Photos from interferometers - when?

[Tacitus]

The two largest interferometers in the world - Paranal and Keck are currently undergoing lengthy commissioning periods. Some early results from the instruments have already been published, but does anyone know when we will start seeing photographs from them? Is it even possible? If it is, the results should be spectacular.

Mike

[ngc3314]

The two largest interferometers in the world - Paranal and Keck are currently undergoing lengthy commissioning periods. Some early results from the instruments have already been published, but does anyone know when we will start seeing photographs from them? Is it even possible? If it is, the results should be spectacular.

Mike

Getting images from even excellent interferometric data can be tricky. In the jargon, the reason is that the giant optical interferometers sample the (u,v) plane so sparsely that only simple structures can be retrieved with enough fidelity for an image. In something more like English (for our home audience today), each pair of mirrors at a single time measures the extent to which the field of view looks like a sine wave on the sky, whose direction is the vector separation of mirrors and whose periodicity (say in arcseconds) gets finer as the mirrors are more widely separated. If you have enough such measurements, postprocessing can give a high-fidelity image (as we've seen from the VLA and other radio interferometers). Many of the things we want to know most urgently from the optical/IR interferometers can be answered from relatively simple image reconstructions (but they will be just that for a while) - sizes, shapes, and limb darkening of stars, faint companions, jet or disk structure in quasars.

The image reconstruction gets dramatically better as one has more baselines, and one can often add baselines just by continuing to observe as the Earth rotates and changes the projection of the mirror separation on the sky. As far as I've seen, the state of the art for optical interferometric imaging is seen in

(1) the Naval Prototype Optical Interferometer (http://ftp.nofs.navy.mil/projects/npoi/index.html); they have six telescopes now interfering, which gives (6*5)/2=15 independent baselines, and show images of some tight multiple stars (I like the sequence showing orbital motion of the formerly spectroscopic central binary in the Mizar system). You can see there some of the artifacts to which such reconstructions are vulnerable (as in the ears on the star images in Eta Virginis).

and (2) the Cambridge Optical Aperture Synthesis Telescope (COAST), (http://www.mrao.cam.ac.uk/telescopes/coast/). Their site shows images of hot spots on Betelgeuse.

The CHARA array on Mt. Wilson (http://www.chara.gsu.edu/CHARA/index.html ) is getting fringes, but I have yet to see pictorial results.

These arrays can all yield "prettier pictures" than the Keck or VLT arrays simply because they have more elements than even the finished versions of the giants, but of course the scientific import may be quite different when one instrument has 100x the light per element to work with. I still want to actually see the evidence for an accretion disk in some quasar or Seyfert galaxy, and this kind of interferometry is probably the best short-term way to see whether they are there...

[Glom]

So the answer is no.

[Kullat Nunu]

The Large Binocular Telescope, which will be an interferometer, will be capable of taking images.

Here is a simulated LBT image of Io.

[Tacitus]

Thanks for the interesting reply. So, in simple terms, the number of photons gathered from the target is simply not enough to build an image from?

The image reconstruction gets dramatically better as one has more baselines, and one can often add baselines just by continuing to observe as the Earth rotates and changes the projection of the mirror separation on the sky. As far as I've seen, the state of the art for optical interferometric imaging is seen in

Is that why they are building outrigger telescopes - to add more baselines for better images?

Thanks, Mike

[ngc3314]

Thanks for the interesting reply. So, in simple terms, the number of photons gathered from the target is simply not enough to build an image from?

The image reconstruction gets dramatically better as one has more baselines, and one can often add baselines just by continuing to observe as the Earth rotates and changes the projection of the mirror separation on the sky. As far as I've seen, the state of the art for optical interferometric imaging is seen in

Is that why they are building outrigger telescopes - to add more baselines for better images?

Thanks, Mike

The photon rate is an issue, but often it's still the distribution of baselines that will limit image reconstruction. This controls a lot of observing programs with the VLA, even working at such long wavelengths that one always has plenty of photons. Googling around a bit, I foun examples of how image quality improves with more baselines (Fourier components) at
http://www.geocities.com/CapeCanaveral/2309/page3.html and http://www.eracnet.org/workshop/doc/basics.pdf; the latter one uses the same illustrations but shows all the math...

The outrigger telescopes serve (I seem to recall) the functions both of adding baselines, and specifically allowing cheaper addition of longer baselines. The correlation measured between them will scale with the product of the mirror areas, so even a smallish outrigger gives decent sensitivity when the big one is 8-10m in diameter.