The Geometry of and Mass Accretion Rate through the Maser Accretion Disk in NGC 4258

Abstract
A maximum likelihood analysis of the NGC 4258 maser positions and velocities reveals a ~2 σ deviation from Keplerian motion in the projected rotation curve of the high-velocity features, corresponding to a ~9 km s-1, or 0.8%, flattening of the LOS velocities with respect to Keplerian motion over the range of the high-velocity masers. While there are a number of potential explanations for this flattening, we argue for pure Keplerian rotation in an inclination-warped disk on the basis of the ability of this model to explain a number of otherwise puzzling features of the system. A radial gradient in the disk inclination of 0.034 mas-1 is not only consistent with the observed rotation curve, but it generates a bowl along the near edge of the disk that naturally explains the otherwise puzzling narrow spread in the declinations of the systemic masers. It also explains the existence and location of an apparently recurring flare among the systemic masers. There is no significant evidence for non-Keplerian rotation in the inclination-warped disk. An additional implication of the inclination warp is that the disk rises in front of and obscures the central engine at a disk radius of about 8.3 mas, or 0.29 pc. By comparing the observed X-ray column to conditions in the disk at this radius, we argue that the disk must be atomic at 0.29 pc. Hence, we conclude that the molecular-to-atomic transition occurs just beyond the outermost maser at 0.28 pc, and from this we infer an accretion rate of ~10-4α M yr-1, where α (1) is the standard dimensionless parameterization of the kinematic viscosity. Our model suggests that most of the observed X-ray column arises in the warped accretion disk at 0.29 pc and that the maser emission is truncated at large radii predominantly as a result of the molecular-to-atomic phase transition originally proposed by Neufeld & Maloney. The inferred accretion rate is consistent with the jet-dominated accretion models of Yuan et al.