New information about Hot Jupiters

[Basil]

Hot Jupiters may be challenging the current theory of planetary formation.

 

http://www.usnews.com/science/articles/2010/04/15/astronomers-find-nine-new-planets-and-upset-the-theory-of-planetary-formation.html

 

The Bacandforthtrian says looking at things from a new angle is always a good idea.

 

 

The Bacandforthtrian is another fine product of Basil's Twisted Imagination, Igg. All rides in reverse.

[Old Man]

Re: New information about Hot Jupiters

 

They must be captures. It seems to me that there are probably more stray sub-stellar gas giants floating around in space than people really think about. But logically, not every accreting body will reach the critical mass necessary for stellar ignition.

[Cancer]

Re: New information about Hot Jupiters

 

ESO press release that spurred the US News item

 

I'll look for a preprint about this tomorrow.

[Cancer]

Re: New information about Hot Jupiters

 

OK, I got the preprint (which is here, the paper by about 20 authors starting with Amaury).

 

I don't believe one of their fits, but two of the others look solid.

 

The measurement is spectroscopic, looking at apparent Doppler shifts of the central star, due to the gravitational pull by the planet. The planet is invisible. Only transiting planets are being studied ... "transiting" here means an eclipse where the planet blunders in front of the star from our point of view. This makes the total light received from the star drop, while the planet is moving across the face of the star. Nothing new to this. Only for a small number of systems are we lucky enough to be nearly on the planet's orbital plane and see these eclipses, but there's enough of them that there's some we can see. Measuring the eclipses with a photometer (which tells you when the eclipses begin & end and how big the planet is, because the planet is dark and the amount of light blocked is just how much of the stellar surface the planet covers) are additional data that you include in the fit.

 

Presumably the star is rotating. The models predict the situation like what we have in our Solar System: the orbits of the planets are all pretty close to the equatorial plane of the star, and the sense of the orbital motion is the same as the sense of rotation for the star.

 

With a transiting system, since we have to be close to the planet's orbital plane to see any eclipse, presumably the star's rotation axis is in the plane of the sky (that is, we are "looking down" directly on the star's equator with no tilt of its rotation axis). So draw a circle on paper to represent the star, and the left half of the star is moving toward us (rotating out of the page), the right half is moving toward us (rotating into the page), and the centerline is moving purely sideways from left to right. There's a smooth gradation in the velocity that any square kilometer of the star's surface has, as you look at this rotating spherical star. The Doppler shift only cares about velocities along the line of sight. With the two halves of the star being equal in size and moving equally if in opposite directions as the star rotates, a single wavelength emitted equally from every square kilometer of stellar surface gets Doppler shifted to shorter wavelengths on the left side, shifted to longer wavelengths on the right side, and there's zero shift on the midline. The result is that a spectral line is broadened, but there's no net change of wavelength, when you see the whole stellar disk of the rotating star. This is called "rotational line broadening".

 

But with an eclipse by a large planet, part of the planet's surface gets blocked. And that part is a circular piece that moves across the stellar disk. If the planet orbits in the same direction as the star is rotating, that blockage moves from left to right. If it's retrograde (that is, opposite the star's rotation), then the blockage moves from right to left.

 

That moving blockage wipes out the symmetry of the rotational line broadening. While the planet is blundering in front of the star, it blocks one side of the star more than the other at all times except that instant the planet's center is on the midline of the star as we see it. That means that part of the rotating-toward-us half isn't balanced by an equal amount of the rotating-away-from-us half. The result is a net shift in the spectral lines of the star as the planet is blundering across the stellar surface. This shift is the "Rossiter-McLaughlin effect". You get a weird blip in the star's Doppler-shifted velocity as the planet crosses in front of the star, and the shape of the blip gives you a handle on the angle between the star's rotation axis and the planet's orbital plane.

 

That's what these guys have tried to measure. In the explanation above I have left out lots of complications that are distractions from the key point (and these guys seem to have handled the complications correctly).

 

They claim three cases (I only believe two) where the fit is good enough to indicate that the planet is going around in the opposite direction of stellar rotation. All of them have circular orbits (more precisely, the fits don't distinguish between the orbital eccentricity and zero).

 

There's a couple of real subtle effects I've thought of as I typed this that I need to check on, but for now I think their claim has to be taken seriously. The big weakness is the amount of observational data they have is small. To get the spectroscopic measurements during the eclipse with enough time resolution to be able see what's happening takes a rather special instrument, and there's only a couple of them that can do it. It seems clear these guys are reporting "first results" from one of them, and that means that AFAICT they only have one eclipse observed for each star. Getting many of them would be Lots Better. I don't think their result will go away with better observations (based on what they show in this preprint), but it's worth a wait-and-think-and-see attitude right now.

[Drhoz]

Re: New information about Hot Jupiters

 

They must be captures. It seems to me that there are probably more stray sub-stellar gas giants floating around in space than people really think about. But logically' date=' not every accreting body will reach the critical mass necessary for stellar ignition.[/quote']

 

except captures are really difficult to do - if it's traveling thru interstellar space it'll have more than enough momentum to head right back off into interstellar space once it's fallen to the depths of the gravity well. Unless something already in the system steals some of the momentum via the slingshot effect and gets accelerated out instead, or it was part of a double system already and it never sees it's twin again.