Regolith Thickness on the Moon and Mercury: Insights from Orbital Imagery and 2D Numerical Modelling

REGOLITH THICKNESS ON THE MOON AND MERCURY: INSIGHTS FROM ORBITAL IMAGERY AND 2D NUMERICAL MODELLING. M. Joulaud1,2, P. Allemand1, J. Flahaut2, V.J. Langlois1, E. Füri2. 1Laboratoire de Géologie de Lyon: Terre, Planètes, Environnement (LGL-TPE), Université Claude Bernard Lyon1/CNRS/ENS, 2 rue Raphaël Dubois 69622 Villeurbanne, France ([email protected]), 2Centre de Recherches Pétrographiques et Géochimiques (CRPG), CNRS/Université de Lorraine, 15 rue Notre-Dame des Pauvres, 54500 Vandœuvre-lès-Nancy, France. Introduction: The regolith encompasses the non-cohesive granular layer of various materials present on most terrestrial planetary bodies of the solar system [1]. This layer is subject to impact cratering and a range of weathering processes. On the Moon and Mercury, the regolith covers most of the surface, making the bedrock largely inaccessible. Consequently, the regolith is the primary window for observing these two airless planetary bodies. The present study investigates regolith properties by combining two approaches: 1) High-resolution remote sensing imagery of the Moon and Mercury is used to analyse the morphology of small craters and infer locally the regolith thickness. Boulder density is manually assessed [2]. 2) 2D numerical modelling is used to assess the role of regolith thickness on cratering processes and to gain a better understanding of the final morphology of craters [3,4]. The study was initially conducted on the Moon as part of the Emirates Lunar Mission (ELM) [2,5,6]. The primary landing site of the ELM was the floorfractured crater Atlas (3.8 Gyr [6]) near Mare Frigoris, while the three backup landing sites were located in the following maria areas: Sinus Iridum (3.4 Gyr [7]), Oceanus Procellarum (1.3 Gyr [8]), and Lacus Somniorum (3.7 Gyr [9]). For comparison with the Moon, two types of terrain on Mercury are investigated to assess the hermean regolith thickness: the old intercrater plains (> 3.9 Gyr) vs. the young smooth plains (3.7-2.5 Gyr) [10]. Finally, a comparison between the remote sensing approach and the 2D numerical modelling is performed to establish a theoretical model of the vertical structure of the regolith on both airless planetary bodies. Data & methods: Remote sensing. On the Moon, LROC NAC high-resolution imagery (0.5-3 m/px) [11] and associated Digital Terrains Models [12] are used to study small craters and boulders. The small crater morphology [13,14] is used to infer the regolith thickness: the interior feature of concentric, flatbottomed, and central mound is measured to derive the regolith thickness from an empiric relationship based on laboratory experiments [13]. Boulders are then manually mapped on the images and compared to existing LRO Diviner [15] and Mini-RF [16] datasets. On Mercury, MDIS NAC high-resolution imagery (1-3 m/px) [17] is used to assess the morphology of small craters. Due to the lower quality and resolution of the data on Mercury compared to that on the Moon, the depth of bowl-shaped craters is used as a lower bound for the regolith thickness of the terrains (as these craters are hypothesised to form exclusively in regolith [13,14]), to compare with the small crater morphology measurements. Numerical modelling. 2D models using a discrete element method [18] are used to simulate a projectile impacting a two-layer granular medium. A brittle, non-cohesive layer (the regolith) overlies on a strong, cohesive layer (the bedrock) [3,4]. The control parameters of the simulations are the impactor’s diameter and velocity, the thickness of the upper layer and the mechanical strength of the two layers. The quantified outputs are the transient cavity depth, the final crater diameter, and the bedrock fragmentation [4]. Results & discussions: Table 1 summarises the results of the regolith thickness and boulder density assessment on the Moon and Mercury for all sites investigated. Site Age [Gyr] Regolith thickness [m] Boulders Atlas Crater 3.8 1.2 (1; 1.7) 7232 Sinus Iridum 3.4 2.9 (2; 3.9) 5664 Oceanus Proc. 1.3 1.7 (1.2; 2.4) 17322 Lacus Som. 3.7 1.8 (1.4; 2.3) 1794 Smooth p. 2.5 – 3.7 4.6 (3.3; 7.8) 2 Intercrater p. > 3.9 3.7 (2.6; 5.5) 10 Table 1: Summary of each site’s age, median regolith thickness with 1st and 3rd quartiles, and number of boulders. Blue: Moon. Orange: Mercury. NB: On Mercury, the “Boulders” value corresponds to the number of images presenting boulders. Regolith thickness. The morphology of small craters indicates that both on the Moon and Mercury, no correlation is observed between regolith thickness and surface age, thereby invalidating the hypothesis that older surfaces present a thicker regolith [19]. When comparing the regolith thickness of the Moon and Mercury, the hermean regolith appears thicker, as already proposed in previous studies [20,21] (Figure 1). On Mercury, the differences in regolith thickness could be attributed to differences in bedrock properties [22]. Figure 1: Violin plots of the regolith thickness (in metres) for the investigated areas, from youngest to oldest (left to right), on both the Moon (Oceanus Procellarum to Atlas crater) and Mercury (Smooth plains and Intercrater plains). Boulders. Manual counting of boulders on the Moon demonstrates a strong correlation with existing radar-derived boulder density data (LRO Diviner [14] and mini-RF [15]). However, no discernible correlation between regolith thickness and boulder density is apparent. On Mercury, the boulder density cannot be quantified precisely due to resolution limitations of the NAC images. Nonetheless, fewer boulders are observed compared to the lunar surface, with a higher concentration of boulders within the intercrater plains, which present older surfaces. Numerical modelling. When simulating an impact of a projectile on a two-layer granular material [9], the evolution of the transient cavity can be characterised by a threshold of regolith thickness, which separates the strength-dominated excavation regime from the gravity-dominated one. This threshold increases with impact velocity. Impacts of different energy can produce the same final crater diameter, depending on the regolith properties. Increasing the impact velocity results in a crater profile (at a different scale factor) similar to that produced by decreasing the regolith thickness, except for differences in bedrock fragmentation. The bedrock fragmentation, which is associated with regolith production, decreases until it reaches zero at a given regolith thickness threshold, at which point the bedrock ceases to be fragmented (within the range of impact velocities studied). The 2D DEM simulations provide important input parameters for numerical models of regolith growth [23,24,25]. Conclusions: The regolith serves as a tool for probing planetary surfaces, providing information on both the impact flux and the target properties. Based on observations of small craters and boulders, the target properties are analysed using remote sensing data and 2D numerical modelling, leading to the following conclusions: - The regolith thickness increases over time until a critical thickness is attained, at which point the regolith becomes self-shielding against further impacts (cover/blanket effect). - In the investigated areas, the regolith thickness does not appear to be correlated with surface age, thereby invalidating the hypothesis that the regolith thickness is a function of surface age. - The production of regolith is currently halted on the Moon and Mercury (probably since 3.9 Gyr?), because of a thick layer of regolith (making the bedrock inaccessible for a significant range of impact energies) and a low impact flux/low impact energy. Consequently, planetary-scale impacts or resurfacing processes are now required to produce new regolith. - A more competent bedrock will result in a thinner regolith with a higher abundance of boulders, while a less competent bedrock will result in a thicker regolith with fewer large boulders. References: [1] Bates R. L. and Jackson J. A. (1980) American Geological Institute. [2] Joulaud M. et al. (2024) Space Sci. Rev., 220, 65. [3] Langlois V. J. (2025) Submitted. [4] Joulaud M. et al. (2024) 55th LPSC. [5] Els S. G. and Al Marzooqui H. A. (2020) MBRSC. [6] Flahaut J. et al. (2024) Space Sci. Rev, 220, 53. [7] Hu T. et al. (2020) PSS 194, 105134. [8] Qian Y. Q. et al. (2018) JGR Planets, 123, 1407-1430. [9] Flahaut J. et al. (2021) 72nd IAC, IAC-21,A3,2A,7,x64211. [10] Ostrach L. R. et al. (2015) Icarus 250, 602–622. [11] Robinson M. S. et al. (2010) Space Sci Rev 150, 81–124. [12] Grumpe A. et al. (2014) ASR 53, 1735–1767. [13] Quaide W. L. and Oberbeck V. R. (1968) J. Geophys. Res. 73, 5247–5270. [14] Bart G. D. (2014). Icarus 235 (2014) 130–135. [15] Paige D. A. et al. (2010) Space Sci Rev 150, 125–160. [16] Bandfield J. L. et al. (2011) JGR, Vol. 116, E00H02. [17] Hawkins S. E. et al. (2007) Space Sci Rev 131, 247–338. [18] Pöschel T. and Schwager T. (2005) Springer. [19] Bart G. D. (2025) Planet. Sci. J. 6, 36. [20] Kreslavsky M. A. and Head J. W. (2015) 46th LPSC. [21] Zharkova A. et al. (2015) AGU Fall, P53A-2099. [22] Chertok M. A. et al. (2023) JGR Planets 128, e2023JE007767. [23] Oberbeck V. et al. (1973) Icarus 19, 87–107. [24] Hirabayashi M. et al. (2018) J. Geophys. Res. Planets 123, 527–543. [25] Zhang M. et al. (2023) JGR Planets 128, e2023JE008035.