Astrophiz 175: Greg Sleap ~ Astro Data Wizardry

LISTEN: https://soundcloud.com/astrophiz/astrophiz-175-astrodatawizardry

SUMMARY: In this fabulous episode of Astrophiz, Greg Sleap, the Software and Systems Team Leader at CIRA, the Curtin Institute of Radio Astronomy, (which is a node of ICRAR, the International Centre for Radio Astronomy Research) discusses the Murchison Wide Field Array (MWA) and its role in capturing radio data from the cosmos.
The MWA is a low-frequency radio telescope array located in remote Western Australia. It is used for various scientific studies, including the search for the signals of the Epoch of Reionization, detecting black hole radio jets, transients, GRBs, FRBs, Pulsars, SETI research, studying supernova remnants, and monitoring space junk.

The MWA has been successful in producing over 300 journal papers and is preparing for future developments, including the use of all 256 tiles, and collaboration with the Square Kilometer Array (SKA) telescope project.

AUDIO TRANSCRIPT

BRENDAN: Welcome to the Astrophiz podcasts! My name is Brendan O’Brien, and we’d like to acknowledge Australia’s first astronomers, the Aboriginal and Torres Strait Islander people, the traditional owners and custodians of the land we are on.

This episode is produced on Yorta Yorta and Pangarang country.

Right now, I have an amazing guest to introduce to you. Today, you’re going to discover and understand the data wizardry involved in capturing the zetabytes of astro data from immense radio telescope arrays, correlating and channelling that data to supercomputers, and then making it available to researchers all over the world, who in turn create all those stunning images that enable us to understand our previously unseeable universe.

This guy, Greg Sleap, and his multidisciplinary team, are absolute wizards at their work!

BRENDAN: Hello, Greg.

GREG: Hi, Brendan.

BRENDAN: Today I’m really pleased to be speaking with Greg Sleap. Greg is the Software and Systems Team Leader at CIRA, the Curtin Institute of Radio Astronomy, which is a node of ICRAR, the International Centre for Radio Astronomy Research in Perth, Western Australia. CIRA manages the Murchison Wide Field Array Project, known affectionately as the MWA, a low frequency radio telescope array of spider antennas in remote Western Australia, which is one of three precursors to the Square Kilometer Array telescope, the world’s largest radio telescope.

The SKA will require the collection, correlation and distribution of unimaginable amounts of radio data pulled in from the cosmos.
For those who love numbers, the SKA-Low telescope will generate about eight terabytes per second, 24 hours per day and 365 days per year, which is then analyzed by astrophysicists all over the world.
And it’s Greg’s job to make sure they get it.
So thanks for speaking with us today, Greg.

GREG: No worries, Brendan. Thanks. It’s a pleasure to be on the podcast.

BRENDAN: Excellent. So before we look at your world of data and your current work for ICRAR at CIRA, can you tell us where you grew up, please, Greg, and where your curiosity and drive to understand and manage data came from?

GREG: Sure. I was born in Canberra, in the Australian Capital Territory in Australia. It’s nice and cold there, not too far from where we are now. And, yeah, as a family, we moved across to Perth when I was about six years old for the warmer climate. And, yeah, I sort of had this fascination and love for space, astronomy, stars, science fiction, ever since I can remember, really. And I think one of the things that inspired me was that I was given a book as a kid, probably when I was about four or five years old, called “You Shall Go to the Moon”. And the book would have been written in the 60s sometime, because it was full of diagrams and things, and the story was basically unfolding about the first trip to the Moon. And of course, obviously in 1969, that became a reality. So this book was predating that.

But I fell in love with it. And I fell in love with the idea that space is a place, it’s not just a background of stars that we see when we walk out and look up at the night sky. It’s real places you can go to and explore. And I must have had my grandmother read that story to me millions of times. I just fell in love with it. And the final page of the book actually is this father and his son standing on the moon in their spacesuits, and they look up and point at Mars, and the dad says, the next stop will be there.

And that really was an exciting way to end the book. So I think definitely my love of space and astronomy absolutely started back then.

BRENDAN: Wow! That’s great. Greg … It sounds like that book was a bit prescient because collectively we’re all doing the same thing right now.
Okay, so please tell us a little bit about those early school days and your earliest ambitions and how those ambitions might have evolved Greg.

GREG: Yeah, so the love of space and astronomy was sort of always in the background, but I quickly started to also have a second love, which was computers. So my dad would bring home from work what was called a Luggable computer. So this is before there were laptops and notebooks and things. So it was basically a briefcase-sized computer that was portable. And he would bring that home, do some work on it, and when he was finished with work, I would ask if I could have a play on it. And he would set me up so that I had something simple like a text editor or something like that. And I would spend hours just playing on this computer.

And that really sparked my love of computing. And even early in primary school, we were one of the first schools to get a BBC Microcomputer. You don’t see them anymore these days. And that computer had a whole range of software on it for kids games and puzzles and that sort of thing, and some programming as well. And so starting year four, I really realized that this is the thing I want to get into.

And so during my school years, we gradually got computer labs at school and at high school. And I would spend a lot of my time writing little programs, modifying existing programs. When I finally got a computer at home, which was Amstrad, that’s another brand you don’t hear of anymore. I would get my mum and dad to buy me computer magazines that had listings of programs in there. And you would actually sit there and type them out into the computer and then run them. And it would be like a game of Solitaire or Pacman or something like that.

And I just loved doing that, and I loved modifying the programs and seeing what I could do if I changed them. And all this time as well, I was also knee deep in astronomy magazines as well, drilling over all of the shiny telescopes and all the beautiful images that were coming back from space, including one of the biggest impacts on me was the Voyager Two mission. When we first saw images of Uranus and Neptune for the first time … ­­­that really had a big impact on me.

But I think probably the biggest impact throughout high school actually was Star Trek The Next Generation. I’m a big Sci-Fi nerd and have always been, but when Star Trek Next Generation came on the little screen, I think it honestly changed my life. It really changed the way I thought about what science is and what engineering is. Star Trek upheld all these values of what it is to be a good person, what it is to actually be in a society that values knowledge and intelligence.

And of course, they’re exploring the stars in spaceships and shooting phases everywhere.
So what was there not to like? Really?

BRENDAN: Fantastic Greg. I share quite a few of those experiences.
For me, I think my favourite is “Blake’s 7” and the computer known as ORAC.
But let’s go on with your career. After school, you completed your Bachelor of Commerce with a focus on it at the Curtin Business School in Perth, Western Australia. And some years later you did your master’s in Astronomy at James Cook University in Townsville, up in Queensland, right up on the other side of Australia, 5000 km away.
But before we talk about that, would you like to mention some of the people or circumstances that influenced your shift from the world of business to astronomy?

GREG:  Yeah, sure thing. So when I got out of university, I’d really gone fully into the IT side of my passions. And at the time I thought that that was really the only place I could take it. So going into corporate IT and working as a software developer would be the only way that I could actually exercise that passion. And I didn’t think there was any chance of me getting any job in astronomy since I wasn’t an astronomer.
And I didn’t realize that there were all these careers possible that didn’t actually require you to be an astronomer but you could work within the field. So initially I didn’t make that connection. But in the early 2000s, there was more and more information about the Square Kilometer Array project which Australia was bidding to host. And that got me thinking that if I did want to start to combine my love of IT with my love of astronomy, that I’d need to actually do something to separate myself from the average applicant. So I realized more and more that modern astronomy is data based astronomy and computing based astronomy.

So to get my foot in the door, I thought, well, what’s better than combining It and astronomy by getting some sort of qualification in astronomy? So I found the Astronomy Master’s course and James Cook uni was where I finished. But I actually started at University of Western Sydney. The course started there and then we transferred. So doing the Masters in Astronomy was a way to get my foot in the door and to be noticed when there are a bunch of resumes out there for some sort of computing job for the SKA, for example.

And having that Master of Astronomy, I thought, would actually help me quite a bit. Interestingly enough, the course was external, and so it was fully online while I was working full time. But I loved every minute of it because my passion in astronomy had been very, I guess, surface level. I hadn’t really dug into it too much, and I never really did the high level maths and physics at high school. So for me, this course was brilliant. It was a ‘Master’s by coursework’, and we covered everything from particle physics, the electromagnetic spectrum, planetary science, stellar evolution, galactic evolution, cosmology, quantum mechanics, just the whole gamut. And I learned so much, and I actually kind of kicked myself for not going for not taking the physics and maths in high school and actually pursuing maybe a more direct astronomy career.

And the interesting thing is that at the time so this is back in I graduated in 2005. The funny part of that course was that there was only a small portion of the syllabus devoted to radio astronomy. So radio astronomy was still quite a niche field, I guess, within Australia at that time. Obviously, there were some prominent radio astronomy observatories operating, for example, like parks, but in this course, it was mainly focused on optical astronomy.

But I am very glad I did it, and it really made me more determined to try to find my way in and get an astronomy job.

BRENDAN: Okay, very good. Look, let’s go into that a bit. You’ve mentioned getting your foot in the door in astronomy, so let’s go back a bit. And you started off as a coder, then a software developer, systems analyst, server management, sys admin, IT team letter in the business and mining industries. But then you landed at the MWA, the Murchison Widefield Array, as their Data Manager. How did you land that job?

GREG: Greg yeah, so the job itself came up in quite an interesting way. I didn’t actually realize that the job was available. So I’d been sitting on the couch, not having a good day, and I realized that on my calendar, on my phone, it was telling me that the current University Astrofest was on that night. And I was feeling lazy, and I thought, I can’t be bothered going anyway. Then I thought about what else I would be doing that night. And at the time there was nothing else going on, so I thought I should just go to Astrofest.

So I took myself down to Astrofest and came across this strange looking piece of mesh on the ground with 16 spider shaped antennas all clutched to it and a little box where all the antennas were connected to. And so I asked the gentleman that was tending that weird thing. What the heck is this and what do you do? Anyway, it turns out that the person that I was talking to was a member of the MWA Operations Team and that we were standing right next to one of the MWA tiles.

Anyway, he went and explained to me all about the MWA and how it worked, and then he asked me some questions about my background. What do I do? So I was explaining that I work in the mining industry, but in an It capacity, and I deal with lots of data and business intelligence and all that sort of stuff.
And he then said, “Oh, well, you’d sound perfect for the job. Why don’t you apply?”
…. And I’m like, “What job?”

So he then pointed me to the relevant job ad on the Curtin University website, and I applied, and that’s how I ended up here in the MWA operations team.

BRENDAN: Yeah, that’s a great story! Okay, so you started off as their Data Manager. And to me, that sounds like an oversimplification of your role that you had there originally, because I looked at your CV, and the job involved software engineering network, admin server, admin procurement, deployment, the management of CIRA’s IT system, all with the purpose of supporting various large scale big data radio astronomy instruments.
Now, I’m pretty sure you didn’t spend all your time wrangling a keyboard Greg.
Would you like to remind our listeners about what the MWA is? And you just mentioned the ‘Spiders’. What did the MWA look like out on site when you went there for the first time you visited? Was that … in what … 2015? 2016?

GREG: Yeah. So I think my first trip up to where the MWA is, which is the Murchison Radio Astronomy Observatory, or shortened to the MRO, which has recently received a name from the Wajarri Yamatji people of “Inyarrimanha Ilgari Bundara”, meaning “sharing sky and stars” in the Wajarri language.

BRENDAN: That’s so cool!

GREG: Yeah. So that’s the site where it’s located. So it’s about three and a half hour’s drive northeast of Geraldton. So for people tho don’t know Western Australia, that’s basically in the middle of WA in terms of the North and South and a few hundred kilometers East from the coast as well. Out in the desert, you’re talking red dirt, lots of heat in summer, so 50C° degrees plus it can easily get to out there and quite cold in winter as well, overnight. It is an amazing place!

So that was my first impression of it. Yeah. 2016 or so went up there because whenever we get new staff, we try to get them up to see the MWA for itself so they can actually get an understanding of the physicality of the instrument.
And when I went up there, the spiders looked a lot smaller than they do when they’re up close, when you’re looking at them out the window of the car or whatever. And we were tasked, actually, with working out why one of the tiles wasn’t working. So a tile is just basically 16 of these spider shaped antennas grouped together. We grouped them together in a “tile”, and so we had to work out why one of these tiles wasn’t working anymore. So we drove out there and then stomped across into the bush for a bit and looked down and saw that the Beamformer box, which is the little box that takes all of the 16 dipoles signals together and allows us to point the telescope in any direction without actually any moving parts.

So that box had been completely broken and ripped out of its location. And around the place where the Beamformer box was lying were these what looked like kangaroo tracks. So we surmised that a random kangaroo had just hopped along, got its foot or leg caught in the cables, and because they’re powerful creatures, just ripped them out and went off on its merry way.

BRENDAN: Okay, Greg thanks. Look, can you just paint the picture of what ‘is’ the MWA, the Murchison Wide Field Array?

GREG: So the MWA, which also, by the way, has a Wajarri Yamatji  name, which is Gurlgamarnu (pronounced ‘Golga-mahn’) meaning “the ear that listens to the sky” which I think is super cool.

BRENDAN: Yeah!

GREG: So it’s a low frequency radio telescope. So that means that we’re looking for radio waves, or we’re sensitive to radio waves that are in the meter scale.
So you’re looking at sort of 1 meter to 2 meters long in wavelength.
And we have 144 of those tiles that I mentioned before.

We actually have 256 deployed, but we can only use 144 at a time, and that’s a limitation of the number of receivers we have. So the receivers we currently have … we have 18 deployed, and they’re very specialized, quite expensive bits of equipment. And the job of those is to take the signals coming in from eight of the tiles and then they digitize that and split it into lots of little frequency channels, in fact, 256 of them.

The astronomer that is currently operating the telescope selects 24 of those 256 and we throw the rest away. The reason for that is it’s just too much data to deal with. And 24 channels is actually quite a nice amount for most of the science cases that we have. So, just getting back a bit, the telescope itself covers the same band as FM radio and TV frequencies. So it’s really important that we minimize any sort of interference which can completely contaminate any sort of signals that we’re receiving from space.

The signals we’re receiving from space are so weak that even just a transmission from many hundred kilometers away can easily overwhelm any sort of signal we’re getting from space. And that also includes things like WiFi, Bluetooth and mobile phones. So when we’re up at the Murchison, we’re inside a radio quiet zone where you’re not allowed to operate any of these radio devices except in an emergency. And that has really helped to make sure that the MWA takes really good data that’s very free of interference, and that lets us actually do a lot better science.

The MWA was actually conceived and funded and built by a collaboration from about 20 different institutions around the world. So we’ve got member institutions in Australia, USA, China, Japan and Canada. And previous members that we’ve had have included New Zealand, India. So it is a truly international project. And Curtin Uni gets the lucky job of being the lead organization to actually operate and maintain the telescope on behalf of the collaboration.

So keeping the radio environment quiet is super important. So much so that CSIRO provide a control building up near the MWA, which houses a lot of CSIRO servers and equipment and also ours. And it’s actually inside a Faraday cage. So to enter the building, you’ve got to go through an airlock to get in and out of the building. And that’s basically you going in and out of the Faraday cage, which basically blocks any stray radio transmissions from going outside of that room.

And it’s very important to mention as well that the MRO actually is not just for the MWA. Of course, the CSIRO, which is the Australian Government’s Science Agency, they have their own telescope up there called ASKAP, which is the Australian SKA Pathfinder. It’s basically made of 36 dishes. And they are also a radio telescope as well. But they’re operating at a different frequency than us, so they’re at higher sort of frequencies, and we’re in the low frequency end.

There’s also ‘EDGES’, which is a little instrument. It’s actually quite modest looking, but its job is to look for the Cosmic Dawn, and that’s operated by Arizona State University. And of course, the Murchison Radio Astronomy Observatory is also the home to the future SKA-Low as well. So you have this one location, it’s radio quiet, and we’ve got all these instruments up there taking advantage of the beautiful remoteness of the location.

The other thing that I did notice when I first went up there was all the different flora and fauna up there. Kangaroos, as we’ve mentioned, there’s wedge-tailed eagles, ants, snakes, Bungarras, which are monitor lizards. And of course, it’s very hot and dusty. But still, it’s an amazing place. And every chance that I get to go up there, I take it, even though most of the time I’m working inside the control building, installing servers and plugging in cables and things of that.
But it’s still just an amazing experience.

BRENDAN: Fantastic. And can you mention some of the science that’s happening that’s coming out of the MWA?

GREG: Yeah, so the MWA is quite a versatile instrument. It’s got quite a wide field of view, and we can actually process or analyse data in many different ways. And so that allows us to do a lot of different science. Some of the primary science that the MWA was intended for is searching for the signal of the Epoch of Reionization. And this is where I’m not an astronomer and probably can’t explain it very well, but effectively it’s the light from the first stars in the universe. So we’re talking 13 billion years ago, which, because they were emitted back that long ago, the universe has expanded since then.

And as the universe expands over those 13.7 billion years, it’s actually redshifted the light from the visible spectrum down into the radio spectrum. So the wavelengths of light have actually expanded. And when you’re talking wavelengths that are in the meter to two-meter range, that’s right bang where we’re sensitive to with the MWA.
So the MWA is actually a really good Epoch of Reionization machine.

BRENDAN: Fantastic Greg!  Does the MWA see black holes?

GREG: Yeah, Brendan. So the MWA is really good at detecting the radio jets that get emitted from supermassive black holes at the heart of most galaxies. They come out as big streaming jets, sort of above and below. And that’s something that we’re really good at detecting. And in fact, a lot of the sources that MWA detects as part of the GLEAM Survey are actually the jets coming out of the supermassive black holes at the heart of galaxies.

But the MWA also does a whole heap of other science as well, including detecting supernova remnants. There’s also Transients, so that’s like pulsars and fast radio bursts. The fast radio bursts is quite a new area for the MWA. And we haven’t had any detections yet, but we’ve had a program of searching for them and we’ve been able to put some limits on what the FRBs would look like through the MWA. And there’s also lots of studies going on with the sun.

The sun is a big emitter of radio waves, believe it or not, and the ionosphere as well. So we can detect ‘plasma tubes’ that are in the Earth’s ionosphere with the MWA.

BRENDAN: Cool.!

GREG: And there’s also a fledgling SETI commensal search going on which uses basically a copy of the MWA data. So when the MWA is observing any other source, a copy of that data goes into a special set of machines which run some SETI pipelines which look for any sort of signals or anything that might be of interest coming from advanced civilizations.
They’re called ‘Technosignatures’. And I’ve been working with the team on that one. So that’s been quite an exciting thing to work on. And MWA is also doing some ‘Space Situational Awareness’. So the really interesting thing is that if you have a radio transmitter, a powerful one, say in the local town of Geraldton, which is about 350 km away, that transmitter is transmitting in all directions, including up.

So when you have any sort of satellite or let’s say some space debris from a rocket that has been spent in orbit, any sort of debris like that, these radio waves will actually bounce off those pieces of debris. And we can detect that with the MWA. And so that allows us to effectively monitor the sky for space junk, which is becoming an increasingly massive problem for launch operators and satellite operators in being able to find a bit of orbit that isn’t contaminated with space junk that could potentially damage or destroy spacecraft.

So I guess that’s a pretty good roundup of all the science. The MWA does Surveys as well, so that’s where you basically point at all of the sky that you have access to, which for the MWA is most of the southern sky and a little bit of the northern sky. And so one of the surveys there that I mentioned before was GLEAM. That’s probably been the biggest survey for MWA and the most used by other astronomers.

But there’s work going on a thing called the SMART Survey, which is basically a southern sky survey of all the pulsars. So the idea is to look for all of these pulsars that’s predicted to be many finds, hundreds, I think, and they’ve already discovered a couple through this survey and it’s going to be a treasure trove of data for scientists to pour through for many, many years!

BRENDAN: Yeah, fantastic!

GREG: And of course, I guess the final thing with the MWA is that it’s going to be very useful for the upcoming SKA-Low. So as the SKA-Low gets built, the MWA will be used to test the SKA-Low, so there’ll be observations done on both telescopes and they’ll be comparing the data that comes back from both and making sure that the SKA-Low is actually working properly, which actually is a lot trickier than it sounds.

BRENDAN: Thanks, Greg. Yeah, GLEAM is a fantastic survey.
The imagery that’s come out of GLEAM is just beautiful. If people want to go and have a look at that, just look up some GLEAM images, you’ll be blown away.
Now, Greg, in previous episodes, we’ve interviewed many ICRA scientists who have used your data from the MWA and from ASKAP arrays to make some beautiful scientific discoveries. We’ve talked with Phil Edwards, with Melanie Johnson-Hollit, Stephen Tingay and Natasha Hurley-Walker, Karen Lee-Waddell and there’s many other scientists there who are leading the MWA and the ASKAP projects.

And even your colleagues, you’ve got an engineer, Mia Walker. If people want to go back and listen to our Episode 167, Mia does a fantastic job and so many of our listeners will have some background knowledge of your data collecting instruments. But before we talk about big data and the Pawsey Centre and the development of the future SKA could you bring us up to date with where is the MWA right now?
How is the infrastructure developing?

What does the data flow look like now? And what lies in the future for the MWA?

GREG: Yeah, great questions. So the MWA is operating great. We have 256 tiles deployed, but we are only using 144 at the moment. But we’re working on getting some new receivers which will allow us to utilize all 256 tiles at once. And what that does is that gives us a lot more sensitivity and also it allows scientists to have a bigger range of baselines.
Now in radio astronomy, when you’ve got an interferometer where you’ve got lots and lots and lots of antennas or dishes, the baselines are a really important factor as to the performance of the telescope.

So if you’ve got long baselines, which is just the distance between any two of the antennas or tiles, so long baselines allow you to see with greater resolution, and short baselines let you see more emissions and nebulous features. And so having a mix of these baselines is really great as well. So once we get to 256 tiles, that will really be unleashing the full potential of the MWA. We’ve been producing all of this great science in all the science areas, and there’s been over 300 papers published on MWA data and many thousands of citations of those papers, which really just goes to show that the MWA is really pulling its weight when it comes to furthering our knowledge of the universe.

We’ve only recently installed a new Correlator for the MWA called MWAX. And the Correlator is a computing cluster that sits up at their Murchison and it takes all of the data from all the receivers. And so each receiver has eight tiles. So we basically get all the data from the receivers. And what the Correlator’s job is to split the data into more fine channels. So you’ve got lots and lots and lots of these little fine channels and it takes the signal from every single tile and multiplies it against every other tile.

BRENDAN: Wow!

GREG: So you can imagine this is a really big computational machine. And what that allows us to do is to basically record what the common signal is from all of the tiles. So when you think about a radio source that might be overhead of the MWA, then all of the tiles should actually see that signal around about the same way and same amount of power and same phase, which is just basically talking about where the wave actually strikes the antenna.

But if you’ve got a source of interference that’s off to the east somewhere, for example, so maybe it’s a car driving past. And we’ve actually seen that cars spark plugs can actually cause interference. So whenever the spark goes off, we see a little spike in the data. So we don’t even want cars to be driving past when we’re observing. And so that spark would come from one direction. And so the tiles that are on that side of the array closest to the car will see that signal more.

The tiles that are further away see that signal less, or they see it with lower power. And so when you do this cross multiplication, those sort of stray signals don’t actually get as much power when you add up all of the signals from all of the tiles. So probably Radio Astronomers listening to this, they’re probably shaking their heads at my explanation. But from an IT guys point of view, I’m happy with that.

So that’s generally what we use the Correlator to do. And I was privileged enough to work on that project. So I was in charge of making sure that the data moves between all the different stages of the Correlator and then once it gets out of the Correlator, then it goes down to our archive at the Pawsey Centre.

The other things that I’ll just mention as to why the new Correlator was so important is, apart from going to 256 Tiles, it also let us operate in all these different modes, which we couldn’t do before, which gives astronomers more choice as to how they want their data.

And it allows us to capture special triggers. So other observatories, especially space-based observatories, might detect some big flash of like a gamma ray burst, a GRB for example. And we know from the study of gamma ray bursts that after the initial flash of gamma rays, there’s a radio emission as well. And that radio emission is usually delayed by a short amount of time because lower power than the gamma rays and the radio rays take a bit longer to propagate through space.

And so we get a trigger from that electronic trigger that tells our telescope that GRB has gone off and to go and point there and take some data so that we can look at the radio emission from the GRB.

So it’s things like that that are really cool.

And also, as I mentioned before, the SETI back end is what’s called a commensal instrument. So it takes a copy of the data and does its own thing. And we’re working on some fast radio burst commensal pipeline as well, so there’ll be another copy of the data that will go and just be there to feed machines that will be looking for fast radio bursts.

And so some components have been out in the field for ten years.
So 2013 is when we started operations. So they’re doing really well, considering, although we do have quite an active field work team who go out there and replace components all the time. We get kangaroos, we get lightning, we get ants who decide that it’s a really fun thing to go up into electronics and then they end up shorting out electronics sometimes as well.

And finally the MWA archive, which is more my thing.
I look after, we’re sitting at 48 Petabytes storage and that’s at the Pawsey Centre and that’s on a mix of tape and disc. And that’s been collected over the last 10 years. And we try to keep as much of that as possible because we store what’s called raw data. So there’s other observatories who don’t have the luxury that we do storing the raw data. So they store science products, which have been processed in a very specific way. And as you go along those processing pipelines, you’re effectively narrowing the reuse of that data, but it massively reduces the size of it. We’re able to keep the raw data and that means that astronomers who initially requested the time on the telescope to take data can process it the way that they want for their science case.

And then once it’s out of the embargo period, then other astronomers can come along and say, well, ‘Hang on a sec. There’s this beautiful data for this target that I’m interested in, but I need to process it in this other way’.

And they can go off and do that. And in fact, one of our researchers is leading a team, … Natasha Hurley-Walker. She’s leading a team that is looking for transients in the MWA archive. So this is just going back through old data that was used by other people for other things and then finding if there’s any radio source that blinks on and blinks off. And she’s already discovered a couple of these things with some Nature papers published about them as well. And there’s still a bit of a mystery because these things blink on and off with strange periods that don’t sort of conform to the current menagerie of things that could be producing them like pulsars, magnetars and other things. So that’s pretty exciting stuff!

BRENDAN: Sensational! Beautiful science!

Well, the MWA has certainly stamped itself in the world of science. Look in your role as CIRA Software and Systems Team Leader. I know that you don’t specifically work for the Square Kilometre Array, but can I ask a couple of questions about the future developments there of the SKA?
What were the big lessons that you’ve learned in the development of the MWA that are going to guide the work in getting the SKA data systems online?
And secondly, I believe the contracts have been awarded and construction of SKA-Low has begun out on that remote site in the desert and scrubland.
And do you know what the construction timeline is and when can we expect the first science data to flow into the SKA?

GREG: Sure. So in terms of the lessons learned, I guess there’s many and they range all the way from the physicality of the mechon and the harsh environment and how to make equipment that can stand the test of time and to, to withstand the elements that are out there. One thing I had mentioned before is that as well as all the heat and the dry and the dust, we also get floods up there too. So when it rains, the rain has nowhere to go and the ground is very hard. So it just ends up forming these river channels that uh, that run through. And so our equipment needs to be able to withstand all of that as well. So that’s probably one of the, the big lessons and it’s one of the lessons that CIRA has been working with the design team on what’s called the AAVs, which is the Aperture Array Verification system.

So it’s ah, one of the pre-construction tasks of the SKA. And so Kyra engineers have been working with a group of Italian engineers to design what the SKA-Low antennas will look like and some of the field electronics and all of that stuff is stuff that they’ve been taking into account and learning from us and going out there with our group to experience those harsh conditions.
Aside from that, in terms of the data challenge and the processing and computing challenge, the SKA scale is a huge jump up from the likes of the MWA. MWA is big enough and keeps us very busy, but SKA is another whole level altogether. So one of the lessons we’ve found with managing the MWA data with quite modest hardware, I guess if you can divide and conquer, then the job is much easier.

Yeah. So instead of building one big computer, we split the job up so that it can run on lots of smaller computers.
So for example, our Correlator that’s up at the MWA, MWAX, it consists of 24 servers, each with lots of rRAM, lots of discs and GPUs. So that’s graphical processor units, which you know, you might find in your home computer. These ones are data centre grade ones. And we process those 24 channels I spoke about each channel goes into a different server. So that way we didn’t need to build one massive computer that could handle 24 channels. We build 24 computers that can handle one channel each.
And so that for us was one of the ways that we can manage with this huge flow of data.

The other thing that we’ve learned, and I think this is pretty common with big data projects, is that you try to keep your processing and your code near the data. Like, don’t try to move the data to where you can process it because it’s just too big.
And the infrastructure to move petabyte uh, scale data sets is just not there.

We, for example, have a 100 gigabit link that goes from the MWA down into the Pawsey Centre and that’s a big link. So a hundred gigabits is the equivalent of about a thousand home NBN connections or home internet connections at a hundred megabits.
So that’s a big connection there but that is still nowhere near enough for the SKA data.

BRENDAN: Amazing!

GREG: We also spent a lot of time optimising the software and hardware that we have. So if you just use things out of the box, they’re gonna be suboptimal. So we spent a lot of time testing, a lot of time assimilating with test data before we actually put anything up at the Murchison.
And there’s many lessons, for example. Yeah, keeping the archive as long as possible.
Um, for us at the MWA, lots of hidden treasures in there that we haven’t discovered yet.

However, with the SKA, the current funding model I guess only really allows the final data products to be stored, not the raw data. And that’s just because the data sets are just too big. And with the cost of storage and moving that size data, what they do is they’ll get the data coming out of the telescope, apply processing to it, which the astronomer who has requested time on the telescope will have input into.  And then at the end of it, the astronomer will get the exact sort of science product that they’re after.

BRENDAN: Yep.

GREG: If they’ve made a mistake, they’ll have to re-observe with the SKA telescope.

BRENDAN: Yep.

GREG: With the MWA however, if you make a mistake processing the data, then you just process it a different way again using that same raw data.

BRENDAN: Yep.

GREG: And so that’s the, the big difference there. But with the SKA, it’s just not possible to use that model. So that’s probably where the MWA and SKA differ quite significantly in how we deal with that data.

BRENDAN: That’s so cool!

GREG: Answering your second question about the schedule and the timeline, the SKA organisation, the SKAO, has published a, I guess, a preliminary schedule of construction. Obviously there’s big caveats because as we know, there’s uh, shortages of workers in the world at the moment post pandemic. There’s also supply chain issues. So at the moment, the idea is that by about early 2024, there should be six stations. Now in SKA terminology a station is the same as what we would call a tile. So a SKA station consists of 256 antennas, whereas one of our tiles is 16 antennas. So there’ll be six stations in early 2024, and then the year after that it’ll grow to 18 stations, two years after that, 256 stations. So that’s quite a big leap in those two years.

BRENDAN: Whoa! Yep!

GREG: And the current plan is that by late 2027, all the stations should be deployed, which will be the full SKA-Low phase one array, and the ‘construction complete’ milestone, I guess, would be nearer to mid or late 2029.

So it’s gonna take quite a while to build and commission and throughout the whole construction process, as each of these milestones is reached with 6 then 18 then 256 stations, the data will actually be flowing throughout that whole time.
It won’t be for end users and scientists, it’ll be for the commissioning team who are there to make sure that the telescope actually works. And there might be some early-adopter astronomers who put up their hand to spend time to debug things, to analyse where there might be issues.

With telescopes like this, it’s incredibly complex and there are lots and lots and lots of wires if you can imagine. And so all it takes is for, you know … a … a wire that should be plugged into X to be accidentally plugged into Y and suddenly those signals are now reversed. And those sort of issues are, are really subtle and difficult to detect. So there’s gonna be a lot of work for a lot of people to iron out all those little bugs and to get the thing working. But it will all be very much worth it when the first science comes through, which I’m not sure exactly when that would be, but I’m guessing that would be around that construction complete time is when the first sort of production science would be occurring.

BRENDAN: Oh! what an astonishing project! You must be so lucky to be working on the MWA, um, as a SKA precursor Greg. Okay. So that’s the background, the projects, the people, the telescopes, the antennas. Look, you’ve been hinting at the magnitude of the data and how it’s managed. And we are probably talking about Zettabytes of data coming in for just the SKA-Low alone. Then you’ve got all of those instruments over in South Africa, the other part of the SKA, can you talk about that big picture of what happens to the data after it’s sucked out of the sky by all of those beautiful antennas, those dipoles that’s correlated with your WMax Correlator and how does it end up on the screens of astrophysicists all over the world? You’ve hinted at it, how do you move such vast amounts of data so far and so quickly and maintain its integrity?

GREG: Sure, yeah. So as, as you mentioned, the … the data coming through the tiles then goes to receivers, then goes into the Correlator, then out of the Correlator. We then have a 100 gig link. It actually goes first to Curtin University. We’ve got, uh, some servers there which uh, maintain some temporary storage. So the idea is that if there’s any issues with the next step in the chain, which is pushing the data into the Pawsey archive, then it’ll just fill up the discs that are sitting in Curtin Uni, then from Curtin, there’s another hundred gig link that goes across to PAWSEY. And once it goes into Pawsey, it goes into one of two archiving systems. So either Acacia, which is based on disc or Banksia, which is based on tape. And we’ve got allocations in both and we shuffle between the two depending on the capacity that we have and where the data is best suited.

The stuff that’s on tape has a little bit of a longer latency to get it out again when an astronomer requests it, but it’s actually pretty quick, usually within sort of 10 minutes. And the Acacia storage is discs, so it’s very fast. So you, you put it in there and you can get it out again reasonably fast. So once it’s sitting in the archive, users around the world, uh, use our MWA all-Sky Virtual Observatory ASVO portal. And that portal is a web-based interface and we also have command line interface as well for people that, uh, love to do the command line stuff and not anything on the web. And that allows users to select or to find the data first and then to go and download the data. And optionally we can do some kind of pre-processing steps to it. So the sort of things that the astronomer will have to do because it is raw data, we have to pre-process it and we can do it for them or they can do it themselves and they do it using supercomputers.

Generally the, the data even at the MWA level is too big for users to process on their own laptops or on their own desktop. Yep. So PAWSEY has various supercomputers available. So there’s GALAXY, uh, which is the MWA sort of dedicated one, and there’s also SETONIX as well. But we have users all around the world. So for the users that work on the Pawsey systems, the data doesn’t actually move very far to get to them. And that’s actually really great. So that’s keeping that concept of keep the code near the data or keep the data near the code. So users that are using Pawsey systems have this benefit of the fact that they’re doing their analysis on computers that are right next to the data. But we have researchers all the way around the world including Japan, USA, China, et cetera. So they actually need to either be able to apply for and, and get computing time at Pawsey.

But if they can’t then they’ll need to actually download the data across the wire. And depending on the internet connections between Pawsey and their supercomputer, that can be either really nice or really painful depending on their institution and the bandwidth that they have available to them. But we’ve found that even like a sort of a regular data set for MWA that has been pre-processed, which means it’s also been reduced a bit, ends up around the 100 to 300 gigabytes. And we’ve found that most of the institutions that work with MWA data do have reasonably high bandwidth internet connections, which means that they can download that size of data in hours, not days.

BRENDAN: Yep.

GREG: Which helps, and I guess this is the, the contrast in between MWA and SKA. So that’s sort of how the MWA astronomers get their data and then process it.

So they work on a supercomputer that does it.
 with SKA, most of the same process applies. So the data goes from the antennas through various other equipment through a Correlator and it will eventually come down to Pawsey in the science data processor, or SDP, so this is an SKA-Low or SKA computing facility which will do a lot of the pre-processing. It will also do some actual processing. So it will actually go and create images for example, or spectra  image cubes, various other products that the scientists use. And then from there the data will be transmitted across to various SKA regional centres. So the budget for the SKA project itself did not include money to fund the final step of processing the data for astronomers.

BRENDAN: Oh.

GREG: So all of the member countries are having to put up their own money and funding to build these SKA regional centres that will both house the data and process the data for astronomers.

So this is a big job and there’s currently one of the team members that works on my team who is also working with the SKA Regional Centre project for Australia … That’s the Aus SRC. And they are um, finding a fascinating project to work on because they’ve got the difficult job of making all this look easy for the astronomers. So they’ve gotta take this data deluge that will come out of  Pawsey for the SKA-Low and there’s an equivalent for the SKA-Mid in Southern Africa.

So they’ve gotta take this data deluge and process the data and basically run a supercomputing facility and design how the software will work, how users will interact with the system. And so all of that is all happening right now so that they’re ready in six or seven years time for this to all come online. And just answering your other question about how we move such vast quantities of data around and so fast, all of that is thanks to high speed networks around the country and around the world.

So within Australia, AARNet is the provider of educational research networks and they’ve basically built a 800 kilometre fibre link between the Meurchison and Perth. And that’s where we send all of our data. The link is actually been built specifically to have capacity for the SKA. So the amount of glass that’s in the ground because fibre optic is basically glass is approximately uh, several terabits per sec as I understand it. It may actually be more now, but certainly as of a few years ago it was a few terabits and one terabit is a thousand gigabits a second.

BRENDAN: Whoa!

GREG: We only use a small fraction of that. Uh, so we just use the a hundred gigabits. But yeah, the SKA load will need all of the bandwidth they can get to get the data down to Pawsey. The other thing we do is we ensure that we haven’t lost any data by taking checksums of the data as soon as we produce it. And a checksum is a mathematical computation done to the data that you can use to then read back the data and recompute the checksum. And if any single bit of the data has changed, the checksum will be different. So as long as the checksum that you computed when you created the data, matches the checksum of the data at the other end. So that could be at Pawsey or if you’re a researcher getting the data onto your supercomputer to start work on it. As long as those checksums match, then we know that the data has not been modified, corrupted, or anything bad happened to it.

BRENDAN: Okay. Checksums are fingerprints. Okay, thanks Greg. <laugh>. Now one thing we haven’t talked much about on this show is the Pawsey Centre. You’ve referred to it a few times now. Will the Pawsey be part of the data infrastructure for the SKA and or will they have to install other supercomputers to process all that SKA data?

GREG: Sure, yes. So Pawsey will be a key element of the SKA-Low process.
Pawsey has been a long time partner and supporter of the MWA. In fact, we sort of say that we grew up together. So the Pawsey Supercomputer Centre started around the same time as the MWA began and the MWA gradually ramped up the amount of data the goes through Pawsey as Pawsey was upgrading its systems to handle that data.

So over time we’ve sort of both grown together and they’ve been such a, a great supporter of the MWA and basically allowed MWA to maintain this massive archive, which of course bears dividends with researchers being able to access that archive and make discoveries. So the science data processor is the component of the SKA that will likely reside in Pawsey. And I’m not 100% privy to this, but I believe that they would be building new infrastructure to support that. So new supercomputers potentially, just because it is such a huge computational effort to do the job of the SDP, which as I mentioned before was pre-processing and processing of the data to get it to a form that can then go to the SKA Regional Centres for the astronomers to access and work on.

BRENDAN: That’s astonishing Greg. And it’s starting to … I’m starting to realise that what we see in those beautiful radio astronomy pictures that astrophysicists produce, there’s so much work and so much technology and so much energy and so many talented people beavering away in the background to make all that happen. I never realised how immense that backend of the MWA is.
So thank you very much Greg. But let’s now bring our listeners up to date with some of the finer points of your current work. So could you tell us some details of a particular problem that you are working on now that’s driving you crazy or is astonishingly beautiful and exciting or perhaps it could be all of those things. What’s going on Greg?

GREG: <laughs> Yeah, so … so I lead a team of of five. So I’ve got four software developers working with me and an ex-astronomer who has turned software developer. And we do a whole range of things as you sort of mentioned before. Everything from system administration of servers and networks all the way to programming, um, supporting software, web development, managing data, all of these things. So there’s a lot of stuff going on and I think the thing that drives me crazy is that I wish I had time, more time for everything because there’s a lot of things that we, we are dealing with.

But I think probably some of the most exciting things that we’re doing at the moment is we’re assisting with the SETI pipeline for the MWA. So that’s actually, uh, a project initiated by a much larger project called Breakthrough Listen. So Breakthrough Listen is a worldwide programme funded by Yuri Milner

BRENDAN: Yep.

GREG:  To …  to tap into the world’s observatories and see if we can find any evidence of techno-signatures from advanced civilizations.

BRENDAN: Yep.

GREG: And so working on that project, I think I really love doing that because you just never know what you’re gonna find. And it’s also a challenge from the IT side because you’ve gotta process a lot of data in a very specific way and for the most part in real time in order to keep up with the data deluge coming out of the telescope. So that to me is exciting and cool. Another thing that we’re working on as well is that currently our, our web portal to download data. So the MWA ASVO does the really good job of download or providing the raw data to users and pre-processed data, but we don’t have any pretty pictures. So many other astronomical observatories have data portals where you can download beautiful pictures.

BRENDAN: Yep.

GREG:  So for us, that’s all done by the scientists usually in their own systems. And so as a data portal we don’t have access to that. What we want to build is an imaging processing step that users can optionally use to get a pretty picture from the data that they’re processing. And in some ways that can be a really good diagnostic tool because if the picture looks correct, then they know that they can spend the many hundreds or thousands of computer hours to process the data properly. So I’m selling it as a data quality tool, but really it’s because I love the fact that we can make pictures out of radio astronomy images.

BRENDAN: Yes!!

GREG: … And I think probably the other thing that’s exciting is, is that project to expand the MWA to use all of the 256 tiles. It’s going to generate four times more data than we had at 128 tiles. And so that’s a data challenge in itself just to be able to keep up with that data rate and make sure that all of our systems can handle that deluge of data.

BRENDAN: Fantastic! Okay. But how you manage both the excitement and the expectations and the deadlines obviously that you’ve gotta navigate on a project such as the MWA?
And I know it’s not gonna evolve into the SKA-Low … that will be separate, but how do you manage both the excitement and the expectations that are on you to get those milestones met, Greg?

GREG: Yeah, well I guess the first thing I would say is that I love my job. So although there can be some time pressure sometimes and, and maybe a, a feeling of having a, a big load of work, I rarely feel stressed as such. And part of that reason as well is because I know that that what I’m doing and what my team is doing is effectively trying to make the lives of our researchers easier and let them make these discoveries and find out cool things about the universe.

And so that’s a really big motivator for all of us. I think another thing that helps is the fact that the work that my team does specifically we’re … we’re a small team but we’ve actually got a whole bunch of people that help us in numerous ways. So for example, the Pawsey Centre, as I mentioned, without them hosting the data, we wouldn’t have an archive.

And also the supercomputers as well for the scientists to process the data on AARNet transports the data Curtin IT help us a lot. So we have a lot of our servers, our web service and databases and things are hosted in Amazon Web services and Curtin IT help us manage those. And of course we’ve got this massive collaboration of, I think the number is about 270 members at the moment of the MWA collaboration who all have various skills in various things that they’re not just astronomers. All of them have talents in computing programming and … and all sorts of things which help us. And we also have a really great management team that help take a lot of the admin off us so that we can get on with the job and do the technical things that we’re asked to do.

BRENDAN: Fantastic! Yeah, I’m beginning to see what a huge enterprise that the MWA is.
Okay. Look … back to you as a person, Greg. We’ve talked about your roles, your responsibilities about the nature of that beautiful science instrument, the MWA. I saw that you’ve got an interest also in astrophotography and we are lucky to have such dark skies way out there in WA’s  Wajarri country and right here near Wangaratta, I think we are both on Yorta Yorta and Pangarang country.
How’s the astrophotography going?

GREG: It’s going great! I love it. I haven’t had much time recently, so I’m the proud father of a 15 month old baby girl Harriet.

BRENDAN: Lovely name.

GREG:  Yeah. And she keeps us very busy at the moment. So we’ve also had very cloudy skies here recently too. So I haven’t had a chance to grab the ‘scope out for a while. But I currently have uh, an eight inch Schmidt–Cassegrain telescope and a 73 millimetre refactor. And those two basically work in concert to allow me to, to get nice widefield images. So Nebula for example, and the eight inch Schmidt–Cassegrain telescope is really good for planetary stuff. Jupiter, Saturn, Mars, the moon, and also for the smaller targets like galaxies. And I’ve got a mono camera with a filter wheel so that I can do red, green and blue and, and a couple of other filters as well that are really good for nebula. So that’s your hydrogen alpha sulphur II and oxygen three. And so with those last three filters you can actually make what’s called the Hubble Palette images.

BRENDAN: Cool!

GREG: So it’s very similar to what Hubble uses. And if you look at the Pillars of Creation, the Eagle Nebula pictures of Hubble, they’re often done in the Hubble palette as well. So I’m able to recreate some of those … nowhere near as good as the Hubble though, but, but at least I’ve got the same colours for the targets. So yeah, so I’m really loving it and being out here in Yorta Yorta and Pangarang country, the skies here are so dark compared to when I was living in Perth, whereas basically in the middle of suburbia and all of the light pollution. So it’s really beautiful being out here. Although it is a bit colder than Perth, I’ve gotta admit

BRENDAN: <laughs>. Fantastic Greg. And I’m sure you will get back to Astrophotography once Harriet’s a little older. But in the meantime, I’m sure you having a wonderful time with your very young child. Thank you so much. Now we’re almost running out of time, Greg, the mic is all yours and you’ve got the opportunity to give us your favourite rant or rave about one of the challenges that we face in science and engineering or in your own field of data management or perhaps our human quest for new knowledge. The microphone’s all yours, Greg.

GREG: Thanks. Since I’ve got a soapbox, I think this is my biggest bugbear at the moment or biggest worry actually, is just the fact that at the moment we all seem to be living in a world where information is in a bubble or in our own bubbles of information where there don’t seem to be clear facts, at least facts that everyone agrees on.

And I think the internet has contributed quite a bit to that. But I think the internet has also contributed a lot of good to the world. But I think at the moment as humans, we’re not really doing a good job at handling this technological marvel that we’ve created and being able to maintain what we would, I guess describe as, you know, what are the facts in the world and how do we deal with the big issues that are affecting us.

For example, climate change, pandemics, you know, there’s Information Wars all over the place and Culture Wars regarding these topics. And I feel like if we were able to actually all agree on the same facts, then I think the world would all probably realise that we need to act on these things before it’s too late. And you know, with my daughter Harriet, I just think of the world that she would be growing up in and I do worry that we’re not gonna be tackling climate change, especially before we make irreversible damage to the planet. Sorry to … to go on a … on a dark note there … but that’s … that’s currently what keeps me up at night.

BRENDAN: Exactly! And it’s something we all have to face together and we owe a duty to those young children that they have a planet to live on. Is there anything else we should watch out for in the near future? Greg, what are you keeping your eye on?

GREG: Well, obviously the SKA, so I’m following that very closely and … and it’s very exciting to see the very first steps of progress towards construction. So I would recommend anyone that’s interested in, in space and astronomy to keep an eye on the SKA because that’s gonna only keep on progressing.

Also, the Australian Space Agency started a couple of years back and are working towards increasing the capabilities of Australia for things like launching and tracking of satellites. So that’s very exciting as well. Further afield, you’ve got the James Webb Space telescope, which I’m pouring over the images every time they come down. And of course the NASA Artemis project as well. So I believe that the first test launch was successful and they’re gonna be launching again, I think it’s later this year or maybe early next year. And we’ll actually see a return to the moon for humanity, which is pretty exciting. And we’ll also have the first person of colour and the first woman setting foot on the moon as well.

BRENDAN: Cool!

GREG: There’s also multi messenger astronomy, which is becoming bigger and bigger every day, which is basically where you combine the astronomy that you would do in your regular domain. So your … your electromagnetic spectrum, so that’s radio waves, optical ultraviolet, infrared, gamma rays, X-rays, microwave. And you combine that with gravitational wave observatories data.

And in fact, just recently there was a big announcement about the Gravitational Wave Background. So this is the background of gravitational waves caused by super massive black holes in the early universe. And they’ve been able to be detected because they’ve been able to use precise pulsar timing over the last 15 years to see really minute changes in the pulsars timing and they’ve made this discovery and I think we are gonna see more and more big discoveries from this combination of, sort of regular astronomy with gravitational wave astronomy.

So that’s absolutely an exciting field to be watching.

BRENDAN: Yep.

GREG: And one of the other things that I thought was pretty cool is that there are currently ideas to put a radio telescope on the far side of the moon. So because it’s on the far side of the moon that never actually faces the earth, you have basically got the entire moon behind you blocking all of the radio frequency interference so you can have an ultra sensitive radio telescope. And I think I would definitely put my hand up to go to one of those field trips and help build it.

BRENDAN: Fantastic, Greg. Yeah, and I suppose we can imagine the moon as a very big Faraday cage?

GREG: Absolutely!  

BRENDAN: Yeah. Okay. Look … we are out of time now, Greg, thank you so much.
On behalf of all of our listeners and especially from me, it’s been really fabulous and what you’ve laid out in front of us is really astonishing. Thank you, especially for your time. It’s been a great pleasure to get to know a bit about you and your work. Maybe we’ll catch up some time for a a coffee in Wangaratta …and for those who’d like to find out more about SKA-Lpw, you can go to tinyurlDOT com forwardslash SKALOWOZ.
And for the MWA you just go to mwatelescopeDOTorg and you can follow Greg on Twitter.
He puts things up there as @Paladinsmeg.

So good luck with all your projects and with all your next adventures, especially that adventure with young Harriet.
Thank you, Greg!

GREG: You’re very welcome, Brendan. Thanks so much for having me on the podcast. It’s been great fun.

BRENDAN: <laughs>. Okay, see you.

GREG: Cool. Thanks a lot Brendan. Catch you later mate. Bye.

< Short blast of Morse Code>

BRENDAN: And remember … Astrophiz is free and ad-free and we always recommend that you check out Dr. Ian Musgrave’s Astroblog and Southern SkyWatch websites. And in two weeks time at the start of the month, we’ll be bringing you Ian’s August SkyGuide here on Astrophiz.

Keep looking up!

<Music fade out>