OrbitalHub

On August 4, 2026, with a single coordinated command sequence, engineers at NASA’s Jet Propulsion Laboratory switched three power-hungry heaters on Voyager 2 for three lower-power substitutes and turned off a redundant device whose residual heat was already doing the job. The maneuver was nicknamed the “Big Bang” by the mission team because it had to be executed in one uninterrupted block: the probe cannot tolerate a half-finished sequence when the new configuration leaves critical fuel lines running on borrowed thermal margins. Forty-eight years after launch, with plutonium steadily depleting inside its three radioisotope thermoelectric generators, Voyager 2 was running on roughly four watts less available power than it had the year before, and the mission was weeks from a forced instrument shutdown. The swap changed that math. The savings should keep all three of Voyager 2’s remaining science instruments returning data from interstellar space for at least another year before the team has to choose which one to power down next.

Voyager 2 crossed the heliopause in November 2018 and is currently the only spacecraft in the solar system making direct in-situ measurements of the plasma, magnetic field, and energetic particle environment beyond the Sun’s magnetic bubble. Voyager 1 crossed the boundary six years earlier, but Voyager 2 is still returning plasma data that Voyager 1 cannot, because Voyager 1’s plasma science instrument failed in 1980 and was never replaced. The two spacecraft together form humanity’s only working sensor network in the local interstellar medium, and every instrument shutdown is permanent. Once powered off, Voyager’s 1970s electronics cannot be brought back. With Voyager 1 already down to two operating instruments after an unexpected February 2025 power drop forced the LECP shutdown in April of that year, losing a third instrument on Voyager 2 would have left the interstellar mission with a thinner, more asymmetric scientific footprint. The Big Bang also matters as an engineering template. Voyager’s power budget is now so thin that similar swaps will be needed every year or two until the RTGs decay below the threshold needed to run any science instruments at all, which NASA estimates will happen sometime in the 2030s. The procedure the JPL team developed for this swap is the same procedure they will repeat on Voyager 1, and likely on both spacecraft multiple more times before the mission ends.

The Voyager 2 swap was first reported by NASA’s Voyager mission blog on August 4, 2026. The Register’s follow-up the next day added the most concrete engineering detail: a NASA spokesperson told the paper that the swap involved “two heaters and a device that was being kept on since its residual heat was in the right location and it used less power than the actual nearby heater.” The trio was replaced with another three components that together still generated enough heat to keep the fuel lines warm, but at lower wattage. Past swaps on the Voyager spacecraft were usually one-for-one. The Big Bang’s innovation was doing three replacements at once and accepting the temporary thermal risk to the fuel lines in exchange for the combined power savings. Voyager 2 carries a single 70-centimeter titanium tank holding about 100 kilograms of hydrazine at launch, used for attitude control and trajectory correction maneuvers; keeping that fuel from freezing in interstellar space is one of the load-bearing thermal jobs on the spacecraft, and it is the constraint that bounds how aggressively JPL can trade heat-generating components for power savings. Voyager 2’s LECP instrument had already been turned off in March 2025; the plasma science instrument was shut down in September 2024; a third instrument on each Voyager has been off since long before that, leaving three on Voyager 2 and two on Voyager 1. Without the Big Bang, NASA confirmed, another Voyager 2 instrument would have had to be switched off before the end of 2026. With the savings from the swap, that decision has been pushed into 2027 at the earliest. JPL plans to run the same swap on Voyager 1 in the coming months. Voyager 1 is currently farther from Earth than Voyager 2, and both probes are still inside the solar system in the strict astronomical sense, even though both are now in the interstellar medium.

The Voyager power budget is governed by three multihundred-watt radioisotope thermoelectric generators, each containing 24 pressed plutonium-238 oxide spheres. At launch in August 1977, the three RTGs together produced about 470 watts of electrical power. The plutonium-238 half-life is 87.7 years, so the electrical output has dropped by roughly a factor of two over the mission’s lifetime. The probes lose about 4 watts of available power per year, and the drop is accelerating slowly as the thermocouple junctions age and the thermal gradient they exploit shrinks. NASA first began drawing on a backup power reservoir in April 2023; without that step, the mission would have lost at least one more instrument sooner. The thermal architecture is what makes the Big Bang clever. Many components on Voyager 2 give off waste heat as a side effect of operating, and several of those components are physically adjacent to fuel lines, propulsion latches, or science instruments that need to stay above their minimum operating temperatures. Replacing a high-power heater with a lower-power one usually means the heat budget has to be rearranged, but if the old heater was already heating something the new heater does not need to heat, the swap can be near-power-neutral. The Big Bang worked because the three devices that were turned off had already been “ghost heating” their neighbors, with the residual heat in the right place. The three replacements are less powerful individually, but together they keep the fuel lines and adjacent subsystems at the required temperature with a net lower draw from the RTGs. The constraint is sequencing. If the swap had been interrupted partway through, the spacecraft would have been left with an underheated fuel line and an instrument bay that was relying on a component that was about to be turned off. JPL engineers therefore treated the Big Bang as a single atomic operation, sent the entire command sequence as one batch, and then waited more than 23 hours for the signal to reach Voyager 2 and another 23 hours for the engineering telemetry to return confirming the new configuration was stable. The same near-24-hour round trip will apply when the team runs the swap on Voyager 1.

Voyager 2 is now expected to keep all three of its remaining science instruments (the magnetometer, the plasma wave subsystem, and the cosmic ray subsystem) returning interstellar data through at least late 2027, with the team planning to run the same Big Bang swap on Voyager 1 within the next several months. The next decision point is whether the LECP instrument that was shut off on Voyager 1 in April 2025 can be brought back online after that swap, which would give Voyager 1 a third working instrument for the first time in roughly a year. Beyond that, the Voyager power budget will continue to tighten by about four watts per year, and JPL will likely need to run a similar coordinated swap every twelve to twenty-four months until the RTGs decay below the threshold needed for any science operation sometime in the 2030s. The Big Bang is not a permanent fix. It is a way of buying one more year of an interstellar science mission that is now forty-eight years past its design life, using only the hardware that is already aboard the spacecraft.

Source: Wikimedia Commons file page.

Two days before the second Tuesday in October, the same room at the U.S. Institute of Peace that has hosted peace talks for three decades filled instead with research administrators, lab directors, and the kind of people who can tell you the difference between a Bayesian prior and a proposal deadline. The occasion was the White House Office of Science and Technology Policy’s “Science: A New Golden Age Summit,” convened on October 8, 2026 to translate a July report to the President into a coordinated set of federal science bets. NASA showed up with two of them, and neither one was a new telescope.

The first was a quiet institutional commitment: a coordinated metascience program inside NASA’s Science Mission Directorate, with a permanent expansion of dual-anonymous peer review across the directorate’s research solicitations. The second was a fellowship: ten Genesis Mission Fellows, accelerated doctoral students who will be trained simultaneously in super intelligence and in a traditional NASA mission discipline, and placed inside NASA laboratories to work on real flight programs. Both came wrapped in Isaacman’s framing of a “new Golden Age” but, more usefully, both came with concrete structure: a named review mechanism that already exists, and a fellowship template lifted almost verbatim from a graduate research program that has been running since 2019.

Why does the peer-review expansion matter? The short version is that the agency has been running the experiment for six years and is now extending it. Dual-anonymous peer review, the form NASA Science began piloting in 2020 and folded into the Future Investigators in NASA Earth and Space Science and Technology (FINESST) solicitation in 2024, breaks the link between who writes a proposal and who grades it. Proposers rewrite their science cases to remove identifying details — author names, institution, even the kind of instrument references that flag a specific lab. Reviewers see only the science and the budget. The pilot was launched during the first Trump administration and is, in the agency’s own framing on its Dual Anonymous Peer Review page, a structural response to the cognitive biases Tversky and Kahneman laid out in 1974. The new initiative commits the Science Mission Directorate to expanding that practice further, into more of its Research Opportunities in Space and Earth Sciences (ROSES) programs, and to folding the operational lessons into a coordinated metascience capability that can study its own review outcomes in close to real time.

That is a quieter change than a new flagship, but it lands harder on day-to-day science. ROSES is the pipeline through which the bulk of NASA’s external science gets funded: a single solicitation umbrella that issues dozens of individual program elements each year, ranging from astrophysics theory to Earth-system modeling to heliophysics data analysis. If DAPR is the new default for most ROSES elements, then the median funded investigator — the one whose lab, whose postdocs, and whose instrument pipeline are not famous — is suddenly competing on a more level surface than they were a year ago. The biases that DAPR targets are not exotic. They are the usual ones: prestige-by-institution, halo effects from earlier work, name recognition that follows investigators across subfields. None of these are accusations against individual reviewers. They are properties of a system that puts people in front of proposals and asks them to choose.

The fellowship sits at a different altitude. Genesis Mission Fellows will be doctoral students whose training explicitly couples super intelligence — the federal government’s preferred term for advanced AI systems that can plan and reason across long horizons — with a traditional discipline, the way the original FINESST design already pairs a graduate student with a faculty mentor on a single research question. NASA will support ten such students in the first cohort, working on NASA mission areas at NASA laboratories. The fellowship inherits the bones of FINESST, including its structure as a research grant rather than a traditional fellowship and its $50,000-per-year award size, and bolts on the SI training component. The Department of Energy launched a parallel fellowship structure the same day, aimed at the same kind of dual-competency PhD, with the same general design and a 100-PhD-per-cohort scale, which makes the two programs the first coordinated federal attempt to grow a domestic SI-fluent research workforce rather than rent one from outside.

What makes this a coherent story is the compute underneath it. The White House fact sheet accompanying the announcements lists $2.4 billion in industry commitments to the Genesis Mission Consortium for SI compute and tooling: NVIDIA at $1 billion, AMD at $500 million, OpenAI at $200 million, Anthropic and Google at $150 million each, AMP and Emerald AI at $100 million each, and AWS, Armada, Crusoe, and Micron at $50 million each. That pool is the resource the new fellowships will be trained on, and the resource NASA’s metascience effort will be able to use to study its own review process. It is also the lever that the July expansion of the Genesis Mission pulled: the $5 billion federal commitment that pulled fifteen-plus agencies into the consortium, of which NASA’s two October 8 announcements are the first research-training deliverables.

The engineering question that makes or breaks this is whether the training tracks actually produce researchers who can do original work in their discipline, not researchers who can prompt an SI system about it. The federal design here leans on the rotation model: a doctoral student spends time inside a national lab, time inside an industry partner, and time inside the academic department that granted the degree. The expectation is that the rotation through an industry lab forces the AI work to meet engineering constraints — latency, cost, auditability, integration with real instruments — while the academic time forces it to meet peer-reviewed publication standards. The model is closer to the National Science Foundation’s Graduate Research Fellowships in structure than to anything AI-specific, and that is the point. The novel content is the SI training, not the fellowship scaffolding.

There are real risks. DAPR shifts work onto proposers, who have to rewrite their cases to strip identity, and onto panels, who have to learn to read proposals that read like abstracts. If the rewrite costs more time than the bias reduction saves, the program fails on its own terms. The fellowship, meanwhile, depends on industry partners actually opening their labs to doctoral students, and on academic departments accepting an SI training component that is not yet a recognized subfield. Both are bets about institutional behavior, not just about money.

What changes because of this is the federal science workforce itself. A decade from now, the average NASA principal investigator will be someone who competed for funding under DAPR and probably was reviewed by an anonymous panel rather than a familiar program officer. A cohort of researchers who finished their PhDs with hands-on super-intelligence experience inside a national lab will be the ones writing the next generation of mission proposals. Whether that produces better science depends on details that are still being written into the program’s implementation, but the direction is set. NASA has spent sixty-seven years building tools to look outward. Yesterday it turned two of those tools — peer review and graduate training — on its own house.

The figure every planetary scientist carries in their head is the Moon-forming impact: a Mars-sized body called Theia slamming into the infant Earth roughly 100 million years after the Sun was born, vaporising rock, blasting a disk of debris into orbit, and leaving behind a moon that still shows the chemical scar of that single catastrophic day. The trouble is that this picture has lived mostly inside models and inside Apollo-era geochemistry, with no direct way to watch an equivalent event happen elsewhere. NASA’s James Webb Space Telescope has now produced the largest mid-infrared sample yet of extreme debris disks, the rare young stellar systems whose warm dust tells astronomers that something inside them has just been smashed to pieces. The team, led by Kate Su at the Space Science Institute in Boulder, Colorado, reports that the sample splits cleanly into two populations whose separation lines up, almost uncannily, with the timing of the impact that built our Moon.

These collisions matter because they are how rocky planets finish being built. A young star starts life wrapped in a gas-rich protoplanetary disk where dust accretes into planetary seeds, then spends the next few hundred million years refining those seeds into fully grown worlds. By the time the gas is gone, what is left is a colder, thinner debris disk populated by asteroid and comet-like bodies. Most of those debris disks are quiet, with low infrared brightness and dust grains sitting at their original sizes, and a planetary scientist would describe such a system as a settled, post-construction environment. A small fraction are not quiet, and these are the extreme debris disks. They have unusually large amounts of warm dust close to the star, in the region where rocky planets orbit our Sun, and the only way to make that much warm dust is to start with recently shattered, vaporised, or melted rock. Estimates from Spitzer and early Webb data put the fraction at roughly one percent of young stars showing observable signs of this phase, a low number than what theorists expect from models of late-stage accretion, and possibly including our own solar system during the brief epoch when Earth was finishing its assembly.

For Su’s team the breakthrough was less about finding the disks and more about characterising them. Webb’s Mid-Infrared Instrument, with its 6.5-metre primary mirror sitting in the cold L2 environment, can read the chemical fingerprints of silicate minerals in the dust at wavelengths no previous instrument had enough sensitivity to pursue. Some of the disks are silica-rich, with spectra full of features from glassy, high-temperature minerals like the volcanic glass obsidian on Earth. Some are silica-poor, dominated instead by forsterite, the same mineral that gives some Hawaiian beaches their green tint. According to Su’s reading of the mineral record, a silica-rich disk is the sign of a high-energy impact between Mars-sized bodies, where collisions vaporise enough rock to leave behind freshly quenched silica glass. A silica-poor disk comes from gentler, lower-energy collisions between Moon-sized bodies, where the debris is just shattered rock rather than vapor. The geometry of the collision, not just its existence, shows up in the infrared spectrum.

The ages of the disks fall in line with their chemistry. Of the 21 disks the team compiled (five drawn from retired Spitzer archives and 16 from new Webb observations), about one third are silica-rich, and they appear only around stars younger than 300 million years. The remaining two thirds are silica-poor, persist across a much wider age range, and show larger swings in infrared brightness. The age boundary is doing real work. Silica-rich impacts, the team argues, are a young-system phenomenon, products of the chaotic last stage of terrestrial planet assembly. Once a system has settled into a stable architecture, the big bodies stop crashing into each other and only the smaller, lower-energy rubble collisions continue. This is what makes the boundary line up with the Moon-forming epoch. Geochemists place the giant impact at roughly 100 million years after the Sun formed, deep inside the silica-rich range. If the Sun went through its own extreme debris disk phase, the chemistry of its dust would have to be silica-rich.

The solar-system timeline beyond the Moon-forming impact also has a place in the picture. Su’s team points to the Late Heavy Bombardment as a possible silica-poor phase for the Sun, a window roughly 600 to 800 million years after formation when the giant planets migrated far enough to destabilise smaller orbits and trigger fresh cratering across the inner system. The age fits the silica-poor range, where the disks are still appearing a billion years into stellar life, and the mechanism fits the irregular brightness swings the team measures, since orbital instability produces bursts of dust rather than steady amounts. Both the Moon-forming phase and the Late Heavy Bombardment phase are inferred events that happened here, and both have plausible counterparts in the catalogue Su has now built. The chart accompanying the paper makes the alignment visible: protoplanetary disks fill the first ten million years, silica-rich disks take over between roughly ten and one hundred million years, and silica-poor disks continue to show up across the full timeline to several billion years.

There is an honest gap in the picture. Of the 21 disks, only three sit in the silica-poor, old-system range that the team needs to confirm the Late Heavy Bombardment analogy. As coauthor Attila Moor of Konkoly Observatory put it, three disks are not enough to test the hypothesis, and Webb time over the next several cycles will determine whether older silica-poor systems are simply rare or whether the picture needs to be redrawn. The current catalogue is enough to show that Mars-sized collisions leave a recognisable chemical fingerprint in the dust, that the fingerprint disappears after a few hundred million years, and that the disappearance happens for a reason that the Solar System’s own record independently supports.

For planetary science the lesson is operational rather than philosophical. When Su’s team says they have used Webb to bring together the big picture of how rocky planets formed and how giant planets evolved, what they mean is that infrared spectra can now resolve the difference between two collision regimes that previously looked the same in dust-mass measurements. That kind of resolution opens a path to testing formation theories on a per-system basis rather than inferring them only from our own neighbourhood. Earth is no longer the only place we can read the chemistry of giant impacts. There are now twenty-one other laboratories in the catalogue, with more on the way, and the odds that one of them catches a collision in progress just got measurably better.

Earth’s magnetosphere is, by planetary standards, a tidy thing. The solar wind blows outward from the Sun, the planet’s magnetic field pushes back, and the funnel-shaped openings near the north and south poles, called cusps, sit almost directly under the Sun at local noon. That is the place where the incoming stream of charged particles most cleanly slips through the magnetic shield and into the atmosphere. Saturn does nothing of the kind. A team led by Dr. Yan Xu of the Southern University of Science and Technology has now mapped the kronian cusp using six years of archival Cassini magnetometer and plasma data, and found that the opening is dragged well into the afternoon. On average it sits between 13:00 and 15:00 local Saturn time, and in some events it extends toward 20:00, three to eight hours from where Earth’s equivalent sits. The culprit, the team argues, is not the solar wind at all. It is Saturn’s 10.7-hour rotation and the heavy load of ionized water vapor that the moon Enceladus pumps into the magnetosphere from its south-polar plumes.

The paper, published on 1 April 2026 in Nature Communications by Xu, Zhonghua Yao, Andrew Coates, Licia Ray, Sarah Badman, Chris Arridge and nine co-authors, is more than a curiosity about a single planet. The result is direct observational evidence for a long-standing theoretical claim: that on rapidly spinning gas giants with active moons, the planet’s own rotation can become the dominant force shaping its magnetosphere, more important than the Sun. If that is true at Saturn, it almost certainly extends to Jupiter, where the same physics plays out at larger scale, and to the growing catalog of exoplanets whose magnetic environments astronomers are now trying to characterize from a distance. “This study provides critical evidence for a long-held theory, that the rapid spin of massive planets like Saturn with active moons replaces the solar wind as the dominant force shaping magnetospheres,” Coates, a professor at University College London’s Mullard Space Science Laboratory, said in the UCL release.

There is a more practical reason the result matters. Saturn’s auroras are unusually bright and unusually variable, and one of the open questions in kronian physics has been why the most energetic emissions appear in the afternoon sector rather than at noon as on Earth (Lancaster University press release, August 10, 2026). A cusp that lives in the afternoon is also where magnetic reconnection, the explosive snapping and reconnecting of magnetic field lines that accelerates particles to keV energies and beyond, is most likely to fire. Knowing where the cusp sits helps the community locate the engine room of Saturn’s auroras, and the figure and full discussion have been widely covered in the planetary-science press (Sci.News writeup).

The analysis uses data from two of Cassini’s instruments, the magnetometer (MAG) and the Cassini Plasma Spectrometer (CAPS), collected between 2004 and 2010. Cassini spent thirteen years orbiting Saturn before its deliberate plunge into the planet’s cloud tops on 15 September 2017, and its archive is still turning up new science more than eight years after end of mission. For this study, the team pulled 67 discrete cusp crossings from the MAG and electron spectrometer data, then used a magnetic-field-line tracing model to localize each event in kronian local time. The result, summarized in the paper’s Figure 4, is a statistical map that looks nothing like Earth’s: instead of a single peak at noon, the kronian cusp occurrence rate is broadly distributed across the afternoon and early evening, with the centroid shifted several hours past local noon.

Two ingredients go into the explanation. The first is rotation. Saturn’s day is about 10.7 hours long, roughly 2.25 times shorter than Earth’s, and the planet’s deep interior is thought to rotate almost as a solid body. That fast spin drags the magnetic field, and everything locked to it, around the planet. The second is internal plasma loading. Enceladus, a small icy moon orbiting deep inside Saturn’s magnetosphere, vents water vapor from fractures near its south pole at a rate that turns out to be significant for the planet’s space environment. Cassini first imaged those plumes in 2005, and the ionized water they produce loads Saturn’s magnetosphere with heavy ions, including O+, OH+ and water-group species. Once ionized, this material is picked up by Saturn’s rotating magnetic field and pulled around the planet, where it contributes pressure that competes with the solar wind.

A 2019 paper in the Journal of Geophysical Research: Space Physics had already shown that Saturn’s equatorial current sheet oscillates with a period tied to the planet’s rotation rate, and that the sheet’s position depends on the balance between the solar wind pushing from outside and the internal plasma pushing from inside. The Xu et al. result extends that picture to the polar cusp, and shows that the imbalance is large enough to move the cusp by hours of local time.

The magnetosphere is the cavity carved out of the solar wind by a planet’s magnetic field. Its outer boundary, called the magnetopause, is the surface where the pressure of the planet’s field balances the pressure of the incoming solar wind. On Earth, the magnetopause is roughly ten Earth radii upstream, and the cusps are narrow funnel-shaped regions near the magnetic poles where field lines converge and dip back into the atmosphere. Charged particles from the solar wind can ride these open field lines directly into the upper atmosphere, which is why the cusps are preferred sites for auroral precipitation and direct solar wind entry into the magnetosphere.

The picture changes when a planet rotates fast and carries a lot of plasma around with it. At Saturn, the magnetosphere is enormous, more than ten times wider than the planet itself, and the heavy load of Enceladus-derived ions gives the internal plasma pressure enough weight to push the magnetopause outward on the night side and compress it on the day side. The fast rotation then drags the whole asymmetric structure around the planet. In the team’s high-resolution simulations, the net result is that the cusps are no longer located near local noon but are pushed several hours toward dusk.

That matters because of magnetic reconnection, the process by which oppositely directed magnetic field lines snap and reconnect across the magnetopause, releasing stored magnetic energy and accelerating particles to high energies. The location of the cusp sets the stage for where reconnection is most likely to occur, and where reconnection occurs is where the most energetic charged particles enter the magnetosphere. A cusp that sits in the afternoon sector therefore predicts afternoon-loaded energetic particles and auroral emissions concentrated on the afternoon and evening side of the polar cap.

The team also addressed a secondary question. Saturn’s magnetic field is famously hard to measure, because the planet’s radio emissions, the usual proxy for rotation rate, are unusually weak compared to Jupiter’s. The kronian rotation period has been a long-running puzzle. The Xu et al. result does not solve it, but it constrains one variable that goes into rotation-period models: the spatial distribution of plasma inside the magnetosphere.

Saturn’s magnetosphere is not a scaled-up Earth. The kronian cusp, the place where the solar wind enters the atmosphere most easily, is shifted into the afternoon because of two things the Earth does not have in any meaningful amount: a 10.7-hour rotation that drags magnetic field lines around the planet, and an active icy moon that loads the magnetosphere with water-group plasma. The Xu et al. 2026 paper compiles 67 such events from Cassini’s first six years at Saturn, and the statistical map of cusp occurrence has its centroid at 13:00 to 15:00 local time, with some events reaching 20:00.

The next test is at Jupiter, where Juno is still returning magnetometer data and where the same physics should produce a similarly skewed cusp distribution at larger scale. Looking further out, the framework should help astronomers reading magnetic signatures of hot Jupiters around other stars, where rapid rotation and intense plasma loading are the rule.

A piece of rock on the inner rim of Jezero Crater was soaked at least three times by very different waters over the course of Martian history, and NASA’s Perseverance rover has now read each of those chemical signatures off the same outcrop. The work, led by Candice Bedford at Purdue University, appears in the journal Communications Earth & Environment and was published online on September 21, 2026 (NASA Mars news, Sep 21 2026; Sci.News coverage). It turns a single strip of terrain along the crater’s former shoreline into one of the more interesting astrobiology sites on the planet, and it quietly resets a long-running debate about how the rocks there formed in the first place.

The location in question is called the Margin Unit, a band of rock that hugs the inner edge of Jezero’s rim where the ancient lake that once filled the crater met dry land. Mission scientists had been puzzling over it since before Perseverance even landed. Orbital instruments had shown strong carbonate signals along that shoreline, and carbonates are the sort of mineral that on Earth usually forms in shallow, life-friendly water, so the natural expectation was that the Margin Unit would turn out to be sedimentary rock, the layered, sandy stuff that is great at trapping microbial remains. Instead, Perseverance’s instruments showed that the band is mostly olivine-rich igneous rock, a coarse-grained material that on Earth typically crystallizes slowly underground from cooling magma. The carbonates are there, but they are not the original rock. They are an overprint, and the new paper unpacks how three separate waters, each with a different temperature and chemistry, left their own overprint on the same parent material over time (Universe Today, Sep 24 2026). That detail matters because the Margin Unit sits inside one of the largest carbonate exposures on Mars, and Jezero itself has been the centerpiece of NASA’s most ambitious astrobiology campaign in decades. If the orbital carbonate signal really does trace a single ancient lake, then the carbonates are essentially a fossil of that lake, and they should record the chemistry of habitable water. If the carbonates instead reflect a much messier history of groundwater, hydrothermal fluids, and lake water acting on the same rock at different times, the carbonates are not a snapshot of habitability, they are a stack of snapshots. Bedford’s team argues it is the latter, and they have the chemistry to show it.

The paper is built on more than 185 rock targets, all measured by SuperCam, the laser-and-spectrometer instrument perched on Perseverance’s mast. SuperCam fires a laser up to 6.5 meters away at a target rock and reads the spectrum of the resulting plasma to determine the rock’s chemistry. Across roughly 265 meters of vertical relief in the Margin Unit, the team at high elevations saw what they expected from a deep, slow-cooled igneous body: coarse olivine crystals with essentially no sign that water had ever reached them. About 265 meters lower, at the elevation where the ancient lake would have lapped against the crater wall, the same parent rock looks transformed. The olivine grains are fractured, and silica has filled the spaces between them. Some of those silica-bearing samples also contain the calcium-sulfate and fluorite veins that mark a third, much hotter stage of alteration (Space.com coverage). The sequence the team proposes runs as follows. Long before any lake existed in Jezero, carbon-dioxide-rich groundwater percolated through cracks in the olivine-rich bedrock. Where that groundwater met the right minerals, it produced carbonate. As the softer host rock around those carbonate-filled fractures eroded away over billions of years, the carbonates were left standing as ridges, which is the texture that first drew orbital attention to the Margin Unit in the first place. Then, in a window that the team cannot date precisely, water that may have been the Jezero paleolake, or a shift in the chemistry of groundwater connected to the lake, washed through the rock again. This second event partly remobilized the early carbonates and dropped silica into the pore spaces, with more silica appearing in rocks that sit below what would have been the old waterline. Finally, a third water event, this one clearly hotter, drove fluids through younger fractures in the eastern Margin Unit and deposited 25-centimeter-thick veins of calcium sulfate and fluorite. Fluorite, in particular, is a fingerprint of hydrothermal activity on Earth, and finding it on Mars is a strong hint that hot water was circulating through volcanic terrain long after the lake was gone.

Bedford has been working with Mars rovers for a decade, and her framing of the result is worth the gravity. “It is very rare that things are as we expect them to be from orbital data,” she said in a NASA press release accompanying the paper, adding that she hopes the work helps reshape how scientists view the history of water in Jezero and across Mars, and ultimately helps reconstruct the changing climate and habitability of early Mars (NASA Mars news). Her co-author Eleni Ravanis, who works on the SuperCam team at the University of Hawaii at Manoa, made a more specific point about what the silica signal is actually telling us: turning olivine into carbonate leaves silica behind, so the more silica you see, the closer the rock sat to the waterline, which makes silica a kind of paleoshoreline marker for a lake that no longer exists. The implication is that the Margin Unit is not just a place where water once flowed, it is a place where water acted multiple times, with each event leaving its own chemical signature stacked on top of the last.

Astrobiologically, that stacking is the most interesting part. On Earth, the reaction between water and olivine can release hydrogen, and there are microbes that can live off that hydrogen. The reaction also leaves behind carbonate and silica, both of which are good at preserving traces of those microbes. So a rock that has been soaked, reheated, and re-soaked in different fluids is not a worse biosignature target than a rock that has only been touched once. It is a record of a longer, more complicated habitability window, and the chemistry preserved inside it is denser. Perseverance has already drilled and cached samples from the Margin Unit, including cores named Pelican Point and Lefroy Bay, and those samples are expected to be among the most prized tubes in the future Mars Sample Return campaign. If any of those cached samples turn out to carry organic material, this paper will be the document that explains why those particular rocks were the right ones to pick up. The image below is a 360-degree Mastcam-Z panorama of Turquoise Bay, a region inside the Margin Unit, captured between October 5 and October 16, 2023, sols 933 through 944 of the mission, and it shows the kind of rolling, weathered terrain that SuperCam has been analyzing ever since (NASA Photojournal PIA26815).

Five weeks after lifting off from Kennedy Space Center on a Falcon Heavy, NASA’s Nancy Grace Roman Space Telescope did the one thing its coronagraph was built to do. On September 22, 2026, the Coronagraph Instrument, a technology demonstrator bolted onto the side of the Wide Field Instrument and shielded from starlight by a system of masks and deformable mirrors, opened to cosmic light for the first time, according to a NASA Roman blog post on September 30 and a contemporaneous Caltech announcement that also noted the milestone on September 27. The team had kept the detectors slightly warmer than they will run during science operations to keep contamination from settling on them, so the first image was a noisy one. Five days later, on September 27, they cooled the detectors and pointed again at a new patch of the Large Magellanic Cloud, and the second frame showed what a star field is supposed to look like. The coronagraph had a focused image, and the team could exhale.

The milestone matters because Roman’s coronagraph is the first active, deformable-mirror coronagraph ever operated in space. Hubble’s coronagraphs were passive: a black spot in the focal plane to physically block starlight, with no way to clean up the diffraction rings and speckles that left a bright star’s halo drowning out any nearby planet. Roman’s coronagraph, designed and built at NASA’s Jet Propulsion Laboratory and operated from Caltech’s IPAC, has two deformable mirrors that reshape themselves about 100 times a second to push the residual starlight into a controlled pattern the team calls a “dark hole.” If the approach works, it will be the technological seed for the Habitable Worlds Observatory, the still-in-formulation flagship NASA wants to launch in the 2040s to image Earth-like planets around Sun-like stars. None of that is possible until a deformable-mirror coronagraph has actually been tested in orbit, which is what made September 22 a quiet but real turning point.

The story behind September 22 starts with launch. Roman left Earth on August 30, 2026 (the mission profile on Wikipedia covers the launch, the cruise to L2, and the planned five-year primary survey), and the Wide Field Instrument and Coronagraph Instrument have been powering on and stretching out one subsystem after another as the observatory cruises toward its halo orbit around the Sun-Earth second Lagrange point. The coronagraph itself woke up on September 1. Mid-month it deployed its mechanical limbs and electronics. The team then spent September 15 through 21 testing the fine-guidance system, a piece of the observatory’s attitude control that has no separate hardware, but instead reads guide-star positions from a small patch of each of the 18 infrared detectors in the Wide Field Instrument. The fine-guidance system reports those positions to the spacecraft’s attitude control about four times per second, and the attitude control nudges the observatory a tiny amount each time to keep it from drifting. Tests confirmed Roman could hold pointing stable to better than one hundred-thousandth of a degree for half an hour at a time with the Wide Field Instrument, and for eight hours at a time with the Coronagraph, a tighter constraint because the coronagraph’s own internal stability loop was also running.

The stability budget for the coronagraph is what makes the second observation more impressive than the first. Even a few nanometers of mirror motion, or a few milliarcseconds of pointing error, would let unwanted starlight leak around the coronagraph’s masks and overwhelm a planet roughly a millionth as bright as its host star. Engineers put the current pointing performance in the same terms they use for any other precision pointing system: keeping a laser focused on a U.S. dime from about 150 miles away, with a goal of stretching that to roughly 230 miles as the commissioning continues. With the fine-guidance loop and the coronagraph’s own internal loop both running, on September 22 the team allowed starlight from a faint star in the Large Magellanic Cloud to fall on the coronagraph’s detectors. The image was blurry by design because the detectors were running warmer than they would in steady-state science, but it confirmed the optical path from the primary mirror, through the coronagraph masks and deformable mirrors, to the focal plane was clear and focused.

Five days later the team ran a second observation, this time after cooling the detectors toward their final operating temperature for better sensitivity, and pointed at a new region of the Large Magellanic Cloud where the team expected many stars in a single frame. The coronagraph saw them. That second observation confirmed the team could reliably move the field of view and keep the coronagraph stable while doing so. Roman’s coronagraph instrument scientist Vanessa Bailey, at JPL, summed up the second observation in plain language: “we’re breathing a sigh of relief.” Alexandra Greenbaum, coronagraph data management lead at IPAC, framed the milestone around what comes next: she’s looking forward to the rest of commissioning on both the operations and the data-processing side.

The technology itself is the science and engineering deep dive here, because that is what makes the milestone a milestone. A coronagraph suppresses starlight so a much fainter planet or disk next to that star becomes detectable. In a passive coronagraph, like the one on Hubble’s STIS instrument or on James Webb’s NIRCam, a physical mask in the focal plane blocks the bright star’s core, but the wave nature of light still spreads some of that starlight into a halo of rings and speckles around the mask edge. Roman’s coronagraph actively corrects the wavefront of the incoming light using two deformable mirrors, surfaces whose shapes can be changed hundreds of times per second by pushing on the back of the mirror with hundreds of small actuators. When the deformable-mirror pattern is right, the speckle pattern from starlight interferes with itself and cancels out across a small region of the image, opening what the team calls the “dark hole.” That dark hole is where a much fainter planet, if there is one, would be visible. The trade is that the dark-hole trick only works for a narrow annulus around the star, and the deformable mirrors must keep running or the dark hole closes within milliseconds. Roman’s coronagraph uses two 64-by-64 actuator mirrors and a set of focal-plane masks that physically block the starlight before the deformable mirrors do their work, and the September 22 observation was a sanity check that all of those pieces are aligned well enough to deliver a focused image at all.

The other piece that makes the coronagraph unusual is the fine-guidance system, which is part of why the September 15-to-21 testing week mattered. Hubble, JWST, and most space telescopes carry a separate guider instrument, basically a small auxiliary telescope that locks onto a guide star and tells the main telescope where it is pointing. Roman does not have one. Instead, the team reuses a few pixels on the edges of the Wide Field Instrument’s 18 infrared detectors to track guide stars, and uses the spectra of those guide stars, not just their positions, to hold pointing steady. That gives Roman a way to point more precisely than a star-image tracker would, because spectra carry more information than a centroid, and it saves mass by removing a separate instrument. Roman’s guiding lead Begoña Vila, at Goddard, said the team is looking forward to validating that spectral-guiding mode in the coming weeks. The September 15-to-21 tests used position-only guiding; spectral guiding is the second tier, and confirming it is one of the remaining milestones of commissioning.

What comes next is a three-month commissioning campaign that will run Roman through a series of progressively harder tests. The coronagraph team will practice keeping the dark hole stable for longer and longer exposures, will test the deformable-mirror control loop under realistic thermal conditions, and will try imaging the dust disks that the coronagraph was originally scoped to study. NASA expects Roman’s first image release in early 2027, with the coronagraph’s first science-quality image landing later in the year once the deformable-mirror dark hole is reliably closing on demand. The September 22 and 27 observations are not those images. They are the “we have a working instrument” sign on the door, and the second one shows that the door is on hinges.