6
$\begingroup$

This question has already been partly addressed here: When will we have the technology to directly observe an exoplanet with significant clarity?

However, my question is focused more on the limitations of space telescope construction. Specifically, I'm curious how large we would need to make a telescope to image the surface of an exoplanet. My napkin math:

If JWST can image a penny at 40km, they can image 19,050,000 km^2 at Proxima b. Therefore, to detect an earth-sized planet visually we would need a JWST equivalent that is 1,000 times larger. However, this probably wouldn't give any details, so we would need an even larger telescope to see anything more.

Making a telescope this large doesn't seem that far beyond the realm of possibility. Sure, JWST was stupidly expensive, but now that we know how to build it, mass-producing parts for a bigger one doesn't sound all that intractable. Moreover, if Starship works, putting that much mass in orbit becomes relatively inexpensive (~$500 million for 1000x JWST mass via napkin math, so probably double/quadruple that).

My questions are:

  1. Do telescopes scale linearly like this? Could we just make a WAY bigger JWST, and assemble it in orbit, using multiple Starship launches? Also, am I over/under estimating the size differential needed to get a decent resolution?
  2. Are they any new proposals for super-massive telescopes that we could assemble using Starship's ability to put tons of payload into orbit for cheap?
$\endgroup$
  • $\begingroup$ I know on the JWST the mirror alignment became a problem. Its not just an issue of scaling up, you also have to maintain mirror precision. One of the reasons JWST has all the mirror segments is so that it can do alignment procedures on-orbit to get good data. I imagine this would be shockingly expensive to do at scale for a several Km telescope. $\endgroup$ – Knudsen Number Apr 9 at 3:40
  • 1
    $\begingroup$ @KnudsenNumber the main reasons why it uses segments is so that it can be folded for launch. $\endgroup$ – Antzi Apr 9 at 5:27
  • $\begingroup$ @Antzi I did say "one of the reasons" not the main reason. Of course it has to be folded to fit in the fairing, but it has 18 segments and only actually folds up into three main pieces. If it was just about the folding you wouldn't need so many segments, but the focusing of the mirror is done be actuators on the back. See here for proof jwst.nasa.gov/mirrors.html $\endgroup$ – Knudsen Number Apr 9 at 14:56
8
$\begingroup$

So the main thing of interest in this post seems to be angular resolution. You're after objects that are relatively bright and small. Angular resolution increases linearly with aperture size. However, you don't necessarily need a giant light-bucket of a telescope to achieve high resolution. You can achieve some impressive resolution through use of interferometry. Interferometry is essentially about combining the light from multiple telescopes. The distance between the two telescopes gives a resolution that's about the same as if you used a single giant telescope with a diameter the same as the separation between the two telescopes. Obviously, as the light-collecting surface area is much lower, the two interferometric telescopes won't be able to pick up faint objects nearly as well as a single massive one.

The next 'flagship' optical space telescope is currently called ATLAST, and is planned for development and launch in the 2030's (which practically speaking, given the usual development cycle, means we probably won't see it till 2050 at the earliest, if it ever even makes it out of development). The design is still in the concept stage, and the primary mirror is slated to be somewhere between 8 and 16 metres in diameter (compare the 6.4 metre diameter JWST and the 2.4 metre Hubble). At it's largest, this would provide an angular resolution about 6.6x that of Hubble.

However, this resolution isn't nearly enough to resolve red giants with, let alone exoplanets.

The VLT, by virtue of the separation of its individual telescopes (about 130 m between 1 and 4 as measured by google maps because I couldn't find an official value), is able to achieve a resolution of 0.002 arc seconds, which is about 50x better than Hubble. We are now beginning to be able to see blurry images of the largest nearby red giants.

Antares - Wikimedia commons

Picture of Antares, a red supergiant, obtained from VLT. Other imaged stars can be found here

So, what do we have to do to obtain an image of an exoplanet of similar quality?

The largest exoplanets are about ~10,000 times smaller than the largest stars. If you want similar resolution to the above picture, you need 10,000 times the effective diameter (separation) of the VLT. That's about 1,300 km, or about the 'width' of the Arabian peninsula from the Red Sea to the Persian Gulf.

I'm no astronomer, so I don't know the practical challenges of building something with that separation (whether in space or on the ground). I do know that the distance between the two telescopes must be known exactly, and not allowed to deviate. Putting two telescopes on (say) either side of the Great Rift Valley, would be a bad idea.

What about in space? Again, problems with the distance between the two telescopes will occur. The distance must remain exactly the same between them and the object that combines and records the light. The linked question in your question has an answer that gives some of the monumental difficulties with designing such a telescope.

In summary In order to image exoplanets, you don't want a single big telescope, you want a set of smaller telescopes with a (very) wide separation between them, so you can do interferometry. But even for interferometry, the odds are stacked against you due to the engineering challenge of having a megametre-long but precisely measured distance between them.

$\endgroup$
  • 1
    $\begingroup$ For placing two telescopes a known distance from each others the surface of the moon seems like a nice option. Also, no pesky atmospher. $\endgroup$ – lijat Apr 9 at 6:07
  • $\begingroup$ I agree. In the far future I can imagine the moon bristling like a hedgehog with such telescopes. $\endgroup$ – Ingolifs Apr 9 at 7:25
  • 2
    $\begingroup$ Could you use lasers to tell exactly how the distance between the telescopes changes, even if they move a bit? $\endgroup$ – Bert Haddad Apr 9 at 7:27
  • 1
    $\begingroup$ Also, how many stations do you need for interferometry? Could you just spread a few hundred out over a wider area, or do you need a concentration similar to the vlt over that area? $\endgroup$ – Bert Haddad Apr 9 at 7:29
  • 1
    $\begingroup$ I had exactly the same idea too. Don't quote me on this, but I believe the problem boils down to the fact we can't detect a photon's phase when it hits a detector. The light itself needs to combine and its waves interfere (hence interferometry) before it enters any sort of detector. I would recommend asking this as a separate question on physics.SE. (otherwise I probably will, given a bit of free time). Ditto your second question. I don't know. I think you just need two, but more can't hurt. it would be a good physics.SE question. $\endgroup$ – Ingolifs Apr 9 at 7:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.