PDA

View Full Version : Discussion: Smallest Extrasolar Planet Found



Fraser
2004-Aug-25, 04:27 PM
SUMMARY: A team of European astronomers have used the European Southern Observatory's HARPS Instrument to find the smallest extrasolar planet ever discovered; it's believed to be only 14 times the mass of the Earth. The planet orbits a star called mu Arae every 9.5 days, which is located 50 light-years away in the southern constellation of the Altar. A planet this size lies right at the boundary between rocky planets and gas giants. But since it's so close to its parent star, it's probably rocky, with a relatively small atmosphere, so it would be classified as a "super Earth".

What do you think about this story? Post your comments below.

John L
2004-Aug-25, 06:46 PM
I wonder if this is a "Hot Jupiter" that has had most of its atmosphere blown off. Other planets have been seen that appear to be losing their atmosphere due to solar wind action. It is possible this planet is a gas giant being slowly blown away in the wind.

antoniseb
2004-Aug-25, 07:00 PM
Originally posted by John L@Aug 25 2004, 06:46 PM
I wonder if this is a "Hot Jupiter" that has had most of its atmosphere blown off.
That makes sense to me. I suspect that if it were still losing a great deal of atmosphere, we'll be able to detect it [as we have with some other hot Jupiters]. It may be that what is left is the part of the protoplanetary disk that accreted into this Hot Jupiter as it migrated toward the star. This close to the star, it would probably be a liquid soup of refractory compounds and glasses. Let's all agree to not try and colonize the place.

om@umr.edu
2004-Aug-25, 07:08 PM
Perhaps this is not "the smallest extrasolar planet ever discovered."

As I recall, smaller planets were found in the 1990s orbiting PSR 1257+12.

With kind regards,

Oliver
http://www.umr.edu/~om

StarLab
2004-Aug-25, 07:40 PM
Well, maybe we won't colonize it - but, lemme throw out this one hypothesis: maybe our earth was once also a "Hot Jupiter?" (like during its planetary formation.)

John L
2004-Aug-25, 07:43 PM
The story addressed these. The Exo-Planet community seems to have unofficially decided that planets orbiting pulsars don't really count since we'll NEVER be able to colonize any of them. Sure, most of the planets around normal stars that have been found have no chance of ever supporting life, but there may be other small planets around the same stars to could support life. You can't get that around a pulsar.

But the planets around that pulsar were the "first" to ever be confirmed to be orbiting another star - a major achievement that got the field really going - even if you wouldn't want to live there.

Planetwatcher
2004-Aug-25, 09:04 PM
Both Oliver, and John L. have good points.
However, I don't know if the pulsar's planets should be excluded simply because we'll NEVER be able to colonize any of them. The same is true for nearly every exo-planet found so far.
I would think it better to catagorize them as exo-planets not orbiting main sequence stars.

As for the new planet at Mu Arae, it is likely only a couple million miles away from the star. I did some checking and found that the star itself is a little bigger then our Sun, but also very slightly cooler.
The life zone would be not quite as far out as Mars is in our Solar System.
But there is another planet in this system, it is Jupiter sized and it is located some 20 or 30 million miles farther out from the life zone. About the same distance as between the orbits of Earth and Venus.
A gas giant that close would seriously disrupt the orbit of a planet in the life zone.

Here is a link to Mu Arae. Mu Arae (http://www.solstation.com/stars2/mu-arae.htm)

John L
2004-Aug-25, 09:40 PM
I read the preprint for the team that made the discovery and found this:


First, given its location and the characteristics of the central star, it is unlikely that this object was in fact a much more massive giant planet which has lost a large fraction of its envelope over its lifetime.

Their reasoning was that there have been found other large mass planets closer in that haven't lost theirs. They argued that it would take a planet smaller than Jupiter initially to lack the gravity to hold its atmosphere against the close solar wind. They had some other lines of reasoning that made sense, too, so it seems this may really be a big terrestrial world.


Planetwatcher,

That's what I meant, basically. A pulsar is definitely not main sequence, and we would NEVER send anything there for close inspection (far future star trek-type travel). The radiation environment around a pulsar would make visting impossible. If you've ever visited the Extrasolar Planets Encyclopedia website, you'll note that they have a separate category for pulsar planets that is kept separate. Most other sites I've seen do the same.

Duane
2004-Aug-26, 02:37 PM
I also disagree with excluding the planets found orbiting pulsars. While I can understand the reasoning of placing them in a separate catagory, I find it odd that they wouldn't include them when discussing the smallest planet yet discovered.

This is a rather interesting system though. The closest planet is in a nearly circular orbit, whereas the other 2 planets are highly eccentric. Seems to make a stronger case for planatary migration in certain circumstances.

Given this star is a tad cooler then our own sun, and given further that an earth-sized planet would be in a stable orbit no closer than 1.3 AU from the star, it appears there is little chance of a habitable planet--although one never knows :)

VanderL
2004-Aug-26, 03:10 PM
Hi there,

Very interesting all of this, but we're forgetting that we still don't know exactly how planetary systems are formed anyway. Our solar system is a very oddball ensemble (compared to what we know about exoplanets) and it would be wise to start our modelling from the "hot Jupiters" outward. I wouldn't be surprised if the situation of the close-orbiting giant is the standard and the circumstances on such a world coud be different from what we expect.

Cheers.

Guest
2004-Aug-26, 04:14 PM
this is wonderful news

John L
2004-Aug-26, 04:42 PM
Originally posted by VanderL@Aug 26 2004, 10:10 AM
Very interesting all of this, but we're forgetting that we still don't know exactly how planetary systems are formed anyway. Our solar system is a very oddball ensemble (compared to what we know about exoplanets) and it would be wise to start our modelling from the "hot Jupiters" outward. I wouldn't be surprised if the situation of the close-orbiting giant is the standard and the circumstances on such a world coud be different from what we expect.
I think its premature to say that our solar system is an oddball. The radial velocity method for detecting planets around other stars - the most common and effective method to date - is strongly biased to short period, high mass planets. To find a Jupiter you would have to measure that star's radial velocity for longer than the planet's year (11 years for Jupiter), and preferably for two or more of that planet's years to get a good solid confirmation. We simply haven't been watching enough stars with sufficient resolution for a sufficient length of time to have found lots of other copies of our solar system. In 20 more years, we may have found thousands of them, but until we do the observation for the length of time necessary, or find a new technique that's faster, then you can't say our solar system is not the norm.

Duane
2004-Aug-26, 07:49 PM
I agree with John L. There are several lines of research going on right now to try and explain the formation of planets, including observations of planets that seem to be accreting right now.

I have seen research that says Earth-like planets should be common, and others that say they should be rare. There is research suggesting that planets can form quite quickly, and other research that says they can't form very fast. I have even seen a recent report that there may be two different mechanisms at work, one that favours the formation of large gaseous planets, the other that favours formation of solar systems similar to our own.

So, I also agree with VanderL that we don't have the whole story behind planetary formation.

om@umr.edu
2004-Aug-28, 12:37 PM
Please keep in mind that rocky, Earth-like planets are composed mostly of even-numbered elements with high nuclear stability. Elements like Fe, Si, O, and S.

Jupiter, on the other hand, consists mostly of elements of low nuclear stability. Elements like H, He, C, N.

It is generally agreed that these two batches of elements came from different regions of highly evolved stars.

Recent observations suggest that these different stellar regions do not necessarily mix in asymmetric stellar explosions.

There would be no need to "blow away" light elements in order to form rocky, Earth-like planets if light elements from the stellar surface never mixed with Fe, Si, O, and S from the stellar interior.

Wouldn't this be a reasonable way to form the rocky, Earth-like planets observed around a pulsar?

Wasn't there also a recent news report about clouds of Si and others of almost pure Fe still present a few hundred years after a supernova explosion?

With kind regards,

Oliver
http://www.umr.edu/~om

VanderL
2004-Aug-28, 01:26 PM
Hi again,

The notion that the solar system is an oddball collection is discussed in this paper (http://xxx.lanl.gov/abs/astro-ph/0407476), I think the authors show convincingly that the bias introduced by the detection method is not a strong argument against the notion that we inhabit an abnormal planetary system.
Clearly the models for planet formation need a refit. Maybe the expulsion theory from the EU folks could help? ;)

Cheers.

om@umr.edu
2004-Aug-28, 01:31 PM
I agree, VanderL.

We do not yet have the detection method to establish if the planetary system here is "ordinary" or "strange."

With kind regards,

Oliver
http://www.umr.edu/~om

VanderL
2004-Aug-28, 02:21 PM
I agree, VanderL.

We do not yet have the detection method to establish if the planetary system here is "ordinary" or "strange."

Hi Oliver,

What I actually meant to say is that the solar system is the wrong place to start modelling when we compare this sytem with the explanet database. So even if we don't have have the means to determine whether Earth-like worlds exist, our system with Jupiter at 5 AU (I'm not sure that's correct) is definitely a rare system. Only a few of the 120 systems show similar planets at the same distance and according to the authors of the above paper our solar system is statistically different from the rest.
My point is that if the current studies show that we are in a planetary sytem that is statisticically different from the rest (taking detection bias into account) we must start modelling from the norm. This norm is maybe not completely established, but it wouldn't hurt to try to explain close-orbiting giant planets first assuming that's the starting point for all systems. Hence my remark that maybe the "expulsion" idea is useful.

Cheers.

om@umr.edu
2004-Aug-29, 02:23 PM
Right, VanderL,

Rocky, Earth-like planets are made of even-numbered elements with high nuclear stability. Elements like Fe, Si, O, and S.

The gaseous giant planet, Jupiter, are made of elements of low nuclear stability. Elements like H, He, C, N.

These two batches of elements came from different regions of a highly evolved star.

Isotope measurements show that they never mixed:

http://www.umr.edu/~om/abstracts2001/windl...leranalysis.pdf (http://www.umr.edu/~om/abstracts2001/windleranalysis.pdf)

Those are some of the experimental constraints that must be addressed by those who seriously attempt to model the formation of planetary systems.

With kind regards,

Oliver
http://www.umr.edu/~om

antoniseb
2004-Aug-29, 02:29 PM
Originally posted by VanderL@Aug 28 2004, 02:21 PM
our system with Jupiter at 5 AU (I'm not sure that's correct) is definitely a rare system. Only a few of the 120 systems show similar planets at the same distance
Actually, there is a strong observational bias that only allows us to detect systems with hot Jupiters. The much larger number of systems in which we have so far observed no planets could all be exactly like ours and we wouldn't be able to detect it yet.

You can't yet assume that our system is rare.

VanderL
2004-Aug-29, 03:05 PM
Hi Antoniseb,

I linked to a paper by Beer et al. (http://xxx.lanl.gov/pdf/astro-ph/0407476) that specifially states that, based on rigorous statistical analysis, our solar system must be considered an outlier.
They discuss in the paper how this conclusion was reached (taking the "detection bias" into account) and what it means for the modelling.

My suggestion is that we need to address the question of close-orbiting giant planets first, and start our models with the assumption that those planets are the standard and not our own solar system.
The standard accretion model assumes giant planets are formed some distance from the star and migrate inwards (which imo is very unlikely because that would take massive amounts of energy) to account for the close orbits.
There is one alternative scenario that I'm aware of (possibly other models exist as well) that specifically proposes that planets are formed by ejection from the parent star (by fissioning) and of course since that is an EU proposition it is highly speculative. If other posibilities exist I would like to hear them.


Oliver, you said



Rocky, Earth-like planets are made of even-numbered elements with high nuclear stability. Elements like Fe, Si, O, and S.

The gaseous giant planet, Jupiter, are made of elements of low nuclear stability. Elements like H, He, C, N.

These two batches of elements came from different regions of a highly evolved star.

Isotope measurements show that they never mixed

I'd like to speculate a bit about the EU model for planet formation and if observations match this ejection idea, which is the exact opposite of what is currently used as model.

In the EU model for planet formation the big planets (they can be considered small stars actually) are ejected from the parent star, while the rocky planets are either a byproduct of this event or are ejected from a gas giant and can be considered moons. Can such a scenario be reconciled with the observation of the unmixed state of the elements?

Cheers.

ASEI
2004-Aug-29, 04:01 PM
Actually, there is a strong observational bias that only allows us to detect systems with hot Jupiters. The much larger number of systems in which we have so far observed no planets could all be exactly like ours and we wouldn't be able to detect it yet.

You can't yet assume that our system is rare.

Exactly. We can only see very massive planets, and only if they are orbiting close into the parent stars right now. So we go looking for extra solar planets and find a few gas giants or giant terrestrials orbiting close to their parent stars. Then we conclude that since we haven't found any tiny earthlike worlds, or gas giants orbiting further out (even though we couldn't detect them if we tried), that our system must be odd, even though it is a dirt-common type of star! QED, right? :rolleyes: :rolleyes: Helooooo, is there any intelligent life out there? :rolleyes: :lol:

VanderL
2004-Aug-29, 05:33 PM
Ok,

I'll try to explain what the authors of paper did when comparing our solar system with the 100-odd exoplanets discovered thusfar. They plotted the distance to the star with the size of the planet and also looked at eccentricities in the orbits.

The bias everyone seems to point to is not an argument, because Jupiter-like planets at 5 AU distance from the star are equally detectable as Jupiter-like planets in closer orbits (mass is more important than distance). It doesn't use Earth as a parameter, only Jupiter-sized planets. If we only look at those equally detectable planets, our solar system is a clear outlier.

This a quote from the paper by Beer et al.


We have noted that the solar system is an outlier in the pa-
rameter space of known planetary systems. This is already
evident from Fig. 1, and we have used a non-parametric sta-
tistical test to show that Jupiter lies at about 2.3 sigma in the tail
of the periastron distribution. Given this, it seems sensible
to enquire whether the solar system is a typical planetary
system in the sense that it actually belongs to the set of exo-
planetary systems observed so far, or whether it might be
intrinsically different, or special, in some way. In particular,
we address the question of the extent to which it is necessary
to assume that all planetary systems formed in the same way
as the solar system. The major issue here is whether theo-
retical models for the formation of exo-planetary systems
should just consider minor modifications of the current the-
ory for the formation of the solar system, or whether some
quite different ideas might need to be investigated.


It can be argued that we need years of collecting more data to be sure, but 2.3 sigma seems enough of a difference to assume the solar system really is different.
So, what is the "standard" model for planet formation?

Cheers.

antoniseb
2004-Aug-29, 05:50 PM
Originally posted by VanderL@Aug 29 2004, 05:33 PM
The bias everyone seems to point to is not an argument, because Jupiter-like planets at 5 AU distance from the star are equally detectable as Jupiter-like planets in closer orbits (mass is more important than distance).
Just because they found one or two doesn't mean that it is just as easy. A Jupiter-like planet at 5 AU around a G2 dwarf will take six to twelve years to detect, depending on where the planet is during the first observation. A planet at 0.05 AU will take a few weeks.

In ten or fifteen years, we will start to get observations of terrestrial planets around nearby stars. Once we start doing that, we can discuss statistics about what is common and what isn't.


So, what is the "standard" model for planet formation?
I don't think there is one single model yet. It will probably be the case that there are several related models with some branch points.

om@umr.edu
2004-Aug-29, 06:12 PM
The standard model (SM) assumes that all our planets formed out of one interstellar cloud, consisting mostly of Hydrogen and Helium. The central region collapsed to form the Sun.

John Wood of Harvard is one of the leading spokesmen for this idea. George Wetherill of the Carnegie Institute tries to model the accretion of planets.

SM claims that all planets were initially Jupiter-like, made mostly of light weight volatile elements.

SM claims that the inner planets lost these volatile elements, leaving the rocky, iron-rich planetary cores that we now call Mercury, Venus, Earth and Mars.

According to the standard solar model (SSM), material did not accrete to form the Sun.

"Strange" isotope abundances of xenon in Jupiter, unlike the xenon in the Sun and the inner planets, is unexplained by SM or SSM.

With kind regards,

Oliver
http://www.umr.edu/~om

VanderL
2004-Aug-29, 07:48 PM
Just because they found one or two doesn't mean that it is just as easy. A Jupiter-like planet at 5 AU around a G2 dwarf will take six to twelve years to detect, depending on where the planet is during the first observation. A planet at 0.05 AU will take a few weeks.

If that really was a major problem for the statistical analysis that was performed in the article, then why was it written in the first place? If it really is impossible to be sure of the numbers, the whole article is crap. What makes the authors confident to a "2.3 sigma" level?
I think the argument that we need more data is always valid, but it shouldn't be used to hide from the real issue here. What is needed is a model that can account for the close-orbiters, assuming it is the norm.

On top of that, the issues raised by Oliver need to be addressed as well, all the data should lead us to a model that explains the characteristics that can be found, here as well as in other systems. Of course there will be many gaps, because planetary systems can become chaotic (or that's what I think) leading to deviations from the norm that will make it impossible to find a good model, as maybe happened when we started with our system as standard. Of course if there is nothing to compare with, our system was the norm. Now we know more, we should move on.

It couldn't hurt to come up with as many models as possible, I'm curious as to which models explain the "hot Jupiters".

Cheers.