Near infrared (NIR) photometry
Introduction
Polar ring (PR) galaxies are considered one of the major keys to infer the shape of dark matter halos (e. g. Schweizer et al. 1983, Whitmore et al. 1987, hereafter WMS). Even if a large number of polar ring candidates and related objects do exist (see the Polar Ring Catalogue-PRC, Whitmore et al. 1990), the “kinematically confirmed” polar rings, i.e. systems for which the rotation of the ring in a polar plane is observed, are rare 10 objects). Among them, the prototype polar ring galaxy NGC 4650a has motivated numerous efforts (Laustsen & West 1980, Schechter et al. 1984, WMS, Sackett & Sparke 1990 (SS), Sackett et al. 1994 (S94)), to put constraints on the 3-dimensional shape of its dark matter halo. So far the results have been contradictory:
Send ojfirint requests to: F. Combes
WMS concluded that the dark matter halo was spherical, while SS, in a more sophisticated study, showed that dark halos with flattening from E0 to E8 where compatible with the observed kinematics. Their best fit was given by an E5 flattened halo, and the spherical case gave an extremely poor fit. As already outlined by Schweizer et al. (1983) and SS the fact that most polar rings are so nearly polar seems to imply that dark matter, if present in a significant amount, is not spherical. Recently S94 proposed a dynamical model for the polar ring system NGC 4650a using new optical observations of the S0 radial velocities and velocity dispersion, which rules out the spherical halo option. Their best fit model is given by a dark halo with an E6 to E7 flattening, with its axes aligned with those of the SO disk, and inside 10 kpc, the amount of dark matter is about 2.2 times the visible mass. Arnaboldi et al. ( 1993a) also favoured a flattened dark matter halo for the only polar ring known around an elliptical: AM2020-504. The best fit occurs for a flattening equal to that of the projected light.
The contradictions in the various dynamical models may be caused by a certain number of intrinsic difficulties. As emphasised by Reshetnikov & Combes (1994), dark matter is known to dominate the gravitational potential well outside the optical radius of galaxies, and the direct determination of the 3-D shape distribution would require the measurement of radial velocities onto two perpendicular planes at larger distances than the galaxy optical radius. This is impossible since two perpendicular gaseous distributions cannot co-exist, because of cloud cloud collisions and dissipation. Therefore polar ring dynamics may give only an indirect measurement of the dark halo tening, since it is based on the extrapolation of the inner mass model, dominated by the luminous component, into the outer regions, where the gravitational potential is sampled only by the HI present in the polar ring. A good knowledge of the potential in these two regions is further hampered by the nearly edge-on orientations of the tracers, introducing an un-welcome coupling of density and velocity distributions along the line of sight.
In this work, we are revisiting the question of the dark halo shape in the prototype polar-ring galaxy NGC 4650a. We present in Sect. 2 new near~infrared (NIR) photometry that gives more
© European Southern Observatory ' Provided by the NASA Astrophysics Data System
F. Combes & M. Amaboldi: The dark halo of polar-ring galaxy NGC 4650a
| | | | 100 150 200
E300
Fig. 1. Near-infrared image of NGC 4650a: in the NIR the dominating component is the central S0, much brighter than the polar ring. East is
up, North on the right
details on the stellar distribution in the central SO and in the polar ring, and helps to select their mass-to-light ratios. In Sect. 3 we propose a new kinematical model which accounts for the observed radial velocities and velocity dispersion along the S0 major axis using only the visible mass, and a constant M/LB of 4. We discuss the main uncertainties at every step of the modelling, mainly the density-velocity coupling along the line of sight, and emphasise the non-uniqueness of any solutions. In Sect. 4 we present a more detailed model in which the ellipticity of the orbits is accounted for, and our best fit, where the dark matter has an oblate distribution with the same axes as the polar ring itself. In Sect. 5 the results and the uncertainties of the modelling are discussed, and the conclusion are derived.
Near infrared (NIR) photometry
Near infrared Kn band images of NGC 4650a were obtained at the 4m. Anglo-Australian telescope on the of April 1994, with the IRIS infrared camera at the f / 15 Cassegrain focus. The detector is a hybrid array of 128 128 pixels: our image scale is 0.61”/pix. The Kn filter was chosen for its high sensitivity and because it greatly reduces the effect of water vapour. Several
standard stars from the “IRIS Users’ Guide” were observed during the night to transform the images into the standard Kband system. Since the polar ring in NGC 4650a subtends an angle on the sky bigger than 1', a mosaic of 3 fields was required to image the whole system: total integration time for each field is 450 seconds. The images for each field were acquired using the offset technique: a cycle was defined so that 5 sub-images were taken together with sky frames (before and after each exposure on the galaxy), and bias frames (at the beginning of each cycle in order to linearise the CCD frames). Linearisation and bias subtraction were done using the FIGARO task irislin at the end of each cycle. Flatfielding was done using the “light-on, lightoff” dome flats, after differencing and normalising them. Each sub-frame was sky subtracted using an averaged sky, derived from the sky frames taken before and after it; the resulting image for each field of the mosaic was finally derived by registering and co-adding all subframes (see Fig. 1). The luminosity profiles were extracted along the position angles of the PR major axis and the host galaxy major axis (as listed in the PRC).
© European Southern Observatory ' Provided by the NASA Astrophysics Data System
F. Combes & M. Arnaboldiz The dark halo of polar-ring galaxy NGC 4650a
Modelling