Persistent liquid water is generally considered a prerequisite for planetary habitability1,2. For all known life, which is to say life on Earth, water is the primary component of cells and serves as an indispensable matrix for biochemical processes, facilitating molecular transport, solvation, and reactivity3,4. On a macroscopic scale, the planetary water cycle helps maintain temperate climates, distributes substantial heat from equator to poles5, and drives the global biogeochemical cycling of bio-essential elements6. Water’s possibly unique suitability for life stems from its exceptional physicochemical properties, including its potent polarity, high dielectric constant, and capacity for extensive hydrogen bonding4,7. While it has been suggested that under certain conditions of temperature and pressure, liquid ammonia and certain organic solvents might share water’s bio-friendly attributes8, such conditions are doubtful on the rocky planets and moons deemed as possible abodes for life in our solar system. Consequently, the search for extraterrestrial life remains focused on worlds that harbor liquid water (Fig. 1).
Fig. 1: Evolution and distribution of potential ocean-bearing planetary bodies in the solar system.
The conventional habitable zone (HZ) is defined as the circumstellar region where Earth-like planets with CO2-H2O-dominated greenhouse atmospheres can sustain liquid water on their surfaces. In more optimistic formulations, the Solar System’s HZ may extend from early Mars to recent Venus. Discoveries of subsurface liquid-water reservoirs and oceans in dwarf planets and icy moons of the outer Solar System have substantially broadened the concept of planetary habitability beyond the classical HZ. This extended view of habitability incorporates endogenous energy sources that can maintain liquid water in planetary interiors or subsurface reservoirs. Planetary bodies are not shown to scale.
This emphasis on liquid water has strongly shaped the astronomical concept of the habitable zone (HZ), which is conventionally defined as the circumstellar region within which an Earth-like planet with the right atmospheric composition can sustain liquid water on the surface9,10. Commonly, the HZ boundaries are estimated as a function of stellar irradiance and orbital distance, assuming radiative balance between absorbed stellar flux and emitted thermal radiation, and a CO2-H2O greenhouse effect11. Under some optimistic formulations, the solar system’s HZ may span from early Mars, which preserves abundant evidence for ancient surface water activities12,13, to recent Venus, which assumes that Venus did not undergo a runaway transition to a hot climatic state until ~1 Ga (i.e., billion years) ago14,15 (Fig. 1).
However, a star-centric HZ definition overlooks the influence of planetary properties, including atmospheric and interior evolution, and some climate feedbacks. For example, recent models suggest that Venus may never have had a liquid-water ocean on its surface because of early climatic evolution and cloud feedbacks16. Conversely, although modern Mars lies within the outer boundary of the conventional HZ, its low mass allowed for substantial atmospheric escape, leading to a tenuous atmosphere that cannot sustain persistent surface liquid water13. Meanwhile, evidence for extant subsurface Martian aquifers remains elusive17,18. Furthermore, the conventional HZ concept neglects endogenous energy sources, such as tidal dissipation and radiogenic heating, which can maintain liquid water independently of stellar insolation. This limitation is underscored by growing evidence of subsurface liquid water bodies or oceans in icy bodies in the outer solar system, including dwarf planets like Ceres and Pluto and moons such as Europa, Ganymede, Callisto, Enceladus, and Titan19 (Fig. 1). Together, these discoveries have broadened the concept of planetary habitability beyond the classical HZ, in both space and time (Fig. 1). This expanded perspective also extends to exoplanetary systems, where an increasing number of candidates may host surface oceans20, potentially accompanied by ocean-bearing exomoons21.
However, life is not constituted by water alone; it requires a specific suite of bioessential elements dissolved within it. Beyond hydrogen and oxygen, a cell’s core chemical formulae include carbon, nitrogen, phosphorus, and sulfur, alongside a range of minor and trace elements22. The molar ratio of C:N:P in marine biomass on Earth, the Redfield ratio, is broadly consistent across diverse contemporary life forms23. Intriguingly, this ratio correlates with the elements’ relative bioavailability in the modern ocean24, underscoring the critical roles of aquatic chemistry in supplying life’s building blocks and life’s ability to exert a strong control on elemental ratios within said aquatic chemistry. Consequently, the chemical composition of a planetary water reservoir is not only a critical factor in assessing its habitability2 but may also directly reflect the life within it. However, not all water is habitable; water activity (aw), controlled by the interactions between ions and dissolved organic molecules in aqueous solution, can inhibit cellular physiology when solutes reach high concentrations25,26,27.
At steady state, planetary water chemistry is predominantly governed by water-rock interactions, processes that are highly sensitive to environmental conditions including temperature, pressure, pH, redox state, and rock composition. Throughout Earth’s geological history, evolving surface conditions have driven profound changes in ocean chemistry, which both shaped and were shaped by the biosphere28,29. Other ocean worlds in our solar system, however, have followed vastly different evolutionary paths dictated by distinct accretion and orbital histories, tectonic regimes, impact fluxes, and atmospheric evolution19. Consequently, the chemical features and evolutionary histories of these extraterrestrial oceans likely differ significantly from Earth’s (e.g., Xu et al30.), with profound implications for their capacity to support life as we know it.
In this review, we first synthesize the chemical features of water reservoirs and their biological implications on the modern Earth (Section 1). Based on this foundation, we then outline constraints on the evolutionary history of Earth’s ocean chemistry and discuss its interplay with the biosphere (Section 2). Next, we evaluate geological evidence for aqueous chemistry on early Mars and discuss its consequences for habitability and biosignature preservation (Section 3). We then examine the inferred chemical environments of four representative ocean-bearing icy bodies, namely dwarf planet Ceres, Jupiter’s moon Europa, and Saturn’s moons Enceladus and Titan, and highlight current knowledge and key uncertainties (Section 4). Finally, we conclude by outlining open questions in planetary ocean chemistry that are paramount to address through future research to fully evaluate habitability across the solar system.
Section 1. Water bodies on the modern earth: a framework for planetary habitability
Earth is predominantly covered by liquid water (~71% of its surface), and ~96.5% of its total surface water inventory resides in the ocean, while only a minor proportion is distributed among saline lakes and freshwater reservoirs6. These surface water bodies are integral components of the global hydrological cycle yet exhibit striking chemical diversity (Table 1), primarily governed by the extent of water-rock interaction and local hydrodynamic conditions. The hydrological cycle begins with rainfall. Rainwater is dilute, with only minor dissolved solids (typically ~10 mg/L), oxidizing, and weakly acidic (pH ~ 4–6), due to equilibration with atmospheric CO2 and O26,31. In surface or near-surface drainage, rain and river waters cause chemical weathering of terrestrial rocks6,31, releasing alkalinity and dissolved salts, including oxidants, which are ultimately transported to the ocean via rivers and subterranean groundwater discharge. In the marine environment, solutes accumulate until they are removed through the deposition and burial of authigenic minerals in local environments32, entering the large-scale, long-term deep element cycle via sediment recycling.
Table 1 Examples of aquatic habitats and the characteristic metabolisms on the modern Earth
Beyond these major surface water reservoirs, a diverse suite of smaller-volume water bodies exists with a very broad array of physicochemical profiles (Table 1). These environments, which include hypersaline lakes, brine systems in sea ice or in sedimentary pore waters, subglacial lakes, and hydrothermal vent systems, can exhibit extreme ranges in temperature, pressure, pH, salinity, and redox conditions33. Their chemical differences commonly reflect isolation from major surface water reservoirs within diverse lithologies, allowing them to sustain steep physicochemical gradients against the backdrop of Earth’s generally mild and oxidizing surface conditions. Although these extreme environments are rare on the modern Earth, they are highly relevant as analogs for environments proposed for the origin of life on the early Earth and potentially in water bodies on other planetary surfaces and interiors. For instance, alkaline soda lakes and low-temperature serpentinization-driven hydrothermal systems have been proposed as key settings for the origin of life on Earth34,35,36 as well as in the subsurface ocean of Enceladus37, due to steep chemical gradients and high availability of nutrients and reductants. Similarly, terrestrial hot spring systems have been proposed as analogs for ancient hydrothermal environments on early Mars38. Consequently, the study of these extreme aquatic systems on Earth provides a critical foundation for assessing the habitability and potential biosignatures of extraterrestrial environments33,39,40.
Microbial life is ubiquitous in modern aquatic environments, with the active assemblage (encompassing a range of functionality and activity level) and the residence time of characteristic microbes largely determined by the physicochemical parameters of water. In sunlit water, oxygenic photoautotrophs—including prokaryotic cyanobacteria (e.g., Prochlorococcus, Synechococcus) and eukaryotic phytoplankton— capture solar energy to drive primary production41 (Table 1). Dissolved organic matter derived from these photoautotrophs sustains heterotrophs, which remain numerically dominant across most aquatic systems42. Going deeper down the water column, light becomes limiting, and thus, photosynthesis declines and then ceases. Organic matter derived from descending photosynthetic biomass fuels heterotrophic respiration as it settles into dark environments43. In oxygen-minimum zones and anoxic sediments, aerobic respiration is replaced by anaerobic respiration, where the oxidation of organic carbon is coupled to the reduction of nitrate, sulfate, or other available oxidants32,44. Anoxic habitats also host microbial consortia capable of conducting anaerobic oxidation of methane coupled to sulfate reduction, a process largely mediated by consortia of anaerobic methanotrophic archaea (ANME) and sulfate-reducing bacteria (SRB)45 (Table 1).
Elevated concentrations of metabolic substrates can be supplied by water-rock interactions (e.g., serpentinization) and associated abiotic synthesis of organic carbon (e.g., Fischer-Tropsch reaction), as well as during radiolysis. In some interface environments, the mixing of reduced (e.g., Fe2+, H2, CH4, H2S, and organics) and oxidized substrates (e.g., O2, Fe(III), and SO42-) can build redox disequilibrium that provides the energetic driving force for chemolithoautotrophic metabolisms46,47,48. These microorganisms dominate the base trophic level on the modern Earth49,50, including anaerobes such as methanogens, ANME, acetogens and SRB, as well as aerobes like sulfur-oxidizing and hydrogen-oxidizing bacteria (Table 1). Even at extreme temperature, pH, pressure and salinity conditions, diverse microorganisms have evolved to thrive in hostile environments such as hydrothermal vents and subglacial lakes33,51 (Table 1). These discoveries have continuously expanded the boundary conditions of life39,40, and thus increase the plausibility of life surviving in other planetary waters33.
Importantly, there is often a large diversity of low-abundance taxa52 and spores53 in addition to the dormant forms mentioned above, and together these constitute an integrated microbial community. At steady state, the overall biomass of each ecosystem is dependent upon the availability of energy or nutrients, which is strongly shaped by local or global hydrological processes, and can vary significantly across modern environments. For example, cell concentrations can reach > 108 cells mL−1 in nutrient-rich eutrophic coastal waters (e.g., Li et al54.; Yuan et al55.), largely sustained by riverine transport and ocean upwelling56,57, but decline to <103 cells mL−1 in nutrient-poor subsurface waters58,59. While saline lakes and freshwater lakes and rivers together have previously been estimated to host ~2 × 1026 prokaryotic cells (i.e., archaea and bacteria)60, the majority of prokaryotic cells reside in the marine pelagic zones (~1 × 1029 cells at 0–4000 m water depth)61. In addition, despite having extremely slow cellular turnover times62,63, marine sediments and terrestrial groundwater can also host significant cell populations (> 1029 cells)61. Comparatively, waters on the modern Earth, characterized primarily by a moderate range of physio-chemical parameters, have built a robust and complex ecosystem with much higher primary productivity than the reconstructed levels for the ancient Earth64 (see also Section 2) and theoretical estimates for extraterrestrial water bodies65,66,67 (Section 3 & 4).
Section 2. Co-evolution of the hydrosphere and biosphere on the Earth
Earth’s surface has hosted a liquid hydrosphere since the early Hadean68,69. The planet’s water is thought to have been delivered by H-bearing planetesimals and meteorites and/or formed during the interaction of Earth’s early magma ocean with nebular gas, though the relative proportions remain debated70,71. Nevertheless, high oxygen isotope ratios in detrital zircons suggest the presence of liquid water near Earth’s surface as early as 4.4 Ga69. Despite significantly lower solar luminosity at this time, a “Faint Young Sun” was likely compensated for by high atmospheric concentrations of greenhouse gases72, which maintained temperate conditions for sustaining a persistent liquid surface ocean. Direct geological constraints on the Hadean ocean’s chemistry, including its pH, salinity, and trace element inventory, are sparse and contentious73,74,75. Furthermore, the habitability of these primordial waters was likely frequently perturbed by intense impact bombardment76,77.
Despite the above considerations, principles of water-rock interaction allow for general inferences about early aquatic environments (Fig. 2). High volcanic outgassing fluxes of CO2 and other acidic volatiles likely rendered most surface waters weakly acidic to circumneutral78,79. However, progressive silicate weathering would have gradually increased ocean alkalinity and pH over time78,79, perhaps buffering at a value around 6.5-7 in the Archean (Fig. 2a). Before the emergence of large continental landmasses, alkalinity and critical greenhouse gases (such as CH480 and potentially H281) released from seafloor basalt weathering and impact ejecta probably served as a principal feedback stabilizing Earth’s early climate72, although these reactions remain poorly quantified77,78,82. These extensive water-rock interactions might also have helped maintain reducing surface waters and an oxygen-poor atmosphere in the early Earth81,83. Ocean salinity in the Hadean and Archean is poorly constrained84 but might have reached levels close to or higher than the modern, given greater volcanic volatile input74 and a reduced evaporite sink in the absence of extensive subaerial continents85. Intense hydrothermal alteration, coupled with limited riverine input, may have led to seawater enriched in Ca2+ but relatively depleted in Mg2+ compared to today78,86 (Fig. 2d). Moreover, surface waters would accumulate reductants like Fe2+ and Mn2+ under a weakly reducing atmosphere devoid of biogenic oxygen 75,87, in contrast to the modern oxic surface waters (Fig. 2b & 2c). In such anoxic and ferruginous seawater, rapid reduction and reaction with abundant hydrothermal Ba2+ and Fe2+ would have kept dissolved sulfur levels in the micromolar range88,89 (Fig. 2d). Ammonium (NH4+) would have been the thermodynamically stable form of dissolved nitrogen90, and small amounts of NH4+ were probably produced in hydrothermal settings91 and via the reduction of lightning-generated nitrogen oxides92. In the absence of biosilicification, the primitive ocean would also have been rich in dissolved silica (Fig. 2d), derived from hydrothermal leaching of the seafloor93, leading to widespread silica precipitation94,95 and silicification of early sediments96,97. High dissolved silica is also proposed to favor reverse weathering in the early ocean, which might help stabilize the long-term evolution of climate and marine pH98.
Fig. 2: Co-evolution of life and Earth’s surface environments.
a Reconstructed evolution of ocean pH. The solid blue lines show median pH estimates from Krissansen-Totton et al78., with the blue shading denoting the 95% confidence interval. Yellow dashed lines indicate the 95% confidence interval from Halevy and Bachan79. b Evolution of ocean redox stratification structure239. c Reconstructed evolution of atmospheric O2 and CO2 levels through time121,240. d Evolution of dissolved bio-essential elements in subsurface seawater and their dominant chemical forms28,86,96,97,241,242,243. The [P] (% modern levels) axis represents the percentage of P concentrations in the ocean relative to modern levels. e Major evolutionary milestones of life, shown by red symbols and lines (upper), together with the evolution of total primary productivity, shown by the blue dashed field and lines (lower)64. f Geological constraints on the emergence of major microbial metabolisms during the Archean Eon29. Dark age bars indicate relatively well-constrained age estimates, whereas lighter age bars indicate more uncertain ranges.
This geochemical backdrop, or at least portions of it, provided the physical setting for life’s emergence. Reducing conditions favored the abiotic synthesis and stability of organic molecules and their precursors99,100. Key nutrients like nitrogen and phosphorus may have been more abundant or in more bioavailable forms than today90,101,102. Furthermore, extensive water-rock interactions produced catalytic mineral surfaces (e.g., phyllosilicates, native metals, metal oxides, and sulfides)34,103,104 that could have facilitated prebiotic chemistry. Meanwhile, large thermal and chemical gradients, sustained by surface radiation, impacts, and geothermal heating, provided the thermodynamic disequilibrium necessary to drive early metabolisms34,104. Phylogenetic reconstructions indicate that the Last Universal Common Ancestor (LUCA) was likely a chemoautotroph, exploiting these abundant chemical energy sources105.
Life probably emerged in the late Hadean or early Archean105 (Fig. 2e), coinciding with a decline in impact frequency106. At the same time, continuous dissipation of Earth’s internal heat would have resulted in a secular decline in geothermal heat flow107, leading to reduced volcanic activity108. In the meantime, continental landmass grew substantially in areal extent109,110; however, much of it may have remained largely submerged until the middle or late Archean111. These shifts drove a decrease in atmospheric pCO2 and a rise in ocean pH and alkalinity78 (Fig. 2a). The emergence of continental landmasses also established modern-style lacustrine and riverine systems, promoting large-scale continental weathering and marine upwelling supplies of nutrients that could support rising biological productivity64,90,112,113 (Fig. 2e). Moreover, the establishment of these environmental settings might have spurred a diversification of microbial metabolisms (Fig. 2f) as more oxidants and nutrients became available29,114. Among them, one critical innovation is the emergence of cyanobacteria probably in the middle to late Archean115,116 (Fig. 2e). However, the atmosphere and surface ocean remained largely anoxic until the Great Oxidation Event (GOE, at ~2.4 Ga; Fig. 2c)117,118, possibly due to either slow biological evolution of efficient O2 production83,119 or long duration to saturate various O2 sinks near Earth’s surface81.
The GOE triggered global oxidative weathering, introducing oxidants to the ocean90,118,120. However, atmospheric O2 levels likely remained low (<10% of present levels) during the Proterozoic118,121, resulting in a stratified ocean: oxygenated at the surface but anoxic and largely ferruginous, with localized euxinic margins (H2S-rich) at depth122 (Fig. 2b). The reasons for this protracted period of intermediate oxygen levels despite cyanobacterial dominance remain unclear, but may be linked to persistently low primary productivity, as suggested by invariant carbon isotope records and significantly negative triple oxygen isotope values123,124. The assembly of Nuna and Rodinia supercontinents125, coupled with overall low continental topographic relief126, might have limited nutrient supply via continental weathering and ocean upwelling126,127, constraining marine productivity128. Despite this apparent environmental stability, the mid-Proterozoic witnessed the emergence of eukaryotes and multicellular organisms129 (Fig. 2e), potentially driven by local habitat heterogeneity or environmental perturbations130,131. The subsequent breakup of Rodinia and increased orogenic activity in the late Proterozoic enhanced continental weathering132, nutrient supply, and oxygen production, culminating in the second oxygenation of the atmosphere and ocean (Fig. 2b, c) and the radiation of complex eukaryotic life in the Ediacaran and early Phanerozoic133 (Fig. 2e).
The rise of complex Phanerozoic ecosystems further shaped biogeochemical cycles. The colonization of land by plants in the Ordovician to early Devonian significantly amplified the continental supply of weathering products134, including bio-limiting nutrients (e.g., phosphorus), to the oceans, promoting organic carbon burial and further oxygenation of the atmosphere and deep ocean135. This established or helped sustain a fully oxygenated ocean with high levels of oxidants (including oxygen, sulfate, and nitrate) but low dissolved iron, similar to the modern (Fig. 2b, d). The end-Permian mass extinction created ecological niches for modern eukaryotic phytoplankton to radiate in the Mesozoic, notably dinoflagellates, coccolithophores, and diatoms136. These groups might greatly enhance the efficiency of the biological pump and carbonate factory, increasing carbon export to deep sediments136,137. Moreover, diatoms began to efficiently strip dissolved silica from seawater, resulting in low silica oceans97. Stepwise, these biological innovations drove the ocean toward its modern chemical state (as detailed in Section 1).
In summary, the chemical evolution of Earth’s hydrosphere and its habitability have been jointly dictated by the co-evolution of geosphere and biosphere (Fig. 2). The secular cooling of the Earth reduced volcanic and hydrothermal activities, while the emergence of land increased continental weathering and marine upwelling supplies of nutrients. These tandem processes drew down atmospheric CO2, raised ocean pH, and altered nutrient inventories (e.g., sulfur, phosphorus, nitrogen, and trace metals), which in turn controlled biological primary productivity and metabolic evolution. The subsequent rise of complex life amplified organic carbon burial, leading to a fully oxygenated surface environment that ultimately ushered in the robust and complex ecosystems of today.
While this Section focuses on the generic evolutionary history of Earth’s oceans, it is crucial to acknowledge the persistently important roles of diverse aquatic environments, such as lakes, rivers, hot springs, and hydrothermal systems. Such environments would have been established as soon as the surface hydrosphere appeared, although their relative abundance may have fluctuated considerably over geological time. Furthermore, evidence suggests that the colonization of terrestrial aquatic habitats may have begun as early as the Archean138,139, with these environments becoming increasingly important following the emergence of complex life (e.g., plants140,141; Fig.2e). The biological activity within these environments, in turn, significantly shaped the habitability of the terrestrial realm138,141. However, compared to the marine record, the long-term chemical evolution of these continental and niche water bodies is poorly constrained, due to a general scarcity of continuous sedimentary archives. This challenge is compounded by their intrinsic chemical heterogeneity, which is highly sensitive to local environmental and hydrological conditions.
Section 3. Evolution of hydrosphere and habitability on Mars
The present-day surface of Mars is forbiddingly cold and dry, making it an unlikely habitat for life. Orbital and rover exploration over the past half century, however, demonstrates that Mars was not always like this142,143. Early in its history, our planetary neighbor boasted rivers and lakes, perhaps even an ocean (Fig. 3). Early orbiters, in particular those of NASA’s Mariner and Viking programs, clearly imaged drainage channels much like those incised by flowing water on Earth144. Ensuing spectroscopic mapping by NASA and ESA orbiters documented the widespread distribution of clays, hematite and other minerals generally formed in association with liquid water145,146. Moreover, observations by the Mars Odyssey orbiter documented high concentrations of hydrogen at or just beneath the Martian surface, especially at higher latitudes147. The interpretation of these observations as a signature for water ice was corroborated by orbital imaging of surface features similar to those associated with near-surface ground ice on Earth17.
Fig. 3: Temporal evolution of key surface and interior processes on Mars.
a Schematic evolution of Martian atmospheric density, which remains poorly constrained. Geological and geochemical evidence suggests that early Mars may have sustained a denser atmosphere than today, at least episodically. b Conceptual evolution history of liquid-water reservoirs on Mars. Early Mars may have hosted relatively stable surface-water bodies, including lakes and channel-forming fluvial systems. As atmospheric lost progressively, stable surface liquid water became increasingly difficult to sustain, and any remaining liquid water was likely restricted mainly to the subsurface. c Timing of major impact events on Mars. Question marks indicate uncertainties in the duration and intensity of impact episodes. d Evolution of major volcanic episodes on Mars. e Representative secondary minerals formed through water-rock interactions on Mars. Based on data from Sasselov et al218., Wordsworth12, Ehlmann et al244., Ramirez and Craddock245.
With the successful landing and deployment of the NASA rovers Spirit and Opportunity in 2004, it became possible for the first time to study Mars and its history in much the same way that field geologists do on Earth, via the physical and geochemical investigation of ancient Martian sedimentary rocks exposed in outcrop, largely associated with meteorite impacts (Fig. 3c). Opportunity’s exploration of sedimentary rocks in Meridiani planum found evidence of ripple marks known to be shaped by flowing water on Earth; Opportunity also discovered abundant evaporite minerals that record rock-water interactions (Fig. 3e). Early diagenetic hematite concretions additionally record iron oxidation associated with ground water percolation. Taken together, evidence from Meridiani planum indicates that Mars hosted a regionally extensive arid, acidic and oxidizing environment more than three billion years ago148,149.
Opportunity’s twin rover Spirit, operating on the other side of the planet, found further evidence of rock-water interactions but within older strata150. Notably, a failed wheel, dragged through surficial regolith in Gusev Crater, fortuitously discovered a white material beneath surficial dust – amorphous silica, likely formed by the alteration of basalt by hydrothermal fluids151. Rovers Opportunity and Spirit, thus, corroborated and enriched orbital evidence that liquid water once existed at the Martian surface. But they did more than that, providing insights into the properties and persistence of early water bodies. Water is key to life as we know it, but not all water bodies are habitable. In particular, aw sets an environmental limit for essentially all species found on Earth27. Ions and dissolved molecules bond weakly with H2O, limiting the amount of water available for physical and biological activities. Every species has a lower limit of tolerance for aw, and calculation of aw in some ancient Martian waters suggests that evaporation could drive it to low levels that may become uninhabitable for most known life on Earth25,26.
The persistence of surficial water is another important consideration for ancient habitability within the Martian environment. Geochemical observations of jarosite, amorphous silica and smectite clays in ancient Martian sedimentary rocks provide compelling evidence for liquid water at their time of formation (Fig. 3e). However, the fact that continued interactions with water generally transform such minerals into one other, more stable species on Earth suggests that on Mars these minerals have not seen much water since they formed billions of years ago152, thus suggesting Mars has been persistently dry for a long time.
In the wake of Opportunity’s and Spirit’s pioneering exploration, two more recent rovers have greatly improved our understanding of Martian surficial environments in both space and time. Since 2012, Curiosity has explored a thick succession of sedimentary rocks exposed in Gale crater, and, since 2020, Perseverance has been investigating sedimentary deposits in and around an ancient delta preserved in Jezero crater. Mars chronology is currently in a state of flux, as confidence in the Late Heavy Bombardment hypothesis has waned (Fig. 3c), but in-situ geochronological analysis indicates that sedimentary rocks within Gale crater are 4.1 – 3.6 billion years old153, older, on stratigraphic grounds than the evaporite-rich rocks in Meridiani planum (Fig. 3e). Deltaic and delta-front deposits in Jezero crater may be comparably old–again, both older and environmentally distinct from those discovered in Meridiani planum. At both newer localities, the rovers discovered lacustrine sedimentary rocks that not only preserve sedimentological and chemical evidence of water, but also have organic molecules, phosphorus, nitrogen, redox gradients, and a more bio-friendly aw, collectively opening the possibility that these environments were, in principle, habitable154,155. Indeed, in 2025, a silty sedimentary rock bed in Jezero crater was found to contain abundant small, irregular patches containing vivianite, an iron phosphate mineral, and greigite, an iron sulfide, interpreted as a potential biosignature156. The pattern is indeed suggestive, although other explanations are possible. Rock samples were cached for eventual return to Earth, and if and when that happens, scientists may be able to discriminate among various hypotheses for the rock’s origins.
With questions of biology in mind, it is useful to return to questions of persistence. In comparison to the Earth, the window of habitability on early Mars may have been cut short. For decades, Martian environmental history was characterized in terms of a warm wet interval that persisted for at least hundreds of millions of years in Mars’ youth. However, today, there is strong reason to interpret the history of Martian surface water as one of episodic warmth and surface water, with clement environments persisting for perhaps tens of thousands of years every so often during the planet’s early history. Several hypotheses have been proposed to explain the episodicity of surface water, including transient warmth and wetness engendered by large impacts and volcanic eruptions (Fig. 3c,d), as well as episodic shifts in obliquity157,158. In any event, drivers of transient warmth and wetness declined in frequency and size, eventually resulting in the Mars we see today.
So, what happened to Mars’ surface water? Isotopic ratios of various gases support the hypothesis that large impacts and hydrodynamic escape removed the primordial atmosphere, and once Mars’ short-lived magnetic field dissipated, solar wind interactions stripped away remaining or secondary atmospheric gases through time159. These data also indicate that much of the Martian atmosphere-surface H2O reservoir was lost rapidly, perhaps as early as ~3 billion years ago160 (Fig. 3a). In contrast to Earth’s subduction-driven mode of volatile recycling between its interior and surface, the lack of a similar mechanism on Mars may have set it on an irreversible path toward desiccation as volatiles were permanently lost to its ancient crust through alteration processes160,161. Some water vapor would also have been lost by photodissociation of water vapor or ice in the upper atmosphere162,163, resulting in hydrogen escape to space while leaving oxygen gas to be consumed by the oxidation of reduced iron at the surface. Of course, as noted above, a considerable amount of H2O remains as ice within surface rocks and regolith, as well as water ice in polar ice caps. Recent geophysical data that are interpreted as requiring a large reservoir of water at depths of 11 to 20 kilometers below the Martian surface164 can be also explained without liquid water165 and so remain controversial. In summary, while present day Mars is sufficiently cold that liquid water is not stable at its surface, during its early history it supported episodic rivers and lakes, some of them at least transiently habitable.
Section 4. Subsurface oceans in icy bodies of the outer solar system
Beyond the terrestrial planets, our solar system hosts a diverse population of ocean worlds among dwarf planets and icy satellites, including Ceres, Europa, Ganymede, Callisto, Enceladus, Titan, Triton, and Pluto (Fig. 1). In stark contrast to Earth’s surface oceans, these extraterrestrial bodies harbor their liquid water reservoirs beneath thick icy crusts, often sandwiched between layer(s) of ice and rock (Fig. 4). Some subsurface oceans are believed to have persisted for at least 107 years166,167,168,169, sustained by heat from radioactive decay and/or tidal dissipation19. There are also strong signs of past or ongoing hydrothermal activities in some planetary bodies, particularly Enceladus170. Moreover, remote geochemical analyses of their surface and plume materials suggest availability of bio-essential nutrients and organic species in the oceans (Table 2). These lines of evidence jointly indicate that these subsurface oceans are likely habitable and potentially conducive to the origin and survival of life. This section synthesizes the up to date understanding of the ocean chemistry and habitability of four key representatives: Ceres, Europa, Enceladus, and Titan.
Fig. 4: Interior structures and possible material exchange pathways in ocean-bearing icy moons.
This schematic illustration compares two major types of ocean-bearing icy moons: a smaller icy moons, such as Enceladus and Europa, in which a subsurface ocean may directly overlie a poorly differentiated silicate interior; and (b) larger, more strongly differentiated bodies, such as Titan and Ganymede, in which high-pressure ice layers separate the ocean from the deep rocky interior. Chemical species produced at the surface or within the interior may be exchanged through ice-shell recycling or subduction, ocean circulation, plume activity, as well as convection and melting within high-pressure ice layers. These possible transport processes regulate ocean chemistry and, consequently, the habitability of subsurface oceans. The structure is not shown to scale.
Table 2 Four representative icy worlds with putative subsurface liquid water bodies in the solar systemCeres: a relic ocean world
Ceres, a dwarf planet in the asteroid belt, is a highly hydrated body. Our understanding of Ceres was revolutionized by the Dawn spacecraft (2015-2018). Thermal evolution models suggest that Ceres once hosted a global subsurface ocean, formed early in its history by the metamorphic dehydration of hydrated materials166. While this global ocean likely froze due to limited radiogenic or tidal heating, evidence points to the existence of regional brines in the subsurface171. These brines are implicated in forming recent geological features and precipitating salts and secondary minerals on the surface. The brine is inferred to be reducing and alkaline, as indicated by observations of Fe(II)-bearing minerals and brucite172 (Table 2). The presence of ammonium-bearing phyllosilicates and carbonate salts (Table 2) further suggests considerable levels of dissolved NH3 and carbonate ions173,174. Moreover, surface analyses indicate a high abundance of organic compounds (potentially up to 30 wt%)175 (Table 2), likely from a large subsurface reservoir. In summary, Ceres may host relict brines whose chemistry has been largely shaped by serpentinization reactions176, analogous to terrestrial serpentinizing systems.
Europa: a tidally-locked ocean world
Unlike Ceres, Europa is robustly inferred to host a global, long-lived ocean sandwiched between its icy shell and partially differentiated core177 (Fig. 1). This ocean likely formed early via dehydration of hydrated minerals167 and is sustained today by intense tidal heating178. While transient water plumes have been observed179,180 (Table 2), the lack of direct sampling means its composition must be inferred from remote sensing of the surface. Current models suggest a water mass dominated by Na+, Mg2+, and Cl-181 with K+ and Ca2+182 (Table 2), but a central controversy surrounds the origin of surface sulfates – whether they are sourced from the ocean or produced in situ by radiolysis of sulfides183,184. This ambiguity propagates into major uncertainties in the ocean’s pH and redox state182,185. Recent observations of surface CO2 (Table 2) have been interpreted as evidence for an acidic and oxidizing ocean186, despite of several alternative interpretations in lack of reliable constraints. However, if true, this condition would not favor the stability of organics from accretion or impact delivery182, which may partly explain the lack of detectable organic materials to date. Moreover, any organics emplaced on the surface would be subjected to intense ionizing radiation from Jupiter187, undermining their survival in surface ice188. The surface radiation could also generate potent oxidizing species, such as O2 and H2O2 at Europa’s surface189 (Table 2). Possible ice tectonics, driven by tidal flexing190 (Fig. 4a), may help transport these oxidants to the subsurface ocean, providing oxidant substrates for possible chemosynthetic life67. However, recent geophysical studies suggest that Europa’s seafloor may not currently host active volcanism191,192 or significant ongoing water-rock interaction167, limiting the release of reductants and buildup of redox gradients necessary to sustain life on Europa67,193,194.
Apart from tectonic activities, ocean circulation is another key control on the transport and mixing of surface-derived oxidants and interior-derived reductants, and thus on the buildup of chemical disequilibria for potential metabolisms. Unlike Earth’s oceans, icy moons’ subsurface oceans lack wind forcing and are thought to be driven primarily by buoyancy contrasts associated with basal heating and ice-shell melting and freezing195,196 (Fig. 4a). Although this circulation may operate under a different and possibly less efficient mixing regime than that of Earth’s ocean, it also governs the extent to which hydrothermal products and other potential biosignatures can be transported from the seafloor to the ice shell or surface197, thereby strongly influencing their detectability.
Enceladus: an active ocean world
Among icy bodies, Enceladus provides the most direct evidence of a subsurface ocean, i.e., the ongoing water plume observed by the Cassini spacecraft in its south polar region198 (Table 2). Geophysical models further indicate a global, long-lived liquid ocean (> 108 years), sustained by tidal dissipation along with radiogenic heating169,199. Detailed analyses of the emitted gases and ice grains by the Cassini’s spectrometers reveal a reducing chemical environment rich in CO2, H2, CH4, and NH3, alongside a diverse suite of organic compounds200,201,202 (Table 2). The salt composition of the ocean, inferred from analyses of the emitted ice grains, is dominated by NaCl, KCl, and NaHCO3/Na2CO3203 (Table 2). Recently, phosphorus, a key limiting nutrient for life on Earth, has been revealed to accumulate to significant abundance (~mM) within Enceladus’s ocean204,205, which is at least three orders of magnitude higher than modern levels in Earth’s oceans (Table 1). Compared with other icy bodies, Enceladus has the most compelling evidence for ongoing high-temperature (> 90 °C) hydrothermal activity with the detection of silica nanoparticles in the plume170, although recent geophysical model suggests the ocean stratification would impede particulate transport from the hydrothermal vents to the plumes of Enceladus197. Nevertheless, intensive water-rock interaction and potential hydrothermal activity would help build chemical and energetic gradients as well as establish mineral interfaces in the Enceladus ocean floor, which are essential for the origin and sustainability of life104.
Titan: an organic-rich ocean world
Titan is unique among icy worlds because of its dense nitrogen-methane atmosphere (~1.5 bar; Table 2) and thick orange haze206,207. The successful Cassini-Huygens mission fundamentally advanced our understanding of Titan’s complex organic chemistry (Table 2) and revealed a possibly global subsurface ocean beneath its icy shell208,209. Radar and landmark-tracking observations identified surface motions decoupled from the solid interior, indicating an outer ice shell floating above a liquid layer210. Gravity and rotational data further support a differentiated interior, consisting of a low-density outer shell overlying a denser rock/ice core with an intervening global ocean208. However, latest analyses suggest that this liquid reservoir, if present today, may be spatially limited or replaced by a partially molten high-pressure ice layer rather than global ocean211. The liquid layer, if it exists, is likely tens to hundreds of kilometers thick, situated between an ice Ih shell and high-pressure ice mantle168. Its composition probably includes water, ammonia, dissolved salts (e.g., Na+, K+, Cl-), and possibly organics delivered from the surface and/or subsurface212,213. Ammonia may act as an antifreeze, which helps maintain the liquid water layer(s) under Titan’s frigid temperatures214,215. Potential energy sources that sustain the ocean include tidal heating and radiogenic decay, although the energy fluxes are modest compared to other active ocean worlds168. Nevertheless, several dynamic processes, such as cryovolcanism, impact-driven melting, and diapiric transport, are proposed to facilitate the exchange of material and energy between Titan’s surface organic reservoir and its interior ocean207,213.
Crucially, Titan’s atmosphere provides an important analog for understanding prebiotic synthesis under reducing atmospheric conditions, both on Titan and on early Earth. Photochemical reactions involving CH4 and N2 produce nitriles and other complex organics, while the formation of an organic haze exerts significant climatic effects – phenomena often invoked in models of Earth’s prebiotic environment following large impacts or other transient reducing episodes76. These atmospheric processes establish a continuous source of chemically reactive organics to Titan’s surface and interior, potentially feeding aqueous environments with prebiotic precursors. Titan’s liquid water layer(s) could therefore be potentially habitable, supporting aqueous geochemistry, redox reactions, mineral dissolution, and organic processing216. However, major uncertainties remain regarding the ocean’s redox balance, energy availability, and inventory of bio-essential elements. Despite these uncertainties, Titan stands out as a complex, organic-rich ocean world. Its subsurface water-ammonia liquid layer(s), coupled to an active surface and thick organic atmosphere, makes Titan a compelling target for exploring habitability beyond Earth217.
In summary, subsurface oceans on dwarf planets or icy moons fulfill several basic criteria of planetary habitability: active and persistent liquid water, available bio-essential elements and energy, suitable physicochemical conditions, and prolonged environmental stability2. However, these environments are inherently more limited than the complex planetary systems of early Earth or Mars, although possibly conducive to some origin of life processes218. While a diverse array of organic molecules has been detected, key biomolecules like amino acids and nucleotides remain elusive. Furthermore, prebiotic pathways that rely on environmental cycles—such as wet-dry cycles for polymerization—are difficult to envision in these perpetually submerged, ice-covered settings (Fig. 4). Nevertheless, if life could have originated in these environments, it is most likely chemotrophic due to the lack of sunlight65,66,219. Speculative theoretical estimates of potential biomass, though highly model-dependent, suggest plausible populations for ecosystems of methanogens on Enceladus65,66 or SRB on Europa67. However, these estimated biomasses are many orders of magnitude lower than those in Earth’s oceans, which may limit the evolutionary rate of potential ecosystems per se on such bodies64. This profound difference, coupled with the formidable technical challenge of probing through kilometers of ice, defines the primary obstacle for future missions: the detection of a potentially simple and sparse biosphere in a vast, hidden ocean.
Moving forward
Across ocean-bearing worlds, several cross-cutting factors emerge as universal controls on habitability. First, fluid-rock interactions act as a geochemical engine, regulating pH, alkalinity, and redox state while mobilizing nutrients and generating chemical energy. Second, redox gradients and chemical disequilibria provide sustained energy sources for metabolism, whether through radiolytic oxidants on Europa or potential hydrothermally-derived reductants on Enceladus (Fig. 4). Third, the persistence and stability of liquid water set the temporal window for prebiotic chemistry and evolution; Mars likely experienced episodic clement intervals and short-lived lakes, whereas Europa and Enceladus may maintain long-lived subsurface oceans sustained by tidal and radiogenic heating; Titan’s water-ammonia liquid layer(s) demonstrates persistence at cryogenic temperatures (Fig. 1). Fourth, the availability and speciation of bioessential elements, including carbon, nitrogen, phosphorus, and sulfur, govern feasible metabolisms and energy yields. Finally, environmental extremes impose physiological limits. On Mars, brines often drove aw below terrestrial thresholds for biological activities; on Europa, higher pressure shapes fluid properties and mineral equilibria; on Titan, ammonia lower the freezing point and modulates buffering capacity. Temperature is a universal constraint: on Earth, the upper thermal limit for life approaches ~120 °C in high-pressure environments where water remains liquid.
Taken together, these controls suggest a multidimensional framework for habitability–integrating the persistence of liquid water, energy from redox disequilibria, nutrient supply and speciation, and the envelope of physicochemical limits (including aw, pressure, and temperature). Factors specific to each planetary body then modulate this framework: Mars’ episodicity and periodically low aw, Ceres’ relict brines shaped by serpentinization, Europa’s oxidant delivery through its ice shell, Enceladus’ hydrothermal flux and phosphate abundance, and Titan’s ammonia-water chemistry and atmospherically-supplied organics. Viewed through this lens, the following unknowns and aspirations naturally resolve into questions about (i) how redox gradients are generated, maintained, and transported; (ii) how nutrients are partitioned and speciated in time and space; (iii) how complex organics form, cycle, and persist; and (iv) how microenvironments at mineral interfaces, within ice, and in porous crust, link chemical conditions to physiological limits.
While the habitability of the ancient Earth and extraterrestrial oceans is often assessed through the lens of modern terrestrial biogeochemistry220, significant knowledge gaps persist. Key questions remain regarding the generation and maintenance of redox gradients, the speciation and bioavailability of essential nutrients and trace metals, and the production and cycling of complex organic matter. Furthermore, our understanding of chemical microenvironments—within biofilms, at mineral interfaces, and in concealed habitats like deep aquifers and subglacial lakes—lags far behind that of bulk solutions. A fundamental challenge is to quantitatively link chemical conditions to the energetic and physiological limits of life, determining how spatial and temporal variability dictates microbial survival and growth (Table 1). Finally, in an era of unprecedented anthropogenic change, the impact of human-driven perturbations on these fundamental aquatic processes and the microbial communities they support represents a critical and poorly constrained frontier of research.
Within our solar system, only on Earth has life persisted for some four billion years and transformed the planet221. The modern Earth has been profoundly shaped by the long-term evolution of its geosphere and biosphere220,222 (Fig. 2); however, major gaps persist in our understanding of this intertwined history. Although a surface hydrosphere was established early, the evolutionary trajectories of bio-essential elements, including key biomolecules and potentially limiting nutrients like phosphorus and trace metals, remain poorly constrained. Very likely, the origin of life and the environmental settings essential for its emergence differed significantly from the habitable conditions of modern organisms1, representing another fundamental frontier. Furthermore, while processes such as photochemistry, impacts, volcanic activities, and glaciation events are known to have heavily perturbed surface conditions, their specific effects on aquatic chemistry are not well characterized. In spite of the seemingly dynamic interplay between the long-term evolution of the biosphere and geosphere (Fig. 2), how and to what extent they actually co-evolved also requires deeper investigation. Compounding these challenges, the geological record is heavily biased towards marine sediments, leaving the evolution and role of non-marine aquatic environments throughout Earth’s history poorly constrained.
Beyond the Earth, ocean-bearing worlds, from early Mars to dwarf planets and icy moons, represent prime targets in the search for habitable environments and extraterrestrial life. However, Mars and possibly Venus are informative examples of bodies that have likely long passed their habitable windows (Fig. 1). Yet, fundamental questions persist regarding the evolution and longevity of their aqueous environments throughout the solar system, as well as the availability of critical bio-elements. These uncertainties are even greater for exoplanetary oceans, hindering a full assessment of their potential for prebiotic chemistry and their possibility for life. Moreover, a paramount challenge remains the detection and robust interpretation of potential biosignatures.
Fortunately, a new generation of missions is poised to transform our understanding. ESA’s Jupiter Icy Moons Explorer (JUICE) mission, launched in 2023, will arrive at Jupiter in 2031 and is designed to conduct detailed observations of three large ocean-bearing moons: Europa, Ganymede, and Callisto223,224. Meanwhile, NASA’s Europa Clipper mission, which is also en route to Jupiter and is expected to arrive in 2030, aims to search for water plumes on Europa and, if encountered, characterize material derived from its subsurface ocean through plume fly-throughs225,226,227. Moreover, NASA and ESA initiated the Mars Sample Return (MSR) program; accordingly, the Perseverance rover has already collected surface samples for potential future return. In parallel, China is implementing the Tianwen-III mission, planned for launch in 2028, which is designed to return drilled Martian samples potentially by 2031228. These Martian samples will provide unprecedented insight into the history of water and the potential for life on Mars. Together, these endeavors will not only constrain the habitability of other worlds but also offer a critical lens through which to re-examine the poorly preserved record of life’s origins and early history on Earth.