Printable Version of Topic

Posted by: rboerner Apr 1 2015, 12:26 AM

Since there are no planned new missions to Uranus and Neptune, I was wondering whether we can expect Voyager-2 like images of the ice giants and especially their satellites from the 30 meter - class telescopes in the 2020s.

If my math is correct, then E-ELT, operating at the diffraction limit in visible wavelengths, would resolve ~50km at closest Uranus distance, and ~100km at closest Neptune.

Furthermore, by using super-resolution techniques (combining several subpixel-shifted exposures into one higher resolution one, similar to what was done with Pluto HST pictures), these resolutions could perhaps be improved by a factor 3.

http://www.unmannedspaceflight.com/index.php?s=&showtopic=7970&view=findpost&p=219181

~16km at Uranus and ~33km at Neptune is worse than the best Voyager 2 data, but it should be sufficient to see geology [on the satellites].

In any event, since Voyager 2 encountered Uranus during its southern summer, and it will be northern summer in 2028, we would not be re-imaging the already known southern hemispheres of the Uranian satellites, but the currently unknown northern hemispheres.

I also get that E-ELT could get ~20 pixels across the disk of Eris using a factor 3 super-resolution enhancement.

Is my basic thinking here valid? I'd especially be interested to hear from someone who understands more of the math of super-resolution and its limits. Could factors much greater than 3 be achieved?

Posted by: Steve G Apr 1 2015, 12:58 AM

I was wondering the same thing about the James Webb Telescope but even though it's 100% more powerful, it's not going to give us 2500 pixel wide pictures of Io or Titan, at least from what I understood in the materials I have read.

Posted by: Gerald Apr 1 2015, 10:54 AM

The more efficient method to get high resolution images is probably http://en.wikipedia.org/wiki/Interferometry applied to an array consisting of at least three ("small") telescopes at the corners of a regular (equilateral) triangle.
That way you can overcome the http://en.wikipedia.org/wiki/Diffraction-limited_system of a single telescope.

Posted by: machi Apr 1 2015, 12:49 PM

~60 km at Uranus and ~90 km at Neptune with E-ELT/EPICS/EPOL.



Superresolution doesn't work beyond Nyquist limit (NL) and most of the planned cameras for future ELTs will have already better sampling than NL.
Superresolution techniques are only used for cameras with sampling worse than NL (for example HST/ACS).
So diffraction limited resolution will be ~60 km at Uranus and ~90 km at Neptune despite fact that resolution per pixel will be higher than that (~25 resp. 35 km/pixel in case of EPICS/EPOL).



It's pretty obvious from published plans that extreme hi-res instruments will not be ready until ~2030 (and those instruments are optimized for exoplanet imaging not for Solar system imaging).
Imaging with near IR AO cameras is viable in 20's but resolution will be "only" ~100-150 km in case of Uranus and 150-250 km in case of Neptune and KBO.
But those NIR cameras are better suited for Solar system objects (larger FOV, better contrast in NIR for planetary atmosphere imaging etc.).



As I wrote, SR technique has no meaning in case of future imagers.
EPICS will be able to image Eris with more than 20 image elements across disc even without SR.
But true diffraction limited resolution will be 2.5× lower.
But I must add that resolution and resolvable details aren't the same thing!.
Prominence of details in final image is dependent on local contrast.
Detail with high contrast can be visible even if his dimensions are well under resolution limit of camera.
But this isn't generally true for most of the geologic features on the Solar system bodies.
Notable exception is the bright spot on the Ceres.



Interferometry is very difficult for faint objects. For most of the outer Solar system objects it's almost impossible to get interferometric image.
More promising technology is Fresnel array. Very big (10's of meters) space-based Fresnel imager can be relatively cheap.