Extragalactic Stellar Astronomy/The Distance Scale

Extragalactic stellar astronomy comprises the large field of detailed studies of individual stars in nearby galaxies (up to several ten million light-years away), in order to understand stars and stellar evolution in different environments, and to infer crucial information on their host galaxies. It has only recently become feasible, with the advent of large ground-based telescopes of the 8--10m class (e.g. the VLT or Keck) and dedicated instrumentation (e.g. UVES, FORS). Quantitative spectroscopy is the technique of choice for performing this task.

Figure 10: Supergiants in the spiral galaxies NGC300 (located at a distance of 6.5 million light-years in the nearby Sculptor Group) and NGC3621 (at 21.5 million light-years in the field) were the first targets for quantitative spectroscopy beyond the Local Group. Spectra of 62 and 10, respectively, blue supergiants were obtained with FORS1 on the VLT (Bresolin et al. 2001, 2002).

A recent breakthrough represents the step beyond the Local Group. The first two target galaxies were NGC300 and NGC3621 (Fig. 10), where medium-resolution spectra were obtained for a large number of objects. Quantitative spectroscopy of stars this distant was performed for the first time. The spectrum synthesis gives confidence that stellar metallicities can be determined with sufficient accuracy from spectra of this quality (Fig. 11). In turn, the abundance data from the individual stars can be used to confirm/expand previous studies of nebulae on the galactochemical evolution of these galaxies, manifesting itself e.g. in the galactic abundance gradient (Fig. 12).

Spectrum Synthesis for one of the NGC300 Supergiants

Figure 11: Spectrum synthesis at 0.2, 0.5 and 1.0 times solar metal abundance (dotted, dashed and solid lines, respectively) for a BA-type supergiant in NGC300 (thick solid line). It is concluded that this star, lying in the outskirts of that galaxy, is metal poor, at approximately 0.2 times solar metallicity. Unlabeled short marks are used for the identification of FeII lines, all others are marked explicitly.

Figure 12: Comparison of stellar (dots) and HII region abundances (diamonds, from Deharveng et al. 1988) as a function of galactocentric radius in NGC300. A typical error bar for the HII region abundances is shown in the lower left. Some HII regions have been observed twice, in this case the corresponding symbols have been connected. The dotted and dashed lines are least-squares fits to the data, representing the nebular oxygen abundance gradient of NGC300 for two different calibrations.

Besides the large potential for galactic studies, quantitative spectroscopy of BA-type supergiants can contribute to another field of active research in modern astronomy. These objects offer also a great potential for constraining the extragalactic distance scale from exploiting even two independent, purely spectroscopic indicators. This is an essential step for the determination of the Hubble constant, a fundamental cosmological parameter that describes cosmic expansion.

The first distance indicator dates back to the findings of Puls et al. (1996) for O stars, that the wind momentum rate (the product of mass-loss rate and terminal velocity of the stellar wind) is related to the stellar luminosity. This wind momentum--luminosity relationship (WLR) is well understood in terms of the theory of radiation-driven stellar winds (details can be found here). Application of the WLR requires the fitting of the wind-dominated H_alpha feature through spectrum synthesis based on hydrodynamic model atmospheres in order to derive the mass-loss rate and observations of the UV metal resonance lines for determining the terminal velocity. A WLR is also found for BA-type supergiants (Kudritzki et al. 1999), where the terminal velocity can also be derived from H_alpha for cases with a well-developed P-Cygni profile. Though, the WLR has to be calibrated empirically with respect to stellar metallicity (photon absorption by the metal lines drives the wind), which is work under progress (Fig. 13).

Figure 13: Wind momentum--luminosity relationship for A-type supergiants in the Galaxy and M31 (at ~solar metallicity). The positions of two objects in NGC300 and NGC3621 are also displayed. From Bresolin et al. (2002).

The second distance indicator related to quantitative spectroscopy of BA-type supergiants is the flux-weighted gravity--luminosity relationship (FGLR). It is based on the absorption strengths of the higher Balmer lines formed in the stellar photosphere which are barely affected by the stellar wind. Massive stars pass quickly through the phase of late B and early A-type supergiants during their evolution to the red supergiant stage, at roughly constant mass and luminosity. This means that in this phase the surface gravity and effective temperature of the star are coupled and they are related to the stellar luminosity through the well-known mass--luminosity relationship. Both parameters are determined from non-LTE spectrum synthesis, in particular from the modelling of the gravity-sensitive Balmer lines (line broadening through the linear Stark effect) and temperature-sensitive ionization equilibria (or an empirical spectral type--temperature relation). Good agreement with the predictions from stellar evolution calculations is found in a first step of the empirical calibration of the FGLR (Fig. 14).

Figure 14: Flux-weighted gravity--luminosity relationship as derived from 37 supergiants of spectral types B8 to A4 from 8 galaxies in the Local Group and beyond. The relationship obtained from stellar evolution models at solar metallicity including the effects of rotation (Meynet & Maeder 2000) is also shown and labeled with the initial zero-age main-sequence masses of the corresponding stellar models. Effective temperatures are given in units of 10⁴K.

However, in order to fully exploit the potential of BA-type supergiants for the study of the Local Universe lots of fundamental work still has to be done. This can only be performed on objects within the Local Group, for which high-resolution spectroscopy is feasible. The Local Group (see Fig. 15 for a schematic view) is an assembly of ~35 gravitationally bound galaxies spread over a region of approximately 4 million light-years diameter. Galaxy groups are the link between field galaxies -- free floating systems in the universe -- and galaxy clusters, comprising up to thousands of gravitationally bound galaxies. Our Milky Way is one of the three dominating giant spiral galaxies of the Local Group. Most of the different galaxy types are found in the Local Group, except for giant ellipticals, lenticular and heavily interacting systems.

Figure 15: Schematic 3-D view of the Local Group. The dashed ellipsoid marks a radius of 1 Mpc (=3.26 million light-years) around the Local Group barycentre. The underlying grid is parallel to the plane of the Milky Way. Galaxies above the plane are indicated by solid lines and below with dotted lines. The dashed circles enclose the presumed M31/M33 and the Milky Way subsystems. Morphological types: large spirals (open symbols), dwarf irregulars (dIrrs; yellow symbols), dwarf ellipticals and gas-deficient dwarf spheroidals (dEs & dSphs; blue-green symbols), dSph/dIrr transition types (violet symbols). From Grebel (1999).

The focus of attention will still lie on the nearest neighbours of the Milky Way, the Magellanic Clouds (Fig. 16 & 17), which are gravitationally interacting with each other and our own galaxy. Each of these systems contains a plethora of blue supergiants accessible to high-resolution spectroscopy even with telescopes of modest size.

Figure 16: The Large Magellanic Cloud (LMC) is the nearest neighbour galaxy to the Milky Way. It is a fairly large galaxy of irregular type. Located at a distance of ~160000 light-years, the LMC is visible by the naked eye on the southern sky, and it is well resolved into stars with even a modest telescope. The bright red patch at the eastern end of the galaxy is the prominent star-forming region 30 Doradus (the Tarantula Nebula).

Figure 17: Two metal-poor dIrr galaxies of the Local Group: the Small Magellanic Cloud (SMC) and NGC6822. The SMC is the second nearest neighbour to the Milky Way, while NGC6822 is located at a ~tenfold larger distance. While the SMC is gravitationally interacting with the LMC (therefore its distorted shape), NGC6822 is an isolated system, separated by several hundred-thousand light-years from any other galaxy within the Local Group. The NGC6822 image is adopted from this site.

In order to study objects in different environments, like isolated dwarf irregular galaxies (e.g. NGC6822, otherwise a twin of the SMC, Fig. 17) or other spiral galaxies, like M33 (Fig. 18), access to the existing large telescopes is required. With multi-object spectrographs (e.g. FLAMES, DEIMOS) becoming available, all the data required for developing extragalactic stellar astronomy into a mature discipline will be harvested in the next years.

Figure 18: Besides the Milky Way and the Andromeda galaxy (M31), M33 is the third large spiral in the Local Group, seen almost face-on. It is this characteristic that predestines this system for detailed studies of galactic properties (e.g. abundance gradients) as a function of galactocentric distance. Vigorous star formation takes place along the pronounced spiral arms, giving rise to numerous HII regions powered by massive, hot stars. As a consequence, a large number of blue supergiants is also found throughout this galaxy.

Of course, while these pages concentrate on our own work, such efforts cannot be undertaken by one person alone. Therefore feel free visit other members of our collaboration group. Thanks to:
Fabio Bresolin (IfA Hawaii), Wolfgang Gieren (Conception), Artemio Herrero (IAC Tenerife), Andreas Kaufer (ESO), Rolf-Peter Kudritzki (IfA Hawaii), Danny Lennon (ING, La Palma), the Munich Hot Star Group (USM), Paco Najarro (CSIC Madrid), Grzegorz Pietrzynski (Conception), Stephen Smartt (IoA Cambridge), Carrie Trundle (Queen's University Belfast), Miguel Urbaneja (IAC Tenerife), Kim Venn (Macalester College Saint Paul), Charo Villamariz (IAC Tenerife), and whoever contributed, but whom we forgot.