Starry Nights and Cosmic Insights: Van Gogh's Real Sky Revealed

0:00 Anna: Welcome to today's episode of Astronomy AstroDailyPod. Let's kick off with something a little different. In June 1889, Vincent van Gogh wrote to his brother Theo that he finished a new study of a starry sky. It became the most famous night sky in art. And for more than a century, people have asked whether those swirling stars were ever real. Avery: A new paper says yes. It points to one particular morning, the pre dawn sky over the south of France around 22 July 1889. We checked the sky for ourselves and we'll tell you what we found and where.

0:39 The argument is still open. Anna: Also today, NASA's Roman telescope moves into its final commissioning phase and Rubin's test camera turns out a seagull. I'm Anna. Avery: Hi, I'm Avery. This is Astronomy daily for Friday 23rd October 2026. Let's get into it. Anna: The starry night has hung in the Museum of Modern Art in New York since 1941. You know it, even if you don't think you do. A dark cypress, like a flame. On the left, a sleeping village. Below, 11 blazing stars, a crescent moon and that great rolling swirl across the middle of the sky.

1:20 Avery: Van Gogh painted it at the Asylum of St. Paul de Montsou in St. Remy de Provence, where he'd admitted himself in May 1889. Art historians usually dated to mid June that year, mainly because of that letter to theo written around 18 June. Anna: The obvious question for us is whether any of it is the real sky or whether it's all feeling and imagination. That question has been argued before. In 1984, the art historian Albert Boim, working with the astronomer George Abell, went looking for a real sky.

1:57 In 1986, Harvard astronomer Charles Whitney looked again. They didn't agree. And in earlier readings, the little group of stars in the middle of the canvas was matched to Aries. Avery: Now, a new study in Scientific Reports published this week takes another run at it. The lead author is Elijah Jonghyun Kim, PhD candidate at UH Kyunghee University School of Space Research in South Korea. He's also chief executive of SL Lab, a company that builds robotic telescopes. We'll come back to that. Anna: What's interesting here is the method, because it isn't one argument, it's four separate checks that all have to agree.

2:41 Avery: First, the view. Kim travelled to St. Remy in 2019 to work out what Van Gogh could actually see from his room and how much sky would fit in that window. That gives a, uh, narrow strip of sky looking roughly east. Anna: Second, the shapes the team took the five central stars in the painting and compared their pattern with real star groups using a technique called Procrustes analysis. Avery: The name comes from the Greek myth of the innkeeper who stretched or cut his guests to fit his bed. In the statistical version, you scale, rotate and slide one pattern onto another until they line up as closely as possible, then measure what's left over.

3:25 The smaller the leftover mismatch, the better the fit. Anna: On that test, the best match was the Hyades, the V shaped cluster in Taurus around the bright star Aldebaran. And here's a nice detail. A, uh, pen drawing Van Gogh made after the painting matched even more closely. Avery: If the Hyades are new to you, they're well worth knowing. They're the nearest star cluster of their kind, an open cluster about 150 light years away. Its hundreds of stars were born together roughly 625 million years ago and are still drifting through space as a family.

4:05 Anna: Because they're close and their distance is so well measured, the Hyades have been a rung on astronomy's distance ladder for over a century. Measure them carefully and you can calibrate stars much further away. The one bright star that doesn't belong is Aldebaran, which sits in front of the cluster at about 65 light years. Avery: Third, the dates. The Hyades only work if they were actually above the horizon in that part of the sky before dawn. Along with a crescent Moon and the brilliant morning star, Venus is that morning star.

4:42 All three come together around the 22nd of July, 1889. Anna: And fourth, the documents. The team reread Van Gogh's letters alongside historical weather records. His letters from that summer keep mentioning cicadas, which suggests a hot spell. And the team argues that fits July better than June, because this is the Avery: kind of claim you can test. We did. We computed the sky over St. Remy for nautical dawn on 22 July, 1889. Anna: The moon was a waning crescent, about 28% lit and roughly 29 degrees up in the east.

5:23 Venus was blazing at magnitude -4.2, about 18 degrees up in the east. And Aldebaran, with the Hyades around it, sat about 20 degrees up, almost beside Venus. Avery: Then we tried mid June, the traditional date. On the morning of 19 June, the moon was a fat, waning gibbous, about 2/3 lit, which is not the slim crescent in the painting. And the Hyades hadn't even risen before dawn. They were still 7 degrees below the horizon. Anna: So on the astronomy, the paper holds Up. A late July pre dawn sky has every ingredient.

6:03 The painting has. A mid June sky doesn't. Avery: Kim's answer is that Van Gogh may have conceived the picture in June and finished it after the 22nd of July. He's careful not to claim more than that. He says the study doesn't definitively rewrite the painting's accepted history and that linking the letters to the timeline relies partly on his own interpretation. Anna: The team also ran the same method on other paintings. They checked the field of view in Starry Night over the Rhone from 1888 against photographs taken on site.

6:37 They suggest the stars in Cafe Terrace at night may be the summer triangle. Kim's line is that the conversation with Van Gogh's other skies is still going. Avery: Now the other side, because there is one she Shape matching on its own wasn't decisive. The team says the Procrustes test couldn't rule out other star patterns that fit similarly well, which is why the history and visibility checks matter so much. 5 stars is not a lot to work with. Anna: Donald Olson of Texas State University, an astronomer well known for dating paintings and literature by their skies, told Science News he has doubts.

7:19 His point is that the narrow field of view the authors assume doesn't fit the much wider view you'd need to include the hills and cliffs in the painting. Avery: And he proposes a simple test. If the theory is right, he says, it should be easy. Ask an experienced amateur astronomer to find the Hyades in the painting and see whether they pick out the star the team identifies as Aldebaran. He doubts they would. Anna: Art historians also point out that the painting's style sits differently from Van Gogh's other June work.

7:52 That's partly why the June date has held for so long. Avery: And a, uh, transparency note. SLLAB is a company and the authors declare a competing interest. The company says it may benefit from related publications, educational programmes, exhibitions and media. That doesn't make the astronomy wrong. To their credit, the team has put its code and data online for anyone to check. Anna: Does it matter what night it was? In one sense, no. Van Gogh wasn't drawing a star chart. The sorrel, the oversized stars and the glowing cypress are about how the night felt, not where Aldebaran was.

8:32 Avery: But Van Gogh's letters show he really did watch the sky. And he wrote about getting up to look at the countryside before sunrise. If he was looking east before dawn that July, he'd have seen exactly what we computed A thin moon, a brilliant Venus and the little V of the Hyades real sky. Painted as a feeling. That's not a bad way to describe what a lot of us do when we go outside at night, and you Anna: can see the same thing for yourself. The Hyades are back in our evening sky right now, and in Skywatch, we'll tell you exactly when to look.

9:10 Venus, meanwhile, is about to slip past the sun and return as the morning star, just as it was over St. Remy. Avery: Our next story is a progress report and a good one. NASA's Nancy Grace Roman Space Telescope has entered science commissioning the last phase before it starts its surveys. Anna: Longtime listeners have followed Roman since before launch. It went up on a Falcon Heavy on 30 August. In September, we covered the Wide Field instrument switching on and last week its focus result. Avery: This week, NASA says the checkout of the observatory and its main instrument is complete.

9:50 Julie McEnery, Roman's senior project scientist at Goddard, put it the observatory is fully functional. It can guide, take exposures, move its data and slew from one target to the next. Anna: So the job now moves from does it work? To how good are the pictures? That sounds like a small step, but for Roman, it's the whole game. Avery: Roman's big surveys look for very small changes. Galaxy shapes stretched by a fraction of a percent by dark matter, supernovae brightening and fading across months, stars in the galactic bulge that flicker for a few hours when a planet passes in front of them.

10:33 Anna: If some frames came out slightly blurrier or brighter than others because of the telescope itself, you couldn't tell a real change in the sky from a wobble in the instrument. Josh Sleeter, the Wide Field Instrument scientist, said any imperfection has to be driven down because even very small flaws could hide the signals thereafter. Avery: One worry going in was guiding Roman holds steady by locking onto individual stars, and the team wasn't sure that would work in very crowded fields or behind thick dust.

11:09 Tom Barkley, the operations project scientist, says test observations in exactly those regions went, um, well. Anna: That matters for the Galactic Bulge Survey in particular, which stares into the most crowded, dustiest part of our sky. That's also the part that rides high over Australia in winter. Avery: Now comes a planned run of test observations to calibrate how the detectors respond to light and to tune the image quality before science operations begin. NASA hasn't put a date on that handover in this update.

11:46 Anna: And a fair caveat this update gives no numbers. It's a milestone, and the quality figures will come later. But one line stood out. Toleder said a typical Wide Field instrument observation will reach objects about 100 million times fainter than the human eye can see. Avery: Every one of those frames covers about 200 times the area of a Hubble infrared image. That's the reason this telescope exists faint and wide at the same time. Anna: From a telescope just getting ready to one that's been rehearsing for two years.

12:22 The Vera C Rubin Observatory has released a new image of the Seagull Nebula. And there's a twist in how it was made. Avery: It wasn't taken with Rubin's giant camera at all. It came from commcam, the commissioning camera, a much smaller engineering stand in used to test the telescope before the real camera went in some context. Anna: On scale, the main camera, LSSTcam is about 3.2 gigapixels, the largest digital camera ever built for astronomy. ComCam is about 144 megapixels. Each frame still covers a patch of sky about two full moons across.

13:03 Avery: CommCam ran for seven weeks from 24 October to 11 December 2024 and took about 16,000 exposures. So this release lands two years almost to the day since that campaign began. Anna: The seagull itself is a sprawling cloud of glowing hydrogen and dust on the border of Canis Major and monocerotis, roughly 3,700 to 3,800 light years away. The COMMCAM frame centres on the bird's head IC2177, where new stars are forming. Avery: And here's why the release is more than a pretty picture. Star forming regions are hard for processing software.

13:47 Wispy nebulosity fills the whole frame. There's no clean background sky, and automated pipelines can mistake glowing gas for noise or noise for structure. Anna: Wire Lab says this field gave the image processing team early practice with exactly that kind of mess, using the same software that now handles the full survey. Avery: To be clear about what this is, an image release from data taken in late 2024. There's no new scientific paper behind it. The science from Rubin is arriving separately. We covered one result.

14:21 The newly found globular cluster Rubin GC1 back in September. Anna: But it's a good reminder that the observatory on, um, Cerro Pechan in Chile, which will map the southern sky every few nights for a decade, was tested and tuned long before most of us saw its first images in June 2025. Avery: And the seagull is easy for everyone. Right now it's about 67 degrees up at its highest from Sydney, around 40 to 45 from Los Angeles and New York, and um 28 from London. Look near Sirius. In the small hours though, you'll want a dark sight and a camera to bring out the red.

15:01 Anna: Time for a few quick hits first ESA's Hera Mission, its second big braking burn, BRM2, about 31 minutes long, was scheduled for early Thursday the 22nd after last week's 93 minute BRM M1 was confirmed within its required parameters. Avery: As we record, ESA hasn't posted a result for that burn. Three smaller transition burns are planned later this month with arrival at Didymus in November. We'll bring you the outcome as soon as it's announced. Anna: Base X now Says Starship Flight 15 is targeted for November and that it will try to catch both stages with the launch the super heavy booster, which it has done before, and for the first time the Starship upper stage that came from the company's third quarter call on the 20th.

15:51 Gwynne Shotwell added that a uh, reflown upper stage and the long delayed ship to ship propellant transfer test are both aimed at early 2027. The word is targeted, not scheduled NASA's Avery: Science Mission Directorate has published an open letter to the science community called Faster to Science. It says the agency is changing how it runs competitions and selects missions so that science gets delivered faster. The details are due at a virtual NASA science town hall on Monday the 26th US time, and we'll report what changes once they're spelled out.

16:29 Anna: NASA's Swift Observatory is still expected to re enter in early November, according to the Associated Press. NASA itself hasn't published a date so Avery: we won't give one and NASA holds its Artemis Update later today US time. That's 2:00pm Eastern on Friday, which is 5:00am Saturday here in Sydney. We'll have what's said on tomorrow's weekend wrap. Anna: What's next on the show? Avery: Time for Skywatch and we'll start with a planet you can't see. Anna: Venus reaches inferior conjunction this weekend passing between Earth and the sun be instant is 3:44 universal time on Saturday the 24th.

17:10 That's 2:44 Saturday afternoon in Sydney and quarter to five Saturday morning in London. In North America it falls on Friday evening, 8:44 in Los Angeles and 16 minutes before midnight in New York. Avery: Venus will be about 41 million kilometres from us, a little over a quarter of the Earth's sun distance, but only about six and a half degrees from the sun in the sky, so it's lost in the glare for now. Anna: And the safety line we include every episode. Never look at the sun or sweet binoculars or a telescope and anywhere near it.

17:49 Without a filter certified to ISO 123122. Sunglasses are not a solar filter. With Venus this close to the sun, the safe plan is to wait. Avery: It won't be long. Within a couple of weeks, Venus climbs into the predawn sky as the morning star, the role it played over St. Remy in July 1889. Anna: On Saturday evening, the nearly full Moon sits close to Saturn. At 9pm local time. They're about 6.5 degrees apart from Sydney and seven from London, a little wider from Los Angeles and New York. At 9 to 11 degrees, Saturn is magnitude plus 0.45 with its rings still opening.

18:39 Avery: Full Moon is at 4:11 Universal Time on Monday the 26th. That's 3:11 Monday afternoon in Sydney. So the Moon rises full that evening. It's ten past midnight on Monday in New York and eleven past nine on, uh, Sunday evening in Los Angeles. Anna: In the UK, clocks go back an hour early on Sunday the 25th. So London's full moon is at 4:11 Monday morning Greenwich Mean Time. North America changes its clocks a week later on the 1st of November. Avery: Since the Hyades are this episode's star cluster, here's how to find them.

19:20 Look for the V shaped group around the bright orange star Aldebaran. The eye of Torresta Bull. Anna: From the northern cities, Aldebaran, uh, rises around 8:30pm local time, and by 11:30 it's 32 to 36 degrees up in the east from Sydney, it rises a little after 10pm and is about 5015 degrees up by 11:30. It gets much better after midnight. And from down here, the V hangs upside down like a, uh, capital A, Avery: as we said in the lead, Aldebaran, uh, isn't actually part of the cluster. It sits in front at about 65 light years.

20:00 Binoculars are the best way to see the Hyades. The cluster spans more than 5 degrees too wide for most telescopes. Anna: A word about the Moon, though. It's nearly full and up, uh, most of the night. And next week the bright waning Moon moves through Taurus. It passes within about a degree and a half of the pleiades around the 28th and stays roughly 10 degrees from Aldebaran. For the fainter Hyades stars, your best dark sky chance is the first week of November after last quarter on the Avery: second, Mercury is a Southern Hemisphere treat.

20:37 This week at the end of civil Twilight, it's about 14 degrees up in the west from Sydney. At uh, magnitude 0.8. From Los Angeles, it's just three degrees up from New York about one and um from London it has already set. Anna: The pre dawn sky belongs to the north. At nautical dawn on Saturday, Mars is about 60 degrees up and Jupiter around 50 from Los Angeles and New York and still above 45 from London. From Sydney they're only 22 to 24 degrees up. Avery: The two planets are 8.8 degrees apart now closing to just 1.2 degrees on 16 November.

21:20 It's worth following through the next three weeks. Anna: Finally, space weather. After a minor G1 storm on Wednesday night, another stream of fast solar wind from a coronal hole is due around Saturday. Earthsky also reports modelling that suggests part of a recent coronal mass ejection could deliver a glancing blow late on the 26th or early on the 27th. Avery: With a full Moon, only a fairly strong display will cut through. But if you're at high latitude north or south, keep an eye on the forecasts in Australia.

21:56 Um, that's the Bureau of Meteorology's Space Weather Forecasting Centre.

22:02 That's the show today, a real sky in the starry night. Or at least a strong case for one Roman getting down to the fine detail, Rubin's test camera, finding a seagull and a weekend with Venus passing the sun. Tomorrow is the weekend wrap, a fresh lead the week that was and whatever NASA says in its Artemis Update. Anna: If you've got a question or a favourite painting with a real sky in it, send it through the contact [email protected] we read everyone. Avery: Show notes and sources are there too.

22:39 I'm, um, Avery. Anna: And I'm Anna. Clear Skies Astronomy Day Stories.