next previous
Up: Metal-poor turnoff and subgiant


5 Discussion and conclusion

While a significant fraction of turnoff and subgiant stars were already shown to exist in the Basel fields SA51 and SA57 from the analysis of combined RGU and proper motion data in the early papers by Buser & Chiu (1981a,b), here we have demonstrated for the first time that, in the absence of proper motion data, a simple variant of the classical three-color analysis of RGU observations indeed allows to derive density gradients and luminosity functions which are consistent with canonical stellar distribution functions of both the large-scale space densities and the luminosities in the local sphere of the Galaxy.

This achievement sharply contrasts with earlier results derived from the original catalogs of RGU field data, which were based on Steinlin's (1968) preliminary standard of the RGU system and which were shown to carry large systematic errors in the basic color and absolute magnitude calibrations (Buser 1988; Buser & Rong 1995). In fact, using the original RGU data base, density gradients and luminosity functions could not be derived consistent with both plausible density models and local luminosity functions for the large majority of Basel survey fields (e.g., Fenkart 1989a-d).

As anticipated in the introduction, the present improvements can be attributed completely to the revised standardization procedure and the extended calibrations set up for the RGU system via synthetic photometry techniques. The comprehensive tie-in with the UBV system then also justifies the iterative procedure adopted in Sect. 3 for determining absolute magnitudes for the lower-abundance stars by feedback from the resulting density and luminosity functions.

However, while in this paper we have been able to demonstrate in a very simple way that indeed, all the above innovations are leading to an altogether positive result, it must also be admitted that no claims of uniqueness can be made for the solution presented in Sect. 4. The main reason is simply that in actual practice the star-by-star treatment prescribed by the classical three-color method (which was adopted here!) does not allow to account for the multiple values of the astrophysical parameters that can be associated with an individual star, given its apparent magnitude, two-color position, and an upper limit in absolute magnitude only. A fortiori, this same procedure neither allows to account for more than one or (at most!) two full iterations of propagating through the resulting density and luminosity functions the different assigned absolute magnitudes of a large sample of randomly selected stars which are scattered in a finite area of the observed (U-G,G-R) two-color diagram.

Therefore, a more exacting quantitative treatment naturally requires a more performing automated analysis which also allows to exploit the full ranges of available astrophysical calibrations and constraints, such as the dependencies of the stellar colors on age, metallicity, and luminosity. Most importantly, the calibration of the observed ultraviolet excess, $\delta_{U{-}G}$, will need to be extended beyond the main sequence to the lower-gravity turnoff and subgiant stars in order to provide a more accurate determination of the metallicity distribution of the intermediate and extreme populations II identified in this paper. These more ambitious tasks are being completed as part of the all-survey analysis; preliminary results obtained for the first-half survey of seven fields are published in Buser et al. (1998a).

Acknowledgements

This work was supported by the Swiss National Science Foundation.


next previous
Up: Metal-poor turnoff and subgiant

Copyright The European Southern Observatory (ESO)