AUSTRALIAN FRONTIERS OF SCIENCE, 2008

The Shine Dome, Canberra, 21-22 February

Session 1: Discussion

Question: With the stations that will be situated in Hobart, Canberra and so on, obviously the population is larger and so you are going to have interference from mobile phones et cetera. Will this interfere with your data collection? If so, are you going to be able to address that?

Steven Tingay: Basically there are technical requirements on the radio frequency interference environment. Those requirements are much stricter for the majority of the antennas that are located in Western Australia, and the requirements will be less restrictive for these outlier stations. There are certain reasons why that is. One is that if you have got interference on just one antenna, it won't actually correlate with the signals on another antenna, so it sort of disappears naturally. The other thing that we can do is to use some clever on-the-fly processing at each station to excise some of the radio frequency interference before it even gets into the system. That is going to be a big area of research and will, hopefully, get implemented on all of the stations.

Question:  Naomi, you showed comparative images, at two different resolutions from older data and newer data, of the leading arm crashing into the Milky Way disk. You said that the next generation will provide a greater resolution by a factor of 10 or more. You can see from the old and the new data that the clouds are breaking up into smaller bits. Do you expect that to keep happening? And what does that say about gas in the Milky Way?

Naomi McClure-Griffiths: That is a very interesting question. So far, every time we have gone to higher resolution the clouds break up into more and more little clouds, and we have never hit the end point where things are cascading down into smaller and smaller scales. But whether or not there is an end point to that actually gives us fundamental information about how energy travels around in the galaxy. So if we can detect the bottom point where things stop breaking up into smaller and smaller clouds, then we learn something about the galaxy. And if we don't, then we learn something about a different type of physics. It is an unknown.

Question: The interaction of the charged particles with the magnetic fields that is governing the structure of the gas clouds is very interesting, and you can see that directly. But I wonder about the other 95 per cent of the 'stuff' out there – namely, the dark matter, which surely will exert very strong gravitational interaction with the things you are trying to look at. Are you going to learn something about that? It is all very well to focus on the 5 per cent, but what about all the rest?

Naomi McClure-Griffiths: Well, dark matter and then dark energy are the Holy Grail of astronomy, and we have a lot of things in the science case for the Square Kilometre Array which say that we are going to get information about these aspects.

We do see the gravitational interaction of dark matter, and particularly looking at atomic hydrogen in other galaxies we see from its rotation that there must be something out there to define its gravitational effects. So we do have some information in that sense, not onto the nature of the item but onto the effect that it has on the other items that we can see.

We can't image what is not visible, so the best that we can hope to do is to look at as many different things as we can see and, hopefully, infer the effects of the other bits of the universe that we can't see. But that is going to bring hard problems.

Question: Does atmospheric turbulence or do inhomogeneities in the ionosphere affect, or frustrate, this sort of astronomy attempt?

Steven Tingay: Yes, they sure do. At low frequencies, below about 1 GHz, the ionosphere and turbulence in the ionosphere, variations with the ionosphere with time of day and annual cycles and all these things are the main effect. Above a gigahertz it is usually the wet atmosphere component that affects the data, and it affects the data by introducing the phase change to the signals at the different antennas, which you want to remove. All you want is the phase signature of the background universe, if you like. So there are pretty standard calibrations techniques that we use to remove those effects.

The flip side is that you can actually learn quite a lot about the ionosphere and the atmosphere after you have done your calibration, because what we are essentially throwing away is information about the turbulence in the atmosphere.

Question: The data that you are talking about sharing is incredibly large. In biology, and in particular in structural biology, we get pretty scared about even minute amounts of data in comparison. With synchrotrons we are talking about trying to put on-line a gigabyte or two of data, and we're having huge difficulty in our community in actually doing that type of thing worldwide. Really, I think, we could learn a lot from the way you guys even currently manage this type of data.

So how do you deal with, for example, things like legacy and keeping data alive and high quality?

Steven Tingay: I think the scale of problem as we have described it is currently, today, not very feasible at finite cost. I think it is fair to say that Moore's Law is our friend – or we hope that it is going to be our friend. But you make a very good point, that there are other disciplines dealing with similar problems, perhaps not quite of the same scale. So I think this is one of the major points in which a cross-disciplinary approach could benefit everyone.

Naomi McClure-Griffiths: Just a bit more on that: although the data rates that we have now are not exceptional by SKA standards, the movie of the cube that I showed was 32 gigabytes. All of our datasets are on-line in an archive that comes off the telescope so that people can access that through crude tools, and could be done a lot better. But there are a lot of efforts around the world associated with something called the Virtual Observatory project to try to keep data on-line, accessible with tools by anybody sitting at their desk doing astronomy. If they want some bit of data that came out a telescope 20 years ago in remote Australia, they can get it from London. But at present I think we're not doing the best job at transferring that around, but we are starting to talk about it and think about it.

Question (cont.): Do you have a central resource, or is this a case where Parkes, for example, set up their own data sharing system and London has a different way but it all kind of works together?

Naomi McClure-Griffiths: It is basically the latter.

Question (cont.): Ash Buckle is leading a similar effort in terms of structural biology, hopefully Australia-wide, and we are finding a similar thing: we have to just set something up and see if it works.

Naomi McClure-Griffiths: Yes, and that means that you reinvent the wheel every single place that you're doing this. Somebody learns how to interact with one archive and they don't know how to interact with another archive, and so there are limitations to the model of everybody setting up their own. The idea of a centralised repository is a wonderful idea.

Question (cont.): Maybe we should have a joint structural biology-astronomy talk some time.

Naomi McClure-Griffiths: Yes!

Steven Tingay: We don't blink at terabytes any more. We might blink at thousands of terabytes. But setting up a many-terabyte archive is really pretty easy these days. The really difficult thing is effective global sharing of that information, and so the major point of infrastructure that I think needs to be addressed most critically is the data networking. You want to be pumping this data along at the very least multi-gigabits per second pipes, and getting access to even that level of infrastructure now is really quite difficult. And it's going to be a key technology for this.

Question:  This is not big science, this is huge science – thousands of antennas stretching over thousands of kilometres. I can understand the scientific case for this very well, but at some point you have had to interact with politicians and make a social case for this. I am curious to know what the social case is.

Steven Tingay: Well, we all like to know where we came from. It depends on what sort of level of social case you are talking about, but at least what I find from going out and talking to people is that astronomy is a massively well-subscribed subject in terms of people's interest. I don't think we are really going to have a problem selling it at that level.

Sometimes I think we go over the top a little bit in selling this as 'huge' science, because it is a couple of billion Euros – which, yes, is a lot of money, but that is a typical NASA planetary mission, funded out of a single agency. So I think we need to be a little bit careful of that, and not oversell it on that basis. The dollar number is not super-frightening. It is some of the new technology that we need to develop which is slightly scary.

Anne Green: We are already engaged with the funding agencies. Another issue, which has not been raised so far, is that there are quite likely to be a lot of R&D spin-offs. This is a science symposium, but there are a lot of technologies that come from developments such as in astronomy and space exploration. (One of them is mobile phones, which are now anathema to radio astronomers!) It isn't just to say that the science we are going to do is, 'Are we alone? Where do we come from? Where are we going?' – those big questions. There are many other ancillary benefits that are coming from that.

Naomi McClure-Griffiths: We have also used education quite a lot. The kids around my neighbourhood were upset when Pluto was demoted as a planet. Astronomy is something that kids seem to get interested in very easily, and so we have used that as a selling point to say, 'Look, we don't expect that we are going to train up a million astronomers, but if we can get some kids interested in science by using this tool that they already think is cool, then we might be able to make more scientists in the future.' And so we have mechanisms laid down in the SKA and ASKAP plans to let kids do observations, to interact with the telescope, to be able to use it as a 'gigantic experimental toolbox' that they would otherwise never have access to. It is a kind of a multi-pronged social approach.

Question: It wasn't clear to me whether SKA or your Pathfinder mission is going to have radio polarimetry capabilities.

Naomi McClure-Griffiths: Both will have polarimetry capabilities. We are setting the specs for the Pathfinder now, and they say we should have very good polarisation purity and be able to do very good polarisation experiments. That technology is, of course, untested, and we hope that we will make certain that that works on ASKAP and then apply it across to SKA. But, for both of them, it is a really key driver that they have polarisation capabilities.

Question: Will you be able, possibly with the high resolution capabilities, to look at other nearby galaxies, in particular elliptical galaxies, and try and observe the gas in ellipticals with a bit more detail, to ascertain how much is there and how different the evolution in those galaxies is from our own galaxy, which is a spiral galaxy?

Naomi McClure-Griffiths: Yes, absolutely. I think the key there is having the bigger bucket of collecting area. We have got more sensitivity with our new telescope so that we can try to look for these very small amounts of gas that we would be looking for in the elliptical galaxies, to compare them with our own. But that definitely is part of the science case for both of them.

Question: Is there a similar venture occurring in South Africa?

Naomi McClure-Griffiths: Yes! (I was waiting for that question.) At the same time, we are working hand-in-hand and neck-and-neck with South Africa, which is building a prototype telescope as well – which has different strengths and weakness. They are designed to complement each other. It would be stupid to build two identical Pathfinder telescopes in the southern hemisphere, so we are trying to make certain that we don't overlap too much there. And that has been really quite an interesting thing, because it has really invigorated the astronomical community in South Africa enormously, to start building this Pathfinder.

Steven Tingay: Well, on the long baselines there are fundamental differences – Africa is a north-south continent, and Australia is east-west. In particular, when you place the constraint on Australia that you don't want too many antennas in the tropics, and therefore you avoid the north of the country, there are some implications there. There are real implications for the quality of imaging that you can do. So I think that is the major difference between the two sites from the long baseline aspect.

Anne Green: When those two sites were shortlisted, there were four sites which were proposed and those two were considered scientifically and technically acceptable. (We were not allowed to rank them, but they were both considered acceptable.) What happens next is an interaction between the funding agencies and the astronomers, and we want to make sure that we choose the best site. But Naomi is right: at the moment it is a win-win for astronomy in Australia, which has a very long and proud history in radio astronomy, and we are helping the South Africans. They have a big technology gap, but they are very keen to catch up.

Question: With the very high angular resolution that you are seeking, what are the implications of the fact that your observing platform – namely, the continent – is not stable at those resolutions, even at quite high frequencies? And to what extent do you need to know the movement of your observing platform relative to the stars?

Steven Tingay: Yes, there are implications. Basically, to make an observation you need to know accurately the position of the thing that you are looking at, and you also need to know accurately the locations of all of your antennas. If antennas are moving around in a relative sense, due to continental drift or seismic activity or any local effects to do with artesian water or anything like that, you have got to make a correction for that. Similarly, you can calibrate out some of the effects of the ionosphere, using relatively standard techniques. And, although it is a bit crude to state this, from those corrections you can basically then go back and refine the positions of the antennas. So, as well as learning a lot about the universe billions of lightyears away, we can simultaneously learn about the ground under our feet and what the continent is doing.

Historically there are two streams of this high resolution science: one is astronomy and the other one is geophysics. And you can, in a sense, do both simultaneously. We basically like to have our antenna positions located with an accuracy of a few millimetres or a centimetre in the global reference system.