Last updated November 28, 2017 at 2:31 pm
Meet the people making the Square Kilometre Array work. ICRAR-based PhD student David Gozzard is one of the team working out how to synchronize potentially thousands of telescopes based thousands of kilometres apart. Here, he tells the story of how he managed to achieve this monumental task.
Four years ago, I started my PhD in experimental physics. I had just signed up to spend the next few years developing signal stabilization systems for telescopes and space science experiments, and I didn’t really have much of an idea about what that would entail, but I was looking forward to finding out.
During my first week as a PhD student my supervisor was away in Beijing for a “kickoff meeting”. I didn’t know what it was about — maybe something to do with football? I spent the week building a proof-of-concept system to synchronize a bunch of big lasers to shoot down space junk, and trying not to break anything in the lab. I wasn’t 100% successful at that last part.
When my supervisor returned, all the previous plans were thrown out. “We’ve got to get to work on the SKA,” he told me.
The Square Kilometre Array (SKA) is a radio telescope that, when completed, will be the largest and most powerful telescope ever built. It will be able to see deeper in to the universe, and with greater sensitivity, than anything that came before it. It will answer questions about how the first stars and galaxies formed after the Big Bang and how they changed over time, whether black holes and spacetime behave as Einstein’s Theory of General Relativity says they should (or not), where the molecules that are the basic building blocks of life come from, and it will collect the faint radio signals that might reveal the existence of intelligent life beyond our solar system. The scale of the project is so great that ten countries have clubbed together to build the SKA, with smaller contributions coming from another ten countries. All in all, 600 engineers and scientists from 20 countries are working to design and build the SKA, and I had just become one of them.
To achieve the goals its designers have set for it, the SKA will have thousands of radio antennas spread over thousands of kilometres. To make all of these separate antennas behave as one giant telescope, the signals from each of them will be combined in a computer that radio astronomers call the “correlator”. But for this to be effective, all of the antennas need to be properly synchronized.
The distances between the antennas mean that the radio waves from space will reach each antenna at different times and this has to be accounted for by the correlator. For the correlator to do this, it needs all of the antennas to be synchronized very precisely to an atomic clock at the heart of the observatory. The reference signal from the atomic clocks is transmitted to each antenna site through fibre-optic cables.
But there’s a problem.
Mechanical stresses on the fibre cables caused by wind and rain and daily temperature changes affect the fibre and degrade the precision of the atomic clock signal reaching the antennas. While smaller telescopes, like the Australia Telescope Compact Array (right), can cope with the smaller amount of degradation over their shorter links, the vast scale of the SKA means that it needs systems that counteract the stresses on the fibre cable.
And that’s where my signal stabilization systems come in.
Along with my supervisor, Sascha Schediwy, I spent most of the last four years developing a system to transmit the atomic clock signal out to the SKA’s antennas with enough stability to synchronize the array to the precision that is needed. The system imprints the atomic clock signal onto a laser, which transmits the signal through the fibre cables to the antennas. Inside our receiver, at the antennas, is a small mirror that bounces a portion of the signal back through the link on a return trip to the transmitter. The electronics in the transmitter detect the returning signal, which has now seen double the effect of the stresses on the link, compares it to what the signal should have been, and then pre-compensates the outgoing signal. For example, if our system detects that the returning signal is being affected by the expansion of the fibre cables as they warm up during the day, it will adjust the transmissions to cancel out the changes resulting from the stressed cables, and the signal arrives at the antenna crisp and clean and ready to use.
If that sounds complicated — good — because it is. It’s taken me and my supervisor nearly four years to build the systems (right), test them in the lab, then test them on existing telescopes, fix any bugs, re-test, and prove that the system can handle the conditions the SKA is going to be subjected to. The SKA is being built in outback Western Australia and South Africa, which generally don’t have weather patterns that are compliant with the needs of the super-precise telescope systems.
To test the stabilization systems in realistic conditions, I have made two trips to the Murchison Radioastronomy Observatory in Western Australia, where the Australian portion of the SKA is being built, a trip to the Karoo region of South Africa, where the South African portion is being built, and two trips to Narrabri in New South Wales, where I used CSIRO’s Australia Telescope Compact Array as a stand-in for SKA hardware. As well as being important for our experiments, each trip has been a fascinating experience. During our second trip to the Murchison, large storms dumped 45 mm of rain in one night, flooding the area, and cutting us off from the telescope. This meant our test program got cut short, but I got to 4-wheel-drive through the flood waters to get back to town. And at the Compact Array, I even got to drive one of the huge 270-tonne dishes when they were being repositioned!
But back to the important part — these field tests showed that the stabilization system worked on real telescopes, and that it could maintain synchronization between antennas to a precision better than 5 parts in one hundred trillion. To put that in perspective, that’s like two clocks disagreeing with each other by only one second after running for 600,000 years! That should satisfy the needs of the SKA.
But my supervisor and I weren’t the only ones working on synchronizing the SKA.
Because the synchronization system is so critical to making the huge telescope work properly, the SKA Organization decided to retain two separate teams working on different synchronization system designs. If one design turned out to be a dead-end, hopefully the other would not. Our colleagues at Tsinghua University in Beijing have also developed a working synchronization system.
It was time to decide whose system would actually be used to build the SKA.
The SKA Organization put together a panel of experts who poured over every aspect of the two designs, and every test report, to decide which was better. Each system had its own advantages and disadvantages. One was cheaper, while the other took up less space; one was easier to install and use, while the other would be easier to upgrade in the future…
The decision was a bit of a surprise — both systems will be used.
The SKA is actually two mostly separate telescopes: SKA-low, being built in Western Australia, and SKA-mid, being built in South Africa. This was done because no one telescope is perfect for doing all the scientific jobs the SKA wants to do. SKA-mid will be better at mapping galaxies and studying how they have changed over the eons, while SKA-low will be better for looking at the earliest stars and galaxies. The design differences of the two synchronization systems mean that each one is better suited to one of the two sites. Tsinghua’s system will be used on the Australia-based SKA-low, while our system is better for SKA-mid in South Africa.
We celebrated the decision with dinner and drinks at a very nice little restaurant a short walk from the university.
But our job is not done yet. Now we have to take our laboratory prototypes and turn them into systems that can be mass-manufactured. We’re going to need 200 of them for the first stage of the SKA alone. So that’s what we’re working on now, and it means more lab work, more tests, and more field trips to South Africa to ensure that, when the SKA is switched on for the first time, the new view of the universe it gives us will be sharp and stunning.
Now join David to discover what happens behind the scenes at his laboratory: