Hydrogen chloride has been detected several times over the same part of the Alba Mons volcano in the northern summer, when the gas usually vanishes from the Martian air. Scientists suspect it leaks from a deep magma chamber, but this is a hypothesis: no eruption, lava or chamber has been seen.
The finding is in a paper by a team led by Alexander Trokhimovskiy, published on 4 November 2026 in Science Advances. Its title calls the signals a "potential sign" of magmatic activity and describes them as "mysterious". The data come from the ACS (Atmospheric Chemistry Suite) spectrometer on Europe's ExoMars Trace Gas Orbiter, which measures what the atmosphere is made of.
Hydrogen chloride (HCl) is the first halogen gas found in the Martian atmosphere. It was discovered in 2021 and had appeared during the global dust storm of 2018. The explanation then was dust chemistry: wind lifts salty dust, which reacts with water vapour and releases chlorine, which forms HCl. Kevin Olsen of the University of Oxford says "water is crucial to this chemistry".
The gas shows up in the dust season and then disappears. Scientists saw a volcanic source as unlikely because seismic data from the InSight lander showed weak volcanic activity, and the orbiter had not detected sulphur gases, which an eruption would release.
Eight detections in 78 observations
Alba Mons does not fit this pattern. Of 78 observations in the northern summer, spread over two Martian years, eight show a clear signal in a narrow strip of longitude around the volcano. The first two detections led to a targeted search the following year, which found six more.
The peak is about 1.5 parts per billion in the lowest 10km of the atmosphere. Elsewhere there are only upper limits. Over the wider seasonal cycle, levels reach 4 to 5 parts per billion.
The signal appears in early northern summer, when hydrogen chloride is normally hard to catch. The authors say it returns every year in the same place, above cracks on the northern slopes.
Fresh cracks below the detection sites
Images from the CaSSIS, HiRISE and CTX cameras show narrow, fresh cracks below the detection sites, some only tens of centimetres wide. In places they form swarms tens of kilometres long, along ridges. The team says such structures have not been described on Mars before. The study says dust or seasonal changes would wipe out narrow cracks like these unless they were new or kept active.
The authors say gases from below ground could rise through the cracks, as they do through fumaroles on Earth. In winter, ice at temperatures down to about minus 120C seals the cracks. In early summer the ice evaporates and lets the hydrogen chloride out. This would explain why the signal appears in one place and in one season.
The flow is put at under 3kg a second, assuming the molecule lives in the atmosphere for about one Martian day. On Earth, volcanoes release from a few kg to more than 100kg a second, so Alba Mons would be one of the weaker sources. A non-volcanic source is also possible: the chlorine could be stored in subsurface ice or salts.
A flow this weak would not leave a strong heat signature, which the authors say explains why searches for volcanic activity by heat have found nothing.
The authors also looked at mud volcanoes and the breakdown of perchlorates, but these are found across the whole northern lowlands and do not explain the clustering around Alba Mons. The link between the cracks and the gas remains an interpretation, because coverage with detailed images is uneven. Hydrogen chloride alone is not a sign of life, and the study does not prove liquid water or a habitable environment.
Co-author Daniel Mège of the Space Research Centre of the Polish Academy of Sciences believes Mars "has kept enough internal heat to sustain volcanic activity until recently". Colin Wilson, the European Space Agency's TGO project scientist, said the findings "could turn out to be the biggest success of TGO so far".
The team suggests looking for similar signals at Tharsis and Cerberus. More detections from another northern summer will be described in a separate paper.