PDA

View Full Version : resolving the image of a star



zorro
2007-Feb-28, 02:34 PM
In telescopic images of star fields (e.g. Hubble images), many stars appear as a disk rather than a point of light. Is this because the telescope is actually capable of resolving the image of the star as a disk, or is it merely due to "fuzziness" in the image?

antoniseb
2007-Feb-28, 02:39 PM
This is due to the fuzziness of the image. There have been a few times that stars have been 'imaged' using special techniques, but these are only a few nearby very large red giant stars. For the most part, today's best telescopes cannot image stellar disks.

zorro
2007-Feb-28, 02:49 PM
Thanks antoniseb. Please forgive my ignorance, but this begs the next question - How much would the resolving power of the best current telescope have to improve to resolve a "regular" nearby star, e.g. Alpha Centauri? 10 times better? 100 times? 1000 times?

Kaptain K
2007-Feb-28, 03:01 PM
The theoretical limit of one of the Keck scopes (10 meter diameter - the largest currently operating scopes) is about 0.01 arc seconds. The angular diameter of Alpha Centauri is about 0.007 arc seconds, so a 15 meter scope should (theoretically) be just able to resolve Alpha Centauri. Resolution is a linear function of aperture, so to get 10 times the resolution, you would need 10 time the aperture, etc.

Hamlet
2007-Feb-28, 03:35 PM
In telescopic images of star fields (e.g. Hubble images), many stars appear as a disk rather than a point of light. Is this because the telescope is actually capable of resolving the image of the star as a disk, or is it merely due to "fuzziness" in the image?

Here (http://hubblesite.org/newscenter/archive/releases/1996/04/image/a/) and here (http://www.spaceimages.com/atofbet.html) are Hubble images of Betelgeuse. To my knowledge, this is as close as we've been able to get in resolving the disk of another star.

Occam
2007-Feb-28, 06:46 PM
Hi Zorro,
what you seem to be referring to is usually caused by exposure time. Because of their very nature, all photographs of star fields are time exposures, simply because the distant subjects are so dim. In order to obtain the stunning pictures we've seen in recent years, especially those from Hubble, the exposure times have been very long. The consequence of this is that naturally brighter stars are over-exposed (the light collected from them exceeds that required) and they appear in the image as blobs instead of points.

mugaliens
2007-Mar-04, 07:05 PM
Sorry, Occam, but Captain K's explanation is technically correct.

A way to increase the aperature, synthetically, is to use multiple exposures all taken over the same period of time from distant locations, such as in various stages of orbit around the Earth. Given enough images, which can be overlayed, and the distance between then, this approach would drastically improve the resolution.

Tim Thompson
2007-Mar-04, 08:10 PM
A way to increase the aperature, synthetically, is to use multiple exposures all taken over the same period of time from distant locations, ...
The distant locations part is unnecessary. All you need to do is dither the telescope pointing by sub-pixel angular changes, so that you oversample the point spread function (http://en.wikipedia.org/wiki/Point_spread_function) (PSF) even more. You can then use a maximum likelihood (http://en.wikipedia.org/wiki/Maximum_likelihood) method (The Richardson-Lucy algorithm (http://en.wikipedia.org/wiki/Richardson-Lucy_deconvolution) is a popular choice) to enhance the resolution by a factor of 2 or 3 (maybe more if you are lucky) by artificially shrinking the PSF. That's what the software I use on Spitzer images does. It works surprisingly well for deconvolving crowded fields, or enhancing marginal sources, by sweeping up the flux in the Airy rings (http://www.matter.org.uk/tem/diffraction_at_aperture.htm). The method we use was originally developed as the HiRes tool (http://irsa.ipac.caltech.edu/IRASdocs/hires_over.html) for IRAS (http://lambda.gsfc.nasa.gov/product/iras/) images (but we do it faster, cheaper, and maybe even better).

That said, individual stars still have such small angular diameters that only a few, at best, could be resolved like this for a single dish telescope. Of course, the HST images of Betelgeuse (http://hubblesite.org/newscenter/archive/releases/1996/04) are the best known, but Betelgeuse (http://www.solstation.com/x-objects/betelgeuse.htm) is not so far away, and is optically as big as the orbit of Jupiter, one of the largest stars known. But one can measure the angular diameter of a star, and see sufficiently large contrast differences across the surface, using an interferometer (http://www.geocities.com/capecanaveral/2309/page1.html). The first successful astronomical interferometer that I am aware of is the 20 foot Michelson Interferometer (http://www.mtwilson.edu/vir/100/20fti/index.php) on the 100-inch Hooker Telescope (http://www.mtwilson.edu/vir/100/) at Mt. Wilson Observatory (http://www.mtwilson.edu/index.php). They were able to measure several stellar diameters (all red supergiants), including Betelgeuse. There are now several productive astronomical, optical interferometers around, and more stellar diameters are being measured. Perhaps the most significant instruments are the CHARA array (http://www.chara.gsu.edu/CHARA/) at Mt. Wilson, the VLTI (http://www.eso.org/projects/vlti/) at the European Southern Observatory, the Palomar optical interferometer (http://pti.jpl.nasa.gov/), and the Keck interferometer (http://msc.caltech.edu/software/KISupport/).

Nereid
2007-Mar-04, 08:47 PM
Two things to add to Tim Thompson's excellent post:

1) the wavelength of light you are observing in is also important; e.g. the HST images of Mira (http://antwrp.gsfc.nasa.gov/apod/ap010121.html) (and Betelgeuse) were taken in the UV, where the FOC's PSF is smaller.

2) there are several other techniques for estimating stellar diameters (and more), such as lunar occultations (http://spiff.rit.edu/richmond/occult/bessel/bessel.html) and speckle interferometry (http://en.wikipedia.org/wiki/Speckle_imaging). Note that the former does not produce an image as we normally understand it, nor do the optical interferometers mentioned above.