Discussion and news about the modern effort to understand the nature of life on Earth, finding planets around other stars, and the search for life elsewhere in the universe

Thursday, October 28, 2010

1 in 4

Having sweated over my previous post I'm hesitant to knock it off the top slot, but this is too good to pass up. Appearing in the 29th October issue of Science is a wonderful paper by Howard et al. The bottom line is that they have made as careful a statistical study of Doppler detected planets as seems practical at this time. The results of monitoring 166 relatively nearby normal stars for five years indicates that an astonishing 1 in 4 such stars may harbor small Earth-mass planets on close orbits.

It's important to be very clear, the planets actually detected in this survey are rocky objects a few times the mass of the Earth, and they are emphatically not in the classically defined habitable zone of these stars - orbiting within 0.25 AU (a quarter of the distance of the Earth from the Sun). However, the extrapolation to true Earth-mass objects is pretty likely to be robust. It'd better be, this claim is effectively saying that there are tens of billions of such planets in our galaxy and that they outnumber giant worlds in close orbits.

This presents a big challenge to certain aspects of how we think planets form. Most current models suggest that there should in fact be a deficit in small rocky worlds in these close-in orbits, since the processes of orbital migration or runaway planet growth tend to thin this population drastically. Something is afoot - and it may also indicate that those holy grail worlds - the Earths in the habitable zone - are far more numerous than we had dared to hope. Indeed, the authors of this paper suggest that Earth-mass planets orbiting at 1AU could be more numerous than their close-orbit cousins.

Howard et al. end their paper by evaluating the implications for Kepler results. It's awfully promising - as many as 260 Earth sized (1 to 2 times Earth radius) planets with 50 day or less orbital periods should be coming our way, and who knows what in the habitable zone of Kepler stars...

3 comments:

kurt9 said...

Their statistics are for systems with planets within a 50 day orbital period. Our own solar system flunks this standard as Mercury has an orbital period of 88 days and is significantly smaller than the Earth.

Caleb Scharf said...

Right, but they extrapolate on the basis of these numbers and some simple physical expectations for planet formation. Obviously the realm litmus test will be to actually detect the Earth mass objects at 1 AU, but we're a hair's breadth away from that (less so with Kepler, more so with Doppler detection).

Prabal said...

We actually did this paper for our journal club several weeks ago - and there were a couple questions that I'd be interested in hearing you expound on Caleb.

They exclude M-type stars in this survey (69 stars) I'm guessing on the grounds that they're interested in constraining the sample to Sun-like stars. Isn't there potentially much to be lost by that selection considering how rich GJ 876 and 581 have been and the overall abundance of M class stars? Considering orbital architecture of those systems I can imagine their extrapolation could be effected as well.

The other issue is whether the biases inherent in RV analysis also biases the power law they come up with - though this really gets to the main thing I'm interested in hearing about - what exactly the development of a planetary mass function would mean for the field in the near future. I understand that developing constraints and general rules for a PMF is useful even with the limited sample of exoplanets we currently have (even as Kepler and other sources being to increase that number), but what exactly will be the immediate boons of a continuously improving PMF? Would it primarily be to allow us to even further constrain survey targets or are we talking deeper understandings of planet formation and exoplanet system architecture even based upon still nascent attempts? There have already been a couple papers on a planetary mass function on arXiv and it'd be nice to see a little discussion on why it is such an exciting topic for the near future.