Constraints on Martian Crustal Lithology from Seismic Velocities by InSight

Abstract

Analysis of data from the seismometer SEIS on NASA’s InSight mission has by now provided a wealth of information on the crustal structure of Mars, both beneath the lander and at other locations on the planet. Here, we collect the P- and S-wave velocity information for kilometer-scale crustal layers available up to now and compare it to predictions by rock physics models to guide the interpretation in terms of crustal lithology. Modeling is performed based on the Hertz-Mindlin model for un- or poorly consolidated sediments, Dvorkin and Nur’s cemented-sand model for consolidated sediments and Berryman’s self-consistent approximation to simulate cracked rocks. Considered lithologies include basalt, andesite, dacite, kaolinite, and plagioclase, and cementation due to calcite, gypsum, halite and ice. We use Gassmann fluid substitution to study the effect of liquid water instead of atmosphere filling the pores or cracks. Below the lander, available constraints are based on Ps-receiver functions and vertical component autocorrelations for SV- and P-wave velocities, whereas SH-reflections and SsPp phases provide additional information on SH- and P-wave velocities in the uppermost 8-10 km, respectively. SS and PP precursors at the bouncing point of the most distant marsquake contain information on crustal velocities at a near-equatorial location far from InSight. Surface wave observations from two large impacts as well as the largest marsquake recorded by InSight provide average crustal velocities along their raypaths, which are distinct from the body wave results. The subsurface structure beneath the lander can be explained by 2 km of either unconsolidated basaltic sands, clay with a low amount (2%) of cementation, or cracked rocks (e.g. basalts with at least 12% porosity). Within the range of lithologies considered, the seismic velocities can neither be explained by intact rocks, nor rocks with completely filled pores, e.g. by ice, nor by fluid-saturated rocks. Below, down to a depth of about 10 km beneath InSight, both P- and SV-wave velocities are consistent with fractured basaltic rocks or plagioclase of at least 5% porosity, depending on crack aspect ratios. About 10% of that porosity needs to have a preferred orientation to explain the observed anisotropy. For porosities exceeding 12%, the measured velocities would also be consistent with water-saturated rocks. The transition to higher velocities at about 10 km depth beneath InSight can be modeled by more intact material, i.e. a porosity reduction by 50% compared to the layer above, which can be achieved by either cementation or a lower initial porosity. The SV-velocities derived by surface waves down to 25-30 km depth, averaging over a large part of Mars, are consistent with basalts of a porosity of less than 5% or nearly intact plagioclase. They could also be explained by rocks with a higher porosity if pores are filled by ice, but that is unlikely for the whole depth range considered. The velocities at larger depth, i.e. below about 20 km beneath InSight and 25-30 km along the surface wave paths, are consistent with intact basalt.