Table of contents

Volume 328 (2011) of the Journal of Physics: Conference Series provides a record of the invited and contributed talks, and of the posters presented at the GREAT-ESF workshop entitled `Stellar Atmospheres in the Gaia Era: Quantitative Spectroscopy and Comparative Spectrum Modelling' (http://great-esf.oma.be and mirrored at http://spectri.freeshell.org/great-esf). The conference was held on 23–24 June 2011 at the Vrije Universiteit Brussel, Belgium. 47 scientists from 11 countries around the world attended the workshop.

The ESA–Gaia satellite (launch mid 2013) will observe a billion stellar objects in the Galaxy and provide spectrophotometric and high-resolution spectra of an unprecedented number of stars observed with a space-based instrument. The confrontation of these data with theoretical models will significantly advance our understanding of the physics of stellar atmospheres. New stellar populations such as previously unknown emission line stars will be discovered, and fundamental questions such as the basic scenarios of stellar evolution will be addressed with Gaia data.

The 33 presentations and 4 main discussion sessions at the workshop addressed important topics in spectrum synthesis methods and detailed line profile calculations urgently needed for accurate modelling of stellar spectra. It brought together leading scientists and students of the stellar physics communities investigating hot and cool star spectra. The scientific programme of the workshop consisted of 23 oral (6 invited) and 10 poster presentations about cool stars (first day; Comparative Spectrum Modelling and Quantitative Spectroscopy of Cool Stars), and hot stars (second day; Quantitative Spectroscopy of Hot Stars). The hot and cool stars communities use different spectrum modelling codes for determining basic parameters such as the effective temperature, surface gravity, iron abundance, and the chemical composition of stellar atmospheres. The chaired sessions of the first day highlighted new research results with spectral synthesis codes developed for cool stars, while the second day focused on codes applied for modeling the spectra of hot stars.

The workshop addressed five major topics in stellar atmospheres research:

• Spectrum synthesis codes

• Radiation hydrodynamics codes

• Atmospheric parameters, abundance, metallicity, and chemical tagging studies

• Large spectroscopic surveys

• New atomic database

The workshop presentations discussed various important scientific issues by comparing detailed model spectra to identify differences that can influence and bias the resulting atmospheric parameters. Theoretical line-blanketed model spectra were compared in detail to high-resolution spectroscopic observations. Stellar spectra computed (i.e., in the Gaia Radial Velocity Spectrometer wavelength range) with 1-D model atmosphere structures were mutually compared, but also to 3-D models from advanced radiation hydrodynamics codes. Atmospheric parameters derived from spectrum synthesis calculations assuming Local Thermodynamic Equilibrium (LTE) were evaluated against more sophisticated non-LTE models of metal-poor stars and the extended atmospheres of giants and supergiants. The workshop presented an overview of high-resolution synthetic spectral libraries of model spectra computed with the synthesis codes. The spectral model grids will be utilized to derive stellar parameters with the Discrete Source Classifier Algorithms currently under development in the Gaia DPAC consortium (http://www.rssd.esa.int/index.php?project=GAIA&page=DPAC_Introduction). They are implemented for training Gaia data analysis algorithms for the classification of a wide variety of hot and cool star types; FGK and M stars, OB stars, white dwarfs, red supergiants, peculiar A and B stars, carbon stars, ultra cool dwarfs, various types of emission line stars, Be stars, Wolf–Rayet stars, etc. A substantial number of oral and poster presentations discussed different techniques for measuring the abundance of various chemical elements from stellar spectra. The presented methods utilize spectra observed with large spectral dispersion, for example for accurately measuring iron, carbon, and nitrogen abundances. These methods are important for ongoing development and testing of automated and supervised algorithms for determining detailed chemical composition in tagging studies of large (chemo-dynamical) spectroscopic surveys planned to complement the Gaia (astrometric and kinematic) census of the Galaxy.

The complete scientific programme is available here. The workshop website also offers the presentation viewgraphs (in PDF format) and some nice photographs of the talks and poster breaks http://great-esf.oma.be/program.php. The papers presented at the workshop and collected here have been edited by A Lobel, J-P De Greve, and W van Rensbergen. The Proceedings essentially follow the order of presentation during the conference program, divided into cool and hot stars (oral papers are followed by poster papers). It also offers a review paper about new research results presented at the conference, including a record of the main discussion sessions. 27 papers passed through the peer review process. The manuscripts were submitted before 15 September 2011 and accepted by the referees before 1 November 2011. We would like to thank the reviewers for their constructive criticism during the preparation of this Volume. It has 225 pages by 27 authors and 135 co-authors, and includes 114 color figures and 21 tables. The articles are freely available at http://iopscience.iop.org/1742-6596/328/1.

We would like to express our gratitude for the financial support from the European Science Foundation and the Research Foundation – Flanders. We thank the Vrije Universiteit Brussel for making the meeting and poster rooms freely available during the conference. We specially thank Mrs Merel Fabré for assistance with the workshop organization and administration at the VUB. We thank the Royal Observatory of Belgium for help with conference grants management, transportation, and administration. We are grateful to all the workshop participants for their valuable contributions and active discussions which made the conference very successful indeed.

Alex LobelRoyal Observatory of Belgium, Ringlaan 3, B-1180 Brussels, Belgium E-mail: Alex.Lobel@oma.be     alobel@sdf.lonestar.org Web: http://alobel.freeshell.orgJean-Pierre De GreveWalter van RensbergenVrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium Editors Brussels, October 2011

CONFERENCE POSTER

SCIENTIFIC COMMITTEE Jose Groh (Germany) Peter Hauschildt (Germany) Ulrike Heiter (Sweden) Alex Lobel (Belgium) Bertrand Plez (France) Norbert Przybilla (Germany) Rosanna Sordo (Italy)

CONFERENCE SPONSORS European Science Foundation (ESF) http://www.esf.org Research Foundation – Flanders (FWO-Vlaanderen)http://www.fwo.be Free University Brussels (VUB) http://www.vub.ac.be Royal Observatory of Belgium (ROB) http://www.astro.oma.be

CONFERENCE PHOTOGRAPH (23 June 2011)

CONFERENCE PARTICIPANTS

All papers published in this volume of Journal of Physics: Conference Series have been peer reviewed through processes administered by the proceedings Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.

The Gaia mission will provide spectroscopic and spectrophotometric observations for a vast number of stars, allowing a characterization in terms of stellar parameters along with the astrometric measurements. The pipeline for astrophysical parameter determination relies on libraries of model spectra for mapping stellar parameters to flux distributions. To maximize the reliability of the parametrization, adequate efforts are required to assess the realism of the model spectra and to calibrate the analysis methods. We describe a spectrum modelling experiment, in which high-resolution optical and infrared spectra of four cool giant stars were analysed by 14 different groups. The resulting variance in derived stellar parameters illustrates the need to be cautious when comparing or combining stellar parameters from different model atmospheres and analysis strategies. The main causes for the differences in parameters derived by different groups seem to lie in the physical input data and in the details of the analysis method.

We describe the first results from our project aimed at large-scale calculations of NLTE abundance corrections for important astrophysical atoms and ions. In this paper, the focus is on Fe which is a proxy of stellar metallicity and is commonly used to derive effective temperature and surface gravity. We present a small grid of NLTE abundance corrections for a sample of Fe I lines and discuss how the NLTE effects influence the determination of effective temperature, surface gravity, and metallicity for late-type stars.

In this contribution, we present the STAGGERGRID, a collaborative project for the construction of a comprehensive grid of time-dependent, three-dimensional (3-D), hydrodynamic model atmospheres of solar- and late-type stars with different effective temperatures, surface gravities, and chemical compositions. We illustrate the main characteristics of these 3-D models and their effects on the predicted strengths, wavelength-shifts, and shapes of spectral lines, highlighting the differences with respect to calculations based on classical, one-dimensional, hydrostatic models, and discuss some of their possible applications to elemental abundance analysis of stellar spectra in the context of large observational surveys.

The ALHAMBRA photometric system was specifically designed to perform a tomography of the Universe in some selected areas. Although mainly designed for extragalactic purposes, its 20 contiguous, equal-width, medium-band photometric system in the optical wavelength range, shows a great capacity for stellar classification. In this contribution we propose a methodology for stellar classification and physical parameter estimation (T_eff, log g, [Fe/H], and color excess E(B – V)) based on 18 independent reddening-free Q-values from the ALHAMBRA photometry. Based on the theoretical Spectral library BaSeL 2.2, and applied to 288 stars from the Next Generation spectral Library (NGSL), we discuss the reliability of the method and its dependence on the extinction law used.

I present the widely used model atmosphere codes ATLAS, MARCS, and PHOENIX, and I compare their output model structures and spectra for cool stars of FGKM-types. While model atmosphere stratifications agree closely with each other in the f-D approximation, this is not the case for spectra. Differences between model spectra from different codes are largest in the blue-UV, but smaller differences appear in all regions, especially in the molecular features of cooler model spectra. I recommend the groups to try to solve these discrepancies together. In the meantime, users must be careful when using these spectra in regimes where they differ.

I discuss here only comparisons of spectra at solar metallicity, and this should be extended to other metallicities. Detailed comparisons with carefully calibrated spectrophotometric data, and high resolution spectra for stars with well known parameters are also of prime importance. It appears that we still need better line positions for molecules. Finally we should remember that f-D models are only a step towards a better representation of reality, and we should keep developing, and carefully test 3-D, NLTE models.

Gaia will observe up to a billion stellar sources. Automated algorithms are under development to derive the atmospheric parameters of all observed spectra, from low resolution optical spectra alone or in synergy with high resolution spectra in the near-IR Ca II triplet region. To do so, a large database of state-of-the-art stellar libraries has been produced for the Gaia community, computed using different codes optimized for specific purposes. The choice to use different spectral codes in different regions of the H-R diagram raises the problem of the coherence of the different spectra, specifically in the transition zones. We present a comparison between the libraries from the point of view of spectra simulations for training the Gaia algorithms. We also present the implementation of these libraries into a Simple Stellar Population code.

Differences between metallicity measurements of dwarfs and giants belonging to the same open clusters, as envisaged, would have tremendous implications, among other things, on the planet formation theory and on the stellar populations in the Milky Way. They also have a high importance on their own right, as they may indicate either that there are important processes presently not taken into account in modelling the atmosphere of evolved stars, or that the composition of the stars changes when they leave the main sequence. However, they have not yet been evaluated based on solid and conclusive data. We review literature data, employing different models and line lists, in order to investigate the extent, the origin, and the nature of the aforementioned differences.

Using BD+46 442 as an example, we describe our analysis of the spectra of post-AGB stars (SpTs F-K, luminosity types I-II, [Fe/H]=−2.0...+0.5), obtained with the échelle spectrograph HERMES on the 1.2 m Mercator telescope. We obtain atmospheric parameters and atomic abundances using hydrogen line profiles and equivalent widths (EW) of weak metal lines. Our oscillator strengths and solar abundances for the majority of elements have been consistently adjusted to match the observed solar spectrum. The EW analysis is performed using F. Castelli's modified ATLAS9 photospheric models and C. Sneden's radiative transfer code MOOG. The resulting abundances are compared to those obtained by using the original R. Kurucz's ATLAS9 models and the WIDTH9 code. The outlined procedure can be employed for the spectroscopic analysis of the high-resolution spectra of warm (super-)giant stars in ground-based follow-up programs of Gaia.

S stars are cool stars of temperatures similar to those of M giants, but their atmospheres are enriched in carbon and s-process elements because of either extrinsic pollution by a binary companion or intrinsic nucleosynthesis and dredge-up on the thermally-pulsing AGB. Despite numerous attempts to link phenomenological spectral classification criteria to physical parameters (T_eff, gravity, C/O, [s/Fe], [Fe/H]), the parameter space of S stars is poorly known and this has prevented accurate abundance analysis of S stars until now. Here we present a large grid of S-star model atmospheres. ZrO and TiO band strength indices as well as VJHKL photometry are needed to disentangle the effective temperature, C/O and [s/Fe]. The stellar parameters derived on the basis of low-resolution spectra and photometry are shown to be fairly accurate when compared to high-resolution data of the same stars. The C/O ratio of S stars is found to be between the solar value (0.5) and 0.99, and not 1 as often claimed in the literature. Consistently with stellar evolution expectations, the C/O ratio increases as the effective temperature decreases.

The frequency of maximum oscillation power measured in dwarfs and giants exhibiting solar-like pulsations provides a precise, and potentially accurate, inference of the stellar surface gravity. An extensive comparison for about 40 well-studied pulsating stars with gravities derived by traditional means (ionization balance, pressure-sensitive spectral features or location with respect to evolutionary tracks) supports the validity of this technique and reveals an overall remarkable agreement with mean differences not exceeding 0.05 dex (although with a dispersion of up to ∼0.2 dex). It is argued that interpolation in theoretical isochrones may be the most precise way of estimating the gravity by traditional means in nearby dwarfs. The use of seismic targets as benchmarks in the context of forthcoming large-scale surveys (such as the follow up of the Gaia mission) is briefly discussed.

The question of the origin and evolution of planetary systems is of fundamental importance for astrophysics. Dusty debris discs are signatures of planetary systems and, therefore, constitute valuable tools to provide new light in our understanding of how planetary systems form and evolve. We present the first results of a spectroscopic programme of a sample of stars with debris discs. High-resolution echelle spectra are used to determine metallicities and abundances. Properties of stars with debris discs, are compared with those of stars hosting planets, as well as 'normal' stars.

Convection plays an essential role in the emerging intensity for many stars that will be observed by Gaia. Convective-related surface structures affect the shape, shift, and asymmetry of absorption lines, the photocentric and photometric variability causing bias in Gaia measurements. Regarding the importance of Gaia mission and its goals, it is mandatory to have the best models of the observed stars. 3-D time-dependent hydrodynamical simulations of surface convection are crucial to model the photosphere of late type stars in a very realistic way. These simulations are an important tool to correct the radial velocities and to better estimate the parallaxes and photometric variability.

Grids of theoretical stellar spectra are fundamental for estimating basic stellar parameters from photometric and spectroscopic data observed in large sky surveys such as SDSS, LAMOST, Gaia, etc. Do the different atmosphere models influence the parameters estimation? We compute the Lick indexes and uvby color indexes using the MAFAGS-OS grid of model atmospheres and fluxes provided by F. Grupp (personal comm.) and the Kurucz grids [1]. A spectrum comparison reveals the behavior of spectra from the MAFAGS and Kurucz grids.

We find that using the (b-y) index, consistent effective temperatures can be determined from both the Kurucz and MAFAGS grids of theoretical spectra. The m1 index, together with color index, can be used to determine the metallicity of F- and G-type stars, but the measurements of the Kurucz and MAFAGS grids show systematic discrepancies for cool stars. The c1 indexes computed with both grids show small discrepancies for T_eff < 6000 K, while for T_eff > 6000 K, the c1 indexes agree well.

The Lick indexes of the Kurucz grid and the MAFAGS grid tend to be in agreement for warm stars with temperatures above 5000 K, while for cool stars with temperatures ranging from 4000 K to 5000 K, the difference of Lick indexes for both models is apparently large.

We also compare the MAFAGS spectrum and Kurucz spectrum of the same temperature, surface gravity, and metallicity using a correlation coefficient for the complete spectrum. For warm stars, the MAFAGS and Kurucz spectra are almost the same, while for cool stars below 5000 K, there are some discrepancies between the MAFAGS and Kurucz spectra that induce internal discrepancies in the parameters determination.

S-type stars are thought to be the first objects, during their evolution on the asymptotic giant branch (AGB), to experience s-process nucleosynthesis and third dredge-ups, and therefore to exhibit s-process signatures in their atmospheres. Until present, the modeling of these processes is subject to large uncertainties. Precise abundance determinations in S stars are of extreme importance for constraining e.g., the depth and the formation of the ¹³C pocket. In this paper a large grid of MARCS model atmospheres for S stars is used to derive precise abundances of key s-process elements and iron. A first estimation of the atmospheric parameters is obtained using a set of well-chosen photometric and spectroscopic indices for selecting the best model atmosphere of each S star. Abundances are derived from spectral line synthesis, using the selected model atmosphere. Special interest is paid to technetium, an element without stable isotopes. Its detection in stars is considered as the best possible signature that the star effectively populates the thermally-pulsing AGB (TP-AGB) phase of evolution. The derived Tc/Zr abundances are compared, as a function of the derived [Zr/Fe] overabundances, with AGB stellar model predictions. The computed [Zr/Fe] overabundances are in good agreement with the AGB stellar evolution model predictions, while the Tc/Zr abundances are slightly over-predicted. This discrepancy can help to set stronger constraints on nucleosynthesis and mixing mechanisms in AGB stars.

It is generally accepted that the atmospheres of cool/lukewarm stars of spectral types A and later are described well by LTE model atmospheres, while the O-type stars require a detailed treatment of NLTE effects. Here model atmosphere structures, spectral energy distributions and synthetic spectra computed with ATLAS9/SYNTHE and TLUSTY/SYNSPEC, and results from a hybrid method combining LTE atmospheres and NLTE line-formation with DETAIL/SURFACE are compared. Their ability to reproduce observations for effective temperatures between 15 000 and 35 000 K are verified. Strengths and weaknesses of the different approaches are identified. Recommendations are made as to how to improve the models in order to derive unbiased stellar parameters and chemical abundances in future applications, with special emphasis on Gaia science.

Simple spherical, non-rotating stellar models are inadequate when describing real stars in the limit of very fast rotation: Both the observable spectrum and the geometrical shape of the star deviate strongly from simple models. We attempt to approach the problem of modeling geometrically distorted, rapidly rotating stars from a new angle: By constructing distorted geometrical models and integrating standard stellar models with varying temperature, gravity, and abundances, over the entire surface, we attempt a semi-empirical approach to modeling. Here we present our methodology, and present simple examples of applications.

The derivation of high precision/accuracy parameters and chemical abundances of massive stars is of utmost importance to the fields of stellar evolution and Galactic chemical evolution. We concentrate on the study of OB-type stars near the main sequence and their evolved progeny, the BA-type supergiants, covering masses of ∼6 to 25 solar masses and a range in effective temperature from ∼8000 to 35 000 K. The minimization of the main sources of systematic errors in the atmospheric model computation, the observed spectra and the quantitative spectral analysis play a critical role in the final results. Our self-consistent spectrum analysis technique employing a robust non-LTE line formation allows precise atmospheric parameters of massive stars to be derived, achieving 1σ-uncertainties as low as 1% in effective temperature and ∼0.05–0.10 dex in surface gravity. Consequences on the behaviour of the chemical elements carbon, nitrogen and oxygen are discussed here in the context of massive star evolution and Galactic chemical evolution, showing tight relations covered in previous work by too large statistical and systematic uncertainties. The spectral analysis of larger star samples, like from the upcoming Gaia-ESO survey, may benefit from these findings.

The LDS (Low Dispersion Spectroscopy) performed in various extended sky surveys with optical telescopes using objective prism and photographic plates offers an interesting opportunity to test simulated low-dispersion spectra for the Gaia BP/RP photometers and to compare them with real data, especially for objects with strong emission lines. We present a review of astrophysics with LDS performed in the past, as well as an overview of existing extended sky surveys (with photographic plates) providing LDS data. Some of them provide almost complete coverage of the northern or southern hemisphere (e.g. the Northern and Southern Mt Wilson - Michigan Hα surveys or the German La Paz Bolivia Southern Spectral Sky Survey). We show examples of these data and discuss a comparison of existing LDS plate data with expected/simulated Gaia BP/RP data. We show examples of real data for objects with very strong and wide emission features confirming that such features will be detectable with Gaia BP/RP. We also discuss the importance of Gaia RP/BP low-dispersion spectroscopy for astrophysical studies.

The Gaia-ESO Survey will produce spectra of about 100 000 stars, using the VLT FLAMES instrument. This includes hot, massive stars in a number of selected clusters. I describe the on-going cluster selection as well as the work package responsible for analyzing the hot-star spectra.

Massive stars are extremely luminous and characterized by mass loss through radiation-driven stellar winds. The radiation emitted at the stellar surface may interact with the wind, making the analysis of the emerging spectrum a very challenging task. In addition to the luminosity, effective temperature, and surface gravity, several other stellar parameters impact the spectral morphology of these objects, remarkably the mass-loss rate and the wind terminal velocity. That is generally the case for OB supergiants, Luminous Blue Variables, and WR stars. CMFGEN [1] comprises the state-of-the-art in non-LTE radiative transfer and has been successfully applied to the above classes of objects over the last decade. The code assumes spherical symmetry, stationary outflow, and both photospheric and wind lines can be treated in non-LTE. Full line blanketing due to hundreds of thousands of spectral lines is included, as well as wind clumping. Here I discuss the assumptions behind CMFGEN and present examples of models and spectroscopic analyses.

We present the IACOB grid-based automatic tool for the quantitative spectroscopic analysis of O-stars. The tool consists of an extensive grid of FASTWIND models, and a variety of programs implemented in IDL to handle the observations, perform the automatic analysis, and visualize the results. The tool provides a fast and objective way to determine the stellar parameters and the associated uncertainties of large samples of O-type stars within a reasonable computational time.

The VLT-FLAMES Tarantula Survey (VFTS) has secured mid-resolution spectra of over 300 O-type stars in the 30 Doradus region of the Large Magellanic Cloud. A homogeneous analysis of such a large sample requires automated techniques, an approach that will also be needed for the upcoming analysis of the Gaia surveys of the Northern and Southern Hemisphere supplementing the Gaia measurements. We point out the importance of Gaia for the study of O stars, summarize the O star science case of VFTS and present a test of the automated modeling technique using synthetically generated data. This method employs a genetic algorithm based optimization technique in combination with fastwind model atmospheres. The method is found to be robust and able to recover the main photospheric parameters accurately. Precise wind parameters can be obtained as well, however, as expected, for dwarf stars the rate of acceleration of the ow is poorly constrained.

We are building a new spectral library with the X-Shooter instrument on ESO's VLT: XSL, the X-Shooter Spectral Library. We present our progress in building XSL, which covers the wavelength range from the near-UV to the near-IR with a resolution of R∼10 000. At the time of writing we have collected spectra for nearly 240 stars. An important feature of XSL is that we have already collected spectra of more than 100 Asymptotic Giant Branch stars in the Galaxy and the Magellanic Clouds.

NGC 2244 located in the Rosette Nebula is a young open cluster composed of seven O-type stars. A first paper focused on the multiplicity of these stars, revealed only one binary system out of the seven studied stars. The minimum binary fraction of this cluster (∼ 14%) differs to the average fraction measured on the nearby clusters (∼ 44%). In order to better constrain this discrepancy, an analysis based on the determination of the stellar and wind parameters of these stars with the CMFGEN atmosphere code was performed. The main results confirm that all the stars have an age between 0 and 5 Myr, and that the N surface abundance appears to be consistent with the evolutionary models for a population of stars of the same age. Moreover, this investigation exhibits the existence of dynamical interactions inside this young open cluster sufficiently strong to eject the hottest component from its centre.

Interest in near infra-red spectroscopy of massive stars has increased dramatically over the last decades, as it offers the possibility to analyze stars embedded in dusty star forming regions and near the Galactic center. We present an analysis of both high resolution optical and, separately, high resolution VLT/CRIRES near-IR spectra in the J, H, K and L-band of nearby dwarf O-type stars. Applying a genetic fitting algorithm approach using state-of-the-art FASTWIND non-LTE atmospheres, we present a comparison of the stellar and wind properties as derived from these two spectral regimes. In this approach we retrieve the effective temperature to within a sub-type and the surface gravity to within 0.2 dex, but find a discrepancy in the mass-loss rates of 0.2 up to 1.0 dex. We find that He II 1012 nm and Brackett-α lines do not yield consistent mass-loss estimates, the former producing much lower values than Hα, while Brackett-α and Hα are consistent.

A long term project to investigate the spectral appearance over the Gaia RVS domain of a large sample of Be stars and interacting binaries has been undertaken. The aim of the Ondřejov project is to create sufficient amounts of training data in the RVS wavelength domain to complement the Bp/Rp classification of Be stars which may be observed with Gaia. The project's current status is described and sample spectra in both the Hα and RVS wavelength domains are presented and discussed.

Highlights of the meeting on Stellar Atmospheres in the Gaia Era: Quantitative Spectroscopy and Comparative Spectrum Modeling (http://great-esf.oma.be and mirrored at http://spectri.freeshell.org/great-esf) held on 23-24 June 2011 in Brussels, Belgium are emphasized. New research results are summarized and a record of the scientific discussions during the meeting is provided, as well as important open questions for future research.