A Disk Around a Young Star
November 6, 2014 9:05 PM   Subscribe

The resolution of a telescope (how much fine detail it can see) is proportional to the wavelength of light divided by the size of the telescope. Since radio wavelengths are many times larger than optical wavelengths, radio telescopes like the GBT and Arecibo are large compared to optical telescopes. Even so, the resolution of radio telescopes is worse, and you rarely see radio images as beautiful as those produced by the Hubble Space Telscope. That has now changed.

Through the magic of interferometery [warning; technical PDF], many small telescopes can be combined to act like a single large one. For the past ten years, the worldwide astronomical community has been building the Atacama Large Millimeter Array (ALMA) in the high desert of northern Chile. The array is almost complete, and in September, ALMA began observing with individual antennas separated by 15 kilometers (in the future, the maximum separation will be 16 kilometers).

What can you do with a telescope that is effectively 15 kilometers across? You can take a picture of the protoplanetary disk around a young star, and see the ridges and gaps in the disk. "Young star" is relative of course; HL Tau is only 1 million years old. But since our sun is about 4.6 billion years old, HL Tau is less than one one-thousandth the age of the sun. Perhaps the biggest surprise is that it is probably young planets that are producing the gaps in the disk; theories have a hard time producing planets so quickly. The science image is impressive enough that you don't really need the artist's impression, but the movie showing the image in context is well worth it.
posted by sedna17 (31 comments total) 44 users marked this as a favorite
 
I see no sign of Bode's Law in that system, which is interesting.
posted by Chocolate Pickle at 9:09 PM on November 6, 2014 [2 favorites]


I was just composing my very first FPP for this very image and you beat me to it.

However, here's my bonus material from that now-stillborn FPP: It will likely be a while before there's anything to colonise in these systems, but RAND Corp have made available for free download Habitable Planets for Man, a seminal popular work on the search for habitable planets. Habitable Planets For Man, by Stephen H. Dole and Planets for Man, by Stephen H. Dole and Isaac Asimov. I remember my excitement when I found the Dole/Asimov edition on the dusty shelves of my local library back in the 60's.
posted by Autumn Leaf at 9:14 PM on November 6, 2014 [11 favorites]


the movie showing the image in context is well worth it.

That movie is like the very thing I dreamed of seeing when I was a kid: to look out at a vast, starry night, choose a star, and just go there and see it. It's the real thing that the thrill I get doing a deep-dive with a real-time fractal zoomer is hinting at. Thank you.
posted by straight at 9:37 PM on November 6, 2014 [1 favorite]


I love this. And I hate this. Thanks for the links sedna17

I took astrophysics 101 two and one half times because fuck yes.

The orbit of the earth provides 186000000 miles of paralax, so there is that. VLA antennae are useful for spotting objects. They are binoculars.

http://en.wikipedia.org/wiki/Astronomical_spectroscopy is a lot of fun and involves some serious imagination.
posted by vapidave at 9:44 PM on November 6, 2014


Eponysterical?
posted by Joe in Australia at 9:45 PM on November 6, 2014 [1 favorite]


Freakin' amazing! (Said in Carol/Cheryl's voice from Archer.)
posted by kalessin at 11:13 PM on November 6, 2014


Interesting. this implies that planets begin forming at the same time the star does. It might even be the case that instead of single protostars forming in molecular clouds, various eddies may form, begin orbiting larger eddies, and then collapse at relatively the same time. That is, we've often described our Jovian planets as failed stars- maybe they're as old as our sun?
posted by happyroach at 12:25 AM on November 7, 2014 [1 favorite]


And yet everyone's going about their daily lives as normal. How can this fucking be?
posted by popcassady at 1:51 AM on November 7, 2014 [5 favorites]


In a similar vein -

In the upcoming movie Interstellar, Astronauts explore a black hole. The visual effects team for the movie sought advice from astrophysicists when it came to how gravity, light, etc. would behave near a black hole, and got the latest mathematical findings to plug into their graphics rendering program. And the results were so unexpected - yet accurate and supported by the math - that the film itself is now going to be featured in scientific papers about, "oh, hey, now we actually can SEE what a black hole looks like."
posted by EmpressCallipygos at 3:09 AM on November 7, 2014 [11 favorites]


Mind blown.
posted by alexordave at 4:18 AM on November 7, 2014


straight: "That movie is like the very thing I dreamed of seeing when I was a kid: to look out at a vast, starry night, choose a star, and just go there and see it. It's the real thing that the thrill I get doing a deep-dive with a real-time fractal zoomer is hinting at. Thank you."

I downloaded that fractal zoomer and put it on autopilot with a newtonian fractal. Then I watched it zoom in, following each identical formation as it delved into another.

I had to stop because the feeling of being a tiny, insignificant speck in the vastness of the universe was overwhelming me. Also, I need to go to work to support my family, which, though insignificant in the grand scheme of things, is still important to me.
posted by double block and bleed at 4:30 AM on November 7, 2014 [1 favorite]


Hey, we have something like this in the Netherlands: LOFAR, very cool technology.
posted by Pendragon at 4:30 AM on November 7, 2014


Fantastic post, thanks sedna17.
posted by metaBugs at 5:00 AM on November 7, 2014


There are other large interferometers (the VLA and LOFAR already mentioned in this thread), as well as others. However, it's generally not quite true that an interferometer acts like a single telescope of the same size. In particular, if you only have a small number of telescopes, then your ability to make a pretty/accurate picture is compromised.

Imagine that you have an unknown picture on a scratch-off lottery ticket. For each pair of telescopes you can scratch off a single short curve. With a small number of telescopes you'll only get hints at what the underlying picture looks like, and you'll have to make some guesses to reconstruct the image. ALMA has so many telescopes that you get to scratch off almost everything and really just see the picture. That explanation is approximate (really there's an extra Fourier transform in there), but it's a fine way to think about it.
posted by sedna17 at 5:43 AM on November 7, 2014 [2 favorites]


I see no sign of Bode's Law in that system, which is interesting.

Though if Bode's Law results from orbital resonance between planets, wouldn't we expect a system to take tens or hundreds of millions of years to settle into a regular pattern of orbits (this system estimated at being a relatively infantile 1M years old)? It seems from the amazing image going around that the system must still be in the "proto-planets clearing huge amounts of debris out of their orbits by accretion" phase.
posted by aught at 6:27 AM on November 7, 2014


I like that "Space artist" is now something a kid can want to grow up to be. It also makes for a cool business card.
posted by echocollate at 7:30 AM on November 7, 2014 [1 favorite]


It might even be the case that instead of single protostars forming in molecular clouds, various eddies may form, begin orbiting larger eddies, and then collapse at relatively the same time.

"Eddies in the space-time continuum."

"Is he?"
posted by scalefree at 7:39 AM on November 7, 2014 [5 favorites]


ALMA is amazing.

The link to the interferometry basics is rather technical - it's meant to guide astronomers proposing for ALMA time - but in hand-wavy terms, the two things we're working with are resolution and sensitivity.

Resolution is the level of fine detail in an image, and it scales as the wavelength divided by the telescope "diameter". A bigger telescope sees more detail, and looking at shorter wavelengths shows more detail. I put diameter in quotes because for an interferometer, what matters is the longest baseline, but it follows the same idea - the wider your telescope spans, the finer your resolution.

Sensitivity is how faint or deep you can see. This scales straightforwardly with your collecting area - the bigger your light bucket, the more photons you'll catch and the better you'll see. ("What big eyes you have, Grandma!" "All the better to see you with, my dear.")

So ideally, you'd want a huge telescope with a very very smooth surface. That way you'd be able to operate down to very short wavelengths, have incredible resolution, and have a big enough bucket to collect lots of photons. Unfortunately, that costs money. (I think the current estimate is that cost scales as the cube of the telescope diameter, but that's subject to the laws of economics as well as engineering physics.) At radio wavelengths, we've built big telescopes. Like, altering the geography kind of telescopes. But there's only so big you can go. (And the Chinese are determined to go bigger!)

So eventually, you resort to a trick: you construct an imaginary telescope of vast diameter, and only fill in parts of the surface. Your sensitivity depends on how much of the surface you could afford to fill in, but your resolution depends on the size of this imaginary telescope - the longest distance between your filled in patches. That's it. The trick is, instead of geometric optics, you have to focus your light by first receiving it at each patch, then bringing all the received signals together somehow (waveguides, optical fiber, FedEx - not kidding) and putting them through your own custom supercomputer. That's incredibly costly, but it does let you "focus" your telescope over a much wider patch of the sky than what would have been the focal plane of your huge telescope.

So the catch that I glided over is in the "bring all the signals together" bit.

As you know, Bob, light behaves as both a particle and a wave, and this wave-particle duality depends on the energy of the photon. At long wavelengths, we think of radio waves. At short wavelengths, we think of gamma ray photons like bullets. There are radio photons too, but each one carries very very little energy, and there are very very many of them, so you can receive a radio wave and capture that signal, and make copies while preserving its phase. (This bit is DEEP. You should say "Woah" here in your best Neo voice.) That's why we can record signals at each VLBA antenna and literally FedEx the tapes over to one central location and do interferometry. (I do this. It's neat.)

At optical wavelengths, this is not so easy: you can't make copies without destroying phase. So instead, you have the much more expensive process of masking out bits of your aperture, or you have the Keck telescope outrigger concept, where you're correlating the photons received at each big telescope against photons received at the small outriggers. But that's just a handful of baselines (2x4 = 8). With the VLBA, because we can make perfect copies of the signal at each antenna, we have 10x9/2 = 45 baselines. At the VLA, that's 27x26/2 = 354 baselines.

So ALMA operates at the sub-millimeter wavelengths, really right at the cusp of this wave-particle duality.

That's what's so astonishing about the technical feat - the telescopes are mirror smooth (since they have to be smoother than at least a quarter of the wavelength they're operating at, if you want any hope of useful focusing) and the electronics in the amplifiers and detectors are operating at the outer reaches of our capability, where weird and wonderful quantum effects are routine. And we take it for granted - it's all engineering these days - and use it to do interferometry at sub-mm and make images of baby planets carving channels in their protoplanetary disks.

Like I said, it's amazing.
posted by RedOrGreen at 7:53 AM on November 7, 2014 [31 favorites]


For some reason I woke up before sunrise this morning, so I lugged my telescope out my back door and killed a little time before starting the day. Jupiter was almost directly overhead, so I looked at Jupiter and its moons for a while, then looked around to see what else was in the sky. The sun was just rising in the east, and Orion was about to set in the west, so I pointed my scope there and looked at the Orion Nebula.

With all the light pollution of Austin, and the lightening sky, it was just a hazy blue/green "cloud" in my eyepiece. It's about 24 light years across. Across.

I was hoping to see the Andromeda galaxy but the time wasn't right. When you look at that thing, the light that reaches your eyes left Andromeda before humans walked the earth.

Space is rad.
posted by spikeleemajortomdickandharryconnickjrmints at 7:57 AM on November 7, 2014 [3 favorites]


>>It might even be the case that instead of single protostars forming in molecular clouds, various eddies may form, begin orbiting larger eddies, and then collapse at relatively the same time.

"Eddies in the space-time continuum."

"Is he?"


Of course the turbulence in molecular clouds is (to first approximation) scale free.
posted by sedna17 at 8:07 AM on November 7, 2014 [1 favorite]


Cool space stuff aside, operating mm wave at that scale is jaw dropping.
posted by Dr. Twist at 8:32 AM on November 7, 2014


Fascinating post. EmpressCallipygos's linked article is a great read as well.
posted by Librarypt at 9:03 AM on November 7, 2014


We could have had pictures like this much sooner if they'd listened to Grote Reber. I hereby nominate Dec. 22 to be Grote Reber day.
posted by Twang at 9:18 AM on November 7, 2014


Somewhat related, here are some optical images of star discs from Hubble.

I feel like I'm seeing a lot more European astronomy work recently. Obviously the Rosetta mission, but more than just that. I wonder if they realized like NASA did that public outreach over the Internet was a worthwhile effort.
posted by Nelson at 10:41 AM on November 7, 2014


Since radio wavelengths are many times larger than optical wavelengths, radio telescopes like the GBT and Arecibo are large compared to optical telescopes.

I just want to add that high-frequency radio has wavelengths in the mm range. This means you can have small radio telescopes for those frequencies, like the one on the roof of the Harvard astronomy dept. in Cambridge MA
posted by vacapinta at 10:55 AM on November 7, 2014


Nelson: I feel like I'm seeing a lot more European astronomy work recently.

This may very well be true - and they're doing excellent work, especially with their much more stable funding situation - but don't let the ESO links in the post mislead you. This result is coming out of NRAO, and Crystal Brogan (quoted in the NRAO release) and Bill Saxton are based at Charlottesville. At least, I think that they're the primary people involved, but I'm not plugged into ALMA and it's a big international collaboration, involving ESO, NRAO in the US, NAOJ in Japan, etc...
posted by RedOrGreen at 11:45 AM on November 7, 2014 [1 favorite]


> ... so you can receive a radio wave and capture that signal, and make copies while preserving its phase. (This bit is DEEP. You should say "Woah" here in your best Neo voice.)

Picture me as Ted, not Neo, saying "Nooo Waa-ay".

Seriously, if you have links or would like to explain a bit in thread, I'd love to hear details on

- how they manage to record phase
- how this depends on the wavelength
- how they manage to 'synchronize' phases in the data from different sites.
posted by benito.strauss at 2:19 PM on November 7, 2014 [1 favorite]


An old friend of mine, Alison Peck, was the deputy project scientist with ALMA for five years through construction and early science and to my great surprise some googling reveals that CBS had a 60 Minutes segment on ALMA on March 9th, and she's one of the scientists in it. Did you know she was on 60 Minutes, RedorGreen? (RedorGreen's an astronomer and we discovered a couple of years ago that we both knew Alison.)

Anyway, she was with the NRAO and you can see that the personnel represent a number of national agencies. The US's NRAO and the EU's ESO dominate, but there are many others. ALMA is truly international because it's pretty ambitious. It's big and expensive.

In the 60 Minutes segment, toward the end they show and discuss some imagery from what I believe is this protoplanetary disk of HL Tauri that's in this post and in the news recently.

We've had decades of great astronomy from radio telescope array interferometery, but ALMA represents a huge multinational project to level up, so to speak.

There's all sorts of limitations and difficulties with ground-based optical astronomy. By the 70s, we'd reached a point where even if we made the mirrors bigger and smoother, we'd still have the limiting factor be atmospheric distortion due to pockets of air in the telescope's line-of-site fluctuating in density. Developments since then have answered that particular problem (and that deserves a post all its own, because that's also very cool), and so with that and other improvements there's been a renaissance in optical astronomy since the 90s, but you've still got these big, expensive and delicate surfaces for observing what is only a small slice of the electromagnetic spectrum.

All sorts of interesting things are happening outside the visible spectrum.

Faster frequency/shorter wavelength (I'll prefer the wavelength description from here on) means higher energy and so a lot of very interesting, energetic things are observable at the x-ray and gamma ray wavelengths above the visible spectrum. However, that's all absorbed by the atmosphere so we're limited to space telescopes for those.

The longer wavelengths, though, make it through the atmosphere pretty well. The VLA, which has been until now the world's premiere array of radio telescopes, observes in the wavelengths from about seven millimeters to about four meters, which is pretty much in the range we popularly think of as "radio" and includes television and wifi signals and such.

In contrast, the Sub-Millimeter Array on the top Mauna Kea in Hawaii, where Alison worked before joinging the team at ALMA, observes between about a half a millimeter to one-and-a-half millimeters, which is way up there right where (microwave) radio becomes infrared. ALMA pushes a bit further into that range, but also observes farther down all the way to ten millimeters, which slightly overlaps with the VLA's shortest wavelengths.

Water in the atmosphere is a factor at these wavelengths and so that's why ALMA is at a very high elevation at one of the very driest places on Earth, and why the SMA is on the top of Mauna Kea.

Radio waves are, of course, light waves (as are x-rays and gamma rays) and so all of this astronomy is observing the light from astronomical objects (you still sometimes come across someone referring to radio telescopes as "listening" which is like fingernails on a chalkboard for me) and so we're both seeing the same sorts of things we see with optical telescopes but also very different things.

Certain frequencies are associated with certain elements and molecules, both in terms of excitation (shining) and absorption, as well as it being more broadly the case that certain ranges of frequencies are associated with certain physical processes. For example, the highly energetic gamma rays are associated with highly energetic physical interactions, such as the collision of particles. So while it's the case that this light outside of the visible range can allow us to see the shapes of things in the way that we see visible objects, it actually tells us quite a bit more than just that -- the specific frequencies of light we're seeing (and not seeing) tell us specific things about the physical processes going on. Furthermore, as I said, different things shine or are opaque to different frequencies, so just as is the case with x-rays, where the x-rays pass through soft tissue but are blocked by bone, different ranges of light show us the shape of the things that are obscured by other things that are opaque to those frequencies.

Taken together, this means that observing using different ranges outside the visible spectrum allows us to see the shape of different things (including in the same region of space) as they are visible in those different ranges, like how we see a person in visible light but their bones in x-rays -- and the simple fact of what those specific wavelengths are, what's shining, what's absorbed, tells us a great deal about what that stuff is made of and what it's doing. This is how we learn so much with radio astronomy. In the 60 Minutes segment, Alison talks about finding sugars and alcohols and that's how we know this, how we know what some of these things are.

Going back to the specifics of ALMA and sub-millimeter telescopes, while it's the case, as I've written, that there's a whole lot of interesting things to see in those longer radio wavelengths, it's also the case that the shorter the wavelengths, the higher the energy. That implies a number of things, but basically in this context it means that sub-millimeter observations give us a window on a whole bunch of stuff that's very interesting to us while also having all the benefits of radio astronomy.

Incidentally, Alison first appears in the 60 Minutes segment talking a bit about ALMA's correlator, which is the special kind of supercomputer that makes interferometric radio telescopes possible. In 1994, Alison and I visited my aunt and uncle, both of whom were programmers, and my uncle worked for Sandia National Labs. They asked her questions about the VLA and then my uncle asked about the correlator, "That's a teraflop machine, isn't it?" And that was about right, depending on how you looked at it. For context, the first general purpose supercomputer that reached a teraflop (a trillion floating point operation a second) wasn't built until two years later, ASCI Red, which, incidentally, was at Sandia. VLA's correlator was a very different kind of computer, so it's not strictly accurate or sensible to talk about it that way. The point is, though, is that it was one of the most powerful computers on the planet.

Fast-forward to today. Now, the computer I'm typing this on has an nVidia graphics card that can manage 1.3 teraflops. The recently upgraged VLA's correlator manages 10 petaflops, which is equal to ten-thousand teraflops. ALMA's correlator reaches 17 petaflops.
posted by Ivan Fyodorovich at 10:38 PM on November 7, 2014 [5 favorites]


Reminds me of the LHC triggers, which is the sexiest high-speed computation device I've heard about recently. Trigger speeds seem to be given in events/second, not flops, so it's hard to compare the two, but considering that they've got torrents of some of the most expensive data on the Earth running through them I imagine they build 'em to be fast.
posted by benito.strauss at 9:37 AM on November 8, 2014 [1 favorite]


> VLA's correlator was a very different kind of computer, so it's not strictly accurate or sensible to talk about it that way. The point is, though, is that it was one of the most powerful computers on the planet.

Yeah, and what it's doing is basically focusing the telescope - instead of just reflecting all the light rays to a focus, we're receiving them at each of the VLA's 27 antennas, piping them to the correlator building, and then correlating each pair of antennas (27x26/2 pairs, or baselines) over each tiny slice of time and observing bandwidth. The data rates are astounding - for those of us used to the earlier days of the VLA, the EVLA and ALMA now spit out wild torrents of data.

benito.strauss: if you have links or would like to explain a bit in thread, I'd love to hear details on
- how they manage to record phase
- how this depends on the wavelength
- how they manage to 'synchronize' phases in the data from different sites.


Okay, I'll tell you about the last one first. It's "easy" - we start with a precise correlator model, which predicts the relative delay between the wavefronts hitting each antenna for each pointing direction in the sky. That gets us close, but the phase offset drifts with time (and weather, and temperature, and ...) - so we look at calibrator sources that are known to be compact. Then our desktop processing software asks this question: assuming this was a point source, what is the correction I need to apply to the recorded phases in order to line them up? In the 80s and early 90s, this question required mainframes and minicomputers to answer. These days my laptop breezes through it, as long as the data set is not too big. And once we've computed the corrections, we apply them to the target sources. Usually, we'll bracket observations of any science target with calibrator scans.

With the VLBA, we have to not just fit for phase offsets but also drift rates and time delays - remember, we're spanning a continent. And you can literally observe Hawaii drifting away from the mainland - the "fringe rates" to Mauna Kea get worse over months until the correlator model is updated with a revised position. ALMA is now implementing its long baselines - 16 km instead of 2000 km, but the wavelengths are mm instead of 20 cm, so similar issues apply.

Your other two questions are hard to answer without getting in too deep. But the best place to learn interferometry is the biannual NRAO Synthesis Imaging Summer School. Back in 1998, when I attended as a grad student, I had the privilege to hear V. Radhakrishnan deliver the last lecture of the school. At the time, it was cryptic and a total head scratcher, but he was explaining the exact issue I was trying to outline above, that quantum effects prevent you from copying and preserving phase at shorter wavelengths, but allow you to do so at longer wavelengths. I don't think I got that from the lecture... Luckily the lecture notes went into a book - the famous White Book that every radio interferometric astronomer has - and Rad's lecture is the last chapter in that book. I'm not sure if that is at all suitable for your technical level, but you might be interested in skimming Chapter 33 (look at Figure 1, on page 676, for example) in the White Book. Here's the chapter, archived on ADS.
posted by RedOrGreen at 1:02 PM on November 8, 2014 [2 favorites]


Thanks for the reply. The linked chapter is at the edge of my understanding (lots of math, but only two years of physics), but it looks like he's kept it at a higher level, not bogged down in details, so I'll give it a shot.
posted by benito.strauss at 3:07 PM on November 8, 2014 [1 favorite]


« Older The Slow Unveiling of James Tiptree Jr.   |   Just some awesome crafts Newer »


This thread has been archived and is closed to new comments