How likely is it that we’ll be able to make images of extra-solar planets? By “image” here I mean something that at can serve as at least a very crude map. Shows at least differences between some “darker” regions and some “lighter” regions…
I mean without actually traveling or sending probes to those solar systems, of course.
If this may be possible someday, how long is it likely to take before it’s possible?
We have some images now. For example, one of the planets orbiting Formalhaut (the “Eye of Sauron”) has been imaged, as well as GQ Lupi b. More recently, there were more sophisticated images taken such as those of HR8799.
Never. We have a hard enough time resolving the planet from the star, what you seem to want to do is to resolve different parts of the planet. This is almost impossible. Almost.
First off, it depends on the planet you want to view. If you are looking for Earth-like planets, it is difficult to even resolve them at all. Coronagraphy, the technique used to image the planets that Crescend linked to, is really only effective for nearby stars and planets that have a large angular separation from their parent star. This technique is used to block the light from the parent star, a step that is necessary as the irradiance contrast between the star and its planet can vary from 10[sup]10[/sup] in the visible spectrum to 10[sup]3[/sup] in the mid IR. But say you were able to block out this light (using Coronagraphy or nulling interferometry), you still need to resolve the different parts (the darker and lighter regions) of the planet. The Rayleigh Criterion states that in order to resolve two point sources with an angular separation of [symbol]q[/symbol], you need a telescope with an aperture diamter of 1.22 [symbol]l[/symbol]/[symbol]q[/symbol] (where [symbol]l[/symbol] is the wavelength). This means to resolve two spots on Earth’s diameter from a measly 15 light years distant (note: only 40 or so stars are this close to Earth), you would need a telescope with a diameter of ~4.6 miles to image in the visible spectrum. Not too bad, but remember that the parent star will be 10[sup]10[/sup] times brighter in this region of the spectrum making viewing the planet next to impossible. If you switch to the IR (say around 12 [symbol]m[/symbol]m) where the contrast ratio is only 1000:1, you would need a telescope over 100 miles in diameter (you would also need to cool your optics with liquid helium as they would glow pretty brightly at this wavelength - check out the blackbody radiation spectra).
There are techniques to get around the need for gigantic telescopes, namely Fizeau Imaging or aperture synthesis. Simply, this technique combines the light from multiple telescopes to create an image that has the same resolution of a telescope with a diameter equal to the distance between the telescopes. The Very Large Array in New Mexico operates on this principle. With this technique it is necessary for the telescopes to measure both the irradiance (intensity) and the phase of the incoming light. With radio waves, this is easy, with visible light or IR; not so much. Luckily, light from distance objects like stars is spatially coherent in much the same way as a laser. You can capture light from stars and combine it to make interference fringes or write holograms. Because of this you do not need to measure the absolute phase of the light, you merely need to interfere it to get the relative phase between the telescopes. Problem solved. Not. Unfortunately, the degree of coherence falls off with the distance between the telescope apertures. So what does all this mean? The ability to use multiple telescopes to create an image requires that the telescopes are reasonably close together, while the need to resolve images far away needs telescopes far apart. The practical result is that it is just not possible to resolve the “darker regions” and the “lighter regions” on any Earth-like planet, even if that planet was in Alpha Centauri. Who knows though, maybe we will build huge free floating interferometric telescope arrays in space that are able to do it. I doubt it will be in my lifetime however.
Forgive my ignorance, but is it possible to get any useful information about these extrasolar planets from the radio spectrum? Or does that only tell us about things that emit their own radiation?
Here are the best (heavily processed) images we have of Pluto and Ceres. Ceres is closer than Jupiter.
There are only a handful of stars (such as Betelgeuse, a red giant relatively close to oour solar system) that show up as discs instead of point sources in the most powerful telescopes. Imaging an extra-solar planet in detail would be almost impossible, as they are thousands of times smaller than these giant stars.
Personally, I believe that it’s possible - but only when we can build the space telescope arrays that L. G. Butts referred to.
And in terms of assigning a date to that accomplishment - I also hesitate to even guess. Experience has shown that constructing even small things in space is expensive and dangerous, and the public will to allocate much effort, resource, money, and risk to human life to explore our universe is fickle at best.
Even using a giant space-based inteferometer, it would probably be impossible to get a detailed image of an extra-solar planet, like the OP is asking for. An interferometer is a nice shortcut to achieving the same resolution as a much larger telescope, but does not have the equivalent light-gathering power. A very long exposure would be required. The planet’s rotation would make it difficult to get a sharp image.
I’m curious what stars have been resolved optically by interferometry. I know Betelgeuse, which is I think the closest supergiant, has, and that there were one or two other stars (but not which ones). Anyone aware of references listing and linking to images of stars resolved by interferometry?
Only things that emit their own radiation. I believe our best chance is in the near IR as there are some good absorption lines their we can use to infer atmosphere composition and the planets will glow almost as brightly as their parent stars in this band. Note that I am not an astronomer and might be wrong about this.
Good point. Didn’t even think about this.
I looked at potential telescope designs for a project in grad school, and the exposure times were measured in weeks.
It was 1868 when Hippolyte Fizeau suggested that a screen with two holes installed in front of the objective lens of a telescope (See Figure 11) would allow measurement of stellar diameters with diffraction limited resolution. I believe the first such device, the Michelson stellar interferometer, was used in 1920. As for a list, I am not sure it exists. Interferometry is a standard technique in astonomy and is, I believe, the best way to measure the size of stars that cannot be resolved into a disk (i.e., pretty much all of them). There is a History of astronomical interferometry here (warning: postscript file) though I cannot look at it on this machine as I don’t have a ps viewer…
I dont think this is an inherent problem.
If you gather the data over many planetary rotations, I suspect you can mathematically tease out the “original image”, removing the rotational blur so to speak.
I think the bigger issue is likely how long do you have to gather light to have enough data to get a decent image.
Even better way around the rotation problem: The OP wants at least some detail on surface features, but he didn’t specify which surface features he wants. So try to image the polar caps.
In theory, yes, but it’s not a simple problem. You’d need to know the rotation period of the planet with a high degree of precision. As the planet moves in it’s orbit, it’s appearance will change from the telescope’s point of view, as it will be lit from a different angle. And if the planet has any weather, you are pretty screwed. I seriously doubt it would be possible to get a sharp image.
The best images we have of Pluto are a result of computer processing. While scientifically valuable, they aren’t very satisfying to look at, only Pluto’s mother could love them. Hopefully the New Horizon’s space probe will give us a proper look in 2015.
Nice try, but the polar caps on Earth and Mars exhibit large seasonal variations.
OK, so we’ll have motion blur (funny though it might seem to get motion blur from a receding or advancing glacier). But the OP isn’t asking for a high-quality image, and it seems to me that being able to distinguish the equator from the poles meets his standard.
How darn “sharp” do you want for planets orbiting stars light years away anyhow?
As for seasons and the planet moving in its orbit, that IMO only becomes an issue when the “exposures” become on the order of months. And even then, if you know the orbit of the planet, which we will, you should be able to correct for that to some extent. And over many “exposures” over many months, weather would mostly be an additional noise component.
I won’t bet my life on it, but I suspect you would NOT need to know the planets rotation rate to deconvolve the data into an image. I certainly do not need to know the period of a randomly sampled signal to reconstruct it using Fourier Analysis if I have enough data. I suspect there is something equivalent to that for interofermetric (sp?) image processing. For that matter, doing photometry of the whole planet image could probably give you that rotation rate anyway.
IMO, its doable without any massive breakthrough in our math, science, technology, or GDP. Its much more of a how much can we do for XYZ dollars? My WAG is that putting as much time, risk and effort into it as we have for International Space Station Useless Black hole would yield some useful and interesting results.
Oh, and ALL of this data is going to require massive computer processing. Its the very nature of interferometry imaging.
