External collaborateurs: Timothy C. Beers (Michigan); Silvia Rossi (São Paulo); Jesper Sommer-Larsen (Copenhagen).
Project description: The kinematics of halo stars potentially provide very strong constraints on the formation history of the Galaxy. At present, the best measurements of the halo's kinematics are obtained from an analysis of the motions of stars found in the near vicinity of the Sun (d < 5 kpc; see Beers & Sommer-Larsen 1995, ApJS, 96, 175). The difficulty of obtaining large samples of more distant in-situ tracers has limited the size (and locations) of samples suitable for analysis of the kinematics in the outer halo of the Galaxy. Never-the-less, some progress has been made. Sommer-Larsen et al. (1997, ApJ 481, 775) have used a combined sample of some 700 FHB stars located at distances from 4 to 20 kpc from the Sun to test a model for the run of kinematics with Galactocentric distance. Their best-fit model is one in which the radial component of the stellar halo's velocity ellipsoid decreases rapidly beyond the Solar circle, from sigmar = 140±10 km/s to sigmar = 89±19 km/s at large radii. This decrease in the radial component is matched by a simultaneous {\it increase} in the tangential component, from $95 \pm 10 $ km/s in the Solar neighborhood to $130 \pm 20$ km/s in the outer halo. These results are best understood in terms of a scenario in which the accretion and merger of smaller ``fragments'' have contributed stars to the outer Galaxy, a picture in concert with present ideas for the formation of galaxies observed at high redshift. While the extant data does a nice job in constraining predictions of the shift from a radially- to tangentially-dominated velocity ellipsoid at distances within 10 kpc of the Sun, {\it confirmatory} evidence can only be obtained from {\it large} samples of stars between 10 and 25 kpc distant, which have the greatest leverage on the model. We have assembled a catalog of large numbers of candidate FHB stars in the magnitude range $15.0 \le B \le 17.5$ from the HES, which explore distances from 8 to 25 kpc. These data include samples of stars in the direction of the Galactic anticenter, which can provide particularly powerful constraints on the model once radial velocities are obtained. Other directions which we hope to explore with the Hamburg/ESO candidate lists are of importance as well. Analysis of the data we have already obtained (Wilhelm {\it et al.} 1998, in preparation) indicates that there may well exist more than one kinematic subsystem of FHB stars in the halo, as defined by their rotation. Several workers have argued for the existence of a counter-rotating component of the halo (the so-called ``high-halo'' region, generally taken to be comprised of stars with distances greater than 4--5 kpc from the Galactic plane). By obtaining radial velocities of a sample of some 200 additional FHB stars at intermediate latitudes in the directions of Galactic antirotation, we hope to establish the existence of this so-called ``retrograde halo population''. For the FHB candidate selection in the HES, an automated, two-step procedure is applied. In the first step, a raw FHB candidate sample is compiled by means of automatic classification (see Christlieb {\it et al} 1998, to appear in: R. Albrecht {\it et al.}, Proceedings of ADASS VII). The classification algorithm has been applied to all extracted, one-dimensional spectra with a signal $>5\sigma$ above diffuse plate background which have not been detected as overlaps by the automatic overlap detection algorithm. In the second step, the sample is cleaned of plate artifacts, a few remaining overlaps, mis-extracted spectra, etc., by visual inspection on the computer screen, and it is checked whether the spectra appear consistent with FHB classification -- {\it i.e.}, they exhibit strong Balmer lines and weak or absent CaII K. Wilhelm {\it et al.} (1998, AJ submitted) describe techniques which we will use to derive estimates of T$_{\rm eff}$, log g and [Fe/H] for the sample of FHB stars we obtain. The Wilhelm {\it et al.} paper shows that, using a weighted combination of (a) UBV colors, (b) Balmer-line widths and (c) CaII K equivalent widths, in conjunction with a metallic-line index to separate Am and Ap stars, it is possible to classify a sample of FHB stars within 10 kpc of the Sun with less than a 10\% mis-identification rate. We expect even fewer misidentifications in the proposed distant sample since nearby disk-type A stars will not be present. Using these techniques it is also been shown that metallicities can be estimated to to within $\pm 0.3$ dex from medium-resolution, $S/N \sim 15$ spectra, with substantial improvement for stars with $S/N = 20$. We have therefore used $S/N = 20$ as our target in planning the observations. At a dispersion of 1 \AA\ per pixel (3 \AA\ per resolution element) we should be able to obtain radial velocities accurate to 15 km/s, which is sufficient for the determination of the line-of-sight velocity dispersions which constrain our model.
Last modified 15.1.1999 by
nchristlieb@hs.uni-hamburg.de