Across eight Mars years of sampling at Gale crater, Curiosity found peak oxygen-18 levels in carbon dioxide 10 to 30 parts per thousand higher in southern spring and summer than in the northern season, a swing that polar ice and sunlight cannot produce on their own.
Peer reviewed · Observational (in situ measurements at one site) · Water explanation is modelled, not observed
The carbon dioxide in Mars’ air gets measurably heavier every southern spring and summer. A team led by Heather Franz at NASA’s Goddard Space Flight Center compiled 72 measurements made by the Curiosity rover from 2012 to 2025 and found the extra heavy oxygen arrives in a seasonal pattern that the freezing and thawing of polar CO2 ice cannot explain on its own. Their proposed explanation centres on CO2 swapping oxygen atoms with water held on soil grains or in thin liquid-like films. That water has not been seen directly, and all the data come from one crater.
What you need to know
- Finding: Curiosity measured more oxygen-18 in atmospheric CO2 during southern spring-summer than northern spring-summer, with the gap between seasonal peaks at 10 to 15‰ in recent years and 20 to 30‰ earlier in the mission. Day and night samples showed no distinguishable difference.
- Method: 34 runs with the Tunable Laser Spectrometer and 38 with the mass spectrometer in the rover’s SAM instrument suite, Mars years 31 to 38, compared with a NASA Ames climate model and simple mixing models.
- Source: Franz and 25 co-authors, “A decadal record of isotopic trends in martian atmospheric CO2”, PNAS, published 16 November 2026, open access.
- Why it matters: If the proposed exchange is real, water in the top metres of Martian soil trades oxygen with the air every year, tying today’s water and carbon cycles together.
- Limitation: The water is inferred from isotope ratios and models, not detected, and a single near-equatorial site cannot separate local, regional and global causes.
Evidence at a glance
| Item | Assessment |
|---|---|
| Evidence status | Peer reviewed (PNAS, edited by Mark Thiemens of UC San Diego). Observational: direct measurements of Mars air by one rover at one site. The water explanation comes from models, not observation. |
| Primary source | Franz, H. B., Webster, C. R., Flesch, G. J. et al., “A decadal record of isotopic trends in martian atmospheric CO2”, PNAS 123 (47) e2615294123, 16 November 2026. Licence CC BY-NC-ND 4.0. |
| Evidence strength | Moderate for the seasonal pattern, because it repeats across several Mars years and two instruments, and the two seasons differ at the 99% level in a Student’s t test. Limited for the water explanation, because the authors show it is feasible in models but the water reservoirs’ size and isotopic make-up are, in their words, highly uncertain or unavailable. |
| Principal limitation | All samples come from Gale crater, near the equator. The authors write that the relative contributions of global, regional and local processes cannot be well constrained by the available data. |
| Coverage at a glance | 2 distinct credible publishers found (Phys.org and Gizmodo) as of 18 November 2026, 09:25 UTC, from SciQuest web searches for the paper title and for “Curiosity CO2 isotopes water”. Excluded: aggregator and rewrite sites. Coverage is not evidence that a finding is correct. |
Sources: Franz et al. 2026, PNAS (full text read); SciQuest coverage search at the time stated.
Is this a big deal?
It is a solid step, not a discovery of liquid water. The new part is the record itself: the authors call it the first long-term record of CO2 isotopes in Mars’ air, built over 13 Earth years by a rover that kept working. The interpretation is careful and openly provisional. Read it as “the air carries a seasonal fingerprint that points toward water in the soil,” and expect other teams to test it.
What scientists found
Isotopes are versions of an element with different masses. Most oxygen is oxygen-16; a small share is the heavier oxygen-18. Scientists report the share as δ18O, in parts per thousand (‰) relative to a standard based on Earth’s ocean water. A rise of 10‰ means the ratio of oxygen-18 to oxygen-16 went up by 1%.
Mars’ air is 95.1% carbon dioxide by volume, according to NASA’s Mars fact sheet. Curiosity’s laser spectrometer measured its δ18O at 24 to 73‰, averaging 52 ± 11‰. Values from southern spring through summer, the half of the Martian year from solar longitude 180° to 360°, were higher than in the northern half, and peaked near 250°. That is almost exactly when Mars is closest to the Sun, at 251°.
The size of the seasonal gap changed over time. Between the peaks of the two seasons it was 20 to 30‰ in Mars years 32 to 34, and 10 to 15‰ in Mars years 35 to 38. Carbon-13 in the same CO2 varied by about 10‰ a year but without the lopsided north-south pattern seen in oxygen.
How they found it
Curiosity landed in Gale crater in August 2012. Inside its body sits SAM, the Sample Analysis at Mars instrument suite, which can pull in a sip of outside air. One tool, the Tunable Laser Spectrometer, shines an infrared laser at 2.78 micrometres through the gas, bouncing it 43 times between mirrors 20 cm apart, and counts how strongly each isotope version of CO2 absorbs. A second tool, a quadrupole mass spectrometer, sorts the molecules by weight.
The team pooled 34 laser runs and 38 mass spectrometer runs from Mars years 31 to 38. A Mars year lasts 687 Earth days, about 1.9 Earth years, so the record spans eight Martian years. The authors lean on the laser data, because background gases inside SAM can skew the mass spectrometer.
They also tested a claim from Earth-based telescopes. Spectra taken in Mars year 28 had suggested CO2 isotopes might swing by 100‰ or more between day and night as gas sticks to cold ground at night and is released by day. SAM usually works at night because of resource limits, so the team scheduled two day-night pairs specially. The laser pair from Mars year 38 gave δ18O of 43 ± 6‰ by day and 34 ± 8‰ by night, which overlap within their uncertainties. The authors suggest the telescope signal, which samples the whole air column, may come from higher in the atmosphere.
Why the usual explanations fall short
When CO2 freezes onto the winter polar cap, the ice takes slightly more oxygen-18 than the air. At 142 K, a typical polar temperature, that preference is only about 5‰. The team then ran the NASA Ames Legacy Mars General Circulation Model, a global weather simulation, and deliberately set the freeze-thaw preference to 100‰, more than ten times the lab value. Even so, after winds mixed the polar air with the rest of the atmosphere, the yearly swing at Gale crater came out at about 5‰. With the realistic value, it would be well below 1‰.
The authors checked that the model moves air correctly by comparing it with Curiosity’s own argon readings, which rise and fall with the polar seasons. The model matched them in timing and size. So the shortfall is not a mixing error. Sunlight splitting CO2 into CO and oxygen adds even less: about 0.2‰ to δ18O, by the authors’ mass-balance estimate.
The proposed answer: water in the soil
CO2 and water can trade oxygen atoms, even in Antarctic ice at −35 °C, inside liquid-like layers on ice surfaces. Think of a currency exchange: CO2 hands over an oxygen atom and takes one back from water, so if the water carries more oxygen-18, the CO2 leaves heavier. The comparison has limits. The swap slows in the cold, taking tens of days near 200 K, and it settles toward a balance rather than piling up without end.
Water clings to mineral grains in the soil, and salts such as perchlorates, found at Gale crater, lower the freezing point of brines, possibly to about 185 K (−88 °C) on Mars. Earlier studies proposed that films as thin as one molecular layer, about 0.2 nanometres or two ten-millionths of a millimetre, could exist across much of the surface. As Mars nears the Sun in southern spring, warming soil could release CO2 that has exchanged oxygen with such water.
The numbers are plausible for a local version of the idea. If the CO2 in and around Gale crater exchanged with water adsorbed in the top 2 m of soil, that water would need a δ18O of about 62 to 70‰ to lift the crater’s CO2 to the recent peak of 65‰. The authors say that range is within what has already been inferred for Gale. A planet-wide version would need water across most of Mars’ soil at 133‰ or more. The authors note that the most demanding cases call for water richer in oxygen-18 than any reservoir yet observed.
Why it matters to you
This will not change anything in your daily life, and it says nothing directly about life on Mars. The paper does not discuss habitability. What it offers is a possible way to track water in present-day Martian soil through the air above it.
For researchers, the record makes it possible to follow Mars’ water and carbon cycles season by season, a timescale the authors say had not been examined this way before. Until other sites are sampled, it remains a clue about one crater rather than a map of Martian water.
What the evidence does not show
It does not show liquid water. The proposed films are “liquid-like” layers that may be as thin as one layer of molecules, and the exchange is inferred, not observed. The authors call their model “an attempt to demonstrate the feasibility” of the process.
It does not show a global pattern. Everything comes from one crater near the equator. The authors write in their significance statement: “Relative contributions from possible sources cannot be well-constrained by the available data.”
The year-to-year drop is also unexplained. The authors could not find an instrument cause for the larger swings in Mars years 32 to 34, but cannot rule one out entirely. Possible Martian causes include the rover’s climb of about 760 m in height since landing, which changes how much outside air mixes in, the global dust storm of Mars year 34, which cut ultraviolet light at the surface by more than 90%, or a one-off release of CO2 from the south polar cap.
How this fits what we already knew
Viking, Phoenix, telescopes and Martian meteorites had already shown that Mars’ air is enriched in oxygen-18 and carbon-13 compared with its rocks, which scientists read as lighter isotopes escaping to space over geologic time. What was missing was how those ratios behave over hours and seasons. The new record supplies that, and finds no sign near the surface of the large day-night swings the telescope study suggested.
Orbiters had also mapped hydrogen near the equator, which points to hydrated minerals and adsorbed water, and a modelling study by Kobayashi and colleagues estimated 2 to 4 kg of water per square metre in the top 2 m of equatorial soil. The Curiosity data give that hidden water a possible role in the air above it. You can read more about the gases in what Mars’ atmosphere is made of.
Independent perspective
SciQuest has not yet obtained comment from a scientist outside the author team, and we found none quoted in other coverage. The paper was peer reviewed, with Mark Thiemens of UC San Diego as editor. We will add independent comment if we receive it.
How other headlines framed it
- Phys.org framed it as a hint of small amounts of surface water and stressed that the exchange was not directly observed.
- Gizmodo led on an annual oddity in the atmosphere that may point to hidden water, with emailed comment from Franz. Its page title, which says Curiosity “finds signs of hidden water on Mars today”, is firmer than the paper, which proposes the water from models.
Neither outlet claimed the result makes Mars more habitable. Some aggregator headlines went further and said the rover detected water, which the paper does not report.
What happens next
Curiosity is still sampling. The authors set out a test: if the smaller recent swings come from changes in air mixing or a one-off polar release, southern-summer values should stay flat or fall; if dust temporarily slowed the exchange, they should climb back. They add that measurements by future missions at other locations on Mars would help show how far the soil-air exchange extends. For the bigger picture, see what we know about whether Mars was habitable.
In one sentence
Ten-plus years of Curiosity air samples show Mars’ CO2 gets heavier in oxygen every southern summer by more than polar ice can explain, and water clinging to the soil is the leading proposed cause, still unconfirmed.
Quick check: did it click?
- Did Curiosity find a difference between day and night CO2? No. The day and night samples overlapped within their uncertainties.
- Why can’t the polar ice caps explain the seasonal swing? In a climate model, polar freeze-thaw diluted by winds gives under 1‰ at Gale, against an observed 10 to 30‰.
- Has liquid water been detected? No. Exchange with adsorbed water or thin liquid-like films is a modelled explanation.
Read next
- What is the atmosphere of Mars made of?: the gases Curiosity is sampling.
- Hydrogen chloride over Alba Mons: another case where Mars’ air hints at something hidden below.
- How far is Mars from Earth?: closest, average and farthest distances, and why they change.
Sources and disclosure
- Franz, H. B., Webster, C. R., Flesch, G. J., Mahaffy, P. R. et al., A decadal record of isotopic trends in martian atmospheric CO2, Proceedings of the National Academy of Sciences 123 (47) e2615294123, 16 November 2026 (open access, CC BY-NC-ND 4.0).
- NASA Space Science Data Coordinated Archive, Mars Fact Sheet (orbital period 686.98 days; atmosphere 95.1% CO2).
- Kasal, K., Over a decade of atmospheric data from NASA’s Curiosity rover hints at small amounts of surface water on Mars, Phys.org, 17 November 2026.
- Lapointe, E., Mars’ Atmosphere Does Something Strange Every Year, and It May Point to Hidden Water, Gizmodo, 16 November 2026.
Disclosure: This article was drafted with AI assistance and checked against the sources above. The 1.9 Earth-year conversion, the 1% meaning of 10‰, the 0.2 nm conversion and the 760 m climb are SciQuest calculations from the cited values; the graphic is a SciQuest original drawn from values in the paper. The featured image is credited in its caption. Spotted an error? See our corrections policy.
Discover more from SciQuest
Subscribe to get the latest posts sent to your email.

Leave a comment