Geology of the InSight landing site on Mars

InSight landed in western Elysium Planitia on November 26, 2018. Because the payload focuses on Mars’ interior, understanding the landing site’s regional setting and shallow subsurface is crucial for interpreting geophysical data (SEIS seismometer, HP3 mole, and RISE tracking). The site lies near the martian crustal dichotomy, between ancient Noachian highlands and the younger northern lowlands. Western Elysium Planitia is bounded by highlands, the Medusae Fossae Formation, lavas from Elysium Mons to the north, and very young Cerberus Fossae lavas to the east. Prior to landing, mapping classified the plains as Early Amazonian–Hesperian, about 200 m of basaltic lava flows over weaker, likely Noachian sedimentary rocks. The key questions include the surficial geology, near-surface stratigraphy and mechanical properties, crater degradation state and age, and how these factors affect and contextualize in situ seismic and heat-flow measurements.

The study builds on: (1) prior orbital geologic mapping that classified the landing region as Early Hesperian Transition (eHt) with uncertain volcanic vs sedimentary origin; (2) extensive context of previous Mars landings (Viking, Pathfinder, MER Spirit/Opportunity, Phoenix, MSL) that characterized soils, rock distributions, and aeolian processes; (3) regional volcanism and tectonism of Elysium and Cerberus Fossae, with very young lava flows down Athabasca Valles and expected seismicity; (4) pre-mission assessments predicting >3 m of impact-fragmented regolith above basaltic lavas and wrinkle-ridge deformation; and (5) established rock size-frequency models and thermophysical methods for inferring particle sizes and cementation, plus crater degradation frameworks and equilibrium crater populations used to estimate surface ages.

• Lander localization: Combined HiRISE, CTX, and HRSC images hierarchically georeferenced to the MOLA cartographic grid to derive precise location and elevation of the lander, heatshield, and backshell/parachute. HiRISE stereo DEMs used to quantify local relief.
• Inertial position from Doppler: Used RISE X-band Doppler tracking over the first 34 sols with a Mars rotation model harmonized to MOLA reference to estimate spin radius, longitude, and derive latitude via matching to MOLA topography; compared inertial and cartographic positions to assess map-tie uncertainty.
• Surface and subsurface observation: Documented landing-induced surface modification (dark spot, scouring, striations, pebble tails) from lander cameras; analyzed pits excavated by retrorockets for subsurface exposure (depths up to ~10 cm, pit wall slopes, clods, cemented matrix).
• Soil mechanical assessment: Interactions of HP3 mole and Surface Support Assembly and lander arm scoop with soil used to infer mechanical properties (e.g., mole pit geometry, absence of ejecta piles, scoop indentation depth). Estimated cohesion via slope stability analysis and from laboratory analogs; derived minimum cohesion bounds consistent with observed steep pit walls and lack of slumping.
• Rock abundance and size-frequency: Conducted rock counts in five representative areas (workspace, near/far RAD spots, NW, and S with largest rocks). Used stereo-derived DEMs and orthomosaics; digitized rock outlines, computed convex hulls and axes; converted oblique measurements to meters using camera angular resolution and range; constructed cumulative fractional area and cumulative number vs. diameter distributions; compared to exponential and related models (1–10% rock abundance curves).
• Thermophysical analysis: Used orbital and lander radiometer data to measure thermal inertia (TI) and infer particle sizes; evaluated seasonal variations to assess vertical contrasts; modeled effect of rock abundance and cement volume on TI.
• Crater morphology and statistics: Performed morphometric analysis of >2000 craters (>20 m diameter) in a 20 km2 region; included 1316 quasi-circular hollows to assess degradation continuum, equilibrium slope, and crater retention age for Homestead hollow.
• Aeolian and seismic indicators: Interpreted spatial distribution of seismically detected convective vortices to infer relative subsurface stiffness contrasts (weaker, sandier fill vs. surrounding terrain).

• Precise location: InSight landed near the center of the ellipse at 4.502°N, 135.623°E, elevation −2613.43 m (MOLA grid). RISE inertial location differs by ~220 m west, consistent with cartographic map-tie uncertainty.
• Geomorphic setting: The lander sits on the western side of Homestead hollow, a degraded ~27 m diameter impact crater with smooth, sandy, granule- and pebble-rich surface and low rock abundance; adjacent terrain to the west is slightly rockier and rougher.
• Crater population and age: Numerous small (1–10 m) craters nearby, many degraded and soil-filled; bright ejecta Corinto secondaries present. Morphometric analysis of 2261 craters (including 1316 hollows) shows a degradation continuum with a −2 equilibrium slope for 20–100 m diameters; Homestead hollow crater retention age ~400–500 Myr.
• Rock abundance and sizes: Rock abundance is low and matches pre-landing expectations. Smooth terrain near the lander has ~1–2% rock abundance; rockier terrain has ~2–4%. Distributions for small diameters (<5 cm) are steep and pebble-rich, similar to Phoenix (~2%) and Spirit (~4%) sites. Largest rock in far RAD spot <5 cm; ~2% of surface covered by rocks >3 cm, too little to affect thermal inertia.
• Landing disturbance: HiRISE shows a ~20 m radius dark spot around the lander with ~35% lower albedo; inner ~5 m slightly brighter. Lander images show radial striations, mm-scale relief, pebble tails, and a ~1 m rolling pebble track, indicating removal of surficial dust and scouring of unconsolidated sand/granules by pulsed retrorockets.
• Near-surface stratigraphy: Exposed by retrorocket pits and HP3 mole pit: microns-thick surficial dust; ~1 cm unconsolidated sand; a variable cm-thick duricrust (5–10 cm observed) of cemented sand/pebbles/rocks with steep, cohesive faces; below, poorly sorted, unconsolidated sand with rocks (likely ejecta lenses). Thermal inertia 160–230 J m−2 K−1 s−1/2 implies dominance of fine sand (~150 μm). Thermal modeling limits cement volume to a fraction of a percent and indicates absence of strong thermophysical contrasts in the upper few tens of cm.
• Soil mechanical properties: Pit wall slopes up to 60–70° require minimum cohesion of ~5–24 Pa to maintain stability at observed scales. Scoop imprint ~0.5 mm deep without slumping implies cohesion of at least 1–1.9 kPa for reasonable densities/friction; lab simulants forming open pits under mole hammering indicate cohesions ~2.5–12.5 kPa. Mole pit shows resistant, horizontally layered crust with overhangs and pebbles in a cemented matrix.
• Process interpretation: Rim vs. interior rock perching/burial indicates early stripping of fines from rims and infill into the hollow, followed by ongoing slow eolian abrasion and punctuated modification by later impacts. Seismically detected convective vortices more frequent to the east suggest weaker, sandier shallow subsurface within the hollow compared to rockier terrain outside.
• Regional context: Bedforms (ripples) are sparse near the lander and largely sequestered at or within rims of relatively fresh craters, consistent with modeled/measured winds; bright bedforms and dusty surfaces indicate little recent aeolian transport.
• Subsurface architecture: An impact-generated regolith ~3 m thick under the lander (consistent with crater depths and ejecta crater inferences) overlies fractured basaltic lava flows; wrinkle ridges in the region support underlying basalt flows ~200 m thick.

The observations confirm that InSight’s landing site geology is dominated by impact, eolian, and limited mass-wasting processes, producing a sand-dominated regolith with low rock abundance over basaltic lavas. This context is essential for interpreting seismic and heat-flow data: low thermal inertia and low seismic velocities are consistent with unconsolidated fine sand layers capped by a thin, cohesive duricrust with minimal cement volume. The degraded-crater setting (Homestead hollow) explains spatial variability in subsurface stiffness and the distribution of convective vortices detected seismically. Sparse active bedforms and dusty surfaces suggest an old, aerodynamically equilibrated surface with limited recent sediment transport, further stabilizing the shallow thermophysical and mechanical properties relevant to SEIS noise modeling and HP3 penetration behavior. The close agreement with pre-landing orbital predictions strengthens the link between remote-sensing indicators (rocky ejecta, wrinkle ridges, TI) and in situ ground truth, improving interpretation for other northern lowland sites.

This work provides the first comprehensive in situ geologic and shallow subsurface characterization of an Elysium Planitia site along the martian dichotomy, documenting: (1) landing in a degraded impact crater (Homestead hollow) with smooth, sandy, granule/pebble-rich soils and low rock abundance; (2) a near-surface stratigraphy of microns-thick dust, ~1 cm unconsolidated sand, a cm-scale cohesive duricrust (5–10 cm observed), and poorly sorted sand with rocks beneath; (3) a ~3 m regolith overlying basaltic lava flows consistent with regional mapping; and (4) surface modification dominated by impact and eolian processes with limited recent activity. These results validate pre-landing remote predictions and provide key inputs for interpreting InSight’s seismic and heat-flow measurements. Future work will refine stratigraphy and mechanical properties using continued SEIS observations (including marsquake and atmospheric signals), further HP3 interactions, improved thermophysical modeling, and integration with regional geologic and geodetic constraints to reduce map-tie uncertainties and generalize to other northern plains sites.

• Direct subsurface exposure is limited to the top ~10 cm (retrorocket pits and mole pit); deeper stratigraphy (meter-scale regolith thickness) is inferred from crater statistics and thermophysical models.
• Observations are spatially constrained to the immediate vicinity of the lander and line-of-sight panoramas; bedrock outcrops were not observed.
• Mechanical property estimates rely on indirect indicators (slope stability, scoop/mole interactions) and analog lab tests with associated uncertainties in bulk density and friction angles.
• Thermophysical inferences assume minimal rock influence and homogeneous near-surface layers; limited seasonal variation restricts detection of strong vertical contrasts.
• Cartographic-to-inertial map-tie uncertainty (~220 m) persists, though characterized, and may affect precise geolocation comparisons.