Slashdot Mirror
Nemesis, the Sun's Binary Star Companion?
0xC2 writes "The Binary Companion or 'Nemesis' theory asserts that a yet-to-be discovered companion to our Sun may actually exist. Recent observations of two nearby stars (assumed companions) show debris disks 'strikingly like the Kuiper Belt int the outer part of our Solar System'. The Binary Research Institute site is devoted to the theory, and presents a concise introduction, list of evidence, and sample calculations in support of the theory. A fascinating read, although the physics and related calculations are not trivial." Has the 'unique theory on the internet' vibe to it, but interesting nonetheless.
... could we possibly find the outer planets by observing their influence on the inner planets' orbits, if there were a freaking brown dwarf in the neighborhood that we didn't know about?
Something like that would've ruined Kepler's whole day.
Actually I am a physicst, and while I'm not an astronomer, I do work with "dynamics" on a fulltime basis.
Not to mention, all but one real astronomer also think this theory is ridiculous. The site linked is by an AMATEUR astronomer, not someone with a formal training in the hard sciences. I'm not contradicting a specialist, I'm contradicting a whackjob internet troll. No, not you - the guy with the binary solar system website.
Yes I do realize binary star systems are not rare. Not detecting it by visible light is exactly WHY classical mechanics comes into play - all you're really doing is dealing with a bunch of forces that go like 1/r^2. Investigating it for the sake of completeness is certainly not folly - however the arguments on the website linked in the article are nonsense.
I can't find it again at the moment - but I saw somewhere that they implied that the inaccuracy of predictions in precession over time was a result of our current theories being flawed, and that the binary theory somehow magically removed this inaccuracy. This is an example of the utter bullshit that anyone with an understanding of nonlinear dynamics would notice immediately. You're dealing with a many-body system here. That's inherently chaotic. That means, it's exponentially sensitive to initial conditions. Therefore, as time goes on your results get worse and worse due to small measurement errors in your initial conditions. NO MODEL can remove this effect and still claim to use newtonian physics - the equations are nonlinear and involve more than three objects interacting - therefore the equations of motion are chaotic. Period.
OF COURSE YOU CAN GET MORE ACCURATE RESULTS WHEN YOU PUT IN AN IMAGINARY EXTRA OBJECT - you can TUNE the parameters of this object arbitrarily to try to fit the experimental data. If I collect a bunch of data from all kinds of experiments, I can easily find a tenth order polynomial and get a very accurate fit to the data. This is also completely meaningless because all those fit parameters have no physical meaning.