Volume 51, Issue 2 e2023GL105490
Research Letter
Open Access

Evidence of Ice-Rich Layered Deposits in the Medusae Fossae Formation of Mars

Thomas R. Watters

Corresponding Author

Thomas R. Watters

Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA

Correspondence to:

T. R. Watters,

[email protected]

Contribution: Conceptualization, Methodology, Validation, Formal analysis, ​Investigation, Data curation, Writing - original draft, Writing - review & editing, Visualization, Supervision, Project administration

Search for more papers by this author
Bruce A. Campbell

Bruce A. Campbell

Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington, DC, USA

Contribution: Methodology, Formal analysis, Writing - review & editing

Search for more papers by this author
Carl J. Leuschen

Carl J. Leuschen

CReSIS, The University of Kansas, Lawrence, KS, USA

Contribution: Software, Formal analysis, Visualization

Search for more papers by this author
Gareth A. Morgan

Gareth A. Morgan

Planetary Science Institute, Tucson, AZ, USA

Contribution: Validation, Formal analysis, Visualization

Search for more papers by this author
Andrea Cicchetti

Andrea Cicchetti

IAPS, Istituto Nazionale di Astrofisica, Roma, Italy

Contribution: Software, Resources, Data curation

Search for more papers by this author
Roberto Orosei

Roberto Orosei

IRA, Istituto Nazionale di Astrofisica, Bologna, Italy

Contribution: Methodology, Validation, Resources, Writing - review & editing

Search for more papers by this author
Jeffrey J. Plaut

Jeffrey J. Plaut

JPL, California Institute of Technology, Pasadena, CA, USA

Contribution: Methodology, Validation, Resources

Search for more papers by this author
First published: 23 January 2024
Citations: 2

Abstract

Subsurface reflectors in radar sounder data from the Mars Advanced Radar for Subsurface and Ionospheric Sounding instrument aboard the Mars Express spacecraft indicate significant dielectric contrasts between layers in the Martian Medusae Fossae Formation (MFF). Large density changes that create dielectric contrasts are less likely in deposits of volcanic ash, eolian sediments, and dust, and compaction models show that homogeneous fine-grained material cannot readily account for the inferred density and dielectric constant where the deposits are more than a kilometer thick. The presence of subsurface reflectors is consistent with a multi-layer structure of an ice-poor cap above an ice-rich unit analogous to the Martian Polar Layered Deposits. The volume of an ice-rich component across the entire MFF below a 300–600 m dry cover corresponds to a global equivalent layer of water of ∼1.5 to ∼2.7 m or ∼30%–50% of the total estimated in the North Polar cap.

Key Points

  • Mars Advanced Radar for Subsurface and Ionospheric Sounding radar sounder data reveals layering in the Medusae Fossae Formation (MFF) deposits

  • Layers are likely due to transitions between mixtures of ice-rich and ice-poor dust, analogous to those in Polar Layered Deposits

  • An ice-rich portion of the MFF deposit may contain the largest volume of water in the equatorial region of Mars

Plain Language Summary

The Medusae Fossae Formation (MFF), located near the equator of Mars along the dichotomy boundary between the lowlands of the northern hemisphere and the cratered highlands of the southern hemisphere, is one of the largest and least understood deposits on Mars. The Mars Advanced Radar for Subsurface and Ionospheric Sounding radar sounder detects echoes in MFF deposits that occur between the surface and the base which are interpreted as layers within the deposit like those found in Polar Layered Deposits of the North and South Poles. The subsurface reflectors suggest transitions between mixtures of ice-rich and ice-poor dust analogous to the multi-layered, ice-rich polar deposits. An ice-rich part of the MFF deposit corresponds to the largest volume of water outside the polar caps, or a global equivalent layer of water of ∼1.5 to ∼2.7 m.

1 Introduction

The Mars Advanced Radar for Subsurface and Ionospheric Sounding instrument (MARSIS) onboard the European Space Agency's Mars Express spacecraft (Picardi et al., 2005) has successfully probed the ice-rich North Polar Layered Deposits (NPLD) and South Polar Layered Deposits (SPLD) (Picardi et al., 2005; Plaut et al., 2007), the Dorsa Argentea Formation (Whitten et al., 2020) the Hematite-Bearing Plains and Etched Plains deposits of Meridiani Planum (Watters et al., 2017), and the Medusae Fossae Formation (MFF) (Watters et al., 2007) (Text S1 in Supporting Information S1). The MFF is the most mysterious, and its origin the most controversial, of these deposits. The MFF deposits are broadly distributed along the dichotomy boundary (Figure 1) and are thought to be either volcanic ashfall deposits (Bradley et al., 2002; Hynek et al., 2003; Kerber et al., 2011; Ojha & Lewis, 2018; Ojha & Mittelholz, 2023), eolian sediments (Tanaka, 2000), dust deposits (Ojha et al., 2018), or an ice-rich deposit analogous to the PLD (Head & Kreslavsky, 2004; Schultz & Lutz, 1988). It has also been suggested that MFF consists of massive deposits of pumice that floated in a northern ocean and accumulated along the dichotomy boundary (Mouginis-Mark & Zimbelman, 2020). Dielectric properties derived from MARSIS and SHARAD radar sounder data do not rule out ice-rich MFF deposits (Campbell & Morgan, 2018; Carter et al., 2009; Orosei et al., 2015; Watters et al., 2007), with a possible 300–600 m thick insulating cover of dry sediment (Campbell et al., 2021).

Details are in the caption following the image

Medusae Fossae Formation (MFF) deposits along the Martian dichotomy boundary. The locations of Mars Advanced Radar for Subsurface and Ionospheric Sounding radargrams are indicated by white lines. Mars Orbiter Laser Altimeter shaded relief map with colorized elevation. The dark shaded areas show the MFF deposit boundaries modified from Tanaka et al. (2014) (multiple unit types were combined).

Central to the question of massive ice within the MFF is whether such a two-layer model is supported by internal density changes that create significant dielectric contrasts, and whether the overall dielectric behavior with thickness can be explained by a self-compacting, ice-poor sedimentary unit of uniform grain size. Early MARSIS work showed reflecting interfaces in just two locales (Watters et al., 2007). An absence of internal dielectric contrasts would be expected for an ice-poor volcanic ash, eolian sediment, or dust unit because the depositional mechanisms involved are less likely to result in significant density/dielectric layering contiguous over horizontal distances comparable to the MARSIS footprints (i.e., 10 s of km). For the case of an ice-poor sedimentary unit, the bulk density is dependent on the particle density, the porosity of the deposit, and the degree of compaction.

We present two-dimensional subsurface profiles from MARSIS sounding data that show strong radar reflections correlated with abrupt density/dielectric changes over lateral distances up to 100 s of km. Clutter simulations confirm that these echoes likely do not come from off-nadir reflections by surface topography. We also derive a new estimate of the maximum thickness of the MFF deposits and compare their bulk real dielectric constant (inferred from the value required to yield a projected basal boundary consistent with the regional slope) to the results of compaction models for ice-free deposits with a range of grain size and initial porosity.

2 Results

MARSIS SS3-mode data acquired over the last decade and targeted Super-Frame Mode SFM data (see Text S2, Figure S1 in Supporting Information S1) acquired over the last 3 years provide new insight into the nature of the MFF deposits. The three largest contiguous MFF deposits are Lucus Plunum (∼5°S, 185°E), Medusae Fossae-Eumenides Dorsum (∼0°N, 200°E), and Amazonis Mensa-Gordii Dorsum (∼0°N, 215°E) where deposits extend to Gigas Fossae (Figure 1). A survey of MARSIS data shows evidence of layering in all three units (Figures 1 and 2). In each case, the depth below the surface is estimated from the time delay Δt using the speed of light corrected by the mean dielectric constant (see Text S3 in Supporting Information S1).

Details are in the caption following the image

Radar data collected in SS3-mode and SFM-mode over Medusae Fossae Formation (MFF) deposits and the North and South Polar Layered Deposits (SPLD). (a) SS3-mode data from orbit 10216, band 3 over Amazonis Planitia where the vertical axis shows round-trip time delay, Amazonis Mensa is indicated by AM, Gordii Dorsum by GD, and valley between them as A-G V (see Figure 1). Subsurface echoes designated AM1 and AM2 are interpreted to be from internal MFF layers (b) the same data from orbit 10216 where the subsurface echoes have been depth corrected to a vertical scale in meters using a real dielectric constant ε′ of 3. Depth corrected radargram shows echo AM3 aligns with echo A-GV1, both interpreted to be from the MFF base. A-GV2 is below the MFF base and may be from earlier deposits of MFF material, (c) data from orbit 06487, band 4 over the SPLD, depth corrected using a ε′ of 3, (d) data from orbit 14019, band 3 over the North Polar Layered Deposits, depth corrected using a ε′ of 3, (e) SFM-mode data from orbit 19681, band 2 over a short segment of Amazonis Mensa in round-trip delay time format, (f) data from orbit 18703, band 4 over Lucus Planum in time-delay format, and (g) data from orbit 18664, band 3 over Eumenides Dorsum in time delay format. Subsurface echoes (white arrows) are offset in time-delay from the surface echo and are interpreted to be either nadir reflections from the interface between layers in the MFF deposits or between the MFF base and the lowland volcanic plains material.

Of the large MFF deposits, the subsurface of Amazonis Mensa is the most striking with multiple echoes interpreted to be evidence of layering (Figures 2a and 2b). At least two, and possibly more, undulating echoes suggest a complex sequence or group of layers in Amazonis Mensa. Multiple internal echoes in similar patterns have been found in both the SPLD (Figure 2c) and NPLD (Figure 2d) (Picardi et al., 2005; Plaut et al., 2007). The most prominent echoes are at Δt ∼ 4.3 μs (AM1) and ∼9.9 μs (AM2) in SS3 orbit 10216, interpreted as two dielectric interfaces above a basal contact at ∼17.0 μs (AM3) (Figure 2a, Figure S2 in Supporting Information S1). The reflector at Δt ∼ 17.0 μs (AM3) is continuous with an ∼4.3 μs (AG-V1) echo beneath deposits in the extensive lower-elevation area separating Amazonis Mensa and Gordii Dorsum (Figures 2a and 2b, Figure S2 in Supporting Information S1). A reflector identified in SHARAD orbit 24211_1 (Lalich et al., 2022) that crosses the valley, further to the north where the MFF deposits thin, correlates with the basal contact found in 10216 and adjacent orbit 13416 (Figures S4a and S4b in Supporting Information S1). The deeper reflector (A-GV2) in the region, also noted in MARSIS track 4117 (Watters et al., 2007), is below that SHARAD-detected base (Figure S4D in Supporting Information S1) and may be linked with earlier deposits of MFF material (Morgan et al., 2015). Evidence of multiple layers in Amazonis Mensa is found in the northwest (Figure S5 in Supporting Information S1) and northeastern flanks of the deposits (SFM orbit 19681, Figure 2e). Here, two subsurface echoes, at Δt ∼ 4.3 μs and at ∼8.5 μs, occur above a basal echo at ∼14.2 μs (Figure S6 in Supporting Information S1).

Although MARSIS and SHARAD both detect the Δt ∼ 4.3 μs (AG-V1) reflector in the low-elevation area and trace it under the Amazonis Mensa massif, SHARAD (Figure 2a) does not detect dielectric interfaces at smaller delays to match those in the MARSIS data. This intriguing result may be connected with the very different wavelengths (∼60–100 m for MARSIS and ∼20 m for SHARAD) and delay resolutions (1-μs for MARSIS vs. 0.1 μs for SHARAD) of the two instruments, potentially leading to different patterns of multi-layer echo enhancement within a delay cell (Campbell & Morgan, 2018; Lalich et al., 2022). In simple terms, we propose that the MARSIS-observed reflections are likely due to a favorable vertical spacing of density changes that reinforce the radar echoes at longer wavelengths over the larger delay cells.

Evidence of layering is also found in the easternmost MFF deposits near Gigas Fossae (Figure 1, Figure S7 in Supporting Information S1), and in Lucus Planum. Two distinct subsurface echoes at Δt ∼ 8.5 and ∼12.8 μs are detected in SS3 orbit 18703 (Figures 1 and 2f) (Figure S8 in Supporting Information S1). These are interpreted, respectively, to be echoes from a dielectric contrast at intermediate depth and what we term the “basal” interface, which may lie between the MFF and local plains deposits or simply the deepest penetration of a layered MFF sequence.

In the thinner, westernmost MFF deposits of Zephyria Planum (∼0°N, 153°E) (Figure 1), evidence of an internal reflector is less definitive. A subsurface echo in SFM orbit 19738 at Δt ∼ 5.7 μs may be from a dielectric contrast or a strong sidelobe of the surface echo (Figure S9 in Supporting Information S1). However, an internal reflector is present in the Zephyria Planum deposits in SHARAD orbit 22011 (Campbell et al., 2021).

Evidence of internal layering is not present in all radargrams traversing MFF deposits, due either to the absence of a significant dielectric contrast along track, or lower MARSIS signal gain resulting from spacecraft altitude or ionospheric loss.

2.1 Maximum Thickness and Loss Tangent

Subsurface echoes are also detected in the thick MFF deposits of Eumenides Dorsum (Figure 1). A subsurface echo in SS3 orbit 18664 at Δt ∼ 5.7 μs is interpreted to be a shallow-depth interface above a deeper (Δt ∼ 20.6 μs) basal echo (Figure 2g, Figure S10 in Supporting Information S1). Deeper subsurface echoes in Eumenides Dorsum are found in MARSIS SS3 orbits 13240 and 15423 (Figures 3a and 3b). Echoes at Δt ∼ 30 μs occur above deeper, apparent basal echoes in these orbits. The basal echoes from MFF dielectric interfaces identified here are at greater time delay than earlier studies which estimated a maximum penetration depth of 2.5 km over Eumenides Dorsum (Watters et al., 2007). Orbits 13240 and 15423 cross the highest elevations of Eumenides Dorsum and show a sharp transition to a diffuse echo pattern that likely demarks the contact between the MFF deposits and the lowlands volcanic plains (Tanaka, 2000; Watters et al., 2007) (Figures 3c and 3d).

Details are in the caption following the image

Radar data over Eumenides Dorsum. Radargrams showing Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS) SS3-mode radargrams from orbits 13240, band 2 (a) and 15423, band 2 (b) with round-trip time delay on the vertical axis. The lower two figures (c) and (d) show the same MARSIS tracks where the subsurface echoes have been converted to vertical distance using a real dielectric constant of 3. The time delays to the top of the diffuse basal echo (∼40 μs) are the largest yet found in sounder data in the Medusae Fossae Formation (MFF) deposits. The runout of power with time delay suggests a very rough subsurface interface at the base of the MFF or localized high loss. A subsurface reflector above the basal interface is interpreted to be an internal layer (∼30 μs).

The measured Δt between the surface returns and the basal echoes range from ∼39.0 to 41.8 μs (Figures S11 and S12 in Supporting Information S1) with a mean of 40.4 μs and a standard deviation of 1.1 μs (n = 8). The maximum relief of MFF deposits above the lowlandsvolcanic plains is estimated using Mars Orbiter Laser Altimeter (MOLA) topography. The relief is measured between the maximum elevation of the MFF deposits and the lowlands volcanic plains along MOLA profiles, with a mean value along the ground track of orbits 13240 and 15423 of ∼3,700 ± 90 m (Figure S13 in Supporting Information S1). This is consistent with MOLA-based estimates of the maximum thickness of MFF deposits of Eumenides Dorsum by Hynek et al. (2003) of at least 3.5 km. The only other deposit the MARSIS radar sounder has penetrated to a depth of 3.7 km is the SPLD (Plaut et al., 2007). The elevation-time delay relationship based on the measurements of maximum relief corresponds to a mean ε' of ∼2.7 ± 0.2 (n = 4) (see Text S3 in Supporting Information S1), in agreement with previous estimates of 2.9 ± 0.4 based on measurements using MARSIS data over thicknesses up to ∼2,500 m (Watters et al., 2007) and ε' of 2–3 using SHARAD data (Campbell et al., 2021; Carter et al., 2009). From the echo power at 3,700 m (using band 3) and previous MARSIS measurements (Watters et al., 2007), the attenuation of the MFF is ∼0.004 dB/m with a range (from the standard deviation of the slope) of ∼0.002–0.005 dB/m. This range in attenuation corresponds to a range in loss tangent of ∼0.002–0.004. A loss tangent of 0.003 is consistent with previous estimates using MARSIS data (Watters et al., 2007) and those obtained using SHARAD data (Campbell et al., 2021; Campbell & Morgan, 2018). Estimates of the loss tangent of the SPLD range from ∼0.001 to 0.005 (Plaut et al., 2007).

We interpret the dielectric interface seen by both sounders at 4.3 μs in the region between Amazonis Mensa and Gordii Dorsum, and extending to greater time delay under Amazonis Mensa, as the contact between the main MFF deposits and earlier plains-forming flows. The MARSIS reflections at Δt ∼ 4.3 and 5.7 μs (∼370 and ∼490 m depth) within Amazonis Mensa are attributed to multiple, roughly horizontal density changes within a 1-μs delay cell that coherently interfere to produce a strong echo at 60–100 m wavelength. We propose that this zone of density changes marks the transition between a 300–600 m thick dry upper cap proposed from SHARAD data and an ice-rich lower unit analogous to the NPLD and SPLD. Reflectors at similar depth in Lucus Planum (Figure S8 in Supporting Information S1), Zephyria Planum (Figure S9 in Supporting Information S1), and Medusae Fossae-Eumenides Dorsum (Figure S10 in Supporting Information S1) may indicate a similar change in ice content.

2.2 Compaction Behavior

The primary alternative hypothesis remains an ice-poor material of some grain size that has low porosity at the upper surface. At grain sizes typical of sand, volcanic ash, or silicate dust particles, a deposit will significantly compact over thicknesses of several kilometers. This self-compaction results in a decrease in porosity and an increase in bulk density, causing a change in the depth-integrated electrical properties (i.e., the average real dielectric constant) of the material (Campbell et al., 2021; Morgan et al., 2015; Ulaby et al., 1988; Watters et al., 2017). The compaction behavior of example materials is evaluated to 3,700 m, the maximum thickness described above (Figure 4) (see Text S4 in Supporting Information S1).

Details are in the caption following the image

Compaction models for three ideal materials that simulate those commonly proposed for Medusae Fossae Formation (MFF) deposits. Porosity (a), density (b), and apparent real dielectric constant (c) curves for three geologic materials: a loose basalt sand (blue large dash), a volcanic ash (red small dash and dot), and a silicate or rock dust (gray small dash). The compaction and density models incorporate the compressibility of the materials Mars gravity, and the apparent real dielectric constant accounts for the two-way delay time through a material with an increasing index of refraction with depth due to compaction (see Text S4 and Table S2 in Supporting Information S1). The error bars show the effect of ±5% variation in the initial porosity. The gray areas in C shows the range in estimates of ε′ for the MFF deposits (2.5–3.3) and the range in estimates of the maximum thickness of the deposits (2.5–3.7 km). No modeled material has predicted real dielectric constants within the observed range (double shaded area).

Typical physical properties of these materials (i.e., compressibility k, initial porosity ϕo, particle density ρp) are given in Table S2 in Supporting Information S1. Dust-sized particles experience the most reduction in porosity and increase in density with depth (Figures 4a and 4b). Variations in the assumed initial porosity of ±5% do not significantly change the compaction profiles, particularly at depths corresponding to the maximum MFF thickness (Figures 4a and 4b).

The real dielectric constant ε′ of a fine-grained material as a function of depth can be estimated given a porosity and density profile, because in a compacting deposit the two-way travel time is related to the integral of the effective speed of light over the unit depth (see Text S5 in Supporting Information S1). Model results indicate that a 3,700 m thick sequence of sand-sized particles has an apparent ε a > 4 ${\varepsilon }_{a}^{\prime }\, > \,4$ while dust or volcanic ash have an ε a > 5 ${\varepsilon }_{a}^{\prime }\, > \,5$ (Figure 4c). This confirms that an ice-poor deposit of uniform grain size cannot readily account for the inferred ε′ of the MFF deposits of 2.9 ± 0.4, and that a large fraction of the thickness beyond a few hundred meters must comprise a minimally compressible material with real dielectric constant of ∼3 (Campbell et al., 2021; Watters et al., 2007). We further considered a deposit of pumice sand, which likewise did not satisfy the observations (see Text S4 in Supporting Information S1).

3 Discussion

Our conclusion, reinforcing the SHARAD evidence for a two-layer model with significant ice-rich material at depth would require deposition of a PLD-like deposit at the Martian equator during periods of high obliquity (Forget et al., 2006; Head & Kreslavsky, 2004; Schultz & Lutz, 1988). This is supported by several lines of evidence: rhythmic layering in MFF deposits in HiRISE image surveys (Khan & Lewis, 2023); modeling of the Martian paleoclimate indicating that at high obliquity the distribution of stable ground ice and regions of ground-ice stability extends to equatorial latitudes (Aharonson et al., 2022); and epithermal neutron data from the Mars Odyssey Neutron Spectrometer (Wilson et al., 2017), suggesting >40% water equivalent hydrogen (WEH) in some locales (Feldman et al., 2011). Recent neutron data collected by the FREND instrument onboard ESA's Trace Gas Orbiter also suggest high WEH in some locations near the MFF deposits (Malakhov et al., 2020). The apparent abundance of WEH in the upper tens of centimeters in the MFF near-surface where it is expected to be dry may be due to local enhanced diffusion of water vapor from the deep, ice-rich portion of the deposit. A layer of dust or pyroclastic ash (Campbell et al., 2021; Wilson et al., 2017) could provide the insulating material that preserves the water ice in MFF deposits.

If the lower unit of the MFF is a mix of basaltic dust and water ice, at an upper range of ε′ of 3.3 estimated for the deposits (Watters et al., 2007), the volume fraction of dust is likely <20% (see Text S6, Figure S14 in Supporting Information S1), a potentially greater fraction than in the PLD (Plaut et al., 2007). The volume of water in the ice-rich layer derived by subtracting 300–600 m of cover from the total volume of the MFF deposits is ∼220,000 to 400,000 km3, or ∼30%–50% of the total estimated water in the NPLD (Brothers et al., 2015). This corresponds to a global equivalent layer of ∼1.5 to ∼2.7 m (Figure 5, Table S3 in Supporting Information S1). Thus, an ice-rich portion of the MFF deposit may contain the largest volume of water in the equatorial region of Mars.

Details are in the caption following the image

Thickness map of the suspected ice-rich portions of the Medusae Fossae Formation (MFF) deposits. MFF deposit boundaries are modified from Tanaka et al. (2014) (multiple unit types were combined and some of the boundaries adjusted). The paleo (pre-MFF) surface was estimated, and 300 and 600 m of dry cover subtracted to estimate the thickness of the proposed ice-rich layer. The total ice volume of is estimated to be ∼2.2 × 105 km3 (600 m removed) to ∼4.0 × 105 km3 (300 m removed) which corresponds to a GEL of water of ∼1.5–∼2.7 m (see Table S2 in Supporting Information S1). The locations of reflectors in orbits 13240 and 15423 (Figures S10 and S11 in Supporting Information S1) correspond to the thickest sections of MFF deposits in Eumenides in both the 300 and 600 m cover maps. The greatest volume of the total water in the MFF deposits is in Eumenides Dorsum (Table S2 in Supporting Information S1).

Acknowledgments

We wish to thank the editor and two anonymous reviewers for their comments and suggestions that greatly improved the manuscript. MARSIS is managed by the Agenzia Spaziale Italiana (ASI) and the National Aeronautics and Space Administration (NASA). The Mars Express mission is managed and operated by the European Space Agency. This work was supported by the Italian Space Agency (ASI) through contract 2019-21-HH.0.

    Data Availability Statement

    Datasets for this research are available in these in-text data citation references: MARSIS (2023) and SHARAD (2023). Processed radargrams and other data are available on the Smithsonian's Figshare site (Watters, 2023).