We briefly review some constraints for stellar models in various mass regimes and evolutionary stages as provided by observational data from spectroscopy to multi-wavelenghts photometry. The accuracy of present generation of stellar models can be significantly improved only through an extensive comparison between theory and observations.

After a brief review of the observational evidences indicating how the populations of Be stars, red/blue supergiants, Wolf-Rayet stars vary as a function of metallicity, we discuss the implications of these observed trend for our understanding of the massive star evolution. We show how the inclusion of the effects of rotation in stellar models improves significantly the correspondence between theory and observation.

The impact on the predicted Teff scale of using the latest MARCS model atmospheres, instead of a fixed atmospheric structure (e.g., the gray T–τ relation) is examined. The former were fitted to stellar interior models at both the photosphere and at τ = 100 to determine the sensitivity of evolutionary tracks and isochrones for [Fe/H] = 0.0 and −2.0 to the chosen fitting point. In the case of solar abundances, the Teff of the giant branch varied by up to 100–150 K, depending on how the outer layers were treated. Much smaller variations were found for metal-poor giants (or main-sequence stars). Interestingly, models for the low solar Z favored by Asplund et al. (Z=0.0125) were unable to reproduce the gap near the turnoff in the C-M diagram of the old open cluster M 67, in contrast to models that assume Z=0.0188.

We report about the results of the Stellar Evolution Challenge, initiated at the Leiden Lorentz Center meeting “Fine-tuning stellar population models” (June, 2006).

It is well established that mass loss from AGB stars due to dust driven winds cannot be arbitrarily low. We model the mass loss from carbon rich AGB stars using detailed frequency-dependent radiation hydrodynamics including dust formation. We present a study of the thresholds for the mass loss rate as a function of stellar parameters based on a subset of a larger grid of such models and compare these results to previous theoretical work. Furthermore, we demonstrate the impact of the pulsation mechanism and dust formation for the creation of a stellar wind and how it affects these thresholds and briefly discuss the consequences for stellar evolution.

We present a new library of stellar evolution models for a large range of masses and chemical compositions, based on an up-to-date theoretical framework. We briefly discuss the physical inputs and the assumptions adopted in computing the models. The last developments of this database are also presented.

In this proceeding we present the procedure that we have adopted to obtain a dataset of Padova94 tracks (Bressan et al. (1993), Fagotto et al. (1994), Fagotto et al. (1994)) interpolated in metallicity. The procedure requires special care to avoid spurious features in the resulting grid, thus we have subdivided tracks in evolutionary phases, we have chosen the suitable interpolation method and the transition masses. Finally, we have compared our interpolated dataset with a similar models, Girardi et al. (2000), obtaining a general good agreement.

We have computed new models for stars of low and intermediate mass, with varying degrees of α-element enhancement factors, using new low-temperature molecular opacities. We present some of the effects found.

The possibilities and problems of using calculated spectra from model atmospheres when analysing stellar populations in galaxies are reviewed. Various types of consistency tests for stellar models are discussed, as well as comparisons with observational data. It is argued that major improvements in the model spectra are possible and worthwhile.

One of the main ingredients of current stellar population models is a library of stellar spectra. Both empirical and theoretical libraries are used for this purpose, and the question about which one to use is still being debated in the literature. Empirical and theoretical libraries are improving significantly over the years, and many libraries have become available lately. It is not clear what are the advantages of using each of these new libraries, and how far behind are models compared to observations. Here we compare in detail some of the modern theoretical libraries availabe in the literature against empirical libraries attempting to detect their weaknesses and strengths. The aim is to be able to compute in the short future a new synthetic stellar library that combines the best qualities of the current available ones, while improving considerably their weaknesses.

We are calculating stellar spectra for types A through K using Kurucz codes, Castelli models, and Kurucz laboratory lines plus guessed identifications for other lines in the spectra. Weighted coadditions of these spectra are being constructed to match spectra observed in integrated light of old stellar systems such as elliptical galaxies and globular clusters. Grids of theoretical spectra, both stellar and composite, that include an enhancement of light elements and span a wide metallicity range will be calculated over 2200Å–9000Å, and will be archived on MAST at the Hubble Space Telescope website. Here we summarize our results and describe how we automate the fit to our grid of an observed high-resolution stellar or globular-cluster spectrum, to determine the stellar parameters or to break the age-metallicity degeneracy.

The stellar population models dramatically progressed with the arrival of large and complete libraries, ELODIE, CFLIB (=Indo-US) and MILES at a relatively high resolution. We show that the quality of the fits is not anymore limited by the size of the stellar libraries in a large range of ages (0.1 to 10 Gyr) and metallicities (−2 to +0.4 dex). The main limitations of the empirical stellar libraries are (i) the coverage of the parameter space (lack of hot stars of low metallicity), (ii) the precision and homogeneity of the atmospheric parameters and (iii) the non-resolution of individual element abundances (in particular [α/Fe]). Detailed abundances measurements in the large libraries, and usage of theoretical libraries are probably the next steps, and we show that a combination between an empirical (ELODIE) and a theoretical library (Coelho et al. 2005) immediately improves the modeling of (α-enhanced) globular clusters.

We spectroscopically characterize the Galactic Bulge to infer its star formation timescale, compared to the other Galactic components, through the chemical signature on its individual stars. O, Na, Mg, Al were obtained for 50 K giants in four fields towards the Galactic bulge from UVES spectra (R=45,000), while Fe was measured in more than 400 stars with a slightly low resolution (R=20,000) and the GIRAFFE spectrograph at VLT. Oxygen and Magnesium show a well defined trend with [Fe/H], with abundances larger than those measured in both thin and thick disk stars, supporting a scenario in which the bulge formed before and more rapidly than the disk. On the other hand the iron distribution peaks at solar metallicity and it is slightly narrower than that measured in previous works. Part of the present results have been published by Zoccali et al. (2006) and Lecureur et al. (2006), and part will be discussed in forthcoming papers.

We have derived chemical abundances for Ca, Ti, Si, Mg, O, Na, Al, Ni, Co and Cr for a sample of stars with peculiar kinematics and probable origin near the bulge. Our sample stars are in the metallicity range = −0.8 ≤ [Fe/H] ≤ +0.6 dex, and have small pericentric distances, Rp ≤ 3.5 kpc, small scale height, with Zmax < 0.16 kpc, and old ages, 9 to 11 Gyr. We have found that the abundance distributions of O, Mg and Al lie between bulge and both thin and thick disks distributions, i. e., with an enhanced pattern relative to thin and thick disk stars, and an underabundant behavior compared to bulge stars. [Na/Fe] ratios ovelarp the bulge distribution in the metal-poor tail, and show similar values compared to thin and thick disk stars for the super-solar metallicity range. Ca, Ti and Si values are similar to those of the thick disk stars in the metal-poor range, while an underabundant behavior is seen relative to thin disk stars for metallicities [Fe/H] > +0.3 dex. Compared to bulge stars, such elements are deficient in our sample stas. For the iron-peak elements Cr and Ni we have found a slightly overabundant behavior relative to both thin and thick disks distributions in the metal-poor range, and a smooth decreasing trend for [Cr/Fe] for stars in the supersolar regime. [Co/Fe] ratios track the solar value in the metal-poor range, and show an underabundant behavior relative do thin disk stars for metallicities [Fe/H] > 0.0 dex.

Recent stellar spectral libraries have sought higher resolution and the accurate determination of specific optical spectral indeces as stellar population indicators. But the value of accurate flux comparisons over wide wavelength regions should still be emphasized, particularly as more and better spectro-photometric data for composite populations becomes available.

Spectra in the range 4000-7000 Å were obtained for a sample of bulge stars using the GMOS-Gemini low resolution spectrograph. The sample stars were selected from a VLT-FLAMES project for the observation of 1000 bulge stars, for which abundance ratios have been derived. Our aim is to study old stellar populations in external galaxies.

Only stellar atmosphere models that span a wide range of optical depth can allow a correct treatment of radiative transfer at all the frequencies of the spectrum.

Cross-checking the reliability of various stellar spectral databases is an important and desirable exercise. Since number of stars in various databases have no known spectral types and some of the libraries do not have complete coverage resulting in gaps. We use an automated classification scheme based on Artificial Neural Networks (ANN) to cross-classify stars in the Indo-US stellar spectral library (Valdes et al. 2004), JHC (Jacoby, Hunter & Christian 1984), ELODIE spectra (Moultaka et al. 2004) and STELIB (Le Borgne et al. 2003). We have also examined the effects of over-training and over-fitting on the classification efficiency of a Neural Network. It is hoped that such a automated data analysis and validation technique will be useful in the future.

We give a progress report on our work to reduce and calibrate spectra obtained for Hubble's Next Generation Spectral Library (NGSL). We will shortly be working with Bruzual and Charlot to incorporate these spectra into their stellar population synthesis code.

Here we present the results of an observational program aimed at providing a stellar library in the K band with an appropriate coverage of physical stellar parameters (effective temperature, gravity and metallicity) to be used for stellar population synthesis models. In particular, we study the behavior of the CO feature at 2.3 μm as a function of the stellar parameters and we will compute empirical fitting functions that can be easily implemented into stellar population models to provide accurate predictions for integrated CO strengths that will help to face outstanding problems in galaxy formation and evolution.