Thursday, September 21, 2006

Bad Astronomy on CNN

Stealing a bit of thunder from the Bad Astronomer, I noticed an article on CNN this morning that had a few errors in it. Nothing major, but I figured it might be fun to address.

The first part of the article I don't like is the title. It's sensationalist. It claims that scientists are rethinking things as if the notions currently held have just been shattered and it's time to start over. Yet later on the article goes on to say the discovery "doesn't necessarily undermine that discovery or other previous research". So there's some sloppy writing right off the bat.

The next problem is at the end of the first sentence when the author says that astronomers are thinking about how stars "disintegrate." Disintigratiobn implies things are being broken down into their most basic forms, namely atoms. In a supernova, the exact opposite happens! Stars, while being extremely complex in a mechanical sense as I'm finding out this semester, are actually made up of pretty much the simplest stuff you can think of. In the large stars, the ones prone to becoming supernovae, most of the atoms are ionized and since it's mostly hydrogen (at least on the visible surface), that means that it's going to be a bunch of protons and electrons floating around. Can't get much simpler than that.

Meanwhile, when a supernova occurs, all these building blocks are smashed together making more complex nuclei which gives rise to all the "metals" in astronomy (astronomers call metal anything with an atomic number greater than helium). So "disintegration" is the wrong word. "Integration" perhaps? But that might stir up bad memories of calculus...

The next, and what I'd say is the biggest, error is saying that the Chandrasekhar Limit is "1.4 times the size of our sun". Size has nothing to do with it. It's all about the mass.

Let's take a closer look at what causes a supernova. In any object, the sun and stars being no exception, there always has to be a balance of forces. The mass of the object, makes a gravitational pull that goes towards the center. Thus, if nothing was stopping it, all objects would collapse down to a single point. Fortunately, there's lots of ways to stop this collapse. All you need is an opposing pressure in the opposite direction. In stars in the main portion of their lifetime, the heat and energy generated in the core pushes the outer layers back to keep things from collapsing.

But what about things like the Earth and stars that have already used up all their fuel? What keeps them from collapsing? The answer is that atoms themselves to the work. The charged electron clouds around the atoms repel one another just like how those colorful balls in McDonalds playpens settle in together.

This is the case for the white dwarf stars that are the progenitors for the type of supernova discussed in the article. So what's with this limit?

What ends up happening is there's a point where atoms just can't hold themselves up anymore. Pile on enough mass and the atoms collapse. So what the magic limit an object without any additional outwards pressure can support? It's generally been held that it's 1.4 times the mass of the sun.

White dwarf stars are the end result of stars similar to our sun. They're normal stars, that as they die, blow off their outer layers slowly creating what's called a planetary nebula. What's left behind is the burnt out core of the star. But at this point, it's not gaining mass and is obviously less than that limit (otherwise it would have gone straight into a supernova which is a different class than the one talked about in the article).

So where does the extra mass come from? It turns out that a very appreciable fraction of stars in the universe exist in binary systems. If one star dies by blowing off its outer layers, and the remaining core is just below 1.4 solar masses, it's possible that, when it's companion starts swelling up to toss out its outer layers, that the first star will grab some of that material. If there's enough to push it over the limit, the white dwarf collapses, resulting in a supernova.

Now that you know the good astronomy behind the article (and where it buggered up), let's take a moment to analyze what the article is saying and what its implications are.

The article discusses a supernova that was discovered in a galaxy ~4 billion light years away (that's pretty darn far) that the progenitor white dwarf was two solar masses before it decided to collapse. This is well over the formerly thought limit of 1.4 solar masses.

I'm not quite sure how the masses were determined in this case. There's lots of tricks, but given the distance, I'm inclined to think that it may just be a case of error involved in observations. But assuming that the presumed mass is correct, this still doesn't throw out the Chandrasekhar limit. It just means something funny's going on. One possibility that the article mentions is that the white dwarf was spinning rapidly which would produce a centrifugal force that would help take some of the weight off. Or perhaps this supernova wasn't formed in the manner we usually attribute to their formation.

But what are the implications if it turns out that the Chandrasekhar limit isn't as steadfast as previously thought or is alltogether wrong? It turns out that this class of supernova is very important in astronomy. Because we think we know what the mass is of the exploding star, we can determine how bright the explosion should be. By comparing that to how bright it is, we can figure out how far away things are. This is one of the key determinations in finding the age of the universe.

So if these "standard candles" turn out to burn a bit brighter than previously though, it means our estimated age would be off too. Alltogether, it's an interesting article even though a few things were off. I'll have to look when I have more time and see if I can find the publication.

11 comments:

Anonymous said...

It was a Nature article; hence, sensationalism. Just posted on astro-ph: http://arxiv.org/abs/astro-ph/0609616 ... it also looks like it just appeared today in Nature.
I haven't read it yet, but often Nature articles have the problem of not actually explaining what they did and exactly how they came to their conclusions, due to limited space and limited knowledge of the field by the audience.

Jon Voisey said...

You're right that it's an oversimplification, but shhh. I don't expect the majority of people to have the background to really have the background to understand all the details (nor do I have the time anymore to write 10 page long posts to explain such things *sigh*). Ultimately, I try to keep this blog about the level of an intro astronomy class.

Anonymous said...

Any thought on the spate of new books by Smolin and others exposing "string theory" for the vacuous fraud that it is.

Kind of like multiverses, etc.

I mean, you smear ID for not being testable. (Although is is*)while crowing about stuff like "mutiverses".



*Simple, create life that develops intelligence. Do it my a mindless, undirected process. Repeat in laboratory when asked. Case closed.)

Jon Voisey said...

I haven't had much chance to look into string theory yet. But from what I have learned about it, "theory" is all together the wrong word to apply to it. Hypothesis might be better.

While much of the parts of String (whatever) are not testable as you note, there are consequences of it that may manifest itself in tests that high energy physicists will be able to perform when the Large Hadron Collider (I think that's the one) comes online in the next few years. If energy seems to be "disappearing", it could be indicative of some of the tiny extra dimentional space that string theory predicts.

The difference between Strings and ID is that String is based on mathematical extrapolation from extremely well tested physical laws which give rise to the curiosity that is the basis for the concept. While still not nearly enough to call it a theory, it's still far more than ID has.

TheBrummell said...

The test of ID you propose, Emanuel Goldstein, would not actually constitute a test of a falsifiable hypothesis.

Imagine we wish to disprove the tired old ID concept of "irreducible complexity", using one of their favourite examples: the bacterial flagellum.

We set up a vat of bacteria, let's use E. coli, all lacking flagella. We apply some selective pressure favouring movement ability - perhaps the richest resources are in a bag, slowly moving through the broth, or something.

We wait X number of E. coli generations - at a peak reproductive rate of cell division every 20 minutes, we can afford to pack in thousands of generations using a year-long operating grant.

So the "test" is: did any of our E. coli develop a flagellum on their own, thus proving that this structure (at least) is not "irreducibly" complex?

The answer is: it doesn't matter. If none of the buggers get flailing, this doesn't mean the structure can't evolve, it just means we didn't give them enough time, or the right mix of selection pressures, or something. Likewise, if any of the buggers do start flailing, this proves nothing, since our undefined designer could have reached into the vat and caused the structure to evolve. We simply have NO WAY of controlling for God, or any other designer without well-defined restrictions and constraints.

Stephen said...

Stars are most just hot air.

We get that on the web too.

Jon Voisey said...

Of course they are.

Steam has less calories.

TheBrummell said...

The test of ID you propose, Emanuel Goldstein, would not actually constitute a test of a falsifiable hypothesis.

Imagine we wish to disprove the tired old ID concept of "irreducible complexity", using one of their favourite examples: the bacterial flagellum.

We set up a vat of bacteria, let's use E. coli, all lacking flagella. We apply some selective pressure favouring movement ability - perhaps the richest resources are in a bag, slowly moving through the broth, or something.

We wait X number of E. coli generations - at a peak reproductive rate of cell division every 20 minutes, we can afford to pack in thousands of generations using a year-long operating grant.

So the "test" is: did any of our E. coli develop a flagellum on their own, thus proving that this structure (at least) is not "irreducibly" complex?

The answer is: it doesn't matter. If none of the buggers get flailing, this doesn't mean the structure can't evolve, it just means we didn't give them enough time, or the right mix of selection pressures, or something. Likewise, if any of the buggers do start flailing, this proves nothing, since our undefined designer could have reached into the vat and caused the structure to evolve. We simply have NO WAY of controlling for God, or any other designer without well-defined restrictions and constraints.

Stephen said...

Stars are most just hot air.

We get that on the web too.

mollishka said...

Because we think we know what the mass is of the exploding star, we can determine how bright the explosion should be. By comparing that to how bright it is, we can figure out how far away things are. This is one of the key determinations in finding the age of the universe.

So, this is a bit of an oversimplification. Type Ia supernovae are not observed to be the same bolometric (intrinsic) luminosity. What is observed is that they all have the same shape light curve: the brighter ones take longer to get bright and then to dim than the less bright ones. This difference is known as the "stretch factor" ... I don't know what the source of this difference is, but then, I don't think anyone does (this is an area of active research... see http://arxiv.org/abs/astro-ph/0609540 for a this-week example). Supernovae are incredibly complex beasties... no one has yet been able to model (as in, in the computer simulation variety) a supernova that, you know, explodes. Suffice it to say it's a myth of old legends that Type Ia supernovae have the same intrinisic luminosity, and that we even know theoretically what that luminosity is.

mollishka said...

It was a Nature article; hence, sensationalism. Just posted on astro-ph: http://arxiv.org/abs/astro-ph/0609616 ... it also looks like it just appeared today in Nature.
I haven't read it yet, but often Nature articles have the problem of not actually explaining what they did and exactly how they came to their conclusions, due to limited space and limited knowledge of the field by the audience.