When do you reject a theory?

Recently I’ve been wondering: when have you amassed enough evidence to reject a theory?  At some point astronomers decided to cast aside the geocentric cosmology in favor of the heliocentric cosmology.  Later, the Big Bang theory replaced static cosmologies.  How did they know they had enough evidence to reject an observationally supported theory?

Last week I found an interesting paper on radio galaxies.  Radio galaxies are normal galaxies with supermassive black holes in their center that shoot out a highly collimated jet of matter moving at 99% the speed of light.  This jet is mostly seen in the radio part of the electro-magnetic spectrum, hence the name “radio galaxy.”  When radio galaxies shoot these jets almost directly at us, things appear quite different; the radiation varies rapidly and is overall much, much brighter than when the jets aren’t directed toward us.  When a radio galaxy’s jets are directed toward us, we call it a blazar.

Radio galaxies can be further divided into two classes: FRI and FRII.  FRII jets (pictured below) are much more powerful and more collimated.  At the end of their jet, the material is hitting the extra-galactic medium (gas outside of galaxies) and heating up, producing a bright hot-spot.  On the other hand, FRI jets (pictured above) are simply less powerful and collimated.  You might think that FRI jets and FRII jets are just two ends of a spectrum, but that’s not the case.  Most jets clearly fall in one group or the other, with a clear dividing line between them.

As you might expect, there are also two types of blazars: BL Lacs and Radio Quasars.  A lot of work has been done to see which type of radio galaxies produce these blazars.  The emerging picture is that BL Lacs are FRI jets pointing at us, and Radio Quasars are FRII jets pointing at us.  Furthermore, there are some good reasons that come from physics that this should be the case.

Unfortunately, recent observations cited in the paper I read suggest that around a third of all BL Lacs are actually FRII galaxies pointing at us, not the expected FRI galaxies pointing at us!

Shall we just throw this whole theory of radio galaxies/blazars out the window now?  Not yet is my best answer.  The theory still explains a lot of concrete observations in terms of straightforward physics.  Furthermore, we don’t have any better theory to replace it with yet.  We might find some evidence, theoretical or observational, that will allow the current theory to remain intact.  What’s more likely, in my opinion, is that we will find more theoretical/observational evidence that suggests the original theory was on the right track but just too simple.

This sort of situation arises all the time in science: a dominant theory explains a wide variety of phenomena in terms of a few physical principles, yet problems remain.  Many times these problems are solved, but occasionally these problems persist, leading to a modification or collapse of the dominant theory.

Some scientists latch on to these problems, despite the supporting evidence, and hold them up as examples for why the scientific community is biased towards the dominant theory.  Critics of the Big Bang, evolution, and human-induced climate change, for example, reject the dominant theory but propose no better theory of their own.

So when do we reject a theory?  First, we have to understand the dominant theory well conceptually (saying that you can’t understand it doesn’t cut it).   We must also understand the entire body of supporting evidence.  Secondly there must be excellent data–which has withstood the test of time and several analyses–that, if true, completely undermines the theory.  Finally, a new theory must be in place that fully explains both the data that undermined the old theory and the body of supporting evidence of the old theory.

Some of the most powerful explosions in the Universe: gamma ray bursts

I just attended an awesome talk given by Dr. Niccolo Bucciantini about one of the most puzzling astronomical phenomena out there: gamma ray bursts. These bursts are just pulses of photons that last from 1 second to ~1000 seconds, and the photons are at extremely high energies. All in all these bursts are some of the most luminous objects in the universe. Yet the puzzling thing is that the bursts are very dissimilar, some are just one simple pulse, others have a variety of flares before they die out. The classic joke is, “if you’ve seen one gamma ray burst, … then you’ve seen one gamma ray burst.” We can at least create two categories: long bursts (100-1000 sec.) and short bursts (1-2 sec). The talk I attended concerns long bursts.

The colossal power and compactness of these objects actually puts some serious constraints on what could be powering them. To date, compact objects (either neutron stars or black holes) are the most efficient way to produce prodigious amounts of energy. As it turns out, many models of gamma ray bursts start with a star in its death throes whose core collapses into a compact object. Then, the rest of the star falls onto the compact object, which then produces copious amounts of the energy. In fact, around 10% of the rest mass energy of the in-falling matter is converted into light (less than 1% of the rest mass energy in converted to energy in nuclear fusion (what powers the stars).

Another interesting property of gamma ray bursts can be deduced from the short variability (in time) in the amounts of energy we detect. Even though compact objects, are, well…compact, it turns out they are just too big to be the causes of such short variability. As Bucciantini put it, “Shaq can’t move as fast as the smaller point guards.” Thus, the light we see is being emitted from a small region, which I’ll call a blob. Unfortunately this causes some serious problems. If you have a bunch of highly energetic photons in a small blob, they will not escape! They will slam into each other, producing an electron and a positron. Thus we won’t see any photons because all of them will produce electron-positron pairs. The solution to this problem is actually quite simple: we can say this blob is moving towards Earth at a speed very close to the speed of light. Then all these photons that we see as being so energetic are only energetic because they were produced in a blob that had a lot energy (due to its motion) in the first place! So if you’re riding along with the blob then the photons aren’t as energetic, meaning they wouldn’t be able to produce electron-positron pairs. This then allows for the highly energetic photons to be produced in a small blob and for them to be able to escape without slamming into each other and metamorphosing into electron-positron pairs. (I’m leaving out a few details. Look here for the whole story of what’s called “relativistic beaming.”) In conclusion, gamma ray bursts are supernova that send out a highly relativistic jet (meaning ~99% of the speed of light) that is pointed right at Earth.

Bucciantini and collaborators are working on a model that has a unique type of neutron star, a magnetar as the “engine” which powers these bursts. A magnetar is just a neutron star with a huge magnetic field (14 orders of magnitude stronger than Earth’s magnetic field). Its field is like Earth’s in that the true “north” pole (or spin axis) does not line up with the magnetic pole. This misalignment coupled with a neutron’s stars extremly fast spin, causes the star to emit lots of light. This light, which is electromagnetic energy, comes directly from the neutron star’s spinning energy. Therefore, by conservation of energy the magnetar actually spins down due to all the electromagnetic energy (light) produced (this is known as magnetic breaking).

What’s so great about all this magnetar stuff? As it turns out we can measure the spinning rate of normal neutron stars that don’t spin down like this, and calculate the amount of “spinning energy.” It turns out this energy is just the amount needed to power a typical long gamma ray burst! Also, when you calculate how long it takes for a typical magnetar to spin down, it takes 100-1000 seconds, precisely the duration of a long gamma ray burst!

The next really exciting part of Bucciantini et al’s model is that the magnetic field in the presence of the collapsing star seems to be able to produce the jet that the short variability arguments support!

All in all, it was an excellent talk about a very interesting and puzzling subject in astrophysics. Kudos to Bucciantini.