MYS0D7EXRJT3793·Mystery · signal

The Mars methane puzzle

NASA/JPL-Caltech, SAM/GSFC · NASA media (generally not copyrighted; check the credit line)

Rover background: 0.41 ± 0.16 ppbv mean, varying 0.24–0.65 ppbv (2012–2017)⁠[6]
Largest spike: 20.5 ± 4 ppbv, June 2019⁠[7]
Day versus night: 0.05 ± 0.22 ppbv by day against 0.52 ± 0.10 ppbv at night, same season⁠[7]
Orbital limit: About 20 pptv annual mean of best limits, 640 occultations over 1.44 Mars years⁠[10]
Instrument: Tunable Laser Spectrometer of the SAM suite on Curiosity, Gale crater⁠[5]
Latest: April 2025 re-analysis proposes leaks from the instrument's foreoptics chamber⁠[12]
Status: Unexplained
Questions: 0 of 5 answered

People David C. CatlingUniversity of Washington; co-author of the statistical and instrumental critiques Kevin ZahnleNASA Ames; co-author of the 2011 and 2025 critiques Sébastien ViscardyRoyal Belgian Institute for Space Aeronomy; lead author of the 2025 re-analysis Christopher R. WebsterJPL; lead of the TLS methane measurements

Catalog keyMYS0D7EXRJT3793

Methane on Mars matters because on Earth more than 90 per cent of atmospheric methane is made by living things and the rest by geochemistry⁠[1]. Sunlight destroys it slowly, so on Mars, standard photochemistry gives it a lifetime of centuries, long enough to spread evenly around the planet⁠[2]. Any methane that comes and goes, or is found in one place and not another, needs both a source and an unknown fast sink.

Reports began in 2004. The Planetary Fourier Spectrometer on Mars Express found about 10 parts per billion on average, varying over the planet⁠[3], and telescopes on Earth reported plumes of thousands of tonnes in 2003⁠[1]. These claims were questioned⁠[4]. From 2012 Curiosity's Tunable Laser Spectrometer measured methane directly in air drawn into the rover at Gale crater: a background of a few tenths of a part per billion, a seasonal cycle, higher values at night, and spikes of 7 and later 20 parts per billion⁠[5]⁠[6]⁠[7].

The ExoMars Trace Gas Orbiter, whose objectives include the search for methane⁠[8], has looked since 2018 with two spectrometers whose limits are 10 to 100 times below the earlier detections, and has found nothing, to a limit of about 20 parts per trillion, a twentieth of Curiosity's background⁠[9]⁠[10]. If both are right, methane must seep out near the surface at night and vanish within the day, by a process not known⁠[11]⁠[7]. If the rover is wrong, the detections are an artefact of the instrument: a 2025 re-analysis found methane in the instrument's foreoptics chamber thousands of times the reported atmospheric levels, enough to explain the readings by tiny leaks⁠[12]. Of the 5 questions on this site, 0 are answered.

Key questionsset by the editors

If the methane is real, where does it come from?No answer ahead yet

Is there methane in the air at Gale crater?No answer ahead yet

All 5 questions

What happened

The early detections were remote. Mars Express reported 10 ± 5 ppbv as a global average, ranging from 0 to 30 ppbv⁠[3]. Ground-based infrared spectroscopy reported methane the same year⁠[13] and, in a 2009 paper on 2003 data, plumes from discrete regions, the largest holding about 19,000 tonnes⁠[1]. Lefèvre and Forget showed that such variations are incompatible with known chemistry unless methane lasts under 200 days, 600 times shorter than predicted⁠[2].

Ground-based infrared spectra of Mars, 2003 (IRTF CSHELL and Keck NIRSPEC)NASA/Trent Schindler, 2003. Public domainArtifact

Curiosity landed in Gale crater in August 2012. Its first 20 months gave a background of 0.69 ± 0.25 ppbv and four measurements over 60 sols averaging 7.2 ± 2.1 ppbv⁠[5]. Five years of data gave a background of 0.41 ± 0.16 ppbv with a seasonal cycle⁠[6]. The Trace Gas Orbiter began science operations in 2018; its first months found no methane, to about 0.05 ppbv⁠[9]. In June 2019 Curiosity recorded its largest spike, 20.5 ± 4 ppbv, and in the same season found methane at night but not by day⁠[7]. The orbiter's limit fell to about 20 pptv by 2021⁠[10]. In April 2025 Viscardy, Catling and Zahnle published a re-analysis questioning the rover's detections⁠[12].

Timeline

Timeline

Squeezed where little happened. Each mark is a published claim, by how it stands:AcceptedMost likelyStill possibleDisputedRuled out

The accounts

The rover's record is presented by the TLS team in a series of papers that report every measurement with its enrichment protocol and confidence interval⁠[5]⁠[6]⁠[7]. The team has also answered objections, for example showing that ozone cannot be mistaken for methane at the instrument's resolution⁠[14]. The orbital record is presented by the ACS and NOMAD teams⁠[9]⁠[10]. The sceptical accounts are by Zahnle, Catling and colleagues, from the 2011 critique of the remote detections to the 2025 critique of the rover data⁠[4]⁠[15]⁠[12].

What was recorded

Curiosity's Tunable Laser Spectrometer measures methane absorption in a sample cell filled with Martian air, either directly or after enrichment, in which carbon dioxide is removed by a scrubber to concentrate methane about 25 times⁠[14]. Each measurement fits three closely spaced lines (the R3 triplet)⁠[12]. The SAM data are archived in the Planetary Data System, with raw products held but not formally certified and derived products released in numbered batches, the latest covering sols up to March 2026⁠[16].

Curiosity SAM-TLS methane measurements, Gale craterNASA/JPL-Caltech, 2012. Public domainArtifact

The Trace Gas Orbiter records solar occultations: spectra of sunlight passing through the atmosphere at the limb; its limits apply above a few kilometres from the surface⁠[10]⁠[11]. Its NOMAD calibrated data are in ESA's Planetary Science Archive; ACS data there are raw only⁠[8]. The Mars Express PFS spectra behind the 2004 detection and the 2019 coincidence with a rover spike are in the same archive⁠[17]⁠[18].

The 2025 re-analysis drew on housekeeping data and unreported experiments as well as published spectra. It found that the foreoptics chamber, through which the laser passes before the sample cell, typically held methane three to four orders of magnitude above the levels reported for the cell, with unexpected pressure changes⁠[12]. In five experiments with fuller data, the three triplet lines gave discrepant methane values, with a probability of about one in a thousand if the errors were as reported⁠[12].

The explanations

Local seepage with a fast sink is the explanation that keeps both records. Methane seeps from the ground at a low steady rate; at night, when the air near the surface is stable and the crater's winds converge, it builds up within metres of the ground where the rover samples it; by day, mixing dilutes it below the rover's limit and the orbiter, which sees only higher layers, never detects it⁠[11]⁠[7]. The day-night difference measured in 2019 supports this⁠[7]. It does not remove the problem: if seepage is widespread, the methane must still be destroyed or locked away far faster than any known process does, or the orbiter would see it accumulate⁠[7]⁠[10]⁠[11]. Transport modelling finds that air in the crater mixes in about a sol, so measurement timing matters⁠[19].

ExoMars Trace Gas Orbiter solar-occultation spectra (ACS and NOMAD)ESA, 2018. CC BY-SA 3.0 IGOArtifact

Instrumental artefact is the explanation that keeps the orbiter's record and drops the rover's. Viscardy and colleagues estimate that leaks of less than 0.1 per cent of the foreoptics methane into the sample cell would produce the reported atmospheric levels and would not show in the housekeeping data, and that the way the three triplet lines are averaged understates the uncertainty⁠[12]. An earlier statistical analysis found the data as a whole not consistent with a seasonal cycle⁠[15]. The rover team has rejected specific confusions, such as ozone⁠[14], and reported day-night differences that a constant leak would not obviously produce⁠[7].

If the methane is real, its source could be geological, such as gas from water-rock reactions or old reservoirs escaping through seeps, or biological, such as microbes below the surface⁠[20]⁠[21]. The abundances measured so far do not distinguish them⁠[20].

Published claims

Published claims and their status

Year	Who	Claimed	Status	Reason
2004	Vittorio Formisano and the Mars Express PFS team⁠[3]	Mars's atmosphere holds about 10 parts per billion of methane on average, varying from 0 to 30 over the planet.	Disputed	The Trace Gas Orbiter, 10 to 100 times more sensitive, finds none, and critics doubt the early detections were secure⁠[9]⁠[4].
2004	Vladimir Krasnopolsky and colleagues⁠[13]	Telescope spectra from Earth show methane in the Martian atmosphere, which could point to life.	Disputed	Ground-based detections must separate Martian lines from Earth's own methane, which critics say was not done securely⁠[4].
2009	Michael Mumma and colleagues⁠[1]	In northern summer 2003, plumes of methane holding up to about 19,000 tonnes rose from discrete regions of Mars.	Disputed	Never repeated, and criticised for its handling of telluric methane; the orbiter now sets limits far below it⁠[4]⁠[10].
2009	Franck Lefèvre and François Forget⁠[2]	The reported variations in Martian methane require it to be destroyed 600 times faster than known chemistry allows.	Accepted	No known sink has been found that is fast enough, which is why variable methane remains hard to explain⁠[11].
2011	Kevin Zahnle, Richard Freedman and David Catling⁠[4]	None of the telescopic or orbital methane detections made up to 2011 is secure.	Most likely	The orbiter's later non-detections fit this view, though the early detections cannot now be retested⁠[9].
2015	Christopher Webster and the Curiosity SAM team⁠[5]	Air at Gale crater carries a background of about 0.7 ppbv of methane, with an episode of about 7 ppbv over 60 sols.	Disputed	Orbital limits are far lower, and a 2025 re-analysis argues the readings may come from methane leaking inside the instrument⁠[10]⁠[12].
2018	Christopher Webster and the Curiosity SAM team⁠[6]	Background methane at Gale rises and falls with the seasons, from about 0.24 to 0.65 ppbv.	Disputed	A statistical re-analysis found the data as a whole not consistent with a seasonal cycle⁠[15].
2018	Yuk Yung and colleagues⁠[20]	Geological, geochemical and biological processes could each produce the methane Curiosity reports, and specific measurements could tell them apart.	Still possible	None of the discriminating measurements, such as carbon isotopes or companion gases, has yet been made⁠[20].
2019	Oleg Korablev and the ExoMars TGO team⁠[9]	The Trace Gas Orbiter found no methane in its first months, to a limit of about 0.05 ppbv.	Accepted	Confirmed by two instruments and tightened to 20 pptv over 1.44 Mars years⁠[10].
2019	Marco Giuranna and colleagues⁠[18]	Mars Express saw methane over Gale a day after a Curiosity spike in June 2013, from a source east of the crater.	Still possible	A single coincident detection from an instrument whose earlier readings are disputed; no later orbital detection has matched it⁠[9].
2019	John Moores and colleagues⁠[11]	Methane seeps from the ground, gathers near the surface at night and is diluted by day, which reconciles the rover and the orbiter.	Still possible	It still needs an unknown fast sink, and the rover's night readings it rests on are themselves questioned⁠[11]⁠[12].
2019	Giuseppe Etiope and Dorothy Oehler⁠[21]	Gas seepage of the kind seen on Earth could produce Mars's spikes, seasonal background and orbital non-detections.	Still possible	An argument by analogy; no seep has been located on Mars⁠[21].
2020	Edward Gillen, Paul Rimmer and David Catling⁠[15]	The Curiosity data show no strong seasonal cycle beyond what noise or random variation would produce.	Disputed	The enriched measurements alone fit a seasonal or a random pattern equally, so the dispute with the rover team stands⁠[15].
2020	Christopher Webster and the Curiosity SAM team⁠[14]	Ozone cannot be mistaken for methane in the rover's measurements at the instrument's resolution.	Accepted	Not challenged; the later critique turns on leaks in the instrument, not on ozone⁠[12].
2021	Christopher Webster and the Curiosity SAM team⁠[7]	Methane at Gale is present at night but not by day, and reached 20.5 ppbv in June 2019.	Disputed	The orbiter cannot see the surface layer at night, but the 2025 re-analysis questions the rover readings themselves⁠[12].
2021	Franck Montmessin, Oleg Korablev and the ACS team⁠[10]	Over 1.44 Mars years the atmosphere above a few kilometres holds less than 20 parts per trillion of methane.	Accepted	Consistent with the earlier ACS and NOMAD non-detections and not contested⁠[9].
2025	Sébastien Viscardy, David Catling and Kevin Zahnle⁠[12]	Methane in the rover's foreoptics chamber, thousands of times the reported atmospheric level, could explain the detections through tiny leaks.	Still possible	A 2025 critique not yet tested on the instrument; no published reply from the rover team was found in October 2026⁠[12].

What would settle it

The 2025 paper proposes a two-step experiment on Curiosity's instrument that would test whether methane leaks from the foreoptics chamber into the sample cell; its result would decide the artefact reading directly. A second, independent instrument measuring methane at the surface at night, at Gale or elsewhere, would test the seepage reading. From orbit, occultations or nadir observations reaching the lowest kilometre at night would test whether methane builds up near the ground. A measurement of the methane's carbon and hydrogen isotopes, if it is real, would point towards a geological or a biological source.