EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 686 (June 2024) A&A, 686 (2024) A77 Full HTML
Open Access
Issue
A&A
Volume 686, June 2024
Article Number A77
Number of page(s) 65
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/202348527
Published online 31 May 2024

© The Authors 2024

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

The role of episodic mass loss in evolved massive stars is one of the outstanding open questions facing stellar evolution theory (Smith 2014). While the upper limit to the masses of stars is thought to be 150 M⊙ (Figer 2005; Oey & Clarke 2005), with numerous claims of 200–300 M⊙ (based on R136a1; Crowther et al. 2010; Banerjee et al. 2012; Brands et al. 2022; Kalari et al. 2022), the masses of hydrogen-deficient Wolf-Rayet (WR) stars do not exceed 20 M⊙ (Crowther 2007). The theory of classical line-driven winds (Kudritzki & Puls 2000), once thought to be responsible for removing the envelopes of massive stars, has been shown inadequate, both on theoretical grounds (due to clumping; Owocki & Puls 1999) and estimations based on spectral lines (Bouret et al. 2005; Fullerton et al. 2006; Cohen et al. 2014), which require reductions in the mass-loss rates by a factor of ~2–3. One alternative is the interaction of binary systems via Roche-lobe overflow, which is predicted to occur in 70% of massive stars and strip the envelopes in ~30% of O stars in the Milky Way, given the high binary fraction of massive stars (~70%; Sana et al. 2012; Duchêne & Kraus 2013; Moe & Di Stefano 2017). Episodic mass loss is possibly the dominant process that operates in all massive stars (including those in binaries or binary products; e.g., Aghakhanloo et al. 2017; Beasor et al. 2020; Decin et al. 2024); however, the physical mechanism responsible for it remains a mystery (see, e.g., the review by Smith 2014).

The importance of episodic mass loss has come to the forefront in both the massive star and supernova (SN) communities. Spitzer images have revealed numerous circumstellar shells surrounding massive, evolved stars in our Galaxy (Gvaramadze et al. 2010; Wachter et al. 2010). Episodes of enhanced mass loss have been recorded not only in luminous blue variables (LBVs), but also in extreme red supergiants (RSGs; e.g., VY CMa, and Betelgeuse; Decin et al. 2006; Dupree et al. 2022) and yellow hypergiants (e.g., IRAS 17163-3907; Koumpia et al. 2020). Moreover, many studies show evidence of circumstellar material (CSM) present around most normal Type II-P SN progenitors at the moment of explosion (Förster et al. 2018; Morozova et al. 2018), hinting that there is an abrupt increase in the mass-loss rate that is synchronized with the explosion (Davies et al. 2022). Such high mass-loss rates of SN progenitors are considered a possible explanation of the “RSG problem” of not observing luminous RSGs as Type II-P SN progenitors (Smartt et al. 2009; Smartt 2015). In cases where the pre-SN, hydrogen-rich CSM is dense enough, its interaction with the SN ejecta can be the main energy source of the event that would be identified as a Type IIn (e.g., Smith 2017; Fox et al. 2020), for which LBVs have been identified as direct progenitors (e.g., Smith et al. 2014). This was the case of the remarkable SN2009ip progenitor that underwent a series of episodic mass-loss events, which were identified as “SN imposter” events, before its final collapse (e.g., Mauerhan et al. 2013; Smith et al. 2022). There is also evidence that at least a portion of superluminous SNe arise from interactions with CSM (e.g., Gal-Yam 2012), with mounting evidence suggesting that these events occur in low-metallicity host galaxies (Neill et al. 2011; Gal-Yam 2019, and references therein).

Models of single-star evolution have adopted empirical, constant mass-loss prescriptions (Meynet et al. 2015; Beasor et al. 2021) or, recently, time-averaged mass-loss rates (Massey et al. 2023), which highly influence the outcome. Recently, there have been improvements in the methodology of measuring RSG mass-loss rates depending on the stellar parameters and evolutionary phase; however, the resulting mass-loss rates differ significantly (Beasor et al. 2020; Yang et al. 2023; Decin et al. 2024).

The ASSESS project1 (Bonanos et al. 2023) aims to tackle the role of episodic mass loss in massive stars by using the fact that mass-losing stars form dust and are bright in the mid-IR. Physically, there are a number of ways a massive star can become a source of significant mid-IR emission. First, dust can form in a dense, but relatively steady, stellar wind. In the most extreme cases, such as in the progenitors of the SN 2008S and the NGC 300-OT 2008 transient (Bond et al. 2009), the wind is optically thick even in the near-IR and the source star is only seen in the mid-IR (Prieto et al. 2008). Second, a very massive star can have an impulsive mass ejection or eruption, with dust forming in the ejected shell of material. Initially, the optical depth and dust temperatures are high but then drop as the shell expands. The most famous example is the “great eruption” of η Carinae in the 19th century (Humphreys & Davidson 1994; Davidson & Humphreys 1997; Smith & Frew 2011), which ejected several solar masses of material. Third, the dust can be located in a circumstellar disk and emit over a broad range of temperatures, as observed in supergiant B[e] stars (sgB[e]) stars (Kraus 2019).

While stars with significant mid-IR emission are intrinsically rare, many of the most interesting massive stars, such as η Car or “Object X” in M33 (Khan et al. 2011; Mikołajewska et al. 2015), belong to this class. Therefore, we searched for analogs of these interesting stars using mid-IR photometry of nearby galaxies along with the existing mid-IR “roadmaps” for interpreting luminous massive stars (Bonanos et al. 2009, 2010), which are based on known massive stars in the Large and Small Magellanic Clouds (LMC and SMC). These studies have identified LBVs, sgB[e], and RSGs from the brightest mid-IR sources, due to their intrinsic brightness and due to being surrounded by their own dust. Previous searches for dusty, evolved massive stars using mid-IR selection criteria have validated the method (see, e.g., Britavskiy et al. 2014, 2015, 2019; Kourniotis et al. 2017).

The ASSESS team has collected published mid-IR photometric catalogs from Spitzer of nearby galaxies within 5 Mpc with high star-formation rates: (a) seven dwarf galaxies within 1.5 Mpc from the “DUST in Nearby Galaxies with Spitzer (DUSTiNGS)” project (Boyer et al. 2015b): IC 10, IC 1613, Phoenix, Peg DIG, Sextans A, Sextans B, and Wolf-Lundmark-Melotte (WLM); (b) 13 galaxies within 5 Mpc (Khan et al. 2015; Khan 2017): M31, M33, NGC 247, NGC 300, NGC 1313, NGC 2403, M81, M83, NGC 3077, NGC 4736, NGC 4826, NGC 6822, and NGC 7793; and (c) five galaxies within 4 Mpc (Williams & Bonanos 2016): NGC 55, NGC 253, NGC 2366, NGC 4214, and NGC 5253. The mid-IR photometry made available by the “Surveying the Agents of a Galaxy's Evolution (SAGE)” surveys of the LMC (Meixner et al. 2006) and SMC (Gordon et al. 2011) has also been searched for undetected, dust-obscured targets in our nearest neighbor galaxies. These catalogs contain mid-IR photometry of over 5 million point sources in 27 nearby galaxies, 19 of which have Pan-STARRS1 coverage (Chambers et al. 2016), providing an ideal dataset for a systematic study of luminous, dusty, evolved massive stars. We compiled mid-IR photometric catalogs for these galaxies, including their counterparts in Pan-STARRS1 (g, r, i, z, and y bands) and other archival photometric catalogs (see Sec. 2.1), to construct the spectral energy distributions (SEDs) of evolved stars in these galaxies out to 24 μm. The single epoch 5σ depth of Pan-STARRS1 ranges from the 22nd magnitude in the g band to the 20th magnitude in the y band, corresponding to absolute magnitudes brighter than −6 in g and −8 in y at 3.5 Mpc, which include the most luminous, evolved targets. de Wit et al. (2023) present the results so far of applying our methodology to the Magellanic Clouds, yielding eight new RSGs, a new yellow supergiant (YSG), a known sgB[e], and a candidate LBV (LBVc).

Based on these catalogs, we selected over 1000 luminous and red sources (selected according to their colors in [3.6]-[4.5]) from these 27 galaxies and are in the process of collecting follow-up low-resolution spectroscopy. In this paper, we present the spectroscopic data obtained from ten southern galaxies and the resulting catalog of evolved massive stars. In Sec. 2, we describe the target selection criteria and observations. In Sec. 3, we present the spectral classification procedure and results for each galaxy. In Sec. 4, we discuss the resulting catalog, and in Sec. 5 we summarize the results.

2 Observations and data reduction

The aim of our observational campaign was to spectroscopically identify dusty massive stars in nearby, star-forming galaxies. This section describes the construction of photometric catalogs, and how these were used to identify potential targets. We then describe the multi-object spectroscopic (MOS) observations with the Focal Reducer and Low Dispersion Spectrograph (FORS2) on the Very Large Telescope (VLT), and the data reduction process.

Table 1

Properties of target galaxies.

Table 2

Distribution of targets per galaxy and per photometric catalog.

2.1 Photometry catalogs

To identify dusty massive stars, we constructed multiwave-length photometric catalogs for all nearby galaxies listed above, for which Spitzer (Werner et al. 2004) IR point-source catalogs are available. Here we describe these catalogs for the ten galaxies targeted in this paper: WLM, NGC 55, NGC 247, NGC 253, NGC 300, NGC 1313, NGC 3109, Sextans A, M83, and NGC 7793. These galaxies were chosen as they were visible from Paranal and contained a sufficient number of targets (see Sec. 2.2) to justify MOS observations. Table 1 provides the properties of the galaxies, including the coordinates of each galaxy, the distance (from Cosmicflows-2; Tully et al. 2013), and metallicity, as well as the radius R (i.e., the size of the galaxy based on visual inspection) used to match the catalogs.

Spitzer point-source catalogs containing sources detected in the IRAC (Fazio et al. 2004) 3.6 and 4.5 μm bands formed the basis of our catalogs. Where available, these were supplemented by IRAC 5.8 and 8.0 μm and MIPS (Rieke et al. 2004) 24 μm magnitudes or their 3σ upper limits (see Table 2 for references). These catalogs were cross-matched with optical and near-IR photometry, as well as astrometric data from the following surveys:

  • Pan-STARRS1 (PS1): g (481 nm), r (617 nm), i (752 nm), z (866 nm), and y (962 nm),

  • VISTA Hemisphere Survey (VHS; McMahon 2012): J (1.25 μm), and Ks (2.15 μm),

  • Gaia DR2 (Gaia Collaboration 2016, 2018): Proper motion in right ascension (RA), declination, and parallax.

A radius of 1″ was used for cross-matching. Cases with multiple counterparts within 1″ were flagged and no additional photometry was added. These targets constitute 0.1% of the sample and were excluded from the priority target lists. For the sources that have a Gaia counterpart and a good astrometric solution (astro_excess_noise < 1 mas), we used the astrometric data to perform a conservative cleaning of foreground sources. Sources were flagged as foreground if they meet one of the following criteria:

  • pmra/pmra_error > 3.6,

  • pmdec/pmdec_error > 3.9,

  • parallax/parallax_error > 3.6.

These cuts were based on values determined for stars in M31 and M33 (three times the observed standard deviation; see Maravelias et al. 2022). Table 2 gives the total number of sources in each catalog, the number of matches in each of the additional photometric surveys, and the number of sources flagged as foreground contamination. It also provides the mean offset between coordinates from Spitzer and each of the other surveys.

Table 3

Definition of the priority system and the criteria for each priority class.

Table 4

Distribution of target priorities per galaxy.

2.2 Target selection

We used the multiwavelength photometric catalogs to construct target lists of potential dusty massive stars. The first criterion was an IR color of m3.6 − m4.5 ≥ 0.1 mag. This selected targets with an IR excess and excluded the majority of remaining foreground stars (which mostly have m3.6 − m4.5 ~ 0 mag). Additionally, an absolute magnitude criterion of M3.6 ≤ −9.0 mag (corresponding to log L/L⊙ ~ 4.5) and an apparent magnitude criterion of m4.5 ≤ 15.5 mag were applied. The former excluded the majority of asymptotic giant branch (AGB) stars (Yang et al. 2020), while the latter prevented strong contamination by background IR sources (i.e., background galaxies and quasars; Williams et al. 2015). However, a small fraction of contaminants remained in our target lists (see Sec. 4 for details).

The resulting target lists contain 1298 priority sources. We assigned priorities 1 to 6 (P1 to P6) to each of the targets (see Table 3), depending on their mid-IR colors, absolute magnitude criteria, and, finally, whether or not there are optical or near-IR counterparts. Priority 1 selected the brightest IR sources with a strong IR excess, which are most likely visible in the optical, while P2 targets are similar to P1 but with moderate IR excess. Table 4 presents the distribution of target priorities per galaxy for all the priority targets available (1298 sources), as well as the number and percentage of priority targets placed in slits and observed (384 targets or 30% of selected targets). We note that there is a selection effect, as the priority targets sample more luminous IR sources with increasing distance. Spitzer color-magnitude diagrams (CMDs) of the target lists indicating the priority of each target and the magnitude criteria are shown in Appendix A. The fields for MOS observations were chosen to maximize the number of P1 and P2 targets that can be observed. This resulted in a total of 20 fields in the ten galaxies. In some cases (e.g., M83; see Fig. 1) the fields (strongly) overlapped to allow as many priority targets as possible to be observed.

thumbnail Fig. 1

Three overlapping fields in M83. P1 and P2 targets are indicated. North is up and east to the left.

2.3 Pre-imaging and mask design

Pre-imaging of each field was obtained using the imaging mode of FORS2 to facilitate the design of the mask for the multi-object spectroscopy. Each field was observed for 3 × 10 s without a filter with the red detector. This prevented the saturation of our brightest targets, while still reaching deep enough for our faintest P1 and P2 targets (V ~ 22 mag) to have sufficient signal-to-noise ratios (S/N). The individual exposures were stacked using the IRAF (Tody 1986, 1993) task imcombine before being loaded into the FORS Instrument Mask Simulator (FIMS)2 software.

We used rectangular slits of 1″ × 6″. The slit length was chosen to provide enough coverage for sky subtraction, while maximizing the number of slits that can be placed on the mask. First, slits were placed on as many priority targets as could be fitted, as each slit blocks out the horizontal 6″ strip at that location (as the spectrum gets dispersed on the detector in the horizontal direction). Any remaining space on the mask was filled in with other “filler” sources from the Spitzer catalogs (FC), and, when space allowed, with random non-catalog filler objects (FNC). An overview of the masks for each field, with the distribution of the total number of slits (956) per galaxy, field, and priority class is given in Table 5. The pre-images with the location of the slits overlaid are shown in Appendix B.

Table 5

Overview of masks.

2.4 Multi-object spectroscopy and data reduction

The MOS observations for eight of the galaxies were obtained in 2020–2021 (ESO program 105.20HJ) and for the remaining two galaxies (NGC 1313 and NGC 3109) in 2022 (ESO program 109.22W2). The 1″ slits with the GRIS_600RI+19 grism yielded a resolving power R = λ/Δλ ~ 1000. The nominal wavelength coverage of the chosen setup was 512–845 nm. The actual wavelength coverage varied from slit to slit, and depended on the horizontal location on the mask (e.g., for slits placed on the left of the mask, the bluest wavelengths were dispersed outside of the detector). Each field was observed for two observing blocks of 3 × 900 s, yielding a total exposure time of 5400 s per field. The observing log is provided in Appendix C.

The data were reduced using the FORS2 pipeline v5.5.73 under the EsoReflex (Freudling et al. 2013) environment. The pipeline applied the following procedures: bias correction, flat fielding, slit identification, flux calibration, sky subtraction, and extraction of the 1D spectra. In most cases, this produced good quality spectra. However, in cases with multiple objects on the slit, overlapping slits, strongly variable nebular emission, and/or vignetting on the top of the charge-coupled device (CCD), the sky subtraction performed by the pipeline left strong residuals. For this reason, we also performed the reduction without sky subtraction and manually selected the object and sky extraction regions from the 2D spectrum. For each slit, the automatically and manually extracted spectra were visually inspected, and the best reduction was chosen. In rare cases where slits strongly overlapped, no spectra of sufficient quality could be extracted. In almost all cases these were filler targets. Additionally, the used grism caused a vertical shift, causing some of the spectra to fall off the top edge of the CCD or in the gap between the two CCDs. The total number of successfully extracted spectra was 763.

thumbnail Fig. 2

Representative spectra of targets classified as RSGs, following a temperature sequence from K-type (top) to M-type (bottom). Shaded areas correspond to regions with TiO absorption; telluric absorption is indicated with a ⊕ symbol.

3 Spectral classification and stellar content

The spectral classification of the observed sources was performed in two stages. Initially, sources were separated into broad classes (i.e., red, yellow, blue, and emission-line object) based on the slope of the spectrum and the presence of emission lines. The wavelength range of our spectra is not significantly affected by extinction; therefore, the color correlates with the effective temperature. Next, each spectrum was scrutinized and given a more robust classification, if allowed by the signal-to-noise ratio. The details of the classification procedure for each type of object are given in Sects. 3.1–3.3. The catalog is presented in Sec. 3.4, and the stellar content of each target galaxy is described in Sec. 3.5.

3.1 Classification of red and yellow objects

We first inspected the red objects for the presence of molecular absorption bands. In cases where no features were seen (i.e., due to the object being faint or having a low S/N), the classification remained as previously assigned (i.e., red object). The majority of red objects exhibited molecular TiO absorption, indicating a star of M-type. To identify which of these are foreground stars, we calculated the shift of the spectral features due to the radial velocity of the star. When the radial velocity was not in agreement with the host galaxy, we classified them as foreground stars (fgd). As giant and dwarf stars would be far too faint to be detected at great distances, we assigned a supergiant classification whenever the radial velocity was in agreement with the radial velocity of the host galaxy. Given that the primary luminosity class diagnostic, the Ca II triplet, was out of range, the radial velocity confirmation served as one of the primary criteria for the RSG classification.

Spectra showing traces of TiO absorption (i.e., at λλ6150 and 7050) were classified as M-type RSGs (M I), whereas those without were classified as K-type (K I). We checked all our M I sources for ZrO, but in the few cases it was detected it was much weaker than TiO, which is consistent with our luminosity class determination. Figure 2 presents representative examples of target spectra classified as RSG. For all classified RSGs, we additionally verified that RSG atmospheric models (i.e., marcs) fit the observed spectra well (de Wit et al., in prep.). This forthcoming work will also include a more accurate spectral classification of the RSG into early or late-K and early or late-M. When molecular carbon features were present in a spectrum (i.e., the Swan C2 bands or CN molecular bands), the object was classified as a carbon star (C-star).

A few stars had a flat continuum and exhibited a variety of spectral lines. The vast majority of these objects were classified as yellow foreground dwarfs, due to their brightness and non-shifted spectral lines. However, six were classified as YSGs (with spectral types A-F, F, or G), and four as YSG candidates (with luminosity class I-III). These showed a strong O I λ7772 absorption triplet, which is sensitive to luminosity changes in the F0–K0 regime (Kovtyukh et al. 2012). Furthermore, Gaia DR3 proper motions and parallaxes for these targets were small and the radial velocities were consistent with their host galaxy. Figure 3 presents all six spectra classified with certainty as YSG; sources of spectral types A, F, or G, are collectively referred to as AFG.

thumbnail Fig. 3

Spectra of all six targets classified as YSGs. The strength of the oxygen triplet at λ7772 indicates a supergiant luminosity class. Other key spectral lines are indicated; telluric absorption is indicated with a ⊕ symbol.

3.2 Classification of blue objects

Initially, we inspected the blue objects for the presence of spectral features. If none were detected, the raw classification remained. The majority of blue objects were not classified further, as the spectra were of too poor quality to detect spectral lines. However, in cases where spectral lines were clearly detected, we were able to assign a spectral class. When only a broadened Hα emission profile was identified in a spectrum, indicating an accelerating shell of mass surrounding a star, the object was classified as a blue supergiant (BSG). Additionally, when Fe II emission was present in the region λλ6200–6500, the object was classified as an LBV candidate. Ultimately, when [Fe II] (~λλ5150–5550) and [O I] λ6300 emission, originating from low-density regions in a circumstellar disk, were present, the object was classified as a sgB[e] star. Maravelias et al. (2023) presented the LBVs and sgB[e]. Blue objects with Hα clearly in absorption along with other metallic lines, such as the O I triplet (λ7772), Si II (λλ6347, 6371), or He I (λλ5876, 6678, 7065), were classified as BSGc (of B or A type). When Hα was in absorption, the supergiant class was only assigned to stars that showed a strong O I λ7772 absorption triplet. In cases where a strong Si II doublet was observed (indicating Teff ≥ 9000; Mashonkina 2020), sometimes paired with strong He I absorption or a Paschen jump around λ8200, we assigned a B-type classification. In cases where there was mainly significant Hα absorption, we assigned an A-type classification. Figure 4 presents examples of targets classified as BSGs.

3.3 Classification of emission-line objects

We used the following procedure to robustly classify objects with a series of narrow emission lines. First, these objects were inspected for common transitions (e.g., [N II], [S II], He I, and [Ar III]). In cases where a strong emission component was present, but we were not able to assign a robust classification, the target was classified as an “Emission Object” (“EmObj”). In the majority of cases, a doublet of [N II] (λλ6548, 6583) and [S II] (λλ6717, 6731) emission was present. If spectra contained only these five features, they were classified as Nebulae (“Neb.”). This classification could either be singular or added to another classification (i.e., RSG + neb). Additionally, if He I and [Ar III] (i.e., λ7136) emission was present in a nebular spectrum, we classified it as an H II region. For many cases of H II regions, a stellar continuum was clearly present. These received an additional classification (H II + star, or H II + star + Hα if the Hα emission profile appeared broadened, revealing a mass-losing star in the center of the H II region). If a spectrum had emission lines along with a blue component and a red component, it was classified as a cluster with blue stars and at least one RSG. Some spectra exhibited extremely redshifted emission lines. If the spectrum resembled an extremely redshifted H II region, the object was classified as a background galaxy. One spectrum of a background galaxy featured a Balmer jump and was therefore classified as a Balmer jump galaxy. Lastly, if a redshifted spectrum portrayed strong asymmetric and broad emission lines, we classified it as an active galactic nucleus (AGN) or a quasar. We used characteristic emission-line patterns for such objects (e.g., Mg II λ2798) to measure their redshift. We present an H II region, a galaxy, an AGN and a quasar in Fig. 5.

We proceeded to analyze the spectra of all the H II regions, along with all the targets exhibiting nebular contamination in their spectrum to measure the flux of the lines Hα, [N II] λ6583, [S II] λλ6716 and 6731. We selected the manually reduced spectra as the target and sky apertures were selected more carefully than in the automatized pipeline version. However, we visually inspected the 2D spectrum to avoid overlapped slits, inhomogeneous nebular emission in the spatial direction, or other artifacts that could compromise the measurement of the flux. We used the IRAF task splot to measure the fluxes and their associated error, establishing a minimum error of 3% in the flux when the signal of the nebular emission was much more powerful than the noise. The results are presented in Sec. 4.3.

thumbnail Fig. 4

Representative spectra of BSGs that exhibit Hα emission. The Hα profiles are a superposition of the nebular line profile broadened at the base by a circumstellar component, indicating significant mass loss for these objects.

3.4 Catalog

Following the procedure described above, we classified all of the 763 successfully extracted FORS2 spectra. We also used the Mikulski Archive for Space Telescopes (MAST) portal to search for available imaging of our targets from the Hubble Space Telescope (HST) using the Hubble Legacy Archive (HLA). When available, photometry from the Hubble Source Catalog (Whitmore et al. 2016) assisted in the identification of our targets. We found that 150 of our 541 classified targets had available imaging and performed a robust visual classification for 80 of these, providing extra information on whether the targets are point-sources (i.e., stars) or extended sources (i.e., clusters, background galaxies). We compared the spectral classifications against the visual classification from the HST images and found them to be consistent; in three cases we revisited and improved the spectral classification4.

Of the 763 successfully extracted spectra, 454 were robustly classified (~60%) as an evolved massive star, an extended object (i.e., H II region or cluster), or a contaminant (i.e., a galaxy or a foreground star), while 87 (or 11%) were given a more general classification (i.e., nebula, blue object or red object). The remainder of the spectra (222 stars or 29%) were not of sufficient quality to be assigned a spectral class. Table 6 presents the distribution of classified targets per galaxy and spectral class, for both robust classifications (440) and candidates (14).

The number of evolved massive stars (including candidates) that were classified is 185 targets (~24%). This is the largest catalog of evolved massive stars beyond the Local Group (e.g., Bibby & Crowther 2012; Grammer et al. 2015; Britavskiy et al. 2019; Humphreys et al. 2019). Out of the 185 targets, the RSGs, LBV candidates and sgB[e] stars are robust classifications (see Table 6), whereas 30–40)% of the BSGs and YSGs are candidates. This is because their luminosity class could not be confirmed from the spectral range and resolving power of our spectra, mainly due to the lack of resolved spectral lines that are sensitive to changes in the surface gravity. The global contamination fraction in our sample is (Nfgd+NC+NAGN+Ngal+NQSO)/Nspec ~ 12%, where Nspec is the 763 spectra.

The resulting catalog of all 541 spectroscopically classified targets is presented in Table 7. The targets are ordered by galaxy RA and by their m4.5 magnitude within each galaxy. The catalog contains the following information: target ID, coordinates (in degrees, J2000), field, target priority, 17 bands of photometry from Spitzer IRAC and MIPS, Pan-STARRS1, VHS, and Gaia DR2, our spectral classification, comments on the spectral classification, the classification from the HST images (when available), along with the HST filters available and comments on the HST images.

thumbnail Fig. 5

Example spectra of targets classified as H II regions or galaxies (including AGN and quasars). Their redshifts were calculated from the emission lines indicated for each of these targets.
Table 6

Distribution of classified targets per galaxy and spectral class.

3.5 Comments on individual galaxies

In the following subsections, we present the results of our target classification in each galaxy.

3.5.1 WLM

The WLM sample mostly consisted of stars without a priority or a low priority. No RSGs were found in this galaxy, despite recent studies on the RSG content of WLM (i.e., Britavskiy et al. 2015; González-Torà et al. 2021, as these targets were not in our priority system).

We observed one sgB[e] (WLM-95), which is the first such source identified in a low-Z galaxy (~0.14 Z⊙) presented by Maravelias et al. (2023, source WLM-1). In addition to strong, broadened Hα and weak Fe II emission lines, significant He I λ5875 and λ6678 absorption is present. This object was previously classified as an emission-line star by Massey et al. (2007) and Britavskiy et al. (2015). We report no significant changes in the spectrum (over a period of 8 yr), as the presence of Fe II and [O I] emission and absence of [Ca II] emission were observed in both of the spectra.

No BSGs, YSGs, or LBV candidates were identified in WLM. Finally, the majority of observed red stars contained the strong carbon absorption bands typical of carbon-rich AGB stars. No H II regions were found.

Table 7

Catalog of spectral classifications.

3.5.2 NGC 55

We classified 42 RSGs in NGC 55, mostly of M I type. De Wit et al. (in prep.) will provide a more detailed classification as well as stellar parameters through MARCS modeling. In addition, six stars with broadened Hα emission were observed, the largest number found in a single galaxy in our sample. Four were classified as BSG (NGC 55-260, NGC 55-287, NGC 55-797, and NGC 55-NC6), two as LBVc (NGC 55-736 and NGC 55-2924), and one as a sgB[e] star (NGC 55-178; see Maravelias et al. 2023). One additional BSG (NGC 55-260) has Hα in absorption, yet shows a strong O I λ7772 triplet. We observed one YSG (NGC 55-222) of spectral type F-G.

The contamination by foreground stars is low for NGC 55, although we detected 17 extended objects. We also identified a background galaxy, an AGN at z = 0.39 and a quasar at z = 1.158, which is coincident with the X-ray source 4XMM 001547.5-391240. The AGN is presented in Fig. 5.

3.5.3 NGC 247

We identified 16 M-type RSGs in NGC 247. As the median RSG type depends on metallicity (Elias et al. 1985; Levesque et al. 2006), we note that the RSG spectral types are in agreement with the previously estimated metallicity (~0.4 Z⊙; Davidge 2021, similar to M33 and the LMC) of the galaxy (see Sec. 4.3).

Various other types of evolved massive stars were observed, including one LBVc (NGC 247-1192) and one sgB[e] (NGC 247-246), both described in Maravelias et al. (2023). NGC 247 hosts a bright YSG (NGC 247-1786). A blackbody fit to the available photometry revealed an effective temperature around 5600 K. We observed one additional YSG (NGC 247-NC1), one candidate YSG (NGC 247-3021) and five BSG or candidate BSG. The contamination by H II regions in NGC 247 is relatively high, on the order of ~20%. Some spectra contained both a bright blue component and at least one RSG, meaning these are most likely clusters (see Sec. 3.3).

3.5.4 NGC 253

We identified 12 RSGs and one sgB[e], NGC 253-739 (Maravelias et al. 2023). All of the RSGs were classified as M-types, except for one potential K-type (NGC 253-3509). For this star, the detection of the TiO bands was hampered by the low S/N of the spectrum (although a best-fit marcs model gave a Teff compatible with a K-type supergiant). The metallicity of NGC 253 (0.72 Z⊙) is consistent with finding mainly M-type RSGs. We obtained a wide range of radial velocities for these RSGs, varying from 70 to 450 km s−1. These values are in agreement with the kinematical maps measured for NGC 253 (Hlavacek-Larrondo et al. 2011). One spectrum (NGC 253-91) revealed a promising YSG candidate, sharing similarities with an early K-type RSG, yet it was too hot to be modeled with marcs models. Finally, we identified one BSG candidate. The confusion by H II regions was more extreme in NGC 253 (38%), compared to NGC 247 (20%), which is a similarly distant and inclined galaxy. Eleven targets revealed a composite spectrum including at least one blue and one red component. Three targets were classified as foreground stars.

3.5.5 NGC 300

We identified 46 RSGs in NGC 300, the majority of which were classified as M-types. Additionally, one BSG (NGC 300-901) and two sgB[e] (NGC 300-67 and NGC 300-389) have been identified due to the presence of Hα emission (for both the BSG and sgB[e] stars), along with Fe II and [O I] λ6300 emission (for the sgB[e]). The sgB[e] NGC 300-67 further displayed strong He I lines in emission (see NGC 300-1 in Maravelias et al. 2023). We observed six additional BSGs and one YSG showing Hα in absorption but with the O I λ7772 triplet being sufficiently strong. Ten targets were classified as H II regions, with a higher number of contamination coming from foreground stars (17). Finally, background galaxies, AGN and a quasar contributed to eight more contaminants.

3.5.6 NGC 1313

One RSG of spectral type M was identified in this galaxy (NGC 1313-310), in agreement with expectations for the metallicity of the galaxy (~0.6 Z⊙). Additionally, one BSG and two BSGc were detected, but no LBVs or sgB[e] stars. NGC 1313 was one of the farthest galaxies in our sample, and as such, about half of the observed targets were classified as extended objects (i.e., clusters and H II regions). Lastly, we observed five foreground stars.

3.5.7 NGC 3109

We identified one target as a RSG of early K-type (NGC 3109-167). This is consistent with the bulk of the RSG types for this metallicity (~0.2 Z⊙). One object (NGC 3109-188) has been classified as an LBVc due to the strong circumstellar Hα and Fe II emission features detected, along with He I lines in absorption (see Maravelias et al. 2023). Apart from one non-priority YSG (NGC 3109-287), most of the non-priority targets were classified as foreground stars. We identified a Balmer jump galaxy and a quasar at z ~ 1.14, which is shown in Fig. 5.

3.5.8 Sextans A

One RSG was identified in Sextans A (SextansA-1906), with an early K-type spectrum, which is compatible with the low metallicity of the galaxy (0.06 Z⊙). The sample of Sextans A mainly contained blue stars, the majority of which were classified as O stars. We classified SextansA-s1 and SextansA-s2 as supergiants (O-B I) based on their Hα emission, although Lorenzo et al. (2022) classified them as O9.5II and O3-O5Vz, respectively. A few spectra contained broadened Hα emission profiles, but we were unable to determine whether these are indeed BSGs. No LBVs, sgB[e], foreground stars or H II regions were observed.

3.5.9 M83

We identified one RSG (M83-682) of spectral type K, which is unexpected at super-solar metallicity (1.6 Z⊙), with no further LBV, sgB[e], YSG, or BSG candidates in this galaxy (see Table 6). The lack of identified massive stars is mostly the result of the increased contamination due to crowding, and decreased S/N due to the large distance to the galaxy (see Sec. 4.2). The majority of spectra taken for M83 revealed characteristics of extended objects (i.e., H II regions and stellar clusters). Although some of the observed H II regions revealed a stellar continuum and a broadened Hα emission profile, indicating the existence of a mass-losing blue star in the center, we were unable to determine its nature. We narrowed down the spectra of many objects to blue stars with some degree of nebular contamination, but the spectra generally lacked the quality to determine whether these are supergiants or LBV candidates.

3.5.10 NGC 7793

The sample in NGC 7793 contains nine RSGs showing clear molecular TiO absorption. We estimate the spectral type to be early M for the majority of these objects. One particular source of interest is NGC 7793-111, for which we noticed strong circumstellar Hα, clear Fe II emission features in the left wing of Hα, and weak [Ca II] λλ7291–7323 emission (although the spectrum suffers from various residuals and artifacts). The area around the [O I] λ6300 line was highly contaminated, and therefore the distinction between an LBV and a sgB[e] classification could not be made. Maravelias et al. (2023) provided additional photometric information, which resulted in a sgB[e] classification. We further classified one BSG (NGC 7793-721) and one YSG (NGC 7793-30) candidate. Six H II regions and several background galaxies and foreground stars were found in the sample.

4 Discussion

To determine which of our classified targets have previous spectral classifications in the literature, we searched for all publications providing spectral types of massive stars in our ten target galaxies and compiled a catalog of these sources. Then we cross-matched our 541 targets against the known massive stars in our catalog and found 31 matches, all of which were among our 185 massive stars, yielding 154 newly classified massive stars.

Table 8 lists the IDs and coordinates of the matched sources from our work and the literature (in degrees, J2000), the newly determined and previous classifications, and a comment field listing the cross-matched ID from the literature. The matched sources include two LBV candidates, a sgB[e] and an Fe star (which we reclassified as sgB[e]; see Maravelias et al. 2023 for more details), 10 RSG for which we provide a spectral type classification for the first time, a cool star5 that was previously classified as a RSG, three H II regions in M83 that have been identified as WR clusters, and several O, B, and A supergiants for which more accurate spectral types are available in the literature. We note an interesting case, NGC 3109-167, which was classified as KI in this work (as weak TiO bands were detected). It has a previous classification as G2I and therefore may exhibit spectral type variability. NGC 300-52 (K-M V) was previously classified as a WR candidate from narrowband imaging (Schild et al. 2003); however, its proper motion from Gaia DR3 confirms that it is a foreground source.

To visualize our results, we proceeded to construct CMDs and figures showing the distribution of the classified sources in each galaxy. In the following subsections, we present a discussion of the CMDs, the spatial distribution plots, and the effectiveness of our target selection method.

4.1 Color-magnitude diagrams and spatial distribution plots

We constructed both cumulative CMDs and CMDs for the individual galaxies in the mid-IR (m3.6 vs. m3.6 − m4.5 and J − m3.6), near-IR (Ks vs. J − Ks), and optical (G vs. GBP − GRP and g vs. g − r), presenting the evolved massive stars (RSG, BSG, YSG, LBV, and sgB[e]) classified in this work, to verify the robustness of our classifications. Figure 6 presents four cumulative CMDs for all our evolved massive stars and H II regions in all ten galaxies. For the construction of these CMDs, we used the distances in Table 1 to convert to absolute magnitude. The CMDs for the individual galaxies are presented in Appendix D. These show apparent magnitude versus color and also include the background galaxies. The CMDs for NGC 300 (Fig. D.5) and NGC 55 (Fig. D.2) are noteworthy, as they contain the largest number of classified targets (110 and 85, respectively).

In the CMDs, we grouped together the candidate supergiants with the confirmed supergiants. We also merged the Galaxies, AGN and quasars, labeling them as “Galaxies” in the CMDs. For the optical CMDs, we used photometry from Gaia DR2 and Pan-STARRS1, when available. For the galaxies with near-IR data (e.g., VHS) we created CMDs similar to Bonanos et al. (2009), which are used to separate the LBVs and sgB[e] from the rest of the sources. Unfortunately, both the optical and the JHKs photometry is scarce for our sample; the Araucaria Project (e.g., Karczmarek et al. 2021) will soon release near-IR photometric maps for some of our galaxies. The Spitzer MIPS [24] band is not reliable for the galaxies studied in our survey because of the low resolution of MIPS (6″), which corresponds to large regions at distances of a few megaparsecs. Therefore, it was not used.

By examining the positions of evolved massive stars on the CMDs, we confirm the location of sgB[e] stars at around J − m3.6 ~ 3.0 mag, in agreement with the location of these stars in the respective CMDs in the Magellanic Clouds (Bonanos et al. 2009, 2010). The emission objects are found to be colocated with the LBVs and sgB[e], with the exception of WLM-1325 (see Fig. D.1) at M3.6 ~ −6 mag. In particular, we note the priority 1 emission-line objects NGC 300-309 and NGC 7793-47 (Figs. D.5 and D.10), and speculate that they could be faint LBVs. We find the majority of RSG and YSG to be outside our priority system (see Sec. 4.2). We finally highlight the BSGs with mid-IR excess (see, e.g., Fig. D.5), which implies the presence of CSM around them. Some massive stars do not appear as priority targets because they did not pass our m4.5 criterion. We detect H II regions mainly in more distant and inclined (edge-on) galaxies, for example NGC 253 and M83, as crowding and contamination of the mid-IR photometry used to select targets increases with distance (see Sec. 4.2).

Given the large number of RSGs, we can explore trends and comment on outliers. We find that RSGs cluster at J − Ks ~ 1.0 mag and lie on a vertical branch spanning ~3 mag in m3.6 and ~2.5 mag in Ks (see Fig. D.2 for NGC 55 and Fig. D.5 for NGC 300), similar to the LMC (Bonanos et al. 2009; Woods et al. 2011; Jones et al. 2017) and SMC (Bonanos et al. 2010; Ruffle et al. 2015). In all these galaxies, the bulk of the population is concentrated within ~2 mag, between −9.5 and −11.5 mag in both M3.6 and MKs. Interestingly, the most luminous RSGs were selected as priority targets, as they have the highest mass-loss rates, and therefore dustier environments and are brighter in the mid-IR. We estimate that we probe RSG luminosities between log(L/L⊙) ~ 4.4–5.6, from MKs (using a bolometric correction BCK = 2.69 mag and adopting the distances from Table 1). As improved distances exist for some target galaxies, the location of RSGs on the CMDs and the Hertzsprung-Russell diagram will be revisited in de Wit et al. (in prep.).

We note that several RSGs have unusually blue mid-IR colors, m3.6 − m4.5 ≲ −0.5 mag, but normal optical colors and spectra, namely NGC 55-790, NGC 55-1040, NGC 55-1047, NGC 253-1534, NGC 55-1734, NGC 247-2912, and NGC 247-3231 (see Figs. D.2 and D.3). We attribute their blue mid-IR colors to their faint mid-IR photometry (i.e., m4.5 ~ 17.5 mag), which is likely contaminated in these edge-on galaxies, and to variability, as the 3.6 μm and 4.5 μm data were obtained in some cases at different epochs. Furthermore, the blue colors could be due to the CO absorption bands at 4.5 μm (Verhoelst et al. 2009). Similarly, NGC 247-3683 and NGC 247-2966 have blue optical colors, but a classification of “MI+Neb.” The most likely explanation for the blue colors of these RSGs is that they are part of a system or cluster including at least one hot star. The blue component was not accessible and therefore undetected by our spectra, but would provide the necessary ionizing radiation to produce the observed emission lines.

Comments on specific targets are given below:

  • LBVs in NGC 55: NGC 55-736 and NGC 55-2924 are not priority targets as they do not satisfy the apparent magnitude selection criterion in m4.5. NGC 55-2924 does not satisfy the absolute magnitude criterion in m3.6 either, which implies that our magnitude cuts exclude fainter LBVs;

  • NGC 300-357 is the only candidate YSG found among our priority targets (P3; the spectrum of the object has Hα in emission);

  • NGC 1313-48 is a remarkable BSG (BI), very luminous in the mid-IR, but lacking optical photometry (a P6 target), which is shown in the mid-IR CMD (Fig. D.6). It is located near the center of the galaxy (Fig. E.6).

We finally constructed spatial distribution diagrams (Figs. E.1–E.10) for the evolved massive stars found in each galaxy of our sample using images from the Digitized Sky Survey 2 (DSS2). We found the evolved massive stars and the H II regions to be mainly located in the plane of the host galaxy or its spiral arms, which is the expected location for such objects. These diagrams helped us correct the YSG classification for NGC 3109-1209 (due to its redshifted spectrum that matched the radial velocity of the target galaxy) to a foreground source, since it is not located on the disk of the host galaxy and has a high proper motion in Gaia DR3.

Table 8

Catalog of matched evolved stars.

thumbnail Fig. 6

IR and optical cumulative absolute magnitude CMDs for all evolved massive stars and H II regions. Background sources are from NGC 300. The dashed lines in the first panel indicate the criteria used for target selection; all our priority targets are located within the box at the top right defined by these lines.
Table 9

Distribution of classified, priority (Prio), and massive stars (MS) per galaxy and spectral class.

4.2 Effectiveness of the selection method

As described in Sec. 3.4, out of the 956 slits, 763 were successfully extracted and of these, 454 were classified (i.e., 47% of the total slits, or 59% of the successfully extracted ones). We identified 185 evolved massive stars among these, which is 19% of the total slits (956), 24% of the successfully extracted targets (763), 28% of the classified targets (541), and 41% of the accurately classified targets (454). This moderate success rate is due to two factors: (i) only 384 out of 956 targets (195 out of the 541) were priority targets, and (ii) many targets, independent of priority class, suffered from reduction errors (e.g., vignetting, overlap, or poor sky subtraction), which prevented a classification.

In Table 9, we present the distribution of stars per galaxy and per massive star class for the total number of classified stars (541), the priority stars (195), the classified massive stars (185) and the massive stars that were in the priority system (44). The breakdown of priority massive stars by spectral type is also given, along with the success rates of recovering massive stars. The average success rate of selecting massive stars was 28%, with NGC 55 and NGC 300 achieving the highest rates (~50%). Excluding Sextans A, which only had 1 priority target, and M83 and NGC 253, which were highly contaminated by H II regions, the average success rate of our priority system in selecting evolved massive stars is 36%, while for NGC 55, NGC 247, NGC 300, and NGC 3109 the success rate ranged from 38–42%. We also find that 15–25% of our classified RSG, BSG, LBVc and YSG were priority targets, whereas for sgB[e] stars the fraction is much higher at 86%. By comparing the CMDs with the selection of priority targets (Appendix A) to the CMDs with the classified sources, we find that many of the “filler” stars turned out to be RSGs or YSGs (see, e.g., the CMD of NGC 300 in Figs. A.5 and D.5). In some galaxies, priority targets with color m3.6 − m4.5 ≥ 1.0 mag were not placed in slits (see, e.g., Figs. A.6 and A.10). This is due to the limitations imposed by MOS observations.

We revisit the cuts applied to limit contamination (see Sec. 2.2), namely the magnitude cut at 4.5 μm to remove background galaxies, the color cut m3.6 − m4.5 ≥ 0.1 mag to avoid yellow foreground dwarfs, and Gaia cleaning to remove foreground sources. The magnitude cut was quite successful, yielding a 4% contamination by galaxies (as a fraction of the 454 classified targets); however, it prevented us from selecting all but the most luminous massive stars in galaxies found at distances greater than 4 Mpc. The Gaia cleaning resulted in a 13% contamination by foreground stars. Future data releases of Gaia will improve the removal of foreground sources. Our selection criteria in M83 yielded a large number of H II regions, as the Spitzer IRAC photometry with its ~2″ resolution gets increasingly contaminated with distance. We note that most H II regions in M83 were P1 and P2 targets (i.e., with optical counterparts). Furthermore, as all targets were observed with similar exposure times throughout the observing campaign, the robustness of the classifications generally drops with distance. Ideally, targets without optical counterparts (P3, P4 and P6) need to be followed up with near-IR spectroscopy. This is particularly important for sources with m3.6 − m4.5 > 1.0 mag.

The completeness of our study of dusty massive stars (i.e., the priority targets placed in slits and observed vs. the total number of available priority targets) in these galaxies ranges from 14% (NGC 1313 and NGC 3109) to 56% (WLM), with an average of 30% (see the last column of Table 4). This does not take into account the number of sources we rejected when creating the original catalog (~0.1%) when de-blending, as they are negligible. The completeness achieved, along with the small number of detected dusty, massive stars per spectral class, limits our ability to determine the frequency of episodic mass loss. However, these percentages can be used to estimate a conservative total number of dusty, massive stars in each galaxy, which will of course be limited by the galaxy inclination, as well as the angular resolution and sensitivity of Spitzer. For example, using the completeness for each galaxy from Table 4 and the corresponding number of priority RSGs from Table 9, we predict that six additional dusty RSGs exist in NGC 55, four in NGC 300 and one in NGC 247 and NGC 7793. These are of interest for characterizing dust-enshrouded RSGs (e.g., Beasor & Smith 2022) in more distant galaxies, especially in NGC 55 and NGC 300, which are expected to contain 17 and 12 dusty RSGs, respectively.

4.3 H II regions

In addition to evolved massive stars, our catalog contains 99 H II regions. The majority of them were found in M83 and NGC 253, where we identified 28 and 24 H II regions, respectively (see Table 6). Since there have been previous studies of H II regions in M83, we cross-matched the 28 H II regions identified in that galaxy, using a 1″ radius, and found three matches in Rumstay & Kaufman (1983, M83-127 with 132, M83-433 with 209, M83-1048 with 247), one match with Long et al. (2022, M83-993 with HII058) and one match with Winkler et al. (2023, M83-1048 with HII-13). Cross-matches with the H II regions reported by Bresolin & Kennicutt (2002), Bresolin et al. (2005), and Bresolin et al. (2009) did not yield any matches within a 1″ radius. We therefore present 24 new H II regions in M83, which can be further studied to, for example, identify shocked material (e.g., supernova remnants) or to derive the abundances (e.g., Bresolin et al. 2016).

Historically, the ratio [S II]/Hα has been used as a diagnostic to distinguish between photo-ionized and shocked emission, using 0.4 as the boundary (Mathewson & Clarke 1973). However, recent works suggest using additional nebular line ratios to better assess the origin of the nebular emission (Kopsacheili et al. 2020). Following this approach, we plotted our emissionline flux ratio measurements (see Sec. 3.3 and Table F.1) on the diagram [N II]/Hα versus [S II]/Hα (see Fig. 7), discovering eight targets with [S II]/Hα higher than 0.4. Among these objects, we distinguish four “RSGs+nebula” (NGC 55-889, NGC 300-NC4, NGC 253-452, and NGC 253-1162), two “blue+nebula” (NGC 55-NC12 and NGC 253-NC4), a H II region (NGC 253-1657), and a cluster (NGC 253-769). Further analysis is needed to reveal the origin of this shocked material (e.g., SN remnants in the field, shock interaction because of strong winds, etc.) and whether any of the four “RSGs+neb” contain bow shocks similar to those around Betelgeuse (Mohamed et al. 2012; Mackey et al. 2012), or IRC-10414 (Gvaramadze et al. 2014). Furthermore, we identified two targets with an enhanced [N II]/Hα ratio with respect to their host galaxies. Their classification is “composite” (NGC 253-1) and “EmObj+neb” (NGC 300-NC21). In particular, the latter shows stellar N II emission in the series of λλ5668 − 5712 and the lines λ6482 and λ6611. These features in emission could suggest a N-rich mass-losing star, which is enriching the gas around it and enhancing the relative abundance of nitrogen over hydrogen.

The ratio [N II]/Hα is a well-known indicator for metallicity in H II regions, supernova remnants (Leonidaki et al. 2013), and star-forming galaxies (Storchi-Bergmann et al. 1994). In Fig. 7, the measurements in each galaxy follow distinct lines depending on their metallicity, with the most metal poor ones located in the lower part of the diagram and the most metal-rich in the upper part. However, the trend does not hold for super-solar oxygen abundances (Kewley & Ellison 2008), which explains why the measurements for our highest metallicity (0.7 Z⊙ and 1.6 Z⊙) galaxies overlap. The case of NGC 247 (0.4 Z⊙) is noteworthy, as the line ratios overlap with those of NGC 55 (0.27 Z⊙) and NGC 1313 (0.35 Z⊙) especially for three H II regions located in the outskirts of the galaxy, instead of those of NGC 300 and NGC 7793 (~0.4 Z⊙), implying a metallicity of ~0.3 Z⊙ for this galaxy. However, the identification of 16 M-type RSG (a narrow effective temperature range of RSG was also noted by Davidge 2021) in this galaxy is consistent with a metallicity of 0.4 Z⊙.

4.4 Other objects of interest

Carbon stars were identified only in nearby galaxies (seven in WLM6, one in Sextans A, and two in NGC 300). The carbon star SextansA-12116 has been reported as a variable (Boyer et al. 2015a), extreme AGB star by Boyer et al. (2017). In more distant galaxies it is likely that we have also observed such stars, but because of their lower luminosities compared to RSGs, we were not able to classify them since the S/N of their spectra was too low. Carbon stars are difficult to filter out because they occupy similar regions as the RSGs in most CMDs. Among the targets classified as galaxies, there are three quasars, two of which have an X-ray counterpart: NGC 55-664 (4XMM J001547.5-391240), NGC 300-349 (4XMM J005449.7-374000), and NGC 3109-299. All quasars have broad emission features and are at redshifts between 1–2.

Another noteworthy target (P1) is the “emission object” NGC 300-309, which turns out to be the supernova imposter SN 2010da or NGC 300 ULX-1 (Binder et al. 2020). It has been explained as an ultraluminous X-ray source with a RSG donor and a neutron star (Heida et al. 2019). We detected Hα circumstellar emission, which remains strong at above 100 times the continuum, in agreement with the latest spectrum in 2015 reported by Villar et al. (2016). Poor sky subtraction prevented us from detecting other lines.

thumbnail Fig. 7

[N II]/Hα vs. [S II]/Hα for all H II regions and spectra classified as nebulae found in eight of our target galaxies, which range from Z=0.2–1.6 Z⊙. The vertical line indicates the value [S II]/Hα = 0.4, above which the emission is shocked. The diagonal shows the 1:1 relation.

5 Summary

We have presented the results of a spectroscopic survey of luminous, evolved massive stars with dust in nearby (1–5 Mpc) southern galaxies, which were observed with FORS2 on VLT as part of the ASSESS project. We developed a “priority system” of applying magnitude and color cuts to archival Spitzer photometry to select dusty, evolved massive stars for multi-object spectroscopy in the galaxies WLM, NGC 55, NGC 247, NGC 253, NGC 300, NGC 1313, NGC 3109, Sextans A, M83, and NGC 7793. Out of the 763 objects observed, we obtained classifications for 541 of them, of which 454 were robust classifications; 87 were given a general classification, and 222 were not of sufficient quality to allow for a classification. The robust classifications include 185 evolved massive stars, 99 H II regions, 10 carbon stars, 21 galaxies (including quasars and AGN), and 61 foreground sources, yielding a contamination fraction of 12%. Of the 185 evolved massive stars, 154 have spectral types reported for the first time. The evolved massive stars include 129 RSGs, 27 BSGs, 10 YSGs, 4 LBVc, 7 sgB[e], and 8 emission-line objects; 24% of these are within the priority system (i.e. they have clear signs of circumstellar dust). We further measured the [N II]/Hα and [S II]/Ha line ratios for all H II regions and nebular emission-line spectra, which yielded eight shocked emission regions, four of which contain a RSG component.

The fraction of dusty massive stars observed with respect to those initially selected in these galaxies is on average 30%, due to the constraints imposed by the MOS observations. We report a success rate of 28% for identifying massive stars among our observed spectra (which include “filler” sources outside the priority system), while the average success rate of our priority system in selecting evolved massive stars was 36% (see Sec. 4.2) due to reduction errors (e.g., vignetting, overlapping spectra, and poor sky subtraction). The efficiency of recovering priority massive stars versus distance given the resolution of Spitzer (1.7″ at 3.6 μm) starts dropping at 3.5 Mpc (Table 9), which implies that the James Webb Space Telescope, with its 0.1″ resolution, will enable the photometric selection of dusty massive stars out to 60 Mpc. Furthermore, both the Euclid mission and the Nancy Grace Roman Space Telescope, with their wide-field capabilities in the optical and near-IR, will provide deep photometry. This, combined with the variability information from the upcoming Vera C. Rubin Observatory's Legacy Survey of Space and Time, will enable the identification and detailed study of all massive stars, including the dusty, evolved ones, in a large number of nearby galaxies. Moreover, machine-learning methods – such as the photometric classifier presented by Maravelias et al. (2022), with its overall accuracy of ~80% – are expected to greatly improve the efficiency of identifying massive stars and therefore increase the sample of dusty, evolved massive stars.

The resulting catalog is the largest catalog of evolved massive stars and RSGs beyond the Local Group. This catalog can serve as a valuable resource for follow-up studies of specific targets or types of objects, such as detailed modeling of RSGs (de Wit et al., in prep.). Follow-up spectroscopy is currently underway for LBVs, certain BSGs and YSGs, and selected RSGs. A forthcoming paper will present the results of our observations of dusty, evolved targets in galaxies located in the northern sky.

Acknowledgements

The authors acknowledge funding support from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (Grant agreement No. 772086). EZ also acknowledges support from the Hellenic Foundation for Research and Innovation (H.F.R.I.) under the “3rd Call for H.F.R.I. Research Projects to support Post-Doctoral Researchers” (Project No. 7933). We thank Elias Koulouridis for helping with the classification of the quasars, Despina Hatzidimitriou for helpful discussions and the referee for a careful reading and a very prompt report with suggestions that improved the manuscript. Based on observations collected at the European Southern Observatory under ESO programme 105.20HJ and 109.22W2. This work is based in part on observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under a contract with NASA. This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement. Based on observations made with the NASA/ESA Hubble Space Telescope, and obtained from the Hubble Legacy Archive, which is a collaboration between the Space Telescope Science Institute (STScI/NASA), the Space Telescope European Coordinating Facility (ST-ECF/ESAC/ESA) and the Canadian Astronomy Data Centre (CADC/NRC/CSA). The Digitized Sky Surveys were produced at the Space Telescope Science Institute under U.S. Government grant NAG W-2166. The images of these surveys are based on photographic data obtained using the Oschin Schmidt Telescope on Palomar Mountain and the UK Schmidt Telescope. The plates were processed into the present compressed digital form with the permission of these institutions. This research made use of Astropy7, a community-developed core Python package for Astronomy (Astropy Collaboration 2013, 2018) and Photutils, an Astropy package for detection and photometry of astronomical sources (Bradley et al. 2022).

Appendix A Target selection from Spitzer color-magnitude diagrams

This appendix provides Spitzer m3.6 − m4.5 versus m3.6 CMDs for each of the galaxies in Figs. A.9 to A.10. All priority targets are indicated, as are targets that had a slit placed on them. The different magnitude cuts that were used to assign target priority are also indicated; the apparent magnitude criterion of m4.5 ≤ 15.5 mag for rejecting background sources causes the diagonal cut. The location of η Carinae is shown for reference.

thumbnail Fig. A.1

Spitzer CMD of WLM.
thumbnail Fig. A.2

Spitzer CMD of NGC 55.
thumbnail Fig. A.3

Spitzer CMD of NGC 247.
thumbnail Fig. A.4

Spitzer CMD of NGC 253.
thumbnail Fig. A.5

Spitzer CMD of NGC 300.
thumbnail Fig. A.6

Spitzer CMD of NGC 1313.
thumbnail Fig. A.7

Spitzer CMD of NGC 3109.
thumbnail Fig. A.8

Spitzer CMD of Sextans A.
thumbnail Fig. A.9

Spitzer CMD of M83.
thumbnail Fig. A.10

Spitzer CMD of NGC 7793.

Appendix B Mask design

This appendix provides an overview of the mask layouts of each observed field. For each field an overview image is presented consisting of the pre-imaging of that field, the locations of the reference stars (double red circles), and the slit locations (black 1“×6” rectangles). The target coordinates are corrected for distortions of the instrument while the pre-imaging is not, and hence the slits may appear slightly off-target in some cases.

Appendix B.1 WLM Field A

thumbnail Fig. B.1

Mask layout for WLM Field A.

Appendix B.2 NGC 55 Field A

thumbnail Fig. B.2

Mask layout for NGC 55 Field A.

Appendix B.3 NGC 55 Field B

thumbnail Fig. B.3

Mask layout for NGC 55 Field B.

Appendix B.4 NGC 55 Field C

thumbnail Fig. B.4

Mask layout for NGC 55 Field C.

Appendix B.5 NGC 247 Field A

thumbnail Fig. B.5

Mask layout for NGC 247 Field A.

Appendix B.6 NGC 247 Field B

thumbnail Fig. B.6

Mask layout for NGC 247 Field B.

Appendix B.7 NGC 253 Field A

thumbnail Fig. B.7

Mask layout for NGC 253 Field A.

Appendix B.8 NGC 253 Field B

thumbnail Fig. B.8

Mask layout for NGC 253 Field B.

Appendix B.9 NGC 253 Field C

thumbnail Fig. B.9

Mask layout for NGC 253 Field C.

Appendix B.10 NGC 300 Field A

thumbnail Fig. B.10

Mask layout for NGC 300 Field A.

Appendix B.11 NGC 300 Field B

thumbnail Fig. B.11

Mask layout for NGC 300 Field B.

Appendix B.12 NGC 300 Field C

thumbnail Fig. B.12

Mask layout for NGC 300 Field C.

Appendix B.13 NGC 300 Field D

thumbnail Fig. B.13

Mask layout for NGC 300 Field D.

Appendix B.14 NGC 1313 Field A

thumbnail Fig. B.14

Mask layout for NGC 1313 Field A.

Appendix B.15 NGC 3109 Field A

thumbnail Fig. B.15

Mask layout for NGC 3109 Field A.

Appendix B.16 Sextans A Field A

thumbnail Fig. B.16

Mask layout for Sextans A Field A.

Appendix B.17 M83 Field A

thumbnail Fig. B.17

Mask layout for M83 Field A.

Appendix B.18 M83 Field B

thumbnail Fig. B.18

Mask layout for M83 Field B.

Appendix B.19 M83 Field C

thumbnail Fig. B.19

Mask layout for M83 Field C.

Appendix B.20 NGC 7793 Field A

thumbnail Fig. B.20

Mask layout for NGC 7793 Field A.

Appendix C Observation log

This appendix gives the observing logs of all pre-imaging (Table C.1) and spectroscopic Mask Exchange Unit (MXU) (Table C.2) observations. For the pre-imaging, the mean airmass and seeing are given for the three (back-to-back) exposures, for MXU airmass and seeing are given at the start and end of the exposure. The reported seeing values come from the differential image motion monitor (DIMM) at Paranal, and refer to the seeing at 500nm at zenith. The image quality of the pre-imaging was measured by computing the median full width at half maximum of all sources detected in the field using the Python module photutils (Bradley et al. 2022).

Table C.1

Log of pre-imaging observations.

Table C.2

Log of MXU observations.

Appendix D Color-magnitude diagrams of classified targets

This appendix provides the mid-IR (M3.6 vs. m3.6 − m4.5 and J − m3.6), near-IR (MKs vs. J − Ks) and optical (MG vs. GBP − GRP, g vs. g − r) CMDs for each galaxy of our sample (see Section 4.1), similar to the cumulative CMDs shown in Figure 6. The CMDs were constructed using data from various surveys (described in Section 2.1) when available and include all evolved massive stars in our sample, as well as the dominant sources of contamination.

thumbnail Fig. D.1

IR and optical CMDs for the classified targets of interest in WLM, i.e., all evolved massive stars, H II regions, and background objects.
thumbnail Fig. D.2

Same as Figure D.1, but for NGC 55.
thumbnail Fig. D.3

Same as Figure D.1, but for NGC 247.
thumbnail Fig. D.4

Same as Figure D.1, but for NGC 253.
thumbnail Fig. D.5

Same as Figure D.1, but for NGC 300.
thumbnail Fig. D.6

Same as Figure D.1, but for NGC 1313.
thumbnail Fig. D.7

Same as Figure D.1, but for NGC 3109.
thumbnail Fig. D.8

Same as Figure D.1, but for Sextans A.
thumbnail Fig. D.9

Same as Figure D.1, but for M83.
thumbnail Fig. D.10

Same as Figure D.1, but for NGC 7793.

Appendix E Spatial distribution diagrams of classified targets

This appendix provides the spatial distribution plots for all galaxies presented in this paper. These figures contain the positions of the same sources that were presented in the CMDs in Appendix D.

thumbnail Fig. E.1

Spatial distribution of classified targets in WLM. The background image is from DSS2.
thumbnail Fig. E.2

Same as Figure E.1, but for NGC 55.
thumbnail Fig. E.3

Same as Figure E.1, but for NGC 247.
thumbnail Fig. E.4

Same as Figure E.1, but for NGC 253.
thumbnail Fig. E.5

Same as Figure E.1, but for NGC 300.
thumbnail Fig. E.6

Same as Figure E.1, but for NGC 1313.
thumbnail Fig. E.7

Same as Figure E.1, but for NGC 3109.
thumbnail Fig. E.8

Same as Figure E.1, but for Sextans A.
thumbnail Fig. E.9

Same as Figure E.1, but for M83.
thumbnail Fig. E.10

Same as Figure E.9, but for NGC 7793.

Appendix F Properties of H II regions and targets with nebular emission

This appendix provides the flux measurements and line measurements for 76 H II regions and 36 other targets with nebular emission, described in Sections 3.3 and 4.3. We list the target ID, the spectral classification, the fluxes of the lines Hα, [N II] λ6583, [S II] λλ6716, and 6731, and the ratios [N II]/Hα and [S II]/Hα.

Table F.1

Flux measurements (10−17 erg cm−2 s−1Ǻ−1) and line ratios.

References

  1. Aghakhanloo, M., Murphy, J. W., Smith, N., & Hložek, R. 2017, MNRAS, 472, 591 [NASA ADS] [CrossRef] [Google Scholar]
  2. Astropy Collaboration (Robitaille, T. P., et al.) 2013, A&A, 558, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Astropy Collaboration (Price-Whelan, A. M., et al.) 2018, AJ, 156, 123 [Google Scholar]
  4. Banerjee, S., Kroupa, P., & Oh, S. 2012, MNRAS, 426, 1416 [NASA ADS] [CrossRef] [Google Scholar]
  5. Beasor, E. R., & Smith, N. 2022, ApJ, 933, 41 [NASA ADS] [CrossRef] [Google Scholar]
  6. Beasor, E. R., Davies, B., Smith, N., et al. 2020, MNRAS, 492, 5994 [Google Scholar]
  7. Beasor, E. R., Davies, B., & Smith, N. 2021, ApJ, 922, 55 [NASA ADS] [CrossRef] [Google Scholar]
  8. Bibby, J. L., & Crowther, P. A. 2012, MNRAS, 420, 3091 [NASA ADS] [CrossRef] [Google Scholar]
  9. Binder, B. A., Carpano, S., et al. 2020, Galaxies, 8, 17 [NASA ADS] [CrossRef] [Google Scholar]
  10. Bonanos, A. Z., Massa, D. L., Sewilo, M., et al. 2009, AJ, 138, 1003 [Google Scholar]
  11. Bonanos, A. Z., Lennon, D. J., Köhlinger, F., et al. 2010, AJ, 140, 416 [NASA ADS] [CrossRef] [Google Scholar]
  12. Bonanos, A. Z., Maravelias, G., Yang, M., et al. 2023, Proceedings of IAU Symposium, 361, in press, [arXiv:2308.04891] [Google Scholar]
  13. Bond, H. E., Bedin, L. R., Bonanos, A. Z., et al. 2009, ApJ, 695, L154 [Google Scholar]
  14. Bouret, J. C., Lanz, T., & Hillier, D. J. 2005, A&A, 438, 301 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  15. Boyer, M. L., McQuinn, K. B. W., Barmby, P., et al. 2015a, ApJ, 800, 51 [Google Scholar]
  16. Boyer, M. L., McQuinn, K. B. W., Barmby, P., et al. 2015b, ApJS, 216, 10 [Google Scholar]
  17. Boyer, M. L., McQuinn, K. B. W., Groenewegen, M. A. T., et al. 2017, ApJ, 851, 152 [NASA ADS] [CrossRef] [Google Scholar]
  18. Bradley, L., Sipőcz, B., Robitaille, T., et al. 2022, https://doi.org/10.5281/zenodo.6385735 [Google Scholar]
  19. Bresolin, F., & Kennicutt, Robert C., J. 2002, ApJ, 572, 838 [NASA ADS] [CrossRef] [Google Scholar]
  20. Bresolin, F., Gieren, W., Kudritzki, R.-P., Pietrzyński, G., & Przybilla, N. 2002, ApJ, 567, 277 [NASA ADS] [CrossRef] [Google Scholar]
  21. Bresolin, F., Schaerer, D., González Delgado, R. M., & Stasińska, G. 2005, A&A, 441, 981 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Bresolin, F., Pietrzyński, G., Urbaneja, M. A., et al. 2006, ApJ, 648, 1007 [NASA ADS] [CrossRef] [Google Scholar]
  23. Bresolin, F., Ryan-Weber, E., Kennicutt, R. C., & Goddard, Q. 2009, ApJ, 695, 580 [Google Scholar]
  24. Bresolin, F., Kudritzki, R.-P., Urbaneja, M. A., et al. 2016, ApJ, 830, 64 [NASA ADS] [CrossRef] [Google Scholar]
  25. Brands, S. A., de Koter, A., Bestenlehner, J. M., et al. 2022, A&A, 663, A36 [Google Scholar]
  26. Britavskiy, N. E., Bonanos, A. Z., Mehner, A., et al. 2014, A&A, 562, A75 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  27. Britavskiy, N. E., Bonanos, A. Z., Mehner, A., et al. 2015, A&A, 584, A33 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Britavskiy, N. E., Bonanos, A. Z., Herrero, A., et al. 2019, A&A, 631, A95 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Chambers, K. C., Magnier, E. A., Metcalfe, N., et al. 2016, arXiv e-prints, [arXiv:1612.05560] [Google Scholar]
  30. Cohen, D. H., Wollman, E. E., Leutenegger, M. A., et al. 2014, MNRAS, 439, 908 [NASA ADS] [CrossRef] [Google Scholar]
  31. Crowther, P. A. 2007, ARA&A, 45, 177 [Google Scholar]
  32. Crowther, P. A., Schnurr, O., Hirschi, R., et al. 2010, MNRAS, 408, 731 [Google Scholar]
  33. Davidge, T. J. 2021, AJ, 162, 152 [NASA ADS] [CrossRef] [Google Scholar]
  34. Davidson, K., & Humphreys, R. M. 1997, ARA&A, 35, 1 [NASA ADS] [CrossRef] [Google Scholar]
  35. Davies, B., Plez, B., & Petrault, M. 2022, MNRAS, 517, 1483 [NASA ADS] [CrossRef] [Google Scholar]
  36. de Wit, S., Bonanos, A. Z., Tramper, F., et al. 2023, A&A, 669, A86 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  37. Decin, L., Hony, S., de Koter, A., et al. 2006, A&A, 456, 549 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Decin, L., Richards, A. M. S., Marchant, P., & Sana, H. 2024, A&A, 681, A17 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  39. Della Bruna, L., Adamo, A., Lee, J. C., et al. 2021, A&A, 650, A103 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  40. Duchêne, G., & Kraus, A. 2013, ARA&A, 51, 269 [Google Scholar]
  41. Dupree, A. K., Strassmeier, K. G., Calderwood, T., et al. 2022, ApJ, 936, 18 [NASA ADS] [CrossRef] [Google Scholar]
  42. Elias, J. H., Frogel, J. A., & Humphreys, R. M. 1985, ApJS, 57, 91 [NASA ADS] [CrossRef] [Google Scholar]
  43. Evans, C. J., Bresolin, F., Urbaneja, M. A., et al. 2007, ApJ, 659, 1198 [Google Scholar]
  44. Fazio, G. G., Hora, J. L., Allen, L. E., et al. 2004, ApJS, 154, 10 [Google Scholar]
  45. Figer, D. F. 2005, Nature, 434, 192 [Google Scholar]
  46. Förster, F., Moriya, T. J., Maureira, J. C., et al. 2018, Nat. Astron., 2, 808 [CrossRef] [Google Scholar]
  47. Fox, O. D., Fransson, C., Smith, N., et al. 2020, MNRAS, 498, 517 [NASA ADS] [CrossRef] [Google Scholar]
  48. Freudling, W., Romaniello, M., Bramich, D. M., et al. 2013, A&A, 559, A96 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  49. Fullerton, A. W., Massa, D. L., & Prinja, R. K. 2006, ApJ, 637, 1025 [Google Scholar]
  50. Gaia Collaboration (Prusti, T., et al.) 2016, A&A, 595, A1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  51. Gaia Collaboration (Brown, A. G. A., et al.) 2018, A&A, 616, A1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  52. Gal-Yam, A. 2012, Science, 337, 927 [Google Scholar]
  53. Gal-Yam, A. 2019, ARA&A, 57, 305 [Google Scholar]
  54. González-Torà, G., Davies, B., Kudritzki, R.-P., & Plez, B. 2021, MNRAS, 505, 4422 [CrossRef] [Google Scholar]
  55. Gordon, K. D., Meixner, M., Meade, M. R., et al. 2011, AJ, 142, 102 [NASA ADS] [CrossRef] [Google Scholar]
  56. Grammer, S. H., Humphreys, R. M., & Gerke, J. 2015, AJ, 149, 152 [CrossRef] [Google Scholar]
  57. Gvaramadze, V. V., Kniazev, A. Y., & Fabrika, S. 2010, MNRAS, 405, 1047 [NASA ADS] [Google Scholar]
  58. Gvaramadze, V. V., Menten, K. M., Kniazev, A. Y., et al. 2014, MNRAS, 437, 843 [Google Scholar]
  59. Hadfield, L. J., & Crowther, P. A. 2007, MNRAS, 381, 418 [CrossRef] [Google Scholar]
  60. Hadfield, L. J., Crowther, P. A., Schild, H., & Schmutz, W. 2005, A&A, 439, 265 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  61. Hartoog, O. E., Sana, H., de Koter, A., & Kaper, L. 2012, MNRAS, 422, 367 [NASA ADS] [CrossRef] [Google Scholar]
  62. Heida, M., Lau, R. M., Davies, B., et al. 2019, ApJ, 883, L34 [NASA ADS] [CrossRef] [Google Scholar]
  63. Hernandez, S., Larsen, S., Aloisi, A., et al. 2019, ApJ, 872, 116 [NASA ADS] [CrossRef] [Google Scholar]
  64. Hlavacek-Larrondo, J., Carignan, C., Daigle, O., et al. 2011, MNRAS, 411, 71 [NASA ADS] [CrossRef] [Google Scholar]
  65. Hosek, Matthew W., J., Kudritzki, R.-P., Bresolin, F., et al. 2014, ApJ, 785, 151 [NASA ADS] [CrossRef] [Google Scholar]
  66. Humphreys, R. M., & Davidson, K. 1994, PASP, 106, 1025 [NASA ADS] [CrossRef] [Google Scholar]
  67. Humphreys, R. M., Stangl, S., Gordon, M. S., Davidson, K., & Grammer, S. H. 2019, AJ, 157, 22 [CrossRef] [Google Scholar]
  68. Jones, O. C., Woods, P. M., Kemper, F., et al. 2017, MNRAS, 470, 3250 [NASA ADS] [CrossRef] [Google Scholar]
  69. Kalari, V. M., Horch, E. P., Salinas, R., et al. 2022, ApJ, 935, 162 [NASA ADS] [CrossRef] [Google Scholar]
  70. Karczmarek, P., Pietrzyński, G., Gieren, W., et al. 2021, ApJS, 253, 42 [NASA ADS] [CrossRef] [Google Scholar]
  71. Kewley, L. J., & Ellison, S. L. 2008, ApJ, 681, 1183 [Google Scholar]
  72. Khan, R. 2017, ApJS, 228, 5 [NASA ADS] [CrossRef] [Google Scholar]
  73. Khan, R., Stanek, K. Z., Kochanek, C. S., & Bonanos, A. Z. 2011, ApJ, 732, 43 [NASA ADS] [CrossRef] [Google Scholar]
  74. Khan, R., Stanek, K. Z., Kochanek, C. S., & Sonneborn, G. 2015, ApJS, 219, 42 [NASA ADS] [CrossRef] [Google Scholar]
  75. Kniazev, A. Y., Grebel, E. K., Pustilnik, S. A., Pramskij, A. G., & Zucker, D. B. 2005, AJ, 130, 1558 [NASA ADS] [CrossRef] [Google Scholar]
  76. Kopsacheili, M., Zezas, A., & Leonidaki, I. 2020, MNRAS, 491, 889 [CrossRef] [Google Scholar]
  77. Koumpia, E., Oudmaijer, R. D., Graham, V., et al. 2020, A&A, 635, A183 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  78. Kourniotis, M., Bonanos, A. Z., Yuan, W., et al. 2017, A&A, 601, A76 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  79. Kovtyukh, V. V., Gorlova, N. I., & Belik, S. I. 2012, MNRAS, 423, 3268 [NASA ADS] [CrossRef] [Google Scholar]
  80. Kraus, M. 2019, Galaxies, 7, 83 [NASA ADS] [CrossRef] [Google Scholar]
  81. Kudritzki, R.-P., & Puls, J. 2000, ARA&A, 38, 613 [Google Scholar]
  82. Kudritzki, R.-P., Urbaneja, M. A., Bresolin, F., et al. 2008, ApJ, 681, 269 [NASA ADS] [CrossRef] [Google Scholar]
  83. Leonidaki, I., Boumis, P., & Zezas, A. 2013, MNRAS, 429, 189 [NASA ADS] [CrossRef] [Google Scholar]
  84. Levesque, E. M., Massey, P., Olsen, K. A. G., et al. 2006, ApJ, 645, 1102 [NASA ADS] [CrossRef] [Google Scholar]
  85. Long, K. S., Blair, W. P., Winkler, P. F., et al. 2022, ApJ, 929, 144 [NASA ADS] [CrossRef] [Google Scholar]
  86. Lorenzo, M., Garcia, M., Najarro, F., et al. 2022, MNRAS, 516, 4164 [NASA ADS] [CrossRef] [Google Scholar]
  87. Mackey, J., Mohamed, S., Neilson, H. R., Langer, N., & Meyer, D. M. A. 2012, ApJ, 751, L10 [NASA ADS] [CrossRef] [Google Scholar]
  88. Maravelias, G., Bonanos, A. Z., Tramper, F., et al. 2022, A&A, 666, A122 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  89. Maravelias, G., de Wit, S., Bonanos, A. Z., et al. 2023, Galaxies, 11, 79 [NASA ADS] [CrossRef] [Google Scholar]
  90. Mashonkina, L. 2020, MNRAS, 493, 6095 [CrossRef] [Google Scholar]
  91. Massey, P., McNeill, R. T., Olsen, K. A. G., et al. 2007, AJ, 134, 2474 [NASA ADS] [CrossRef] [Google Scholar]
  92. Massey, P., Neugent, K. F., Ekström, S., et al. 2023, ApJ, 942, 69 [NASA ADS] [CrossRef] [Google Scholar]
  93. Mathewson, D. S., & Clarke, J. N. 1973, ApJ, 180, 725 [NASA ADS] [CrossRef] [Google Scholar]
  94. Mauerhan, J. C., Smith, N., Filippenko, A. V., et al. 2013, MNRAS, 430, 1801 [NASA ADS] [CrossRef] [Google Scholar]
  95. McMahon, R. 2012, in Science from the Next Generation Imaging and Spectroscopic Surveys, 37 [Google Scholar]
  96. Meixner, M., Gordon, K. D., Indebetouw, R., et al. 2006, AJ, 132, 2268 [NASA ADS] [CrossRef] [Google Scholar]
  97. Meynet, G., Chomienne, V., Ekström, S., et al. 2015, A&A, 575, A60 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  98. Mikołajewska, J., Caldwell, N., Shara, M. M., et al. 2015, ApJ, 799, L16 [CrossRef] [Google Scholar]
  99. Moe, M., & Di Stefano, R. 2017, ApJS, 230, 15 [Google Scholar]
  100. Mohamed, S., Mackey, J., & Langer, N. 2012, A&A, 541, A1 [Google Scholar]
  101. Morozova, V., Piro, A. L., & Valenti, S. 2018, ApJ, 858, 15 [Google Scholar]
  102. Neill, J. D., Sullivan, M., Gal-Yam, A., et al. 2011, ApJ, 727, 15 [NASA ADS] [CrossRef] [Google Scholar]
  103. Oey, M. S., & Clarke, C. J. 2005, ApJ, 620, L43 [NASA ADS] [CrossRef] [Google Scholar]
  104. Owocki, S. P., & Puls, J. 1999, ApJ, 510, 355 [Google Scholar]
  105. Patrick, L. R., Evans, C. J., Davies, B., et al. 2017, MNRAS, 468, 492 [NASA ADS] [CrossRef] [Google Scholar]
  106. Prieto, J. L., Kistler, M. D., Thompson, T. A., et al. 2008, ApJ, 681, L9 [NASA ADS] [CrossRef] [Google Scholar]
  107. Rieke, G. H., Young, E. T., Engelbracht, C. W., et al. 2004, ApJS, 154, 25 [Google Scholar]
  108. Ruffle, P. M. E., Kemper, F., Jones, O. C., et al. 2015, MNRAS, 451, 3504 [NASA ADS] [CrossRef] [Google Scholar]
  109. Rumstay, K. S., & Kaufman, M. 1983, ApJ, 274, 611 [NASA ADS] [CrossRef] [Google Scholar]
  110. Sana, H., de Mink, S. E., de Koter, A., et al. 2012, Science, 337, 444 [Google Scholar]
  111. Schild, H., Crowther, P. A., Abbott, J. B., & Schmutz, W. 2003, A&A, 397, 859 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  112. Smartt, S. J. 2015, PASA, 32, e016 [NASA ADS] [CrossRef] [Google Scholar]
  113. Smartt, S. J., Eldridge, J. J., Crockett, R. M., & Maund, J. R. 2009, MNRAS, 395, 1409 [CrossRef] [Google Scholar]
  114. Smith, N. 2014, ARA&A, 52, 487 [NASA ADS] [CrossRef] [Google Scholar]
  115. Smith, N. 2017, in Handbook of Supernovae, eds. A. W. Alsabti, & P. Murdin, 403 [Google Scholar]
  116. Smith, N., & Frew, D. J. 2011, MNRAS, 415, 2009 [NASA ADS] [CrossRef] [Google Scholar]
  117. Smith, N., Mauerhan, J. C., & Prieto, J. L. 2014, MNRAS, 438, 1191 [NASA ADS] [CrossRef] [Google Scholar]
  118. Smith, N., Andrews, J. E., Filippenko, A. V., et al. 2022, MNRAS, 515, 71 [NASA ADS] [CrossRef] [Google Scholar]
  119. Spinoglio, L., Fernández-Ontiveros, J. A., Malkan, M. A., et al. 2022, ApJ, 926, 55 [NASA ADS] [CrossRef] [Google Scholar]
  120. Storchi-Bergmann, T., Calzetti, D., & Kinney, A. L. 1994, ApJ, 429, 572 [NASA ADS] [CrossRef] [Google Scholar]
  121. Tody, D. 1986, SPIE Conf. Ser., 627, 733 [Google Scholar]
  122. Tody, D. 1993, in Astronomical Society of the Pacific Conference Series, 52, Astronomical Data Analysis Software and Systems II, eds. R. J. Hanisch, R. J. V. Brissenden, & J. Barnes, 173 [Google Scholar]
  123. Tully, R. B., Courtois, H. M., Dolphin, A. E., et al. 2013, AJ, 146, 86 [NASA ADS] [CrossRef] [Google Scholar]
  124. Urbaneja, M. A., Kudritzki, R.-P., Bresolin, F., et al. 2008, ApJ, 684, 118 [NASA ADS] [CrossRef] [Google Scholar]
  125. Verhoelst, T., van der Zypen, N., Hony, S., et al. 2009, A&A, 498, 127 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  126. Villar, V. A., Berger, E., Chornock, R., et al. 2016, ApJ, 830, 11 [CrossRef] [Google Scholar]
  127. Wachter, S., Mauerhan, J. C., Van Dyk, S. D., et al. 2010, AJ, 139, 2330 [NASA ADS] [CrossRef] [Google Scholar]
  128. Werner, M. W., Roellig, T. L., Low, F. J., et al. 2004, ApJS, 154, 1 [NASA ADS] [CrossRef] [Google Scholar]
  129. Whitmore, B. C., Allam, S. S., Budavári, T., et al. 2016, AJ, 151, 134 [NASA ADS] [CrossRef] [Google Scholar]
  130. Williams, S. J., & Bonanos, A. Z. 2016, A&A, 587, A121 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  131. Williams, S. J., Bonanos, A. Z., Whitmore, B. C., Prieto, J. L., & Blair, W. P. 2015, A&A, 578, A100 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  132. Winkler, P. F., Long, K. S., Blair, W. P., & Points, S. D. 2023, ApJ, 943, 15 [NASA ADS] [CrossRef] [Google Scholar]
  133. Woods, P. M., Oliveira, J. M., Kemper, F., et al. 2011, MNRAS, 411, 1597 [Google Scholar]
  134. Yang, M., Bonanos, A. Z., Jiang, B.-W., et al. 2020, A&A, 639, A116 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  135. Yang, M., Bonanos, A. Z., Jiang, B., et al. 2023, A&A, 676, A84 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
4

HST images for NGC 247-143, NGC 247-157, and NGC 300-274 indicate galaxies, rather than the approximate “star,” “red,” and “cool star” classifications from the spectra, which were found to be galaxies at redshifts z = 0.133, 0.156 and 0.163, respectively.

5

Reduction artifacts prevented a more accurate classification.

6

WLM-1120 corresponds to star 10 in Britavskiy et al. (2015).

All Tables

Table 1

Properties of target galaxies.

In the text
Table 2

Distribution of targets per galaxy and per photometric catalog.

In the text
Table 3

Definition of the priority system and the criteria for each priority class.

In the text
Table 4

Distribution of target priorities per galaxy.

In the text
Table 5

Overview of masks.

In the text
Table 6

Distribution of classified targets per galaxy and spectral class.

In the text
Table 7

Catalog of spectral classifications.

In the text
Table 8

Catalog of matched evolved stars.

In the text
Table 9

Distribution of classified, priority (Prio), and massive stars (MS) per galaxy and spectral class.

In the text
Table C.1

Log of pre-imaging observations.

In the text
Table C.2

Log of MXU observations.

In the text
Table F.1

Flux measurements (10−17 erg cm−2 s−1Ǻ−1) and line ratios.

In the text

All Figures

thumbnail Fig. 1

Three overlapping fields in M83. P1 and P2 targets are indicated. North is up and east to the left.
In the text
thumbnail Fig. 2

Representative spectra of targets classified as RSGs, following a temperature sequence from K-type (top) to M-type (bottom). Shaded areas correspond to regions with TiO absorption; telluric absorption is indicated with a ⊕ symbol.
In the text
thumbnail Fig. 3

Spectra of all six targets classified as YSGs. The strength of the oxygen triplet at λ7772 indicates a supergiant luminosity class. Other key spectral lines are indicated; telluric absorption is indicated with a ⊕ symbol.
In the text
thumbnail Fig. 4

Representative spectra of BSGs that exhibit Hα emission. The Hα profiles are a superposition of the nebular line profile broadened at the base by a circumstellar component, indicating significant mass loss for these objects.
In the text
thumbnail Fig. 5

Example spectra of targets classified as H II regions or galaxies (including AGN and quasars). Their redshifts were calculated from the emission lines indicated for each of these targets.
In the text
thumbnail Fig. 6

IR and optical cumulative absolute magnitude CMDs for all evolved massive stars and H II regions. Background sources are from NGC 300. The dashed lines in the first panel indicate the criteria used for target selection; all our priority targets are located within the box at the top right defined by these lines.
In the text
thumbnail Fig. 7

[N II]/Hα vs. [S II]/Hα for all H II regions and spectra classified as nebulae found in eight of our target galaxies, which range from Z=0.2–1.6 Z⊙. The vertical line indicates the value [S II]/Hα = 0.4, above which the emission is shocked. The diagonal shows the 1:1 relation.
In the text
thumbnail Fig. A.1

Spitzer CMD of WLM.
In the text
thumbnail Fig. A.2

Spitzer CMD of NGC 55.
In the text
thumbnail Fig. A.3

Spitzer CMD of NGC 247.
In the text
thumbnail Fig. A.4

Spitzer CMD of NGC 253.
In the text
thumbnail Fig. A.5

Spitzer CMD of NGC 300.
In the text
thumbnail Fig. A.6

Spitzer CMD of NGC 1313.
In the text
thumbnail Fig. A.7

Spitzer CMD of NGC 3109.
In the text
thumbnail Fig. A.8

Spitzer CMD of Sextans A.
In the text
thumbnail Fig. A.9

Spitzer CMD of M83.
In the text
thumbnail Fig. A.10

Spitzer CMD of NGC 7793.
In the text
thumbnail Fig. B.1

Mask layout for WLM Field A.
In the text
thumbnail Fig. B.2

Mask layout for NGC 55 Field A.
In the text
thumbnail Fig. B.3

Mask layout for NGC 55 Field B.
In the text
thumbnail Fig. B.4

Mask layout for NGC 55 Field C.
In the text
thumbnail Fig. B.5

Mask layout for NGC 247 Field A.
In the text
thumbnail Fig. B.6

Mask layout for NGC 247 Field B.
In the text
thumbnail Fig. B.7

Mask layout for NGC 253 Field A.
In the text
thumbnail Fig. B.8

Mask layout for NGC 253 Field B.
In the text
thumbnail Fig. B.9

Mask layout for NGC 253 Field C.
In the text
thumbnail Fig. B.10

Mask layout for NGC 300 Field A.
In the text
thumbnail Fig. B.11

Mask layout for NGC 300 Field B.
In the text
thumbnail Fig. B.12

Mask layout for NGC 300 Field C.
In the text
thumbnail Fig. B.13

Mask layout for NGC 300 Field D.
In the text
thumbnail Fig. B.14

Mask layout for NGC 1313 Field A.
In the text
thumbnail Fig. B.15

Mask layout for NGC 3109 Field A.
In the text
thumbnail Fig. B.16

Mask layout for Sextans A Field A.
In the text
thumbnail Fig. B.17

Mask layout for M83 Field A.
In the text
thumbnail Fig. B.18

Mask layout for M83 Field B.
In the text
thumbnail Fig. B.19

Mask layout for M83 Field C.
In the text
thumbnail Fig. B.20

Mask layout for NGC 7793 Field A.
In the text
thumbnail Fig. D.1

IR and optical CMDs for the classified targets of interest in WLM, i.e., all evolved massive stars, H II regions, and background objects.
In the text
thumbnail Fig. D.2

Same as Figure D.1, but for NGC 55.
In the text
thumbnail Fig. D.3

Same as Figure D.1, but for NGC 247.
In the text
thumbnail Fig. D.4

Same as Figure D.1, but for NGC 253.
In the text
thumbnail Fig. D.5

Same as Figure D.1, but for NGC 300.
In the text
thumbnail Fig. D.6

Same as Figure D.1, but for NGC 1313.
In the text
thumbnail Fig. D.7

Same as Figure D.1, but for NGC 3109.
In the text
thumbnail Fig. D.8

Same as Figure D.1, but for Sextans A.
In the text
thumbnail Fig. D.9

Same as Figure D.1, but for M83.
In the text
thumbnail Fig. D.10

Same as Figure D.1, but for NGC 7793.
In the text
thumbnail Fig. E.1

Spatial distribution of classified targets in WLM. The background image is from DSS2.
In the text
thumbnail Fig. E.2

Same as Figure E.1, but for NGC 55.
In the text
thumbnail Fig. E.3

Same as Figure E.1, but for NGC 247.
In the text
thumbnail Fig. E.4

Same as Figure E.1, but for NGC 253.
In the text
thumbnail Fig. E.5

Same as Figure E.1, but for NGC 300.
In the text
thumbnail Fig. E.6

Same as Figure E.1, but for NGC 1313.
In the text
thumbnail Fig. E.7

Same as Figure E.1, but for NGC 3109.
In the text
thumbnail Fig. E.8

Same as Figure E.1, but for Sextans A.
In the text
thumbnail Fig. E.9

Same as Figure E.1, but for M83.
In the text
thumbnail Fig. E.10

Same as Figure E.9, but for NGC 7793.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.