Chemical weathering over hundreds of millions of years of greenhouse conditions on Mars

Mars today is cold, dry, and oxidized, yet ancient terrains host valley networks and paleolakes indicating a past warmer, wetter climate. Leading hypotheses invoke strong greenhouse warming of a thicker early atmosphere, potentially with reducing gases (H2, CH4) mixed with CO2. Key open questions include whether Mars experienced a single major warming event or multiple episodic excursions, whether the climate was globally warm or regionally variable, and how redox conditions evolved. Hydrated alteration minerals on Mars, particularly compositional stratigraphy with an upper Al-rich clay layer overlying Fe/Mg smectites, have been interpreted as paleo-weathering profiles formed by top-down leaching and can record both climate and redox conditions via Fe mobility. This study conducts a global assessment of 203 such stratigraphic exposures to constrain their geology, stratigraphy, ages, and implications for early Mars climate and redox state.

Prior orbital spectroscopy work identified widespread phyllosilicates and compositional stratigraphy on Mars, notably at Mawrth Vallis, Nili Fossae/NE Syrtis, Valles Marineris plateaus, and other Noachian terrains. These sequences, typically Al-rich clays over Fe/Mg smectites, have been linked to pedogenic weathering under precipitation-driven, top-down leaching. Fe mobility under reducing conditions has been proposed based on gradational color/compositional transitions that crosscut bedding, suggesting anoxic weathering. Previous regional studies established the abundance and diversity of clay minerals, associations with hydrated silica, jarosite/alunite (acidic conditions), allophane/imogolite (neutral to weakly acidic), and occasional carbonates (neutral conditions), implying varied aqueous chemistries. Age constraints and stratigraphic relationships suggested Noachian to Early Hesperian timing for alteration, but global synthesis, additional detections, and systematic age bracketing remained needed to resolve whether profiles represent a single or multiple climate excursions and to evaluate global versus local processes.

The authors compiled 203 exposures of compositional stratigraphy (including 54 new detections) from prior literature and global hydrous mineral catalogs. They integrated multiple datasets in a GIS environment with co-registered products: CTX mosaics (6 m/pixel) for geologic context; HiRISE panchromatic and IRB color images (~0.25 m/pixel) to map sub-meter color/compositional variations and contacts; THEMIS day/night IR mosaics (~100 m/pixel) for thermophysical context; CRISM VNIR hyperspectral data (~0.4–3.9 µm; 18 m/pixel in FRT mode) for mineral identification; and topography from blended MOLA/HRSC grids (200 m/pixel) plus high-resolution DEMs derived from CTX/HiRISE stereo using Ames Stereo Pipeline. CRISM I/F data were atmospherically corrected with the CRISM Analysis Toolkit (v7.4). Spectral parameter maps were computed to detect Fe/Mg phyllosilicates (e.g., D2300) and Al–OH-bearing phases (e.g., BD2200), with spectral ratios to spectrally bland regions used to validate mineral identifications (e.g., nontronite, saponite, montmorillonite, kaolinite, hydrated silica, allophane/imogolite). HiRISE color patterns (bluish/white Al-rich vs red/brown Fe/Mg-rich) were used to extrapolate compositional mapping beyond CRISM coverage and to assess nature of contacts (gradational vs sharp) and their relation to bedding. Geological contexts were cataloged (crater rims/walls/floors, intercrater plains, valley networks, massifs/knobs, basins). Ages were constrained primarily via crater size–frequency distributions on associated, better-exposed cap or host units using CTX mosaics and Crater Tools in ArcGIS, adopting Ivanov’s production function and Hartmann–Neukum chronology, with stratigraphic superposition for relative timing. Counting areas exceeded 1000 km² where possible, focused on craters >1 km diameter (minimum ~0.8 km where necessary), avoiding obvious secondaries. Regional stratigraphic columns were compiled across key provinces to compare mineral sequences and ages.

• Inventory and distribution: 203 compositional stratigraphy deposits identified across the southern highlands, including 54 new detections. Latitude range primarily 40°S to 30°N; elevation range approximately −3000 to +6000 m. Nearly 88% of exposures occur in Noachian units (65 in Early Noachian, 85 in Middle Noachian, 28 in Late Noachian, 18 in Hesperian–Noachian transition units).
• Stratigraphic pattern: 201/203 (>99%) exposures show a single transition of Al/Si-rich materials overlying Fe/Mg smectites. Only two locations exhibit Fe/Mg smectites over Al-rich clays (Meridiani Planum and south Coprates Chasma), suggesting rare multiple pedogenic episodes.
• Nature of contacts: HiRISE color indicates gradational transitions that crosscut bedding, consistent with top-down leaching and Fe loss under reducing conditions; white/bluish Al-rich horizons over red/brown Fe/Mg smectite horizons are typical.
• Geological contexts: Profiles occur in diverse settings (impact structures, intercrater plains/basins, valley network interfluves, knobby/chaotic terrains, massifs). Observational bias likely underrepresents true areal extent.
• Felsic terrains: On massifs around Hellas Basin, felsic (plagioclase-rich, anorthositic) units show Al-clay (kaolinite) alteration at high elevations, with Fe/Mg smectites lower, indicating precipitation-driven weathering of felsic protoliths in addition to more common mafic hosts.
• Age constraints: Oldest host terrains with profiles date to ~3.97 Ga (Thaumasia Planum explosive volcanoes; also ~3.83 Ga). Youngest clear example is in Orson Welles crater chaotic deposits, with host layered fill modeled at ~3.18 Ga and crater rim/ejecta ~3.57 Ga, indicating Late Hesperian weathering. In general, most ages cluster ~3.8–3.6 Ga.
• Duration to form profiles: Based on thicknesses (commonly 50–60 m, up to ~120 m) and terrestrial clay formation rates (~0.01–0.05 mm/yr), individual profiles could form in ~10^6–10^7 years, a small fraction (<1–2%) of the Noachian timeframe over which weathering occurrences are observed (~700–800 Myr span).
• Environmental inferences: Mineral assemblages (Al clays with hydrated silica, jarosite/alunite, allophane/imogolite, occasional carbonates) imply variable pH and redox conditions, but pervasive Fe mobility and color patterns support episodes of anoxic/reducing greenhouse conditions enabling top-down Fe leaching.
• Global significance: Occurrence over wide elevation range (>9 km) and broad geographic/chronostratigraphic distribution supports globally pervasive, precipitation-driven chemical weathering in early Mars’ history.

The global prevalence of a single Al-over-Fe/Mg clay transition suggests a common formation mechanism: top-down, precipitation-driven chemical weathering accompanied by Fe mobility under reducing atmospheric conditions. This pattern addresses whether events were local or global, indicating a global climatic process affecting much of the southern highlands. The rarity of multiple stacked transitions within single outcrops might imply a single climate transition; however, the wide span in modeled ages (Early Noachian to Late Hesperian) is difficult to reconcile with a one-time event. A chemical resetting scenario is therefore favored: multiple warming episodes occurred, but younger events dissolved and transported Fe downward, erasing evidence of earlier Fe-rich horizons and leaving a single apparent transition per exposure. Observational limitations (exposure length, resolution) may also mask multiple transitions. The linkage of profiles to diverse host lithologies, including volcaniclastic and impact-generated materials with high porosity and reactivity, suggests that lithologic properties likely accelerated weathering rates, enabling profile formation within million-year timescales during warm episodes. The coexistence of Al clays with minerals diagnostic of acidic (jarosite/alunite) and neutral to mildly acidic (hydrated silica, allophane/imogolite) conditions, plus occasional carbonates, indicates spatial or temporal variability in aqueous geochemistry. Correlation with valley networks and lake basins suggests weathering broadly coincided with pulses of fluvial activity in the Late Noachian/Early Hesperian. Overall, the results constrain plausible climate mechanisms to reduced greenhouse-induced warm excursions of 10^6–10^7 years, consistent with transient warming by H2/CH4 with CO2 and potentially modulated by volcanism, impacts, or obliquity-driven methane release.

A global survey of >200 martian weathering profiles reveals a near-ubiquitous compositional stratigraphy of Al/Si-rich horizons over Fe/Mg smectites across Noachian terrains and into the Hesperian, spanning elevations from approximately −3 to +6 km. Most profiles exhibit a single mineralogical transition, yet age constraints from ~3.97 to ~3.18 Ga imply weathering occurred over hundreds of millions of years. The findings are best explained by multiple episodes of precipitation-driven, top-down chemical weathering under reducing greenhouse conditions, with geochemical resetting during later events erasing older Fe-rich horizons. Individual profiles likely formed within ~10^6–10^7 years, providing tighter constraints on the duration of warm, wet intervals than geomorphic features alone. The presence of profiles in felsic and mafic contexts underscores the global nature of the process. Future work should focus on improving exposure mapping, in situ and returned-sample geochronology to distinguish alteration ages from host deposits, higher-resolution stratigraphic mineral mapping to detect potential multiple transitions, and integrating climate models with redox-sensitive weathering processes to refine the timing and drivers of early Mars climate excursions.

• Observational biases: Weathering profiles are most readily detected on steep, dust-free slopes; many occurrences likely remain obscured or unresolved, and exposures commonly span only tens to hundreds of meters, limiting detection of multiple transitions.
• Age uncertainties: Crater counting on small areas and ancient terrains entails significant uncertainties, especially for craters <4 km; ages often bracket host or cap units rather than the alteration event itself, providing only upper limits to weathering ages.
• Resolution and spectral limits: CRISM spatial resolution may miss thin or closely spaced mineralogical horizons; BD2200 (Al–OH) parameters can be noisy for small exposures; compositional heterogeneity may be under-resolved.
• Stratigraphic ambiguity: Gradational contacts that crosscut bedding support pedogenesis but complicate distinguishing diagenetic overprints or multiple events; sedimentary reworking (transport, burial, exhumation) can overprint primary weathering signals.
• Lithologic variability: Differences in porosity, permeability, and reactivity (e.g., ash, impactites, glass vs crystalline rocks) affect weathering rates and profile thickness, complicating timescale estimates and inter-site comparisons.