Friday, December 18, 2009

When two might mean nothing, or it might be the future of physics

Yesterday saw the culmination of weeks of building buzz surrounding the announcement of results from one (of many!) searches for dark matter in the laboratory.  The result: two "events" that look like dark matter, but there's a 23% chance that these are just background events.  In science, 23% chance of being wrong is too high, so most people are seeing this as, at best, an ambiguous result.  Read more about the announcement here, or take it straight from the team's mouth (in PDF format).

What was all the excitement about?  Let me explain.  No, there is too much.  Let me sum up. Buttercup is marry Humperdink in little more than half an hour. (Sorry).  The story begins in the 1930s, when astronomer Fritz Zwicky noticed that galaxies in galaxy clusters were moving faster than could be explained by the gravity of visible stars, so he proposed some sort of invisible matter was responsible.  Most people (wrongly!) dismissed this idea until the 1970s and 1980s, when astronomer Vera Rubin discovered that the outer parts of spiral galaxies were orbiting faster than could be explained by the visible stars and gas alone.  About the same time, X-ray telescopes confirmed that Zwicky's galaxy clusters were more massive than the visible material alone could explain.  Meanwhile, comparison of the Big Bang theory's predictions of the amounts of different elements produced during the Big Bang with the actual amounts we see in the Universe said that much of this dark matter could not be made of normal atoms.  Skipping over many equally-important observations, the summary is that astronomy needs some sort of matter in the Universe that is different from the atoms stars, planets, and people are made out of.

Around the same time, particle physicists were exploring possible extensions of the Standard Model, a theory explaining the relationships between all the normal particles in nature.  There has been some evidence that the Standard Model is not the end of physics.  Neutrinos, ghostly particles that can pass through a light year of lead without hitting anything, were found to have mass, while in the Standard Model they had been assumed to be massless.  The Standard Model also does not explain where particles come from in the first place, and it doesn't explain why the universe has much more matter than antimatter.  So, theorists started speculating and calculating additional theories.  Many of these naturally produce a particle that doesn't interact with normal matter except by gravity and very, very rare collisions.  By a happy coincidence (or maybe because the physicists were being guided by the observations), the predicted properties of most of these new particles match the properties of dark matter inferred by astronomers.

This is all fine and good, but the only way to prove these theories is to actually find dark matter in the laboratory.  This is a huge challenge, because dark matter only very rarely interacts with real matter, and there are many other things that interact with particle physics detectors in the same way as dark matter, including cosmic rays, radioactive decays of atoms in the Earth, or even neutrinos from nuclear power plants.  Not only is looking for dark matter candidates searching for a needle in a haystack, but it is searching for a needle in a haystack made of needle-shaped, shiny gray hay.

Many groups are looking for dark matter in laboratory experiments, and they are all working to figure out how to limit this background.  You can put it deep underground and shield it with lead from ancient sailing ships (so the radioactive elements have decayed).  You can cool it really cold.  You can put all kinds of fancy detectors.  This is amazingly hard, and lots of smart people are working very dilligently on it.  If you are the first to unambigously detect dark matter, you'll probably get a Nobel Prize.  Or, if you can convince the world that you should have detected dark matter but you didn't, you'll get a Nobel Prize and poke a big hole in our understanding of the Universe.  Either one of these would be fascinating results.

So, a few weeks ago, one of the teams looking for dark matter in a mine shaft deep under Minnesota, the Cryogenic Dark Matter Search II (CDMS-II), suddenly got very quiet.  They scheduled two simultaneous talks for yesterday.  They wouldn't talk to the press.  Rumors started swirling that they had detected dark matter.  Rumors were made (and denied) that a paper announcing discoveries had been accepted by Nature, perhaps the most prestigious journal in worldwide science.

I watched the CDMS II talks yesterday over streaming video, at least until the video froze RIGHT AS THEY WERE ANNOUNCING FINAL RESULTS.  Grr.  But thanks to the live-blogging at Cosmic Variance, I learned that the result was: two possible detections, with an expected false alarm rate of about 1 detection.  Because of the way that probability and statistics works, detecting two false positives when expecting one is not that rare. The quoted statistics are that there is a 23% chance that these detections were both false alarms.  As I said at the start, 23% is not good enough to claim anything definitive.  We prefer numbers smaller than 1%  and preferably much, much smaller.  Five detections would have met this criteria, but two does not.

The CDMS II team is now working on a bigger, more sensitive version of their experiment.  And other teams will be scrambling to try and get their dark matter detection experiments working all the faster.  This is well and good, as competition speeds the results.  I am disappointed that the hype was allowed to build over an ambiguous result.   A clear detection, or even a clear non-detection, would have been big news and worthy of the hype.  This was a very interesting result, and showed that experiments are now good enough to detect dark matter in the near future, but it was not the Earth-shaking result that the preparations suggested.  I don't know who is responsible for the hype, or if it just grew unexpectedly and nobody knew what to do about it, but it makes me grumpy.

Congrats to the CDMS II team on an awesome technological feat.  Now go prove or deny that dark matter exists.  That's what we all really want to know.

Note: This was all about dark matter, not dark energy, which is even weirder and may not be related to dark matter.  But it might be.


  1. Interesting. Would like to link to the article from let us know :)

  2. Sure! Anyone can link to any of my posts. I appreciate the traffic.