Posted by
Hemos
on from the it's-like-the-earth-baby dept.
jrb writes "For the first time, a small planet (i.e. non-gas giant sized) has potentially been found outside the solar system, helped by a gravitational lensing effect that magnified it. The BBC is carrying the full story. "
Science is inductive. A theory is proposed, then grows in acceptance as a large body of data is found to agree with the theory, while none is found that contradicts it.
We've had theories of solar system evolution, planetary spacing, formation of asteroid belts, and others, for a long time. But these theories are very unsatisfying because we only have one datapoint, our own solar system.
This may not seem important, if you believe that the solar system is completely typical. By the nature of a bell curve, most likely our system is in most respects. But there is very little evidence it is yet. We were already pretty certain that gas giant planets are not uncommon. This is the very first evidence that small rocky planets may not be uncommon either.
According to my calculations the feat is very roughly equivalent to detecting a speck one micrometer in radius at a distance of two kilometers, so I'm impressed anyway.
Just trying to wrap up some points that were being put out here, and maybe answer some questions in the meantime with my intermediate knowledge of astronomy.
Finding this planet, if the evidence leans towards that theory, is a big deal, as so far all we have found arround other systems are very big gas giants. One solar system is really bad for statistical analysis. We could be the fluke of the universe, so just because it happened here doesn't mean it had to happen somewhere else.
Nearly all planet observations, an really all astronomical observations, are objects infered by the bizare behavior of well lit objects. Faint changes in spectrum of a single bright object, means that it is probably a binary system... etc. The best analogy I have heard about astronomy is a goldfish trying to figure out what the world outside its pond is like. It can never go there, but can learn from indirect observations.
I assume that the gravity lensing difference between the two stars can easily be picked out because although they both throw a lot of light, they don't have the same spectrum, and probably not even the same redshift. You can then subtract out the closer star because you very carefully observed it when there was nothing significant behind it.
I don't claim to be an expert on such issues, but hopefully someone got something from my little rant here.
-- There is no silver bullet. Plus, werewolves make better
neighbors than zombies or vampires anyway.
After a report in this week's new scientist about rocky planets being formed by gamma ray bursts, I was a wee bit worried. If this is such a planet, there's no need to divide the Drake Equation by a thousand.
Having been lucky enough to do a literature review on this topic recently (as part of my 3rd year u/g course) I can clear up a couple of issues;
The method used works as follows; when gravitational field of the planet warps the space around it, any light from the star that might otherwise have 'missed' the telescope/eye/pinhole camera (!) would be 'bent' back to the aforementioned instrument.
Hence we do _not_ see the planet, rather the effect of the planet on a star which is how all extrasolar planet detection methods (except one which has failed to date) work.
We have no instruments capable of resolving a planet, but NASA & ESA both havbe projects that in 2020-2060 will be able to do so at IR frequencies. Hence the BBC picture is wrong. All it pointed out was the star.
This method is not repeatable, since it relies on a chance that a background star acts as the source and the planet in orbit around an unseen star all line up for us.
You might think, 'doesn't the planet star lens the background one?' - it does! The additional blip caused by the planet on the light curve is what gives it away.
The typical distance to the background star (usually in the galactic plane) is 100 parsecs, the planet's parent star is usually half this distance for geometric reasons.
Hence it's really far away! We can tell virtually nothing about the planet apart from it's mass (which won't help diffrentiate between tiny gas giants and big terrestrial types).
If anyone want's more info (or even <gasp> a copy of my lit review, written for an intelligent person) then email me. Dosvidania tovarish!
-- --
Thus conscience does make cowards of us all - Hamlet
The rest mass of a photon (particle of electro-magnetic radiation, which includes all frequencies of visible light) is 0 - not 0."some very small amount" but just plain 0. And, in the absence of electro-magnetic fields, a photon has speed c in all relativistic frames of reference. (hence "c" is the "speed of light", which is "invariant in a vacuum").
The simplest way to think of what is actually happening is to think of space itself being "bent" by gravity and so the path of light through that space is not straight in the classical, Euclidean/Cartesian sense.
Another way to describe what is happening, without having to understand what is meant by space being "bent" (after all, any N-dimensional manifold can be embedded in a 2*N [-1?] - dimensional Euclidean space) is that light travels along a path in space-time with minimum seperation, where seperation is a 4-dimensional measure, somewhat akin to distance, determined by the metric tensor of the space-time traversed. In the presence of a gravitational field caused by mass (actually any gravitational field - but thats an even weirder subject), the metric tensor differs from that of Euclidean 4-space, so the path of a photon is NOT a Euclidean straight line.
(Of course, the simplest approach is just to say that gravity bends light and not try to explain why;-)
Re:The only problem with looking for RF signals...
by
dentin
·
· Score: 3
Actually, I think your estimate is a little low. I've been fiddling with some of these questions for a while now, and even have a sort of simulator that generates star systems. It looks (as near as I can tell) that about one in 10 star systems has a planet with liquid water, reasonable gravity, and appropriate temperature range.
Also, current thinking is that the odds of life happening on such a planet is fairly high... on the order of 10 to 50% (from various abiogenesis experiments). Of course we only have one real data point, but the evidence seems to point to the idea that life isnt that hard to make.
The probability of intelligence is significantly lower, but once they have intelligence the probability of rf technology is effectively 1, so that term vanishes as well. I don't think intelligence is that rare, and that after 5 billion years of evolution, I would put this factor around.5. You are free to use your own value of course, but given that semi-intelligent creatures abound on this rock, I don't think intelligent creatures who can use tools are that far off.
Probability of emission frequency if fairly low as well, however not quite as low as you would think: There are certain bands that are the best for transmitting in. Most of the spectrum is filled with broadband noise, and there are a few marker frequencies that would be the most efficient/effective to transmit on. Instead of 1e-6, I'd be a bit more conservative and put it at 1e-4.
Of course there is one more term you forgot to mention: the length of time an alien race might transmit such a signal. This is pretty much anyone's guess, but id place it at no more than 500 years - which is a really short period of time. This factor should be divided by the average age of the stars we will be looking at, which would be about 5 billion years. This factor alone works out to 1e-7.
So the net result is.1(star with planet)*.1(planet with life)*.5(tool/rf using life)*1e-4(proper frequency band)*1e-7(prob we will catch them transmitting)
This works out to about 5e-14 per star, which is still pretty low, but not 1e-24. Also, we can get rid of the 1e-4 factor by improving our detection technology. Additionally, the 1e-7 number may be significantly larger if electromagnetics end up being the only way to communicate across large distances. I wouldn't expect much EM radiation from the planet though, as eventually everything would go to cable/fiber optics instead of radiated waves.
So while the odds are still highly against us, they arent quite as bad as you depict and we can increase them over time.
-dentin
-- Alter Aeon Multiclass MUD - http://www.alteraeon.com
Not the first Earth-sized planets, either
by
J05H
·
· Score: 3
Just the first one found around a main sequence or nearly-main sequence star. In 1989, three Earth sized (well, one is Mars sized, but close enough) planets were discovered orbitting a pulsar. They are obviously dead planets, like their star, but they always fail to be mentioned, especially in the mainstream media. Anyway, check out the Extrasolar Planets Encyclopedia for more info on all of this.
Finding planets this way is a really haphazard way of doing it. Stars rarely line up well enough to make gravitational lensing really viable as a method of detecting another planet. Another method they've been using is watching the Doppler shift of a selected star. Any star with an object revolving around it exibits a regular 'wobble' in the shift. Make a guess at the mass of the star, apply some centuries old math to it, and voila! You know how many objects are orbiting the star, how massive they are and how far away from the star!.
Astronomy is a science where you can not repeat your experiment (the universe). Whenever you get a result, in this case the result is that we live on a planet, you have to spend a long time considering any possible biases. The fact that we'd be dead if we weren't on a planet is a pretty big bias towards finding ourselves on one, even if it's the only planet in the universe.
As for it being pretty obvious that there are other planets out there, 1000 years ago it was pretty obvious that the earth was flat.
Slashdot Mirror
Science is inductive. A theory is proposed, then grows in acceptance as a large body of data is found to agree with the theory, while none is found that contradicts it.
We've had theories of solar system evolution, planetary spacing, formation of asteroid belts, and others, for a long time. But these theories are very unsatisfying because we only have one datapoint, our own solar system.
This may not seem important, if you believe that the solar system is completely typical. By the nature of a bell curve, most likely our system is in most respects. But there is very little evidence it is yet. We were already pretty certain that gas giant planets are not uncommon. This is the very first evidence that small rocky planets may not be uncommon either.
According to my calculations the feat is very roughly equivalent to detecting a speck one micrometer in radius at a distance of two kilometers, so I'm impressed anyway.
Just trying to wrap up some points that were being put out here, and maybe answer some questions in the meantime with my intermediate knowledge of astronomy.
I don't claim to be an expert on such issues, but hopefully someone got something from my little rant here.
There is no silver bullet. Plus, werewolves make better neighbors than zombies or vampires anyway.
After a report in this week's new scientist about rocky planets being formed by gamma ray bursts, I was a wee bit worried. If this is such a planet, there's no need to divide the Drake Equation by a thousand.
The method used works as follows; when gravitational field of the planet warps the space around it, any light from the star that might otherwise have 'missed' the telescope/eye/pinhole camera (!) would be 'bent' back to the aforementioned instrument.
Hence we do _not_ see the planet, rather the effect of the planet on a star which is how all extrasolar planet detection methods (except one which has failed to date) work.
We have no instruments capable of resolving a planet, but NASA & ESA both havbe projects that in 2020-2060 will be able to do so at IR frequencies. Hence the BBC picture is wrong. All it pointed out was the star.
This method is not repeatable, since it relies on a chance that a background star acts as the source and the planet in orbit around an unseen star all line up for us.
You might think, 'doesn't the planet star lens the background one?' - it does! The additional blip caused by the planet on the light curve is what gives it away.
The typical distance to the background star (usually in the galactic plane) is 100 parsecs, the planet's parent star is usually half this distance for geometric reasons.
Hence it's really far away! We can tell virtually nothing about the planet apart from it's mass (which won't help diffrentiate between tiny gas giants and big terrestrial types).
If anyone want's more info (or even <gasp> a copy of my lit review, written for an intelligent person) then email me. Dosvidania tovarish!
-- Thus conscience does make cowards of us all - Hamlet
Well, uh, no.
;-)
The rest mass of a photon (particle of electro-magnetic radiation, which includes all frequencies of visible light) is 0 - not 0."some very small amount" but just plain 0. And, in the absence of electro-magnetic fields, a photon has speed c in all relativistic frames of reference. (hence "c" is the "speed of light", which is "invariant in a vacuum").
The simplest way to think of what is actually happening is to think of space itself being "bent" by gravity and so the path of light through that space is not straight in the classical, Euclidean/Cartesian sense.
Another way to describe what is happening, without having to understand what is meant by space being "bent" (after all, any N-dimensional manifold can be embedded in a 2*N [-1?] - dimensional Euclidean space) is that light travels along a path in space-time with minimum seperation, where seperation is a 4-dimensional measure, somewhat akin to distance, determined by the metric tensor of the space-time traversed. In the presence of a gravitational field caused by mass (actually any gravitational field - but thats an even weirder subject), the metric tensor differs from that of Euclidean 4-space, so the path of a photon is NOT a Euclidean straight line.
(Of course, the simplest approach is just to say that gravity bends light and not try to explain why
Actually, I think your estimate is a little low. I've been fiddling with some of these questions for a while now, and even have a sort of simulator that generates star systems. It looks (as near as I can tell) that about one in 10 star systems has a planet with liquid water, reasonable gravity, and appropriate temperature range.
.5. You are free to use your own value of course, but given that semi-intelligent creatures abound on this rock, I don't think intelligent creatures who can use tools are that far off.
.1(star with planet)*.1(planet with life)*.5(tool/rf using life)*1e-4(proper frequency band)*1e-7(prob we will catch them transmitting)
Also, current thinking is that the odds of life happening on such a planet is fairly high... on the order of 10 to 50% (from various abiogenesis experiments). Of course we only have one real data point, but the evidence seems to point to the idea that life isnt that hard to make.
The probability of intelligence is significantly lower, but once they have intelligence the probability of rf technology is effectively 1, so that term vanishes as well. I don't think intelligence is that rare, and that after 5 billion years of evolution, I would put this factor around
Probability of emission frequency if fairly low as well, however not quite as low as you would think: There are certain bands that are the best for transmitting in. Most of the spectrum is filled with broadband noise, and there are a few marker frequencies that would be the most efficient/effective to transmit on. Instead of 1e-6, I'd be a bit more conservative and put it at 1e-4.
Of course there is one more term you forgot to mention: the length of time an alien race might transmit such a signal. This is pretty much anyone's guess, but id place it at no more than 500 years - which is a really short period of time. This factor should be divided by the average age of the stars we will be looking at, which would be about 5 billion years. This factor alone works out to 1e-7.
So the net result is
This works out to about 5e-14 per star, which is still pretty low, but not 1e-24. Also, we can get rid of the 1e-4 factor by improving our detection technology. Additionally, the 1e-7 number may be significantly larger if electromagnetics end up being the only way to communicate across large distances. I wouldn't expect much EM radiation from the planet though, as eventually everything would go to cable/fiber optics instead of radiated waves.
So while the odds are still highly against us, they arent quite as bad as you depict and we can increase them over time.
-dentin
Alter Aeon Multiclass MUD - http://www.alteraeon.com
Just the first one found around a main sequence or nearly-main sequence star.
In 1989, three Earth sized (well, one is Mars sized, but close enough) planets were discovered
orbitting a pulsar. They are obviously dead planets, like their star, but they always fail
to be mentioned, especially in the mainstream media. Anyway, check out the Extrasolar Planets Encyclopedia for more info on all of this.
gigantino.tv - Heavy but weighs nothing.
Finding planets this way is a really haphazard way of doing it. Stars rarely line up well enough to make gravitational lensing really viable as a method of detecting another planet. Another method they've been using is watching the Doppler shift of a selected star. Any star with an object revolving around it exibits a regular 'wobble' in the shift. Make a guess at the mass of the star, apply some centuries old math to it, and voila! You know how many objects are orbiting the star, how massive they are and how far away from the star!.
.sig: Now legally binding!
It's not silly. It's a perfectly valid point.
Astronomy is a science where you can not repeat
your experiment (the universe). Whenever you
get a result, in this case the result is that
we live on a planet, you have to spend a long
time considering any possible biases. The fact
that we'd be dead if we weren't on a planet is
a pretty big bias towards finding ourselves on
one, even if it's the only planet in the universe.
As for it being pretty obvious that there are
other planets out there, 1000 years ago it was
pretty obvious that the earth was flat.
Ale.