Hydrothermal activity generated by impact melt emplacement on the rim of Ritchey crater, Mars

Abstract

Impact-induced hydrothermal systems have the potential to sustain long-lived aqueous environments throughout the history of Mars, yet their nature and distribution are not well-understood. While post-impact hydrothermal alteration on Mars has traditionally been studied at central peaks, we reported similar processes within a well-preserved impactite stratigraphy across the inner rim of Ritchey crater. This stratigraphy comprises a sheet unit overlying fragmented breccia, consistent with the emplacement of impact melt rocks on ballistic ejecta deposits, similar to features observed in complex craters on Earth. Analysis of CRISM hyperspectral data revealed alteration minerals including serpentine, chlorite, Mg-carbonate in fractured bedrock, veins, and erosional windows underneath the sheet unit. These alteration minerals are unrelated to post-impact fluvio-lacustrine facies or pre-impact target bedrock. Instead, their formation is most plausibly attributed to the emplacement of hot impact melt and subsequent groundwater percolation through fractured bedrock at the crater rim. The widespread distribution of the alteration minerals on the inner rim suggests that impact cratering can create extensive habitable environments. Our findings underscore the potential for alteration minerals at crater rims to preserve biosignatures, a key objective of the Mars 2020 mission at Jezero crater.

Introduction

In hypervelocity impact events, the substantial kinetic energy of the impactor undergoes conversion into internal energy, where shock-induced high pressures and temperatures generate impact melt^1,2. On water-bearing or icy planets^3,4, and even on asteroids⁵, the ensuing propagation of heat and water after meteoric impact culminates in the development of impact-induced hydrothermal systems, which have the potential to support diverse forms of life⁶.

Numerical simulations have predicted impact-generated hydrothermal systems on Mars driven by heat from impact melt, triggering basin-wide groundwater percolation^7,8,9 and subsequent chemical alteration¹⁰, consistent with observations from terrestrial impact systems¹¹. These hydrothermally altered sites potentially fulfill important requirements for habitability, i.e., liquid water, nutrients, and energy^3,12,13. Moreover, potential biosignatures on Mars, if present, were likely to be preserved in chemically precipitated minerals (e.g., carbonates and silica) from various settings, particularly in association with hydrothermal activity^3,14,15. However, the extent, mechanism, and distribution of such hydrothermal systems in impact craters on Mars are unclear. Most previous research has searched for signatures of hydrothermal alteration within central peaks, where a series of alteration minerals associated with vein-^16,17 or mound-like features^18,19 have been identified in orbital datasets. Less commonly, alteration minerals at crater rims were detected in many craters^20,21,22, including Jezero²³, but no clear relationship with the impact or impact-related hydrothermal alteration was established. As the Mars 2020 Perseverance rover is exploring and sampling the crater rim of Jezero, it is crucial to scrutinize the origin of alteration minerals for potential links to post-impact hydrothermal systems. One challenge in making this link is that ancient Martian craters like Jezero are often heavily eroded, so lacking convincing geological evidence for heat sources. A sufficiently large, young, and well-preserved impact crater would reveal more about the nature of impact-induced hydrothermal alteration within crater rims.

Ritchey crater is located ~200 km south of Valles Marineris and is a complex crater 78 km in diameter, with exceptionally well-preserved central uplift, terraces, rim, and ejecta (Fig. 1A). Crater counting on the Ritchey ejecta yields a retention age of 3.46 Ga, suggesting an Early Hesperian impact age²¹. At the crater center, a smooth, coherent layer and a dark, rough unit were speculated to be a preserved impact melt deposit²¹. Although the crater floor is covered by aeolian deposits, the inner rim exhibits well-exposed and intact bedrock. Fluvial channels and related fan deposits²¹ emanating from the wall indicate at least some post-impact aqueous activity occurred.

Fig. 1: Geology of Ritchey crater.

A Ritchey overview, CTX image mosaic (see Methods). B Eastern crater rim area of Ritchey, showing the smooth and dark-toned sheet unit. HiRISE grayscale and enhanced color images (ESP_011846_1515 and ESP_012914_1515) on CTX basemap. Cyan area: regions of interest (ROI a) used to extract average CRISM spectra in Fig. 3. Inset 1 at the bottom left shows the increasingly brecciated texture toward the margin of the sheet unit. C Close-up view of the sheet unit margin. Note the brecciated slope/escarpment of the sheet unit (see Fig. 2C for a 3D view). White arrows: fragmented breccia underlying the sheet unit. Yellow arrows: NE-SW ridges in the light-toned basement unit. Green arrow: dark-toned dike-like material in the sheet unit. Inset 1 shows decameter-long vertical structures (black arrows) across the escarpment of the sheet unit that resembles degassing pipes³⁶ or hydrothermal channels³⁷ in melt-rich breccia on Earth. Some of these vertical structures show side branches with varying shapes (white arrows). Inset 2 shows a carbonate-bearing basement rock with fractures or possible vein fillings (red arrows) related to the fault zone, see Fig. 3 ROI d for spectrum. D Comparison of the brecciated escarpment of the sheet unit (see Fig. 2A for 3D view) and a smooth escarpment of bedrock. The breccia and altered bedrock units are stratigraphically confined between the sheet unit to the NW and the dark-toned underlying bedrock unit in the SE. E Cross section of the impactite stratigraphy, see E–E’ in Fig. 1B.

The excellent preservation of Ritchey crater offers a unique opportunity to study impactite-alteration stratigraphy on the rim for large, complex craters on Mars. Here we use Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) hyperspectral visible/near-infrared²⁴ and High-Resolution Imaging Science Experiment (HiRISE²⁵) images to identify alteration minerals, impactite stratigraphy, and post-impact geological processes on the rim area, in order to investigate their relationship to impact-induced hydrothermal activity. Through these observations, we seek to provide an important framework to understand the alteration processes and astrobiological potential of martian crater rims.

Results

Geomorphology and stratigraphy

On the inner rim of Ritchey crater, the most common stratigraphic sequence presents as fractured light-toned bedrock overlain by a fragmented breccia unit (Fig. 1C, D), and capped by a darker-toned sheet-like unit (Fig. 1C, D). Outcrops of light-toned massive bedrock at the base of the sequence exhibit a ridged texture, with smooth escarpments along their slope (Fig. 1C, D). The fragmented breccia unit contains large, rounded to unsorted boulders (up to ~70 m) embedded in a medium-gray toned matrix (Fig. 1C, inset 2). On the top, the sheet unit forms a ten- to decameters-thick coherent layer that can be traced laterally across several km (Fig. 1B). With a homogeneous and smooth surface, the sheet unit is increasingly brecciated toward the margin (although the brecciated margin may be covered by dust), embedded with smaller blocks (usually <10 m) than the underlying fragmented breccia unit (Figs. 1C, D, 2A, C). Some light-toned, elongated (up to 40–50-m long), and vertical pipe-like structures are observed inside the sheet unit (Fig. 1C).

At its marginal slope/escarpment, the sheet unit sometimes grades to a rough-surfaced pitted unit (Fig. 3B and S1E) or a ridged unit (Figs. 1C, 3F, G) that overlies the light-toned basement rocks. The pitted unit is a less regular and coherent, patchy material (Fig. 3B), comprising closely-spaced, decameter- to hundred-meter-sized pits that form “soap froth” networks with negative topography whose rims do not appear to be raised above the surrounding terrain. Blocks and brecciated fabrics are embedded at the interior walls of the pits (Fig. 3B inset). The rims of the pits are not surrounded by ejecta, so unlikely to be impact craters. Between the pitted unit and the basement, a ridged unit is present, showing elongated, near-linear ridge features (Fig. 3C, F) and brecciated fabrics at the side slopes (Fig. 2A, C).

Fig. 2: Comparison of the impactite stratigraphy between Ritchey and terrestrial craters.

A, C 3D view of the sheet unit overlying the fragmented breccia unit at the eastern rim of the Ritchey crater, HiRISE image on HRSC DEM (architected from Fig. 1C, D). B, D The impactite stratigraphy at Aumühle outcrop, Ries crater (B, Germany) and Mistastin lake impact structure (D, Canada). Photo D by Cassandra Marion in 2021 at Coté Creek (https://craterexplorer.ca/mistastin-impact-crater/).

Spectral data

CRISM spectra of typical basement rocks in the Ritchey crater rim show strong pyroxene absorption bands (0.9–1.0 μm, 1.9–2.0 μm) most consistent with low-Ca pyroxene (LCP), with no associated alteration mineral detections²¹. The sheet-like unit also shows broad absorptions centered near 0.9 and 1.9 μm consistent with LCP, but with a shoulder near 1.2 μm potentially due to mixing with olivine or glass²⁶ (Fig. 3I, spectra a and g). The sheet unit appears to be largely unaltered, as no hydration (e.g., 1.9 or 1.4 μm) bands and only a weak, possible Al/Si-OH (2.21 μm) band is consistently detected (Fig. 3I, spectrum a; Fig. S1C, D). The pitted unit has similar spectral features (Fig. S1B, D) as the sheet unit. In contrast, the fragmented breccia (Fig. 3E) does not show pyroxene bands, and instead resembles olivine or glass with a broad band centered >1.0 μm, and a weak possible band at 1.9 μm, suggesting minimal hydration (Fig. 3I, spectrum e).

Fig. 3: Details of the sheet unit, ridged unit and ejecta, and spectra for ROIs.

A The continuous cauliflower-like, discontinuous wrinkles, and flow features (white arrows) of the melt at the northern rim of the Ritchey, HiRISE image (ESP_013125_1520_RED). B Close-up view of the box area at Fig. 3A from HiRISE IRB (ESP_013125_1520_IRB) color image. The yellow to gold-colored material indicates the pits within the sheet unit, similar to the pitted materials described by refs. ^34,35, whereas the basement rocks are white and at lower elevations. C Ridged unit at the western rim, HiRISE image PSP_007363_1510_RED. Green area indicates the ROI for spectral analysis. Boxed area shows the brecciated fabrics at the slope of a ridge. D Altered basement rocks that contain carbonates at the eastern inner rim (PSP_009090_1515_RED). White arrows: ridged area; black arrows: putative mound structure. E continuous ejecta that underlies the ridged unit (ESP_011846_1515_RED). Black arrow: blocks and breccia embedded at the slope/escarpment of the ridge. F Occurrences of serpentine (cyan ROIs) near the ridged unit at the southern crater rim (ESP_017371_1505_RED). White arrows: ridged surfaces. G HiRISE image ESP_013125_1520. Note that the ridged units occur at the margin of the sheet units. H Occurrence of chlorite. Note that the ROI h is located below the brecciated slope of the sheet unit. I CRISM ratio spectra from numbered ROI annotations in Figs. 1, 3C–H (note that spectra a and b do not represent areas in Fig. 3A, B, respectively). Black spectra are from standard minerals in the NASA RELAB database⁷⁷ (https://www.planetary.brown.edu/relab/, see Methods). Processed CRISM data are provided in Supplementary Data files.

Strong signatures of alteration are detected in the ridged unit and altered bedrock. A clear 2.29–2.31 μm band and weak 1.39–1.42 μm band in the ridged unit suggest the presence of Fe/Mg-bearing phyllosilicates such as nontronite and saponite (Fig. 3I, spectra c and d). Many outcrops also show evidence for Mg-carbonate based on an additional 2.51–2.53 μm band (e.g., Fig. 3I, spectra c–f) and 3.4–3.5 μm band, 3.9 μm band drop related to CO₂ overtones and combinations (see Supplementary Data for the full-length spectrum).

Localized signatures of other alteration minerals are also present in the crater rim. Serpentine was identified in the ridged network at the southern rim (Fig. 3F) according to a diagnostic 2.12 μm band, along with 2.3 and 2.5 μm Fe/Mg-OH bands (Fig. 3I, spectrum f). Furthermore, chlorite was detected based on 2.26, 2.00, and 2.3 μm bands (Fig. 3I, spectrum h) at the southern rim (Fig. 1A). At the northern rim, possible prehnite was identified adjacent to a carbonate outcrop (ROI I in Fig S1), based on the diagnostic 1.48 μm absorption and a 2.35 μm Mg-OH band drop.

Discussion

A well-preserved impactite stratigraphy

We hypothesize that the light-toned bedrock, brecciated outcrops, and smooth sheet units on the rim of Ritchey crater correspond to altered bedrock, continuous ejecta, and melt-rich breccia, respectively. The same stratigraphic sequence of melt-rich breccia (e.g., suevite) conformingly overlying continuous ejecta is an important feature (Fig. 2A, C) in double-layered impactites typically produced by complex craters on Earth^27,28. This unique impactite stratigraphy has been verified from well-preserved terrestrial impact craters, such as Ries (Fig. 2B), Mistastin Lake (Fig. 2D)²⁹, and many others²⁸. Similar sheet-like melt-rich breccia units have been identified at Hargraves crater³⁰ and Negril crater³¹ on Mars and in the Orientale Basin on the Moon³². The discontinuous and wrinkled layering of the ridged unit (northern rim-wall, see black arrows in Fig. 3A) is consistent with convergent returning flow³³ and/or movement of the impact melt along the steep wall area during its emplacement and cooling.

The pitted units in other Martian craters are hypothesized to result from degassing of volatile-rich components after the emplacement of the melt sheet^34,35. At Ritchey, the presence of elongated vertical features (Fig. 1C) within the sheet unit bears a resemblance to degassing pipes observed in the Ries suevite³⁶ and hydrothermal veins in the Chicxulub suevite³⁷. Such additional morphological evidence may support the interpretation of the sheet unit as the impact of melt-rich breccia. Alternatively, these morphologies could be erosional features by wind or mass wasting, although they do not possess typical broom-tail-like shapes as exhibited by recurring slope lineae (RSL³⁸).

While dark-toned smooth units could be produced by a variety of processes, the morphology and distribution of the Ritchey sheet unit are most consistent with an impact melt sheet. Notably, the newly identified sheet and ridged units at the wall-to-rim area likely correlate with the impact melt deposits reported at the central peak of the Ritchey crater²¹.

Although lava flows are abundant in the Thaumasia Planum region, the complete, ramparted rim of Ritchey (Fig. 1A) has protected the crater from external mantling infills. In addition, the sheet unit consistently possesses a brecciated slope/escarpment, whereas the slope of the nearby mafic basement is indurated, intact, and smooth (Fig. 1D). Hence, it is unlikely that the sheet unit covering all azimuths of the crater rim is a volcanic flow deposit. These sheet units also have no clear correlation with sedimentary fan deposits on the wall-to-rim areas (Fig. 4A, B). Some sheet units display incised channels (Fig. 4C), suggesting that the sheet units predate the erosional features by fluvial²¹, eolian, or slope failures.

Fig. 4: Evidence that precludes alteration minerals to be pre-existing materials.

A Mapping results of sheet unit (preserved impact melt), pitted unit, and ridged unit (partially eroded impact melt). Previously recognized fan deposits and channels are outlined²¹. The sheet units show almost no contact with channels and fan deposits, except at rare localities (Fig. 4C). B This same map on the Eastern rim of Ritchey with overlap in CRISM MTRDR Fe/Mg phyllosilicate browse products (PFM⁷⁵). The cyan color indicates Fe/Mg phyllosilicate and Mg carbonates. Alteration minerals at the wall-to-rim area (excluding the crater floor) are distributed around the sheet units. See Fig. S2 for phyllosilicate and carbonate browse products of all CRISM MTRDR scenes overlapping on the sheet-fan-channel mapping results. C Box area in Fig. 4B. The channels²¹ partly incised the sheet units. D Same view as C, overlapped with CRISM MTRDR PFM browse products.

Mineralogy of the impact melt sheet

The weak 2.21 μm hydroxylated absorption in the sheet unit commonly results from Al-OH or Si-OH bonds (or both), and is one of the most typical features of glassy, silicate-rich impact melt such as the Ries suevite³⁶—the type locality of impact-melt breccia on Earth³⁹. The wide 0.9 μm LCP bands in the sheet unit (spectrum g in Fig. 3I) shows some asymmetry that could be consistent with glass mixing, although the wide 1.9 μm band of crystallized LCP dominates over glass in lab mixtures. Judging from the depth and width of 0.9 and 1.9 μm LCP bands in the sheet unit, laboratory quantitative mixing experiments suggest the LCP component comprises at most 10–15% of the LCP-glass mixture²³. While none of these observations independently and uniquely identify glass within the impactite stratigraphy, taken together, they are consistent with a composition that is some combination of basement-derived LCP, glass, and glass recrystallized to LCP. A mixed composition of LCP and glass is also plausible for impact melt-bearing breccia, since significant recrystallization (devitrification) of the target material may have occurred when the melt material is cooling⁴⁰. Using this logic, homogeneous, resistant bedrock units with LCP-dominated spectra associated with large impact basins elsewhere on Mars have also been inferred to indicate partially recrystallized impact melt⁴¹.

Alteration of minerals and impact-induced hydrothermalism

The identified Fe/Mg phyllosilicates are consistent with smectite-dominated spectra previously detected in the central peak and crater wall area²¹. Compared to the crater rim, the central uplift of Ritchey shows a more limited diversity of alteration minerals, including Fe/Mg smectites, chlorite, and hydrated silica, whereas serpentines and carbonates were not detected²¹. Previous study also identified a similar sheet unit and interpreted it as impact melt, however, only in the central peak area²¹.

We found that the sheet unit has a wide distribution across the central peak to the crater rim (Fig. 4A). Concomitantly widespread are the alteration minerals that co-occur with the sheet units (Fig S2). Both detailed spectral analysis (ROI a, g, e in Fig. 3I) and spectral parameter maps (Fig. 4B, D) show a distinct lack of strong alteration in the sheet unit and breccia, both of which present a stark contrast to the pervasive alteration detected in the underlying ridged unit and altered bedrock. The absence of hydration in the sheet unit indicates little-to-no aqueous alteration at least within the upper layer of the melt-rich breccia, possibly due to the dehydration during the emplacement of the melt. Only weak alteration in the underlying breccia unit (Fig. 3E) is potentially explained by quick infiltration by the porous ejecta material. Instead, most phyllosilicate and carbonate signatures are exposed within erosional windows on the ridged unit and altered bedrock (Figs. 3,  4B, D), pointing to significant aqueous alteration around the sheet unit. Bulky, non-fractured intact basement rocks without contacts with sheet unit and ridge unit also do not possess alteration signature.

Fluvio-lacustrine processes might have also led to the formation of phyllosilicates. However, this scenario seems unlikely, as although fluvial channels have dissected the entire impactite sequence (e.g., Fig. 4C), there is no morphological evidence of fluvio-lacustrine origin, such as bedded or layered features, for phyllosilicates or carbonates at the crater wall. Likewise, in the rim of the Ritchey crater, because the alteration minerals (Fig. 3) are associated with the sheet units (Figs. 3, 4), they must postdate the sheet units. The lack of alteration minerals in basement rocks that have no contact with the sheet units (or degraded sheet units) further reinforces the interpretation of melt-induced hydrothermal alteration to form alteration minerals.

We therefore hypothesize that phyllosilicate-carbonate assemblages beneath the impactite stratigraphy formed via post-impact hydrothermal alteration driven by the contact with the hot melt-rich breccia. Some carbonate-rich bedrock outcrops exhibit dark fractures (Fig. 1C). These features could be consistent with mineral-filled veins or impact melt dikes. Spectral signatures indicate that the area of interest is rich in carbonate and phyllosilicate (Fig. 3I), resembling carbonate veins in the hydrothermally altered ultramafic rocks on Earth (e.g., listwanite in Oman^42,43,44). Alternatively, the dark material could be impact melt injected into basement fractures (dike suevite), as the materials appear to be affiliated with the sheet unit (Fig. 1C, green arrow). This morphology is similar to an inferred injection dike within a megabreccia block in Holden crater on Mars⁴⁵. Finally, it is possible that the low albedo of the features are shadows or sand-infilling topographic lows. Regardless, the production of serpentine, saponite, chlorite, carbonate, and prehnite by hydrothermal alteration of alkaline fluid at fractured mafic/ultramafic areas is in line with hydrothermal modeling^10,46 and has been verified by remote sensing data on both Earth⁴² and Mars^47,48.

Widespread habitable environments around the crater rim

The preserved impactite and alteration stratigraphy at the rim of Ritchey only captures a fraction of the original hydrothermal system triggered by the impact melt sheet. The initial volume of impact melt correlates with crater diameter, and can be calculated according to scaling laws^49,50. For a crater like Ritchey, models predict the production of ~380 km³ melt⁵¹, with a thickness estimated between 333 and 500 m⁸, much thicker than its present form (up to decameters-thick at the central crater and the rim areas). As the major heat source^8,9, the initial melt sheet of Ritchey likely possessed a greater volume than presently observed (Fig. 4A), resulting in a basin-wide hydrothermal alteration extending across the crater rim.

The hydrothermal alteration minerals in the rim of Ritchey crater (Fig. 3) are largely associated with the ridged units and altered bedrock in close contact with the sheet units (Figs. 3, 4), suggesting that hydrothermal fluid flow was concentrated in these units. For complex impact craters, the crater wall commonly encompasses extensive normal and concentric faults due to shockwave propagation and gravitational collapse^1,52. The widespread, heavily altered ridged units may represent extensional fault zones^41,53, which would enhance the permeability of the shallow crust and facilitate groundwater percolation⁵⁴. Additionally, the possible degassing pipes in the melt-rich breccia and carbonate-bearing (potential vein filling), fractured, altered basement rock (Fig. 1C) further support convective hydrothermal percolation between the impactite and the underlying bedrock. Permeability studies of terrestrial melt-rich breccia samples also reveal a low matrix permeability, suggesting that fluid flow through the melt-rich breccia itself would be more limited⁵⁵ than in the fractured basement bedrocks, decreasing the likelihood of aqueous alteration.

We hypothesize that the diverse alteration minerals in Ritchey track the evolution of the aqueous system following the cooling of the impact melt (Fig. 5). Impact melt glass forms at a temperature higher than 1870 °C⁵⁶. During the emplacement of the melt sheet on the rim area, the incorporation of cooler in situ fragmented materials into the melt-rich breccia leads to a lower temperature of >750–900 °C compared to the central peak and basin-filling melt sheet^28,45,57.

Fig. 5: Schematic model (based on ref. ⁵²) of the early post-impact phase with active groundwater hydrothermal percolation (A and B) and alteration processes at the microscale.

A Early post-impact phase with a hot central uplift and melt sheet covering the impact crater. The temperature of the central uplift is high enough to only allow steaming, instead of hot water circulation and spring discharge. Aqueous percolation started from the crater wall via the basement fractures once the temperature cooled down to below the boiling point. B Cross-section of melt sheet-induced hydrothermal alteration. The heat of the melt sheet was transmitted downward to melt the cryosphere and alter the water-saturated basement rocks. C Process of low-temperature, alkaline fluid altering an olivine-rich rock similar to Ritchey basement. The hydrothermal vein on Earth may preserve potential biosignature at a microscale⁶⁵ in the forms of (1) microstromatolite; (2) carbonate minerals (such as dolomite); (3)  fossilized microorganisms.

The emplaced melt is thicker in the crater basin than on the wall-to-rim area, and thus maintain a hot crater floor over time. Furthermore, uplifted, hot materials at the central peak provide additional heat to the area. Together, these effects result in a decreasing gradient in temperature^8,45 and alteration strength⁴⁶ from crater center to crater rim. Indeed, the variation in temperature and water availability between the crater center and crater rim profoundly changes the diversity in alteration minerals⁴⁶. As the melt is cooling down, hydrothermal alteration would have driven a series of low-temperature alteration reactions below 400 °C: (1) Chlorite-related alteration commonly between 300 to 400 °C (e.g., chlorite-bearing assemblage at Sudbury impact structure on Earth⁵⁸); (2) Serpentinization (<150 to 400 °C), one of the most important reactions for astrobiological interests because it mobilizes critical elements and delivers chemicals (such as hydrogen and methane) to support critical microbial metabolism similar to the earliest biogeochemical cycles on Earth⁵⁹; (3) low-temperature (<100 °C) carbonation to form saponite and carbonate⁴⁴. At low temperatures, variable water-rock ratio is essential to produce the observed mineral diversity: serpentine and chlorite form at low water availability, whereas the formation of Fe/Mg phyllosilicate requires more water^10,46. The implied water-rock ratio variability at Ritchey may suggest variation in permeability⁴⁶ of target bedrock due to different degrees of impact fracturing as well as the change in melt thickness across the crater.

Hydrothermal modeling of a complex crater comparable to Ritchey predicts that the hydrothermal alteration below 200 °C occurs on the crater rim for at least 20,000 years⁸. Such subsurface, low-temperature hydrothermal sites at crater rims (Fig. 5) could have provided potential habitable environment outside the central peaks, where most hydrothermal fluids were suggested to emerge previously^16,18,19,60. On Earth, a diverse microbial community, including thermophiles, hyperthermophiles³, and chemoautotrophs⁶¹, is found in the modern hydrothermal settings related to basaltic lava flows⁶², hydrothermal vents⁶³, and deep alkaline ultramafic environments^64,65. Molecular biomarkers have indicated colonizers in terrestrial impact-induced hydrothermal systems include, but are not limited to prokaryotes such as sulfate reducing bacteria^66,67 and cyanobacteria⁶⁸, or even some microeukaryotes like fungi⁶⁹. Minerals like hydrated silica and carbonate precipitated from hydrothermal fluids could preserve biosignatures from these environments¹⁵ (Fig. 5). Notably, the Perseverance rover on the Mars 2020 is exploring and collecting samples from the rim of Jezero, another large Noachian crater on Mars that likely also produced hydrothermal systems. Indeed, for future sampling campaigns, crater rims would be more accessible than any preserved central peak materials, if they exist, within the crater. Rock samples from impact hydrothermal deposits would be a high priority as they could have high biosignature preservation potential⁷⁰.

Methods

Visible images and topography

Landform investigation was mainly based on the enhanced false-color greyscale (RED) and IRB (infrared-blue) HiRISE (High-Resolution Imaging Science Experiment²⁵) visible images up to 25 cm/pixel. The 20 m/pixel Context Camera mosaic⁷¹ (see also https://murray-lab.caltech.edu/CTX/) was also used as a lower-resolution basemap to substantiate the areas where HiRISE are not covered. Digital elevation model (DEM) data up to 20 m/pixel were acquired from HRSC (High-Resolution Stereo Camera⁷²), supplemented by the ~300 m/pixel-resolution MOLA (Mars Orbiter Laser Altimeter⁷³) DEM data. A three-dimensional terrain model was generated using the commercial software ArcGIS Pro (Esri).

CRISM data

Determination of mineralogy was carried out using CRISM visible/near-infrared (~0.3–2.6 μm) hyperspectral image data (Compact Reconnaissance Imaging Spectrometer for Mars^24,74) at the most recently calibrated MTRDR level (Mapped Targeted Reduced Data Records^75,76). Eight FRT (full resolution targeted, 18 m/pixel) images and two HRL (half resolution long, 36 m/pixel) images were analysed using the CAT (CRISM analysis toolkit) in ENVI. A set of spectral parameters containing spectral reflectance at diagnostic wavelengths is compiled in the delivered MTRDR data, and these parameters are combined as RGB components to highlight particular minerals⁷⁵ such as carbonates (CR2 browse product) and phyllosilicates with Fe and Mg (PFM browse product), as shown in Fig. S2. All parameters in selected browse products were viewed using a linear stretch of 50-98% to enhance spectral features and suppress noises. Average reflectance spectra were extracted from each Region of Interest (ROI) containing several tens to thousands of pixels. These spectra were then ratioed (Supplementary Data 1) against spectrally neutral, bland areas (Supplementary Data 2) at similar elevations to further suppress atmospheric and instrumental effects. Spectra were compared to standard laboratory spectra (Supplementary Data 3) from NASA RELAB facility⁷⁷ at Brown University: magnesite (F1CC06B by Jack Mustard), saponite (CASA58, Edward A. Cloutis), serpentine (LALZ01, Takahiro Hiroi), chlorite (LACL14, Takahiro Hiroi), and prehnite (LAZE03, Edward A. Cloutis), and published results (glass-pyroxene mixtures from ref. ⁷⁸).