Finding Twin Earths Is Harder Than We Thought

Matt_dk writes "Does a twin Earth exist somewhere in our galaxy? Astronomers are getting closer and closer to finding an Earth-sized planet in an Earth-like orbit. NASA's Kepler spacecraft just launched to find such worlds. Once the search succeeds, the next questions driving research will be: Is that planet habitable? Does it have an Earth-like atmosphere? Answering those questions will not be easy. 'We'll have to be really lucky to decipher an Earth-like planet's atmosphere during a transit event so that we can tell it is Earth-like,' said Kaltenegger. 'We will need to add up many transits to do so — hundreds of them, even for stars as close as 20 light-years away.'" The abstract of their paper offers a link to the complete paper as a 17-page PDF; here is a short description from 2007 of the same researchers' work, outlining the type of spectral signature that an Earth-like atmosphere would be expected to show.

Comments

In effect, what they are saying, is

[Jane Q. Public]

that it will take hundreds of years to tell if they are truly Earth-like. And that is complete nonsense.

Once we find a sufficient collection of candidate planets using this instrument, we can devise a different device/experiment to narrow down whether they are Earth-like. That should take maybe a few years to ten years.

That is more-or-less the pattern we have been following, and it has been successful so far. I see no reason to change.

Re:In effect, what they are saying, is

[Nyeerrmm]

From what I can tell in a brief skim, it really does pose a fundamental limit given current technology. The problem is that with the largest mirror we can imagine getting up into space, and with the highest sensitivity sensors, the signal-to-noise ratio is still too low to get a usable measurement without taking hundreds of measurements.

They plan to detect the chemicals in the atmosphere by measuring the absorption bands in the starlight as some of it passes through the atmosphere. This is presumably going to be a lot more sensitive than trying to detect the light from the planet directly, since you have a lot more photons to carry the information. The signal to noise ratio in this case is really limited by the unfortunate fact that light energy is discretized and you can't make finer measurements than a single photon. Thus a large mirror with a high-quantum efficiency (95%) sensor, is really the best you can do.

The only hope to improve this is to either get bigger mirrors, which really depends on improving space access and is unlikely to give order of magnitude improvements, or to implement an as yet unrealized method that is able to get more information. If it were a problem of angular resolution there are plenty of interesting tricks you could use to improve it. Unfortunately I can't think of anything better, and it doesn't seem anyone else has yet either. Of course, that doesn't mean no one will... but its not as simple as just designing the next mission.

Actually... random 3:30 am idea... if you did something in thermal-IR, and measured the absorption of the blackbody emissions of the planet by the atmosphere you might be able to get something working. The intensity would be a lot lower than looking at the stars light, but the dimming due to absorption would be much larger percentage-wise... although it would take some heavy math to show if it would actually give you a better SNR. Of course, there are plenty of holes here: among other things, my knowledge of atmospheric chemistry and absorption is very limited, and this would all depend on being able to resolve the star separate from the planet, and would thus rely on some complicated interferometric methods....... and you'd have to block out the star light to be able to get the planet light as anything more than noise... and probably the number of photons in thermal IR from a planet are too low to be able to even see it on its own... but maybe I'm wrong and it could work, or something else can.

Important distinction:

[RyanFenton]

There's an important distinction between it being hard to find an earthlike planet, and there not being an earthlike planet to find at all.

Our mechanisms for finding planets are all in wobbles in the wavelengths from the light of stars. And because of that, we tend to only see the big wobbles, because small wobbles tend to get lost in the noise.

It would be nice if we could shine a flashlight and get a real look out there, but in most cases, we'd never see what we shone light upon in our lifetimes.

The universe is a HUGE freakin place, filled mostly with stuff we can't get a good clear look at yet.

Entire worlds like ours are are both all we know, but at the same time, are too small for us to even notice in the grandness just outside our atmospheric window.

Ryan Fenton

Re:Wow!

[Jurily]

Second, even if they did, how in the world do you conclude that would be "less distressing"?? One does not follow from the other.

Well, duh. If they have advanced AI, they probably have internet as well. Which means we can view alien porn while we're being wiped out.

Bullets vs. Energy beams

[firmamentalfalcon]

Bullets hurt people because of human blood circulation (loss of blood) and the size of our organs (heh). If robots were built differently or little green men evolved differently, bullets would most likely be ineffective. There is no reason that there is only one wire connecting processor to leg and opening one loop should not hurt the other parts of the circuit. Also, there is almost no reason why the processor needs to be 15 cm big, or the leg motor has to take up the whole length of the leg. There is also no reason why the robots or green guys have to be human size.

However, as long as they are still made of molecules, high amounts of energy should still be able to separate the molecules that they are composed of, and hopefully eliminate them.