I’m still confused about what I discovered in the Kepler Telescope database. Perhaps talking it over here will help.
Perhaps talking about it online will help. First the good news:
1) There is never any question about distinguishing between surface spots on the star and transiting planets – the distinction happens to be trivially easy, which is startling in itself.
2) There is never any question about a transit being masked by statistical fluctuations in the database. There are no events that look sort-of-like transits. Every transit is either a deep drop in the light intensity or repeated very often very regularly, or both.
3) By far the majority of transits observed are from binary star systems, to the extent of at least ten star transits per planet transit.
4) It is impossible to distinguish between the transit of a “brown dwarf” and a large planet. That’s because brown dwarfs are actually computed to be smaller than Jupiter. Anything significantly large than Jupiter is a star. Proxima Centauri, a pretty average red dwarf, has a radius only 1.40 times that of Jupiter. So by “planet” astronomers mean “planet or brown dwarf”.
Now, the problem.
There are two ways to calculate the size of a planet from the light curve. In one, the direct method, the maximum drop in light intensity is proportional to the square of the planet’s radius. In the second method, the speed of a planet in orbit together with the time the planet takes to go from just touching the edge of the star to the time of full opacity is proportional to the planet’s diameter.
The problem is that the two methods don’t agree. The planet diameter calculated from the speed in orbit is significantly larger than that from the maximum drop in light intensity. I’ve come up with several hypotheses, but no good one.
1) Errors in the telescope or software that cause a change in the shape of the light curve. I’m pretty sure I can rule that one out.
2) Partial eclipses. This can only account for at most a small percentage of cases.
3) The planet calculated speed is wrong because the planets are all in strongly elliptical orbits. We know that most binary stars are in strongly elliptical orbits, so that makes sense. But if this was true then the planet diameter calculated from the speed in orbit would tend to be SMALLER than that from the maximum drop in light intensity, not larger, because a planet is more likely to be transiting across the face of the star when it is closer to the star.
4) There’s a lot of light leakage into the telescope from background objects. I know that there’s some light leakage because sometimes the same object is seen in the light curve of more than one star, ie. the light observed by Kepler is not from one star but is the blended result from many stars. This is the one hypothesis that seems most likely, but it is also the hypothesis that has the most drastic consequences – if true, then astronomers are mistaking hundreds of binary stars for planets.
5) Each star has a fuzzy edge – the central star brightness is constant but then the brightness drops slowly across the edge. This doesn’t agree with what we see for the Sun, and doesn’t agree with mathematical models of stellar edge darkening.
6) Large planets are transparent.