I’m currently on a bus to Sydney, from which I will be flying to Munich tomorrow morning at the start of yet another month-long trip; you’d think I would have learned by now to keep them shorter. It’ll be a long couple of flights. I’ll have a five-hour layover in Singapore, so if the airport there is as space-age as the likes of Tom Friedman would have me believe, I’ll have some fun photos to share. But for the time being, there’s not much else I can do except type.
For the record and before I get into much more science, the last couple of weeks have been fairly dull. SkyMapper has been broken — the dome azimuth wasn’t tracking and so we were getting pictures of the inside of the dome, instead of the sky — but it is fixed and back in operation observing SN fields tonight. Meanwhile, Fang has introduced pixel masks to get rid of some of the more pernicious artifacts in our images (aficionados: crosstalk from saturated stars on the other amplifier of each CCD); I’ve finished the code to enable upper limits in historical light curves, as well as automatic cross-matching of pipeline detections to catalogs of known things like asteroids or variable stars. It’s not too pretty right now and I can feel another refactoring coming on soon, but for now it works.
Today’s colloquium speaker was Onno Pols from Nijmegen (“NYE-may-gen”), to talk about a subject near and dear to my heart: type Ia supernova progenitors. I’ve mentioned these before, so those who know me or my work will know that the question is a troublesome one. The progenitors of core-collapse SNe, including almost all SNe II and some SNe I, are well known to be massive stars, the evolution of which in isolation I would say is understood relatively well; the tough part is that theorists are having trouble making the stars actually explode in simulations (but they’re working on it).
In contrast, while we believe SNe Ia are exploding balls of carbon and oxygen and have little trouble blowing up simulated versions of them (though the precise triggering process is still not understood in detail), there is virtually no way normal SNe Ia could be produced by single stars of any mass evolving in isolation. We have to understand how stars in binary systems evolve, and this can be complex because the evolution of each star can feed back on the other: mass loss as a stellar wind, orderly mass transfer to one star in a stream fed from the other’s Roche lobe overflow, formation of a common envelope enshrouding the cores of both stars in a single atmosphere… These effects are studied by evolving each of two stars using well-understood single-star evolution codes, then putting the abovementioned processes in by hand in a simplified framework called binary population synthesis.
Picking a system with the right set of parameters (the masses of two stars, and one orbital separation) to lead to a SN Ia requires some care, and the results necessarily also depend on other assumptions you’re making to simplify the problem. The current story is that it is still fiddly and there are only narrow ranges of system parameters which will result in a SN Ia. Depending on who’s doing the figuring and how long the system’s evolution lasts before explosion, our best theoretical models predict only 1-10% of the number of SNe Ia that we should see. So something’s gotta give — either some assumptions in our models are wrong, or many SNe Ia come from evolutionary scenarios we haven’t even explored yet.
If binary population synthesis wasn’t hard enough to understand, Onno’s group is now looking at trinary population synthesis: evolution of systems of three stars. It turns out stable systems of three stars can exist if the third star is placed at least five times as far away from either of the first two as those two are from each other; what you have then is basically a binary system, but with a distant influence which can affect the evolution of that binary system in profound and bizarre ways.
Some of the plots that Onno showed, for example, discussed the existence of Kozai cycles in which the eccentricity (or ellipticity, or aspect ratio, they’re all related) of the inner binary’s orbit can oscillate wildly with periods of (only!) hundreds of thousands of years, greatly increasing the strength of tidal forces at periastron (the point of closest approach). This means that the size of the inner binary system’s orbit tends to shrink when it would otherwise grow, or shrink much more quickly than you might expect, bringing the stars into contact or sometimes even making them collide.
If you thought such a collision might cause some fireworks, you’d be right. It might even cause dramatically enhanced rates of exotic supernovae like the ones I study. For example, Cody Raskin’s simulations of colliding white dwarfs tended to produce explosions that look a lot like the insanely bright SN Ia 2007if, which I studied in this paper. But we wrote them off at first because white dwarfs were tiny stars and the only way you could get them to collide is by putting a lot of them in a very small space, like the core of a globular cluster. After hearing Onno’s talk, I’m not so sure anymore.
If Onno and his group are right, this could revive quite a few otherwise moribund channels to production of SNe Ia — but at the cost, I think, of losing almost all predictive power on the rates of such events. In a world where anything can happen, we still need to know why particular things do. There is some diversity among “normal” SNe Ia, but not a huge amount, and the real weirdos like SN 2007if are still rare. The facts of what we observe are bound to put some pressure back on Onno’s models, and the challenge is to find a formulation of their formalism that can express it.
[ed:] In the process of writing this post I found another recent article by Cody Raskin citing my own work on SN 2007if. While ADS-stalking citations to one’s own work can obviously be a form of vanity, I’m starting to find that all too often it shows me what interesting and useful work I might have accidentally missed while swimming through the vast literature. I’m keen to read this now, but I have to go to sleep for the time being. Maybe on the plane tomorrow…