View Single Post
  #3  
Old January 13th 05, 09:53 AM
Mike Dworetsky
external usenet poster
 
Posts: n/a
Default



"Nick Mason" wrote in message
...
I was listening to Radio 4 earlier this evening and they were talking
about NASA and their Kepler mission and locating planets around distant
stars using the transit method where the light diminishes due to the
passage of a planet in front of the star.

I didn't hear all the program but it set me thinking, doesn't this
depend on us, Earth, and the distant planet being in the same plane so
that we see the transit? If the distant planet were orbiting so that
from Earth we never saw it transit the star we wouldn't know it existed,
would we?

Am I missing something here?

Isn't it statistically more likely that any such planet would orbit in a
way that we wouldn't see it rather than be aligned correctly so that we
can?


Yes, such transits would be statistically rare. The astronomers wanting to
do this experiment are after information on the general statistics of
planets, rather than completeness of discovery. Think of it this way:

Question to be answered: What percentage of stars of spectral class G0V-G5V
have planets of "Jupiter class"? (Similar analysis for other spectral
types...)

Assumption: The inclinations of such systems to the line of sight to Earth
are random.

(Simplified) Analysis: The chance of any one star of these classes with a
planet, producing a transit we can observe, is 1% (say), because the range
of inclinations is very restricted. Let's monitor 100,000 such stars as a
statistical sample. If we see 1,000 stars with transits, the implication is
that nearly all such stars have J-type planets. If we see 100, only 10%
have such planets. Etc.

If such planets are relatively common, then they would get statistics on how
many stars had 2, 3 or 4 such planets (each one would produce a different
transit on a given star).

By observing in a rich star field, many stars can be monitored
simultaneously.

There is also the possibility with very precise measurements (needs large
telescope) that terrestrial planets could be detected, but the amount by
which the light would drop is very small.

--
Mike Dworetsky

(Remove "pants" spamblock to send e-mail)