Rotational signature of the Milky Way stellar halo

We measure the rotation of the Milky Way stellar halo on two samples of Blue Horizontal Branch (BHB) field halo stars from the Sloan Digital Sky Survey (SDSS) with four different methods. The two samples comprise 1582 and 2563 stars respectively and reach out to ~50 kpc in galactocentric distance. Two of the methods to measure rotation rely exclusively on line-of-sight velocities, namely the popular double power-law model and a direct estimate of the de-projected l.o.s. velocity. The other two techniques use the full 3D motions: the radial velocity based rotation estimator of Sch\"onrich, Binney & Asplund (2012) and a simple 3D azimuthal velocity mean. In this context we a) critique the popular model and b) assess the reliability of the estimators. All four methods agree on a weakly prograde or non-rotating halo. Further, we observe no duality in the rotation of sub-samples with different metallicities or at different radii. We trace the rotation gradient across metallicity measured by Deason et al. (2011) on a similar sample of BHB stars back to the inclusion of regions in the apparent magnitude-surface gravity plane known to be contaminated. In the spectroscopically selected sample of Xue et al. (2011), we flag ~500 hot metal-poor stars for their peculiar kinematics w.r.t. to both their cooler metal-poor counter-parts and to the metal-rich stars in the same sample. They show a seemingly retrograde behaviour in line-of-sight velocities, which is not confirmed by the 3D estimators. Their anomalous vertical motion hints at either a pipeline problem or a stream-like component rather than a smooth retrograde population.