Slashdot Mirror
The Square Kilometer Array
EyesWideOpen writes "A very ambitious project to build the world's largest radio telescope, named the Square Kilometer Array or SKA, is in its early design stages. As its name suggests the SKA will be one square kilometer in size if it gets built. The SKA consortium (consisting of Cal Tech, Cornell, SETI, the Max Planck Institute and Beijing Astronomical Observatory to name a few) hopes to build the telescope by 2010. "If they succeed the SKA will be so big and precise it will jump the world's current best, the American Very Large Array in New Mexico, by a factor of 100, both in sensitivity and resolution." It's interesting to note that the project is based on technology that will only exist in three, five or seven years -- to account for data rates of tens to hundreds of terabytes per second and storage in the petabytes -- so they're counting on Moore's law to hold true."
Given that the wavelength of 'visible' light is approximately half a million times shorter than radio wave wavelengths, the collecting area has to be much larger to get the same antennae gain.
An interesting corollary of this is that the naked eye is (very roughly) as powerful (at visible light wavelengths) as Arecibo is (at radio wavelengths). See the The seti league pages for more info...
Simon.
Physicists get Hadrons!
If this will ever get funded (they recently got some money to make first studies) it will be a telescope the size of half the Netherlands. This is of course not a filled aperture, but a sparse one operating at very low frequencies (10-250 MHz, on both sides of the FM frequencies). It will consist of some hundred small "antenna parks" spread around the country and uses a lot of computer power to generate images. It could be a precursor for SKA.
karma police: arrest this man, he talks in maths; he buzzes like a fridge, he's like a detuned radio. [radiohead]
I believe a petabyte is 1,000 GigaBytes, or 1,000,000 MegaBytes... No, a Petabyte is 1024 Giga Bytes... See http://www.tuxedo.org/~esr/jargon/html/entry/quant ifiers.html or do a search on google...
Moore's Law is about the density of transistors in integrated circuits, not their speed or cost.
Mea navis aericumbens anguillis abundat
Fortunately it's only compared to the VLA in regards of resolution. Single radiotelescopes have no chance in hell to get to extreme resolutions. Resolution is all in the diameter, or baseline. Nothing you can do about, it's just basic physics. Fortunately you can have big holes in your telescope, or inversely just a few parts of the surface. Excactly the principle of the VLA and VLBI in radio frequencies and the VLTI for light. You can even find a simulation applet here
In fact the earth itself is getting too small to get more resolution. Going into space is indeed being looked into, but not in the sense of a satellite like the Hubble orbiting the earth. That would hardly be worth the effort where radio astronomy is concerned. Having a baseline as long as the distance between the earth and the moon, now that would be an improvement. Plus, if it's built on the side that's always turned away from the earth, the telescope will be shielded from all the annoying interference created by all the radiochatter on earth, while it's still possible to look at the same piece of sky as an earth based telescope.
In the visual spectrum, Darwin from ESA looks set to become the next record holder . A first technology demonstration/development flight in the form of SMART-2 is currently under development.
No. There hasn't been a single installation so far with a more than a 1/30 km of collecting area.
What's the difference between what is referred to as the baseline in a VLBA, and what we're talking about here? If you increase the baseline, you increase the "aperture", right? But that doesn't increase the sensitivty, right? Is the real advantage of a huge array of dishes designed and operated as one telescope (as opposed to an ad hoc assembly) the things that are involved in this story -- i.e., data communication bandwidth and control?
:), then the resolution will be dictated by how big Australia is. About 1 milliarcsecond, or about 1000 times better than the average pre-interferometer resolutions you could get with optical telescopes on the ground, and 100 times better than hubble, keeping in mind that a radio telescope of the same size as an optical telescope will always have a resolution many thousands of times less (the ratio of the wavelength of optical light to radio light).
Interferometers are very differnt beasts to normal radio telescopes. Single dish scopes look at a single area of the sky, and their sensetivity is proportional to the collecting area (square of diameter). Their angular resolution is proportional to the diamater. When I say the are pointing at a single area of sky, the telescope is actually looking at one point the size of the angular resolution - you may choose to look for a long time, gathering a spectrum (or looking at a pulsar) of that single point, or you may scan the telscope back and forth slowly to generate an image (with resolution equal to the angular resolition of the telescope).
With interferometers, you have a bunch of telescopes. The fundamental unit is no longer a single dish - it is now every combination of 2 dishes. At ATNF narrabri, there are 6 dishes, so there are 15 combinations (5 + 4 + 3 + 2 + 1) (I remember once having to step through each baseline individully, for each frequency for each observation we made, for each.... something else, to mitigate some interference manually, to get the best possible image I could generate for some nifty work I was asked to do) of pairs. The resolution is now a function of the distances between all the pairs.
You generate an image immediately, by getting the fourrier transform of the signals from the pairs, as the earth rotates. To generate the optimal image, with an East West synthesis telescope (such as Narrabri) where the X -resolution is (almost) the same as Y-resolution, you have to let the earth rotate a half turn, ie you sit there imaging for 12 hours. I have gotten away with observing for 4 before, but that was a very specific project. Other telescopes can sometimes do a "snapshot" mode, where you observe for a few minutes or hours, without too much loss of information. But basically, you don't have to scan the telscopes anymore, the centre of the image is where you point the telescopes, and the size of the image can be as big as the resolution would have been if you were using just one telescope.
The resolution you get is effectively from the farthest separated dishes, and the biggest structure you can see is from the resolution of the closest dishes (this all comes from the fact that you have to perform an inverse fourrier transform of the data coming from the pairs, and there are bits missing from the fourrier plane, where there aren't telescope pairs). With a single dish, you can see structures of any size bigger than the resolution. But an interferometer is missing all these bits where telscopes aren't situated, and in particular, has effectibely a hole in the middle of the "telescope" the size of the distance between the closest dishes. So there is an upper limit on the size of structures you can see (as well as a lower limit).
So occasionally, there have been tricks where you combine the high resolution data from interfereters with the low resolution data from a single dish, and you generate a very accurate and imformative image. This was done for generating a map of the Large Magellanic Cloud (no URL handy). But this needs a lot of work and telescope time, both hard to come by.
The sensetivety goes only as the size of the physical collecting area. So 1 square kilometer indeed is much better than the previous 1/30 or so sqaure kilometers we have had in a single setup. Note that, if the telscope is set up in Western Australia, (where I certainly hope it will
I apoligise in advance for confusing you all, but it is kindof a complex topic, and no doubt my head will explode now as well!