CFHT Information Bulletin Number 37, Semester 97II

Megacam Observations of the Stellar Halo.

Greg Fahlman, UBC

The Stellar Halo is the spheroidal distribution of population II stars which is traced, at least approximately, by the distribution of "halo" globular clusters. It constitutes, at most, only a few percent of the total mass of the Galaxy but the metallicity, kinematics and other collective properties of these stars carry the imprint of the physical processes leading to the formation of the Galaxy. With the remarkable results revealed by the MACHO micro-lensing collaboration, the motivation for achieving a thorough understanding of the spatial distribution and other in situ characteristics of the population II stars is very high.

The recent discovery of the Sagittarius Dwarf Galaxy (SDG) and the realization that this object may have contributed at least four members to the halo globular cluster population has heightened speculation that the stellar halo may not be as homogeneous as is commonly assumed. Streams of stellar debris stripped from bound and accreted companions will persist indefinitely, tracing the orbits of these companions. The discovery and study of such structures will reveal much about the formation and evolution of the Galaxy.

The "pencil beam" survey carried out with FOCAM by Richer and Fahlman (Nature, 358, 1992) was the first attempt at CFHT to probe the in situ characteristics of the distant pop II stars. The results were consistent with a de Vaucouleurs density law, and a mass-spectrum similar to that observed in globular clusters. (There was no evidence for a stellar component associated with the Dark Halo.) A factor of 3 difference in the halo star counts between two 7'x7' fields around the North Galactic Pole is suggestive of an inhomogenous spatial distribution but much wider survey areas need to be probed. Megacam represents an enormous gain over FOCAM (a factor of 64) and offers a wonderful opportunity to obtain a detailed characterization of the stellar distribution in the Galaxy.

The requirements for the survey can be briefly summarized:

  1. We want to observe old, metal poor main-sequence stars at a distance of 100 kpc. This implies a limiting stellar magnitude of V = 25.5 with S/N 10.
  2. Multi-color observations are nededed to establish absolute magnitudes (from colours) and to aid in star-galaxy separation. The limiting magnitudes in the other bands are chosen to match the colours of the most distant detectable main-sequence star: Three bands are a minimum requirement with (B,V,R,I), a preferred option.
  3. The image quality over the field should be high enough to permit star-galaxy separation over most of the available magnitude range: FWHM 0.5". Simple image structure staistics (like a flux-to-peak ratio) can separate stellar candidates from most galaxies provided the PSF is very well characterized. The colours derived from the multi-band survey provide a further discriminant.
  4. Multiple fields need to be observed to look for inhomogenities based; e.g., on reconstructions of the SDG orbit, and to determine the geometrical parameters which describe the averaged (or smooth) Halo.

None of the requirements, except perhaps the choice of certain critically placed fields, are particularly unique to this project. The number of stars compared to faint galaxies in high latitude fields is very small and deep galaxy surveys are completely compatible with this work.

The data analysis can follow two routes. (1) Non-parametric methods may be used to directly estimate the density distribution and the stellar luminosity function under the assumption that a common luminosity function applies throughout the halo. To apply this confidently, the thick disk component either has to be removed or incorporated into the analysis. (The thin disk pop I stars are not a problem here.) (2) A parametric model of the Galaxy including thin- and thick-disk components as well as a stellar halo can be adopted and the parameters constrained with a maximum-likelihood (ML) analysis which incorporates all the data. The two approaches are complementary in the sense that inhomogenities will reveal themselves more directly through the non-parametric appraoch whereas the ML analysis will provide a characterization of the average (smoothed) stellar distribution.

CFHT Information Bulletin Number 37, Semester 97II
Copyright © 1997, Canada-France-Hawaii Telescope