We’ve Seen the Sun’s Inner Corona For the First Time... and It's Not What We Expected

At a glance

The Sun's outer atmosphere reaches 10 million degrees Celsius while the surface just 1,500 km beneath it sits far cooler, and nobody knows why. That is the coronal heating problem, one of the oldest open questions in astrophysics, and the answer is thought to hide in the innermost slice of the corona, the exact region no instrument has ever been able to watch. Total eclipses show it for a few minutes every eighteen months or so, usually from somewhere inconvenient. Coronagraphs, the artificial eclipse machines invented in the 1930s, get closer but still cannot see the innermost band because the occulting disc sits too near the camera and light diffracts around its edge.

In December 2024 the European Space Agency launched a fix that sounds absurd on paper: put the occulting disc on a completely separate spacecraft and fly the two of them 144 metres apart, autonomously, to millimetre precision, through a wildly elliptical orbit of Earth. That is Proba-3, and in this Astrum video Alex McColgan walks through both halves of it: the engineering, which had never been attempted and worked on the first try, and the first science, which has already broken a model. Proba-3 can now see down to 70,000 km above the solar surface, 0.1 solar radii, against a previous floor of 0.7. What it found there is an inner corona far busier than anyone expected, full of plasma blobs and streams running at 14 to 520 km per second, three to four times faster than previous instruments measured in that region. It also spent a month lost, because the autonomous spacecraft could not survive a weekend on its own.

The problem: the atmosphere is hotter than the fire

The video opens on the corona itself, the Sun's outer atmosphere, and lands the paradox immediately. This fiery hell can reach temperatures of 10 million degrees Celsius. But just 1,500 km beneath that inferno, the surface is significantly cooler. Alex states the obvious objection plainly: this arguably makes no sense. Temperature is supposed to fall as you move away from a heat source, not climb. Surely that is the wrong way round.

This fiery hell can reach temperatures of 10 million degrees Celsius. But just 1,500 km beneath this inferno, the surface is significantly cooler. Alex McColgan, 0:05

It is not a curiosity. It has a name, the coronal heating problem, it is one of the greatest unanswered questions in astrophysics, and it has been baffling scientists for decades. Over the years theories and clues have accumulated about the cause, but as Alex puts it, it is hard to get answers about something you cannot easily see, other than the occasional fleeting glimpse during a solar eclipse. The inner corona, the part where the answer most likely lives, has therefore remained largely unobserved.

The fix nobody had tried: two spacecraft, one telescope

Then, in December 2024, that changed. ESA launched two tiny satellites that together are revolutionising our understanding of the Sun, working in perfect synchrony to make the once invisible visible. Alex frames the episode as two stories running in parallel: how these two minisats opened a window into the Sun's most inaccessible region, and the extraordinary engineering that makes flying in millimetre precision formation possible at all.

On the 5th of December 2024, ESA's Proba-3 lifted off on a PSLV-XL rocket from the Satish Dhawan Space Centre, Sriharikota, India. It did not garner headlines. It was quietly set to change the face of solar exploration and break a few engineering records on the way, because Proba-3 is not one satellite but two. This pair of fully autonomous minisats have to fly in millimetre perfect formation through a highly elliptic orbit of Earth, all while positioned, in Alex's unit of measure, one and a half football pitches apart. As he says, that is easier said than done.

This pair of fully autonomous minisats have to fly in millimeter-perfect formation through a highly elliptic orbit of Earth all while it's positioned one and a half football pitches apart. And that's easier said than done. Alex McColgan, 1:48

Why go to that trouble? Because of what sits at the other end of the telescope. Three questions have haunted solar physicists for decades, and the answers to all of them are believed to be hiding in the same uncharted region:

• Why is the Sun's atmosphere hotter than its surface?
• What drives the charged particles in the solar wind?
• How do billions of tons of magnetised plasma erupt from our star in one go?

Same address for all three. The corona.

Why the corona is so hard to see

Observing the outermost part of the Sun's atmosphere has always been a challenge for one brutally simple reason. The corona is a million times dimmer than the star's visible surface, the photosphere. Against that glare it simply gets lost.

It's a million times dimmer than our star's visible surface, the photosphere, meaning this region gets somewhat lost in the Sun's light. Alex McColgan, 2:44

Observations are not impossible, though. There is the eclipse. Total solar eclipses offer an excellent snapshot of the corona, and over the centuries they have delivered real science, not just spectacle. They led to the discovery of the corona itself, and to the detection of ionised iron within it, which was the first evidence of the corona's extreme temperatures. That single detection is why we know the number at the top of this page at all.

The problem with total eclipses is arithmetic. They last a few brief minutes. They are rare, occurring only every 18 months or so. And their paths of totality are often over remote areas or open ocean, which is deeply impractical when the instruments you want to point at the sky are high powered telescopes, and telescopes are not what you would call portable.

The problem with total eclipses, though, is that they only last for a few brief minutes and they're rare, only occurring every 18 months or so. Alex McColgan, 3:21

Bernard Lyot and the artificial Moon

There is an alternative, and it is nearly a century old. In the 1930s a French astronomer called Bernard Lyot started using an occulting disc to act like an artificial Moon, blocking out the Sun's surface to create a mini eclipse on demand. The effect on the science was immediate. Observing time went from minutes to hours.

In the 1930s, a French astronomer called Bernard Lyot started using an occulting disc to act like an artificial moon, blocking out the sun's surface to create a mini eclipse. Instantly, observing time went from minutes to hours. Alex McColgan, 3:50

These new instruments were called coronagraphs, and with them we have been able to keep constant tabs on the corona: tracking its temperature and composition, and watching for changes in the flow of the solar wind.

But, as Alex says, yes, there is a but. Ground based coronagraphs have to compete with two things Lyot could do nothing about. The first is daytime brightness. The second is atmospheric light scattering, the sky itself smearing sunlight across the field of view. Together they make the dimmest parts of the corona, which are exactly the parts closest to the Sun, almost impossible to see. So the inner corona remained mostly unseen and unstudied even after coronagraphs existed.

The logic of the next step follows in one line. To solve the mystery of the Sun's corona we needed a better view, and for that we needed to get rid of at least some of the background light. We cannot make the Sun dimmer. The easiest remaining target is the atmospheric scattering. So the coronagraph has to leave the atmosphere.

Coronagraphs in space, and the wall they hit

Space based coronagraphs have, as Alex puts it, revolutionised our coronal understanding, and the video gives the receipts.

On the 14th of December 1971 an early mission of this kind, the Orbiting Solar Observatory 7, saw a coronal mass ejection for the first time in history. Since then we have seen tens of thousands of them. The images most people picture when they picture the Sun come from the two workhorses that followed: the Solar and Heliospheric Observatory, SOHO, launched in 1995, and the Solar Terrestrial Relations Observatory, STEREO, launched in 2006. Alex admits a personal fondness here: he loves how you can watch solar weather events erupting from the surface when the star itself is blocked out, and how each craft specialises in a different region above the Sun, so that stitched together they produce one large image, each of them holding its own part of the puzzle.

And then, once again, the but. Even in space, the glare of the Sun can be too strong. SOHO in particular cannot see closer than 1.7 solar radii, and most of the interesting phenomena start much closer than that.

SOHO, in particular, cannot see closer than 1.7 solar radii, and most of the interesting phenomena start much closer than that. Alex McColgan, 6:37

Parker Solar Probe and Solar Orbiter, both flying far closer to the Sun, also carry coronagraph instruments, and they suffer the same problem plus two more of their own: internal light scattering, and diffraction that degrades the image when the occulter, the bit doing the blocking, sits too close to the camera lens. That last one is not a software issue or a manufacturing tolerance. It is geometry. On a probe three metres tall, the occulter is at most three metres from the lens, and there is no way to design around that. Scientists needed yet another way to look at the Sun.

An aside from the sponsor

The video pauses here for a sponsor read, and it is tied to the subject rather than bolted on. A total solar eclipse is coming to parts of Europe on the 12th of August, with parts of the northern United States getting a partial. Alex notes you probably cannot fit a space telescope in your pocket, but Dwarf Lab sells a more portable route: the Dwarf Mini, a compact app controlled telescope about the size of a light novel. With the official solar filters correctly installed, its solar mode and automated tracking find and hold the Sun centred, and photo, burst, and video time lapse modes let you record the event several ways. Its live stacking feature combines multiple frames into one crisp image, which makes it a beginner friendly way to shoot the Moon, galaxies, and other deep space objects the rest of the year. Two warnings he does not skip: practise with it before eclipse day, and use the Dwarf Lab solar filters plus certified eye protection, because looking at the Sun directly will damage your eyes. The offer named in the video was 10% off the solar eclipse travel kit through the 31st of July, with code ASTRUM5 at the Dwarf Lab store after that.

Creating an eclipse: move the occulter to another spacecraft

Back to the problem. If a three metre space telescope is not getting the job done, and the reason is that the occulter cannot get far enough from the lens, the solution writes itself. Put the occulter further away. Specifically, put it on another spacecraft.

The solution? To put the occulter further away, specifically on another spacecraft. Alex McColgan, 10:01

To do that, engineers had to create the world's first pair of fully autonomous satellites that can simulate artificial solar eclipses on demand, and, most crucially, do it without other solar radiation, so called parasitic light, leaking into the image.

Proba-3 is the fourth in ESA's Project for Onboard Autonomy, a line of experimental missions designed to work without constant human monitoring from the ground, and this is the most ambitious of them. Two small satellites, one carrying the coronagraph and the other the occulter, fly 144 metres apart, holding their relative positions with millimetric precision so the pair behaves as a single giant spacecraft. That distance is the whole point: it is what gets around the problem every other space coronagraph has with its two halves being too close together. And they are meant to hold it autonomously.

Two small satellites, one containing the coronagraph and the other the occulter, fly 144 m apart, holding their relative positions with millimetric precision to simulate a single giant spacecraft. Alex McColgan, 10:53

Alex concedes this might not sound overly impressive, then insists that it is, and that it has never been done before. To prove the point, he gets into the nitty gritty of how the mission actually works, and this is where the video becomes an engineering documentary.

The orbit: a roller coaster loop with a five and a half hour window

First up, the satellites take a highly elliptical path around Earth, and the reason is propellant. In a standard low Earth orbit the constant need for corrective thruster firings would have burned the propellant in as little as half an hour, making for a very short mission. So an elongated ellipse was chosen instead: 600 km above Earth at its nearest point, out to 60,500 km at its farthest, once every 19 hours and 36 minutes.

Alex's analogy is a roller coaster loop. At the bottom of the loop, near Earth, the satellites are moving fastest. As they climb around the loop they slow from 10 km per second to 1 km per second. Because they are moving so much slower up there, they spend disproportionately more time at the top, at the farthest point of the orbit, the apogee. That is not a side effect. It is the design. Slow means long, and long means observing time.

Near Earth the pair are set for a safe flyby, deliberately not trying to hold formation. Then, as they approach apogee, a signal puts them into active formation, and forming up takes about two hours of preparation before the corona can be observed. The payoff is a window of as much as five and a half hours of artificial eclipse out of every nearly twenty hour orbit.

Millimetre precision: a handoff chain from constellations to one millimetre

Then, like self driving cars, only way more precise, the satellites align themselves. They do it with a suite of absolute and relative positioning technologies: GPS receivers, radio links, optical cameras and LEDs, a laser link, and even shadow position sensors. The elegant part, and the part Alex emphasises, is that each of these hands off to the next with increasing precision. It is a ladder, not a single sensor.

Then, like self-driving cars, only way more precise, the satellites align themselves using a suite of absolute and relative positioning technologies. Alex McColgan, 13:11

It starts with cameras that recognise the constellations around them, called star trackers. These let each spacecraft know exactly which direction it is pointing at any given moment. During the lower portion of the orbit, below 20,200 km, the spacecraft use GPS signals and continuously exchange ranging information and data over inter satellite radio links.

Then comes the vision based system, which is about the two craft relative to each other rather than either of them relative to the universe. A wide angle camera on each spacecraft tracks a pattern of flashing LEDs on the other, giving a rough first read on distance and orientation. A narrower camera then locks onto a smaller LED target and refines that to around 1 cm of accuracy.

But even a centimetre is not precise enough. So the occulter spacecraft fires a laser at a reflector on the coronagraph, which bounces it straight back. That system, the fine lateral and longitudinal sensor, narrows positioning down to a single millimetre. And finally, to make sure the shadow is falling exactly where it needs to, photodetectors monitor its edges in real time. If it drifts even slightly, a correction fires instantly.

One more design decision underlines how allergic this mission is to vibration. To keep the whole system as stable as possible, neither spacecraft has any moving parts at all, apart from a single rotating filter wheel on the coronagraph. Everything else is locked in place.

With the two spacecraft aligned perfectly, a precisely controlled shadow is cast from one satellite onto the other, blocking our star's fiery disk and creating an artificial eclipse that allows sustained observation of the corona. Usually, from Earth, the corona can only be glimpsed naturally for a matter of minutes during a total solar eclipse. The idea with Proba-3 was to reproduce that same effect for hours at a time.

Inner corona: revealed

For the first time ever, these tiny satellites opened the chance to observe the corona on a consistent basis. But it was all uncharted territory. Even the slightest slip in the satellites' formation and the eclipse is gone, and with it the opportunity to study the corona. As the team turned the spacecraft on for calibration, no one really knew if it would all work.

The specific fear was the one that had haunted every coronagraph before it. Just as stray light can ruin images on Earth, it could here too. The team genuinely were not sure whether light would diffract and leak around the edges of the occulter, which would have made the whole 144 metre gambit pointless.

They had nothing to worry about. As the first data came through on the 25th of March 2025, it was clear the Proba-3 team had achieved the millimetre precision needed. Andrei Zhukov, the coronagraph's principal investigator, gave the video its most human line.

It was so unbelievable that it just worked from the first time. Andrei Zhukov, principal investigator for ASPIICS, 16:40

Then the number that justifies the mission:

Before Proba-3, we could see as close as 0.7 solar radii from the sun's surface. Now, we can see down to 70,000 km above the sun's surface, which is just 0.1 solar radii. Andrei Zhukov, 16:44

Put simply, no other space based coronagraph can observe the light scattering off particles in the Sun's corona this close to the Sun. Proof of concept complete, it was time to move on to the science.

The two instruments

The mission carries two scientific instruments.

The first is the coronagraph, ASPIICS, the Association of Spacecraft for Polarimetric and Imaging Investigation of the Corona of the Sun. It images both the quiescent and the eruptive corona, which is to say both when the Sun is calm and during specific solar weather events.

The second is DARA, the digital absolute radiometer, designed to measure total solar irradiance, the total energy output of the Sun arriving at a given area. It is the less glamorous of the two and it is the one that ties Proba-3 into the long term record of how much the Sun actually puts out.

While calibration was ongoing the team kept results under wraps. Then, on the 16th of June 2025, images from the first artificial solar eclipse were released. They were spectacular to look at, and they hinted at the valuable data still to come. By December 2025, a year into the mission, the satellites had created more than 50 artificial eclipses and provided hundreds of hours of invaluable observational data about the Sun's corona.

So what has it found?

The stakes are worth restating before the results. The corona is where the solar wind picks up speed before streaming outward across the entire solar system, eventually washing over the Proba-3 spacecraft themselves and reaching Earth. It is a continuous stream of charged particles flowing outward from the Sun in all directions, filling the whole solar system. The corona is also where most coronal mass ejections are born, massive eruptions of magnetised plasma that, aimed at Earth, can knock out satellites, disrupt GPS, and overwhelm power grids. This is not abstract astrophysics. It is the physics behind space weather forecasting.

The composite image. The video's best single demonstration is a composite built from three sets of images taken over an hour and a half by instruments on different missions, all watching the same coronal mass ejection on the 16th of July 2025. Three layers, three spacecraft:

• Yellow at the centre: the Sun's disk and lower atmosphere, captured in ultraviolet by an extreme ultraviolet telescope aboard Proba-2, a previous ESA solar mission.
• Red on the outside: the outer corona, as seen by the LASCO C2 coronagraph on SOHO.
• Green in between: the inner corona, imaged for the first time with this kind of clarity and continuity by Proba-3's ASPIICS, filling in the gap that had existed between those two instruments for decades.

The result is that you can watch the CME forming at the Sun's edge, expanding through the inner corona, and pushing outward into the outer atmosphere in one continuous, unbroken view. Not three separate movies stitched at the seams, but one event followed the whole way out.

The first science paper. In April 2026 the first science from Proba-3 was published in the Astrophysical Journal Letters, and it broke something. It revealed that solar wind structures in the inner corona can travel up to four times faster than previously thought.

The supporting imagery shows the Sun in ultraviolet, artificially coloured yellow, captured by Proba-2, surrounded by a grayscale region captured in visible light by Proba-3. You can see the solar wind moving away from the Sun in all directions, but in some regions, such as around the bottom of the frame, material also falls back towards the Sun. In the second half of the sequence a coronal mass ejection becomes visible, expanding out towards the right.

What we have learned, Alex summarises, is that the inner corona is far busier and more turbulent than previously expected. It is filled with tiny, constantly moving blobs and streams of plasma, some flowing outward and some actually flowing back toward the Sun. Those structures move at speeds ranging from 14 to 520 km per second, three to four times faster than previous instruments had measured in this region.

This means our existing models of how solar wind accelerates have been significantly off. Alex McColgan, 19:31

Understanding these small scale dynamics is the key to all three of the original questions: how the slow solar wind forms, how the corona gets so hot, and how solar eruptions are triggered. All three feed directly into our ability to forecast space weather, which is the practical reason anyone funds this.

Contact lost: the month the coronagraph vanished

Then the video turns and admits the mission has not been smooth, which is the section that makes it worth watching rather than reading a press release. Alex flags it directly: it sounds as if everything has gone well, but as ever, there have been hiccups.

Earlier in 2026, the Proba-3 team lost the coronagraph spacecraft. It vanished. Uncontactable for a whole month.

Zhukov explains the failure chain himself, and it is the mundane kind that kills spacecraft:

Well, in Andrei's own words, there was a software glitch, and it lost the orientation towards the sun. And well, when the solar panel is not looking towards the sun, then the battery is not charging. And if the battery is not charging, then the power is down, and the spacecraft is in trouble. Andrei Zhukov, relayed by Alex McColgan, 20:10

Software glitch, lost Sun pointing, solar panel facing the wrong way, battery not charging, power down. Facing away from the Sun, the coronagraph could not even enter safe mode to preserve its battery. And it happened at a weekend, so the operators were not working and it was not corrected in time.

Alex does not let that slide, and the observation is sharp: this might sound surprising, but remember, Proba-3 is meant to be autonomous. It should be able to survive a weekend unsupervised. Apparently not.

The recovery is the neatest detail in the whole episode. The team found the missing half by using the camera on the occulting satellite, the one instrument in the universe purpose built to look at the other spacecraft, and once they had it they commanded it to turn back towards the Sun. At worst the two had drifted 80 km apart, which for a pair designed to fly less than 150 metres from one another is an enormous separation, roughly 550 times their intended gap.

Both satellites are now online and back together. The team turned both spacecraft back on and it appears there has been no lasting damage. As of early June they are back on their observing schedule.

Perhaps we're a little way off fully autonomous after all. Alex McColgan, 21:10

The future of Proba-3

Even with a month long sabbatical, the numbers are already remarkable. Proba-3 has collected more than 250 hours of high resolution video of the corona over 57 artificial eclipses. Alex's way of putting that in perspective is the line the whole mission deserves: the average total eclipse on Earth lasts about 7.5 minutes, and Proba-3 has already captured 2,000 times that.

To put this into perspective, the average total eclipse on Earth lasts for about 7.5 minutes. Proba-3 has already captured 2,000 times that. Alex McColgan, 21:27

He is careful not to overclaim on the science. In that time it has not quite solved the three great coronal questions. But for the first time in the history of solar science there is an instrument dedicated to doing so, watching the solar corona with the consistency, resolution, and sensitivity needed to see what is actually happening. And we really are only at the start.

Zhukov and his team are already preparing the next batch of images and discoveries. In December 2026 Proba-3 will pass its nominal mission period of two years, and after that point the mission opens up: researchers from around the world will be able to request specific observations they want Proba-3 to make, handing the instrument to the greater global physics community.

Looking beyond solar science, Alex closes on the other legacy. Proba-3 is quite simply an engineering marvel. No one expected its precision formation flying to work as well as it does, and the mission has gone a long way to proving the experimental concept of autonomous satellites flying in precise unity. He says he is looking forward to seeing what future missions use this technology, and whether it can unlock any of the other big questions still lingering out in the cosmos.

The video signs off with two housekeeping notes: Astrum is live streaming the 12th of August solar eclipse at 6:30 pm BST, answering questions live, and the Astrum Patreon exists to make the channel less reliant on sponsors and algorithms, so the videos can stay independent, high quality, and consistent, created for curiosity rather than clicks.

Where it stands

The video is unusually honest about the limits, and it is worth collecting them in one place rather than leaving them scattered.

What is solid. The engineering is proven, not promised. Millimetre formation flying between two autonomous spacecraft 144 metres apart worked on the first attempt, with no stray light leaking around the occulter, and it has now produced 57 artificial eclipses and more than 250 hours of high resolution video. The inner edge moved from 1.7 to 1.1 solar radii from centre. That band of the corona genuinely had no dedicated instrument before, and now it does.

What is new but early. The 14 to 520 km per second result is one paper, published in April 2026, out of a mission barely past its first year. Alex says more publications are already on the way, which is another way of saying the picture will shift. The claim that existing solar wind acceleration models are significantly off is a strong one, and the sensible expectation is that it gets refined rather than reversed.

What is not solved. None of the three great coronal questions have been answered. Why the corona is hotter than the surface, what drives the solar wind, and how CMEs are triggered are all still open. What changed is that there is now an instrument watching the right region continuously instead of a few minutes of eclipse every eighteen months.

What is a real risk. The month long blackout is the honest counterweight to the engineering triumph. A mission whose entire premise is autonomy lost half of itself to a software glitch and a weekend without staffed operators, and drifted 80 km out of a 144 metre formation before anyone noticed. Nothing was permanently damaged, but the failure mode is exactly the one the Project for Onboard Autonomy line exists to eliminate. Alex's dry closing on it, that we are perhaps a little way off fully autonomous after all, is the fair verdict.

Key takeaways

• The coronal heating problem is the core puzzle: the corona reaches 10 million degrees Celsius while the surface 1,500 km below is far cooler, and the answer is thought to lie in the inner corona, which is a million times dimmer than the surface glare in front of it.
• Eclipses are too short and too rare. A few minutes every 18 months or so, usually over ocean or remote terrain, which is useless for instruments that cannot travel.
• Bernard Lyot solved the availability problem in the 1930s with the occulting disc, turning minutes into hours, but ground coronagraphs still lose the innermost corona to sky brightness and atmospheric scattering.
• Space coronagraphs hit a geometry wall. SOHO cannot see closer than 1.7 solar radii, and Parker Solar Probe and Solar Orbiter share the problem because on a three metre probe the occulter is inevitably close enough to the lens for diffraction to wreck the image.
• Proba-3's answer is to put the occulter on a second spacecraft, 144 metres away, flown autonomously to millimetric precision. It is the fourth in ESA's Project for Onboard Autonomy line and the most ambitious.
• The orbit is the enabler. 600 km at perigee, 60,500 km at apogee, 19 hours 36 minutes per lap, slowing from 10 to 1 km per second so the pair linger at apogee. That buys up to 5.5 hours of artificial eclipse per orbit, after a 2 hour set up.
• Positioning is a handoff chain, not one sensor: star trackers, then GPS and inter satellite radio below 20,200 km, then a wide angle camera tracking flashing LEDs, then a narrow camera on a small LED target to 1 cm, then a laser bounced off a retroreflector to 1 mm, then photodetectors watching the shadow edge for instant correction. Neither craft has any moving parts except one filter wheel.
• It worked on the first try. First data on 25 March 2025, no light leaking around the occulter. The inner edge dropped from 0.7 to 0.1 solar radii above the surface, or 70,000 km.
• Two instruments: ASPIICS the coronagraph, imaging both the calm and the erupting corona, and DARA the digital absolute radiometer, measuring total solar irradiance.
• The first science broke a model. Inner corona plasma structures run at 14 to 520 km per second, three to four times faster than previous instruments measured there, with some material falling back towards the Sun. Existing solar wind acceleration models are significantly off.
• Autonomy is not there yet. A software glitch cost the coronagraph its Sun pointing, the battery drained over an unstaffed weekend, and the spacecraft was uncontactable for a month, drifting up to 80 km from its partner before the occulter's camera found it.
• The record so far: 57 artificial eclipses, 250+ hours of high resolution coronal video, which is roughly 2,000 times the length of an average total eclipse on Earth. In December 2026 the mission passes its nominal two years and opens to observation requests from researchers worldwide.

Chapters

Timestamps are clickable. Click one and the player jumps there and keeps playing while you read.

• 0:00 Proba-3
• 2:36 The Corona
• 5:13 Coronagraphs in Space
• 9:47 Creating an Eclipse
• 13:11 Millimetre Precision
• 16:04 Inner Corona: Revealed
• 19:44 Contact Lost
• 21:15 The Future of Proba-3

Notable quotes

This arguably makes no sense. How can the temperature increase as it moves away from the source? Surely that's the wrong way round. Alex McColgan, 0:12

Known as the coronal heating problem, this is one of the greatest unanswered questions in astrophysics, and it's been baffling scientists for decades. Alex McColgan, 0:23

It's hard to get answers about something you can't easily see, other than the occasional fleeting glimpse during solar eclipses. Alex McColgan, 0:35

This mission didn't necessarily garner the headlines, but it was quietly set to change the face of solar exploration and break a few engineering records in the process because Proba-3 isn't just one satellite, but two. Alex McColgan, 1:20

It's a million times dimmer than our star's visible surface, the photosphere, meaning this region gets somewhat lost in the Sun's light. Alex McColgan, 2:44

In the 1930s, a French astronomer called Bernard Lyot started using an occulting disc to act like an artificial moon, blocking out the sun's surface to create a mini eclipse. Instantly, observing time went from minutes to hours. Alex McColgan, 3:50

SOHO, in particular, cannot see closer than 1.7 solar radii, and most of the interesting phenomena start much closer than that. Alex McColgan, 6:37

The solution? To put the occulter further away, specifically on another spacecraft. Alex McColgan, 10:01

Two small satellites, one containing the coronagraph and the other the occulter, fly 144 m apart, holding their relative positions with millimetric precision to simulate a single giant spacecraft. Alex McColgan, 10:53

Then, like self-driving cars, only way more precise, the satellites align themselves using a suite of absolute and relative positioning technologies, from GPS receivers and radio links to optical cameras and LEDs, a laser link, and even shadow position sensors. Alex McColgan, 13:11

But even a centimeter isn't precise enough. Alex McColgan, 14:37

It was so unbelievable that it just worked from the first time. Andrei Zhukov, ASPIICS principal investigator, 16:40

Before Proba-3, we could see as close as 0.7 solar radii from the sun's surface. Now, we can see down to 70,000 km above the sun's surface, which is just 0.1 solar radii. Andrei Zhukov, 16:44

What we've learned is that the inner corona is far busier and more turbulent than previously expected. It's filled with tiny, constantly moving blobs and streams of plasma, some flowing outward and some actually flowing back toward the sun. Alex McColgan, 19:12

This means our existing models of how solar wind accelerates have been significantly off. Alex McColgan, 19:31

There was a software glitch, and it lost the orientation towards the sun. And well, when the solar panel is not looking towards the sun, then the battery is not charging. And if the battery is not charging, then the power is down, and the spacecraft is in trouble. Andrei Zhukov, 20:10

This might sound surprising, but remember, Proba-3 is meant to be autonomous, so it should be able to survive a weekend unsupervised. Apparently not. Alex McColgan, 20:25

Perhaps we're a little way off fully autonomous after all. Alex McColgan, 21:10

To put this into perspective, the average total eclipse on Earth lasts for about 7.5 minutes. Proba-3 has already captured 2,000 times that. Alex McColgan, 21:27

It hasn't quite solved the three great coronal questions, but for the first time in the history of solar science, there is an instrument dedicated to doing so. Alex McColgan, 21:36

No one expected its precision formation flying to work as well as it does, and the mission has gone a long way to proving the experimental concept of autonomous satellites flying in precise unity. Alex McColgan, 22:21

Resources mentioned

The video and the channel

• Astrum, Alex McColgan's space documentary channel, and this video, "We've Seen the Sun's Inner Corona For the First Time... and It's Not What We Expected".
• The Astrum Patreon, pitched in the outro as the way to keep the channel independent of sponsors and algorithms.

The mission

• Proba-3, ESA's two spacecraft solar coronagraph, fourth in the Project for Onboard Autonomy line. Launched 5 December 2024.
• ASPIICS, the Association of Spacecraft for Polarimetric and Imaging Investigation of the Corona of the Sun, the coronagraph instrument. Principal investigator Andrei Zhukov of the Royal Observatory of Belgium.
• DARA, the digital absolute radiometer, measuring total solar irradiance.
• The European Space Agency, which built and operates the mission.
• The PSLV-XL launch vehicle, flown by ISRO from the Satish Dhawan Space Centre, Sriharikota, India.

Earlier and companion instruments

• Orbiting Solar Observatory 7, which recorded the first coronal mass ejection on 14 December 1971.
• The Solar and Heliospheric Observatory (SOHO), launched 1995, and its LASCO coronagraphs, whose C2 supplied the outer corona layer of the composite CME image.
• The Solar Terrestrial Relations Observatory (STEREO), launched 2006.
• Parker Solar Probe and Solar Orbiter, both carrying coronagraphs that hit the occulter distance limit.
• Proba-2, whose extreme ultraviolet telescope supplied the yellow central layer of the composite.

Science and background

• The coronal heating problem, the puzzle the mission was built to attack.
• Bernard Lyot, inventor of the coronagraph in the 1930s.
• Coronium, the story of the ionised iron detected during eclipses that first revealed the corona's extreme temperatures.
• The solar wind and coronal mass ejections, and the NOAA Space Weather Prediction Center that forecasts their effects on satellites, GPS, and power grids.
• The Astrophysical Journal Letters, where the first Proba-3 science appeared in April 2026.
• NASA's eclipse resources for the 12 August total solar eclipse over parts of Europe, which Astrum is live streaming at 6:30 pm BST.

Sponsor

• Dwarf Lab and the Dwarf Mini, the pocketable app controlled telescope with solar mode, automated tracking, and live stacking. Code ASTRUM5, and use the official solar filters plus certified eye protection.

The one idea to walk away with

Every advance in looking at the corona has been the same trick executed at a longer range. Lyot put a disc inside the telescope. Space missions carried the disc above the atmosphere. Both stalled at the same wall, because a disc a few metres from the lens diffracts light around its own edge and floods the very region you are trying to see. Proba-3's answer was not a better disc or a better sensor. It was to take the disc off the telescope entirely, fly it 144 metres away on its own spacecraft, and hold that gap to a millimetre with a laser. The result is that a band of the Sun's atmosphere that no instrument had ever properly watched turns out to be far more violent than the models assumed, and the mission that revealed it also spent a month unable to find half of itself. Both facts are the same story: this is genuinely new territory, and we are only at the start of it.