A long-standing and controversial claim by the DAMA collaboration in Italy that it has observed dark matter has received fresh support from a US-based experiment. Like DAMA, the CoGeNT collaboration says that it continues to see a seasonal variation in the number of events registered in its detector. Such a variation would be expected if the Milky Way galaxy were shrouded in a "halo" of dark matter, but several other dark-matter searches have failed to see the effect.

Dark matter is thought to make up at least 80% of matter in the universe, yet it interacts only very weakly with ordinary atoms and molecules. While physicists can see its gravitational effects on large objects such as galaxies, dark-matter particles themselves have proved very elusive.

The DAMA detector is located deep underground at the Gran Sasso National Laboratory, where in 1998 the collaboration first claimed to have detected dark matter. Since then, the result has been reinforced with ever more data from the detector. The experiment is unlike most dark-matter searches, which look for individual collisions between incoming dark-matter particles and detector nuclei. Instead, DAMA looks for – and has found – an annual modulation in all of the events registered by its detector. The modulation is expected to arise because the solar system is believed to be ploughing through a dark-matter halo enveloping the Milky Way. In the northern summer, the tangential velocity of the Earth as it orbits the Sun is in the same direction as the motion of the solar system. As a result, the number of collisions detected by DAMA should peak in summer and drop in the winter.

Surprise confirmation

Other physicists do not dispute DAMA's observation. However, some argue that the group, led by Rita Bernabei of University of Rome Tor Vergata, has not done enough to rule out the more mundane causes of that modulation. Juan Collar of the University of Chicago and colleagues built CoGeNT at the Soudan Underground Laboratory in Minnesota specifically to try and settle the issue. The team uses about 100 g of germanium as its detector medium. Although this produces far fewer data than DAMA, which uses 250 kg of sodium iodide, CoGeNT has a much improved capacity to detect the low-mass dark-matter particles capable of explaining the modulation. After running for 15 months, Collar and co-workers announced in 2011 that CoGeNT had, to their surprise, also seen the annual modulation in collisions.

Since then, CoGeNT has gathered more data and the team has just released an analysis of data collected up to April 2013. The study involves slicing those data up in two ways: by collision energy and by location within the detector. The idea is that for any annual modulation to be caused by weakly interacting massive particles (WIMPs) – the preferred candidate for dark matter – the modulation should only be seen in events taking place inside the bulk of the detector and having energies of less than about 2 keV. That was exactly what the researchers found.

Collar describes his group's data analysis as "very basic", calculating the statistical significance of the result to be a modest 2.2σ. This means that there is about a 2% chance of it being a statistical fluke. He argues, however, that the result becomes more interesting when compared with those of other groups, such as DAMA and CRESST, which is also based at Gran Sasso. The different results, he says, yield conflicting WIMP masses and interaction strengths only if WIMPs are assumed to move about randomly inside the galactic halo. However, if, as the latest results imply, some WIMPs behave differently – by streaming in a certain direction, for example – then the different sets of results overlap quite neatly, he explains.

Constraining the possibilities

Collar is careful not to overstate the importance of his group's work, describing the new results as "not evidence" of dark matter but as "constraining the possibilities" of dark matter's identity. He points out that his group's annual modulation might yet be explained away by a currently unknown source of systematic error.

The CDMS collaboration, whose detector is also located in Soudan, has produced its own results with silicon detectors that were reported last year – three events that look like collisions of WIMPs with about the same low mass as those DAMA says it has seen – but with germanium detectors has found no sign of an annual modulation. Meanwhile, two xenon-based experiments – XENON at Gran Sasso and LUX at the Sanford underground lab in South Dakota in the US – have seen nothing at all.

Confusing and messy picture

Dan Hooper, a particle theorist at Fermilab near Chicago, says that the conflicting results present a "confusing and messy" picture, arguing that "it is hard to understand why LUX didn't see more events than it has", if the other experiments really are seeing WIMPs. However, it is possible, he points out, that LUX is simply not that sensitive to low-mass WIMPs. "Until the low-energy calibrations of liquid xenon are made in the presence of an electric field," he says, "there is still a chance that some of these experiments might be seeing dark-matter particles."

According to its co-spokesman Richard Gaitskell of Brown University in the US, LUX has published calibrations with an electric field that go below collision energies of 3 keV. Gaitskell says that "multiple experiments" are now working on calibrations at even lower energies and should make their results available in the first half of 2014.

In fact, Collar and a number of other physicists have built a small xenon detector at Fermilab in order to scrutinize LUX's claimed exclusions. The detector, which uses a yttrium–beryllium neutron source to mimic the recoil of low-mass WIMPs, should produce results within a few weeks.

The next step for Collar will then be the assembly of a 4 kg successor to CoGeNT, called C4, which he hopes to have up and running by the end of 2014. He is also looking forward to results from the GAIA satellite, which was recently launched by the European Space Agency. Designed to chart the position and velocity of about 1% of the stars in the Milky Way, GAIA might, he says, reveal the existence of stars travelling towards Earth and so, in turn, provide independent confirmation of any streaming dark matter.

The results are described on arXiv.

More about the nature of dark matter can be gleaned from this video by Luke Davies of the University of Bristol.