What is the most common planetary system architecture in the Universe?

Clearly, since the radial velocity technique is biased in favor of finding the most massive planets in the shortest orbits around solar type stars (stars for which the luminosity is large enough to get a good spectrum), it would be misleading to contend that the most common planetary system architecture in the Universe is a G-type system containing a few Jovian mass planets having orbital periods less than that of our Jupiter.

One important trend in the current extrasolar planet data is that although there is a great diversity of system layout, there appear to be more small planets than large planets-- even though large planets are the easiest to detect. This indicates that a substantial fraction of stars have Neptune and below mass planets.

Here is my hypothesis: Since M dwarf stars are the most common type of main sequence star, then most planetary systems have an M dwarf as their parent star. Second, since smaller planets form more easily and in greater abundance than larger planets, the most common planetary system architecture is an M dwarf with several terrestrial planets and/or dwarf planets (no Jovian planets).

The Kepler mission, a NASA observatory designed to detect planets as small as Mercury around M dwarfs, will test my hypothesis starting in 2008. Also, don't count out microlensing. I predict that before the Kepler data is released in 2013-15, microlensing searches will have already told us what fraction of M dwarfs have planets down to 5-15 earth masses.

We need more data to say anything for certain.

Currently the most typical extrasolar system seems to contain a Jupiter-mass planet in an eccentric orbit somewhere in a region corresponding the inner Solar system. Hot Jupiters form a clear minority, even though they're much easier to detect. So they're clearly relatively rare objects. "Super-hot Jupiters" that orbit a star in only a couple of days seem to be even more unusual. Most likely the eccentric planets are also biased. There are already some more distant planets known, but they too have very eccentric orbits.

Neptune-mass planets may be relatively common around red dwarfs. It is now clear that Jupiter-type gas giants are rare around red dwarfs.

Most small stars are considerably older than our Sun, which means they have also much less metals (in astronomy, elements heavier than helium). A star's metallicity seems to be clearly connected to how likely the star has planets, so it could be that most old stars have no planets. Recent search of protoplanetary disks in the Orion Nebula suggests that about 60% to 70% of stars there have disks. So not even all young stars necessarily have planets.

__________________
Science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool.
-- Richard Feynman

Is it not unreasonable to suggest that most bodies seem to form in a binary relationship. I guess it is because the first body forms, then concentrates a debris ring that rapidly accretes into a partner body. If sufficient material remains after the second body and is not disturbed by the orbits of the previous pair, then a third can form and so on. I believe that the most common planetary structure will be a dominant sun partnered by smaller star/dwarf, surrounded by an unconsolidated debris cloud.

Well for detection of earth sized planets with this technique they need to detect a radial speed of about 1 m/s vs 10 m/s or higher they have at the moment. The lick obsrvatory a beleive is building a rocky planet decting scope that will some next year.

also I wonder if kelper will be delayed or not because of the reshuffling in nasa.

Long ago I read that most of the proto planets were ejected by sling shot manuver from our solar system. If so, and typical of the last 13.7 billion years, There should be about 10E25 of these free fling planets within about ten billion light years of Earth. 95% are likely less massive than Mercury and/or mostly frozen volitiles, thus would be classed dwarf planets or other objects, if they were in our solar system. Perhaps half of them are temporarilly sharing their orbit between two or more otherwise planetless and companionless stars, compact stars and/or brown dwarfs. Neil

I like the idea but shouldn't our star have captured a foriegn planetoid by now?

Agreed.

I think you make a leap here by stating "no Jovian planets". All we can say reasonably, given your assumptions, is "more terrestrial planets than Jovian planets".

That would be awesome.

__________________
"Barbarism is the natural state of mankind. Civilization is unnatural. It is a whim of circumstance. And barbarism must always ultimately triumph" -- Conan

Who says it hasn't?

If the Sun captured a planetoid, you couldn't reasonably expect it to be on the ecliptic nor in a circular orbit. It also wouldn't be orbiting inside of Neptune because it would have been ejected long ago by the gas giants.

Any one of the KBOs could be a captured, foreign planetoid. There's no way of ever knowing for sure, I think.

__________________
"Barbarism is the natural state of mankind. Civilization is unnatural. It is a whim of circumstance. And barbarism must always ultimately triumph" -- Conan

Some of the planet hunting teams have cautioned NOT to put too much weight into the early discoveries of a large numbers of epistellar jovians. They're finding a lot of them now because they are the easiest to find. It doesn't necessarily follow that they'll be the most common. Long duration planets will obviously take longer to weed out, because the needed observations to determine their orbital parameters from the radial effect on the star require multiple orbits for a confirmation. For a planet with a five year orbit, that's a good 10-15 years for a lock that you've got a planet. Multiple planet systems will require even more time because you've got to tease out the data by filtering the most obvious planets out, and working the rest from the remainder.

Right.

Some of the eccentric planets may actually orbit in relatively circular orbits as additional, yet to be detected planets cause scattering in radial velocities. For example, all the four planets in the µ Arae system orbit in relatively circular orbits. Originally the first two planets were thought to be rather eccentric.

__________________
Science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool.
-- Richard Feynman

I did not know that... That sounds very promising for us SETI fanatics!

__________________
"Barbarism is the natural state of mankind. Civilization is unnatural. It is a whim of circumstance. And barbarism must always ultimately triumph" -- Conan

Same here.
This is a very interesting thread, nice one folkhemmet!

The hysterical thing about the two most common methods of planet finding is that they operate almost like Heisenberg's Uncertainty Principle.

OGLE can give you detailed information about the planet's physical structure, but almost nothing about its orbit, while Radial Velocity can tell you a great deal about its orbit, but almost nothing about its physical makeup (most of the mass stats given by radial velocity are the minimum possible, since there's no direct observation of the body itself, there's a range of possibilities based on mass and distance.)

Here is an interesting new paper which seems to support my hypothesis about the most common planetary system architecture.

astro-ph/0609140 [abs, ps, pdf, other] :
Title: Planet formation around low mass stars: the moving snow line and super-Earths
Authors: Grant M. Kennedy, Scott J. Kenyon, Benjamin C. Bromley
Comments: accepted by ApJ Letters

We develop a semi-analytic model for planet formation during the pre-main sequence contraction phase of a low mass star. During this evolution, the stellar magnetosphere maintains a fixed ratio between the inner disk radius and the stellar radius. As the star contracts at constant effective temperature, the `snow line', which separates regions of rocky planet formation from regions of icy planet formation, moves inward. This process enables rapid formation of icy protoplanets that collide and merge into super-Earths before the star reaches the main sequence. The masses and orbits of these super-Earths are consistent with super-Earths detected in recent microlensing experiments.