Relativity and Cosmology

Splitting the project is an odd decision given that Oz has a vanishingly low population density (near the center of the site), stable government, rule of law, a national optical broadband system being rolled out and a large science/technology infrastructure.

If the project has not been split, the decision would probably have gone to Africa. This is what the Site Selection Advisory Committee recommended earlier.

I think sense has prevailed; the significant investments already made by Oz and SA are now fully utilized. Actually, according to the Site Options Working Group* final report, phase 1 of the project will now deliver more science than specified (for phase 1) without any increase in cost.

-J

* Check under Project Documents

Do you have any idea who the prime contractor for the build in Oz will be mate?

Thanks for that Jorrie.

Do you have any idea about the timeline for this project?

SKA timeline according to http://www.skatelescope.org/uploaded/38221_SKA_Project_Execution_Plan.pdf:

phase 1: 2018, phase 2: 2022, full performance: 2024

-J

So, according to that great reference, construction start for SKA1 is 2016 and SKA2 is 2018.

I'll have to find something else to do till that starts or find a design role...

Plenty of good reading there Jorrie. Thanks.

Thanks for this article.

I'm an electronic engineer and I'm very interested in understanding the methods used to make such a telescope. Of course the phased arrival of an EM wave at multiple receivers can be used (with extremely careful timing) to reconstruct properties of the emitter. But I have never seen a detailed explanation of the "how" its actually done and how its specifically done on such an large array.

For instance how are the individual antenna calibrated (for relative intensity measures, for time delays to/from the computers - which I presume could even be variable with temperature / time of day ! ) so that computers can reconstruct the image? How are they used across continents ?? How are the details of spacial images (like diffuse clouds of gas for instance) rather than simple point emitters (distant stars) reconstructed from the array?

If there's a good source of papers or information on this I'd appreciate it. I don't need spoon feeding :-) just pointing to the kitchen :-)

thanks

Hi ss3e55, I think the secret lurks in calibration, calibration, calibration...

And of course data time-stamping with local, synchronized atomic clocks. This Wiki article gives some good information and references: http://en.wikipedia.org/wiki/Very_Long_Baseline_Interferometry

-J

thanks Jorrie, this is a good start and the wiki leads onwards to more articles...it might be a while but I can see the start of my understanding.

Of course the accuracy is highly dependent on the atomic clocks used to timestamp the signals. I can also see that systematic errors (or periodic ones due to cooling down of antennae from the Sun going below the horizon for instance) can be eliminated by careful recording of a well known standard source (recorded simultaneously on all receivers)...but some of the steps to reconstitute full images are still tough to understand.

For instance at first I presumed each antenna's signal becomes a single pixel in that image, but that implies being able to point each antenna separately to the most minutely fine tuned angle differences , whereas as I think about it now, I think an image might be built by steering the "entire" array and sampling (because only the array as a whole as a fine enough beam resolution needed to resolve a single pixel)? Perhaps the earth's drift and rotation are used to help "scan" a line of pixels...with a flyback and reposition for another line...who knows? :-)

Well thanks for the first steps...I'll see how far my curiosity takes me :-)