Each night, 4,000 antennas belonging to the MWA in remote Western Australia scan the southern sky for solar flares and other cosmic phenomena. In under three years of operation the telescope has acquired more than seven petabytes of data – the equivalent of 90 years of HD video.
With a new-generation radio telescope and supercomputing might, light from the Epoch of Reionisation and the first stars – when the Universe transformed from an opaque realm – may soon be in the sights of Curtin astrophysicists and astronomers.
With science rapidly extending our journeys into the Universe, many of us now gaze at the twinkles of starlight emitted billions of years ago – but no more than 13.7 billion years ago – and wonder what and who else is out there.
Australia is an international leader in radio astronomy research, and, recognising the advances our scientists are making in this field, the Australian Research Council (ARC) is supporting home-grown, awe-inspiring projects for cosmic discoveries. After being awarded a $1 million ARC grant in late-2015, the Curtin Institute of Radio Astronomy (CIRA) has hummed with the anticipation of extending their next-generation radio telescope that will bring scientists 13 billion years closer to seeing the formation of the first stars and galaxies, and the Universe’s ‘Epoch of Reionisation’.
Known as the Murchison Widefield Array (MWA), the telescope is a $50 million international science endeavour from Curtin and 16 organisations across several countries. From its location within the radio-quiet zone of the Murchison Radio-astronomy Observatory (MRO), the MWA has been online and scanning the southern sky since 2012. It’s not a traditional, rotating parabolic dish, but an ‘aperture array’ telescope comprising 4,000 antennas that resemble knee-high robot spiders, coupled to a formation of receiver systems.
Precisely arranged over seven square kilometres of Australian scrubland, the antennas scan the night sky for luminous objects such as exploding supernovae, and massive flares from planets in other solar systems. And because it has no moving parts, the MWA is a highly agile instrument that can be steered to a new source of interest – a cosmic gravitational wave, for example – within 10 seconds of CIRA being alerted.
Big, fast data
The colossal volume of data from the antennas, which streams at 320 gigabytes a second (GBps), demands the supremacy of supercomputers to create a comprehensive picture of the night sky. However, because storage is unfeasible the data must be processed in real time, which leads to another challenge: the limitation on power available for computation, dictated by the observatory’s remote location. To overcome this, the MWA archive team successfully exploited the capabilities of graphics processing units – commonly found in games consoles and smartphones – and reduced the input data rate to a manageable size of 5 GBps.
After collation at the telescope’s correlator, data are streamed to the storage facility at the Pawsey Supercomputing Centre in Perth, via a 10 gigabits per second (Gbps) fibre optic link. There, the centre’s Galaxy supercomputer provides offline processing of data and its distribution to the University of Melbourne and international research centres including the Massachusetts Institute of Technology in the US, the Victoria University of Wellington in New Zealand and the Raman Research Institute in India.
The information is then in the hands of astronomers and physicists working to unlock the Universe’s infinite secrets. MWA-acquired data already underpins more than 30 science research papers, with the project’s impact resulting in the CIRA team in 2015 receiving a prestigious Thomson Reuters Citation and Innovation Award.
The new grant from the ARC will launch MWA phase 2, involving quadrupling the telescope’s footprint to 28 square kilometres via two new arrays of antennae tiles. These will instantly escalate the calibration and sensitivity of the telescope and improve its capability tenfold.
This also means a corresponding increase in the difficulty of data-processing, explains Dr Randall Wayth, Director of the MWA project. “By expanding the size of the array we see the same sky but with increased resolution, which means we have to output data from the correlator at four times the current rate.” Consequently, in addition to its engineering team that is working on the new arrays and enhancements to all components, CIRA has 12 postdoctoral researchers dedicated to science data processing and algorithm development. And necessity being the mother of invention, Wayth adds, the team is already working on innovative ways to compress, store and distribute the data.
Their mission has been made easier by the recent trial of a new 100Gb/sec high-speed data link from the MRO. Completed by researchers at the Cisco Internet of Everything Innovation Centre at Curtin, the link means that once signals are converted to light at the observatory they’ll travel as light all the way to Perth. The trial also verified the digital data packets were arriving on time and in sync, which is critical for the billion-dollar Square Kilometre Array low-frequency radio telescope (SKA-low) also to be located at the Murchison observatory.
Phase 2 and the Epoch of Reionisation
Following the Big Bang, the Universe was briefly in an ionised state before cooling into clouds of hydrogen gas, and the ‘dark ages’ descended. About 500 million years later, Wayth explains, the first luminous objects began reionising the universe into the transparent one we see today, where most of the hydrogen is ionised or locked up in stars.
The MWA is designed to detect a specific wavelength of electromagnetic radiation emitted during this Epoch of Reionisation (EoR), which will lead to a fount of discoveries about the entire cosmic dawn.
Between the Earth and the feeble light of the first stars, however, lies the entire Universe – 13.7 billion years of astrophysical objects and their radio emissions. Black holes at the centres of radio galaxies accumulate and swirl matter around them in accretion discs, and eventually eject the matter in jets larger than their host galaxies. And old stars are exploding here and there, leaving glowing remnants of their former glory.
“The EoR signal is less than one millionth the size of the foreground noise,” explains CIRA researcher Dr Natasha Hurley-Walker. “And right on our doorstep there’s the Milky Way, giving off a constant radio glow as its huge magnetic fields arc through the interstellar plasma.”
Hurley-Walker’s work for the past three years has been the GaLactic and Extragalactic All-sky MWA (GLEAM) survey that is identifying the radiation that camouflages the EoR. The project’s statistics are staggering: two million CPU-hours spent processing 500 terabytes of data. Because “one astronomer’s foreground is another’s science”, the survey is the most downloaded and processed set of observations.
“GLEAM has led to a catalogue of 300,000 newly observed galaxies, and given us a better look at bubbles that encapsulate young stars, for example,” she says.
Another major advantage the MWA has over dish telescopes is its wide field of view – the entire Southern Hemisphere sky. Together, this feature and the MWA’s high sensitivity enable in-depth studies of compact objects that emit radio signals of precise regularity.
Internationally recognised in pulsar astronomy, CIRA astrophysicist Dr Ramesh Bhat is studying intergalactic phenomena that last only milliseconds – such as radio emissions from fast-spinning pulsars, which will help lead to the detection of ‘ripples’ in the fabric of space–time known as gravitational waves.
The study of such exotic objects require recording and processing data at very high time resolution, and with aggregate data rates of 30 terabytes per hour, it’s the most data-intensive functionality of the MWA. This is, however, an important investment, says Bhat. “Pulsars and gravity is a headline science theme for the SKA, given its potential to uncover new vistas in extreme physics,” he says.
While the MWA is in itself a science celebrity, the SKA is the long-awaited main attraction – the largest and most progressive telescope ever conceived – with the global science community anticipating the SKA’s venture into the ‘Cradle of Life’. With an ability to peer inside dust rings that surround young stars and see how other Earth-like planets form, the SKA could detect sources of radio transmissions that indicate the existence of intelligent life elsewhere in the Universe.
However, the SKA cannot proceed without the accomplishments of the MWA. In addition, the MWA will be operating in conjunction with the Aperture Array Verification System – an international SKA-Low prototype spearheaded by Curtin. According to Professor Peter Hall, Co-director of CIRA and Engineering Director at the International Centre for Radio Astronomy Research, this will enable Curtin to deliver critical design review information for the SKA.
Curtin’s end-to-end capability is rare and valuable in the international SKA community, says Hall. “We’ve pursued a carefully planned and successful path in low-frequency astronomy and engineering, marrying the development and operation of the MWA with science and technology trailblazing for the SKA-low.”