Free access article

A&A 381, 51-64 (2002)
DOI: 10.1051/0004-6361:20011266

Confidence levels of evolutionary synthesis models

II. On sampling and Poissonian fluctuations

M. Cerviño^{1, 2, 3}, D. Valls-Gabaud¹, V. Luridiana^{4, 5} and J. M. Mas-Hesse<sup>6

1</sup>  UMR CNRS 5572, Observatoire Midi-Pyrénées, 14, avenue Édouard Belin, 31400 Toulouse, France
²  Centre d'Études Spatiales des Rayonnements, CNRS/UPS, BP 4346, 31028 Toulouse Cedex 4, France
³  Max-Planck-Institut für extraterrestrische Physik, Giessenbachstrasse, 85748 Garching, Germany
⁴  Instituto de Astronomía, UNAM, Apdo. Postal 70-264, 04510 México D.F., Mexico
⁵  European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching bei München, Germany
⁶  Laboratorio de Astrofísica Espacial y Física Fundamental (LAEFF-INTA), PO Box 50727, 28080 Madrid, Spain

(Received 25 April 2001 / Accepted 6 September 2001)

Abstract
In terms of statistical fluctuations, stellar population synthesis models are only asymptotically correct in the limit of a large number of stars, where sampling errors become asymptotically small. When dealing with stellar clusters, starbursts, dwarf galaxies or stellar populations within pixels, sampling errors introduce a large dispersion in the predicted integrated properties of these populations. We present here an approximate but generic statistical formalism which allows a very good estimation of the uncertainties and confidence levels in any integrated property, bypassing extensive Monte Carlo simulations, and including the effects of partial correlations between different observables. Tests of the formalism are presented and compared with proper estimates. We derive the minimum mass of stellar populations which is required to reach a given confidence limit for a given integrated property. As an example of this general formalism, which can be included in any synthesis code, we apply it to the case of young (  Myr) starburst populations. We show that, in general, the UV continuum is more reliable than other continuum bands for the comparison of models with observed data. We also show that clusters where more than 10 ⁵  have been transformed into stars have a relative dispersion of about 10% in Q(He ⁺) for ages smaller than 3 Myr. During the WR phase the dispersion increases to about 25% for such massive clusters. We further find that the most reliable observable for the determination of the WR population is the ratio of the luminosity of the WR bump over the H luminosity. A fraction of the observed scatter in the integrated properties of clusters and starbursts can be accounted for by sampling fluctuations.

Key words: galaxies: starbust -- galaxies: evolution -- galaxies: statistics -- methods: numerical

Offprint request: M. Cerviño, mcs@laeff.esa.es

© ESO 2002

The OpenURL standard is a protocol for transmission of metadata describing the resource that you wish to access. An OpenURL link contains article metadata and directs it to the OpenURL server of your choice. The OpenURL server can provide access to the resource and also offer complementary services (specific search engine, export of references...). The OpenURL link can be generated by different means.

• If your librarian has set up your subscription with an OpenURL resolver, OpenURL links appear automatically on the abstract pages.
• You can define your own OpenURL resolver with your EDPS Account. In this case your choice will be given priority over that of your library.
• You can use an add-on for your browser (Firefox or I.E.) to display OpenURL links on a page (see http://www.openly.com/openurlref/). You should disable this module if you wish to use the OpenURL server that you or your library have defined.