MYS0CQY1PJR4XT7·Mystery · signal

Phosphine in the clouds of Venus

ESO/M. Kornmesser/L. Calçada & NASA/JPL/Caltech · CC BY 4.0

Claimed: 14 September 2020, Nature Astronomy⁠[1]
Line: Phosphine 1–0 rotational transition at 266.94 GHz, seen in absorption⁠[1]
Telescopes: JCMT, Maunakea (June 2017); ALMA, Chile (5 March 2019)⁠[1]
Abundance claimed: About 20 ppb (2020); about 1–4 ppb planet average from reprocessed ALMA data (2021)⁠[9]
Strictest limit: Below 0.8 ppb at 75–110 km, SOFIA, November 2021⁠[8]
Latest: JCMT-Venus monitoring: four campaigns completed, a phosphine re-detection announced for September 2026⁠[13]
Status: Unexplained
Questions: 0 of 4 answered

People Jane S. GreavesAstronomer, Cardiff University; led the detection David L. ClementsImperial College London; leads reporting of JCMT-Venus Ignas A. G. SnellenLeiden astronomer; led the first ALMA re-analysis Sara SeagerMIT planetary scientist; co-author of the detection and source studies

Catalog keyMYS0CQY1PJR4XT7

Phosphine is a molecule of one phosphorus and three hydrogen atoms. On Earth it is made by anaerobic microbes and by industry; in the atmospheres of Jupiter and Saturn it is dredged up from hot, hydrogen-rich depths. Venus has neither condition: its clouds are concentrated sulphuric acid in an oxidising atmosphere, where any phosphorus should be bound to oxygen⁠[1]. In September 2020 Jane Greaves and colleagues reported an absorption line at 266.94 GHz, the frequency of phosphine's lowest rotational transition, in spectra of Venus from the James Clerk Maxwell Telescope in Hawaii and the Atacama Large Millimeter/submillimeter Array in Chile. They inferred about 20 parts per billion and found no abiotic process that could make that much⁠[1].

The claim drew re-analyses within weeks. Venus is very bright at these wavelengths, and a weak line has to be picked out from instrumental ripples in the spectrum's baseline. Snellen and colleagues showed that the polynomial used to flatten the ALMA baseline could create lines as strong as the claimed one at other frequencies, and found only a 2-sigma feature in their own reduction⁠[2]. Thompson found no significant line in the JCMT data under bootstrap tests⁠[3]. ALMA found that its standard calibration was unsuitable for so bright a target and re-processed the data⁠[4]. Others argued the line was sulphur dioxide high in the atmosphere⁠[5]⁠[6], and infrared and airborne searches set upper limits below the claim⁠[7]⁠[8].

The detection team has kept finding the line: about 5 sigma and 1–4 ppb in the reprocessed ALMA data⁠[9], consistent depths in 2017 and 2019 after correcting for sulphur dioxide⁠[10], a reading of phosphine in 1978 Pioneer Venus mass spectra by another group⁠[11], and new JCMT data from a monitoring programme begun in 2022⁠[12]⁠[13]. Every remote detection so far comes from the original team; the independent searches with other instruments have reported upper limits⁠[7]⁠[14]⁠[8]. The question has become one of method: how to separate a weak, broad line from instrumental ripple in the spectrum of a very bright planet. Of the 4 questions on this site, 0 are answered.

Key questionsset by the editors

If phosphine is present, what produces it?No answer ahead yet

Is there a real absorption line at 266.94 GHz in the Venus spectra?No answer ahead yet

All 4 questions

What happened

The JCMT observations were made with the RxA3 receiver over five mornings in June 2017, under project S16BP007⁠[1]. Seeing an absorption feature at the phosphine frequency, the team asked for Director's Discretionary Time on ALMA and observed Venus on 5 March 2019 with 44 antennas, under project 2018.A.00023.S, "Confirming Phosphine in the Atmosphere of Venus"⁠[1]⁠[15]. The paper appeared on 14 September 2020 and reported detections of up to about 15 sigma and an abundance of about 20 ppb⁠[1].

ALMA observation of Venus in Band 6, 5 March 2019ESO, 2019. CC BY 4.0Artifact

Between October 2020 and early 2021 several re-analyses appeared⁠[2]⁠[3]⁠[7]⁠[5]⁠[16]. In November 2020 the team posted a re-analysis of ALMA data processed with a revised calibration, reporting phosphine at about 5 sigma at a lower abundance⁠[9]. In July 2021 Nature Astronomy published an independent Matters Arising, the team's reply and an addendum to the original paper⁠[6]⁠[9]⁠[4]. SOFIA searched for higher-frequency phosphine lines on three flights in November 2021 and found none⁠[8]; the team disputed the result⁠[17]. The team began a long-term monitoring programme at the JCMT, JCMT-Venus, in February 2022⁠[12].

Map of the worldOpen the full map

+−

James Clerk Maxwell Telescope, Maunakea, HawaiiSubmillimetre telescope

Mapped

NASA Infrared Telescope Facility, Maunakea, HawaiiInfrared telescope

Mapped

Atacama Large Millimeter/submillimeter Array, Chajnantor plateau, ChileRadio interferometer

Within 10 km

MappedA regionNatural Earth

Timeline

Timeline

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

The accounts

The detection team's own accounts are the discovery paper, its addendum and reply, and the 2024 review of possible sources⁠[1]⁠[4]⁠[9]⁠[18]. David Clements's 2024 proceedings paper gives the team's account of the controversy and of public and scientific reactions, and states that the discovery papers did not claim evidence of life, only of phosphine⁠[12]. The critical accounts are the independent re-analyses of the same data⁠[2]⁠[3]⁠[6] and the radiative-transfer study that relocated the line to the mesosphere⁠[5].

What was recorded

The primary record is two sets of millimetre spectra: the JCMT single-dish spectra of June 2017 and the ALMA interferometric data of 5 March 2019, both public in their archives⁠[19]⁠[15]. In both, the claimed line is an absorption a few parts in ten thousand deep against the planet's continuum⁠[10]. The difficulty is the baseline: the spectra carry ripples from standing waves and calibration of comparable size, and the line's significance depends on how they are removed⁠[2]⁠[3]. ALMA adds a second difficulty, since an interferometer responds poorly to the broad, smooth emission of a planet's disk⁠[16].

JCMT spectra of Venus at 266.94 GHz, June 2017ESO, 2017. CC BY 4.0Artifact

Other instruments looked at other altitudes and lines. TEXES on the NASA Infrared Telescope Facility, in March 2015 archival data near 950 cm⁻¹, found no phosphine to a 3-sigma limit of 5 ppbv near the cloud tops⁠[7]. Venus Express SOIR occultation spectra gave upper limits above the clouds⁠[14]⁠[20]. SOFIA's GREAT receiver set a limit of 0.8 ppb from 75 to 110 km⁠[8]. From below, the 1978 Pioneer Venus Large Probe mass spectrometer recorded signals at 51.3 km that Mogul and colleagues assign in part to phosphine⁠[11].

The JCMT-Venus programme has collected far more data than the discovery set, with the Ū'ū receiver covering phosphine, water (through HDO) and sulphur dioxide together⁠[12]. Four multi-week campaigns had been completed by mid-2026 and a fifth was approved for July 2026; the team reports variability of sulphur dioxide and water over years and with local solar time, and a re-detection of phosphine⁠[13].

The explanations

The first question is whether the line is real. Critics showed that high-order polynomial baselines can manufacture lines of the claimed strength, that the ALMA data at first carried a calibration problem, and that with other reductions the feature falls to about 2 sigma in ALMA and to insignificance in the JCMT data⁠[2]⁠[3]⁠[4]. The detection team re-reduced both data sets with revised methods and reports the line at about 5 sigma in ALMA and at 6 to 8 sigma in both after removing sulphur dioxide contamination⁠[9]⁠[10].

If the line is real, it could be sulphur dioxide, which also absorbs near 266.94 GHz. Lincowski and colleagues found that a typical mesospheric sulphur dioxide profile fits the JCMT feature, and that the line forms above 80 km, where phosphine would be destroyed too quickly to persist⁠[5]. The detection team estimated the sulphur dioxide contribution at about 10 per cent in the JCMT spectrum and 2 per cent or less in ALMA⁠[10].

If the line is phosphine, the abundance is in question. Estimates have fallen from about 20 ppb to a planet average of 1 to 4 ppb, with local peaks of 5 to 10 ppb⁠[1]⁠[9]; SOFIA's limit of 0.8 ppb applies at 75 to 110 km⁠[8], and the team has proposed that phosphine varies with time and altitude, destroyed by sunlight in the upper atmosphere and more abundant deeper in the clouds⁠[9]⁠[12]. Its source would then be unknown chemistry, or, as the team says is an extreme possibility, an aerial microbial biosphere; no known abiotic route produces enough⁠[21]⁠[18].

Published claims

Published claims and their status

Year	Who	Claimed	Status	Reason
2020	Jane Greaves and colleagues⁠[1]	JCMT and ALMA spectra show phosphine in Venus's clouds at about 20 ppb, a gas with no known abiotic source there, which may point to life.	Disputed	Independent re-analyses found the line could be a processing artefact or sulphur dioxide, and no independent group has reproduced it⁠[2]⁠[6]
2020	Ignas Snellen and colleagues⁠[2]	The 12th-order polynomial used to flatten the ALMA spectrum creates spurious lines, and an independent analysis finds the feature at only about 2 sigma.	Disputed	The detection team recovered the line in reprocessed data with other methods, so the dispute is not settled⁠[9]
2020	Mark Thompson⁠[3]	Re-analysed with other baselines and a bootstrap, the JCMT spectrum shows no significant line, and polynomial fitting produces false positives.	Disputed	The team's re-test of the JCMT line found it at 4.8 sigma against the baseline uncertainty, about 3 sigma on a stricter definition⁠[4]
2020	Thérèse Encrenaz and colleagues⁠[7]	Infrared spectra from 2015 show no phosphine, with an upper limit of 5 ppb, four times below the millimetre value.	Most likely	Other instruments also set limits below the claim, but the detection team argues they sample other altitudes and times⁠[8]⁠[9]
2021	L. Trompet, Ann Carine Vandaele and colleagues⁠[14]	Venus Express solar-occultation spectra show no phosphine above the cloud tops and set upper limits there.	Most likely	The limits apply above the clouds, where the detection team expects phosphine to be destroyed, so they do not test the cloud-level claim directly⁠[9]
2021	Jane Greaves and colleagues⁠[9]	Reprocessed ALMA data show phosphine at about 5 sigma, at a planet average of 1 to 4 ppb, and the JCMT line cannot be sulphur dioxide.	Disputed	Independent groups reading the same data attribute the feature to calibration or sulphur dioxide⁠[6]⁠[5]
2021	Alex Akins, Andrew Lincowski, Victoria Meadows and Paul Steffes⁠[16]	ALMA's response to a bright, extended planetary disk makes a weak, broad absorption line hard to detect reliably.	Disputed	ALMA issued a revised calibration for the Venus data, and the detection team still recovers the line from it⁠[4]⁠[9]
2021	Andrew Lincowski, Victoria Meadows and colleagues⁠[5]	The 266.94 GHz line forms above 80 km, and ordinary mesospheric sulphur dioxide fits the JCMT feature.	Disputed	The detection team used a simultaneous sulphur dioxide line to estimate contamination at about 10 per cent for JCMT and 2 per cent or less for ALMA⁠[10]
2021	Geronimo Villanueva and 26 co-authors⁠[6]	Independent analyses of the ALMA and JCMT data find no phosphine; the ALMA feature reflects calibration and the JCMT feature could be sulphur dioxide.	Disputed	The detection team's reply and later monitoring report the line again, but no independent group has⁠[9]⁠[12]
2021	Rakesh Mogul and colleagues⁠[11]	Mass-spectrometer data from the 1978 Pioneer Venus probe at 51 km contain phosphine and other reduced gases.	Still possible	The assignment rests on re-reading 1978 data, and no later measurement inside the clouds has tested it⁠[11]⁠[18]
2021	William Bains, Sara Seager and colleagues⁠[21]	No known abiotic process, in the atmosphere, at the surface or below, makes phosphine on Venus in the amounts reported.	Still possible	It matters only if the detection is confirmed, and a later review by overlapping authors leaves open that Venus chemistry is incomplete⁠[18]
2022	Jane Greaves and colleagues⁠[10]	Sulphur dioxide contaminates the phosphine line by about 10 per cent at JCMT and 2 per cent or less at ALMA, and the corrected spectra agree.	Disputed	Critics hold that the feature itself is not secure, which the correction does not address⁠[6]⁠[2]
2022	Martin Cordiner, Geronimo Villanueva and colleagues⁠[8]	Three SOFIA flights found no phosphine, with an upper limit of 0.8 ppb between 75 and 110 km.	Most likely	The detection team disputes the analysis and argues for variation with altitude and time, but has not had it confirmed⁠[17]
2023	Jane Greaves and colleagues⁠[17]	The SOFIA data, analysed differently, do not exclude phosphine and can be read as recovering it.	Disputed	No independent analysis of the SOFIA data supports the recovery, and the published limit stands⁠[8]
2024	William Bains, Sara Seager and colleagues⁠[18]	No known geochemical or photochemical route makes phosphine near the observed level, so either Venus chemistry is incomplete or an aerial microbial biosphere makes it.	Still possible	It depends on the detection, which is unconfirmed, and the biological option is untested⁠[6]
2024	David Clements (JCMT-Venus)⁠[12]	Multi-year JCMT monitoring recovers phosphine, at a preliminary 0.3 ppm near 55 km from an experimental method.	Still possible	The method is experimental and the result preliminary; no independent group has reproduced it⁠[13]
2026	Wei Tang, David Clements, Jane Greaves and colleagues⁠[13]	Four JCMT-Venus campaigns give a re-detection of phosphine, alongside variable sulphur dioxide and water.	Still possible	Reported in a conference abstract whose robustness was still to be discussed; no paper or independent confirmation yet⁠[13]

What would settle it

The decisive test is an independent detection: a group outside the original team recovering the line, or phosphine's other transitions, from new data with a reduction method fixed in advance and tested on injected fake lines. Publication of the JCMT-Venus spectra with their reduction code would allow that test on the largest data set. A spectrometer inside the atmosphere would settle it directly: phosphine at the parts-per-billion level in the clouds would be measured, or excluded, by a mass spectrometer or tunable-laser instrument on a descent probe. If phosphine is confirmed, the next question, whether its source is chemical or biological, would need the same in-situ measurements of the clouds' composition.