Revealing the Stratigraphic Architecture and Composition of the North Polar Basal Unit on Mars with Multiband Radar Analyses

Introduction:  The basal unit (BU) is an ice-rich sedimentary deposit within the Planum Boreum (PB) on Mars lying between the Late Amazonian North Polar Layered Deposits (NPLD) and the Late Hesperian Vastitas Borealis interior unit  (Fig. 1, [1-4]). It consists of two subunits, rupēs and cavi [1-5], and represents a record of polar geologic processes and climate events spanning most of the Amazonian Period (~3.3 Ga, [4, 6]). Despite numerous recent studies, several key questions remain unanswered regarding the BU nature [6, 7]:

• Structure and stratigraphy. How are the cavi and rupēs units related? What is the geometry of the erosional unconformity between them? What is their extent and volume?
• Climate and composition. Ref. [5] hypothesized that the cavi unit is made of alternating sand sheets and pure water ice remnants of former polar caps, and thus is a record of the interplay between volatiles and sedimentary processes. In comparison, very little is known about the rupēs unit. What are its volatile and lithic components? Does the nature of the lithic fraction in rupēs differ from that of cavi?

It is now possible to answer these outstanding questions thanks to advancements in data processing techniques and the extensive, dense coverage of radar sounding profiles across PB.

Figure 1: (a) Topographic map of PB and surrounding terrains. (b) Stratigraphy of PB units, modified after [4]. The white line delineates the location and orien-tation of the profile in Fig. 2.

Methods: We use a recently released set of Mars Advanced Radar for Subsurface and Ionosphere Sounding (MARSIS, [8]) profiles that solves ionospheric distortions, resulting in improved quality of radar returns [9]. Thanks to its low operating frequencies (1-5 MHz), MARSIS is capable of penetrating through the entire thickness of the BU (Fig. 2), and thus is the key to fully reconstruct the distribution of the rupēs and cavi units underneath PB. We used the Seisware interpretation suite to map radar reflectors corresponding to the base of the BU and the contact between the cavi and rupēs units across over 500 MARSIS profiles at 3-5 MHz. We then applied inversion techniques to determine the real and imaginary parts of the dielectric permittivity of the rupēs unit following previously established approaches [5, 6, 10, 11] to obtain new insights on its composition.

Figure 2: (a) Original and (b) interpreted sample of MARSIS profile 9548 (location in Fig. 1).

Results: We found that the rupēs unit extends underneath the entirety of the western half of PB and part of Olympia Planum as one continuous unit, and has a upper unconformity with cavi following a pole-facing sloping geometry (Fig. 2). The rupēs unit occupies a volume of 191,000 km³, representing ~53% of the BU. We measured a real dielectric permittivity across the rupēs unit of ε’ = 4.00±0.85 (3 MHz), ε’ = 4.14±0.84 (4 MHz), and ε’ = 4.08±0.78 (5 MHz). We find that the permittivity is spatially heterogeneous (driving apparent uncertainty) and increases moving towards Hyperborea Lingula (at the floor of Chasma Boreale, Fig. 1), where it reaches its maximum values exceeding ε’ = 6. We measured a loss tangent across the rupēs unit of tanδ = 0.017±0.006 (3 MHz), tanδ = 0.013±0.006 (4 MHz), and tanδ = 0.012±0.006 (5 MHz). The loss tangent also increases towards Hyperborea Lingula, where it reaches tanδ >0.02. We note that the base of Hyperborea Lingula is difficult to detect, especially at 5 MHz.

Figure 3: Ternary diagram with possible ice and lithic mixture for the rupēs unit with plotted results of die-lectric permittivity inversions. The white shade repre-sents overlapping real permittivity and loss tangent results.

Discussion: Our initial analysis of the rupēs unit complex permittivity suggests that its composition differs substantially from that of the cavi unit [5], with large loss tangent values indicating the presence of significant amounts of lithic materials. Basalt alteration products with large loss tangents (i.e., tanδ >0.02), such as ferric oxides and/or hydrated minerals [13, 14] are required to explain the high loss tangent measurements, while the strong frequency dispersion of water ice imaginary permittivity [e.g., 12] explains the observed frequency dependence. We find a best match between real dielectric permittivity and loss tangent inversion results using a mixture of 10-15% gypsum and basalt alteration products and 85-90% water ice (Fig. 3). This is further supported by detections of ferric oxides on Mars [e.g., 15], and poly-hydrated Ca, Mg, and potential Fe sulfates at BU exposures [16, 17]. Rupēs materials may have been transported from lower latitude sources [4], where aqueous alteration is more viable than at polar latitudes. However, the strong spatial heterogeneities suggest that significant alteration occurred in situ during the Amazonian period, as previously proposed by [16], perhaps facilitated by warmer high-obliquity periods predicted to occur during the last 3 Gyr [18]. Finally, the high loss tangent measured in Hyperborea Lingula explains the lack of rupēs basal detections by SHARAD [5, 6, 19] despite the relatively low thickness (i.e., 150-200 m) of the rupēs unit at that location [4, 5].

Acknowledgments:  This work was supported by NASA MDAP grant 80NSSC22K1079. SHARAD is provided and operated by the Italian Space Agency (ASI). We are grateful to SeisWare Inc., for providing software licensing.

References: [1] Byrne and Murray (2002) JGR: Planets. [2] Fishbaugh and Head (2005) Icarus. [3] Putzig et al. (2009) Icarus. [4] Tanaka et al. (2008) Icarus. [5] Nerozzi and Holt (2019) GRL. [6] Nerozzi (2021) Icarus. [7] MEPAG Science Goals document. [8] Jordan et al. (2009) PSS. [9] McMichael et al. (2017) 2017 IEEE RadarConf. [10] Campbell et al. (2008) JGR: Planets. [11] Grima et al. (2009) GRL. [12] Fujita et. al. (2000) Physics of Ice Core Records. [13] Stillman and Olhoeft (2008) JGR: Planets. [14] Mattei et al. (2022) EPSL. [15] Bibring et al. (2006) Science. [16] Massé et al. (2012) EPSL. [17] Das et al. (2022) Icarus. [18] Laskar et al. (2004) Icarus. [19] Seu et al. (2007) JGR: Planets.