Rolf-Peter Kudritzki

Research Interests

- Extragalactic stellar astronomy

- Supergiant stars as tracers of chemical evolution and distances of galaxies

- Extragalactic distances

- Stellar spectroscopy

- Stellar atmospheres and radiative transfer

- Radiation driven winds of hot stars

Extragalactic Spectroscopy of Supergiant stars: Mass - Metallicity Relationship, Chemical Evolution and Distances of Galaxies



I use quantitative stellar spectroscopy of blue supergiant stars (BSG) and red supergiant stars (RSG) like Rigel and Betelgeuse in Orion to determine the chemical evolution of star forming galaxies. I also use these objects to measure distances to galaxies and, thus, to contribute to a precision determination of the Hubble constant. This is a close collaboration with many colleagues in Hawaii, the U.S. mainland, Chile and Europe: Miguel Urbaneja, Fabio Bresolin, Ben Davies, Maria Bergemann, Zach Gazak, Norbert Przybilla, Matt Hosek, Wolfgang Gieren, Gregorz Pietrzynski, Chris Evans, Lee Patrick, Betrand Plez, Iting Ho, Jabran Zahid, Ralf Bender, Stella Seitz, Andreas Burkert, Travis Berger and many others. In the following I discuss

a) optical spectroscopy of BSG in distant galaxies

b) the mass-metallicity relationship of star forming galaxies

c) distances to galaxies from spectroscopy of BSG

d) J-band spectroscopy of RSG and star super clusters (SSC) in distant galaxies

e) constraints on the chemical eveolution of galaxies from stellar metallicities and metallicity gradients




Blue supergiants (BSG) in the spiral galaxy NGC 3621 at a distance of 20 million light years. The BSG can be easily identified as individual targets. With modern telescopes such as Keck on Mauna Kea and the ESO VLT at Cerro Paranal it is possible to take spectra of these objects with excellent signal-to-noise.



An ESO VLT/FORS spectrum of a BSG in NGC 3621 compared with spectra of BSG in the solar neighborhood. There is no difference in quality. This clearly demonstrates that precise spectral diagnostics of such stars in distant galaxies is feasible.


The mass-metallicity relationship of star forming galaxies



Observed mass-metallicity relationship of galaxies from spectroscopic studies of blue supergiant stars. The red square belongs to M81. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



The same relationship as above, but now overplotted with the average relationships obtained by Kewley and Ellison (2008) obtained from strong HII-region emssion lines. The 10 different overplotted relationships belong to different strong-line calibrations. Obviously, the calibration uncertainty of strong lines introduces enormous uncertainties for this method. These uncertainties are poorly understood. Calibration (1) corresponds to Tremonti et al. (2004) and (7), (8) represent two calibrations by Pettini and Pagel (2004). Our science goal is to use extensive blue supergiant spectroscopy to develop an improved calibration of HII-region emission line strong line methods. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



The most recent result for the supergiant mass-metallicity relationship obtained from the quantitative spectral analysis of BSGs and RSGs (blue and red circles). Results from HII region studies using strong line diagnostics with different calibrations are shown as dashed curves. The green curve is the relationship found by Andrews and Martini (2013) and uses stacked SDSS galaxy spectra for each mass bin and the diagnostics of auroral lines. For details, see Bresolin, Kudritzki, Urbaneja et al., 2016, ApJ 830, 64.


Quantitative optical spectroscopy of blue supergiant stars (BSGs)



Wide field image of the Sculptor galaxy NGC 300 at 2 Mpc distance. We have studied the stellar content in much detail and have spectroscopically determined metallicity, metallicity gradient, extinction and distance from the blue supergiants marked by blue squares or circles.



NLTE model atmosphere spectra for one of the A supergiants in NGC 300. Different metallicities ranging from 1/20 solar to twice solar are assumed.



NLTE model atmosphere spectrum with a metallicity a factor of three smaller than solar compared with the observed spectrum of one of the A supergiant in NGC 300.



Stellar metallicity gradient in NGC 300. Note that this is the first determination of such a gradient based on the abundances of iron group elements. See Kudritzki, Urbaneja, Bresolin et al., 2008, ApJ 681, 269.


Precision distances to galaxies with BSG quantitative spectroscopy



BSG as precision extragalactic distance indicators. We use a relationship between absolute bolometric magnitude and stellar gravity log g and effective temperature Teff, which was predicted by Kudritzki, Bresolin and Przybilla, 2003, ApJ 582, L83. The figure shows this flux-weighted gravity - luminosity relationship of blue supergiants in galaxies in the Local Group and beyond. With gravities g and effectiv temperatures T determined from the spectrum this relationship can be used to determine precise extragalactic distances. Note that spectroscopy of the supergiant stars provides an independent determination of interstellar reddening and extinction. The method is, thus, free from uncertainties introduced by inadequate estimates of reddening. Details and theoretical background of the FGLR are described in Kudritzki, Urbaneja, Bresolin et al., 2008, ApJ 681, 269. A stellar evolution discussion of the FGLR is given in Meynet, Kudritzki, Georgy, 2015, Astron. Astrophys., 581,36.



Flux-weighted gravity - luminosity relationship of BSG in all galaxies studied so far out to a distance of 7.6 Mpc. See Kudritzki, Urbaneja, Gazak et al., 2013, ApJ Letters 779, L20 or Hosek, Kudritzki, Bresolin et al., 2014, ApJ 785, 151 for further details.



A new calibration of the FGLR using spectroscopy of BSG in the Large Magellanic Cloud. Details are explained in Urbaneja, Kudritzki, Gieren et al. 2017, submitted to ApJ (see preprints).


More examples of extragalactic BSG studies



Blue supergiant stars in the disk of the spiral galaxy M81 at 3.5 Mpc. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



M81 at 3.5 Mpc. Left: Selection of blue supergiant targets from the color-magnitude diagram. Right: Location of selected targets within M81. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.





Metallicity of blue supergiants in the disk of M81. Metallicity determination of object C20 as an example. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



Metallicity of blue supergiants in the disk of M81 as a function of galactocentric distance. We find slight super-solar metallicity in the center and a very shallow gradient. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



Metallicity of blue supergiants in the disk of M81 compared with planetary nebulae (PNe data from Stanghellini et al. 2010. The lower PN metallicity indicates strong chemical evolution of the disk over the last 5 Gyrs. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.





The Hertzsprung-Russell diagram (top) and the (log g, log Teff)-diagram of blue supergiant stars in M81. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



The distance to M81 from the observed FGLR compared to the Kudritzki et al. (2008) calibration. A distance modulus of m-M = 27.7 pm 0.1 mag is obtained. See Kudritzki, Urbaneja, Gazak et al., 2012, ApJ 747, 15.



A study of BSG in NGC 3621 at 6.5 Mpc



Hubble Space Telescope ACS images of BSG in NGC 3621. The circles correspond to 1 arcsec. Note that these objects are so bright that they shine out anything else. The fainter objects in these stamp images are more than four magnitudes fainter. Their brightness corresponds to Cepheid stars. This is one reason why BSG are potentially better distance indicators than Cepheids. Below we show spectral fits of the metal lines in the spectra of two BSG.







Examples of the model atmosphere fits of metal lines for 2 NGC 3621 targets. Note that we determine true metallicity not only oxygen abundance, as one does with HII regions. See Kudritzki, Urbaneja, Bresolin et al. 2014, ApJ 788, 56 for details.



Determination of the distance to NGC 3621 using the FGLR-method. See Kudritzki, Urbaneja, Bresolin et al. 2014, ApJ 788, 56 for details.



The "maser galaxy" NGC 4258 at 7.6 Mpc, an important anchor point of the extragalactic distance scale.



A BSG in the "maser galaxy" NGC 4258 and preparations for a spectroscopic study using the Keck 1 telescope on Mauna Kea and the LRIS spectrograph. See Kudritzki, Urbaneja, Gazak et al., 2013, ApJ Letters 779, L20 for details .



A BSG determination of stellar metallicity in the "maser galaxy" NGC 4258. This is fundamental for using this galaxy as an extragalactic distance scale anchor point. See Kudritzki, Urbaneja, Gazak et al., 2013, ApJ Letters 779, L20 for details .


J-band spectroscopy of red supergiant stars (RSG)



Now I turn to red supergiants (RSG) as Betelgeuse in Orion. These are the brightest stars in the universe at infrared light. Together with my colleagues Ben Davies and Zach Gazak I have developed a method to use low resolution spectra in the near-infrared J-band to determine their chemical composition. In this way, we can use RSG as cosmic abundance probes in distant galaxies.



We tested the method in the nearby massive star association Per OB1 which comprises the massive double star cluster h and chi Persei and contains many RSG.



Spectral fit of near-IR J-band spectra of RSG in Per OB1. The metallicity obtained is similar to the sun and agrees with the metallicity obtained from massive hot stars of spectral type B and A and also BSG. For details see Gazak, Davies, Kudritzki et al., 2014, ApJ 788, 58. The spectral analysis is based on detailed non-LTE radiative transfer calculations which are described in the papers by Bergemann, Kudritzki et al., 2012, 2013, 2015 published in ApJ.



The multi-IFU spectrograph KMOS at the ESO VLT is an ideal instrument for our extragalactic RSG studies. We have done a pilot study in the spiral galaxy NGC 300 at 2 Mpc, for which we have also done an investigation of BSG.



Selection of RSG in NGC 300 from color magnitude diagrams.



Location of the RSG studied with KMOS/VLT in NGC 300. The size of a KMOS field is indicated by the grey circle. The rectangles indicate HST fields.



ESO VLT KMOS spectra of RSG in NGC 300 and spectral fits with Non-LTE calaculations. For details see Gazak, Kudritzki, Evans et al., 2015, ApJ 805, 182.



Stellar metallicities as a function of galactocentric distance in NGC 300. Blue symbols are the BSG from the study by Kudritzki et al., 2008, shown above. The black symbols are the new results from near IR J-band spectroscopy of RSG. The two regression curves are almost identical and indicate excellent agreement. This cleary demonstrates the accuracy and power of extragalactic stellar astronomy. This research field has now truly matured to become a precision tool to investigate the chemical evolution of galaxies.



Same as above but the metallicities of HII_regions analyzed with collisionally excited auroral lines included. There is an 0.1 dex offset of the auroral line metallicities, which is found in nearly every galaxy studied with stellar and HII ayuroral line techniques. This is very likely the result of oxygen in HII-regions depleted by dust formation.



The near IR cryogenic multi-object spectrograph MOSFIRE at the Keck telescope on Mauna Kea is the other instrument for our extragalactic RSG chemical evolution studies.



First J-band spectra of RSG in the Andromeda galaxy M31 obtained with Keck/MOSFIRE and first determinations of metallicities.


J-band spectroscopy of super star clusters (SSC)



Super Star Clusters are another outstanding tool for chemical evolution studies of galaxies



The J-band light of SSCs is dominated by the red supergiants. 95% of the J-band light comes from RSGs as soon as the clusters are older than 7 million years.



A simulation of the spectral energy distribution of a young SSC and a SSC at the age of 15 million years. For clusters older than 7 million years the IR part of the SSC spectral energy distribution is dominated by RSG. See Gazak, Davies, Bastian, Kudritzki et al., 2014, ApJ 787, 142.



A simulation of the J-band spectrum of a SSC at the age of 15 million years. The contribution from RSGs clearly dominates. This is the combined light of fifty to hundred RSGs for a cluster with hundred thousand solar masses. Thus, SSCs are tremendously bright and can be studied specroscopically at enormous distances. See Gazak, Davies, Bastian, Kudritzki et al., 2014, ApJ 787, 142 and Gazak, Davies, Kudritzki et al., 2014, ApJ 788, 58 for details.



Using the spectra of RSG in Per OB1 of our previous study to simulate a an integrated SSC spectrum. The analysis gives the same metallicity as for the individual RSG in Per OB1. This demonstrates that the analysis of integrated SScC spectra is an accurate and p[owerful tool. From Gazak, Davies, Kudritzki et al., 2014, ApJ 788, 58.



Super Star Clusters in the central region of the grand spiral M83



A J-band pilot study of SSCs in M83 and NGC 6946 published in Gazak, Davies, Bastian, Kudritzki et al., 2014, ApJ 787, 142.



SSC metallicity in M83 (Gazak et al., 2015) compared with BSG metallicity (Bresolin, Kudritzki et al., 2016, ApJ 830, 64) as a function of galactocentric radius. Excellent agreement is found confirming the power of extrgalactic stellar spectroscopy.



SSCs in the Antennae galaxies at 20 Mpc distance observed with KMOS at the ESO VLT.



KMOS/VLT J-band spectra of SSCs in the Antennae galaxies at 20 Mpc and a first determination of metallicity (Lardo, Davies, Kudritzki et al., ApJ 812, 160).



The enormous potential of J-band RSG spectroscopy reaching far beyond the Coma cluster of galaxies.


The chemical evolution of galaxies: The measurement of the spatial distribution of metallicity over the disks of star forming galaxies and constraints on galactic winds and accretion from the cosmic web




The spectroscopic measurement of the distribution of metallicity over the disks of star forming galaxies and can be used to constrain the evolution of galaxies and the mass-loss through galactic winds and the mass-gain through accretion from the cosmic web. We apply a simple chemical analytical chemical evolution model, which predicts the spatial distribution of metallity as a function of stellar and interstellar medium (ISM) gas mass column densities. The model has two free paramters, the mass-gain through infall and the mass-loss through outflow, both measured in units of the star formation rate. Using the spectroscopically measured metallicity distribution and the observed stellar mass and ISM gas mass column density distributions we can constrain accretion mass gain and galactic wind mass-outflow. The details are described in Kudritzki, Ho, Schruba et al., 2015, MNRAS 450, 342.


Example: the spiral galaxy NGC 2403


An image of the NGC2403, the companion galaxy of M81.



The observed spatial distributions of ISM neutral and molecular gas mass column densities



The observed spatial distributions of stellar and and total ISM gas mass column densities



Model fit (blue) of the observed (black) galactocentric metallicity distribution. The best fit values for accretion mass-gain and galactic wind mass-loss are indicated on the right. The red curve is a model calculations for a "closed box" model, which ignores accretion and winds.



Fit isocontours constraining accretion mass-gain and galactic wind mass-loss.



Accretion and mass-loss for eleven well constrained galaxies.



A new result from the spectroscopic analysis of 58 blue supergiants in the Sculptor group galaxy NGC 55 (see Kudritzki, Castro, Urbaneja et al., 2016, ApJ 829, 70).



Example model atmosphere fits of BSG line absorption line spectra to determine metallicity.



NGC 55 (red circle) has an extreme value of accretion when compared with the other galaxies of the Kudritzki et al., 2015 study.



There is an anti-correlation of accretion in units of star formation rate with stellar mass.



Since star formation rate increases with stellar mass along the "galaxy main sequence", this means that accretion is only weakly dependent on stellar mass.



Galaxies surrounded by huge neutral hydrogen disks.


Many galaxies are surrounded by huge disks of neutral hydrogen. In these outer disks the star formation rates are very low and the metallicity distribution does not show a gradient.



Bresolin, Kudritzki, Urbaneja et al., 2016, ApJ 830, 64 studied the metallicity distribution of the grand spiral M83 to understand the transition from an inner to an outer disk.



Galactocentric metallicity dsitribution in the disk of M83 obtained from spectroscopy of BSG, SSC and HII-regions.



Observed mass column density distributions for stellar and ISM gas mass used for the chemical evolution model.



Final chemical evolution model reproducing the observed metallicity distribution (orange). The model assumes strong central galactic wind outflows which then leads to metal enriched accretion in the outer disk. The closed box model (blue-dashed), which ignores outflow and accretion, fails badly in the outer disk. For details see Bresolin, Kudritzki, Urbaneja et al., 2016, ApJ 830, 64.



Some older theoretical work ....
  • Winds from Hot Stars. An Introduction
  • Winds, ionizing fluxes and spectra of the first generation of very massive stars in the early universe.
  • back to homepage
  • Last revision: January 22, 2017