Quantitative Spectroscopy of Supergiants in the Local Group Dwarf Galaxy IC 1613: Metallicity and Distance
Travis A. Berger, Rolf-Peter Kudritzki, Miguel A. Urbaneja, Fabio Bresolin, Wolfgang Gieren, Grzegorz Pietrzyński, Norbert Przybilla
DDraft version May 21, 2018
Preprint typeset using L A TEX style AASTeX6 v. 1.0
QUANTITATIVE SPECTROSCOPY OF SUPERGIANTS IN THE LOCAL GROUP DWARF GALAXY IC 1613:METALLICITY AND DISTANCE
Travis A. Berger , Rolf-Peter Kudritzki , Miguel A. Urbaneja , Fabio Bresolin , Wolfgang Gieren ,Grzegorz Pietrzy´nski , & Norbert Przybilla Institute for Astronomy, University of Hawaii, 2680 Woodlawn Drive, Honolulu, HI 96822 Institut f¨ur Astro- und Teilchenphysik, Universit¨at Innsbruck, Technikerstr. 25/8, 6020 Innsbruck, Austria Departamento de Astronom´ıa, Universidad de Concepci´on, Casilla 160-C, Concepci´on, Chile University Observatory Munich, Scheinerstr. 1, D-81679 Munich, Germany Millennium Institute of Astrophysics, Santiago, Chile Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, ul. Bartycka 18, PL-00-716 Warszawa, Poland
ABSTRACTWe present the spectral analysis of 21 blue supergiant stars of spectral type late B to early A withinthe Local Group dwarf galaxy IC 1613 based on VLT-FORS2 low resolution spectra. Combiningour results with studies of early B type blue supergiants we report a wide bi-modal distribution ofmetallicities with two peaks around [Z] ∼ –0.50 dex and [Z] ∼ –0.85 dex. The bi-modal distributioncorrelates with spatial location, when compared with column densities of neutral hydrogen in IC1613. While the low [Z] objects appear in regions of relatively high ISM HI column densities or closeto them, the high [Z] supergiants are found in the central HI hole almost devoid of hydrogen. Thissuggests varied chemical evolution histories of the young stellar populations in IC 1613. Utilizing theFlux-Weighted Gravity-Luminosity Relation (FGLR), we determine IC 1613’s distance modulus as m − M = 24.39 ± Keywords: galaxies: distances and redshifts, galaxies: evolution, galaxies: individual (IC 1613), stars:fundamental parameters, supergiants INTRODUCTIONBecause of their extreme brightness, individual bluesupergiant stars can be studied spectroscopically ingalaxies out to ∼
10 Mpc. With multi-object spectro-graphs attached to 8 to 10m telescopes we can deter-mine stellar parameters ( T eff , log g , metallicity, etc.) ofindividual objects, and, with further analysis, proper-ties of the galaxies themselves, such as metallicities andmetallicity gradients. This analysis is especially impor-tant when trying to understand nearby galaxies’ chemi-cal evolution and star formation history (Kudritzki et al.2015, 2016; Bresolin et al. 2016). In addition, high preci-sion distances to these galaxies can be obtained throughthe flux-weighted gravity-luminosity relation (FGLR)method (Kudritzki et al. 2003), which uses the blue su-pergiant temperatures and gravities obtained via spec-tral analysis (Kudritzki et al. 2008, 2016).IC 1613 is a dwarf irregular galaxy and a memberof the Local Group at a distance of about 720 kpc(Pietrzy´nski et al. 2006). Its metallicity and chemicalevolution have been the subject of controversy. Photom-etry of the intermediate age population resulted in a log- arithmic metallicity relative to the sun of [Z] ∼ –1.3 dex(Cole et al. 1999; Rizzi et al. 2007). Kirby et al. (2013),through the analysis of red giant spectra, report [Fe/H]= –1.19 ± ± α -elements silicon, magne-sium and oxygen) of [Z] = –0.82 ± ∼ –0.7 dex. Such results have a r X i v : . [ a s t r o - ph . GA ] M a y led to speculation that the ratio of α to iron elements([ α /Fe]) of the young stellar population in this galaxyis smaller than the solar value, as already indicated inTautvaiˇsien˙e et al. (2007). In this paper, the authorsinvestigated three red supergiants of spectral type Min IC 1613 and found [Z] ∼ –0.67 dex and [ α /Fe] ∼ –0.1. If the higher metallicity based on the inclusion ofiron group elements were confirmed, this would have in-teresting repercussions for the evolution history of thisgalaxy. It would also solve the puzzle that the observedmassive stellar winds in this galaxy seemed to be toostrong for the very low metallicity as derived from H IIregions (Tramper et al. 2011, 2014).In this paper, we analyze ESO VLT FORS2 high qual-ity ( S/N ∼
70 or better) spectra of 21 late B–early Asupergiants (ASG from here on). The major motivationfor this work is to obtain more information about themetallicity of the young stellar population in IC 1613. Inparticular for the cooler objects among the ASG, opticalmetal line spectra are dominated by iron group elements,which will yield metallicities also based on iron group el-ements. We also determine the distance to IC 1613 usingthe FGLR method and compare with Cepheid and tipof the red giant branch (TRGB) distances to test theFGLR at low metallicity. OBSERVATIONS AND ANALYSIS METHOD2.1.
Spectra and Photometry
The ASG observations took place on the nights of 2003October 26 and 27 as part of the comprehensive spec-troscopic survey of massive stars in IC 1613 by Bresolinet al. (2007). The European Southern Observatory VeryLarge Telescope UT4 with the Focal Reducer and LowDispersion Spectrograph 2 (FORS2) was used and threefields were observed with 19 movable slitlets each 1”wide and 21” long. The spectra cover the range from3700–5900˚A with a spectral resolution of 5˚A (FWHM).For all details of the observations and their reductionswe refer to Bresolin et al. (2007), which also containsa discussion of the spectra including spectral classifica-tion.In addition, as already discussed in the introduction,Bresolin et al. (2007) carried out a quantitative spectralanalysis of nine early-type BSG but did not include thecooler objects in their analysis. In this paper, we nowconcentrate on the analysis of remaining cooler ASG.The list of objects studied is given in Table 1. The objectidentification is taken from Bresolin et al. (2007). Threeobjects, which were originally included in our spectralanalysis, are not listed in Table 1: objects C4 and C13have stellar gravities (log g ) too high for the grid of non-LTE spectra used for the analysis (see below) and objectA10 has an effective temperature ( T eff ) too large for the model grid.Our analysis also requires photometry in order to de-termine interstellar reddening, extinction and apparentbolometric magnitudes, which are then used for a deter-mination of distance through the FGLR method. Forthis purpose we use the B- and V-band photometry ob-tained by Garcia et al. (2009) with the Wield Field Cam-era at the La Palma Isaac Newton Telescope. The pho-tometric data are also provided in Table 1.2.2. Analysis Method
We follow the same procedure described in Kudritzkiet al. (2013, 2016) and Hosek et al. (2014) to deter-mine the stellar parameters T eff and log g and stellarmetallicity ([Z]), defined as [Z] = log Z/Z (cid:12) , where Z (cid:12) is the solar metallicity. We utilize a grid of LTE line-blanketed atmospheres with synthetic spectra obtainedwith non-LTE line formation calculations using the elab-orate atomic models described in Przybilla et al. (2006).The grid is comprised of temperatures from 7900–15000K, while log g varies between 0.8 and 3.0 dex in cgs units(the exact upper and lower limits depend on T eff ). [Z]ranges from –1.30 to 0.50 dex. Details of the grid aredescribed in Kudritzki et al. (2008, 2012). Figure 1 inKudritzki et al. (2008) illustrates the extent and spacingof the model atmosphere grid.The synthetic spectra are then compared with thenormalized observed spectra to obtain stellar parame-ters and metallicity. In a first step we use the observedBalmer lines H , , , , , , to constrain log g as a func-tion of T eff . At fixed values of T eff we determine thegravity which fits the Balmer lines best. Figure 1 pro-vides some examples of the best-fit model spectra fordifferent targets’ Balmer lines. log g is usually deter-mined with an accuracy of 0.05 to 0.10 dex at a fixedvalue of T eff (Table 1 provides the corresponding val-ues for the individual targets). Since the strength ofthe model Balmer lines decreases with increasing tem-perature, the log g fit values increase with temperature.Figure 2 shows two examples of the relationships be-tween T eff and log g obtained in this way. Next, wemove along the gravity-temperature relationship deter-mined from the Balmer lines and compare observed andcalculated flux as a function of metallicity in 11 spectralwindows, which are dominated by metal lines. We assessthe quality of the fit by calculating a χ value at each T eff and [Z]. Then, we calculate the minimum of χ andisocontours in ∆ χ around the minimum, which thenprovide a determination of T eff and [Z] together with anestimate of the uncertainties. We obtain log g from thegravity-temperature relationship determined in the firststep. Figure 3 shows the corresponding ∆ χ isocontoursfor two targets. We emphasize that through the χ fitof many spectral lines in many spectral windows, the re- Figure 1 . Balmer line fits for targets A1 (top) and C1 (bottom). The red curve represents the normalized object spectrum,while the black curve is a normalized synthetic spectrum. The synthetic spectra were computed for T eff = 8750 K, log g =1.90 dex, [Z] = –0.50 dex (target A1) and T eff = 8300 K, log g = 1.60 dex, and [Z] = –0.85 dex (target C1). The x-axis is thedisplacement from the line center measured in ˚A. eff [10 K]2.42.22.01.81.61.4 l og g ( c g s ) eff [10 K]2.42.22.01.81.61.4 l og g ( c g s ) Figure 2 . Fit curves of the Balmer lines in the gravity-temperature plane for targets A1 (left) and C1 (right). Along thesecurves the calculated Balmer lines agree with the observations. The red squares correspond to the fit parameters of Figure 1. eff /10000K − − − − − − [ Z ] eff /10000K − − − − [ Z ] Figure 3 . Determination of T eff and [Z] from isocontours ∆ χ in the metallicity-temperature plane obtained from the comparisonof synthetic with observed spectra (see text). Plotted are ∆ χ = 3 (red), 6 (blue), and 9 (black), respectively. Left : target A1; right : Target C1.
Figure 4 . Examples of metal line fits for targets C1 (top) and A1 (bottom) in three spectral windows using the final modelparameters. Each of the metal lines are labeled accordingly. The black curve represents the normalized object spectrum, whilethe red curve is a normalized synthetic spectrum. sultant stellar metallicity reflects the contribution frommany elements, including the iron peak and α -elements.Figure 4 shows the fits obtained for some of metal linewindows for two targets.Once temperature, gravity and metallicity are ob-tained, we use the synthetic non-LTE spectral energydistribution (SED) of the final model to calculate intrin-sic colors B-V and determine the interstellar reddeningE(B-V). By assuming R V = A V /E(B-V) = 3.3 for theratio of extinction to reddening we then constrain thevalue of extinction A V . Apparent bolometric magni-tudes m bol for each star are then obtained from the de-reddened V-magnitudes and the bolometric correctionBC calculated from the SED. The values of T eff , log g ,[Z], E(B-V), BC and m bol for each object analyzed aregiven in Table 1. The table also contains the stellar fluxweighted gravity defined as log g F = log g – 4log T eff +16, which we will use for a determination of distances.The results compiled in the table will be discussed inthe following sections. BLUE SUPERGIANTS OF EARLY SPECTRALTYPEFor the discussion of metallicities and the determina-tion of distance we will include the BSG in IC 1613 stud-ied by Bresolin et al. (2007), which were observed withthe same spectroscopic setup at the same time as ourASG and then analyzed in detail using non-LTE modelatmospheres. Most recently, Camacho (2017, PhD the-sis) and Garcia, Camacho, Herrero et al. (2018, to bepublished, see Section 1) have extended this work us-ing VIMOS multi-object spectroscopy obtained with theESO VLT. They re-analyzed most of the targets investi-gated by Bresolin et al. (2007) but also added additionalobjects to the sample. For the objects in common theresults are very similar to Bresolin et al. (2007). In ourcomparison we use the results by Camacho, but add twomore targets (A7 and A18) from Bresolin et al. (2007),which were not included in the Camacho study. Thework by Camacho uses the same photometry as we havebeen using for the determination of reddening and bolo-metric magnitudes (Garcia et al. 2009). For the twotargets added from Bresolin et al. (2007) we have re-determined reddening and bolometric magnitudes againusing the photometry by Garcia et al. (2009) for consis-tency. INTERSTELLAR REDDENINGThe measurement of interstellar reddening is crucialfor the determination of accurate distances. As is wellknown, purely photometric stellar distance determina-tion methods can be affected significantly by interstellardust. Therefore, a fundamental advantage of the quan-titative spectral analysis of blue supergiant stars is that
Figure 5 . Histogram of the distribution of interstellar red-dening E(B-V) of the ASGs in Table 1. it provides direct information about interstellar redden-ing. For an actively star forming galaxy such as IC 1613,the massive stars can be imbedded in a dusty environ-ment so we expect a wide range of interstellar reddening.Figure 5 displays the distribution of E(B-V) among theASG targets of Table 1. Indeed we find a wide distribu-tion with reddening values larger than the Milky Wayforeground value of 0.03 mag (Pietrzy´nski et al. 2006).The average value is 0.10 mag. A similar average value,0.09 mag, has been found by Pietrzy´nski et al. (2006)by comparing distance moduli obtained from period–luminosity relationships of Cepheids at different wave-lengths ranging from the NIR to optical wavelengths,but no individual values could be determined. The dis-tribution of reddening values agrees well with the com-prehensive photometric study by Garcia et al. (2009). EVOLUTIONARY STATUSThe spectroscopic determination of the ASG stellarparameters summarized in Table 1 and of the BSG stud-ied by Camacho (2017) and Bresolin et al. (2007) allowus to discuss the evolutionary status of these blue su-pergiants. A good direct approach for such a discus-sion is the “spectroscopic Hertzsprung-Russell diagram”(sHRD) introduced by Langer & Kudritzki (2014).The sHRD displays flux-weighted gravities against ef-fective temperature and is equivalent to the classicalHertzsprung-Russell diagram (HRD) except that it isindependent of the distances adopted. Figure 6 showsthe sHRD for our objects and demonstrates very nicelythat the blue supergiants are in an evolved stage ofstellar evolution. The hotter BSG objects are some-what younger and closer to the original zero-age-main-sequence and the cooler ASG objects are more advancedin their evolution. We also see that most of the coolerASG objects have lower initial zero-age-main-sequence eff [K]2.52.01.51.0 l og g F ( c g s ) Figure 6 . Spectroscopic Hertzspung–Russell diagram of bluesupergiant stars in IC 1613. Plotted is the flux weightedgravity log g F versus the logarithm of T eff . The ASG fromthis study are shown as red squares for metallicities [Z] largerthan –0.7 and as red circles for metallicities lower than orequal to –0.7 (see discussion in § (cid:12) (from the bottom of the figure totop) are also displayed. The calculations of the tracks includethe effects of rotation with initial main-sequence rotationalvelocities of about 180 km/sec. masses than the hotter BSG counterparts in the sHRD.This is the result of a selection effect related to the V-band magnitude limit of the spectroscopic observationsand the much higher bolometric correction of the ear-lier spectral types. More broadly, Figure 6 demonstratesthat the early type and late type objects belong to thesame stellar population, although there is a small agedifference of ∼
20 million years between the hotter, moremassive objects and the cooler, less massive objects ac-cording to the stellar evolution calculations by Brottet al. (2011), which are also illustrated in the figure. METALLICITYThe average metallicity of the total sample of allblue supergiants, ASG and BSG, is [Z] = –0.69 ± ± C o un t Figure 7 . Distribution of blue supergiant metallicities [Z] inIC 1613.
Upper : the 21 ASG, middle : the 13 early typeBSG, and bottom : the total sample. licity derived from the analysis of red giants is signifi-cantly smaller than the blue supergiant value, red gi-ants are, on average, significantly older than blue su-pergiants. Therefore, we expect such discrepancies inmetallicity between these two populations.Figure 7 shows the distribution of metallicities [Z].The result is surprising, because the [Z] distributions ofthe ASG and the hot BSG are significantly different.While the hot objects are narrowly distributed with aclear peak at [Z] ∼ –0.85, the ASG show an additionalcomponent peaking a higher metallicity of [Z] ∼ –0.50.We stress that this difference can hardly be the result ofdifferent stellar age.The bimodal ASG metallicity distribution is an inter-esting result in need of further investigation. For [Z] ≤ –0.7 it is similar to the hot BSG [Z] distribution,but there is an additional component with metallicitieshigher than –0.7. In Figure 8 we investigate whetherthe occurrence of this higher metallicity component isrelated to the spatial location within the galaxy. We Figure 8 . Spatial distribution of blue supergiants in IC 1613. The 21 ASG are plotted either as blue circles ([Z] ≤ –0.70) orred crosses ([Z] ≥ –0.70). The 13 BSG, which all have metallicities lower than [Z] = –0.7, are displayed as blue triangles. Inaddition, we overplot the HI regions from Silich et al. (2006) in translucent grey. Darker color corresponds to higher HI columndensities. plot both groups, BSG and ASG, but distinguish be-tween metallicities higher and lower than [Z] = –0.7.The result is striking. The higher metallicity objectsare concentrated in the central region of the galaxy andspatially separated from supergiants of lower metallicity.Most interestingly, there is also an indication of a cor-relation with the existence of neutral HI ISM gas. Weoverplot the extent of the HI regions observed by Silichet al. (2006) within IC 1613 as translucent, grey con-tours where darker colors indicate higher column densi-ties. This reveals a significant pattern: the higher metal-licity objects appear to cluster in the central areas of lowHI column densities, while the lower metallicity objectstypically appear within the ring-like regions of higher HIcolumn densities or close to them.From the viewpoint of chemical evolution the inter-pretation is straightforward. In the HI gas-depleted re-gions, the star formation process has proceeded more rapidly and has consumed the HI gas to form stars,which, through supernova explosions and stellar winds,have enriched the metallicity in these regions. Thus,there appear to be two populations of ASGs in IC 1613:those born from the higher metallicity ashes of stars inthe center of the galaxy, and those born from the lowmetallicity HI region gas and dust in the outer portionsof the galaxy. Although both the metal-poor HI regionsand the central, metal-rich portions of the galaxy showsigns of recent star formation, they remain separated intheir metallicities, indicative of different chemical evo-lution histories. FGLR AND DISTANCE DETERMINATIONKudritzki et al. (2003, 2008) have shown that the flux-weighted gravity log g F = log g – 4log T eff + 16 of bluesupergiant stars is tightly correlated with their absolutebolometric magnitude. As explained in their work, this F (dex)201918171615 m bo l ( m ag ) Figure 9 . FGLR for IC 1613. The green line represents theLMC calibration by Urbaneja et al. (2017) fitted to the datain a simple least square procedure to obtain the distancemodulus. The red points correspond to ASGs, while theblue points represent the BSGs. is because blue supergiants cross the HRD at roughlyconstant luminosity and mass as they evolve away fromthe main sequence (see the evolutionary tracks of Fig-ure 6 for some examples). This correlation, the FGLR,qualifies blue supergiants as excellent distance indica-tors and has already been used to determine the dis-tances to many galaxies (Kudritzki et al. 2012, 2014,2016; Bresolin et al. 2016; Hosek et al. 2014; Urbanejaet al. 2008; U et al. 2009).Combining the flux-weighted gravities and apparentbolometric magnitudes of the ASG and BSG, we obtainthe observed blue supergiant FGLR of IC 1613 in Figure9. In order to determine the distance to IC 1613, we usethe new FGLR calibration by Urbaneja et al. (2017)resulting from a detailed spectroscopic analysis of 90blue supergiants in the Large Magellanic Cloud (LMC): M bol = a (log g F − .
5) + b (1)if log g F ≥ log g break F , and M bol = a low (log g F − log g break F ) + b break (2)if log g F ≤ log g break F , with b break = a (log g break F − .
5) + b, (3)where log g break F = 1.30 dex, a = 3.20 ± b = –7.90 ± a low = 8.34 ± g F values. This yields individual distance moduli for each object. Thedistance is then determined from a weighted mean thataccounts for the observational errors of log g F and m bol .We obtain a distance modulus of m − M = 24.39 ± ± ±
43 kpc. Here, the systematic error accounts forthe uncertainties of the calibration FGLR parameters.The green curve in Figure 9 represents the calibrationFGLR (Urbaneja et al. 2017) shifted to our determineddistance.The FGLR distance is in agreement with the value of m − M = 24.29 ± m − M = 24.38 ± m − M = 24.38 ± m − M = 24.37 ± σ = 0.6 mag is larger than thetypical values of σ = 0.25–0.40 mag encountered forother galaxies. As the stellar evolution simulations ofthe FGLR by Meynet et al. (2015) indicate (comparetheir Figures 8 and 10), this might be an effect of thelow metallicity at which the effects of rotational mixingare more pronounced and lead to a significant devia-tion from constant luminosity for rotating stars whenthey cross the HRD (Georgy et al. 2013). Conversely,the FGLR of blue supergiants in the metal poor galax-ies WLM (Urbaneja et al. 2008) and NGC 3109 (Hoseket al. 2014) shows a scatter comparable to galaxies withhigher metallicity, when a fit using the new calibration ofEquations (1) to (3) is applied (Kudritzki et al., 2018, inpreparation). We also note that the evolutionary tracksby Brott et al. (2011) for low metallicity stay at roughlyconstant luminosity when crossing the HRD and do notshow the behavior found in the work by Georgy et al.(2013). DISCUSSION AND FUTURE WORKThe results obtained in the combination of this workand of the investigation by Camacho and collaboratorsindicate a spatially inhomogeneous chemical evolutionand star formation history of IC 1613. This is in agree-ment with the detailed study by Bernard et al. (2007)based on resolved-star V- and I-band photometry carriedout with the 2.5m Isaac Newton Telescope. This study0revealed that over the last 0.15–2 Gyr the star forma-tion rate in the central region, now void of ISM neutralhydrogen, has been significantly higher than in the sur-rounding ring-like structure of higher hydrogen density.Then, over the last 100 Myr, very likely because of theconsumption of central ISM hydrogen gas, star forma-tion in the surrounding ring has become larger than inthe central region. The reduced ISM hydrogen densityin the center may simply be the result of enhanced starformation. However, dynamical processes such as stellarwinds and supernovae explosions may have contributedto the reduced ISM hydrogen density, too, as the com-plex velocity and density structure of IC 1613 indicates(Valdez-Guti´errez et al. 2001; Lozinskaya et al. 2003;Silich et al. 2006).Enhanced star formation, which consumes interstellargas to form stars, accelerates chemical evolution throughthe death of massive and intermediate mass stars. Themetallicity of the ISM and the young stellar populationis then a function of the ratio of stellar mass to gas mass.The HI surface density in the central cavity is about afactor of two lower than the average surface density inthe surrounding ring structure, as the isocontours shownin Lake & Skillman (1989) indicate. On the other hand,the stellar surface density is slightly enhanced in thecentral region (Bernard et al. 2007) so that the ratio ofstellar mass to gas mass is higher by about a factor ofthree to four in the center compared to the ring. Aschemical evolution models show, see for instance Ku-dritzki et al. (2015) Figure 1, such a difference in theratio of stellar to gas mass can easily explain the dif-ference in metallicity of 0.35 dex (between the high andlow metallicity ASG) as an effect of chemical evolutioncaused by enhanced star formation.Of course, this interpretation needs to be checked in-dependently, ideally through spectroscopy of other typesof massive stars or HII-regions in the central neutral hy-drogen cavity. So far, all the O-stars (Garcia et al. 2014;Tramper et al. 2014; Bouret et al. 2015) and M super-giants (Tautvaiˇsien˙e et al. 2007) studied by means of aquantitive spectral analysis are located in the ISM gas-rich parts of IC 1613. The same is true for the veryfew HII-regions, for which a direct determination of theoxygen abundance through a detection of auroral lineshas been possible (Bresolin et al. 2007). This work willbe challenging, though. There are many OB star associ-ations in this region (Garcia et al. 2009; Borissova et al.2004), but the members of these associations are rela-tively faint and long exposure times for optical or UV spectroscopy will be required. There are also many HII-regions (Valdez-Guti´errez et al. 2001), but as pointedout by Bresolin et al. (2007) the general problem ofquantitative HII-region spectroscopy in IC 1613 is thatthe surface brightnesses are low and the detection of au-roral lines is difficult. This will require a dedicated studywith very long exposure times, preferably with integralfield spectroscopy at large telescopes.An alternative and very promising spectroscopicmethod to obtain metallicities of massive young stars ismedium resolution J-band technique of red supergiantstars developed by Davies et al. (2010) and Gazak et al.(2014). This technique has already been applied to avariety of galaxies with very good results (Gazak et al.2015; Patrick et al. 2015, 2017). A differential studyof red supergiants in the HI cavity and the surroundingring area would be an excellent independent check of ourblue supergiant result.We note that in our discussion of star formation his-tory and chemical evolution we have only considered thespatial distribution of neutral hydrogen in the ISM. Ifthe central cavity would contain a significant amountof molecular hydrogen of much higher density than inin the surrounding ring, then our interpretation wouldhave to be revised. At this point, we are not aware ofdeep enough molecular gas observations with sufficientspatial resolution, which would allow us to disprove orto confirm our proposed scenario.The determination of a new independent distance toIC 1613 by means of the blue supergiant FGLR tech-nique based on the new calibration by Urbaneja et al.(2017) yields good agreement with the Cepheid and theTRGB methods. It confirms the result by Hosek et al.(2014) that the FGLR method also seems to work at lowmetallicity. This is an encouraging result and furtherestablishes the FGLR method as a new complementarytool to investigate the extragalactic distance scale.We thank the reviewer for the helpful and construc-tive comments. We also thank Ines Camacho, MiriamGarcia and Artemio Herrero for providing us with theirresults prior to publication and for a detailed discussionof our work. WG and GP gratefully acknowledge finan-cial support for this work from the BASAL Centro deAstrofisica y Tecnologias Afines (CATA) PFB-06/2007.WG also gratefully acknowledges financial support fromthe Millennium Institute of Astrophysics (MAS) of theIniciativa Cientifica Milenio del Ministerio de Economia,Fomento y Turismo de Chile, project IC120009.[h!]1
Table 1 . IC 1613 A SupergiantsStar T eff log g log g F [Z] V B - V E(B-V) BC m bol [K] cgs cgs dex mag mag mag mag mag(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)A1 8825 ±
175 1.95 ± +0 . − . -0.49 ± ± ± ± ±
70 1.62 ± +0 . − . -0.42 ± ± ± ± ±
70 1.40 ± +0 . − . -0.40 ± ± ± ± ±
250 1.90 ± +0 . − . -0.70 ± ± ± ± ±
500 2.67 ± +0 . − . -0.45 ± ± ± ± ±
70 1.72 ± +0 . − . -0.60 ± ± ± ± ±
150 1.14 ± +0 . − . -0.96 ± ± ± ± ±
200 1.64 ± +0 . − . -0.95 ± ± ± ± ±
200 1.87 ± +0 . − . -0.93 ± ± ± ± ±
350 2.49 ± +0 . − . -0.50 ± ± ± ± ±
200 1.90 ± +0 . − . -0.50 ± ± ± ± ±
350 2.56 ± +0 . − . -0.55 ± ± ± ± ±
70 1.60 ± +0 . − . -0.89 ± ± ± ± ±
250 1.86 ± +0 . − . -1.25 ± ± ± ± ±
50 1.60 ± +0 . − . -0.85 ± ± ± ± ±
400 1.95 ± +0 . − . -0.90 ± ± ± ± ±
100 1.71 ± +0 . − . -0.85 ± ± ± ± ±
350 1.70 ± +0 . − . -0.85 ± ± ± ± ±
150 1.42 ± +0 . − . -0.45 ± ± ± ± ±
200 2.12 ± +0 . − . -0.55 ± ± ± ± ±
500 2.68 ± +0 . − . -0.45 ± ± ± ± Note —Object identifications from Bresolin et al. (2007).
REFERENCES
Bernard, E. J., Aparicio, A., Gallart, C., Padilla-Torres, C. P., &Panniello, M. 2007, AJ, 134, 1124Borissova, J., Kurtev, R., Georgiev, L., & Rosado, M. 2004,A&A, 413, 889Bouret, J.-C., Lanz, T., Hillier, D. J., et al. 2015, MNRAS, 449,1545Bresolin, F., Kudritzki, R.-P., Urbaneja, M. A., et al. 2016, ApJ,830, 64Bresolin, F., Urbaneja, M. A., Gieren, W., Pietrzy´nski, G., &Kudritzki, R.-P. 2007, ApJ, 671, 2028Brott, I., de Mink, S. E., Cantiello, M., et al. 2011, A&A, 530,A115Cole, A. A., Tolstoy, E., Gallagher, III, J. S., et al. 1999, AJ,118, 1657Davies, B., Kudritzki, R.-P., & Figer, D. F. 2010, MNRAS, 407,1203Garcia, M., Herrero, A., Najarro, F., Lennon, D. J., & AlejandroUrbaneja, M. 2014, ApJ, 788, 64Garcia, M., Herrero, A., Vicente, B., et al. 2009, A&A, 502, 1015 Gazak, J. Z., Davies, B., Kudritzki, R., Bergemann, M., & Plez,B. 2014, ApJ, 788, 58Gazak, J. Z., Kudritzki, R., Evans, C., et al. 2015, ApJ, 805, 182Georgy, C., Ekstr¨om, S., Eggenberger, P., et al. 2013, A&A, 558,A103Graczyk, D., Pietrzy´nski, G., Thompson, I. B., et al. 2014, ApJ,780, 59Hosek, Jr., M. W., Kudritzki, R.-P., Bresolin, F., et al. 2014,ApJ, 785, 151Kingsburgh, R. L., & Barlow, M. J. 1995, A&A, 295, 171Kirby, E. N., Cohen, J. G., Guhathakurta, P., et al. 2013, ApJ,779, 102Kudritzki, R. P., Bresolin, F., & Przybilla, N. 2003, ApJL, 582,L83Kudritzki, R. P., Castro, N., Urbaneja, M. A., et al. 2016, ApJ,829, 70Kudritzki, R.-P., Ho, I.-T., Schruba, A., et al. 2015, MNRAS,450, 342Kudritzki, R.-P., Urbaneja, M. A., Bresolin, F., Hosek, Jr.,M. W., & Przybilla, N. 2014, ApJ, 788, 562