The Billion-Dollar Telescope Race

The Great telescope in Birr was the largest in the world for 72 years from 1845 to 1917, when the 100 inch Hooker telescope at Mt. Wilson was finished.

When do you think that the first images of another world (larger than a pixel or two) will be captured. Or is the distances involved just two far to capture images more than a pixel or two?

What about Hubble, does that not count?

Surely it doesn't make that much sense to continue building large telescopes on earth when you can position one above the atmosphere?

Cookie_Monster wrote: »

What about Hubble, does that not count?

Surely it doesn't make that much sense to continue building large telescopes on earth when you can position one above the atmosphere?

Hubble is very small compared to these monsters, just 2.4 meters (about the same as the Hooker telescope from 1917) and it has cost about €10 billion so far, more than all these colossal next generation telescopes put together.

It will always be easier and cheaper to build a large telescope down here than launch it into orbit, at least until we are all living in semi-detached asteroids.

nokia69 wrote: »

Hubble was built to fit inside the shuttle cargo bay so with the launchers being built today a far bigger replacement would be possible

The shuttle cargo bay was frickin enormous, precisely to carry spy satellites the size of Hubble into LEO. There is no active launcher which can carry as big a load to orbit.

The SLS, MCT and Falcon Heavy will all be bigger, when/if they launch.

Zubeneschamali wrote: »

i see several references to gravitational waves - anyone know how a radio telescope is supposed to detect gravitational waves?

Yes. It's quite a fascinating approach. You know how LIGO works, by detecting the strain induced by gravitational plane waves from distant sources. A passing gravitational wave changes the length of space. The longer the distance (displacement) between two points the greater the change of length (also a displacement) between them caused by the passing wave. The strain (usually denoted h) is the change in displacement divided by the overall displacement:



Since you have metres above and below the line, the strain itself is a dimensionless quantity. Certain astrophysical objects of interest are known to produce strains of . The arms of the LIGO instrument (and therefore the effective displacement between its mirrors) are on the order of thousand kilometre scales, since light is bounced up and down its length several hundred times. So we have:



This is a very small change in displacement we have to measure. One way we could do it is with timed light pulses. We fire a light beam along a length, and count the pulses as they arrive at the other end. If a gravitational wave changes the length, the timing of the pulses changes as they arrive more or less frequently. LIGO uses a laser light beam with a wavelength of about one micron. But when you look at the size of our , it is only a billionth of a single wavelength!

This is why LIGO has to use interferometry. Instead of directly measuring timed light pulses, it measures the relative change in pulse duration in two orthogonal arms which can be done with the required precision. (As you know, it involves interfering the two slightly out-of-phase beams from the two arms to produce a signal).

One of the consequences of using interferometry instead of direct pulse counting is that the LIGO device becomes dependent on the frequency of the gravitational waves. Suppose the wavelength of the gravitational wave was only half the effective length of a LIGO arm. The arm could be partly stretched and partly compressed along its length, resulting in a null signal. Conversely, the wave mustn't be too much longer than the LIGO arm either. We can't measure strain directly using interferometry, only its second time derivative. (If you remember your Doppler effect, it's for the same reason that the siren of the stationary ambulance or ambulance moving at constant speed is at a fixed pitch -- we only measure a changing pitch when it is accelerating, usually due to change of parallax as it whips past us. An unchanging or slowly changing strain is equally no use to us). Fortunately, some parts of the inspiralling black hole mergers that LIGO was designed to detect produce waves of appropriate wavelength.

Even though LIGO has to use interferometry for practical reasons, we could in principle have a detector that works by pulse counting. Imagine you are using a GPS receiver. It works by triangulating timed signals from an array of satellites. Suppose a gravitational plane wave now passes by, changing the length of space in its direction of propagation. Our receiver will detect a change in the timing of the received signals. Real world GPS is not nearly sensitive enough to measure such a change, but if it was it would have a number of advantages over LIGO.

Firstly, we wouldn't have the wavelength dependence of an interferometric LIGO. As long as our time measurement was accurate enough we could measure anything. Secondly, by using a whole array of GPS satellites we would have better directionality. Only the component of each signal in the direction of the gravitational plane wave would be affected. LIGO has a version of this, but it is based on only having two detectors separated by about 3,000 km. An array would allow sources to be better pinpointed.

Ok, here comes the clever bit. Replace your GPS satellites by pulsars. These rapidly rotating objects produce extremely precisely timed radio pulses. A particular class of them -- the millisecond pulsars -- are the most precisely timed of all, with regularity approaching or exceeding that of an atomic clock. Now, enter the new Chinese FAST radio telescope. Firstly it is expected to detect huge numbers of previously unknown pulsars, about five thousand in our own galaxy and perhaps some in other nearby galaxies. About ten per cent of these will be millisecond pulsars. Pulsar rotation rates change over time, but this can be compensated for by observing over a long period of time to calculate the rundown rate. Add to this the fact that a large aperture telescope gives improvements not just in spatial resolution, but in temporal resolution. FAST is expected to yield a pulse timing accuracy of 30 nanoseconds after some years of pulsar observations.

Now when gravitational waves start passing through the space between us and these observed pulsars they change the timing in a way that we can detect. Admittedly, our thirty nanosecond accuracy is extremely rough. Dividing it into the speed of light, we find it corresponds to a change of displacement of ten metres. Compare that to one thousandth of a trillionth of a metre that LIGO has to measure. But our displacement -- the distance to the pulsar -- is vast, so our strain sensitivity can still be reasonable. Of course, a single gravitational wave will have a wavelength much smaller than this. But we'll still be able to measure high strain, very long wavelength gravitational waves. This type is produced by supermassive black hole mergers between colliding galaxies. These may circle each other initially very slowly, producing waves of one cycle in a period of months or even years (on the order of , and wavelengths in light years!). The upside of such long wavelengths and low frequencies is we only have to observe our pulsars to make timing measurements every few weeks or so. The detection rates for supermassive black hole mergers will tell us how (or whether) galaxies were built by these events during the early evolution of the universe.

See the following graph (courtesy of Nan Rendong, chief scientist for FAST) for a comparison of the strain and gravitational wave frequency dependence of LIGO and its planned successors, plus those for FAST with the specified pulsar timing array accuracy (PTA):




Read more here:

• Gravitational Wave, Pulsar and Pulsar Timing Array on Wikipedia
• "The Five-Hundred-Meter Aperture Spherical Radio Telescope (FAST) Project", Rendong Nan et al., International Journal of Modern Physics D.

An article by one of the developers of the Thirty Metre Telescope about how its adaptive optics will work. Resolution of the TMT will be around 0.01 arcsecond -- more than ten times better than the Hubble Space Telescope -- which is fairly amazing for a ground-based scope.

http://live.iop-pp01.agh.sleek.net/physicsworld/reader/#!edition/editions_Astro-2016/article/page-16973

We have binocular vision for 3D, so how far apart would optical telescopes need to be for a 3 D image of Jupiter? or Proxima for that matter?

Could LOT work in conjunction with Hubble to do this?