The Oort Constants Measured from Proper Motions

The Oort constants describe the local variations of the stellar streamingfield. Classically, they are determined from stellar proper motions. We discussproblems arising in this procedure. A large, hitherto overlooked, source ofsystematic error arises from longitudinal variations of the mean stellarparallax. Together with the solar reflex motion, these variations createcontributions to the longitudinal proper motions mu_l*(l) that areindistinguishable from the Oort Constants at the 20% level. Fortunately, we cancorrect for this "mode mixing" using the latitudinal proper motions mu_b(l). Weuse ~10^6 stars from the ACT/Tycho-2 catalogs with proper motion errors of ~ 3mas/yr. We find significant deviations from expectations based on a smoothaxisymmetric equilibrium disk, in particular non-zero C for old red giantstars. We also find variations of the Oort constants with the asymmetric driftof the sub-sample. These correlations are different in nature than thoseexpected for an axisymmetric Galaxy. The most reliable tracers for the ``true''Oort constants are red giants, which are old enough to be in equilibrium anddistant enough to be unaffected by possible local anomalies. For these stars wefind, A ~ 16, B ~- 17, A - B ~ 33, and C ~- 10 km/s/kpc with internal errors ofabout 1-2 and external error of perhaps the same order. These values areconsistent with our knowledge of the Milky Way (flat rotation curve and Omega=A- B ~ 28 +/- 2). Based on observations made with the ESA Hipparcos astrometrysatellite. (Abridged)Comment: Submitted to Ap