Bad Astronomy
The entire universe in blog form

July 11 2016 9:30 AM

An Alien Planet Orbits in a Triple-Star System … and We Have Photos

Astronomers have discovered a truly remarkable planet. Called HD 131399Ab,* it’s in a triple-star system: It orbits a star orbited by another binary pair of stars!

I know, I know: Pics or it didn’t happen, right?

Yeah, well, here you go:

photos of planet
Actual, direct images of the exoplanet HD 131399Ab. See text for description.

Wagner, et al., from the paper

Yes, those are actual images of the planet! The frames A-D show the planet (labeled “b”) at different infrared wavelengths, with the primary star’s position marked by a crosshair (the star’s light has been removed using various observational and processing techniques to better see the planet), and the larger panel E is a composite showing the star, the planet, and the binary.

There’s a lot of very cool stuff going on here, so let me explain.

As a whole, the star system is called HD 131399, and it’s what’s called a hierarchical triple: Two stars orbiting each other in a binary pair, which in turn orbits another star. The primary, most massive star is called HD 131399A, and is hotter and more massive (about 1.8 times more) than the Sun. The binary is composed of a star very much like the Sun and another star that’s cooler, redder, and less massive (0.6 times the Sun’s mass). The binary is pretty far out, orbiting the primary at a distance of about 40–60 billion kilometers. That’s about 10 times Pluto’s distance from the Sun, to give you a sense of scale.

The system is part of a loose cluster of stars called an association located roughly 300 light-years from Earth. This is important: We know, from studying those stars, that the association is young, probably around 16 million years old. Why is that important? Because when planets first form they are very hot, and it takes a long time for them to cool. The more massive a planet is, the longer it takes.

The light given off by a hot planet depends on its temperature, and as I said that depends on its mass and age. We know the age, so by examining the light from the planet, its mass can be derived. In this case, the astronomers found the planet has a mass of about four times that of Jupiter! While that’s big, it’s very firmly in the planetary mass range (even if the planet is older, and therefore more massive, it’s still very likely to be a planet and not a low-mass star or brown dwarf).

solar system to scale
Our own solar system overlaid to scale on the HD131399 system. The new exoplanet orbits its star farther out than Pluto does the Sun. Note the colors have been added for clarity and are not necessarily real.

Wagner, et al.

So how do they know it’s actually orbiting the primary star, and not a background object? I love this part: The astronomers used archived images taken over many years to measure the actual motion (what we call the proper motion) of the system. All the stars in the sky are orbiting the center of the galaxy, but that motion is hard to measure directly because stars are so far away. But some stars are close enough to us that it can be detected (it’s like driving down a road; nearby trees seem to zip past, but a distant mountain appears to pass much more slowly).

By mapping the motion of the stars and the planet, they found that the planet appears to move along with the stars across the sky, showing pretty conclusively it’s actually a member of the system. However, the motion isn’t exactly the same, and that’s because the planet is orbiting the star, and that motion is seen on top of its velocity through space!

Although it’s hard to determine the exact orbital shape and distance, it’s likely the planet is about 12 billion km from the star, and takes about 550 years to orbit it once. Even though it’s more than twice as far from the star as Pluto is from the Sun, its temperature is 575°C (1070°F), still very hot from its formation.

This makes it unique among planets seen so far: It has the widest known orbit for an exoplanet in a triple system. In fact, its orbit is so wide that it’s quite possible the gravity from the binary yanks on it. Over time, the orbit may be unstable! The astronomers ran some simulations and showed it’s about as likely the planet’s on a stable orbit as not. It’s so young that we may be seeing it before it gets ejected from the system … or it may be fine and dandy where it is for the next few hundred million years. Here’s a video animation showing what all those orbits look like:

And there’s more: As far as we know, it’s not possible to form a big planet on such a wide orbit in situ. It’s far more likely it formed much closer in, but an interaction with another massive planet (one too close to the primary to be spotted yet) flung it out to its current distant position. But there’s another possibility, which I find completely fascinating: It may have formed as a planet around the binary stars, and got flung out (either by the stars themselves or another possible planet orbiting them) to its present orbit.

If that’s the case, the planet doesn’t orbit the star(s) it formed around! How weird is that?

Incidentally, very careful observations of the planet show that its atmosphere is likely to have methane and water vapor in it, common in gas giants. It’s amazing we can tell that at all from our distance of 3,000 trillion kilometers, but exoplanet astronomers are getting pretty dang good at what they do.

And one final thing: What does the sky look like from this planet? Or, say, any moons it might have, since the planet is a gas giant.


From more than 12 billion kilometers away, the primary star would only be about 1/500th as bright as the Sun is from Earth. That’s still far brighter than the full Moon in our sky, so it would be a tiny but very bright dot. The binary stars would be much fainter, even at closest approach (depending on the exact orbit of the exoplanets, which to be fair we don’t know), and would change in brightness noticeably over the centuries as the planet orbits its primary.

The distance separating the two stars in the binary isn’t explicitly mentioned in the paper, but judging from the images I’d guess they could be split by eye from the planet. But that would change as well; the two stars orbit each other, and depending on that orbit they might get close enough together to appear as one star over the years, then pull apart over time.

I wonder. If life ever arose in such a system, would they have an easier time than we did deriving the laws of motion in the cosmos? It would be hard to argue everything revolves around their planet when a counterexample sits for all to see in the sky!

It’s fun to speculate over such matters, but this system does show us something we no longer need to speculate about, after decades of exactly that: Stars are very good at making planets, and do so under an amazing array of conditions. Planets are everywhere, even where we used to think they couldn’t be.

I’ve said it before, but it bears repeating; The Universe is more clever than we are. I wouldn’t have it any other way.

* The name comes from the star being the 131,399th entry in the Henry Draper stellar catalog. Stars in multiple systems are designated by brightness using capital letters (A, B, C, and so on), while planets are designated using the name of the star they orbit appended with a lower case letter (starting with “b” for the first planet found, then “c” and so on in order of discovery).

July 10 2016 9:30 AM

Namibian Milky Way

Our planet is gifted with a lovely atmosphere of nitrogen and oxygen, which does a terrific job of keeping us alive.

It also provides us with a very common metaphor: “Different as day and night”. And they are; during the day, sunlight passing through our atmosphere gets scattered—blue more than any other obvious color—making the sky itself appear to be blue and very bright. Contrast that with the black sky at night, dark enough that we see stars. They’re there during the day, too, but the scattered sunlight is so strong and so bright that it washes out the stars.

So how can the photo above show what looks to be the Milky Way and thousands of stars over a daylit scene?

Because it doesn’t. Yes, that’s the Milky Way, and yes, those are myriad stars, but no, it’s not daylit. It’s moonlit!

The photo was taken by photographer Florian Breuer in Namibia on June 10. It was a few hours after sunset, with the crescent Moon low in the sky. He took four 30-second exposure photos and stitched them together to make this panorama; the phenomenal dark and clear southern Africa skies allowing the glow of the stars and our home galaxy through. But the low Moon was also bright enough that, in the long exposures, the landscape was illuminated, and the tree even cast a shadow!

This is not at all how your eye sees the scene were you there, but cameras work differently than our eyes. This is something I wish more people understood; when you take a photo of something you are automatically changing how it’s seen. Colors, brightnesses, perspective, and much more are recorded in ways that alter them. So while this image may look fake, it’s not. At least, it was photographed faithfully to the scene (with minor contrast adjustments), but it’s not how it would look to us standing there.

But nothing is. This is a very important lesson in life, one I iterate whenever I post an optical illusion or eyewitness account of some oddity, but it bears repeating here, too: What you see is almost never what you get.

But that doesn’t make shots like this any less real, or any less beautiful. Breuer took a lot of photos in Namibia, and trust me, you should take a look. Our world is packed with beauty, and sometimes it takes a camera’s—and photographer’s—eye to show it to us.

July 9 2016 9:00 AM

A Gathering Storm

I was sitting in my office at home writing a post for the blog yesterday when I glanced out the window. Although the Sun was shining on my yard, the sky itself was a dark, ominous gray filling the view. Afternoon storms in Colorado are common, but this one seemed different. So I went outside to look, andthis is what I saw:

July 8 2016 8:00 AM

We Have to Be Better Than This

I had an astronomy article lined up to post Friday, but after the events of the past two days it feels wrong to go about business as usual. Tomorrow, maybe, but not today.

As I write this, we still don't have a lot of information about the sniper attack on the otherwise peaceful Black Lives Matter protest in Dallas, and I refuse to speculate. I saw a lot of hate flowing on social media Thursday night, and a lot of speculation based on ignorance and fear. The speculation is natural, perhaps, but we have to be better than that. We have to be.

I plan on spending today listening, reading, and hoping. If the situation warrants I'll update this page.

July 7 2016 8:45 AM

Looking Down on the International Space Station. Literally.

Well, here’s something you don’t see every day … because you can’t see it every day. Or even any day. It’s the International Space Station moving across the cloudy Earth, as seen from above.

NASA Earth Observatory images by Jesse Allen, using Landsat data from the U.S. Geological Survey

That animation was created using images taken on June 19 by the Operational Land Imager on Landsat 8, an Earth-observing satellite, as they both passed over India. ISS orbits about 400 kilometers above the Earth, but Landsat 8 is 300 kilometers higher, so it looks “down” on the station. The orbits are tilted differently, too: The orbit of ISS is tipped by about 51° to the Earth’s equator while Landsat 8 is tipped 98° (it passes very close to directly over the poles, but because the orbital tilt is more than 90°, it technically moves backward relative to Earth, called a retrograde orbit).

Because of their different altitudes and tilts, the two objects move pretty quickly relative to each other, zipping by at several kilometers per second. And that’s where this gets interesting.

Did you notice the clouds look different in each frame of the animation? That’s because the OLI camera is set up to look at Earth in nine different “colors,” from blue out to the infrared. Each frame of the animation shows the scene in a different color, which is why they look different.

You might think OLI employs a bunch of filters to take the images, but it doesn’t. The setup is actually rather complicated, but it uses several detectors spread out in two lines, like two rows of seats in a theater, and each uses what’s called a pushbroom technique to build up the image row by row in each color. When all is said and done, each color image is created at a slightly different time, fractions of a second apart. That’s not enough time to see any actual cloud motion, but more than enough to see the very fast relative orbital motion of the space station.

A common question is, if the images are in different colors, why isn’t the final picture in color? That’s because each image is stored as a series of numbers, digitally, and separate from each other. So the blue image is stored in one data set, and the green in another. It’s only after you combine them (using, say, Photoshop, where you can create a blue layer for the blue image, a green layer for the green, and so on) that you actually get a color photo. If you want more info, take a look at this primer for how Hubble color images are made. It’s actually pretty interesting.

I found the animation on the NASA Earth Observatory website, which is fantastic (and it has lots more info about the animation, too). I check it every day, and you should too. Our planet is lovely, and all sorts of wonderful things can be seen happening on it from space. Taking a look may give you a better appreciation for this world we live on, and I can’t recommend that strongly enough.

One final thing: Landsat was launched in 2013, and a camera on board the booster took pretty amazing footage of it as it was deployed into orbit. Check it out:

July 6 2016 9:00 AM

I’m a Loner, Dottie. A Dwarf Galaxy.

Poor UGC 4879. Doomed to wander the Universe forever, alone.

Or is it? Dun dun dunnnn.

UGC 4879 is a dwarf galaxy, a small collection of stars, gas, and dust located about 4.5 million light years away from us. The Milky Way, our galaxy, is part of a small clutch of galaxies called the Local Group, which is a few million light years across. UGC 4879 is probably located just outside that group, far enough away to be on its own.

In fact, the nearest galaxy to UGC 4879 is another dinky thing called Leo A, and even that’s over two million light years distant from it. Those other galaxies you see in the image are much much farther away; hundreds of millions or even billion of light years past it. It really is pretty isolated.

July 5 2016 12:58 PM

Follow-Up: Jupiter’s Moons Dance as Juno Approached

Monday night, for the first time in nearly two decades, humanity added a new moon to Jupiter: The spacecraft Juno successfully entered orbit around the giant planet, beginning a nearly two-year-long mission to study Jupiter’s interior and determine its origin story.

You can read all about that in my post from early Tuesday morning. Too late for me to include it, though, NASA released a nifty video showing Juno’s view of Jupiter and its four large moons as it approached the system. It’s really quite entrancing:

It’s slow at first, but you can see the disk of Jupiter swell as Juno got closer, from 16 million kilometers distant on June 12 to only about 5 million kilometers on June 29. Juno approached “from the side and above,” if you will; in other words instead of heading straight out to Jupiter it was on a path that took it on a long, curving trajectory that had it sidling up to the huge world. That’s why Jupiter is half lit; the Sun is off to the right, nearly 800 million km away.

It also approached from above Jupiter’s orbital plane, so we’re looking down on the moons, and can see their nearly circular orbits as ellipses (like the lip of a drinking glass looking oval shaped when seen from an angle). Over the 19 days of the animation the moons move, and you can see which is which by how fast each moves: Io takes 1.8 days to circle Jupiter, Europa 3.6, Ganymede 7.2, and Callisto 16.7.

A few things to note in the animation that are cool: One is that you can see Io, Ganymede, and Europa disappear as they enter Jupiter’s shadow on the left. They seemingly flick out of existence briefly as they are eclipsed by the gigantic planet, which blocks the Sun’s light. This can sometimes be seen from Earth, but due to the geometry (Jupiter is farther out from the Sun than Earth) the eclipses appear to happen very near the disk of the planet itself, like we’re peeking around the edge. Because Juno approached Jupiter from the side, the eclipses are obvious and easily seen well away from the planet. Amazing.

You may have noticed that Callisto flickers and jumps a bit as it moves. This isn’t real. As Emily Lakdawalla explains, it has to do with the optics of the JunoCam that was used to take these images. It wasn’t designed to see very tiny point sources of light, so as the image of Callisto fell between pixels it appears to flicker.

Finally, as Jupiter grows, you can start to see features on it, like the light zones and dark belts so familiar to anyone who has peered at the planet through a telescope. Juno will practically skim the cloud tops of Jupiter over its mission, taking incredible images of the dynamic and chaotic cloud structures. That will be breathtaking.

We won’t see those for a while though, since the spacecraft will be doing a thorough shakedown over the next few weeks. But stay tuned. We’re about to see Jupiter in a whole new way, and, hopefully, learn how it got to be what it is today.

July 5 2016 1:53 AM

Jupiter Has a New Moon. And We Put It There.

At 03:53 UTC on Tuesday (or 11:53 p.m. Eastern time on Monday), Jupiter got a new moon. And its name is Juno.

At that time, the main engine of the spacecraft cut off, having burned for 35 minutes and two seconds. When it did, Juno was on a looping, highly elliptical 53.5-day orbit, the first spacecraft to orbit Jupiter since Galileo did in 1995.

The engine burn was tense. Thirty-five minutes is a long time for a spacecraft burn; after 20 minutes it had slowed Juno enough to be in orbit, but not the correct one. It had to continue for another 15 minutes to put the spacecraft on the correct orbit. It worked essentially perfectly. The burn time was off by just one second. That will have no real effect on the orbit.

It was fun to see the folks at the control room(s) in jubilation when the signal came back! A lot of my friends on Twitter erupted in applause as well. Even Google got in on the celebrations with a new doodle:

Juno Google doodle
Google made this doodle celebrating Juno's arrival at Jupiter.


But Juno wasn’t out the woods just yet. After Jupiter orbit insertion, it had to turn to face the Sun so it could power up its trio of huge 10-meter long solar panels, aim its antenna at Earth, and begin sending telemetry (the recorded data onboard). It was nearly an hour later before word was given that Juno had indeed made the maneuver.

We now have a working spacecraft orbiting the mightiest planet in the solar system.

There’s still much to do. The science instruments need to be turned on, checked out, and begin taking their data. Everything appears to be healthy on board, which is great; the radiation environment around Jupiter is nothing less than terrifying, and it’s good that we’re starting from a functioning spacecraft.

And it’s still not done with this first phase. The current 54-day orbit is just the initial orbit; on Oct. 19 the engine will burn again, changing the shape of Juno’s trajectory so that it enters a 14-day orbit. At that point the real science will begin.

Juno will study Jupiter’s internal composition, revealing critical clues on how it formed. Jupiter contains most of the solar system’s water, and the amount it has will tell us how it got there, almost certainly by impacts of rock and ice as Jupiter formed. The science is in the details, though, and to understand where those original rock/ice planetesimals came from we need to know better the water inside Jupiter.

We also aren’t sure how Jupiter formed. Did it go from the bottom up, as several large objects that collided and merged, growing hugely? Or did it start from the top down, forming directly from the disk of material orbiting the young Sun 4.56 billion years ago? These two models predict different internal structures for Jupiter (as I discuss in Crash Course Astronomy: Jupiter, the latter predicts Jupiter may not even have a rocky core), and Juno should be able to investigate Jupiter well enough to distinguish between the two.

And that’s only part of what it will do at Jupiter; much more will occur as well.

Juno traveled 2.8 billion kilometers on its looping road to Jupiter, and is now 870 million kilometers from Earth—even light itself, the fastest thing in the Universe, takes 48 minutes to get to Earth. Juno took five years to get there, and its mission will last at least until early 2018.

Make no mistake: This is an incredible achievement. Staggering. We humans watched the skies for millennia, asked ourselves questions about it, and when we faced those question honestly and openly we learned how to answer them using mathematics and physics.

Eventually our engineering allowed us to observe those celestial objects better, and then we learned how to go to them. Because we were curious and because we were brave we currently have spacecraft orbiting a great number of objects in our solar system.

And that now includes one more. Congratulations to everyone on the Juno team. Very, very well done.

July 4 2016 9:00 AM

Happy Fourth of July!

Uncle Sam Goat says have a fun and safe Fourth of July!

But ... as Matt Inman of the Oatmeal points out, the fifth of July is one of the busiest nights of the year for animal shelters because lots of pets go missing on the fourth, terrified by fireworks. My own dogs have had a rough week due to neighbors' fireworks. Like Matt says, give them extra belly rubs and treats, because they have no idea humans are enjoying themselves and have quite a different view of the evening.

I'll be checking in on my ungulate family members tonight. Cud-chewers or not, keep your own in mind.

Edited to add: Yes, that is my actual goat, and his name is actually Sam.

July 3 2016 9:00 AM

A Cosmic Tadpole Swims Into Hubble’s View

Things change. Even galaxies.

They started off as monumentally huge clouds of gas, which then collapsed. A lot of things happened at that point, including star birth, interactions with other galaxies, the formation of a giant black hole in the center, perhaps the triggering of events that lead to the creation of fantastic spiral arms, and more.

Eventually, you get galaxies as we see them today: some big, some small, some spiral, some elliptical, some with no real shape at all.

But some are still fond of the olden days, and haven’t changed much since then. These few small galaxies are either much as they were 10 billion years ago, or perhaps they formed more recently and are just now going through some of those early stages all the other galaxies did long ago.

One such galaxy is Kiso 5639, a dwarf irregular galaxy roughly 80–90 million light-years from our own. It’s probably a flat pancakes shape, but we see it edge-on, so it looks elongated. One end—the “head”—is much brighter than the other, making it look like a tadpole or comet.

It’s small, only about 10,000 light-years long. Our Milky Way is fully 10 times wider! So Kiso 5639 really is diminutive.

Tadpole galaxies like it are pretty rare in the Universe today, making up less than 1 percent of all galaxies. But in the early Universe they were much more common, more like 10 percent of all galaxies. We can actually see these galaxies; they’re so far away their light took more than 10 billion years to reach us, so we see them as they were when the cosmos was young.

Why do we see so few now? Probably because this tadpole shape is only a stage galaxies go through as they age, and it doesn’t last long before collisions and acts of cannibalism cause galaxies to grow.

Well, most galaxies. We do see ones like Kiso 5639 today. These are likely the remnants left over that never got close enough to other galaxies to be affected by them, so their evolution was stunted.

Kiso 5639
The same image, but also using filters that accentuate hydrogen gas where stars form (red). Clearly, there's a lot going on in that head.

NASA, ESA, and D. Elmegreen (Vassar College), B. Elmegreen (IBM's Thomas J. Watson Research Center), J. Sánchez Almeida, C. Munoz-Tunon, and M. Filho (Instituto de Astrofísica de Canarias), J. Mendez-Abreu (University of St. Andrews), J. Gallagher (University of Wisconsin-Madison), M. Rafelski (NASAGoddard Space Flight Center), and D. Ceverino (Center for Astronomy at Heidelberg University)

Kiso 5639 is asymmetric, as you can see. The head (which is roughly 2,500 light-years across) has quite a bit of gas, and stars are actively forming there. The astronomers who took this image examined the detailed colors of the stars using Hubble Space Telescope and found the stars in the head are less than 1 million years old while in the rest of the galaxy stars range from millions to more than 1 billion years old.

Not only that, but the head has a higher ratio of hydrogen and helium than the gas in the rest of the galaxy (or, if you want to think of it the other way, the tail of the galaxy has a higher abundance of heavier elements than the head).* Heavy elements are created in stars as they are born and age, and they scatter them back into their host galaxy when they die. Gas in older galaxies therefore tends to have a high abundance of these heavier elements like oxygen and iron.

But gas between galaxies has no stars, and therefore has gone unenriched for eons. It’s still mostly hydrogen and helium, and it’s thought that tadpole galaxies move through space head first, collecting this primitive intergalactic gas and getting it piled up there. This is triggering the star formation in the head, and also explains why the elemental abundance is so different there than in the tail.

Galaxies like this are wonderful to study; they’re like a relic of the past, a glimpse into an adolescent stage of growing up that we otherwise wouldn’t be able to study in detail. Seeing Kiso 5639 just 80 million light-years away, and not 12 billion, makes it a ripe target for study. And if our own galaxy went through such a stage—which it very likely did—then it’s like a time machine into our own past.

*Correction, July 3, 2016: I originally misstated that the tail has a lower abundance of heavy elements. I had a brain cloud there; I meant "higher."