Dark Matter is Just Messing With Us Now

The state of play in dark matter searches just refuses to settle down. Just a few weeks ago, the XENON100 experiment released the best-yet limits on WIMP dark matter (a two-dimensional parameter space, "mass of the dark matter particle" and "cross section with ordinary matter"). These limits seemed to firmly exclude the hints of a signal that had been trickling in from other experiments. But... the story isn't over yet. Remember that XENON, like CDMS and other experiments, tries to find dark matter by making a very quiet experiment and picking out individual events where a dark matter particle bumps into a nucleus inside the detector. There is a complementary strategy, looking for annual modulations in the dark matter signal: rather than being very picky about what event is and is not a DM interaction, just take lots of events and look for tiny changes in the rate as the Earth moves around the Sun. Dark matter is like an atmosphere through which we are moving; when we're moving into a headwind, the rate of interactions should be slightly higher than when our relative speed through the ambient dark matter is smaller. The DAMA experiment was designed to look for such a modulation, and it certainly sees one. The problem is that lots of things modulate on a one-year timescale; as Juan Collar explained in a guest post here, there were many questions about whether what DAMA is detecting is really dark matter. Now one of Juan's own experiments, CoGeNT, has seen (very tentative) hints of an annual modulation itself! CoGeNT had already teased us with a hint of a dark matter signal, which (like DAMA) seemed to imply lower masses (about 10 GeV, where 1 GeV is the mass of a proton) rather than the usual masses for weakly-interacting dark matter favored by theorists (hundreds of GeV). But the competitor experiment CDMS, and later of course XENON, seemed to put the kabosh on those claims. The CDMS result was especially hurtful to CoGeNT's claims, as both experiments use germanium as their detector material. Theorists are very clever at inventing models in which dark matter interacts with one substance but not some other substance (see e.g.), but it's harder to invent models where dark matter interacts with one substance in one experiment but not the same substance in some other experiment. Yesterday Juan Collar gave a talk at the April Meeting of the APS, where he revealed something about CoGeNT's latest findings. (I don't think there's a paper yet, but it's supposed to come very soon, and they are promising to share their data with anyone who asks.) Now, unlike for their earlier results, they are explicitly looking for annual modulation. And ... they see it. Maybe. Well, not really enough to take it seriously, but enough to be intrigued. Or, in science-speak: it's a 2.8 sigma result. It doesn't seem to have hit the news very hard, but there are writeups by Valerie Jamieson and David Harris. The CoGeNT folks have 442 days of data, with a rate of about three events per day. Ordinarily, a tasteful physicist would claim that a 2.8 sigma result doesn't even rise to the level of "intriguing"; you need three sigma to count as "evidence," and five sigma for "discovery," by the accepted standards of the field. The reason this is even blogworthy (a low bar indeed) is that it's the first attempt to check DAMA by looking for an annual modulation signal, and the result matches the phase of DAMA's oscillation, and is claimed to be consistent with its amplitude (the experiments use different materials, so it's hard to do a direct comparison). Also, of course, because the team was looking to bury DAMA, not to praise it: "We tried like everyone else to shut down DAMA, but what happened was slightly different." On the other hand, what you would need to explain this purported signal is at first glance still very much incompatible with XENON's limits. In the end: probably still nothing to get too excited about. But at least it will keep the pot boiling a while longer. Don't fear; the experiments are getting better and better, and temporary confusions eventually evaporate. Or are swept away by the dark matter wind.