Slashdot Mirror
Overwhelmingly Large Telescope Closer to Reality
An anonymous reader submits: "The 100m OWL telescope proposed a few years ago by the European Southern Observatory group (ESO) may actually be built. Currently, the largest aperture for a telescope is the Very Large Telescope (VLT) at a 'very tiny' 16.4m by comparison. This monster is predicted to have a light gathering resolution of about 40 times the Hubble Space Telescope and a sensitivity several thousand times greater. Among many other things, it should be powerful enough to detect and gather spectroscopic data of extra-solar planets in order to determine the atmospheric composition and any signatures for life, like oxygen." We mentioned the OWL in this previous article too.
The larger an object is in orbit, the more likely it is to be damaged by random chance and debris. We really need to clean up earth orbits before we start putting more stuff up there.
Should we have more space-based telescopes? Absolutely. But for now, it's much cheaper and safer to have large telescopes down here, even if they do have to account for atmospheric distortion.
That what was all this school was for... to teach us how to solve our own problems. -- janeowit
One is being planned. However, there's absolutely no way you're going to put a mirror that big in space. So if you care more about number of photons and less about resolution (for example, if you're taking spectra of distant point sources like quasars or planets), it's better (and cheaper!) to do it from the ground.
[TMB]
I have ground an eight-inch mirror. If you rub two glass plates with carbo between in a random fashion, the grinding and polishing process naturally produces a spherical surface. We actually want a parabolic surface, but the difference on an f8 mirror of this size is about half a wavelength. You can do this parabolizing by the same back and fourth process, but by pressing down a bit harder on the end of the stroke, to remove more material from the centre of the plate on top. It's a wonderfully low tech process that gives a very accurate result.
Now, if you scale up the mirror, then things get harder. The errors in a larger mirror scale up, so you have to take off many wavelengths thickness,so people have to use interferometers and computer controlled polishing machines.
Adaptive optics made parabolization easier. If your mirror is made up of segments that are a bit smaller than my eight inch mirror, then the differences between a spherical element and a paraboloidal element are no longer worth worrying about.
When you get to the size of the OWL, the difference in a 10 cm tile between a spherical surface and a flat surface is hardly worth worrying about. You could use float glass if it came in stress-free 10cm squares. You can make accurate plastic elements that would do the job. If you can stamp out computer controlled mirror elements, then maing a mirror the size of a football field no longer seems so impossible.
The next big thing is to make the telescope track a celestial object. This thing is going to be about the size of the great pyramid, and the mirror has to stay in shape to a fraction of a wavelength. They reckon they can do it for a billion (10e9) euros. I remember (maybe wrongly) that the Mount Palomar telescope cost about 400 million dollars, back in the late twenties, early thirties.
I am not sure yet that the thing can be built for the price, but it is beginning to look like it might. Cor, juice!
The farther you try to look, the longer the exposure. For instance, the deep field pictures that came back from Huble some years ago had a few days of exposure time. Of course, you better have a mighty solid research project to justify monopolising the telescope for such long times while other labs are waiting for their turn.
Powerful telescopes are built on top of high montains and away from air routes, but not for the reason you think. The field of view of telescope is so narrow, you don't have to worry about things crossing in the way. Rather, monitors lights on the tips of air planes would generate enough background lighting to screw an exposure. Those devices realy are that sensitive to light. In fact, telescope operators tend to play trick on neibouring villagers, telling them on which days they forgot their porch light on. The lights leave a tale tell background whiteness on the pictures.
This post was compiled with `% gec -O`. email me if you need the sources
The purpose of the 100m telescope is just that, to build a very large aperture telescope. This will increase your light gathering ability and angular resolution. This you cannot accomplish (as another poster suggested) in the same manner that they do with radio astronomy (i.e., time-tag the data and put the picture together later during post-processing) because you'll never get accurate enough clocks to make those measurements.
Consider that to make a decent image you need an optic that is accurate to a fraction of a wavelength (lets use 1/10 to make the math easier). To make a radiotelescope image you are dealing with wavelengths of about a meter, so you need to tag the wavefront to about 10 centimeters, which given the speed of light is 3x10^10 cm/s, means you need clocks that are synchronized to a few hundred picoseconds. You can do this with atomic clocks. However, in the light band, if you have a wavelength of 500 nm, you need to tag your wavefront to about 50 nm, which means you need to synchronize your clocks to about 10^-16 seconds. I don't know what kind of improvement you are expecting out of the next generation of atomic clocks, but it isn't going to be six orders of magnitude. And I'll even go out on a limb and suggest that you aren't going to have clocks that accurate in our lifetimes.
A 100m telescope is good science any way you look at it.