Could telescopes eventually clearly show extra-solar planets?

October 8, 2006 6:42 AM

Could telescopes eventually clearly show extra-solar planets?

The further away objects are, the more difficult they are to see: My understanding of why this is so, is that less light from them reaches us. And/or that the light from them diverges, so you need a lens bring the object into focus.

My understanding of telescopes is that they have large lenses that improve focus, and more light-sensitive digital sensors that discern very faint objects. And positioning them in space avoids interference from the Earth's atmosphere.

Since both lenses and light-sensors are continually improving, is it plausible that we will eventually have telescopes that allow us to view planets 100s of light years away as clearly as those in our own solar system? Or is there a physical limit to how much we could improve telescopes?

posted by snarfois to science & nature (12 comments total)

Aside from all of the technical issues with finding a planet in the first place, I think we might eventually run into issues just based on the idea that there's not enough photons to go around.

You probably know that right now, all we can do is detect gas giants, and that's only in situations where the planet's orbit is parallel to our line of sight to its parent star. Smaller, Earth-like planets would be a lot harder.

Of course, if we don't get enough photons coming to us bouncing off of that planet, we can always get a bigger lens. Aside from financial and engineering concerns, I guess the only thing that would really stop you is if there were a great deal of interstellar dust in the way.
posted by adipocere at 6:54 AM on October 8, 2006

The physical limit is how big you can build your telescope. To actually see extrasolar planets at the same resolution that we see, say, Jupiter from Earth (or from the Hubble) would require a really, really big telescope (or really big array of telescopes). What it comes down to in the end is how many photos you can gather.
posted by moonbiter at 7:03 AM on October 8, 2006

Actually, NASA is planning for a space telescope designed to directly detect and image planets around nearby stars. It would actually be two telescopes working together as an interferometer. (I have some doubts about the design, but that's what they've proposed.)

But it wouldn't be anything like as clear as the images we get of planets inside this solar system. For certain physical reasons your ability to discern detail at a distance is a function of the width of your light-capturing system (i.e. your mirror or mirrors). As a practical matter, that means using interferometers. The best optical interferometer in the world right now is the VLTI in Chile. The largest radio interferometer ever created was twice the radius of the orbit of the earth, when the same radio telescope observed the same target 6 months apart and combined the data. (That isn't possible with optical wavelengths.)

What NASA is hoping is that they'll be able to pick out planets as specks of reflected light. But resolving them into discs with surface detail is out of the question.
posted by Steven C. Den Beste at 7:15 AM on October 8, 2006 [1 favorite]

(interim answer, before someone who remembers their physics and/or is up to speed with modern astronomy gets here...)

There are two major factors with a telescope; the angular resolution, and the area for light collection. Up until now, optical telescopes have mainly been made with a single reflecting mirror, so the two have been intimately related (ie a big mirror gets you more light collection and better resolution.)

In radio astronomy because the wavelengths involved are larger it has been possible to de-couple the two factors; you can combine the output of two seperate telescopes to achieve a higher resolution. This leads to Very Long Baseline Interferomtery, which can get you a telescope with the equivalent 'size' of the continental US. (qv The VLBA) On the other hand you have telescopes like Arecibo telescope which can detect extremely weak signals due to it's large area.

As technology has improved, it's become possible to use interferometry at visible wavelengths. The VLT in Hawaii is a terrestrial example which, while it should be able to achieve stunning resolution, won't be able to image planets directly.

Beginning (I believe) with the Space Interferometery Mission, NASA are exploring doing optical interferomery in space. SIM is an integrated 9m baseline, but they're looking into developing probe 'swarms', which'd be able to keep station with one another to the precision required to get seriously large (100+ m) baselines, with a corresponding increase in resolution. (Plus the optical benefits of being in space.)

(Wikipedia also talks about Labeyrie's hypertelescope, which is new to me)
posted by Luddite at 7:59 AM on October 8, 2006

So, in summary (forgot that bit) - it's theoretically possible, probably using optical interferometry and large mirrors, either in space, on the moon, or on earth, but the technology is still hard and expensive. Assuming the technology is developed incrementally (James Webb telescope, and larger sucessors, experimental station keeping satellites, gradually larger baselines) I'd guess 40+ years.
posted by Luddite at 8:21 AM on October 8, 2006

This recent Scientific American article explains how and why to build new, high resolving power telescopes.
posted by meehawl at 9:03 AM on October 8, 2006

It's already happened: The First Image of an Extra Solar Planet , in this case an image that is believed to show a planet about 8 times the mass of jupiter in orbit around a brown dwarf.
posted by jepler at 9:06 AM on October 8, 2006

Imaging extrasolar planets is tough. As I understand it, they often prove the existence of planets by measuring the gravitational effects exerted on starts. Any big planet will cause a small amount of 'wobble', and this tells them that there's a planet out there.
posted by chrisamiller at 11:20 AM on October 8, 2006

As far as getting an image such as we have of Jupiter, where we can see surface features, I can't see any easy way of separating out photons from the star the planet is revolving around from the photons reflected from the planet no matter how big your telescope is.

Did we have any good images of Mercury before flybys?
posted by jamjam at 11:52 AM on October 8, 2006

Thanks for the insightful answers! While I guess I won't see an extrasolar planet as anything more than a disc in my lifetime (unless Labeyrie gets his way), it looks like there will be steady improvements to look forward to.

I had a secondary purpose with this question: I wanted to test a hunch that I'd get far better answers here than on Yahoo Answers. Thanks for proving the hunch right.
posted by snarfois at 12:49 PM on October 8, 2006

There were no good images of Mercury before the Mariner 10 flyby, and there still aren't any except the ones Mariner 10 returned.

But apparently there has been some since then by the Mt. Wilson 60-inch scope.

Still, we won't really have a good idea of what Mercury looks like on until the MESSENGER mission is well under way.
posted by Steven C. Den Beste at 3:43 PM on October 8, 2006

I work on TPF-C, one of the two Terrestrial Planet Finder missions, and I'd like to note that interferometry is not the only proposal NASA is making to look at planets--a coronagraphic space telescope is also in the works.

Coronagraphs work by taking all the light we don't want (the starlight), and either blocking it or sending it into places we don't care about. The simplest coronagraph is the Lyot coronagraph, in which the starlight is focused through a lens, the center is blocked with a small stop, and the remaining light is passed through another lens and aperture, to leave the starlight removed and faint surrounding objects unaffected. This is what was first used to image the sun's corona (hence the name), and is still used in solar spacecraft like SOHO. (More details and history here.)

Unfortunately, a simple coronagraph like this wouldn't be good enough to see a planet--in the visible spectrum, a star is expected to be 10 billion times brighter than any planets around it. There are a number of proposals being bandied about on how to achieve this "10^10 contrast" for TPF, ranging from the baseline (band-limited masks) to the extremely complex (external occulters), and many options in between. The exact design is yet to be decided, but all the above options would let you "look" at a planet. (You'd probably see it on only a few pixels of the CCD at best, so it wouldn't make a very exciting image, but you can tell a lot of things from the spectrum, like atmospheric composition.)

TPF-I (the interferometry mission Steven C. Den Beste mentioned above, though with five craft, not two) I don't know quite so much about, but I figured it's worth mentioning that the two are complementary. TPF-C is in visible wavelengths, TPF-I is designed for IR. (Well, if they ever get off the ground--funding for unmanned science got cut by NASA in favor of manned missions. Good thing Europe and Japan are interested in finding planets, too.)

Also, ground-based planet-finding is a hot topic--see HiCIAO at Subaru or the Gemini Planet Imager for examples. (It's pretty much impossible to see "Earths" from the ground, but "Jupiters" are reasonable--they're a couple of orders of magnitude brighter.) I suspect we'll be seeing Jupiter-sized planets around main sequence stars (not just brown dwarfs) within a couple years, and Earth-size planets once someone gets a telescope (interferometer or coronagraph) up to space.
posted by Upton O'Good at 3:52 PM on October 8, 2006 [1 favorite]