Slashdot Mirror
World's Largest Telescope Begins Production
JohnnyNapalm writes "The Aggie Daily News is reporting today that the first mirrors have been cast for the world's largest telescope. The result of cooperation from some of the foremost institutions in education and science in the nation, the Giant Magellan Telescope stands to operate at a resolution 10 times larger than the Hubble. The project, set to be constructed in Chile, is slated for completion in 2016."
Am I alone in feeling that we haven't even used hubble to the fullest extent of its abilities? Not sure why this is a priority right now.
and not the US Gov't, then THEY get to choose when to pull the plug.
Not some accountant.
My mom says I'm cool.
Largest ptical telescope, perhaps. Arecibo Observatory is still the biggest single telescope, though there are even larger arrays.
After all, I am strangely colored.
The only problem is, where shall we find a Giant Magellan who can operate such a big telescope for us?
Will it be able to show the moon landings?
That's true, but it's simple economics, unfortunately. To build an earthbased telescope, it is cheaper by a factor far outweighing the costs of hoisting the equivalent mass into orbit. On a side note, most ground telescopes utilize correctional algorithms to help with atmospheric distortion. Sure, stars sparkle down here, and dont up there, but NASA's purse is a bit light lately.
Photos
yes, but with adaptive optics and reprocessing the difference is not as big, and the operating costs are nowhere close, and if they add enough additional capabilities (can't do ir/uv in atmosphere, but some radio could help) it might be useful. Few earth telescopes will ever rival hubble however, the enormous field of view coupled with the amazing contrast allowed by its orbit really can't be matched on earth, at least not without additional processing.
for actual scientific purposes, and not pretty pictures it should be as useful.
The first rule of USENET is you do not talk about USENET.
nope, many small mirrors are easier to manufacture by far (large mirrors have too many defects to be useful, or cheap), and adaptive optics requires that the mirrors have a modifiable geometry to properly compensate for the atmospheric interference. the break in the mirrors do not reflect light, so as long as the angles are correct it is not noticable.
The first rule of USENET is you do not talk about USENET.
It does use adaptive optics. Have a peek at the tech section of the GMT site, here: http://www.gmto.org/tech_overview>
From the aforementioned link: The GMT secondary mirror is composed of seven thin adaptive shells, with each segment mapping to a single primary mirror segment. The adaptive secondary will provide diffraction-limited performance over modest fields of view and ground-layer adaptive optics over a field of ten to twenty arcminutes in diameter.
-scott
My other sig is a Glock
I guess scientists suffer from penis envy too.
"Oh yea, well my telescope is bigger then yours!"
Or would it go...
"Hey baby, look, _my_ telescope has a lens that's 40,000 pounds, it'll be 84 feet of girth, and when it gets heated up it takes 3 months to cool"
Price of putting the 2.5-meter Hubble Space Telescope in orbit, and installing its corrective glasses:
:)
Somewhere on the order of $2-4 Billion.
Price of building both 10-meter Keck Telescopes on Mauna Kea:
About $200 Million.
Soooo... for the cost of one orbiting telescope (and that wasn't even counting the later servicing missions), you could build 20-40 terrestrial telescopes, each with four times the diameter.
Oh, and as a data point... expected price of building the 30-meter Telescope:
About $1 Billion.
Launching stuff is way more expensive than getting it places on boats or trucks.
Village idiot in some extremely smart villages.
However, put that same sophisticated, adaptive telescope in space, and I'd be willing to bet you'd see yet further improvement. Besides which, even a composite telescope is going to be limited in size, due to the fact that you can't correct all of the errors - the machinary won't be capable of altering the positions of the mirrors accurately enough, or measuring their positions for computational correction.
In space, you should be able to build much larger composite mirrors, as you should be able to place things much more accurately. If the mirrors are made in space, then that would be even better, as there wouldn't be so many flaws that would need correcting.
Instead of building one super-giant telescope, they might be better building two or three slightly smaller giant telescopes, then hooking them up as an interferometry array. The reason being that you'll get diminishing returns on how much light you can get in anyway, an array allows you to obtain greater accuracy and you can use the array as multiple single telescopes if there is nothing requiring the extra detail.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
For something a little closer to completion than 2016, check out the Southern African Large Telescope. Scheduled to open in November, and will be the biggest optical telescope in the southern hemisphere.
Regards,
-Jeremy
>not actually deforming the mirror segments
IIRC The mirror segments were deformed during construction. The mirror segments need to be ground to the correct shape (with a pretty tight definition of correct). I belive they were deformed in such a way that the actual shape ground was an easy one to do. When the mirrors were released they sprung back to their original overall shape but with the surface ground to what was needed for the final mirror. Neat way of getting around the problem.
I had the wonderful experience of being an undergraduate in astrophysics at UC Santa Cruz, where a grant in adaptive optics was paving the way for ground-based telescopes. By shining a laser straight where the telescope is pointed, aberrations and distortions from the atmosphere can be measured and exactly countered by the telescope, effectively cancelling atmospheric effects to a remarkable degree. Check out http://cfao.ucolick.org./ The main telescope was outside of San Jose, CA, which might seem a strange location for a telescope due to its proximity to a large city. But since all of the streetlamps in San Jose are sodium (whose spectral properties are well known and simple), those features can be subtracted from any measured spectra and in conjunction with adaptive optics, the telescope outside northern california's largest city produces world-class astronomy. This telescope being built should be pretty neat. I wonder how they will deal with gravitational aberrations. Plus scientists won't ever need to face the threat of government letting their instrument "deorbit" while still producing good data.
They intend to use adaptive optics to compensate for the effects of the atmosphere, though if you read the science case they haven't quite decided on how to do this yet.
I can't think of many non-military organisations which have bigger budgets
I can. 2005 Numbers:
Department of Health & Human Services: 584B
Department of Education: 56.5B
Department of Veterans Affairs: 32.5B
Department of Housing & Urban Development: 32B
Department of Homeland Security: 29B
Department of State: 27.5B
Department of Energy: 23.8B
Department of Agriculture: 21.4B
Department of Justice: 20.2B
NASA: 16.1B
Cheaper Departments include: Treasury, Transportation, Labor, Interior, Drug Administration, EPA, and Commerce. They generally run 8-15 billion each.
Source: Washington Post
I don't read AC A human right
It's one piece of glass, with a single, smooth surface on the front, 8.4 m in diameter. The hexagonal "pieces" are holes on the backside. It basically looks like a big honeycomb. This design gives you great stiffness and strength, with only 20% the weight that a solid mirror would have.
Liberal (adj.): Free from bigotry; open to progress; tolerant of others.
The Steward Observatory Mirror Lab had an open house yesterday for observatory personnel, which I attended.
The spin-cast oven is huge. In these pictures, you only see the top portion of it, it actually fills the floor below as well. I believe this is the only large spin-cast mirror facility in the world. The idea behind spin-casting is that, by spinning the molten glass as it is slowly cooled, you automatically get a paraboloid top surface. This makes the final shaping of the mirror much easier, since the first-order shape is already there.
Actually, in the case of the GMT, it will use seven mirrors, six of which are off-axis. The off-axis mirrors will obviously have a more complicated surface than a typical on-axis paraboloid. The mirror being cast now is an off-axis mirror; it is a proof-of-concept that they can grind an eight-meter chunk of glass to an off-axis paraboloid shape with a surface RMS of 20 nanometers (!).
In a few months when the mirror has cooled and solidified, it will be removed from the oven, cleaned, ground, and eventually, polished. The stress-lap polisher is very impressive. It has a network of stress actuators above it, which can dynamically change the shape of the polisher's surface as it travels across the mirror.
It's interesting that the "Aggie Daily News" was chosen as the linked story, which makes it sound like UT Austin and Texas A&M are the major players in the GMT, along with a handful of other, unnamed institutions. In fact, the Carnegie Institute is the impetus behind the project, and the U of Arizona is providing the mirrors. I think this UA News article is much more informative.
Liberal (adj.): Free from bigotry; open to progress; tolerant of others.
The largest dimension on the Apollo landers is 9.07 m, diagonally between the landing legs (it's a 21 foot square). The moon's closest approach to the earth is 363,104 km. Divide those two numbers, and you get the angular size in radians: 2.36e-8 radians, or 0.00487 arcsecond.
With a 27m diameter, the diffraction limit on telescope resolution is 10^8 cycles/radian. So if there were no atmosphere, it would be barely possible.
With an atmosphere, there are problems. A typical good seeing limit is 1 arcsecond (corresponding to 1864 m on the moon). The best on the planet (Dome C, in the middle of Antarctica, at a very good time) is 0.2 arcseconds (373 m on the moon).
It's possible for adaptive optics to get good enough to allow that much correction, but it's going to be difficult.
Hoewever, if you want optical evidence for the Apollo landings, just shine a bright light at the moon and look for the retroreflector arrays the astronauts left there. That experiment has been done a zillion times in the last few decades.
You just need a very, very accurately aligned surface. For moderate sizes, this is easiest to achieve by just making the whole mirror a solid piece. As the mirror gets larger, the stiffness needed to hold its shape to within a fraction of a wavelength of light gets more and more difficult to achieve. Above 5m (the 200 inch Hale telescope), you have to change to active supports, where you measure and compensate for gravity sag rather than just trying to be stiff enough.
There's still a tradeoff, though. The mirror needs to be stiff enough to hold its shape between support points. A thinner mirror weighs less, but requires more active supports.
Past 8m (27 ft), it's just not feasible to build or transport a single mirror, so folks cut it into pieces, effectively sending the mirror thickness to zero in some places. This further complicates the active support system, but it's possible.
The gaps between mirror segments are no different than the shadow of the spider supporting the secondary mirror; they're just places you don't get any light from. They don't pose any particular imaging problem.
You can cut the round mirrors into hexes and pack them closer; this reduces the overall size of the telescope and makes it easier to build. But after considering the posibilities, the GMT folk decided to leave the mirrors round and accept larger gaps. This actually makes a slightly better telescope. No cutting means more mirror area. Wider edge-to-edge gives a better diffraction limit on resolution. And it turns out the round edges produce a nicer point spread function.
There is already a Magellan project, a 2 telescope optical interferometer
Grandparent has a point though. This naming convention is a poor choice. What will they call the next one? "OMG The Really REALLY Big Ginormous Magellan Telescope"? And the one after that?
Marketroids (and apparently the ivory tower residents responsible for naming telescopes) need to learn from the debacle of USB Hi-Speed vs Full-Speed. Future-proof the meaning of your technology's name by assigning it based on absolute, and NOT relative, criteria. "Giant" has no real meaning. "25.6m" (the resolving power of the GMT) does have a meaning that will persist into the future.
1) Absorption. The atmosphere absorbs in many wavelengths of interest, including the UV and parts of the IR. There are some projects that can never be done on the ground.
2) Background emission. The atmosphere "glows" at a number of wavelengths; this acts as a source of background contamination and reduces your sensitivity.
3) Blurring. The stars twinkle. This reduces the sharpness of ground-based images by an enormous factor (for GMT in the optical, excluding AO, by a factor of about 200).
People keep mentioning "adaptive optics" as a way to overcome the blurring from the atmosphere. But the harsh truth is that AO doesn't work all that well, for situations where you actually need to get rid of the effects of the atmosphere. Sure, it sharpens up pictures of binary stars pretty well, but it leaves a bunch of uncalibrated "scruff" near the star that e.g. makes it impossible to look for planets near that star. Another limitation of AO is that it requires a bright star to guide on - although lasers are becoming available. Mind you, the laser stuff seems to have even worse issues with calibration. Finally, AO has a very limited effective field of view; you can only correct over a small patch at a time. It makes it hard to do wide-field surveys that way.
Sooo, the upshot is that you need both, and will continue to need bothy for a long time. That being said, I wish the GMT guys lots of luck.
Human genome = 3 billion base pairs = 6 GBit. Windows + Office = 20 Gbit. Which is more impressive?