Dark Matter in Galaxies



The existence of Dark Matter is still one of the big challenges in modern astrophysics. Since its early inference Jan Oort, who found the stars in the local neighbourhood to be moving somewhat fast as compared to the gravitational potential they should give rise to, and by Fritz Zwicky in 1933 who found that galaxies in clusters are apparently moving too fast, this phenomenon has accompanied astronomy and astrophysics ever since. The pioneering work by Vera Rubin and her collaborators on optical (Hα) galaxy rotation curves proved the prevailance of Dark Matter in galaxies, which was later on followed up by the radio (HI) observations, first systematically conducted by Albert Bosma. It was in the mid 90’s when the study of Dark Matter experienced kind of a revival: with the advent of numerical simulations of structure formation with Cold Dark Matter (CDM) à la Navarro, Frenk & White (1996) it became possible to juxtapose observations and theory for a critical evaluation – and this is still going on. When fitting an analytical density profile to the resulting density distributions, these cosmological simulations predict a density profile (the so-called “NFW profile”) of the form




Here, ρc is the critical density, rs the characteristic radius, and δc is a dimensionless constant. On large scales (clusters, superclusters, cosmological filaments), CDM simulations have proven very successful, while on smaller (galaxy) scales, they were faced with a number of problems, namely

Another density profile that appears to fit the observed rotation curves a lot better was worked out by Burkert (1995), who tried to find the best-fitting density law complying with the observed rotation curves of dwarf galaxies, which are known to be dominated by Dark-Matter throughout. The Burkert profile is an empirical law that resembles a pseudol-isothermal halo. In contrast to the CDM profiles, it has a central core and is characterized by the core radius r0 and by the central density ρ0.




The solution of these problems has been sought for in observations, which always suffer from artefacts. They could naturally also be due to the (unknown) properties of the Dark Matter (e.g. finite cross-sections to interactions with baryons, self-annihilation, etc.). It is at this point where the whole subject becomes (at last) exciting: if we are able to map the density profiles of the dark halos of galaxies with sufficient precision, we should be able to contribute significantly to disclose the nature of Dark Matter! This is one of the goals of our investigations. In an ongoing study of galaxy kinematics using HI and Hα rotation curves, we aim at exploring the central density distributions of the dark halos. We need sensitive measurements of the 21-cm HI line, which we achieve using the WSRT and the VLA in the northern and the ATCA in the southern sky. From the observed HI data cubes, i.e. the measured brightness temperature as a function of position and velocity, Tb(ξ,η,v), the rotation curve is derived and may be combined with optical (Hα) or interferometric CO measurements to achieve higher spatial resolution in the central regions. The observed rotation curve delivers a mass decomposition via




In case of a spheroidal distribution of the Dark Matter, the density profile of the dark halo may be obtained via




where the Poisson takes its most simple form. Our results so far favour pseudo-isothermal dark halos, rather than distributions predicted by CDM simulations. The example below shows measured rotation curves (Ha and HI combined) of three low-luminosity galaxies. In the upper row, Burkert halos have been fitted, while the lower one exhibits fits top NFW density profiles. The individual lines indicate the total rotational velocity (solid), the dark halo (long-dashed), the stellar (dotted), and the gaseous (short-dashed) component. It is obvious that the former yield a better fit. In particular, the NFW profiles require unrealistic mass-to-light ratios (taken from Gentile et al. 2004).




In the course of our studies, we also aimed at studying the outer peripheries of galaxies. Whatever the amount and property of the Dark Matter, galaxies should eventually exhibit a decline in their rotation curves. Alas, such a decline has not been encountered hitherto, in spite of sporadic claims. In order to derive reliable rotation curves, it is indispensable to deproject the observed velocities. This is an easy thing to do in case of flat disks; all one needs is the position angle and inclination of the galaxy under study. However, nature is more complicated. When studying galaxies out to large distances using the 21-cm line of neutral hydrogen one almost invariably encounters deviations from flat disks, such as warps or lopsidedness. A prominent example of such a warp is visible in outer HI disk of the spiral galaxy NGC5055, as shown in the following picture (Battaglia et al. 2006; see also Tom Osterloo’s web pages for spectacular examples of the extent of gaseous disks of galaxies). Even without the illustrating sketch, the warp is readily seen.




Our newly developed code tirific (Gyula Józsa) permits to investigate galaxies with warps or other deviations from a simple flat disk. This technique allows a reliable parametrization of the warp geometries by directly using the HI data cubes. We know three types of warps, which are illustrated below.






Our tools allows us to find the rules which warps 'obey to'. Warps usually commence at the galacto-centric radius where the stellar disk fades away. This is also coincident with a sudden decrease of the surface brightness of the HI emission, or mass density of the gaseous disk. Most notably, warps also occur in isolated galaxies, so that in these cases they cannot be tidally induced by companion galaxies. Our analyses of warps permit to disentangle seemingly complicated cases: strong warps with simple kinematics may mimic complex systems, such as for instance polar-ring galaxies. A possible and attractive explanation of the warp phenomenon is that we are dealing with two distinct dynamical systems, an inner disk governed by the gravitational potential of the baryons (which dominate there), and an outer disk obeying to the potential of a (flattened) dark halo. If the two systems have some misalignment of their angular  momenta, this gives immediate rise to a warp. We have recently illustrated our findings and conjectures in popular scientific form (Klein et al. 2005). Our relevant publications in professional journals may be found here. Some examples of warp analyses are shown below. The three rows show three galaxies each, viz. NGC2541, UGC3580 and NGC5204 (LTR). In the first row their HI distributions (contours) are superimposed onto optical (i'-band) images. This is followed by the kinematic and structural parameters. In the third row, we finally display an “opaque view” of their gaseous disks, resulting from the model cubes.






The close agreement of our models and observations may be demonstrated by the following movie, which shows the observed and fitted model data cube of NGC5204; one should watch this movie several times. The HI column density, represented in grey-scale, has iso-velocity contours superimposed. As the movie proceeds, one sees contours of HI brightness as a function of velocity (we are passing through the HI cube). Red contours represent the observations, while blue ones reflect the model cube. The correspondence of the two is obvious. The blue/red lines with crosses represent the line of nodes and its normal (kinematic major and minor axes).





We had furthermore started collaboration with both, theoretical astrophysics (numerical simulations, Andi Burkert) and particle physics (theoretical particle physics, Manuel Drees). The goal of this project was to mutually deliver pertinent information that can be used to constrain the parameter space of the Dark Matter candidates. An excellent and comprehensive review of the particle candidates of Dark Matter was published by Bertone, Hooper & Silk (2005). Such a collaboration among physicists and astrophysicists is particularly useful in view of LHC at CERN. This accelerator will doubtlessly contribute to this field by producing candidate particles for Dark Matter. Furthermore, there are quite a few experiments (being) set up to directly detect Dark-Matter particles via recoil, such as CDMS or XENON. Other experiments take an indirect route, by observing secondary (decay) products such as neutrinos (ANTARES, ICECUBE), or Čerenkov radiation produced by γ-rays as they pass through the earth’s atmosphere (e.g. H.E.S.S.).


Our research has seen financial support by the Deutsche Forschungsgemeinschaft.