Spiral galaxies rotate too fast. If they would only consist of the visible (baryonic) mass we observe in them and Newton’s Law of gravity is correct, then they would not be stable and should quickly fly apart. That they don’t has been one of the first indications that the galaxies (and the Universe as a whole) either contains large amounts of additional but invisible “dark matter”, or that the laws of gravity don’t hold on the scales of galaxies. One possibility for the latter, Modified Newtonian Dynamics (MOND), proposes that gravity needs to be stronger in the low acceleration regime present in galaxies (for more details see the extensive review by Famaey & McGaugh 2012 and Milgrom’s Scholarpedia article). That the rotation curve (i.e. the function of circular velocity of the galactic disc with radius) of our Milky Way galaxy follows the same trend as the rotation curves of other spiral galaxies has been known for a long time, too. So it appears to be a bit surprising that the Nature Physics study “Evidence for dark matter in the inner Milky Way” by Fabio Iocco, Miguel Pato and Gianfranco Bertone makes such a splash in the international press. That the MW should contain dark matter is not news, but nevertheless the paper got a huge amount of press coverage.
One thing emphasized a lot by the press articles (and press releases) is that the authors claim to have found proof for the presence of dark matter in the ‘core‘, ‘innermost region‘, or even ‘heart of our Galaxy‘1, not just in the intermediate and outer regions. This might be worrisome for modified gravity theories like MOND, which predict that regions very close to the center of the Milky Way should be in the classical Newtonian regime, i.e. the rotation curve should be consistent with that predicted by applying Newton’s law to the observed mass distribution. The underlying reason is that due to the higher density of baryonic matter in the center of the Milky Way the gravitational acceleration of the baryons there already exceeds the low-acceleration limit. But only once the acceleration drops below a certain threshold the non-Newtonian gravity effect kicks in. Interpreted naively (i.e. assuming Newtonian dynamics), this would mimic dark matter appearing only beyond a certain radial distance from the Galactic Center.
Without even going into the details of checking their assumed Milky Way models, the way the observational data is combined and whether there are systematic effects, a simple look at figure 2 in Iocco et al. already reveals that their strong claim unfortunately is not as well substantiated as I would wish.
The plot’s upper panel is what is of interest here. It shows the angular circular velocity in the Milky Way disk versus the Galactocentric radius. The red points with error bars are observed data for different tracers. The grey band is the range of velocities allowed for the range of baryonic mass distributions in the Milky Way considered by Iocco et al. (that are all consistent with observations). If there would be only baryonic matter and Newtonian Dynamics, the rotation curve of the Milky Way should lie somewhere in this area.
First of all, the figure shows that they did not consider any data in the region within 2.5 kpc. That makes sense because that region will be dominated by the bar and bulge of the Milky Way. Stars in the bulge don’t follow circular orbits, so one can’t measure circular velocities there.
So, what is the core, heart or ‘innermost region’ of the Milky Way? Lets try to come up with something motivated by the structure of our Galaxy. The Galactic disk is often modeled by an exponential profile, with a scale length of about 2.2 kpc. What if we say the core of the MW is everything within one scale length? Immediately there’s a problem with the claim by Iocco: They are not even testing data on this scale.
Lets ignore the phrase ‘core’ or ‘heart’ of the Milky Way and focus on the more general formulation they also use in their paper’s title: “Evidence for dark matter in the inner Milky Way”. Looking at their Figure again, we can see that the data start to leave the grey band at a distance of about 6 kpc from the MW center. Thus, within 6 kpc (almost three scale radii of the Milky Way disk!) the purely baryonic models encompass the data. Consequently, here is no need to postulate that dark matter contributes significantly to the dynamics. The figure clearly shows that there is no need, and therefore no evidence for dark matter within 6 kpc of the Galactic Center, which is as generous a definition of ‘inner Milky Way’ as it gets in my opinion. The authors themselves even write that ‘The discrepancy between observations and the expected contribution from baryons is evident above Galactocentric radii of 6-7 kpc’. In this regard it doesn’t matter whether the majority of the possible baryonic models predict a lower rotation curve: as long as the data agree with at least one baryonic model that is consistent with the observed distribution of mass in the Milky Way, there can not be evidence for dark matter.
I really don’t understand why they then claim to have found proof of dark matter in the innermost regions of the Milky Way. My suspicion is that the authors and their press releases seem to have a (literally) quite broad interpretation of the term ‘innermost region’. Judging from the context, they seem to subsume everything within the solar circle of ~ 8 kpc (the distance of the Sun from the Galactic Center) as ‘innermost’. I don’t think it is an appropriate definition, after all it makes the vast majority of the baryonic mass of the Milky Way part of the innermost region. Half the light of an exponential disk is already contained within less than 1.7 scale length (1.7 x 2.2 kpc = 3.7 kpc for the Milky Way), and all of the bulge/bar is in there, too. But if we nevertheless roll with it for the moment we can see that yes, between 7 and 8 kpc there seems to be need for dark matter … or for a MOND-like effect.
So, lets have a look at one MOND rotation curve constructed for the Milky Way (from McGaugh 2008) to see where we expect to find a difference in Newtonian and MONDian circular velocities. The expected Newtonian rotation curve is shown as a black line in the plot, equivalent to the purely baryonic rotation curves making up the grey band in the figure of Iocco et al.. The rotation curve predicted by MOND is shown as a magenta line and the observed circular velocities are the small squares.
The plot immediately reveals that a discrepancy between the Newtonian and the MONDian rotation curves is expected already at small radii, well within 6 kpc. The findings of Iocco et al. that there appears to be some mass missing within the solar circle therefore do not disagree with the MONDian expectation, in contrast to what one of the authors is quoted saying in a Spektrum article. Furthermore, the plot demonstrates that the need for dark matter (or MOND) in the region inside the solar circle was already well known before this new study.
So, in summary, the study doesn’t show all that much new or surprising, the claimed ‘evidence’ for dark matter in the innermost Milky Way is not present in their data (unless you define ‘innermost’ very generously) and some apparent dark matter contribution within the solar circle is not even unexpected based on MOND predictions.
1: The press releases of the TU Munich and Stockholm University even call it a ‘direct observational proof of the presence of dark matter in the innermost part our Galaxy’ (which is clearly wrong, there is obviously nothing direct about it and the innermost part would imply the very center of the Milky Way).
See the overview of topics in The Dark Matter Crisis.