Perseverance Shows Jezero Carbonates Formed in Groundwater

NASA’s Perseverance rover has traced three separate water pulses through igneous rock that orbiters had tagged as a Jezero Crater lakeshore. The Purdue-led SuperCam study, published on 21 September 2026 in Communications Earth & Environment, shows the famous carbonate band formed first in carbon dioxide-rich groundwater, not as lake sediment.

That finding cuts into the orbital story that helped put the rover on this rim. The same carbonate fingerprint stretches across the Nili Fossae country around Jezero, so the misread is bigger than one outcrop.

The Lakeshore NASA Mapped From Orbit

Jezero is a 45-kilometre crater on the north-west side of the 1,200-kilometre Isidis impact basin, in a region already known for olivine and carbonate. Orbiters picked up a narrow inner-rim band with an especially strong carbonate signal, informally named the Margin Unit, and the Mars 2020 plan treated it as a high-value shoreline to sample.

Carbonates on Earth often form in shallow lakes and seas. From space, a bright carbonate strip along an old water line looked like the edge of Jezero’s vanished lake. When the rover reached that inner edge in September 2023, the team expected layered sand and silt. It found coarse crystalline rock that had cooled from magma, later chewed by water.

Before we arrived at the Margin Unit, the main hypothesis was that the carbonate seen from orbit formed from interaction with the lake that existed in Jezero Crater. But now we know that this location became a sort of crossroads for aqueous systems.

Candice Bedford, research scientist at Purdue University, NASA release

Bedford, who led the study, added that Jezero sits inside a vast carbonate exposure, so the lesson travels. Roger Wiens, also at Purdue, leads the SuperCam team. Briony Horgan, a Purdue planetary scientist, was campaign science lead while the rover worked the unit.

Igneous Crystals Kept a Water Ledger

Perseverance climbed about 870 feet (265 metres) of elevation across the Margin Unit. High on the slope, SuperCam saw coarse olivine-rich rock with almost no sign that water had touched it. Those grains grew large because the magma cooled slowly underground, then erosion stripped the cover and left the body in the sun.

The paper draws a hard line at elevations above -2350 metres, where texture and chemistry still read as that slow-cooled olivine rock. Below the hypothesized second terrace of the old Jezero lake, the same body is a different object. Olivine grains are fractured. Silica sits in the gaps. Carbonate fills cracks that now stand as ridges after softer rock wore away.

Bedford compared the fractures to pipes and the carbonate to limescale that blocked them. The crystals still remember the magma. The fills remember the fluids that came later, which is why an igneous protolith turned out to be a useful ledger rather than a dead end.

East Margin slabs, seen on sols 911 to 1000 at about -2423 to -2411 metres, show centimetre-scale layering and possible cross-beds, and the paper notes physical reworking near Neretva Vallis and the western fan. The chemistry still points to a crystalline olivine-rich starting rock, later moved and cemented. West Margin work ran sols 1021 to 1161. North Margin ran sols 1192 to 1211. The High Margin, sols 1245 to 1280, reaches -2338 to -2159 metres and stays closer to the dry, coarse original.

Three Fluids Left Three Mineral Records

SuperCam cannot date the pulses. It can put them in order. Bedford’s team reports three main alteration episodes in the Margin unit, each with a different chemistry and a different mineral dump.

THE THREE WATER PULSES

Eleni Ravanis, a planetary scientist at the University of Hawaii at Manoa and a co-author, said some Margin Unit rocks also contain silica, that turning olivine into carbonate can leave silica behind, and that SuperCam sees more of that silica in rocks that sat below the water line. The lake is still in the story. It is not the author of the carbonate orbiters used as a shoreline stamp.

The last pulse is the odd one out. Fluorite usually forms when hot water runs through volcanic rock, so the thick eastern vein is a hydrothermal mark, not a beach. On Earth, that kind of system can feed microbes. On Mars it is still only a geological clue.

Why Jezero’s Carbonate Signal Misled Orbiters

The orbital lakeshore case rested on a deep 2.5-micrometre carbonate absorption along the inner rim, tied to the highest estimated stand of the paleolake. The rover data break that simple link. The first carbonate formed in the crust, in groundwater that reacted with olivine. Later fluids moved some of that carbonate and dropped silica. A later hot pulse cut new fractures.

The paper also notes that parts of the regional olivine-bearing unit outside Jezero show similar carbonate spectra, which already pointed toward groundwater or hydrothermal solutions rather than a unique beach. Jezero is the place where a rover could test that idea on the ground. The test says the bright rim is a crossroads of fluids inside an igneous body, not a postcard of one lake.

That is the part that travels. Any CRISM carbonate patch sitting on olivine-rich rock now has to answer a question Jezero just made ugly: is this a shoreline, or is it crustal water written into magma?

SUPERCAM AT THE MARGIN

• Targets: More than 185 bedrock points across the unit, plus alteration features the mast could reach.
• Methods: Laser-induced breakdown spectroscopy for chemistry, with Raman and infrared reflectance on the same head.
• Reach: Quantitative chemistry on targets between 2 and 6.5 metres from the mast.
• Cutoff: Above -2350 metres the olivine rock still looks largely dry; the water record concentrates below the old terrace.

When the team wants a chemistry reading, it can command laser chemistry from 6.5 metres, then compare plasma, Raman and infrared on the same rock. SuperCam is co-led by Purdue, Los Alamos National Laboratory, and IRAP and CNES in Toulouse. A panorama built from its data between 8 and 16 October 2023 is the public face of the campaign.

Perseverance Already Cached Three Margin Cores

The rover did not only look. It drilled. NASA’s sample catalog lists three Margin Unit cores still stored on Perseverance, taken while the team still described the eastern slabs in shoreline language. Those tubes are now the physical copy of Bedford’s sequence, waiting for a lab that can date the minerals and hunt for organic traces SuperCam cannot settle.

THE MARGIN UNIT CORES

• Pelican Point: Sealed 25 September 2023 on sol 923 at the Hans Amundsen Memorial Workspace, a 6.10-centimetre rock core.
• Lefroy Bay: Sealed 21 October 2023 on sol 949 at Turquoise Bay, a 4.70-centimetre core from a more carbonate-rich patch in the orbital maps.
• Comet Geyser: Sealed 12 March 2024 on sol 1088 at Bunsen Peak, a 5.78-centimetre core logged as silica-cemented carbonate.

Pelican Point, Lefroy Bay and Comet Geyser remain on the rover. NASA has sealed 30 of the 38 sample tubes it sent, and the High Margin abrasion at Emerita Mesa did not add a fourth core. Until a later mission picks the cache up, the three-pulse story stays a remote-sensing argument plus three sealed sticks of rock.

The catalog still labels Pelican Point and Lefroy Bay as sedimentary, a hangover from the shoreline plan. Comet Geyser is already filed as silica-cemented carbonate, which fits the second pulse better than a simple beach sand.

Fluorite Marks the Last, Hottest Pulse

The eastern vein is the detail no orbital map could have sold as a lakeshore. Calcium sulfate is common enough in Jezero. Fluorite is not. It is the mineral that says the last water was hot, late, and moving through younger cracks after the carbonate ridges and the silica pores were already in place.

WHAT WE KNOW

• The order: Groundwater carbonate, then silica-rich fluids, then a hot sulfate-fluorite vein.
• The host: Slow-cooled olivine-rich igneous rock, later exposed and locally reworked.
• The lake: Still a plausible actor in the second pulse, especially below the old water line.

WHAT IS UNCONFIRMED

• The ages: SuperCam has the sequence, not the clocks.
• The second pulse: Direct lake water, a change in groundwater, or both.

Bedford, after 10 years on Mars rover work, said Mars constantly throws surprises and that orbital data rarely match what a rover finds. She wants the study to reshape how the history of water is read in Jezero and across the planet, and to give context when the cached cores reach Earth. The cores are already in the belly of the rover. The dates are not.

Frequently Asked Questions

Did Perseverance Find Signs of Life in the Margin Unit?

No. The study does not claim microbes lived there. On Earth, water reacting with olivine can free hydrogen that some microbes use as food, and carbonate and silica can trap traces of that chemistry, which is why the site stays on the habitability list without a biosignature in hand.

How Does SuperCam Identify Rock Chemistry From a Distance?

The mast head fires a laser that turns a pinprick of rock into plasma, then reads the light to get elemental chemistry, while Raman and infrared reflectance on the same instrument pick up mineral structure, all on targets as close as 2 metres.

Which Margin Unit Samples Are on the Rover?

Three sealed cores, Pelican Point, Lefroy Bay and Comet Geyser, and NASA’s public catalog still calls the first two sedimentary while listing Comet Geyser as silica-cemented carbonate, a mismatch that now sits beside Bedford’s igneous-plus-fluids reading.

Did the Ancient Jezero Lake Create the Carbonates Orbiters Mapped?

The first carbonate in this unit formed from CO2-rich groundwater inside olivine rock, and similar carbonate spectra on the regional olivine unit outside the crater already favoured crustal fluids, so the orbital shoreline stamp cannot be read as lake precipitate by default.

Can the Three Water Events Be Dated?

Not from SuperCam. Order is fixed by which mineral cuts which, and absolute ages need laboratory clocks on the returned cores, which is why the three tubes are now the part of this story that still has to leave Mars.