Atmospheric oxidation drove climate change on Noachian Mars

The study addresses when and how early Mars transitioned from a relatively warm, reducing environment with icy highlands to the modern cold, arid, oxidizing, latitude-controlled climate. Prior models indicate that early Mars temperatures were elevation-dominant with ice accumulating at highlands and poles. Yet the timing and mechanism behind the shift to a latitude-dominant climate remain uncertain. Evidence for reducing gases (e.g., H2) from volcanism, serpentinization, and impacts suggests a strong greenhouse effect could have warmed early Mars, so progressive atmospheric oxidation could cool climate. The authors hypothesize that surface iron (Fe) abundance patterns in ancient terrains, measured by Mars Odyssey Gamma Ray Spectrometer (GRS), record coupled redox and climate transitions because Fe mobility is sensitive to pH, aqueous chemistry, redox, and temperature. They aim to test whether widespread Fe depletion reflects anoxic, low-temperature weathering and leaching (e.g., seasonal freeze-thaw) and to determine whether the temperature distribution mode shifted from elevation-dominant to latitude-dominant during the Noachian, coupled with atmospheric oxidation.

• Climate mode transition mechanisms proposed include decreasing atmospheric CO2 or H2, and reduced cloud radiative effects. Modeling shows elevation-temperature correlation at higher pressures and latitude control at lower pressures, implying pressure/greenhouse changes could drive a shift from elevation-dominant to latitude-dominant temperature distributions.
• Multiple sources of H2 on early Mars (volcanoes, serpentinization, impacts) could sustain a reducing greenhouse, warming climate. Episodic outgassing and impacts likely produced intermittent reducing atmospheres and transient warming.
• GRS data previously indicated relatively low Noachian surface Fe compared to younger terrains and global averages, suggesting secondary processes affected surface Fe beyond igneous controls.
• Acidic leaching alone is inconsistent with widespread Fe depletion (mass balance, lack of Fe–Cl correlation, and relatively high Th under low pH would be expected). Under oxidizing conditions Fe precipitates as ferric oxides; under reducing conditions Fe(II) is soluble and mobile, facilitating leaching. Paleosols on Earth and Mars analogs show Fe loss under reducing conditions.
• Infrared detections (CRISM) of clay-rich paleosols do not spatially coincide with all Fe depletion regions, consistent with Fe depletion occurring under near-freezing conditions that inhibit phyllosilicate formation.

• Data and spatial binning: The global geologic map (Tanaka et al., 2014) was used to assign an apparent relative surface age (Early, Middle, Late Noachian, or Noachian–Hesperian transition, NHT) to each 5°×5° GRS bin, requiring the dominant unit to cover at least half the bin; mixed or insufficiently covered cells were excluded. GRS provides elemental abundances for the upper ~30 cm.
• Topography: Mean elevation per bin was extracted from MGS MOLA data.
• Correlation analyses: Pearson correlation coefficients (r, with 95% CI) were computed between Fe abundance and elevation, and between Fe abundance and latitude, for terrains of different epochs. To minimize confounding, the Early and Middle Noachian terrains were additionally analyzed in three latitude bands (>20°N, 20°S–20°N, <20°S). Latitude analyses focused on the southern hemisphere to reduce elevation effects from northern topographic variation.
• Statistics: Fe abundances were approximately normally distributed. ANOVA and z-statistic tests were used to compare means across elevations, latitudes, and age categories. Reported as statistically significant at 95% CI when rejecting the null hypothesis.
• Estimating Fe leaching intensity: The potential original Fe abundance of Noachian crust was approximated by ALH84001 bulk Fe (~14.3 wt.%). Late Noachian leaching was estimated as the difference between 14.3 wt.% and the Late Noachian surface Fe. Early and Middle Noachian leaching intensities were estimated by differences between successive epochal means (Early minus Middle; Middle minus Late) to isolate epoch-specific depletion. The influence of sand/dust cover (~14.9 wt.% Fe) was considered to bias depletion estimates low.
• Data access: GRS elemental maps and MOLA topography are available via JMARS or NASA PDS; source data provided in the article’s Source Data file.

• Spatial trends in Early and Middle Noachian: Fe abundance decreases with increasing elevation between −3.0 km and +3.0 km; within this range Fe declines from ~15.4 wt.% to ~13.3 wt.% as elevation increases. Regions above +3.0 km have relatively stable Fe (~14.0 wt.%), close to the global surface average (~14.3 wt.%). Trends are pronounced in the tropics (20°S–20°N) and northern high latitudes (>20°N). In southern high latitudes (>20°S), areas 0–3 km have lower Fe (~13.0 wt.%) than >3 km (~14.0 wt.%).
• Spatial trends in Late Noachian and NHT: Fe abundance decorrelates with elevation but decreases with increasing latitude, especially in the southern hemisphere. Correlation coefficients (southern Mars): Late Noachian r≈+0.52; NHT r≈+0.77 for Fe vs latitude. The correlation between Fe and elevation weakens over time through the Noachian.
• Temporal evolution of surface Fe: Epochal mean Fe increases progressively toward younger Noachian terrains and the global average: the increase is ~0.58 wt.% from Early→Middle Noachian, ~0.15 wt.% from Middle→Late Noachian, and ~0.10 wt.% from Late Noachian→global average. NHT terrains have higher mean Fe (~15.4 wt.%), ~1.2 wt.% above Late Noachian and ~1.1 wt.% above global average.
• Statistical significance: ANOVA shows significant Fe variation across elevations in Early+Middle Noachian (F(5,539)=39.02, p<0.001) and across latitudes in Late Noachian+NHT (F(10,61)=6.67, p<0.001). Across age categories, ANOVA is significant (F(4,810)=41.37, p<0.001). Z-tests are significant at 95% CI except Early Hesperian vs NHT and Late vs Middle Noachian comparisons.
• Interpreted processes: Low-temperature conditions facilitated Fe depletion under reducing conditions via seasonal freeze–thaw, enhancing anoxic weathering and Fe(II) leaching. Above ~+3 km, inferred permafrost limited leaching; between −3 and +3 km, seasonal freeze–thaw enabled efficient leaching; below −3 km and at warmer conditions, higher weathering rates likely enhanced top-down leaching. The decoupling from clay detections suggests Fe depletion often occurred near 0 °C where clay formation is inhibited.
• Fe leaching intensity by epoch: Estimated minimum leaching is ~0.58 wt.% (Early Noachian), ~0.15 wt.% (Middle Noachian), and ~0.10 wt.% (Late Noachian) relative to the inferred original crustal Fe (or adjacent epoch). Depletion likely underestimated due to higher-Fe sand/dust cover (~14.9 wt.%) superposed on Noachian terrains.
• Climate mode and redox transition: The systematic shift from Fe–elevation correlation (Early/Middle Noachian) to Fe–latitude correlation (Late Noachian/NHT) indicates a transition from elevation-dominant to latitude-dominant temperature distribution during the Noachian. Concurrent decline in leaching intensity implies gradual atmospheric oxidation. The jump to higher Fe at NHT suggests a marked shift to oxidizing conditions depositing ferric minerals at the surface.
• Broader implications: Progressive atmospheric oxidation likely weakened greenhouse warming (notably H2-driven), cooled climate, and drove ice migration from highlands to polar regions, leading to the modern bipolar, cold Mars.

The findings support the hypothesis that Noachian surface Fe depletion reflects anoxic, low-temperature weathering and leaching under a reducing atmosphere. The observed decrease of Fe with elevation in Early/Middle Noachian and with latitude in Late Noachian/NHT links Fe mobility to temperature, indicating that the dominant control on surface temperature evolved from elevation to latitude. This climate mode transition coincides with decreasing Fe leaching over time, consistent with progressive atmospheric oxidation and weakening greenhouse effect (particularly reduced H2), and potentially reduced atmospheric pressure. The NHT increase in surface Fe suggests oxidation of dissolved Fe(II) and deposition of ferric minerals at the surface, marking a significant redox shift likely associated with reduced impact-driven H2 after the Late Heavy Bombardment. Episodic volcanism and impacts plausibly produced intermittent reducing, warm intervals, consistent with rover observations of wet–dry cycles and transient warm conditions. Collectively, the spatial-temporal Fe patterns provide geochemical evidence that atmospheric oxidation was a principal driver of climate cooling and the shift to a latitude-dominant, bipolar climate, and they help constrain the timing of coupled redox–climate transitions in the Noachian.

This study leverages global GRS-derived surface Fe abundances, mapped by age, elevation, and latitude, to reveal that early Mars experienced: (1) widespread Fe depletion driven by anoxic weathering and freeze–thaw leaching under a reducing atmosphere; (2) a progressive climate mode transition from elevation-dominant to latitude-dominant temperature control during the Noachian; and (3) a coupled redox evolution culminating in a pronounced oxidation during the Noachian–Hesperian transition, coincident with higher surface Fe from ferric mineral deposition. These results indicate that atmospheric oxidation contributed to climatic cooling and redistribution of ice from highlands to polar reservoirs, setting the stage for the modern cold, bipolar Mars. Potential future research directions include refining epochal Fe estimates by correcting for sand/dust cover and mixed-age bins, expanding high-resolution mineralogical coverage to better co-register with GRS footprints, integrating additional geochemical proxies to quantify redox histories, and coupling improved climate–photochemical models with geologic constraints to resolve the relative roles of pressure loss versus greenhouse weakening.

• Remote sensing depth: GRS Fe measurements reflect only the upper ~30 cm, potentially missing deeper geochemical signatures.
• Spatial resolution and binning: 5°×5° bins are large and often contain mixed geologic units and ages, introducing uncertainty; only bins with a dominant unit were used, but age heterogeneity remains possible.
• Areal coverage: Limited CRISM coverage hampers direct comparison between clay detections and Fe depletion zones.
• Sand/dust cover: Surface dust (~14.9 wt.% Fe) can mask depletion, likely causing underestimation of Fe loss, especially on older terrains with longer accumulation times.
• Statistical scope: Insufficient data to decompose Late Noachian and NHT by latitude bands analogous to Early/Middle Noachian; some correlations are modest due to confounding by extreme elevations/latitudes.
• Assumptions for leaching estimates: The use of ALH84001 (~14.3 wt.% Fe) as a proxy for original Noachian crustal Fe and epoch-to-epoch differences introduces uncertainty.
• Interpretive ambiguity: Multiple factors (pressure, greenhouse composition, obliquity) can influence the transition from elevation- to latitude-dominant temperature control; disentangling their relative contributions remains challenging.