Speaking of Hubble...

SKA: The Sound of Data

July 6, 2012 by Alberto Conti
Artist's impression of the SKA dishes. CREDIT: SKA Organisation/TDP/DRAO/Swinburne Astronomy Productions

Artist's impression of the SKA dishes. CREDIT: SKA Organisation/TDP/DRAO/Swinburne Astronomy Productions

SKA is coming. No it’s not the Jamaican music genre of the late 50s, even if it might sound like music to radio astronomers’ ears. SKA is the Square Kilometre Array, a radio telescope currently in development. SKA is set to revolutionize radio astronomy, perhaps in the way the Hubble Space Telescope did for the visible-light portion of the electromagnetic spectrum.

SKA, as the name implies, will be able to collect radio waves over approximately one square kilometer or a million square meters. The signal from thousands of antennas spread over about 1,900 miles (3,000 km) will work together as one gigantic, virtual instrument capable of extremely high sensitivity and resolution, continuing radio astronomy’s tradition of providing the highest-resolution images in all astronomy. Coupled with its operational frequencies, SKA’s sheer size will make it 50 times more sensitive and 10,000 times faster at surveying the sky than any other radio telescope currently in existence. In addition, thanks to new technological advances, SKA will even enable multiple users to observe different pieces of the sky simultaneously!

This amazing instrument is being designed to address some of the most pressing questions in astrophysics, fundamental physics, cosmology and particle astrophysics:

  • When and how were the first stars formed?
  • What is the nature of dark energy and dark matter?
  • What and where are the conditions for life?
  • Was Einstein correct about General Relativity?
  • Where does cosmic magnetism come from?

The James Webb Space Telescope will also address some of these questions, and astronomers are already thinking about how these two observatories can be used together to provide deeper insight into cosmic objects. With such broad scientific objectives, SKA has the potential of truly transforming the exploration of the universe at radio wavelengths. However, the astronomical community faces huge challenges stemming from its construction and from the sheer volume of data it is expected to produce.

To maximize the potential for new discoveries, SKA needed to find a suitable site. The southern hemisphere of our planet seemed an ideal place given its minimal radio interference, and last month a split decision was announced by the SKA organization members — Australia, China, Italy, the Netherlands, New Zealand, South Africa and the UK — to build core receivers in South Africa, Australia and New Zealand.

Australia and New Zealand will host its core site at the Murchison Radio-astronomy Observatory (MRO) in Western Australia, with the most distant stations located in New Zealand. South Africa will host another core site located about 75 km north-west of Carnarvon, with distant stations in Botswana, Ghana, Kenya, Madagascar, Mauritius, Mozambique, Namibia and Zambia. Each site in Australia and South Africa will host SKA arrays at increasing distances and decreasing density. The densest region, known as the core, will contain approximately half of the total collecting area of the SKA arrays in a circle of just over 3 miles in diameter (about 5 km). A mid-region extending to about 112 miles (180 km) from the core will host arrays placed randomly at an ever-decreasing density from the center. Finally, an outer region extending to about 1,900 miles (3,000 km) from the core will comprise five spiral arms along which dishes, grouped into stations of 20, will be located. The separation of the stations increases towards the outer ends of the spiral arms in a pattern, borrowed from nature, which maximizes spatial coverage (see how and why in this great talk by Dr. Rick White).

Layout of the SKA antennas on the ground. CREDIT: SKA Organisation/Swinburne Astronomy Productions

Layout of the SKA antennas on the ground. CREDIT: SKA Organisation/Swinburne Astronomy Productions

The first phase of construction will take place from 2016 to 2019, with the goal of providing about 20 percent of the total collecting area at low and mid frequencies. This will ensure that SKA can start operating well before full construction is complete. The rest of the construction will be finished by 2024.

However, there is another dimension to the challenge that is SKA. Such a large number of arrays (potentially up to 4,000), with the need to sample celestial phenomena at ever increasing spatial and temporal resolution, will require very high performance central computing engines and a capacity that rivals the current global Internet traffic. In fact, once the raw data is processed, SKA will be able to produce in a single year an amount of data comparable to all the traffic of the entire Internet in 2011!

This is entirely new territory. This huge increase in scale requires a revolutionary approach — not only to traditional radio telescope design, but also to data storage and handling technologies, and computational resources.

SKA will require numerous advances in computing technology:

  • Computers capable of 1,000 more computations per second than current supercomputers
  • Software that mimics the behavior of our brain with millions of connections and billions of computations
  • Improvement in power efficiency to allow large number of machines to run at full capacity constantly

SKA’s future therefore hinges on the project’s ability to forges partnerships with both research and industry leaders in high-performance computing. The payoff for industrial partners seems obvious: an ideal test bed for systems that process large volumes of data from geographically dispersed sources with extreme energy requirements. The hope is that SKA’s spinoffs will spur technology development in areas like high-performance computing and large data-storage warehouses, but also renewable energy generation.

Let the music start: SKA!