Over the past couple days a paper on dark matter has been discussed in a number of new sites, such as the LA Times. The paper is being published in PRL but its preprint was posted to the arXiv in February.
The paper claims to see what looks like a line at 3.5 keV in the x-ray spectrum of the Andromeda (M31) galaxy. It’s been getting a lot of press coverage since it claims to see something that would be evidence for low mass dark matter in the form of sterile (right-handed) neutrinos with (in this case) a mass of about 7 keV. This mass is too low to be seen in most searches other than x-ray measurements. The proposed model is a sterile neutrino that decays to a neutrino and a photon, so the photon carries the energy for half the sterile neutrino mass. The lifetime would need to be very long in order to satisfy cosmological constraints, but the particles don’t need to be completely stable.
Despite all the positive coverage in the press, I haven’t heard much buzz from within the dark matter or even the broader HEP community. I think there are several reasons for this. Claims of finding evidence from dark matter from projects such as this – using methods from observational astronomy – seem to happen at least once a year and never pan out in the end. The statistical power of this result isn’t particularly strong – a point even mentioned in the abstract – so there’s no reason to get excited about a weak preliminary result when none of the many other weak positive results have led to the discovery of dark matter. Furthermore, if you actually look at the spectra (Figure 1 in the preprint), you see that the bump identified as a possible dark matter decay line is actually a slight deviation from a falling background spectrum. Conclusively finding a small bump on top of a large and nonconstant background can be extraordinarily difficult. Even a slight miscalibration or a slight difference between the background model and the true background can cause such a feature to appear in the residuals. If things like model uncertainty are taken into account – and it doesn’t look like that was done in this work – it’s quite possible that the apparent line is not just statistically weak but actually statistically negligible. This of course doesn’t mean that the signal can’t be dark matter. The paper is looking for a signal that is inherently difficult, maybe even impossible given our current knowledge about x-ray astrophysics, to definitively measure.