Gooch's dad

OK, I'm really bored at work, and I ended up at Extrasolar Visions, which is a rather cool site with information about some of the extrasolar planets that have been discovered so far.

All of these extrasolar planets are Jupiter-sized gas giants, some a bit bigger, some smaller. I guess we don't find them by seeing them, but by the wobble in the star they are orbiting.

Being an optics geek, I thought I'd figure out how big of a telescope we would need to see an earth-sized planet orbiting Gliese, which is 15 or so light years away.

As many of you know, this is determined by the Rayleigh criterion and diffraction theory. The minimum resolvable angle is

theta = 1.22*lambda/d

where d is the diameter of the telescope entrance aperture.

So to just barely resolve an earth-sized planet orbiting a star at that distance, we'd need a telescope with a mirror over 2 km diameter.

Shadowy Man will probably point out a half-dozen bad assumptions I'm making, and I think the first one would be that the star light would just swamp any detector that tried to resolve the reflection from the planet. Fair enough, I figure there's a way around that.

What if we want to really get a good look at the planet? To resolve features 2 km apart would require a mirror of roughly the diameter of the earth.

Now, that doesn't mean that we would have to *have* that big of a mirror. It might be possible to do an optical version of Simulated Aperture Radar. I gather that is what the radio telescopes do, too--gather data far apart with sufficient timebase accuracy that you can 'put the data back together' (as I recall it involves a Fourier transform, no big deal).

So how long until we can see with that accuracy? It sort of reminds me of the 19th century predictions (I don't recall who made them) that we could never communicate with Martians, because we would have to have signal flags as large as Britain. The possibility that other means of communication would arise was not considered.

I'd bet that within 50 years we'll be able to resolve planets of that size, at that distance.

Shadowy Man

Take a look at the following two NASA Missions:


Space Interferometry Mission

and

Terrestrial Planet Finder

Shadowy Man

Yes, one way around it is with aperture masks.

Check out Debes & Ge (astro-ph/0301051) , for example.

I was Debes' freshman physics TA.

Gooch's dad

Whoa....fantastic stuff! Silly me, using 19th century optics on the problem.

So can they use the interferometric techniques to assemble good images?

It looks like 50 years was way too long--they're talking about SIM launching in 2009, and finalising plans for the Terrestrial Planet Finder in 2006.

What sort of resolution do you think they can achieve?

Shadowy Man

A quick jaunt through abstracts in the literature provides me with the following information on SIM:

This is old info, and changes to the design may have been made, but in essence I think it is correct. Other abstracts talk about microarcsecond astrometry.

I think TPF will do spectroscopy. So, it will look for absorption from various molecules as "biomarkers".

Lobstrosity

Also check out the Keck interferometer (google it for more links if you wish):

http://spacelink.nasa.gov/NASA.Proje...nterferometer/

The Keck telescopes are currently the largest telescopes in existence. Each one is 10 m in diameter. They are located 14,000 feet up on the summit of Mauna Kea (on the Big Island of Hawaii) and use adaptive optics to remove atmospheric distortions. They are currently working to link the two telescopes interferometrically; since the baseline between the telescopes is around 80 m, this will provide the effective resolution of a 100 m telescope. Though a space interferometry mission is far more exciting, the Keck inteferometer should be complete and generating data much sooner.