Quote:
Originally Posted by Warren Platts
Why would you need data points along the complete orbit? For example, 2M1207b (the first extrasolar planet directly imaged by a telescope) is roughly Jupiter massed, and separated by 46 AU from 2M1207 giving it an oribital period of ~2,400 years. Obviously we haven't watched it that long.
|
And therefore we have no idea of its orbit. In fact, we don't even know its current distance from its parent star. The distance you give is the current separation projected on to the plane of the sky, and the period is a guess made by adopting that distance as the semimajor axis of the orbit.
But I agree that in general we don't need to watch an entire orbit to compute its characteristics: just a long enough segment to allow the rest to be calculated with useful confidence. However, I think
Kullat Nunu is suggesting that to be sure the observed pattern in the data
is an orbit, we probably need to observe at least one cycle.
Grant Hutchison