The Red Giant Branch luminosity function bump ^*

¹ INAF - Osservatorio Astronomico di Padova, vicolo dell'Osservatorio 5, 35122 Padova, Italy
² Dipartimento di Astronomia, Università di Padova, vicolo dell'Osservatorio 2, 35122 Padova, Italy
³ INAF - Osservatorio Astronomico di Collurania, Via M. Maggini, Teramo, 64100, Italy
⁴ Astrophysics Research Institute, Liverpool John Moores University, Twelve Quays House, Birkenhead, CH41 1LD, UK
⁵ INAF - Osservatorio Astronomico di Roma, Via Frascati 33, 00040 Monte Porzio Catone, Italy
⁶ European Southern Observatory, Karl Schwarzschild Str. 2, 85748, Garching b. München, Germany

Corresponding author: M. Riello, riello@pd.astro.it

Received: 2 June 2003
Accepted: 5 August 2003

Abstract

We present observational estimates of the magnitude difference between the luminosity function red giant branch bump and the horizontal branch (), and of star counts in the bump region (), for a sample of 54 Galactic globular clusters observed by the HST. The large sample of stars resolved in each cluster, and the high photometric accuracy of the data allowed us to detect the bump also in a number of metal poor clusters. To reduce the photometric uncertainties, empirical values are compared with theoretical predictions obtained from a set of updated canonical stellar evolution models which have been transformed directly into the HST flight system. We found an overall qualitative agreement between theory and observations. Quantitative estimates of the confidence level are hampered by current uncertainties on the globular cluster metallicity scale, and by the strong dependence of  on the cluster metallicity. In case of the  parameter, which is only weakly affected by the metallicity, we find a very good quantitative agreement between theoretical canonical models and observations. For our full cluster sample the average difference between predicted and observed  values is practically negligible, and ranges from -0.002 to -0.028, depending on the employed metallicity scale. The observed dispersion around these values is entirely consistent with the observational errors on . As a comparison, the value of  predicted by theory in case of spurious bump detections due to Poisson noise in the stellar counts would be ~0.10 smaller than the observed ones. We have also tested the influence on the predicted  and   values of an He-enriched component in the cluster stellar population, as recently suggested by D'Antona et al. ([CITE]). We find that, under reasonable assumptions concerning the size of this He-enriched population and the degree of enrichment, the predicted  and  values are only marginally affected.