Back in April, a study of the motion of hundreds of stars in the Milky Way found no evidence of a massive dark-matter halo (CERN Courier June 2012 p11). The finding came as a surprise and did not long withstand the assault of sceptical scientists questioning the results. A new study based on the same data set, but proposing a different underlying assumption, now reconciles the observations with the presence of a dark-matter halo in line with expectations.
One of the first pieces of evidence for dark matter was that the rotation velocity of stars in the Milky Way remains constant instead of decreasing with distance from the Galactic centre. This flat rotation curve implies the presence of an extended distribution of dark matter, whose mass compensates the decreasing stellar density in the outer regions of the Galaxy. The presence of a similar dark-matter halo is implied by the flat rotation curve observed in almost every spiral galaxy but its actual shape and density distribution is difficult to predict.
To determine the amount of dark matter in the vicinity of the Sun, a team of Chilean astronomers measured the motions of more than 400 red giant stars up to 13,000 light-years from the Sun, in a volume that is four times larger than ever previously considered. Visible matter in the form of stars and gas is dominant in the plane of the Galaxy but at higher elevation above the Galactic disc, dark matter should dominate. The rotational velocity of stars at different Galactic heights should thus result in a measure of the local density of dark matter in the solar neighbourhood.
To their surprise, Christian Moni Bidin of the Universidad de Concepción and colleagues found no evidence at all for a dark-matter halo. They obtained an upper limit of 0.07 kg of dark matter in a volume the size of the Earth, whereas theories predict a mass in the range of 0.4–1.0 kg. This difference of about an order of magnitude led some astronomers to query the validity of the analysis.
Jo Bovy and Scott Tremaine of the Institute for Advanced Study, Princeton, claim that they found a fault in one of the assumptions made by Moni Bidin and colleagues. The problematic assumption is that the average rotational velocity <V> is constant with distance from the Galactic centre at all heights above the plane of the Galaxy. For Bovy and Tremaine, this assumption applies to the circular velocity Vc but not to <V>. The difference is rather subtle, but it is a well identified effect known as the “asymmetric drift”, which arises from a sub-population of stars with elliptical orbits that have on average a lower velocity than Vc. The result is a difference between <V> and Vc that evolves with the height above the Galactic plane and would have led the Chilean researchers to underestimate the density of dark matter.
With their modified assumption that the circular velocity curve is flat in the mid-plane, Bovy and Tremaine obtain a local dark-matter density of 0.3±0.1 GeV/cm3, fully consistent with estimates from the usual models. They also claim to demonstrate that this assumption is motivated by observations, while the previous one was implausible.
As with the OPERA result on the faster-than-light neutrinos (EXO, MINOS and OPERA reveal new results), this is another example of an unexpected result being later disproved. It seems that submitting the problem to the scientific community in the form of a paper is an efficient way to identify quickly the origin of the disagreement. The strength of the scientific community as a whole is to be able to solve major issues more effectively than a single research group.