A compromise resolution for the Square Kilometre Array radio-telescope project was announced over the weekend.
The two bidders for the $1.9 billion project, an Australia-New Zealand consortium and a consortium of African nations led by South Africa, will now split the project between them.
This leaves New Zealand on the outer edge of the project and it is doubtful that any radio dishes will be installed in this country at least for the first phase of the project.
Computerworld flagged the possibility of a split-site option earlier this year, and quoted project director for the Australia-New Zealand bid, Brian Boyle, as casting doubt on the viability of such a decision.
“A significant split of the SKA site by either frequency or collecting area is not supported by the SKA science case,” he said at the time.
“Moreover, the SKA Science and Engineering Committee has considered the arguments for splitting the array on a number of occasions and found no scientific justification to do so – indeed it concluded that a split array would damage the science of the SKA. In addition, there would be an inevitable cost overhead of constructing and operating two major sites rather than one.”
Australia and New Zealand will jointly be involved in the low-frequency side of the wide-ranging observation project, and New Zealand may not get any dishes in the first phase. New Zealand’s science representative on the SKA board, Melanie Johnston-Hollitt, told Radio NZ that “depending on the technology choice for the low-frequency component, we might see dishes extend from Australia into New Zealand, but we have to go through a design phase to determine whether or not that’s the case.”She reiterated that view to Computerworld.
Johnston-Hollitt was a member of the “options working group” charged specifically to look at the scientific validity of splitting the site. After a week meeting in the Netherlands, the working group decided, she says, that a split could be advantageous, because some investment has already gone into building infrastructure in Australia and South Africa. “The science we can do [with the split option] is more than we could have done with the single-site option in Phase 1,” she says.
The working group discovered that by incorporating the existing infrastructure the project could “do more science” more cost effectively than with the single-site option “which would have meant obliterating all the investment on one of the sites,” Johnston-Hollitt told Computerworld.
“That was an unexpected result,” she adds, and changed previous views on the viability of the joint site.
Even under the single-site plan, New Zealand would have been very unlikely to have hosted any dishes in Phase 1, she says.
For Phase 2 of the project, deploying thousands of dishes, Johnston-Hollitt says, the placement of that equipment will depend on countries’ performance in handling Phase 1; so there is a chance of greater NZ participation.
The original plan for the low-frequency component was to have 250 dishes that would only have been able to communicate one reading – one pixel – at a time. CSIRO in Australia has been working on “phased array feeds” that can survey a portion of the sky at once. These will now be incorporated into the two-site configuration. “That’s something in addition to what we would have had under the original single-site plan,” Johnston-Hollitt says.
Experts are looking positively on the possibility of siting major data processing facilities in New Zealand to handle some of the huge data flows - as much as 100 terabits/second – which will come from observation of radio sources over the whole sky, including some of the furthest and oldest stars known.
New Zealand is well-placed for such a role, having already set up a supercomputer to participate in the Murchison wide-field array project, a precursor to the SKA, Johnston-Hollitt says.
That computer was jointly funded by Victoria University, other NZ universities, the Ministry of Economic Development and IBM.