Wednesday, August 19, 2009

The continuing mystery of Type Ia supernova

After two days at a conference on stellar death and supernovae, it's become pretty obvious that the more we learn about Type Ia supernovae, the less we seem to know or understand. It's fairly amazing that after 40 years of intensive work on the problem, our understanding is still so muddy.

First, what do we know? Type Ia supernovae are a type of exploding star. In these explosions, we do not see any hydrogen or helium, which are the most common elements in the Universe. We also see lots of silicon and iron, which are the products of explosive nuclear reactions involving carbon and oxygen. These two facts make it seem likely that the exploding star is a white dwarf. White dwarfs are the nuclear ashes remaining after most stars finish their lives; they're made mainly of carbon and oxygen, and they have very little hydrogen and helium.

We also understand pretty well what happens after the explosion. Type Ia supernovae have a relationship between how bright they are and how long they take to fade away. This is very useful, because if we see Type Ia supernova in the distant universe, and if we can measure how long it takes to fade away, we know how bright it was. This allows us to figure out how far away the explosion was (because things appear fainter the further away they are). This relation allowed astronomers to find some of the first evidence for "dark energy."

But there are some mysteries. First, how do you get a white dwarf to explode? The typical white dwarf is very stable and resistant to explosion. If we can take a white dwarf and add more and more material to it, eventually it will get big enough that gravity will squeeze the white dwarf, cause it to heat up, and set off nuclear reactions. We think.

But how can you add material to a white dwarf? There are several ways. If you take two white dwarfs and merge them together, you might get a big enough white dwarf to explode. But colliding white dwarfs is hard, and we've never actually seen two white dwarfs that are going to collide and be massive enough to explode. This doesn't mean it can't happen, because it is hard to confirm that you have two white dwarfs in such a system, but until we find such a system, this is just an idea.

Another way to grow a white dwarf is to have it orbiting a normal star. If the white dwarf is close enough to the normal star, the white dwarf's gravity will siphon gas off of the normal star and onto the white dwarf. We actually do see several stars where this is happening; these are called cataclysmic variables. But in cataclysmic variable stars, there is usually too little material being transferred to build up the white dwarf to an explosive mass. There are some special types of these cataclysmic variables that may work, with names such as supersoft X-ray sources and recurrent novae, but we still can't be sure that they will explode.

Either of these methods, colliding white dwarfs or cataclysmic variables, take time to work, perhaps billions of years. But about 5 years ago, astronomers started discovering more and more evidence that the number of type Ia supernovae in a galaxy depends on how fast the galaxy is making new stars. That was a complete shock -- how can a process that takes billions of years to work know how fast a galaxy is making stars right now? Maybe there is a way of adding material to brand new white dwarfs in very short times.

Supernova scientists have just been coming to terms with that finding and developing explanations, when yesterday one of the teams counting Type Ia supernovae stirred the pot further. They said, essentially, that maybe they had made a mistake, and the rate of Type Ia supernovae in a galaxy doesn't depend on its current rate of making stars, but is just an artifact of how we were doing the observations. That claim caused a lot of controversy and discussion, and needs a lot of further exploration.

So, in short, we don't know exactly what causes Type Ia supernovae, and we have mixed signals as to how long it takes to start getting these supernovae. I'm cool with this, because it means there is a lot of work yet to be done, and lots of ways I can contribute to understanding these supernovae.

But there should be some concern, too. Scientists are using Type Ia supernovae to try and understand dark energy and to see whether it changes over time in our Universe. But can we believe the results that supernovae give us if we don't understand the explosions even moderately well? There are certainly other lines of evidence that dark energy exists, so the Type Ia supernovae are clearly pretty good for this purpose. But if Type Ias change at a slight level over time, that change in the supernovae could be interpreted as a change in dark energy instead. For this reason, we need to keep studying Type Ia supernovae and the stars that they come from, whatever those stars might be.


  1. Very compelling post; however, a couple of points need to be made:
    1) We need to make a distinction between what is "evidence" [for dark energy] and what is an "observation". Type Ia supernovae are/ have been used as very accurate extra-galactic standard candles; "Dark Energy" and "Dark Matter" are terms that are often thrown back and forth with very little understanding with two sides in a discussion talking about 2 very different things. Those terms are self-discriptive in that we still really don't know *why* the Hubble constant is changing at verh high Z; we observe the effect and attribute it to this "Dark Energy". "Dark Matter", on the other hand, is a catch-all term that is used to explain what could be described as a gravitational anomaly. The mass necessaary for the the *observed* rotational velocities of spiral galaxies cannot be accounted for - essentially Newtonian gravitation/ mechanics falls short. There is a long way between "evidence" for something that we still have no idea about and an "observation";

    2) We have a fairly decent idea how Type Ia supernovae occur: a parasitic what dwarf accretes hydrogen from an orbiting companion in a binar system; the accreting hydrogen causes the white dwarf to approach the Chandrasekhar Limit and, as it does, the temperatures necessaray for p-p chain reactions to occur (10M K) are reached. The surface hydrogen that has been accreting is ignited in a runaway thermonuclear fusion reaction and, since it is a Carbon-Oxygen white dwaft, the newly formed Helium joins in as the conflagration continues, raising the temperatures even higher, igniting the Carbon and so forth with the energy released unbinding the star. We don't observe any hydrogen or helium lines because it was all converted to heavier elememts, unlike a Type II supernova where we *do* observe hydrogen lines (the expanding hydrogen shell from the star's ARGB phase).

  2. Thanks for your comments Tom. The exact progenitors of Type Ias are far from settled; here at the conference there have been vigorous discussions as to whether the single accreting white dwarf model, the merging double white dwarf model, both, or neither are responsible for some/all/none of the observed Type Ia supernovae. Most everyone agrees that you have to start with a white dwarf near or at the Chandrasekhar mass, and then the supernova is ignited by the start of carbon fusion in the interior of the white dwarf. But how you get the white dwarf to the Chandrasekhar mass is definitely still a matter of debate.

  3. Thanks for the clarification, Kurtis; I wish I could have attended the conference! I look forward to reading your papers didn't realize you were so widley published. Brilliant!