted model illustrates a decreased
crater production when compared to a constant 6
mbar Martian atmosphere.
The small crater catalogs from the two rover
traverses can be compared to modern cratering rates
as measured by Daubar et al., (2013). Using erosion
rate estimates and approximate crater resurfacing
calculations of ~50 Ma for the Curiosity craters and
~500 Ma for the Spirit craters, both catalogs can be
compared to the empirical global Martian production
function [6,17]. The resulting cumulative sizefrequency plot is shown in figure 5.

Figure 5. Cumulative size–frequency diagram of the current
empirical global Martian production rate (black) from Daubar et
al., (2013) scaled to the area-time function discussed in [6] and the
Curiosity (purple) and Spirit (green) crater production functions
estimated from multiplying the survey areas by the small crater
lifespans determined from local degradation rates. Craters are in
√2 diameter bins.

The current production rate calculated by Daubar
et al., (2013) was determined by observing new impacts in before and after CTX imagery and is therefore unable to resolve craters smaller than D ≥ 2.0 m
[6]. The Curiosity craters indicate a cratering rate at
Gale crater that is similar to modern estimates, but
the Spirit craters indicate a much slower cratering
rate at Gusev. These results illustrate the regional
dependance of crater formation and retention across
Mars. The increased erosion rates at Gale crater may
be removing information about past atmospheric
properties, while Gusev may show evidence of
slowed crater production due to an increase in atmospheric pressure resulting from a prolonged increase in the Martian obliquity starting ~5 Ma (Figure 1). It is still important to use caution when comparing datasets of small craters from across Mars as
there are many variables involved in crater formation
and retention [6].

Conclusions: Observational, small crater surveys from the surface of Martian rover traverses can
be used to compare with modern cratering rates and
theoretical models for cratering productions to determine constraints on recent Martian atmospheric
fluctuations driven by changes in obliquity. Small
crater frequency distributions serve as an indirect
geologic proxy for changes in the Martian atmosphere. Future work will be conducted to refine the
theoretical crater production and retention rates for
Gale crater and the Gusev Plains, accounting for
their respective elevations and rates of degradation.
References: [1] Laskar J. et al. (2004) Icarus,
170, 343-364. [2] Laskar, J. et al. (2002) Nature, 419(6905), 375-377. [3] Horz F. et al. (1999)
Science, 285, 2105-2107. [4] Popova O. et al. (2003)
Meteoritics & Planet. Sci., 38, 905–925. [5] Williams J. P. et al. (2014) Icarus, 235, 23-36. [6] Daubar, I. J. et al. (2013) Icarus, 225, 506–516. [7] Williams J. P. et al. (2018), Meteoritics & Planet. Sci.,
53, 554–582. [8] Hoffman M. E. et al. (2019) LPSC
L, Abstract #3147. [9] Hoffman M. E. et al. (2019)
9th Mars Conf., Abstract #6371. [10] Shaffer S. J. et
al. (2021) LPSC LII, Abstract #2760. [11] Vasavada
A. R. et al. (1993) JGR., 98, 3469–3476. [12]
Bridges, N. T. et al. (2012) Nature, 485(7398), 339342. [13] Golombek M. P. et al. (2014) JGR, 119,
2522-2527 [14] Grant, J. A. et al. (2022) Earth and
Space Science, 9(2), e2021EA001953. [15] Newsom,
H. E. et al. (2015) Icarus, 249, 108-128. [16]
Golombek, M. P. et al. (2006) Journal of Geophysical Research: Planets, 111(E12). [17] Melosh H. J.
(1989) Oxford University Press. [18] Ward, W. R. et
al. (1974) Journal of Geophysical Research, 79(24),
3387-3395. [19] Michael G. G. and Neukum G.
(2010), Earth Planet. Sci. Lett., 294, 223–229.