The isolated elliptical galaxy NGC 5812–MOND or dark matter?

Abstract

There exist isolated elliptical galaxies, whose dynamics can be modeled without resorting to dark matter or MOND, for example, NGC 7507. Such objects lack understanding within the current framework of galaxy formation. The isolated elliptical NGC 5812 is another object to investigate a possible role of isolation. We use globular clusters (GCs) and the galaxy light itself as dynamical tracers to constrain its mass profile. We employ Gemini/GMOS mask spectroscopy, apply the GMOS reduction procedures provided within IRAF, measure GC velocities by cross correlation methods and extract the line-of-sight kinematics of galaxy spectra using the tool pPXF. We identify 28 GCs with an outermost galactocentric distance of 20 kpc, for which velocities could be obtained. Furthermore, 16 spectra of the integrated galaxy light out to 6 kpc have been used to model the central kinematics. These spectra provide evidence for a disturbed velocity field, which is plausible given the disturbed morphology of the galaxy. We construct spherical Jeans models for the galaxy light and apply tracer mass estimators for the globular clusters. With the assumptions inherent to the mass estimators, MOND is compatible with the mass out to 20 kpc. However, a dark matter free galaxy is not excluded, given the uncertainties related to a possible nonsphericity and a possible nonequilibrium state. We find one globular cluster with an estimated mass of $1.6\times 10^7{M}_{\odot }$, the first Ultra Compact Dwarf in an isolated elliptical. We put NGC 5812 into the general context of dark matter or alternative ideas in elliptical galaxies. The case for a MONDian phenomenology also among early-type galaxies has become so strong that deviating cases appear astrophysically more interesting than agreements. The baryonic Tully Fisher relation (BTFR) as predicted by MOND is observed in some samples of early-type galaxies, in others not. However, in cases of galaxies that deviate from the MONDian prediction, data quality and data completeness are often problematic.