The stellar content of giant H II regions in NGC 7714

In this work we investigate the stellar content of three circumnuclear giant H II regions in the starburst galaxy NGC 7714. We model the stellar population that best reproduces the observational constraints given by the Ha image and the optical spectroscopy from 3710 to 9700 Angstrom. In this paper we address a robust method for analyzing the stellar content of giant extragalactic H II regions (GEHRs) as a first step in a strategy that should allow us to understand better how star formation proceeds. To test the power of the method, we have chosen three very well-studied regions in which many observational constraints are known. The models reproduce simultaneously the observed sizes derived from the H alpha image, the emission-line spectrum from the gaseous component, and the optical features from the massive stars such as Wolf-Rayet bumps in emission, and the near-infrared calcium triplet, CaT, in absorption when detected. We fit the stellar populations through evolutionary synthesis models plus the photoionization code CLOUDY, under the assumption of a shell geometry for the regions. This approach allows us to derive the physical properties of the star clusters, such as mass, age, and metallicity, inside these GEHRs. The method is based on previous work of a complete grid of photoionization models for giant H II regions ionized by evolving star clusters of different metallicities (from 0.05 to 2.5 times solar) and ages between 1 and 5.4 Myr. For the present work, we have followed the cluster evolution further, even after the ionization phase has come to an end, in order to reproduce the observed values of the continuum luminosity at 9000 Angstrom and the equivalent widths of H beta in emission and CaT in absorption. From the results, we find in two of the three studied regions that a single ionizing stellar burst is sufficient to explain all the observational constraints if no reddening is affecting the cluster continuum. Otherwise, the observed values of the H beta equivalent widths imply the existence of. an older component. In this latter case, we find that a model in which two bursts of star formation are considered, a young ionizing one, with ages 3-5 Myr and an older one of 7-9 Myr, that can reproduce the observations. In the third region, the presence of CaT in the near-IR indicates the presence of a non-ionizing population, whose origin is thoroughly discussed.