Thursday, October 23, 2008

One star can just be an oddity, but three makes a class

In May, my collaborators and I announced we had discovered a new kind of star. Our findings were bolstered this week when another group found two more stars of the same kind. This was a breath of fresh air for us, because one object can always be explained as "weird" or "unique." But three objects means that we've truly discovered a new class of star.

As a reminder, our discovery was of variations in light from a relatively newly-discovered type of white dwarf, one with carbon atmospheres. White dwarfs are the leftover ashes from stars that have used all of their nuclear fuel. 99.9% of white dwarfs have a thin atmosphere made of hydrogen or helium, but one in a thousand white dwarfs seem to have pure carbon atmospheres. This is hard to explain, because hydrogen is everywhere, and it is so light, it will float to the top of any white dwarf star. So, the stars are mysterious to begin with.

Then, we found that one of these carbon-atmosphere white dwarfs changes its brightness every 417 seconds. We had predicted this, but were still surprised to actually find it. However, our prediction was for a very specific type of variation caused by the surface of the star sloshing back and forth, and we aren't sure that we are seeing this sloshing. Since our discovery, we've thought of lots of other things that could be causing the changes in the star's brightness, from a companion star blocking some of the light, to starspots, to the star ripping material off of an unseen companion. None of these ideas really fit the data we have, and we need new and different observations to really understand this star. Those observations will have to wait, though, because the star is currently behind the sun, out of view until next spring.

So, we still aren't sure exactly what we discovered, and we thought that we had, perhaps, discovered a one-of-a-kind star. With 10 billion stars in our galaxy, even very rare stars should be around. For example, if a certain type of physics should happen in only one in a million stars, then ten thousand stars in our Milky Way galaxy will show that strange physics. Perhaps our star was some one-in-ten-billion oddity.

Some answers in that regard came out last week, when University of North Carolina graduate students Brad Barlow and Bart Dunlap announced the discovery of two more variable carbon atmosphere white dwarfs. Moreover, these two new stars have some of the same peculiarities as our star. So, while I would argue we still aren't positive exactly what we are looking at, we now know that the star my collaborators and I discovered is not a one-of-a-kind star. These stars are a unique class of object.

So, what are these stars? I think there are three viable hypotheses, which I'm ordering by the likelihood I'd give each option. Other people probably would disagree with me, but if they want to argue, they can get their own blogs.

  1. A "sloshing" white dwarf -- This was our first idea, so I'll stick with it as most likely. The main way to test this idea is to look even harder. Most pulsating ("sloshing") white dwarfs slosh at several different frequencies. If we can detect more than one frequency in these white dwarfs, that will seal the deal. One group is already claiming to have detected such additional frequencies, but their paper is not yet published, and it is important enough that we'd like to independently verify it. I'd give this a 60%-70% chance of being the case.
  2. A rapidly-spinning white dwarf with a spot -- If the white dwarf is spinning every 417 seconds, and if it has a starspot (like a sunspot) on it, then we would expect to see the variations we actually do detect. But, most white dwarfs spin very slowly (over periods of months or years if at all), and no starspots have ever been detected on white dwarfs before. But, there's a first time for everything, and it may be possible to relate the fast spinning with the strange atmosphere. I'd give this scenario a 20%-30% chance of being right.
  3. A white dwarf stealing material from an unseen companion white dwarf -- This was our original alternative explanation, and we can't yet rule it out. The exact details of the mass transfer would contradict most ideas on what would happen if two white dwarfs came very close together. But, maybe hypothesis is wrong, as it's never been tested outside of a computer. For that reason, I think this idea has a less than 10% chance of being right.
  4. Something Else -- Maybe there's another explanation. I think we covered the obvious ones, but maybe I overlooked something?

At any rate, these carbon atmosphere white dwarfs are proving to be very mysterious!


  1. Hello. Hope you don't mind, but I'm curious about something.

    I could be wrong, but to my knowledge, the theory goes that starspots (at least in main sequence stars) are caused by magnetic field lines intersecting a star's photosphere. As such, they're typically associated with magnetic stellar activity (such as flares, CMEs, etc). In turn, magnetic activity is expected of stars with convective envelopes, due to the flows of plasma believed to create such magnetic field lines.

    But white dwarfs are believed to consist primarily of electron degenerate matter, which would be highly compacted and unlikely to be very convective.

    Is there another mechanism in white dwarfs (such as an extreme temperature gradient)...?

    As I say -- just curious. I know quite little about white dwarfs. :)

  2. You are quite right on all you say. The missing ingredient is that the atmospheres of white dwarfs are not degenerate, and can form convection zones at specific temperatures. These convection zones are the reason that white dwarfs of the right temperature can pulsate.

    There are white dwarfs with strong magnetic fields, but it is thought that these fields are not formed by convection. Their origin is mysterious, but may be a relic from their parent star (just like neutron star magnetic fields are). But these magnetic fields should prevent convection, which would prevent the pulsations.

    In fact, two of the three variable carbon-atmosphere white dwarfs have these strong magnetic fields. This is one reason we think that star spots may be the cause. But star spots would not explain the third, non-magnetic star's variations. This is why we are a bit mystified right now!

  3. Fascinating. Thank you! I didn't realise white dwarfs could be properly variable. It certainly sounds like an interesting problem to be working on.

    I hope you don't mind my asking a little more, but I'm curious about these white dwarf atmospheres, especially the carbon rich ones. Obviously they must show carbon atomic lines in their spectra... Do they ever show molecular lines? C2, for instance?