Does Bennu contain a near-surface, fine-grained layer?

Introduction:

Recently, Bierhaus et al. (2023) argued for the presence of a global, near-surface layer (~1-4 m thick) of fine-grained material (particles between approximately 100 μm and 1 cm) on asteroid Bennu, based partly on the presence of small (20 m diameter) craters with smooth interiors. This layer may have evolved due to percolation of fines into the subsurface by dynamic processes such as seismic shaking from impact cratering and/or mass movement. If present, a subsurface layer with a distinct particle size-frequency distribution (PSFD) may have different physical properties (e.g., cohesion, friction angle) that affect the evolution of Bennu’s surface—a stronger, more cohesive layer at depth could, for example, stabilize the near-surface and inhibit creep-style mass movement relative to an unlayered target.

We have therefore undertaken a study to test the hypothesis of a global, near-surface, fine-grained layer on Bennu, and to determine how such a layer would affect mass movement on the asteroid’s surface. Our investigation includes: (1) mapping smooth exposures, characterizing their geologic setting and relative stratigraphy, and determining whether they are consistent with such a layer; (2) carrying out laboratory experiments simulating mass movement in granular targets of varying properties and structures; and (3) complementing laboratory experiments with numerical simulations of surface creep under Bennu’s gravity using the Soft-Sphere Discrete Element Model (SSDEM) code PKDGRAV.

Confirming the presence of layering and its role on surface evolution would further our understanding of both rubble pile asteroid interior structure and rubble pile resurfacing—such understanding could help to address a question regarding the mismatch between rubble pile NEA surface ages (~few to tens Ma) and the proposed breakup of their parent bodies (~1 Ga) (Bierhaus et al., 2022; Bottke et al., 2015; Cho et al., 2021).

Results:

Mapping Smooth Regions: Our first task involves identifying all exposures of smooth material on Bennu. We identified smooth regions using two methods, (1) morphologic identification using the ~5 cm/pix Detailed Survey global mosaic (Bennett et al., 2021), and (2) automated identification of regions with low facet tilt variation (<4º)—a proxy for surface roughness—on 20 cm DTMs. This dual mapping technique (Figure 1) yielded abundant smooth exposures >1 m diameter, with complementary results across both methods. Both methods identified smooth exposures in similar regions; the automated DTM search found more small exposures, while the morphologic mapping better distinguished apparent exposures of fine-grained materials from other features such as smooth surfaces of boulders. We find that smooth exposures are concentrated within the Smooth Unit of the global geologic map (Jawin et al., 2022), with relatively fewer exposures in the Rugged Unit. Our next step is to determine the stratigraphic position of smooth exposures—are fine-grained regions consistent with exposure of a subsurface layer, or is smooth material present at the surface?

Figure 1. (A) Distribution of morphologically smooth exposures (blue circles) compared to geologic unit boundaries (yellow outline). Smooth regions are concentrated in the Smooth Geologic Unit. (B) Distributions of low-tilt regions found with OLA, closely matching the results of morphologic mapping.

Laboratory experiments: Our second task involves a series of experiments simulating mass movement in granular targets. For each experiment, we fill a Plexiglass box with different target materials and slowly raise one edge of the box using a hand crank. The box is tilted until the target material fails (Figure 2). We record each run using high-speed cameras and take still images and 3D scans throughout each run. We vary unlayered and layered targets (2 and 3 layers) using lava sand/gravel, play sand, and colored fishtank gravel.

Our results show evidence of size sorting associated with mass movement, with coarse material accumulating on the distal ends of deposits, and finer-grained material becoming exposed upslope. Failure events are observed to enhance the percolation of fines.

Figure 2. Time series of 2-layer experiment. This shows coarse lava gravel (>1/4”) on play sand.

Numerical assessments: Our third task involves a suite of SSDEM simulations similar to the setup of our laboratory experiments, but with a pre-defined range of particle sizes and SFD under Bennu gravity. Initial results show that PSFD significantly influences the bulk mechanical properties of the granular beds, in particular smaller particles can reduce the overall shear strength of the bed, although bulk cohesion remains negligible for cm-scale grains and becomes detectable only when particle sizes decrease to mm-scale. Additionally, our simulations show that slow landslides can effectively drive downward percolation of fine particles via kinetic sieving of fines. Even low-velocity landslide events can reorganize regolith under Bennu’s extremely weak gravity and low effective normal stress, leading to modification of the surface PSFD and stratification of the granular bed (Figure 3). In addition, segregation efficiency appears to increase with particle size contrast and increasing granular bed depth. Taken together, these results show that slow, creep-style mass movement is an efficient mechanism for producing vertical particle size heterogeneity, which could potentially lead to layering within Bennu’s near-surface—although does not exclude a well-mixed interior.

Figure 3 (Left) Modeling results showing changes in surface particle SFD and (right) size segregation due to landslides under Bennu gravity

Current status of findings:

Our findings show multiple lines of evidence that mass movement can size-segregate particles by moving larger particles to the surface and driving finer-grained materials into the subsurface. However, our results do not currently indicate that fine-grained materials will concentrate in a subsurface layer—rather, they appear to become mixed with coarser particles.

References: Barnouin, OS et al. Nat. Geosci. 12, 247-252 (2019). Bennett, CA et al. Icarus, 357, 113690 (2021). Bierhaus, EB et al. Nat. Geosci. 15, 440–446 (2022). Bierhaus, EB et al. Icarus 406, 115736 (2023). Bottke, WF et al. Icarus 247, 191–217 (2015). Cho, Y et al. JGR: Planets 126, e2020JE006572 (2021). Jawin, ER et al. Icarus 381, 114992 (2022). Scheeres, DJ et al. Science Advances 6, eabc3350 (2020).