AUSTRALIAN FRONTIERS OF SCIENCE, 2008
The Shine Dome, Canberra, 21-22 February
The radio Universe at high resolution and the Square Kilometre Array
by Professor Steven Tingay
 |
Steven Tingay has a science degree from The University of Melbourne and a PhD in astronomy and astrophysics from the Australian National University (Mount Stromlo and Siding Springs Observatory). He has been involved in the international SKA project as a Chair of the Simulations Working Group for a number of years. He was a Bolton Fellow at the Australia Telescope National Facility for three years and a Resident Research Associate at the Jet Propulsion Laboratory in California for three years. Steven has recently been appointed at the Curtin University of Technology in Perth, Western Australia, where he will be working in the field of radio astronomy, in particular in VLBI- and SKA-related research. Prior to this, he was an Associate Professor in the Centre for Astrophysics and Supercomputing at Swinburne University of Technology in Melbourne for five years. |
Naomi McClure-Griffiths and Anne Green have given a really good introduction to what the SKA is all about at the base level. Naomi has talked about science pertaining to the Milky Way galaxy, and this is science that requires what we call a relatively low angular resolution. I am going to talk about aspects of the SKA and the Australian prototype telescope that address much higher angular resolution science. So this is looking at smaller objects in fine detail.
(Click on image for a larger version)
This slide shows a static version of Naomi's movie. As Naomi described it, there is a centrally condensed inner core of antennas or stations, as we call them and then, progressing outward, there are locations where there are localised concentrations of antennas. These extend out for up to 3000 kilometres.
The thing to remember here is that the resolution, the amount of detail that you can see, in objects on the sky is proportional to the separation between these stations. Stations that are a long way apart give you very high angular resolution, allowing you to see a lot of detail. Antennas that are closely spaced give you relatively low angular resolution.
(Click on image for a larger version)
The SKA specification as it stands at the moment is for stations to have separations of at least 3000 km. That is, roughly speaking, the width of the Australian continent. So, to do a back-of-the-envelope calculation: if an observation uses a radio frequency of 1 gigahertz that is, a wavelength of about 20 cm in length with a baseline length of 10 km, that translates to an angular resolution of a millidegree. Just to get a sense of scale: your clenched fist at arm's length subtends an angle of about 10°. A millidegree is about the width of an apple seen from about 6 km away. So that is what I would describe as relatively low resolution.
What I am interested in is the 3000 km baselines. Now we are talking of the order of microdegree resolution. That is the equivalent of an apple at 6000 km distance. So these are the types of angular scales that we are interested in, in probing the universe, and later in the talk I will give a couple of examples of the types of things that we are interested in looking at.
In summary: at base level, short baselines give you low resolution, and they are good for nearby and/or intrinsically large objects like the Milky Way and structures within the Milky Way. The long baselines give you high resolution, and they are good for distant and/or intrinsically small objects.
(Click on image for a larger version)
This could be one realisation of what the SKA looks like in Australia. The large dots are what we consider as remote stations, and you can see the concentration of antennas way out in WA. Overlaid in yellow is the population distribution of Australia, so you can see that we have purposely avoided areas of high population density maybe with the exception of Hobart.
(Click on image for a larger version)
The types of things that we are interested in looking at, and are capable of looking at, are some of the most extreme physical phenomena in the universe. These are processes that release massive amounts of energy in relatively small volumes, talking on the scale of astronomy in the cosmos things like the relativistic jets that originate at the very heart of very distant galaxies, originate from a black hole and accretion disk system.
We have on this slide, at the top right, a schematic of such an object. At the very centre you have a supermassive black hole, maybe 107, 108, 109 solar masses worth of material. You have extreme gravitational fields, and this attracts gaseous material from its environment; most of it is gobbled up by the black hole but some percentage of it gets ejected from the system in highly relativistic bipolar jets. These jets of material travel well in excess of 99 per cent of the speed of light, and they radiate strongly at radio wavelengths. These things are like beacons across the universe, so the types of high resolution instruments we are talking about can probe the environment very close to the supermassive black holes. You can learn about the black hole itself, you can learn about the relativistic jet and also about the material that these jets are slamming into.
Another example is the case of exploding stars, or supernovae. These are stars that have reached the end of their life. They collapse, and very quickly rebound, and this expends a massive amount of energy in the volume of a star, which is not very big. So we can look at these exploding stars in galaxies tens of millions of lightyears away.
On the left of this slide you see the result of one of these explosions. This is a sequence of images with time, showing the expansion of the exploding shell of material as it evolves.
The remnant of these supernova explosions is generally neutron stars. Some of these neutron stars appear as 'pulsars'. The neutron star has a very strong magnetic field attached to it, and generates very highly columnated beams of radiation. The neutron stars also spin at very rapid rates, up to once every few milliseconds or so. Because they are so massive and they rotate in such a regular fashion, these are the universe's most accurate clocks.
At the bottom right of this slide we have a sort of toy model of what a pulsar looks like. Every time the beam of radiation sweeps past our line of sight, you see a pulse, and so you can time these things and do tests of general relativity in extreme gravitational fields.
The SKA will detect every pulsar in our galaxy, and you need very high resolution to be able to measure their distances accurately, in a model-independent way.
So, in summary of the broad scale of science: we are interested in phenomena that are at the extreme limits of physics speeds near the speed of light, and extraordinarily extreme gravitational fields from these massive objects. We are also talking about objects that generate this radiation over the entire electromagnetic spectrum, from radio waves up to very high energy gamma rays (γ-rays).
(Click on image for a larger version)
I will show a couple of examples. This slide shows a little bit about supernovae in distant galaxies. What we can learn through the study of these things are the star formation rates and the supernova rates in these galaxies, it can tell us something about the gas dynamics in the galaxy, and we can use the supernovae as a probe of the ionised medium within the galaxies. Broadly speaking, this allows us to look at a galaxy tens of millions of lightyears away and determine some of its properties, so that we can compare them against the properties of our own galaxy. And that's a very fundamental thing to do.
We can also learn about the physics of the relativistic expansion of the supernova remnant and the interaction it has with its ambient medium.
(Click on image for a larger version)
At the top left is an image of a galaxy, taken through an optical telescope. I think a really good illustration of the multi-wavelength aspect of astronomy nowadays is provided by the image at the bottom left. This is an image of the same galaxy as before, or several images overlaid. The red shows the ionised hydrogen gas content of the galaxy, quite different by comparison with the optical image. The green is the infrared emission; the blue is from a radio telescope, and is a probe of the molecular gas in the system. So you have got matter in many different phases. You have got molecular gas, you've got atomic gas, you've got stars, you've got dust and you need a suite of instruments to be able to reveal all of this.
In the centre of the slide we have an image at radio wavelengths. What we have done with existing telescopes is to make high resolution images which actually identify individually each of these exploding star remnants, as shown in the figure at the top right. In fact, for one of these (TH9) we have resolved it into the shell of material which is shown bottom right, and over time we have watched this expand.
(Click on image for a larger version)
The second example is just a little bit about black holes in distant galaxies. As I have said, these are the beacons of the universe. With existing telescopes we can see the bright examples of these types of sources to very, very distant regions of the universe. The SKA will detect much lower-luminosity examples of this type of object, just about to the edge of the observable universe. So they are going to be a very important probe of what is happening on cosmological distance scales, and will, hopefully, tell us something about the role that massive black holes play in the evolution of galaxies. Because they expend all of this energy, and basically are a mechanism for transporting energy from the centre of the galaxy into the galactic environment, they can potentially have an influence on how galaxies evolve. So that is an interesting thing to look at.
(Click on image for a larger version)
Each of these blue dots is a radio source, and that image was made with a low-resolution instrument. So what we do is to go back with a high resolution instrument and survey each of these tiny little dots. If we can detect it with a high resolution instrument, that tells us that there is basically a black hole at the centre of that galaxy. Looking at the statistics of what percentage of galaxies have black holes and correlating that with other characteristics can potentially tell us a lot.
These are some of the detections we made of the high resolution radio emission from these things. The SKA is going to go massively deeper and therefore pick up many more of these things and give us a much better idea of the statistics.
(Click on image for a larger version)
So that is a bit about the science. I just want to finish by giving you a bit of an idea of the scale of the technology that is required.
When you have got antennas all over the country, obviously you need to connect them up somehow, and that is going to be done, hopefully, with very high speed fibre optic data networks. The sorts of data rates we are talking about are quite mind-boggling. If you have got 100 telescopes, you need to roughly get data at a few terabytes per second from each of maybe 100, or 200, telescopes into a central processor. So just one of these data streams is already millions of times faster than the telecommunications company will tell you that you get with broadband. (It is actually much worse than that.) But a fundamental requirement is actually going to be for a very high capacity national fibre optic network.
Once you have got all that data into a central processing machine, you need to be able to combine all those signals and get the final data output. To do that we simply need to build something which has computing power a few times that of the biggest supercomputer in the world. And, as Naomi said, you have got to locate this next to power and all these other things. So that is a bit of a challenge.
But the job is not finished. You have the final data products and you have got to do something with those, and even though there is a massive reduction in the volume of data going through this processing stage, what you are left with is perhaps an image which you need to manipulate and analyse, and a single image may be a terabyte. Generally, images are stored in memory and you manipulate them in memory, so to hold one image you would need the memory equivalent of 1000 fairly high end desktop PCs.
The types of observations that Naomi is talking about can potentially make hundreds or thousands of images simultaneously, and so you need to deal with that data flow: data archiving and data mining are going to be very important. The SKA is going to be an ICT (information and communications technology) telescope.
So the scale of the problem is huge. How are we going to address it?
(Click on image for a larger version)
We are currently engaged in extending the Australian SKA Pathfinder (ASKAP) across the country. We are building a mini version of the 3000 km SKA.
The ASKAP is going to be built in Meekatharra, Western Australia. Pictured here are some of our existing radio telescopes, including the Mopra compact array near Narrabri, the famous Parkes dish, and one of our telescopes in Hobart. What we are doing is connecting those telescopes up in real time at high data rate, initially 1 gigabit per second but eventually 10 gigabits per second. We are going to combine the signals of the three in New South Wales with the signals from the ASKAP telescope; it is all going to come together in Perth, where we have built quite a large computing facility which will do all the on-the-fly data processing. This allows us to demonstrate (a) that we can do a 3000 km real-time connection at high data rate; and (b) that we can generate large data sets and we can develop the new algorithms and imaging techniques that will be required to deal with the SKA when it comes along even though we are two or three orders of magnitude away from the sort of data volumes that the SKA is going to produce.
(Click on image for a larger version)
In very quick summary: the high resolution SKA, or any high resolution instrument, probes matter and energy under the most extreme conditions in the universe.
The multi-wavelength aspect to all this is very, very important, and the SKA will complement very well multibillion dollar missions that NASA is flying, for example the next general gamma ray telescope satellites.
Finally, we are addressing the technical needs right now, with existing instruments and for immediate scientific yield so, we are trying to extract some science now, as well as prove the technology in order to be prepared for the much larger-scale problems ahead.
That is a pretty broad overview. There is much more information at the international SKA website and also at the Australian SKA website.
