The largest telescope ever made

What will the square kilometre array teach us?
24 November 2020
Presented by Ben McAllister, Adam Murphy
Production by Adam Murphy, Ben McAllister.


An image of a spiral galaxy


This month on Naked Astronomy, we're taking a closer look at the largest telescope ever to be built, the Square Kilometre Array. The SKA spans continents, with some of it in South Africa, and some of it in Australia, but how does that work? And also, what is the SKA going to show us about our universe. To find out, Ben McAllister and Adam Murphy spoke with Phil Diamond, the Director General for the SKA project...

Ben - You’re probably familiar with the concept of a telescope - humans have been making them for at least hundreds of years, and using them to learn about the Universe beyond our planet.

Adam - At its core, a telescope is a device which we point at space to gather light from distant objects, and use it to form images of far away things

Ben - When you think about a telescope, unless you’ve got some kind of astronomy background, you are probably thinking about an optical telescope - that is to say, something like what Galileo used - a series of lenses and mirrors which capture and magnify visible light arriving to Earth from space.

Adam - Indeed, optical telescopes are very important scientific tools, but they are by no means the only major kind of telescope. In fact, a lot of ground-breaking science has been done using what are called “radio telescopes” - a sort of cousin to the types of devices which Galileo would have been familiar with.

Ben - Light comes in a variety of different flavours. The visible light which produces colours like red, green, blue, etc - is only one very narrow and specific region of the broader spectrum of light.

Adam - Things like x-rays, gamma rays, UV, infra-red and importantly, radio waves -

Ben - Like the ones used to broadcast fine programs like The Naked Scientists -

Adam - Indeed! These different kinds of radiation are all different kinds of light, at different points along what is called the electromagnetic spectrum.

Ben - You can think about it as a whole bunch of other colours, just like red and green and blue, but which are in the wrong part of the spectrum for our eyes to detect, because our eyes evolved to detect the kinds of light the sun puts out. But, rest assured, those other kinds of light are light all the same.

Adam - We refer to the different types of light by a property we call their “wavelength”. Lots of things in space which we are interested in learning about don’t emit light in the visible part of the spectrum - they emit light of different wavelengths.

Ben - In particular, lots of interesting heavenly bodies emit light in the form of radio waves - which brings us to radio telescopes.

Adam - We can’t see the radio wave light emitted by, for example, massive hydrogen gas clouds in space using our eyes, or an optical telescope - it simply passes through undetected. Instead, we need a different kind of telescope specially designed to detect radio waves.

Ben - Radio telescopes often look like a big antenna sitting inside a dish - if you’ve ever had a satellite dish on your house, or seen a radio station tower, you get the idea. The dish structure reflects radio waves onto the antenna, where they can be detected and converted into images.
Adam - But of course, because we can’t SEE the radio part of the spectrum, we need to look at the data in different ways, and measure the amount of light of a given wavelength that arrives at the antenna from a given direction. We can also use clever data processing to generate images that we CAN make sense of with our eyes - representations of what we MIGHT see if we could detect radio waves with our retinas.

Ben - Radio telescopes open our eyes to see all kinds of different colours of light coming from space, not just the ones we are used to.

Adam - We’ve been building radio telescopes since the 1930s, and they’ve played a major role in some of the largest discoveries in the history of astronomy.

Ben - It was radio telescopes that allowed us to detect the Cosmic Microwave Background - the leftover radiation from the dawn of the Universe, which has a wavelength far too long to be detected with optical telescopes.

Adam - Radio telescopes are also responsible for many of our observations of the big structures in the Universe. We mentioned clouds of hydrogen gas before - these are important, because hydrogen is the most abundant element in the Universe, and wherever there is a lot of matter - say in a galaxy for example - there is likely to be a lot of hydrogen. This means that by looking for hydrogen, we can learn a lot about large, heavy structures.

Ben - It turns out hydrogen gas has a natural, characteristic wavelength of light that it emits - the famous 21-cm Hydrogen Line.

Adam - This special wavelength is squarely in the range detectable by radio telescopes, meaning we can use them to look for this characteristic radiation, figure out where there are large amounts of hydrogen gas, and learn more about the structure of big things like galaxies.

Ben - In fact, it was radio observations of hydrogen gas which allowed us to measure the way that galaxies like ours rotate. This work, in the ‘60s and ‘70s, led us to discover one of the biggest mysteries in the cosmos today: the understanding that over ⅚ of all the matter in our galaxy, and most galaxies, is actually this mysterious, invisible stuff called dark matter - a topic we will certainly cover in a future episode.

Adam - And, more recently - you may have heard of the Event Horizon Telescope in 2019, which produced the first ever image of a black hole. You guessed it - Event Horizon is a network of radio telescopes.

Ben - Event Horizon is a great example of one of the neat things about radio telescopes. The power of a radio telescope increases with the physical size of the area where it collects radio waves, but at a certain point, it becomes very impractical to make radio telescopes any bigger. There is a limit to how big they can get. However - if you’re clever, you don’t actually need to make one really large dish - you can combine signals from a network of smaller dishes spread out across the globe, just like the Event Horizon Telescope did.

Adam - The higher the total combined collecting area of all of your dishes is, the better. But funnily enough, having your dishes spaced out over a larger area - a longer “baseline” as astronomers call it - can also lead to more powerful imaging.

Ben - Using a clever data analysis technique called interferometry, scientists can combine signals from radio telescopes which are very far away from each other, and increase the resolution of the image compared to if they had just had a large antenna in one place.

Adam - However, synchronizing and processing all of the data from a large, distributed network of antennae can present an enormous engineering and computational challenge.

Ben - We’ve been building larger and larger and more distributed telescopes for decades - getting more and more powerful, and learning an ever-increasing amount about the Universe.

Adam - And all of that progress has been leading to an incredible device which is in development right now - the Square Kilometer Array.

Ben - The SKA will be the largest and most powerful telescope of any kind, ever developed by human beings. It will be a truly staggering, mind-bending achievement of science and technology.

Adam - Originally proposed in the ‘90s, and in the works since then, the SKA will consist of thousands upon thousands of receivers, spread out across multiple countries. It’s a collaborative, international effort to build one of the most sensitive scientific instruments of all time.

Ben - It’s definitely a work in progress - but scientists are confident we’ll get there sometime this decade, and when we do, we’ll be able to see the Universe like we’ve never seen it before.

Adam - So now that we know some of the basics of radio astronomy and the SKA, we’re ready to hear about it in a bit more depth - how far off is it, and what will it actually look like?

Ben - But don’t just listen to us blathering on, we were lucky enough to chat to one of the most qualified people on Earth to talk about it.

Adam - Phil Diamond is the Director General of the SKA Project, and he had lots of fascinating things to say.

Phil - So the sites have been selected. The sites for the two telescopes were selected back in 2012. And on those sites in both countries, the national radio astronomy communities, have built what we call precursor radio telescopes. So in South Africa, they built MeerKAT, which is 64 dishes, which is now operational. In Australia. CSIRO, the national research organisation, has built what is called ASKAP. The Australian SKA Pathfinder. That's 36, 12 metre dishes with very innovative radio cameras, at the focus. And then an international consortium of universities, led by Curtin in Perth, have built the Murchison Widefield Array, which is a precursor low frequency telescope. So what these projects have done, as well as building telescopes to do science, is they've established virgin sites effectively. There was nothing on these sites before these telescopes arrived.

Ben - Okay. So we've talked about the SKA, it's big radio telescope, it's distributed. What does it actually look like? Like what does a part of an SKA look like?

Phil - In South Africa, what we call SKA-mid, we are going to develop much more attractive names for these things.

Ben - Acronyms are very important!

Phil - But at the moment we call them SKA-mid and SKA-low. That's because SKA-mid, will work at what we call mid frequency range for radio astronomy. It will consist of 197 dishes. Each dish bowl is about 15 metres in diameter. They are large single dishes. The Lovell Telescope that I can see right out of my window from my office right now, is 76 metres in diameter, but building a massive array of 76 metre antennas is far too expensive. And building a huge single dish, many kilometres across is technologically challenging, even impossible. What we're doing is building an array of smaller dishes, 197 of them in South Africa, all connected by fibre optics, back to a central, really we call it a correlator, but it's a special purpose computer, it's a lot of digital hardware. So that's what SKA-mid will look like. SKA-low, which operates at much lower frequencies, between 50 and 350 megahertz, so encompassing the FM radio band for example, will have a different technology, because dishes don't work so well at low frequencies. We are reverting then there to an older technology, although much evolved, much improved. There they are what we call log periodic dipole antennas. So they're like two metre tall, almost Christmas tree-like antennas. They are cousins, evolved cousins of the old TV antennas we used to have on the side of our houses before, you know, Sky satellite dishes appeared. So we'll have almost 132,000 of these low-frequency antennas, spread across the Western Australian desert. It's actually 131,072 of them.

Ben - The technologies, they would look like stuff we would recognise, right? In Australia, we have stuff that looks like, as you said, old TV antennas, and in South Africa, things that look like the more modern satellite dishes that some people might have, I suppose.

Phil - That's exactly right. You know, the metal hardware will be familiar. It's the scale of it is different, much different from anything ever built before. We'll be building many more of these things, but the really innovative part, the challenging part, is the digital signal processing, and the software to cope with the enormous volume of data that these machines will generate. Just as an example, from these periodic dipole antennas, the raw data generated from all of the antennas is about six times the global internet traffic. So it's of the order of two petabits per second, which is about six times the rate at which data flows across the global internet.

Ben - Of course, because they're constantly acquiring.

Phil - We're constantly acquiring that data. Now, of course, the data is all in one format. It's on our network. We immediately put it through the first level of digital signal processing to reduce the volume of the data, and send it off to correlators. So we control every aspect of it. It's not the chaotic nature that the global internet is, in a very controlled environment.

Ben - Anyone who grew up on the internet can attest to that chaos!

Adam - Yeah, chaotic is the word!

Phil - But it's handling that scale of data. I think that is one of the challenges.

Ben - So you've got all these dishes and antennae pointed at the sky, constantly acquiring all of these radio waves from space that we can then use to do astronomy. You said there's like, a ridiculous amount of data. What kind of images do you get out of those radio waves?

Phil - There's a couple of different ways in which the data is handled. So radio images of the sky is one, because of the volume of data, there's the software that needs to be modified, upgraded, made to work on modern supercomputers. We will be producing radio images of the sky. So we'll be able to overlay these radio images with images from say, the Hubble space telescope. Another type of data we'll be processing is what's called time series. From each of the antennae, from each of the, what we call the baselines, the connections between them. We will be getting these time series of data. And this is another thing that astronomers will use to look for particular types of objects in the sky, such as pulsars. Pulsars are these rotating remnants of stars that have exploded in a supernova. And they emit these sort of lighthouse beams of radio emission. They're nature's clocks, superb clocks in the sky, and the way those data are received and processed is through taking the time series of data and looking for the time signals that emerge from that data. Pulsar astronomers can measure those times incredibly accurately. And there's all sorts of fundamental physics that emerges from pulsar astronomy. Similarly, there's these fast radio bursts, which are a relatively new object of which we know very little. SKA will be finding hundreds, thousands of these things. So hopefully we'll be able to understand the nature of these, but they are incredibly short, you know, millisecond bursts of very powerful radio signal from somewhere in the sky. We don't actually know what the objects are that emit these things yet. There are various theories, but none are yet pinned down. But actually what all this demonstrates is that we are not fixed on the form of object that we will see in the particular types of data. So we have deliberately designed the SKA to be an exploration machine as well.

Adam - So you mentioned there kind of, some things the SKA will be looking at, but are there any, I suppose, like mission statements, any things in particular you're going to point it at first, or anything you're particularly excited to point it at?

Phil - The science case for SKA, if I it is.

Ben - Hahaha! For audio listeners, Phil has just picked up two very large thick books, the size of two encyclopedias each.

Phil - Yeah. So they actually, they weigh 9.8 kilograms combined.

Adam - Oh, good exercise then. Yeah.

Phil - Yeah. So that's 2000 pages describing the science that astronomers hope to do with the SKA, but in there, we have identified what we call key science programs. So the SKA must be able to deliver these key science projects, as well as do an enormous other range. But at the very least, it must be able to do the key science projects. One is using pulsars, to use pulsars to understand the nature of gravitational waves. A few years ago, LIGO the laser interferometer gravitational wave observatory detected gravitational waves from massive objects coalescing, and emitting gravitational waves across the universe. That's one way of detecting them, another way of detecting gravitational waves from much heavier objects, supermassive black holes at the hearts of galaxies, is by observing the effect of the gravitational waves on the pulsars that sit across the Milky Way, across our galaxy. We'll be using the SKA to monitor the signals coming from a network of pulsars. These are very accurate clocks. Nature's clocks remember, distributed across the sky. And as a gravitational wave crosses our galaxy, it will perturb the clock signals and we'll be able to detect that perturbation, and therefore measure the gravitational wave.

Ben - So like a galaxy wide gravitational wave observatory, as opposed to a few kilometre tunnel?

Phil - Yeah. It's using the galaxy as a telescope, as a detector, which I find mindblowing. Another key area is to use hydrogen. So with the SKA, one of the reasons we're going from 50 megahertz all the way up to 15 gigahertz, is to be able to have that full range of frequency, to detect hydrogen all the way back, almost to the Dawn of the Universe. So we want to see what happens in those early years of the universe. When the universe started to become transparent to radiation, and watch the first stars, the first galaxies evolve all the way up until the present day. So essentially, we'll be acting as a time machine.

Ben - Is the point just that because hydrogen is everywhere in the universe, it's extremely abundant. Being able to see the characteristic, radiation that only hydrogen emits, allows you to see a lot of the interesting structure. And because the signals of the hydrogen in the early universe are still kicking around in the universe, they're just changed a bit, we can still see those really early signals and kind of reconstruct what happened in the early universe?

Phil - No, that's exactly right. Our images of the early universe have come from projects, spacecraft like Planck, an ESA mission that ended a few years ago, or its precursors. These satellites produced snapshots of what the universe looked like when it was about 400,000 years old. Bear in mind, it's almost 14 billion years old now. So we have these snapshots and there's a huge amount of physics being gained from those snapshots of the universe as a baby, as a child, really. And what we want to be able to do with hydrogen, which as you say, is everywhere, is make a movie of the universe, from its childhood, growing up through adolescence, to the mature universe that we now live in. There's going to be some fantastic things come from that, that ability to just observe that time period.

Ben - I can't wait to see what we learn.

Phil - Those are just two of the key science projects encapsulated in the 2000 page science case, there are many more. We'll be trying to understand the fundamental nature of magnetism in the universe. Again, that's best studied by radio astronomy. We'll be looking for the signatures of biomolecules, which are potentially the origins of life on Earth, and therefore potentially elsewhere in the universe. Many other things as well.

Ben - Why is it called the Square Kilometre Array? It sounds like it's much, much bigger than a square kilometre in size.

Phil - The physical extent of the telescopes are much bigger. So in South Africa, these 197 dishes are going to be spread across 150 kilometres. And in Western Australia, the stations of the dipole antennas spread across 65 kilometres, but there's a lot of gaps, a lot of space in there, between the dishes and the antennas. In a sense the SKA, square kilometre is a misnomer right now. What we're designing and shortly building, is phase one of the project. And the combined surface area of the dishes and antennas will be, in that case, will approach a square kilometre. That was the original ambition, that is still our aspiration, is to go for what we call the full SKA. So the current thing that we are building is a step along that road. So as I said, it's a bit of a misnomer, but it does reflect our future ambition and what it is, if you add up the surface area of each of the individual dishes, and they're equivalent in the antennas, it's a mathematical concept for the antennas. It should, with the full SKA, approach a square kilometre.

Adam - If you've got a load of antennae, just kind of sitting on the ground. That's just taking in all the radio waves that are coming in. How do you pinpoint that, you know, okay, this bit of radio wave came from this bit of the sky?

Phil - Imagine that the dish is a radio version of an eye. With your eye, when you look in a certain direction, you can't necessarily see what's off to the side. You know, you have your peripheral vision. You can't see what's behind you, or way off to the right or the left. So, you know, your eye has a beam effectively, that it can see. And of course, you know, you move your head to look at other things, and that's exactly analogous to what we do with the dishes. The dishes, they have a beam shape, approximately equivalent to the size of the full Moon. In fact, even larger. To point to different areas of the sky, yeah, we just move all of the antennas to point in that direction. That gives us the field of view that we are looking at, is the beam of an individual dish. But then to see the detail within that dish, that's where we rely on the array, the interferometric nature of the SKA, because we'll have individual dishes separated by up to 150 kilometres. And combining the signals from those dishes, synthesises, another beam, which is much, much smaller because of the separation of the dishes. And so that acts like a zoom lens within the field of view. So we can see very fine detail on the many thousands, hundreds of thousands, millions of objects within the field of view. And that's how we build up the radio picture of the sky. The low frequency antennas are somewhat different. Each individual antenna effectively sees the whole sky all the time. But what we do is electronically create beans from the individual stations, and steer them around the sky, according to the object the astronomers want to look at. With those antennas, we can have multiple beams operating, pointing in different directions at the same time. Because it really just depends upon the software and the digital electronics in the backend. So they might look like simple Christmas tree antennas sitting in the desert there, but there are very sophisticated systems behind them.

Ben - So just roughly then, how would that, that work? I mean, if you've got the radio waves hitting the antennae from all directions at all times, how do you tell like, the algorithm or the digital signal processing, which parts of the radio waves incoming are relevant?

Phil - What happens is an astronomer, or more likely a large team of astronomers, will put in a proposal to do a particular project, you know, to observe a particular area of the sky for a long time, to get a very deep image, or to observe the whole sky to get less deep images. But a survey of the sky, whatever, they'll put a proposal in. That goes through a selection process, the science will be compared with science from other proposals. And if it is selected, it will get time on the telescope. Then the SKA team will translate the astronomers instructions, to point the telescopes in different places at different times. And we'll get the data. In the old days, we used to have what we called mercury delay lines, which were actually tubes filled with mercury, that were actually adjusted, but each individual pair of antennas, so that they all effectively weighted the beams towards a particular part of the sky. We now do that in digital electronics.

Ben - So it's like adding a delay or something from one antenna to the next, that creates this artificial beam shape.

Phil - Exactly.

Ben - Okay. And does that relate to what you were talking about with the interferometric nature of the dishes, in terms of how far apart, having the dishes further and further apart enables you to, by cleverly combining the signals get finer resolution across the sky?

Phil - Yes, that's exactly right. So my analogy is one of the zoom lens, a big single dish is a wide angle lens. An interferometer with widely separated dishes is a zoom lens gives you that fine level of detail in the sky

Ben - I don't want to hold you to anything. But roughly, what do you think is the timescale for when we're going to be turning on even the first stage of the SKA?

Phil - We're just about to create the SKA observatory as a living entity. This is, there was a treaty that was signed in March of 2019. The treaty has been ratified by a requisite number of countries. We're going through the final administrative stages. And within a few weeks, we should announce the formation of the observatory itself, which is the successor to the current organisation that we have. So everyone will, all the staff will move across to the observatory, which is like ESO or CERN. It's an inter-governmental organisation, the governing council of the SKA observatory, which has members from the participating nations, I am hoping will give us the permission to start construction activities in the middle of 2021. Our target date is actually the 1st of July of next year.

Ben - That's very exciting.

Phil - Yeah, that is, after all of the years we spent developing and designing. That's very exciting. So if we start construction activities, which is going out to tender, to industry on the 1st of July, I hope we will see the first hardware on the ground, the first testing of real hardware on the ground two years later. Early science, which is the commissioning of science modes with significant fractions of the two telescopes in about 2025, and end of construction, 2027, 2028.

Ben - So that ended up being a more complicated question than I intended it to be, but just very quickly. So let's say it's 2028, we've got the SKA, we're doing all kinds of incredible science. We spend a little while using this incredible tool and we learn a lot of amazing things. Let's speculate for a moment. Let's say you could have something even more high concept, even more ambitious. What would be in your mind, the future telescope beyond the SKA, and what kind of technology would it rely on?

Phil - Well, I've mentioned the full SKA, phase two expansion across Southern Africa and across Western Australia, ultimately astronomers are thinking of a dark side of the Moon, telescopes up there. Now that really is unlimited budgets required.

Ben - Yeah, absolutely.

Phil - So to put things up there, almost certainly robotic technology to deploy and operate these things, a three-year postdoc up on the dark, far side of the Moon might be interesting.

Ben - It sounds like fun.

Phil - It sounds like fun, but yeah, that's probably the ultimate I could imagine at the moment. One could also imagine free-floating radio telescopes with, you know, several hundred thousand kilometre baselines out in space, but that's even more technological challenges associated with that. I think they are for a future generation.


Add a comment