Triaxial orbit-based modelling of the Milky Way nuclear star cluster

We construct triaxial dynamical models for the Milky Way nuclear star cluster using Schwarzschild's orbit superposition technique. We fit the stellar kinematic maps presented in Feldmeier et al. The models are used to constrain the supermassive black hole mass M-center dot, dynamical mass-to-light ratio Y and the intrinsic shape of the cluster. Our best-fitting model has M-center dot = (3.0(-1.3)(+1.1)) x 10(6) M-circle dot, Y = (0.90(-0.08)(+0.76)) M-circle dot/L-circle dot,(4.5 mu m) and a compression of the cluster along the line of sight. Our results are in agreement with the direct measurement of the supermassive black hole mass using the motion of stars on Keplerian orbits. The mass-to-light ratio is consistent with stellar population studies of other galaxies in the mid-infrared. It is possible that we underestimate M-center dot and overestimate the cluster's triaxiality due to observational effects. The spatially semiresolved kinematic data and extinction within the nuclear star cluster bias the observations to the near side of the cluster, and may appear as a compression of the nuclear star cluster along the line of sight. We derive a total dynamical mass for the Milky Way nuclear star cluster of M-MWNSC = (2.1 +/- 0.7) x 10(7) M-circle dot within a sphere with radius r = 2 x r(eff) = 8.4 pc. The best-fitting model is tangentially anisotropic in the central r = 0.5-2 pc of the nuclear star cluster, but close to isotropic at larger radii. Our triaxial models are able to recover complex kinematic substructures in the velocity map.GvdV acknowledges partial support from Sonderforschungsbereich SFB 881 'The Milky Way System' (sub-project A7 and A8) funded by the German Research Foundation. RS acknowledges funding from the European Research Council under the European Union's Seventh Framework Programme (FP7/2007-2013)/ ERC grant agreement n. 614922.Peer reviewe