The cored distribution of dark matter in spiral galaxies

TL;DR

This study tests dark matter halo shapes in five spiral galaxies by combining high-resolution Hα and HI rotation curves, derived with the new MET/WAMET method that accounts for warps and profile asymmetries. Model-data-cube comparisons validate the rotation curves and reveal that cored halos with core radii near the optical size best reproduce the observed kinematics, while cuspy NFW or Moore profiles are strongly disfavored. The results consistently show core densities around $\rho_0 \sim 10^{-24}$ g cm$^{-3}$ and challenge standard CDM expectations, with HI scaling and MOND providing inconsistent fits in several cases. The work demonstrates the critical role of warp-aware rotation-curve extraction and extended HI data in constraining dark matter distributions in galaxies.

Abstract

We present the HI data for 5 spiral galaxies that, along with their Halpha rotation curves, are used to derive the distribution of dark matter within these objects. A new method for extracting rotation curves from HI data cubes is presented; this takes into account the existence of a warp and minimises projection effects. The rotation curves obtained are tested by taking them as input to construct model data cubes that are compared to the observed ones: the agreement is excellent. On the contrary, the model data cubes built using rotation curves obtained with standard methods, such as the first-moment analysis, fail the test. The HI rotation curves agree well with the Halpha data, where they coexist. Moreover, the combined Halpha + HI rotation curves are smooth, symmetric and extended to large radii. The rotation curves are decomposed into stellar, gaseous and dark matter contributions and the inferred density distribution is compared to various mass distributions: dark haloes with a central density core, $Λ$ Cold Dark Matter ($Λ$CDM) haloes (NFW, Moore profiles), HI scaling and MOND. The observations point to haloes with constant density cores of size $r_{core} \sim r_{opt}$ and central densities scaling approximately as $ρ_0 \propto r_{core}^{-2/3}$. $Λ$CDM models (which predict a central cusp in the density profile) are in clear conflict with the data. HI scaling and MOND cannot account for the observed kinematics: we find some counter-examples.

Figures (15)

• Figure 1: Optical DSS images (greyscale) superimposed with the H i total intensity map (contours). The first contour is the "pseudo 3-$\sigma$" defined similarly to VS:01 and is equal to: $1.0 \times 10^{20} {\rm atom~ cm}^{-2}$ for ESO 116-G12; $1.1 \times 10^{20} {\rm atom~ cm}^{-2}$ for ESO 287-G13; $1.7 \times 10^{20} {\rm atom~ cm}^{-2}$ for ESO 79-G14; $1.5 \times 10^{20} {\rm atom~ cm}^{-2}$ for NGC 1090; $3.4 \times 10^{20} {\rm atom~ cm}^{-2}$ for NGC 7339; the remaining contours are (15, 30, 45,...) $\times~ \sigma$, except for NGC 1090 and NGC 7339 where they are (6, 12, 18,...) $\times~ \sigma$. The beam is shown in the upper left corner.
• Figure 2: Radial distribution of the neutral hydrogen surface density ($\Sigma_{\rm HI}$): the filled triangles denote the approaching side, the empty triangles the receding side and the filled circles the average. The solid line refers to the ratio ${\rm M}_{\rm HI}(<r)/r$, where ${\rm M}_{\rm HI}(<r)$ is the H i mass inside radius $r$.
• Figure 3: A typical spectrum of the galaxy ESO 79-G14 at an intermediate galactocentric distance. The arrows indicate the systemic velocity as well as the positions of the velocities derived from the MET/WAMET method, the first-moment analysis and by fitting a single Gaussian to the profiles.
• Figure 4: Model of the radial velocities in the plane of the galaxy ESO 79-G14 and the regions that would be intercepted by the beam in the edge-on case.
• Figure 5: Five representative observed channel maps (upper panels) with the correspondent channel maps of the model data cubes (lower ones). The heliocentric radial velocities are indicated above each plot. The central map has a velocity closest to systemic. The cross indicates the centre of the galaxy. Contours are $-4\sigma, -2\sigma, 2\sigma, 4\sigma$, then 10, 15, 20, 30, 50 mJy beam$^{-1}$ for the ATCA galaxies, and $-4\sigma, -2\sigma, 2\sigma, 4\sigma$, then 6, 10, 15, 20, 30, 50 mJy beam$^{-1}$ for the VLA galaxies. The beam is shown in the lower left corner.