One Flash in the Xenon: The Dark Matter Event Nobody Can Explain

0:00Welcome back to Astronomy Daily. It's Friday, September 4th, 2026. I'm Anna and this is Series 5 episode 185. And I'm Avery. Anna, I want to start today with a number 2.6. 2.6 Sigma, which is not a discovery. You said that very fast. I said it fast because it's the most important sentence in the story. But here's the rest of it. A dark matter detector, a mile underground in South Dakota, recorded a single flash of light. In exactly the place a dark matter particle was supposed to show up. And the collaboration spent months trying to make that flash go away and could not do it. One event. One event. That's our lead. Including why we can't explain it and we found dark matter are very different sentences. Then India has put its first imaging satellite into geostationary

1:01orbit. A real first and a capability nobody else in the region has. The first private mission to Venus has been grounded, not by Venus but by a rocket that hasn't flown yet. And a lovely piece of physics out of Syracuse. Why a star that keeps getting torn apart by a black hole puts on a fainter show every time. Plus the sky this weekend both hemispheres and a new sunspot worth knowing about. Let's get into it. Start me at the beginning. Who announced what and where? The LZ collaboration Lux Zeppelin presented a result this week at TEV Particle Astrophysics 2026 in Chiba, Japan, which wraps up today. Brown University released it on Tuesday. The US Department of Energy has published its own account and the paper has gone to physical review letters. And LZ is the big xenon one. LZ is the big xenon one. Ten tons of ultra pure liquid xenon in a

2:03tank at the Sanford Underground Research Facility in Leeds, South Dakota. That's the old home state gold mine and the detector sits about a mile down roughly 1480 meters of rock overhead. Why underground? Because the enemy isn't darkness. It's noise at the surface. You're rained on constantly by cosmic rays. A mile of rock filters nearly all of that out. Then they wrap the xenon in a water tank and a veto detector for stray neutrons and build it all from material screened for radio activity to absurd levels. The art of this field is subtraction. You spend 20 years removing every signal you can explain. Then look at what's left. And what are they hoping is left? A WIMP. A weakly interacting massive particle, the leading dark matter candidate for about 40 years. A heavy particle left over from the early universe that has mass. So it pulls on galaxies gravitationally, but ignores light and ignores ordinary matter almost all of the time. Almost all of the time being the operative

3:08phrase. That's the whole bet. If a WIMP occasionally bumps into an atomic nucleus, a big enough tub of xenon sitting quietly for long enough should eventually record one. The nucleus recoils and you get two flashes, a prompt one, then a second from electrons drifting up through the liquid. Together they tell you where in the tank it happened and whether you hit a nucleus or just knocked an electron loose. Okay, tell me about the event. It's in data taken between March 2023 and April 2024, 220 live days. That data set has been analyzed before. LZ published world leading limits from it. What's new is that a team went back and searched a much wider range of possible interactions than the standard analysis covers, including higher energies. And one event turned up. A nuclear recoil in a region where the expected background is very close to zero. Higher energy. Is that where you'd expect

4:11dark matter? No. And that's the first genuinely odd thing. The simplest WIMP models put your first signal at low energies. This is up the other end. Taken at face value, it points to a particle of at least 200 giga electron volts, more than 200 times the mass of a proton. Interacting in a way, the simplest models don't predict. So it's not the WIMP anyone ordered. It is not the WIMP anyone ordered, which cuts both ways and will come back to that. Give me the statistics honestly. 2.6 sigma globally, 3.4 sigma locally, and the difference between those two numbers is the most useful thing I can teach anyone today. Go on. Local significance asks, at this exact energy for this exact mass, how surprising is this event? Fairly surprising. Global significance asks the fairer question. I searched a whole range of masses and energies. So how surprising is it that somewhere in that range I found one

5:13odd thing. Account for the haystack and the surprise drops. That's the look elsewhere effect. Honest experiments quote both, and LZ did. And 2.6 sigma means what in plain terms? Roughly a half a percent chance known background's produced it. Which sounds compelling until you remember the bar. Particle physics calls something a discovery at 5 sigma, about 1 in 3 and a half million. 2.6 is nowhere near it, and physicists have watched 3 sigma results evaporate for decades. Did they try to kill it? For months. Cosmic rays, neutrons from the rock, radioactivity in the detector materials, instrumental artifacts, all modeled. Aaron Manalai say at Berkeley Lab, who chairs LZ's institutional board said it's the first example in any experiment he's worked on of an outlier that appears valid in every way. That's a striking thing for an experimentalist to say out loud. What does the spokesperson say?

6:16Rick Gates skull at brown is the spokesperson, and he's been about as disciplined as you can be. His line, with only one event we don't want to get ahead of ourselves. We are not claiming to have seen dark matter. And separately, we're very intrigued to see this event in the data, in the region where we expect dark matter to show up, and the competing backgrounds are very low. Both things at once. Both things at once, and that's the correct posture. Sam Erickson at Bristol led the analysis and made the point that matters. Dark matter events are expected to be so rare that only a handful could mark the first detection. You can't dismiss one event for being single, but you can't build a discovery on it either. Is there a UK end to this? A significant one. Imperial College London did much of the work characterizing the event, and Henrique Arauho there put it beautifully. We need to analyze more data to be sure, but these are certainly interesting times.

7:16Now, you promised the caveat about it not being the expected two ways to read an unexpected signal. The generous one, nature isn't a blige to be simple, and 40 years of not finding dark matter may be exactly because we searched the tiniest places first. The unkind one. When a result lands where no model predicted, an unmodeled background is a very live explanation. The reason you haven't modeled it is that you didn't know it was there. Has this field been burned before? Repeatedly, Dama in Italy has claimed an annual dark matter signal for over 20 years that nobody else can reproduce. Xenon 1T reported an excess in 2020 that caused enormous excitement and was most likely tritium contamination, a hydrogen isotope at a level almost too small to measure. That's the standard to hear this week against. To Elziz credit, they've published this as an anomaly, not a discovery. So what settles it?

8:17More Xenon and more time. Elziz has already banked substantially more data than went into this analysis and is running toward a thousand live days. If it's real, the rate is set by physics and more events follow, the significance climbs. If it's a fluke, it decays as exposure grows, and a proposed successor, X-L-Z-D, would hold 10 times the Xenon. This resolves itself in data, not argument. Southern hemisphere angle? Because dark matter feels like a northern hemisphere sport. It has been, and that's changing for a genuinely clever reason. There's now an underground lab in Australia, SUPL, the Stauwell Underground Physics Laboratory, a kilometer down a working gold mine in Western Victoria. It's the first underground physics lab in the Southern hemisphere, built by the University of Melbourne with the arc center of excellence for dark matter particle physics and Ansto, and its first experiment, Sabre South, moves in late this year.

9:18And why does the hemisphere matter for dark matter of all things? Because of Dama. Its claim is that the signal rises and falls once a year. As Earth's motion around the Sun adds to and subtracts from the solar system's motion through the galaxy's dark matter halo. The trouble is that plenty of ordinary things cycle annually, too. Temperature, radon, cosmic ray rates, and in Italy they all peak in summer, alongside the claimed signal. The seasons are flipped, and the dark matter signal isn't. Run a near identical detector in the southern hemisphere, and a real galactic signal peaks in the same calendar month that does in Italy. While a seasonal artifact peaks six months out, elegant piece of experiment design, and the only place on Earth you can do it is the one we happen to broadcast from. So what should people take away from today? Three things. Elzy has found something it cannot explain,

10:19in the right place, and said so honestly. One event is one event, and 2.6 sigma is a long way from a discovery, and the answer is already being collected. The detector is running right now. The honest headline is, dark matter hunters find something they can explain and refuse to over claim it. A good day for science, even if it isn't the day. Onto our second story today, and this one is a genuine national first. Overnight our time, 255 in the morning, Indian Standard Time on the 4th, which was 525 yesterday evening, US Eastern. ISRO launched EOS05 on a GSLV Mark 2 out of Sri Hari Koda. And it's the orbit that's the story, not the rocket. Exactly. EOS05 is India's first dedicated imaging satellite headed for geosynchronous orbit. Everything India has flown for Earth observation until now has been in low orbit,

11:21a few hundred kilometers up. Spell out the difference for people. A low orbit imaging satellite is a sprinter. It races around the planet in 90 minutes and gives you a superb, very high resolution snapshot of a strip of ground. And then it's gone, and you wait. Depending on the orbit, you might get another look in a day or several days. Whereas geostationary is a stair. Geostationary is a stair. 36,000 kilometers up, matching Earth rotation. So from the ground, the satellite appears to hang motionless over the same piece of the planet permanently. You don't get a revisit time because you never leave. And that changes what you can use it for. Completely. ISRO's framing is persistent coverage of the subcontinent, and the applications are obvious once you say it that way. A cyclone forming in the Bay of Bengal. You watch it develop continuously instead of getting one frame a day.

12:24A flood, you see the water advance. A firefront, a landslide, a border. It's a dual-use satellite, civil and military, and India hasn't been shy about that. There's a tradeoff though, surely. There is. And it's worth being honest about it. You are imagining from 100 times further away than a low orbit satellite, so the resolution is inevitably coarser. You are not reading number plates from geostationary orbit. What you're buying is time, not detail. And for disaster response, time is usually the thing you're short of. How did the launch go? Clean. The spacecraft is about 2,367 kilograms, an ISRO chairman V. Narayanan said it was successfully and precisely injected into its planned orbit. From here, EOS05 works its way up to its final station over the coming days. And the GSLV has had a mixed history.

13:25It has, which is part of why this matters to ISRO. The GSLV Mark II has now flown 12 times for 10 successes. That's a vehicle that has visibly matured, and it's the one carrying India's heavier missions to high orbit. Small country club this. Very small. Dedicated high-resolution imaging from geostationary orbit is a capability only a handful of nations have ever fielded. India has just joined that list, and it did it with its own rocket from its own spaceport. Third story, and it's a frustrating one. The first privately funded mission to another planet is still on the ground, and it's going to stay there for a while yet. This is the Venus Life Finder. That's it. It's an MIT-led mission driven by Sarah Seeger, flying in partnership with RocketLab, and it is beautifully, almost aggressively simple. A small probe, one instrument, a few minutes of useful life.

14:29One instrument. That's it. One instrument. It's called an autofluorescence nifolometer, which is a mouthful for a fairly elegant idea. You fire an ultraviolet laser into the cloud droplets as you fall through them. Certain organic molecules absorb ultraviolet light and re-emitted at a different wavelength. They floresce. So the instrument is looking for a glow that ordinary sulfuric acid chemistry shouldn't produce. And why the cloud specifically? Because the surface of Venus is out of the question. 460 odd degrees, 90 atmospheres. But between about 45 and 60 kilometers up, the temperature and pressure are close to conditions at sea level on Earth. Extremely acidic, but not thermodynamically hopeless. That's the only plausible habitable niche on the planet. And it's what the 2020 phosphine claim put back on the table. A result that is still genuinely disputed, and which this mission is designed to go and settle rather than argue about.

15:31So why isn't it flying? Neutron. The mission moved onto RocketLab's new medium lift rocket, and neutron hasn't flown yet. It was originally talked about for 2024, slipped to 2026, and it's still in qualification. The launch date on RocketLab's own website now simply says to be confirmed. How is Seager taking it? With more grace than I would. Her line was, we are awaiting neutron readiness. And she went on to say she continues to have high hopes for the mission, and for its role in demonstrating what private enterprise can do in space exploration. There's an irony in there somewhere. There's a real one. The whole pitch of this mission was speed, that a small, focused, privately funded probe could go and answer one sharp question years before an agency flight could be approved, built, and launched. And it's now waiting on launch capacity, which is the one part of the problem private industry was supposed to have solved.

16:33Meanwhile, Venus is getting crowded. It is. NASA's DaVinci and Veritas and Europe's Envision are all in the pipeline for around the end of this decade. The Venus life finder was meant to be the scrappy one that got there first. And that lead is quietly evaporating on a launch pad in Virginia. Last news story and its pure astrophysics. Published in the Astrophysical Journal on Tuesday, led by Ananya Bandopadjai, a doctoral student at Syracuse University, with Benjamin Amand and Eric Kaufflin, plus collaborators at Leeds, MIT, and the Space Telescope Science Institute. And the puzzle is about stars that survive being eaten. Partly eaten. When a star wanders too close to a supermassive black hole and is ripped apart entirely, that's a title disruption event, one enormous flare and it's over. But there's a smaller class where the star is only partly stripped on each pass,

17:36survives, and comes back around. So it flares over and over. Over and over on a schedule. The famous one is a Sassan 14KO, which flares roughly every 114 days and has done so for years. There are now something like seven or eight of these known. And what's the problem? The flares get dimmer each time, which sounds intuitive, less star left to strip, except the simulations kept saying the opposite. Strip material from a star and it puffs up, and a puffier star is easier to strip next time round. Models kept producing flares that got brighter, and the sky kept producing flares that got fainter. So what's the missing ingredient? Spin. This team ran hydrodynamic simulations of a high mass main sequence star being repeatedly disrupted by a black hole of about a million solar masses. And the key move was giving the star a fast rotation before the first encounter,

18:41spinning in the same direction as its orbit. Why does that change the outcome? Because ordinarily the encounter itself spins the star up, and that spin up is part of what makes the next pass more violent. If the star arrives already rotating at a decent fraction of its breakup speed, there's very little extra spin to give it. The debris then falls back to the black hole, spread over a longer stretch of time, instead of arriving in one lump, and the same material dribbling in over longer makes a fainter more drawn out flare. Spread the fuel out and the fire is lower. That's it exactly, and with tens of percent of breakup rotation, prograde, the simulations reproduce the dimming that's actually observed. Does it tell us anything about how the star got there in the first place? It does, and that's the bonus. A fast spinning star on a tight orbit around a supermassive black hole fits the hills mechanism.

19:41A binary pair strays too close. The black hole keeps one star and flings the other away at enormous speed. The captured one lands exactly where you need it. So the spin isn't an arbitrary knob. It's a fingerprint of how these systems get built. Right, let's get you outside. And the moon is doing us a favor this weekend. Last quarter today. Last quarter today, September 4th, which means it doesn't rise until around midnight, so the entire evening is dark. If you have been waiting for a night to actually look at something faint, this is the weekend. Southern hemisphere first? Southern hemisphere first, because we get the best of it. From Sydney, the sun sets about 20 to 6 now. And once it's properly dark, the center of the Milky Way is high overhead. Sagittarius and Scorpius almost directly above you. From mid-southern latitudes, the galactic core passes near the zenith. So you're looking through the least possible atmosphere. Northern listeners get the same object low and murky above the southern horizon.

20:45It's the one thing we can be smug about. And September is the last good month before it sinks westward. What do you actually look at? Find the teapot of Sagittarius with the naked eye and let your eye drift up out of the spout. That steam is the galactic center. Binoculars turn it into star clouds and dark lanes. And the lagoon nebula is sitting right there, along with a dozen globular clusters. Planets? Venus in the west after sunset, brilliant, unmistakable, low, and building toward greatest brilliancy on the 18th. Saturn is up most of the night in Aquarius, climbing toward opposition on October 4th. And from here, it passes far higher overhead than it does for Northern observers. Predon? Jupiter is the Predon showpiece, well up in the east before sunrise. And on Sunday morning, the 6th, Mars sits just a few degrees below a thin, waning, crescent moon. That's a lovely one for a phone camera if you're up early.

21:47Now North America, because there's a proper event coming. On Tuesday the 8th, the moon occults Jupiter, the planet passes behind the lunar disk. The footprint favors northeastern Asia, where it happens in the dawn sky, and eastern North America, where it happens after sunrise in broad daylight. Daylight, which brings us to the standing reminder, and it applies directly here. If you are observing anywhere near the sun, hunting for Jupiter in a bright sky, or looking at the sunspot I'm about to mention, any filter you use for direct solar viewing, must be certified to the ISO12312-2 standard. Not sunglasses, not welding glass of unknown grade, not smoked glass, not a phone screen. ISO12312-2, and check the certification is real. Sweeping binoculars or a telescope across a daylight sky is exactly the situation where people injure themselves permanently,

22:49and it takes a fraction of a second. And the day after, there's one for us. The 9th, the moon occults regulus, the brightest star in Leo, and that footprint runs across the South Pacific, New Caledonia best placed. Not Australia, unfortunately, but if you're in that track, a first magnitude star vanishing off the edge of the moon is one of the sharpest things you'll ever see. Instantaneous. You mentioned a sunspot. A new one, active region 4524, which rotated into view over the northeastern limb this week and has been busy. It fired an M3 flare peaking at 1920 universal time on Wednesday the second, plus a stack of smaller ones. The coronal mass ejection from that flare isn't aimed at us, but a filament eruption the same day throughout material that may deliver a glancing blow around Monday the 7th. Aurora chances? Modest and honest, quiet conditions through the weekend, so nothing to promise tonight or Saturday.

23:50Monday is the one to watch, and if anything comes of it, the people with a shot are Tasmania and Southern New Zealand down here, and the northern tier of the U.S., Canada, and Scotland up there. Watch the space weather feeds rather than the headlines. And one for northern observers with binoculars. The double cluster in Perseus. Two open clusters side by side in the same binocular field. Nagadai, it's a smudge. In binoculars, it's one of the best sites in the sky, and a moonless evening is exactly when to try it. And that's episode 185. A dark matter detector, a mile underground, has recorded one flash of light a cannot explain, and has been admirably careful about what that does and doesn't mean. India has put its first imaging satellite on station over the subcontinent. The first private mission to Venus is stuck waiting on a rocket that hasn't flown, and a star that keeps surviving a black hole shines a little fainter each time because of how fast it was already spinning.

24:52Full show notes, links to every primary source, and the whole back catalog are at astronomydaily.io. You can find us on X, Instagram, and TikTok at AstroDailyPod. And if you've got a question or a correction, we genuinely want it. There's a contact form on the website. And if today's episode was useful, the single most helpful thing you can do is send it to one person who'd enjoy it. It's the weekend. Get outside and look up while the moon's out of the way. Until tomorrow, clear skies. Clear skies. The star is the toe. The star is the toe.