Cav's Place

As the Kepler team is bent on secrecy, at one point even attempting to obtain approval not to release any information until the primary mission finishes, which is scheduled for the end of 2012, we can only be thankful for such slips that can shed just a little bit more light on their finds. Sure, they released something, but that’s just raw data about 306 of the 706 planet candidates they had at the time, which may be interesting for very dedicated amateur astronomers but otherwise doesn’t count for much unless others would be able to follow up on that set of data, which would require plenty of time on good enough telescopes, and that is in very short supply.
So I’m quite pleased that such a piece of information came out, no matter how it did, though the people involved most certainly are not. We’re basically just talking about a single slide, which is questionable in itself, but it is nevertheless an update. I would like it very much if they’d take this as a cue to be more open and release information as soon as it becomes available, therefore also being able to properly explain what that information means and how it should be taken, but fear they’ll unfortunately go the opposite way, becoming even more guarded and erecting more safeguards in order to prevent even such slips from happening again…

I looked for and watched that speech after a notice stating that the team was working hard on responding to the reactions caused by it, but before reading what the fuss was actually about. What I can say is that, while the guy is certainly not the best person to give speeches, the confusion wasn’t really his fault. In fact, I wasn’t sure what they were supposed to be busy responding to at the time I finished watching his speech, so I then had to look for other articles in order to figure out what the misunderstanding was. Sure, he said “Earth-like” a number of times, but he was obviously referring to data regarding the size of planets, so the thought of taking that term as meaning “habitable” didn’t cross my mind for even a single moment. In fact, the terms used for comparison were listed under the sizes on that chart, with “like Earth” being clearly defined as “having a radius less than twice that of Earth”.
Yet he still came under fire and likely had to pay more of a price than we’re led to believe for what is in fact a misunderstanding generated and propagated by others. I’m sure the main culprits are some tabloids that thought it’d make a nice headline to say there are so many planets just like Earth, likely pretending to misunderstand on purpose, just to make it sound more interesting, which is what such publications usually do. But he had to publish a statement meant to clarify the matter and probably had to face various sanctions as well, not to mention the mistrust of his colleagues from now on, who will likely keep blaming him for… Well, for the fact that they didn’t make that data public sooner and in a more “official” and organized manner, which would have allowed them to explain it better as well.

Yet, as I said before, that slide is pretty questionable in itself, because it lists a total of 1160 planet candidates at the top, but if you estimate the number of planets in each category according to the bar sizes and the numbers listed on the left you only get to a total of about 270, even lower than the 306 covered in a paper analyzing the candidates about which data has been released last month. That seems to suggest either that the numbers are wrong or that the slide is based only on a specific selection of planet candidates, both of these significantly reducing the credibility and relevance of the information offered.
But, if correct, that single slide does provide a fair bit of additional information, starting with that number at the top, which appears to suggest that about 454 new candidates have been taken into consideration in the few weeks between the official release of data and this speech. That number could perhaps be reduced to about 300 if the total includes the roughly 150 candidates that are mentioned in the paper as being known at the time of the release but labeled as likely to be false positives and therefore not included in the count, but it’s a significant increase even so. Yet the most significant piece of information is just the size distribution displayed, which contradicts the paper mentioned above, which indeed stated that most planet candidates had a radius no more than four times that of Earth but listed those with a radius two to four times that of Earth as clearly outnumbering the smaller ones. The paper clearly stated that only candidates with a radius at least 25% larger than Earth’s were included in the data that was made public, even though smaller ones have been detected, so it’s easy to assume that the percentage of small planet candidates detected is significantly greater than what the release suggests, but this is the first time that assumption is confirmed in any way, which is a major development!
(Interestingly, that paper is somewhat questionable as well. The first reason is that apparently they did away with the three transit requirement, listing promising candidates after only two transits, though I can’t quite understand what can be so promising about two similar dips in a star’s brightness coming at apparently random intervals. The second reason is that there are at least 12 planet candidates that shouldn’t be there even if you only count two transits, meaning a single full orbit, because the data covers 43 days of observations and those ten have estimated orbital periods longer than that. Even if we’re to assume that the first transit happened during the first moments of observation, the four with the longest periods couldn’t have even covered half their orbits during that time, as we’re talking about estimates of roughly 87, 119, 207 and a staggering 10389 days, respectively, so not even the infinitesimal dip in brightness that may be noticed when a planet is directly behind a star could have possibly been observed and taken as confirmation in those cases. I really don’t see how a single dip in a star’s brightness can be considered a planet candidate, nor how can one claim to reliably estimate that potential planet’s orbital period based solely on the time it took to pass in front of the star, if that’s indeed what happened.)

Since they announced that they won’t even be releasing information on the 400 candidates held back from the announced 706 until 2011, I don’t exactly hope to see many things happening in the near future, but I will be watching, just in case. With the Earth-based observation season for that patch of sky ending in September according to what I heard, they could at least release an updated list based on those first 43 days of observations towards the end of the year, confirming what they could confirm by then and eliminating those proven to be false positives, but we’ll see. At this rate I really wonder how many decades the team’ll need to go through all the three and a half years of observations that’ll be generated by Kepler during its primary mission if they insist on holding on to the data so much and not allowing others to help…