Several Extrasolar Planets May Be Optical Illusions

person-0.9a writes "Seems that a few of the extrasolar planets detected via their sun's wobble might be written off according to a CNN article. You can also read it about it in New Scientist."

Comments

Very Small number of planets

[HimalayanRoadblock]

Only a small number of planets are cast into doubt, those being ones around "new" stars. A very large majority of the extrasolar planets found are around very old stars.

clarification of illusion

[beaverfever]

In the article it's described not as an optical illusion - it was a spectral illusion.

I can now go to my grave knowing that at least once in my life I used the term "spectral illusion" in a serious discussion.

Re:Anyone who knows more care to elaborate?

[renard]

Sun spots can masquerade as Doppler shifts by altering the spectrum of the star systematically as the star rotates. You think it's the planet's period, but it's actually the star's rotation (at the latitude of the sunspot). That's why Greg Henry's discovery of photometric variability with the same period as the "planet" is so damning.

Planetary transit searches will be subject to sunspots as a noise source (star's light changing by small amounts, erratically), but are unlikely to be fooled by them. A planetary transit causes the star's light to dim in a distinct flat-bottomed way that a sunspot cannot mimic.

Note also that no one has discovered a planet yet using the transit-search technique. The transiting planet of Henry & Charbonneau was known to exist already thanks to the spectroscopic surveys.

-Renard

Re:Of course

[joh3n]

Not quite. The telescopes are indeed strong enough to see the planet. In fact, you can look up in the sky and see a few of them with the naked eye. The problem is that the star is too damn bright in comparison. Also, due to atmospheric blurring effects, it is very hard to seperate the planet from the star as a light source.

Re:Anyone who knows more care to elaborate?

[TMB]

(preface: yes, I Am An Astronomer)

The important thing to realize is that you're measuring the spectrum integrated over the entire surface of the star WEIGHTED BY THE FLUX (read: brightness) AT THAT PART OF THE SURFACE.

The star is rotating, so part of the star is moving away from us and is red-shifted, while part is moving toward us and is blue-shifted. If the surface were all radiating at the same level, then we'd always see part of the spectrum blue-shifted and part red-shifted.

Now stick a big starspot on. The starspot is fainter than the rest of the star. When the starspot is on the approaching side of the star, there is less blue-shifted light and so the net spectrum appears slightly red-shifted. Half a period later, the starspot is on the receding side of the star so there is less red-shifted light, and the net spectrum appears slightly blue-shifted.

The end result is a spectrum that systematically shifts back and forth, very similar to that of a star with an orbiting planet.

[TMB]

It was already in doubt

[Anonymous Coward]

Of the two groups which (nearly simultaneously) dectected the apperent wobble 2 years ago, one wrote:

If photemetry or "S" value measurements continue to show periodicities similar to the observed Doppler velocity period, this would suggest that the source of variation is intrinsic to the star rather than an orbiting planet.... We are not yet completely convinced of the planet-companion interpretation for the velocity variations of HD 19622