![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
This was the base of a presentation I gave two years ago for astrobiology. Since it was focused on the same topic I am studying for this semester's project I thought it might help to refresh me by retyping it here. Too bad so much of the information in this is out of date, incomplete or even wrong. Click on childish abstract for for reading 'fun'.
An extrasolar planet is essentially any planet which does not revolve around the sun. The term 'planet' itself, however, is not so firmly established. It has come into dispute over the past few years, since the actual candidates put forward have turned out to be less than identical to the objects with which we are familiar. The fact that the International Astronomical Union has no formal definition of what a planet actually is has provided ample scope for disagreement.
One of the more popular methods to distinguish between 'planets' and 'not planets' is the way in which it formed. The essence of this idea is that anything formed in orbit about a star is a planet, while anything formed elsewhere is not. In particular, planets are formed as secondary products of star formation, while non-planets (actually, they'd be brown dwarfs, wouldn't they?) are formed by direct condensation from nebulae, like stars. The smallest sized brown swarfs that can be formed like this possess roughly ten times the mass of Jupiter, so anything smaller found in interstellar space would by this definition be a planet, and was probably ejected from a star. There have been, so far, no confirmed reports of such objects being found and they will therefore be ignored from now on. This mas (ten times Jupiter's) is also often used to define an upper limit for what is described as a planet when considering objects found orbiting a star, with anything larger being considered a brown dwarf and thereby avoiding all questions of origin. This distinction is useful because, although it is conceivable that planets with more than ten times the mass of Jupiter could be formed, is that it is either ery difficult or impossible to tell from an object's spectrum whether it is a small brown dwarf or a very large planet. This system therefore makes things much simpler (some would say too simple) than they would otherwise be.
While there has been much discussion of upper limits for planetary masses, the smaller end of the scale seems to have been largely ignored. This may be at least partly because there is little hope of detecting such objects in the near future and there is therefore little sense of urgency to the issue Problems also arise from the fact that there is difficulty finding a point to draw the line that does not seem arbitrary. It seems natural to define a planet as any non-fusing object (a fusing object being a star, by definition) formed in orbit about a star, but this size range would extend all the way down to individual grains of dust. Most people would agree that counting each such speck as a planet is ludicrous, and want a more satisfactory definition. Most proposed definitions, such as all objects above a certain size orbiting the star directly (say, 3000km) fall into the aforementioned arbitrary category. By this method, for example, Pluto would no longer qualify as a planet. Another method uses a more easily justified property of the candidate objects: whether or not it possesses sufficient mass to pull itself into a sphere.. Again, the object must be orbiting the star itself to qualify as a planet rather than a satellite. Two advantages of this method are 1) That by making use of the fundamental properties (mass and composition) of the object in question it seems less capricious than using size alone and 2) Pluto remains a planet, which has sentimental value.
Since there is no actual official definition of what a planet is, the range from ten times Jupiter's mass to whatever minimum mass allows for spherical objects will be used. This will probably not be an issue during this essay, in any case.
At this point it seems like a good idea to demonstrate the relevance of 'extrasolar planets' to the field of astrobiology. This is not difficult. While it may be possible for microbes to survive exposure to space, there is no known life-form which could claim space as its home. If it is possible for life to exist in space, that life would still need to have got its start in a more hospitable environment provided by a planet, and from there it is difficult to imagine what set of circumstances might prepare it for - and deliver it to - space. Even in the idea of 'panspermia', which suggests spores of life travelling through space from star to star, the spores are merely dormant, and do not actually 'live' between worlds. They still need a planet at each end. So in any case, if there is to be life outside the solar system it will require a planet on which to live, and if we want to find it we will need to detect these planets.
There are four main methods for detecting extrasolar planets. The most successful of all these is the radial velocity or Doppler method. As the name suggests, the Doppler method involves measuring the shifting of a stars [sic] spectral lines caused by the changing velocity as it revolves about a common centre of mass with its planet(s).
Detecting a planet by this method requires the ability to resolve changes in velocity one quarter that caused by the planet being sought. Jupiter causes a motion of twelve metres per second, so to find it would require an accuracy of around three metres per second. Currently, this is almost the best available, so planets identical to Jupiter are near the limits of current detectability. The first planet detected around a sun-like star by this method, 51 Pegasi b, produces a motion of fifty-five metres per second in its host star. The current limit of this technique is one to two metres per second and it is unlikely to improve much. There is a fundamental limi to the radial velocity method caused by natural oscillations in the star producing an effective motion of around one metre per second. Planets producing smaller effects than this, such as earth [sic] (at thirty centimetres per second) are lost in the noise.
Pulsar timing is a variation of this method in which the Doppler shift is measured in the time of arrival of a pulsar's pulses rather than the shifting of spectral lines. Because pulsars are naturall extremely precise it is much easier to detect planets by this method than by others. The first extrasolar planets detected were found this way in 1991/1992 orbiting PSR 1257 + 12. The three planets found in the early nineties are all far smaller than any other known extrasolar planets, with the smallest having roughly the same mass as earth's [sic] moon [sic]. The other two are 2-3 times earth's [sic] mass. It is generally believed that these worlds could not have survived the supernova that created the pulsar, so they were likely formed later.
The astrometric method involves making accurate measurements of a star's position to detect the presence of planets. Stellar positions are compared against the positions of field stars in extremely accurate photographs. It is possible in principle to detect even small planets with this method. However, when stars are photographed with current instruments, they show up as fuzzy blobs generally larger in size than the position changes being looked for. This means that greater resolution is required before astrometry can become a useful tool for finding planets.
Interferometry is expected to help solve this problem. The Keck interferometer will use four 1.8 metre telescopes seperated by one hundred metres to simulate a telescope with a diameter of 100 metres. It will be able to measure stellar positions to within 20 microarcseconds. In contrast, the position change in the sun induced by Jupiter, as seen from a distance of ten light years, is 0.6 milliarcseconds. The Keck interferometer will be able to detect Jupiter-like planets around even distant stars and may be able to detect Uranus-like (one twentieth Jupiter's mass) planets around nearby stars. Earthlike planets will still be undetectable and the turbulence of earth's [sic] atmosphere means that further improvements will probably require space-based interferometers.
NASA plans to launch the Space Interferometry Mission in 2009. This will involve two telescopes on the ends of a ten metre boom and will have a resolution of one microarcsecond. It is suggested that this telescope will be able to detect planets with four earth [sic] masses to a distance of 33 light years and may be able to detect earth-massed [sic] planets around some closer stars. The Space Interferometry Mission is largely intended as a 'proof of concept' for larger telescopes to follow.
The radial velocity and astrometry methods complement each other well. Astrometry has difficulty finding close, fast moving planets, while slower, more distant worlds may be forever beyond the reach of the radial velocity method. This means that together the two methods may be able to provide a near complete survey of nearby planets.
The third method for finding planets is the transit method. It involves watching a star for periodic dips in brightness which may indicate the passage of a planet across its disc. The odds of a planet being in such a favourable alignment depend on the distance between planet and star, with the shorter the period (and the closer the planet) the better. A planet in a three day orbit has a 10% chance of transiting as seen by a random observer. At two weeks the odds drop to 4%, while if the period is as long as the earth's [sic, again] (one year) the chance of a transit being seen by a random observer are only 0.5%. Only four extrasolar planets are known to transit so far, three of which were discovered from their transits and one has the shortest period known (29 hours).
NASA plans to launch a satellite called Kepler by 2005 (the only official date I could find, but it has surely been delayed, as SIM was to be its prototype) to use the transit method to discover new planets. Kepler is intended to observe single patch of sky containing one hundred thousand stars over a period of four years. After this time it is expected to observe two or three transits of each candidate for confirmation. The transit method does not detect planets of any particular size preferentially, meaning that Kepler should be able to determine the true frequency of earthlike (and other) planets.
Direct imaging is the fourth method of detecting planets. It probably will not be used to search for new systems, but it could identify new planets in already known systems, and provide detailed information about previously identified worlds. NASA is planning to construct a large space-based interferometer using several components, each larger than the Hubble telescope, spread across hundreds of metre from each other for this purpose, called the Terrestrial Planet Finder. It is expected to launch some time in the next ten to fifteen years. The ESA (European Space Agency) has a similar project called Darwin. Direct detection is expected to be difficult even with these telescopes, since the host star is about one hundred million times the brightness of the planets in visible light, and ten million times brighter in infrared, where the two are closest to each other in brightness.
So far, the more than one hundred extrasolar planets known fall into four general groups. Starting with the least unfamiliar, there are small, rocky worlds similar to earth [sic]. Other than those in the solar system, the only ones known are the inner planets of the pulsar PSR 1257 + 12.
Also recognisable are the 'familiar giants', with long periods and near-circular orbits. This category includes the four outer giants of the solar system but only a few outside it.
The 'eccentric giants' are more unusual. These objects have higly eccentric orbits with periods rangin from ten to one thousand days. Their average mass is several times that of Jupiter and this is the most common type of planet found so far. There are more than ninety known so far and they seem to be found aruond 7% of all sunlike stars.
The last group are the famous 'hot jupiters'. These planets have very short periods, ranging from a week down to three days or less. Their orbits are typically only one eighth the size of Mercury's and their temperature is typically around 1500K, which is the same temperature as red-hot coals. It is likely that tidal interactions with the host star produce synchronous rotation in hot Jupiters. Tidal effects are also believed to be responsible for their more circular orbits. Their mass is typically near that of Jupiter and they seem to be found around 1% of all sunlike stars.
For the purposes of astrobiology, the most interesting extrasolar planets are the ones that might support life. These would be in the aforementioned 'small, rocky' category. However, apart from the so-called 'pulsar planets', whose parent stars would make life impossible, these are currently undetectable. Attention has therefore focussed [sic] on whether the known planets could allow habitable worlds to share their system.
This means that either these giant worlds must orbit their parent star in such a way as to permit an earth-sized body to have a stable orbit within the star's 'habiitable zone' (the region where radiation from the star would warm a planet sufficiently for it to possess liquid water) or, if the planet itself resides within the habitable zone, very large moon might provide a suitable environment.
Unfortunately for the latter case, orbiting close enough to its star to sufficiently warm its moons also means it is close enough that gravitational interactions with the star would also tend to strip the planet of any moons over a relatively short period. For the other case usually only systems with either (or both) 'hot jupiters' or 'familiar giants' are candidates, since the orbits of the 'eccentric giants' are usually too disruptive. However, since the 'hot jupiters' are generally explained as the product of interactions during their formation - with either the nebular disk in which they formed or their fellow planets - this means they must have passed through the habitble zone on the way to their current position, and more than likely disrupted any potentially habitable worlds along the way. This means that very few of the systems discovered so far are at all hospitable to life.
One of the more popular methods to distinguish between 'planets' and 'not planets' is the way in which it formed. The essence of this idea is that anything formed in orbit about a star is a planet, while anything formed elsewhere is not. In particular, planets are formed as secondary products of star formation, while non-planets (actually, they'd be brown dwarfs, wouldn't they?) are formed by direct condensation from nebulae, like stars. The smallest sized brown swarfs that can be formed like this possess roughly ten times the mass of Jupiter, so anything smaller found in interstellar space would by this definition be a planet, and was probably ejected from a star. There have been, so far, no confirmed reports of such objects being found and they will therefore be ignored from now on. This mas (ten times Jupiter's) is also often used to define an upper limit for what is described as a planet when considering objects found orbiting a star, with anything larger being considered a brown dwarf and thereby avoiding all questions of origin. This distinction is useful because, although it is conceivable that planets with more than ten times the mass of Jupiter could be formed, is that it is either ery difficult or impossible to tell from an object's spectrum whether it is a small brown dwarf or a very large planet. This system therefore makes things much simpler (some would say too simple) than they would otherwise be.
While there has been much discussion of upper limits for planetary masses, the smaller end of the scale seems to have been largely ignored. This may be at least partly because there is little hope of detecting such objects in the near future and there is therefore little sense of urgency to the issue Problems also arise from the fact that there is difficulty finding a point to draw the line that does not seem arbitrary. It seems natural to define a planet as any non-fusing object (a fusing object being a star, by definition) formed in orbit about a star, but this size range would extend all the way down to individual grains of dust. Most people would agree that counting each such speck as a planet is ludicrous, and want a more satisfactory definition. Most proposed definitions, such as all objects above a certain size orbiting the star directly (say, 3000km) fall into the aforementioned arbitrary category. By this method, for example, Pluto would no longer qualify as a planet. Another method uses a more easily justified property of the candidate objects: whether or not it possesses sufficient mass to pull itself into a sphere.. Again, the object must be orbiting the star itself to qualify as a planet rather than a satellite. Two advantages of this method are 1) That by making use of the fundamental properties (mass and composition) of the object in question it seems less capricious than using size alone and 2) Pluto remains a planet, which has sentimental value.
Since there is no actual official definition of what a planet is, the range from ten times Jupiter's mass to whatever minimum mass allows for spherical objects will be used. This will probably not be an issue during this essay, in any case.
At this point it seems like a good idea to demonstrate the relevance of 'extrasolar planets' to the field of astrobiology. This is not difficult. While it may be possible for microbes to survive exposure to space, there is no known life-form which could claim space as its home. If it is possible for life to exist in space, that life would still need to have got its start in a more hospitable environment provided by a planet, and from there it is difficult to imagine what set of circumstances might prepare it for - and deliver it to - space. Even in the idea of 'panspermia', which suggests spores of life travelling through space from star to star, the spores are merely dormant, and do not actually 'live' between worlds. They still need a planet at each end. So in any case, if there is to be life outside the solar system it will require a planet on which to live, and if we want to find it we will need to detect these planets.
There are four main methods for detecting extrasolar planets. The most successful of all these is the radial velocity or Doppler method. As the name suggests, the Doppler method involves measuring the shifting of a stars [sic] spectral lines caused by the changing velocity as it revolves about a common centre of mass with its planet(s).
Detecting a planet by this method requires the ability to resolve changes in velocity one quarter that caused by the planet being sought. Jupiter causes a motion of twelve metres per second, so to find it would require an accuracy of around three metres per second. Currently, this is almost the best available, so planets identical to Jupiter are near the limits of current detectability. The first planet detected around a sun-like star by this method, 51 Pegasi b, produces a motion of fifty-five metres per second in its host star. The current limit of this technique is one to two metres per second and it is unlikely to improve much. There is a fundamental limi to the radial velocity method caused by natural oscillations in the star producing an effective motion of around one metre per second. Planets producing smaller effects than this, such as earth [sic] (at thirty centimetres per second) are lost in the noise.
Pulsar timing is a variation of this method in which the Doppler shift is measured in the time of arrival of a pulsar's pulses rather than the shifting of spectral lines. Because pulsars are naturall extremely precise it is much easier to detect planets by this method than by others. The first extrasolar planets detected were found this way in 1991/1992 orbiting PSR 1257 + 12. The three planets found in the early nineties are all far smaller than any other known extrasolar planets, with the smallest having roughly the same mass as earth's [sic] moon [sic]. The other two are 2-3 times earth's [sic] mass. It is generally believed that these worlds could not have survived the supernova that created the pulsar, so they were likely formed later.
The astrometric method involves making accurate measurements of a star's position to detect the presence of planets. Stellar positions are compared against the positions of field stars in extremely accurate photographs. It is possible in principle to detect even small planets with this method. However, when stars are photographed with current instruments, they show up as fuzzy blobs generally larger in size than the position changes being looked for. This means that greater resolution is required before astrometry can become a useful tool for finding planets.
Interferometry is expected to help solve this problem. The Keck interferometer will use four 1.8 metre telescopes seperated by one hundred metres to simulate a telescope with a diameter of 100 metres. It will be able to measure stellar positions to within 20 microarcseconds. In contrast, the position change in the sun induced by Jupiter, as seen from a distance of ten light years, is 0.6 milliarcseconds. The Keck interferometer will be able to detect Jupiter-like planets around even distant stars and may be able to detect Uranus-like (one twentieth Jupiter's mass) planets around nearby stars. Earthlike planets will still be undetectable and the turbulence of earth's [sic] atmosphere means that further improvements will probably require space-based interferometers.
NASA plans to launch the Space Interferometry Mission in 2009. This will involve two telescopes on the ends of a ten metre boom and will have a resolution of one microarcsecond. It is suggested that this telescope will be able to detect planets with four earth [sic] masses to a distance of 33 light years and may be able to detect earth-massed [sic] planets around some closer stars. The Space Interferometry Mission is largely intended as a 'proof of concept' for larger telescopes to follow.
The radial velocity and astrometry methods complement each other well. Astrometry has difficulty finding close, fast moving planets, while slower, more distant worlds may be forever beyond the reach of the radial velocity method. This means that together the two methods may be able to provide a near complete survey of nearby planets.
The third method for finding planets is the transit method. It involves watching a star for periodic dips in brightness which may indicate the passage of a planet across its disc. The odds of a planet being in such a favourable alignment depend on the distance between planet and star, with the shorter the period (and the closer the planet) the better. A planet in a three day orbit has a 10% chance of transiting as seen by a random observer. At two weeks the odds drop to 4%, while if the period is as long as the earth's [sic, again] (one year) the chance of a transit being seen by a random observer are only 0.5%. Only four extrasolar planets are known to transit so far, three of which were discovered from their transits and one has the shortest period known (29 hours).
NASA plans to launch a satellite called Kepler by 2005 (the only official date I could find, but it has surely been delayed, as SIM was to be its prototype) to use the transit method to discover new planets. Kepler is intended to observe single patch of sky containing one hundred thousand stars over a period of four years. After this time it is expected to observe two or three transits of each candidate for confirmation. The transit method does not detect planets of any particular size preferentially, meaning that Kepler should be able to determine the true frequency of earthlike (and other) planets.
Direct imaging is the fourth method of detecting planets. It probably will not be used to search for new systems, but it could identify new planets in already known systems, and provide detailed information about previously identified worlds. NASA is planning to construct a large space-based interferometer using several components, each larger than the Hubble telescope, spread across hundreds of metre from each other for this purpose, called the Terrestrial Planet Finder. It is expected to launch some time in the next ten to fifteen years. The ESA (European Space Agency) has a similar project called Darwin. Direct detection is expected to be difficult even with these telescopes, since the host star is about one hundred million times the brightness of the planets in visible light, and ten million times brighter in infrared, where the two are closest to each other in brightness.
So far, the more than one hundred extrasolar planets known fall into four general groups. Starting with the least unfamiliar, there are small, rocky worlds similar to earth [sic]. Other than those in the solar system, the only ones known are the inner planets of the pulsar PSR 1257 + 12.
Also recognisable are the 'familiar giants', with long periods and near-circular orbits. This category includes the four outer giants of the solar system but only a few outside it.
The 'eccentric giants' are more unusual. These objects have higly eccentric orbits with periods rangin from ten to one thousand days. Their average mass is several times that of Jupiter and this is the most common type of planet found so far. There are more than ninety known so far and they seem to be found aruond 7% of all sunlike stars.
The last group are the famous 'hot jupiters'. These planets have very short periods, ranging from a week down to three days or less. Their orbits are typically only one eighth the size of Mercury's and their temperature is typically around 1500K, which is the same temperature as red-hot coals. It is likely that tidal interactions with the host star produce synchronous rotation in hot Jupiters. Tidal effects are also believed to be responsible for their more circular orbits. Their mass is typically near that of Jupiter and they seem to be found around 1% of all sunlike stars.
For the purposes of astrobiology, the most interesting extrasolar planets are the ones that might support life. These would be in the aforementioned 'small, rocky' category. However, apart from the so-called 'pulsar planets', whose parent stars would make life impossible, these are currently undetectable. Attention has therefore focussed [sic] on whether the known planets could allow habitable worlds to share their system.
This means that either these giant worlds must orbit their parent star in such a way as to permit an earth-sized body to have a stable orbit within the star's 'habiitable zone' (the region where radiation from the star would warm a planet sufficiently for it to possess liquid water) or, if the planet itself resides within the habitable zone, very large moon might provide a suitable environment.
Unfortunately for the latter case, orbiting close enough to its star to sufficiently warm its moons also means it is close enough that gravitational interactions with the star would also tend to strip the planet of any moons over a relatively short period. For the other case usually only systems with either (or both) 'hot jupiters' or 'familiar giants' are candidates, since the orbits of the 'eccentric giants' are usually too disruptive. However, since the 'hot jupiters' are generally explained as the product of interactions during their formation - with either the nebular disk in which they formed or their fellow planets - this means they must have passed through the habitble zone on the way to their current position, and more than likely disrupted any potentially habitable worlds along the way. This means that very few of the systems discovered so far are at all hospitable to life.
