A high-profile claim to have seen a potential signal of dark matter passing through the Earth is encountering some skepticism as independent scientists get to grips with the published data.
On 3 June, Juan Collar of the University of Chicago, and his colleagues on the CoGeNT experiment in the Soudan mine in Minnesota published data that they claimed was consistent with the signature of dark matter particles passing through their germanium detectors.
Although the group stopped short of claiming the detection of dark matter, its statement that its results were consistent with that interpretation drew intense media attention, beginning in May when Collar reported the findings at the American Physical Society meeting in Anaheim, California. This was primarily because of a tantalizing modulation – a seasonal rise and fall – in the frequency of their candidate detections, that seemed consistent with a previously unreproduced claim to have detected dark matter by the DAMA experiment at Gran Sasso in Italy.
Now, in a paper submitted to the arXiv preprint server, Thomas Schwetz-Mangold of the Max Planck Institute for Nuclear Physics in Heidelberg, Germany, and Jure Zupan of the University of Cincinnati in Ohio, agree that there is a statistically significant modulation in the CoGeNT data, but question whether it is consistent with the dark matter explanation. “I’m not convinced it’s coming from dark matter,” says Schwetz-Mangold, “there is some tension with that explanation.”
The issue, Schwetz-Mangold says, is that the best fit to the modulation observed by CoGeNT is not consistent with the overall spectrum of events as a function of energy. Specifically, the amplitude of the modulation is actually stronger than would be expected if a dark matter particle producing the overall spectrum of events was really the cause.
Collar responds that Schwetz-Mangold does not appear to be working with the CoGeNT data. “He seems to be playing some basic games with the information in the plots in our preprint,” he says. Collar adds that it may be true some others are reading too much into the CoGeNT data and that its agreement with DAMA should not be overemphasized.
Schwetz-Mangold says Zupan and he did ask for and receive the data.
Dan Hooper, a theorist at Fermilab near Batavia, Illinois, also noted the excess in a paper posted online on 6 June. However, Hooper comments that the limited amount of data CoGeNT has taken so far could explain the discrepancy between the observed signal and that expected from a dark matter particle. “There’s a little bit of tension but it’s like flipping a coin and getting two heads,” he says. Hooper says he’d start to be concerned about the problem if the pattern continues once CoGeNT has collected about twice as much data. Even then, he says, a modified dark matter model might be able to account for it.
Gilles Gerbier of the Commissariat a l’Énergie Atomique’s Institut de Recherche sur les Lois Fondamentales de I’Univers in Saclay, France, who works on the EDELWEISS-II dark matter experiment, says he’s also concerned by the strength of the modulation reported by CoGeNT. He says the paper contains an apparent error in showing a predicted modulation with an amplitude greater than others are obtaining using a basic dark matter model, which makes the prediction’s fit to the data look better than it should. Collar did not respond to a request for comment on this. Hooper says in his understanding the prediction curves in the CoGeNT paper are not in error but have an amplitude that has been fit to the data.
Gerbier suspects that the modulated signal observed by CoGeNT is caused by background events, meaning it is due to chance. “I think they’re going too far in claiming it’s consistent with dark matter,” he says.
Update 30 June: Collar comments that there is no error in the CoGeNT paper, and that it will shortly be updated to clarify what the predicted modulation curve represents. His remarks about Schwetz-Mangold were based on a talk given on 15 June, not the forthcoming preprint.
Image: The Soudan mine in Minnesota.