Exploring Halo Substructure with Giant Stars. II. Mapping the Extended Structure of the Carina Dwarf Spheroidal Galaxy

Abstract

As part of our survey for substructure in the Milky Way halo as traced by giant stars and to look for tidal stellar debris in the halo viable for measurement of the Galactic mass potential with the Space Interferometry Mission (SIM), we explore the distribution of stars beyond the nominal tidal radius of, but still associated with, the Carina dwarf spheroidal galaxy. We make use of the photometric technique described in the first contribution to this series to identify giant star candidates at the distance and metallicity of the Carina dwarf spheroidal across the entire extent of a photometric survey covering some 2.2 deg2 on and around Carina. These Carina-associated giant candidates are identified by a combination of (1) their M-DDO51 colors, which are a measure of both surface gravity and metallicity at given M-T2 colors, and (2) by locations in the color-magnitude diagram commensurate with the Carina red giant branch in the core of the galaxy. The distribution of the giant candidates bears resemblance to the outer isopleths of Carina presented by Irwin & Hatzidimitriou. However, in contrast to previous statistical star-counting approaches, we can pinpoint actual, remotely situated Carina stars individually. Because we can exclude the foreground veil of dwarf stars, our approach allows greater sensitivity and the ability to map the detailed two-dimensional distribution of extended Carina populations to much larger radii, while utilizing smaller aperture telescopes, than other techniques. Moreover, we identify candidate lists of widely displaced Carina-associated stars bright enough for spectroscopic studies of large-scale dynamical and metallicity properties of the system and for astrometric study by SIM. We obtained spectroscopy for three such extratidal stars and from their radial velocities conclude that all three are associated with Carina. While a single King profile matches our derived Carina core profile, we confirm previous claims for a break in the falloff at about 20'. Beyond this radius, a more gradual falloff as r-γ, with 1 < γ < 2, to r ≥ 80' is found. If the existence of profile breaks is a signature of the predominance of unbound stars, and if we adopt the nominal tidal radius of 28' previously found for Carina, then it would appear that we have identified a substantial population from Carina. If these r 20' stars are truly now unbound from the galaxy, we estimate from the relative stellar distribution a fractional destruction rate for Carina from tidal stripping of order = 0.27 Gyr-1. This is among the highest rates expected for the Milky Way dwarf spheroidals apart from Sagittarius. The existence of such extended populations of Carina-associated stars may have important implications for the existence of large dark matter contents in dwarf spheroidals, as well as for the evolution of the Milky Way halo. Finally, we find that the background density of what are likely to be predominantly random, metal-poor halo field giants maintains a rather flat count-magnitude relation out to the distance of Carina, in keeping with R-3 laws for the Galactic halo.