3 ms·

Small telescopes are very useful for searches for extremely faint, diffuse galaxies because it turns out that these searches are surface brightness limited rath

Small telescopes are very useful for searches for extremely faint, diffuse galaxies because it turns out that these searches are surface brightness limited rather than being magnitude limited like most surveys. Surface brightness is a little counterintuitive because the minimum surface brightness a survey can find is actually independent of telescope size. Using a 8-m telescope is no better for finding a diffuse galaxy than a telephoto lens. The only thing that can improve your signal to noise ratio is observing time and observing time is much cheaper on a small telescope. The basic reason for this is that as you increase the size of your telescope, not only does the collecting area increase, but the area of the image increases as well. So if you're looking at a diffuse source like a galaxy, the increased number of photons gets spread out over a larger area and the total number of photons per pixel remains the same. This isn't a problem for point sources (like stars) because even as the image size increases, all the photons from a point source still fall in the same pixel so the source appears brighter to the detector. As peter303 also mentioned, refracting telescopes don't suffer from the same artifacts that reflecting telescopes do, which is why these surveys use small telephoto lenses rather than small reflectors.

sandworm101 10y ago
So would you get similar surface brightness performance out of a larger telescope by decreasing the size of the detector, by adding a couple lenses to focus the light onto a smaller area? It's been a while since I took astronomy courses and I try to keep up with the basic physics.
- antognini 10y ago
  Yes and no. In principle it would be possible to do something like that, but you could only do it for a part of the image and you would lose the focus. You'd basically be trying to transform an extended source into a point source. In practice it's easier just to use a smaller telescope for a longer period of time. Incidentally, this is related to a puzzle: can you burn a paper with moonlight if you have a sufficiently big magnifying glass? It turns out that you cannot (at least as long as you're using a normal magnifying glass that brings light to a focus), because the highest temperature you can produce is the temperature at the surface of the moon. In the limit of having an infinitely big magnifying glass, the view from the sheet of paper would be the same as the view from the moon, and so the temperatures would be the same as well [1]. [1]: Making various assumptions about radiative thermal equilibrium, etc.
  - tgb 10y ago
    Very cool. I don't have any questions to ask you but I want you to keep talking since this is all very interesting. Edit: this xkcd does a wonderful elaboration on the fire - by - moonlight topic https://what-if.xkcd.com/145/ https://what-if.xkcd.com/145/
    - antognini 10y ago
      I had not seen that xkcd article before! Thanks! It's the most lucid explanation of the problem that I've seen. The only other interesting surface-brightness-related fact I can think of at the moment is that surface brightness fluctuations can be used to measure the distance to galaxies. This is one of the cases where it helps to have a really big telescope. The basic idea is that although galaxies are extended sources, they're really just a collection of point sources --- they're just a bunch of stars grouped together in an area of the sky. Now suppose you have two similar galaxies (at least they have similar stellar densities) and one of them is close by and the other is far away. The nearby galaxy will have relatively few stars per pixel and the distant galaxy will have many more stars per pixel. Since the stars are randomly distributed, the number of stars in any given pixel will be given by a Poisson distribution. A consequence of this is that in the nearby galaxy there will be a lot of variation in the flux from one pixel to the next, whereas in the distant galaxy the image will be much more smooth. So even though the average flux per pixel from both galaxies is the same, you can still tell which one is close and which one is distant based on the surface brightness fluctuations.
      - sandworm101 10y ago
        But without actually imaging stars, how do you know that two galaxies have similar stellar densities? The OP was about low-star/high-darkmatter galaxies. I guess you would have to hope that the entire thing was rotating, and at an angle, that doppler effects could determine the size/mass.
        antognini 10y ago
        That's where you have to make some assumptions given the galaxy type. As with a lot of distance techniques in astronomy, there is a lot of uncertainty. This is one of the reasons that surface brightness fluctuations are not a very popular method.
Sharlin 10y ago
[T]hese surveys use small telephoto lenses rather than small reflectors. Well, small telephoto lenses relative to most telescopes, pretty darn big relative to most camera lenses ;)
- wiredfool 10y ago
  Cheap telescope. Very expensive camera lens.
  - dekhn 10y ago
    I bought a $1000 refractor telescope, then got a prime focus adapter for my camera. It's basically a big 1200mm zoom lens.
bladedtoys 10y ago
Would it be possible to software process images taken by large numbers of amateurs of the same point in the sky to get even better results even cheaper?
- antognini 10y ago
  Yes, in principle. In practice, there would be so much variation in the sensitivities of the detectors that it would be a nightmare to do the image reduction. In normal observing you have to do "flat fielding" every night (or at least every few nights) where you take an image of the sky at twilight so that you can calibrate the different sensitivities of the different pixels in the detector. Then all the pixel scales will be different (i.e., a single pixel on one telescope will correspond to a slightly different area than on another telescope), so you'll have to figure out how to scale all the images to overlay them all. It's just easier to do it all yourself.

sandworm101 10y ago

So would you get similar surface brightness performance out of a larger telescope by decreasing the size of the detector, by adding a couple lenses to focus the light onto a smaller area? It's been a while since I took astronomy courses and I try to keep up with the basic physics.

antognini 10y ago
Yes and no. In principle it would be possible to do something like that, but you could only do it for a part of the image and you would lose the focus. You'd basically be trying to transform an extended source into a point source. In practice it's easier just to use a smaller telescope for a longer period of time. Incidentally, this is related to a puzzle: can you burn a paper with moonlight if you have a sufficiently big magnifying glass? It turns out that you cannot (at least as long as you're using a normal magnifying glass that brings light to a focus), because the highest temperature you can produce is the temperature at the surface of the moon. In the limit of having an infinitely big magnifying glass, the view from the sheet of paper would be the same as the view from the moon, and so the temperatures would be the same as well [1]. [1]: Making various assumptions about radiative thermal equilibrium, etc.
- tgb 10y ago
  Very cool. I don't have any questions to ask you but I want you to keep talking since this is all very interesting. Edit: this xkcd does a wonderful elaboration on the fire - by - moonlight topic https://what-if.xkcd.com/145/ https://what-if.xkcd.com/145/
  - antognini 10y ago
    I had not seen that xkcd article before! Thanks! It's the most lucid explanation of the problem that I've seen. The only other interesting surface-brightness-related fact I can think of at the moment is that surface brightness fluctuations can be used to measure the distance to galaxies. This is one of the cases where it helps to have a really big telescope. The basic idea is that although galaxies are extended sources, they're really just a collection of point sources --- they're just a bunch of stars grouped together in an area of the sky. Now suppose you have two similar galaxies (at least they have similar stellar densities) and one of them is close by and the other is far away. The nearby galaxy will have relatively few stars per pixel and the distant galaxy will have many more stars per pixel. Since the stars are randomly distributed, the number of stars in any given pixel will be given by a Poisson distribution. A consequence of this is that in the nearby galaxy there will be a lot of variation in the flux from one pixel to the next, whereas in the distant galaxy the image will be much more smooth. So even though the average flux per pixel from both galaxies is the same, you can still tell which one is close and which one is distant based on the surface brightness fluctuations.
    - sandworm101 10y ago
      But without actually imaging stars, how do you know that two galaxies have similar stellar densities? The OP was about low-star/high-darkmatter galaxies. I guess you would have to hope that the entire thing was rotating, and at an angle, that doppler effects could determine the size/mass.
      - antognini 10y ago
        That's where you have to make some assumptions given the galaxy type. As with a lot of distance techniques in astronomy, there is a lot of uncertainty. This is one of the reasons that surface brightness fluctuations are not a very popular method.

Sharlin 10y ago

[T]hese surveys use small telephoto lenses rather than small reflectors. Well, small telephoto lenses relative to most telescopes, pretty darn big relative to most camera lenses ;)

wiredfool 10y ago
Cheap telescope. Very expensive camera lens.
- dekhn 10y ago
  I bought a $1000 refractor telescope, then got a prime focus adapter for my camera. It's basically a big 1200mm zoom lens.

bladedtoys 10y ago

Would it be possible to software process images taken by large numbers of amateurs of the same point in the sky to get even better results even cheaper?

antognini 10y ago
Yes, in principle. In practice, there would be so much variation in the sensitivities of the detectors that it would be a nightmare to do the image reduction. In normal observing you have to do "flat fielding" every night (or at least every few nights) where you take an image of the sky at twilight so that you can calibrate the different sensitivities of the different pixels in the detector. Then all the pixel scales will be different (i.e., a single pixel on one telescope will correspond to a slightly different area than on another telescope), so you'll have to figure out how to scale all the images to overlay them all. It's just easier to do it all yourself.