EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 663 (July 2022) A&A, 663 (2022) A74 Full HTML
Free Access
Issue
A&A
Volume 663, July 2022
Article Number A74
Number of page(s) 9
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/202243118
Published online 14 July 2022

© ESO 2022

1. Introduction

In the magnetospheric accretion scenario, the accretion of material from a circumstellar disk in low-mass pre-main sequence (PMS) stars is funneled by the stellar magnetic field, which disrupts the disk at a few stellar radii (Hartmann et al. 2016, and reference therein). Our understanding of this process is still not entirely clear. A key parameter describing the star–disk evolution is the rate of mass accretion, that is, the rate at which mass from the circumstellar disk is transferred onto the central PMS star (see, e.g., review by Hartmann et al. 2016). In particular, it is important to evaluate the relation between mass accretion rate, stellar mass, and age, how the mass accretion rate changes as a star approaches its main sequence (MS), and how the metallicity or in general the chemical composition of the parent molecular cloud could impact the formation and evolution of the star.

Usually, mass accretion rates are derived from the analysis of continuum veiling, ultraviolet (UV) excess emission, or indeed through a detailed study of the profile and intensity of hydrogen emission lines (e.g., Hα, Paβ, Brγ), which requires medium- to high-resolution spectroscopy for each individual object. Even with modern multi-object spectrographs at the largest ground-based telescopes, these methods can be applied to relatively nearby star-forming regions (d ≲ 1−2 kpc), because of crowding. For this reason, the properties of low-metallicity PMS stars located in extra-galactic star-forming regions remain poorly known.

In the last decade, De Marchi et al. (2010) developed an efficient method based on Hubble Space Telescope (HST) photometry that allows the identification of hundreds of PMS stars simultaneously, and the determination of their physical parameters, including effective temperature, luminosity, age, mass, Hα luminosity, accretion luminosity, and mass accretion rate, with an uncertainty of between 15% and 20%, comparable to that allowed by spectroscopy.

This method has been successfully applied not only to regions of the Milky Way (Beccari et al. 2010, 2015; Zeidler et al. 2016), but also to regions of the Small (De Marchi et al. 2011, 2013) and Large Magellanic Clouds (e.g., Biazzo et al. 2019; De Marchi et al. 2017; Spezzi et al. 2012). This method combines V (F555W) and I (F814W) broadband photometry with narrow-band Hα (F656N or F658N) imaging to identify the stars with excess in Hα emission and to determine their associated Hα emission equivalent width, EWHα, the Hα luminosity and the accretion properties of the PMS stars selected.

In this work, we use this method to select and study the PMS populations of the stellar system LH 91 (Lucke 1974) in the northeast outer edge of the super-giant shell LMC 4 in the Large Magellanic Cloud (LMC). This area, investigated with Hα and radio observations by Book et al. (2009), also covers LH 91 I in the southeast of LH 91 (Kontizas et al. 1994) and LH 95 in the north of LH 91 (Lucke 1974).

The most recent work on LH 91 was presented by Gouliermis et al. (2002) using ground-based BVR and Hα photometry. Studying the Hα topography of the area, the authors found that LH 91 is loosely related to an H II region, which seems to be large and rather diffuse. In agreement with Lucke (1974), the authors confirm that LH 91 does not seem to represent a “classical” stellar system in which the stars are physically related to each other. Analyzing the color–magnitude diagram (CMD) in the B and V band, the authors estimated the color excess E(B − V) = 0.16 ± 0.04 using the reddening-free Wesenheit function. Moreover, fitting the Geneva isochrones (Schaerer et al. 1993) derived adopting metallicity Z = 0.008, Gouliermis et al. (2002) derived the age of the system, finding it to be younger than 10 Myr, similar to that of LH 95 and LH 91I, and in agreement with for example Braun et al. (1997, 2000). Instead, Kontizas et al. (1994) estimated an age of about 20 Myr. Finally, Gouliermis et al. (2002) also estimated the age of the background field, the population of the observed area around LH 91, to be older than 50 Myr and up to 1.25 Gyr.

This paper is organized as follows: in Sect. 2 we describe the HST photometric observations, in Sect. 3 we illustrate the analysis needed to identify the PMS stars and to estimate the luminosity and the equivalent width (EW) associated to the Hα excess. In Sect. 4 we measure the physical properties of the stars selected. In Sect. 5 we determine the accretion properties of the selected PMS stars, that is, the accretion luminosity and mass accretion rate, and we show the relation between the mass accretion rate and the stellar properties of the PMS objects, such as their mass and age. We also compare our results with the findings for other star-forming regions in the LMC with the same metallicity, and in particular with LH 95, the closest region to LH 91 for which accretion properties of PMS candidates have been derived (Biazzo et al. 2019). We present our conclusions in the last section.

2. Photometric observations

The LH 91 region was observed with the Wide Field Camera 3 (WFC3/UV) on board the HST in the broad-band filters F555W and F814W, and in the narrow-band filter F656N, the latest centered on the Hα line. The data were collected as part of HST programs #12872 (PI: Da Rio) and #13009 (PI: De Marchi). A short logbook of the observations is shown in Table 1.

Table 1.

Logbook of the observations.

The data were reduced using the standard DAOPHOTII (Stetson 1987) procedure. A list of 10 to 20 well-sampled and isolated stars were used to model the point spread function (PSF) on the F555W and F814W images, and a deep photometric catalog of stars was derived via PSF fitting on the images acquired with the broad-band filters. The final magnitude of each star in a given filter is estimated as the mean of the photometric measures in each individual image taken with that filter, while the standard deviation is taken as the associated error. Aperture photometry was then used to extract the F656N magnitude for each star detected in the optical bands. The choice of performing aperture photometry on the narrow-band images is driven by the fact that such images are characterized by very little stellar crowding. As such, the aperture photometry is the ideal choice as it allows accurate estimation of the magnitude free from any uncertainty that is unavoidably associated with the choice of PSF model. We stress here that the background is locally estimated and subtracted in an annulus around the start. The final catalog of the overlapping fields contains 9423 objects, of which 6980 have a measure in the F656N band. These F555W and F814W band observations are among the deepest ever taken toward the LH 91 region. The instrumental magnitudes in F555W, F814W, and F656N were calibrated to the VEGAMAG photometric system using the zero-point values made available by the Space Telescope Science Institute1.

3. Data analysis

3.1. PMS star identification

We applied the method developed by De Marchi et al. (2010) to identify the PMS stars characterized by an active mass accretion process. We measured the physical and accretion properties of these objects (i.e., Hα luminosity, Hα emission EWHα, mass accretion rate, and accretion luminosity) using photometric data. We refer to De Marchi et al. (2010) for a detailed discussion of the method, while in this work we describe some fundamental steps.

We selected PMS stars on the basis of their Hα excess emission (White & Basri 2003). First of all, we identified the Hα excess emitters in the (m555 − m656) vs. (m555 − m814) color–color diagram shown in Fig. 1. The magnitudes were corrected for the extinction contribution of the Milky Way considering the values mag and E(m555 − m814)MW = 0.1 (Fitzpatrick & Savage 1984). To this aim, we selected from our catalog in F555W, F814W, and F656N bands all those stars whose photometric uncertainties, that is δ555, δ814, and δ656, are less than 0.05 in each individual band. A total of 254 stars satisfied these conditions (gray filled dots in the color–color diagram in Fig. 1), out of 9423 sources in the whole catalog. These are typically MS stars that do not present an appreciable Hα excess. With these stars, we define a reference sequence (dashed black line) with respect to which the excess Hα emission is computed. The dotted blue line of Fig. 1 represents the theoretical color relationship obtained using the Bessell et al. (1998) model atmospheres for MS stars with the chemical and physical parameters appropriate for the LMC (effective temperature Teff in the range of 3500–40 000 K, surface gravity log g = 4.5, and metallicity index [M/H] ≃ −0.5, Colucci et al. 2012). The agreement between our reference sequence and the theoretical one is evident at mF555W − mF814W <  1. The discrepancy between the models and the data at mF555W − mF814W >  1 can be attributed to small number statistics and to the fact that the majority of these objects are red giants, with different physical characteristics from those assumed in the models.

thumbnail Fig. 1.

Color–color diagram of the selected stars in the field of LH 91. All magnitudes are already corrected for the extinction contribution of our Galaxy, mag and E(m555 − m814)MW = 0.1. The arrow shows the reddening vector of E(m555 − m814) = 0.25 and E(m555 − m656) = 0.13 for the adopted LH 91 extinction law. The dashed line represents the median photospheric (m555 − m814) color for the 254 stars with δ555, δ814, and δ656 <  0.05 (gray filled dots). The dotted line shows the model atmospheres of Bessell et al. (1998) computed for the three WFC3/UVES filters. The PMS star candidates with Hα emission excess at the 3σ level are represented with large red dots. The large green dots are the brightest PMS star candidates, with m555 <  20 mag. Error bars are also shown.

To select the most probable PMS stars, after the exclusion of the 254 stars taken as reference, we first selected the targets with photometric uncertainties in each individual band as follows: δ555 and δ814 <  0.1 mag, and δ656 <  0.3 mag, for a total of 1309 objects. As highlighted by De Marchi et al. (2010), the contribution of the Hα line to the m555 magnitude is negligible, and therefore we can define the magnitude of the excess emission as:

(1)

where the superscripts “obs” and “ref” refer to the observation and reference sequence, respectively. We then considered the stars with ΔHα exceeding at least three times the combined mean photometric uncertainties in the three bands δ3:

(2)

A total of 187 stars satisfy these conditions; they are indicated with large red dots in Fig. 1. This means that 187 stars have (m555 − m656) colors exceeding that of the reference template at the given (m555 − m814) color by more than three times the combined uncertainties on their (m555 − m656) values. The large green dots in Fig. 1 are the targets selected with m555 <  20 mag, which we exclude from our following analysis as we are interested primarily in low-mass PMS candidates. Our final sample of PMS candidates is therefore composed by 181 targets. As in these bands the reddening vector due to LH 91 runs almost parallel to the median of the reference sequence (De Marchi et al. 2010), the color–color diagram provides a robust identification of stars with Hα excess even before correction for LH 91 reddening.

3.2. The color–magnitude diagram

We applied the correction for the extinction contribution of the Milky Way and LH 91 to the magnitudes in each band. For the Milky Way, we report the values in the previous section. We estimated the extinction for LH 91 from the value of E(B − V) = 0.16 ± 0.04 color excess in the photometry of Gouliermis et al. (2002) and converted into AV assuming the average LMC reddening law R555 = A555/E(m555 − m814) = 2.97 calculated by De Marchi & Panagia (2014). As Gouliermis et al. (2002) found that the density of the ambient medium in LH 91 is similar to the value for LH 95, and as Da Rio et al. (2009) did not find a significant level of differential extinction while studying the upper MS stars of the latter, we also consider the differential reddening to be negligible in LH 91. We show the CMD (m555 − m814)0 vs. (m555)0 in Fig. 2. The small black dots are the targets of the whole sample, namely 9423 stars. To estimate the age of the system, we fit the CMD with the isochrone models for Z = 0.007 – which is typical of young LMC stars (e.g., Colucci et al. 2012) – taken from the PAdova-Trieste Stellar Evolution Code (PARSEC, Bressan et al. 2012) and distance modulus (m − M)0 = 18.55 (Panagia et al. 1991; Panagia 1999). The turnoff at m555 ∼ 20.5 mag and the red clump at m555 ∼ 19.5 mag and m555 − m814 ∼ 1.0 are best matched by a 1.5 Gyr isochrone (dashed light-blue line), in agreement with the age of the background field stars evaluated by Gouliermis et al. (2002). Stars with Hα excess show a wide apparent spread towards young age and could be divided in two groups, separated by an isochrone at 8 Myr (solid green line).

thumbnail Fig. 2.

Color–magnitude diagram of the field of LH 91. All magnitudes are already corrected for the extinction contribution of our Galaxy and LH 91. The small black dots are the targets of the whole sample (9423 stars). Small gray-filled dots are the stars with photometric uncertainties < 0.05 mag in each band. The large red dots represent the PMS star candidates with Hα excess emission at the 3σ level. Solid green and dashed light-blue lines show the theoretical isochrones from Bressan et al. (2012) for ages 8 Myr and 1.5 Gyr, respectively, metallicity Z = 0.007, and a distance modulus (mV − MV)0 = 18.55.

3.3. From Hα color excess to Hα luminosity

To avoid contamination by stars with significant chromospheric activity, we also imposed constraints on EWHα, selecting only stars with EWHα ≥ 10 Å, because according to De Marchi et al. (2010) this is a reliable cutoff to separate accretors from those not accreting .

For details of the method used here to derive EWHα from the photometry, we refer to De Marchi et al. (2010, 2011, 2013). Here we recall that, as the width of the Hα line is narrow with respect to the width of the filter, the measure of EWHα is given by the difference between the observed Hα line magnitude and the level of the Hα continuum (ΔHα). If we assume that the stars used to define the reference sequence have no Hα absorption features, their (m555 − m656) color represents the color of the pure continuum. Consequently, we calculated the EWHα from the following relation:

(3)

where RECTW is the rectangular width of the F656N filter. The uncertainty on the EWHα measure is dominated by the uncertainty on the Hα magnitude.

Moreover, because of the width of the F656N filter, the small contribution due to the emission of the forbidden N II line at λ6548 is included in ΔHα. Therefore, following the prescriptions by De Marchi et al. (2010), we estimated corrections ranging from 0.2 to 1.4 Å, to be subtracted from the EWHα of our targets. Figure 3 shows the EWHα measured for the selected low-mass PMS candidates as a function of the de-reddened m555 − m814 color.

thumbnail Fig. 3.

Hα equivalent width of the selected low-mass PMS candidates, as a function of the de-reddened (m555 − m814) color. The red squares represent the values of the PMS stars younger than 8 Myr, the blue dots are the oldest PMS stars.

We performed a preliminary study of the EWHα distribution of the PMS candidates at different ages using the isochrone at 8 Myr as a discriminating factor. We divided the sample into stars older (blue dots) and younger (red squares) than 8 Myr (Fig. 3).

The values of EWHα for the sample range from ∼3 Å to ∼17 Å, with a median of ∼9 Å that applies to both the whole sample and the two subgroups. After the selection on the EWHα, a total of 75 objects satisfy the conservative condition (EWHα ≥ 10 Å). The median value of the EWHα is about 12 Å, regardless of age, smaller than the values found in other star formation fields in the LMC, such as LH 95 (EWHα ∼ 30 Å, Biazzo et al. 2019) and SN 1987A (EWHα ∼ 20 Å, De Marchi et al. 2010). The difference could be due to the paucity and to stellar mass range of our sample. Moreover, the figure shows an almost clear separation in color between the two subgroups in LH 91 with the exception of the target with (m555 − m814)0 ∼ 0.4 and EWHα ∼ 10 Å. As the coordinates of this target correspond to those of a massive star in the 2MASS catalog, it could be a Be star. A similar separation in color between the two subgroups of PMS stars was found in LH 95 (Biazzo et al. 2019).

The Hα emission line luminosity LHα can be determined from the absolute sensitivity of the instrumental setup, the photometric zero point (ZP), the distance of the stars, and from the magnitude in the Hα band:

(4)

The values of the photometric properties of the instruments were taken from Ryon (2018), namely the inverse sensitivity PHOTFLAM = 1.714 × 10−17 erg cm−2 s−1 Å−1, and the zero point in the VEGAmag system for the Hα filter, ZP = 19.84 (Calamida et al. 2021). Assuming a distance of 51.4 ± 1.2 kpc (Panagia 1999) and considering the rectangular width of the F656N filter RECTW = 17.679 Å, we determined the Hα luminosity for the 75 targets, finding a median value of about 8.7 × 1030 erg s−1 (0.2 × 10−2 L⊙).

This value is slightly lower than the one found by Biazzo et al. (2019) in the LH 95 association (∼1.2 × 1031 erg s−1, 0.3 × 10−2 L⊙). This is not surprising because Gouliermis et al. (2002) found that the mean Hα intensities of the HII region related to LH 95 (DEM L 252) is about two times higher than the corresponding intensity of the HII region associated with LH 91 (DEM L 251).

In addition, we can compared our result with the median Hα luminosity of other regions of the LMC, namely 30 Doradus Nebula and SN 1987A field; also in these cases our value is lower, the mean LHα estimated in these regions is ∼4 × 1031 erg s−1 (∼10−2 L⊙, De Marchi et al. 2010) and ∼1.5 × 1032 erg s−1 (∼4 × 10−2 L⊙De Marchi et al. 2017) respectively.

The uncertainty on LHα is dominated by the uncertainties on the Hα photometry, on the distance (∼5%) and on instrumental setup (∼3%) (see De Marchi et al. 2010). The total uncertainty on LHα is about 16%.

4. Physical parameters of the PMS candidates

4.1. Effective temperature and bolometric luminosity

We evaluated the effective temperature of the PMS candidates by comparing the theoretical models with the m555 − m814 color of our sample corrected for the reddening due to the Milky Way and LH 91, as explained in Sect. 3.2. To convert the color to Teff we used the models of Bessell et al. (1998) for 3500 K ≤ Teff ≤ 40 000 K, log g = 4.5, and metallicity index [M/H] = −0.5 dex. As the models of Bessell et al. (1998) are not available for temperatures lower than 3500 K, we used the Teff − (V − IC) calibration by Pecaut & Mamajek (2013), with the assumption that the calibrated m555 and m814 magnitudes coincide with the V and IC magnitudes (see Biazzo et al. 2019).

To obtain the luminosity of the stars L⋆, we considered the magnitude m555 corrected for the interstellar extinction, a distance to LH 91 of 51.4 kpc (Panagia 1999), and a bolometric solar magnitude of 4.74 mag (Pecaut & Mamajek 2013). The uncertainty on the effective temperature and stellar luminosity are dominated by the uncertainties on the magnitudes and distance. In Fig. 4, we show the location of the PMS candidates in the HR diagram, with the relative uncertainties, which in some cases are smaller than the symbol size. We highlight that the majority of the PMS candidates are close to the MS and we could only identify them thanks to the information on their Hα excess. We also plot in Fig. 4 the theoretical isochrones for ages of 2, 4, 8, 16, 32, and 64 Myr for Z = 0.007 (Bressan et al. 2012). The red squares represent the PMS candidates younger than 8 Myr, while the blue dots the older ones.

thumbnail Fig. 4.

HR diagram with the location of the low-mass young (red squares) and old (blue dots) PMS candidates. The theoretical isochrones (Bressan et al. 2012) are calculated for ages of 2, 4, 8, 16, 32, and 64 Myr (lines from right to left, respectively) and Z = 0.007. The small black dots are the targets of the sample of our catalog.

From the HR diagram, it appears that LH 91 is characterized by a more or less continuous star formation, from a few million years to ∼60 Myr, with a smaller number of PMS candidates for ages younger than 8 Myr.

From the effective temperature and the luminosity of the PMS stars, we derive the stellar radius R⋆ of these stars, which we use to estimate the mass accretion rate of the selected PMS objects in Sect. 5. Typical mean errors on R⋆ are around 7% and include both uncertainties on Teff and L⋆.

4.2. Mass and age

We derived the mass and age for each target by comparing the location of each star in the HR diagram (Fig. 5) with theoretical PMS evolutionary tracks. We adopted the PARSEC tracks for metallicity Z = 0.007 (Bressan et al. 2012) from 0.1 M⊙ to 3.0 M⊙. We followed the approach discussed in Romaniello (1998) and refined by De Marchi et al. (2011, 2013). According to these authors, we define a grid in luminosity and temperature consisting of evenly spaced cells with sizes comparable to the typical observational errors. Given an evolutionary track of a star of a certain mass, we identify the cells crossed by the star during its evolution. For each cell, we extrapolate information associated with the evolutionary track, namely mass and age. The information is then be associated with the observed star belonging to a particular cell (for further details, see De Marchi et al. 2017). Figure 5 shows the masses of the PMS candidates spanning from 0.2 M⊙ for the cooler objects up to 1.0 M⊙ for the hottest ones. The median value of the sample is about 0.8 M⊙. In the figure, we divided the PMS stars in two subsamples: the younger PMS candidates with an age of less than 8 Myr (red squares), and the PMS candidates older than 8 Myr (blue dots). The sizes of the dots and squares are proportional to the mass accretion rate, which we determine and discuss in Sect. 5. Here, we simply want to investigate whether and how the rate of mass accretion is correlated with evolutionary phase and stellar mass. As one can see in Fig. 5, the targets with the highest mass accretion rates (the largest symbols) are the youngest PMS stars, while the mass accretion rate decreases at older ages. Furthermore, the stars with higher mass have higher mass accretion values (Ṁacc) at all ages.

thumbnail Fig. 5.

HR diagram of our PMS candidates. Red squares and blue dots represent young and old PMS stars, respectively. The size of the symbols is proportionate to the rate of mass accretion, as in the legend. We adopted the PARSEC tracks for metallicity Z = 0.007 (Bressan et al. 2012) from 0.2 M⊙ to 1.2 M⊙ (dashed lines).

In Fig. 6 we show the histograms of the mass (upper panel) and age (lower panel) distribution of the PMS candidates with the bin sizes compatible with the uncertainties on mass and age, respectively. The black line corresponds to the whole sample, the dashed red line corresponds to the young PMS candidates, while the dotted blue line represents the old PMS candidates. The older PMS stars include preferentially higher mass stars, with the mass distribution presenting a peak at ∼0.7 M⊙. The young PMS objects show a continuous distribution in mass, with no evident peak, but this is probably mostly due to the paucity of the subsample.

thumbnail Fig. 6.

Histograms of the stellar mass (upper panel, bin of 0.05) and age (lower panel, bin of 0.2) for the 75 low-mass PMS candidates in logarithmic scale. The red dashed and blue dotted lines represent the distribution of the young and older populations, respectively.

The age distribution could suggest a separation between older and younger PMS stars, with a gap in the range between 5 and 10 Myr. The younger population shows a continuous distribution in age up 5 Myr. The older population constitutes about 90% of the objects, with ages between 10 and ∼60 Myr and a peak at ∼50 Myr.

5. Accretion properties

In the following subsections we describe how we determined the accretion properties of our sample of PMS candidates and we present our study of their relation with the physical properties of the stars.

5.1. Accretion luminosity

The luminosity of the Hα line generated along the funnel flows of circumstellar gas during the magnetospheric accretion process can be used as a tracer to estimate the accretion luminosity. To determine the accretion luminosity of our sample of PMS candidates, we adopted the relationship obtained by De Marchi et al. (2010), who analyzed the data of a group of T Tauri stars in Taurus-Auriga compiled by Dahm (2008):

(5)

The median of the accretion luminosity of our 75 PMS stars is 0.12 L⊙. The uncertainty on Lacc is dominated by the uncertainty on LHα, which is about 16%, related to the photometric error on the Hα magnitude. There is also a systematic error to take into account due to the uncertainties on the ratio Lacc/LHα (Dahm 2008; De Marchi et al. 2011), but as the relation is the same for all stars, this uncertainty does not interfere with the comparison between the targets.

In Fig. 7 we show the accretion luminosity vs. L⋆ of the PMS candidates, the blue dots and red squares representing the old and young ones, respectively. In each star formation region, Lacc increases with stellar luminosity, but the range and dispersion of the data are quite different. For comparison, we show also the data of LH 95, with gray filled dots, SN 1987A with green empty diamonds, and 30 Dor with black empty dots. In LH 91 and LH 95, the dispersion in Lacc seems to decrease with the increase of L⋆. The accretion luminosity spans a range between 0.1 and 1 L⋆, with the peak of the distribution at about 0.3 L⋆ for LH 91. In Sect. 3.3, we shown that the median LHα in LH 91 is lower than that found in LH 95, and therefore it is not surprising that the values of the accretion luminosity in LH 91 are also slightly lower than those of the PMS objects in LH 95 (∼0.17 L⊙). This result could be due to two main factors: in LH 95 the mass range of the sample is larger (0.2–1.8 M⊙), and at the same mass the stars are younger. The samples of 30 Dor and SN 1987A are richer than LH 91 and LH 95, and the range of the accretion luminosity is larger, from 0.1 L⋆ to values higher than 1.0 L⋆. For a comparison, we focus on the range in stellar luminosity in common between the regions, namely −0.65 L⋆ and 0.0 L⋆. We evaluated the median accretion luminosity only for the regions 30 Dor, LH 95, and LH 91 finding values of ∼0.17 L⊙, 0.22 L⊙, and 0.13 L⊙, respectively. Unfortunately, the range in star luminosity of SN 1987A does not cross-match with those of LH 91 and LH 95, and therefore we cannot make a direct comparison.

thumbnail Fig. 7.

Accretion luminosity as a function of stellar luminosity. Blue and red dots represent the older (age greater than 8 Myr) and younger (age smaller than 8 Myr) PMS candidates of LH 91, respectively. The gray filled dots, green diamonds, and black empty dots are the PMS of LH 95 by Biazzo et al. (2019), SN 1987A by De Marchi et al. (2010), and 30 Dor by De Marchi et al. (2017), respectively. The dashed lines show the linear Lacc − L⋆ relationship for different values of the coefficient, as indicated.

In Fig. 8 we show the accretion luminosity versus the effective temperature in logarithmic scale of the old (blue dots) and young (red squares) PMS stars, together with the sample of LH 95 (Biazzo et al. 2019; gray filled dots). This plot is very similar to the HR diagram (Fig. 5). While a separation between the old and young candidates in Teff is evident in the LH 95 sample (see Fig. 9 in Biazzo et al. 2019), in LH 91 there is a continuous distribution in Teff, the PMS stars with the highest accretion luminosity being close to the old subgroup.

thumbnail Fig. 8.

Accretion luminosity vs. effective temperature. The blue dots and red squares are as in Fig. 7, the gray filled dots represent the PMS stars of LH 95 by Biazzo et al. (2019).

5.2. Mass accretion rate versus stellar age

Finally, we derived the mass accretion rate Ṁacc of our PMS candidates from the free-fall equation (Koenigl 1991; Calvet & Gullbring 1998):

(6)

where G is the gravitational constant, M⋆ and R⋆ are the mass and radius of the PMS candidates, and Rin is the inner radius of the accretion disk. Rin depends on how exactly the accretion disk is coupled with the magnetic field of the star, and so its value is quite uncertain. We adopt Rin = 5R⋆, following Gullbring et al. (1998). The median value of the mass accretion rate of our sample is ∼4.8 × 10−9 M⊙ yr−1, with higher values for the younger population (∼1.2 × 10−8 M⊙ yr−1), and lower values for the older candidates (∼4.7 × 10−9 M⊙ yr−1). The values we find are slightly lower than those found by Biazzo et al. (2019) for LH 95, as shown in Fig. 9, where the median rate is about 7.5 × 10−9 M⊙ yr−1.

thumbnail Fig. 9.

Observed mass accretion rate versus age in LH 91. Yellow filled squares and empty black triangles represent the targets with mass greater and smaller than the median mass of the PMS candidates sample, respectively. The gray-filled dots are the PMS candidates in LH 95 (Biazzo et al. 2019). The error bars on the age and mass accretion rate are reported. When the uncertainties are smaller than the symbol size, they are not visible. The arrows represent lower limits. The dashed yellow and solid black lines represent the regression fit of the two subsamples of LH 91.

The mass accretion rate in LH 91 is also lower than the median value measured in the field of SN 1987A (2.6 × 10−8 M⊙ yr−1, as found by Romaniello et al. 2004, and 2.9 × 10−8 M⊙ yr−1 as measured by De Marchi et al. 2010) and in 30 Dor by (∼8 × 10−8 M⊙ yr−1; De Marchi et al. 2017). The uncertainty on Ṁacc is dominated by the uncertainty on LHα, which is of about 16%, but we have to consider also the contribution of R⋆ (7%, including a 5% systematic uncertainty on the distance modulus), stellar mass (∼7%), the intrinsic uncertainties due to the evolutionary tracks (2%−6%, for more details see the Appendix A of Biazzo et al. 2019), and knowledge of the relation Lacc–LHα, which in this case is not very accurate (a factor of ∼2; De Marchi et al. 2010). Finally, the contribution of other sources of systematic error – such as physical processes different from accretion (e.g., chromospheric activity or ionization of nearby massive stars) or nebular continuum – that could affect the determination of Ṁacc are considered to be negligible (De Marchi et al. 2010). In summary, the combined statistical uncertainty on Ṁacc is of about 20%.

A snapshot of the mass accretion rate as a function of the age is shown in Fig. 9. We divided the sample in two subsamples with the stellar mass larger (yellow filled squares) and smaller (empty black triangles) than the median stellar mass (∼0.8 M⊙). The gray filled dots are the PMS candidate in LH 95 (Biazzo et al. 2019). As expected, the accretion appears to decrease with time, in line with the predicted evolution of viscous disks (Hartmann et al. 2016), but there is a large spread in mass accretion rate at a given age. We performed a linear fit to the two subsamples, and find a similar slope: −0.31 ± 0.07 for the high masses and −0.39 ± 0.04 for the low masses. These values are in agreement within the error with those evaluated in other MCs regions (De Marchi et al. 2011, 2013, 2017; Biazzo et al. 2019). This plot also shows that the mean mass accretion rate of the PMS stars in LH 91 is slightly lower than in LH 95, because our sample is composed mainly of older stars (more so than 30 Myr), close to the MS when the accretion process is less powerful.

5.3. Mass accretion rate versus stellar mass

Figure 10 shows the mass accretion rate as a function of the stellar mass of our PMS candidates. Younger PMS candidates are represented by red squares, while the older PMS candidates are marked with blue dots. The gray filled dots represent the PMS candidates in LH 95 (Biazzo et al. 2019). From Figs. 9 and 10 it is evident that the mass accretion rate is typically higher for the younger and more massive stars. Only two low-mass stars (with masses of about 0.3 M⊙ and 0.4 M⊙) have a high mass accretion rate (2.2 × 10−8 M⊙ yr−1 and 1.4 × 10−8 M⊙ yr−1 respectively). Figure 10 reveals a large spread in Ṁacc values for a given stellar mass. This is hardly surprising considering the large spread of ages (see also Rigliaco et al. 2011). Moreover, the older sample of PMS candidates in LH 91 reaches lower values of mass accretion rate when compared to the stars at similar masses in LH 95.

thumbnail Fig. 10.

Distribution of Ṁacc as a function of M⋆. The symbols are as in Fig. 7.

Again, it is interesting to note how the stars of any given mass appear to be younger in LH 95 than in LH 91. This could simply be the result of different evolutionary stages between stars in LH 91 and LH 95, with the former sample being more evolved than the latter. This could in turn justify the smaller number of accretors and the lower values of the mass accretion rate in LH 91 compared to LH 95. This difference might in turn be caused by other physical differences in star formation environment, for example the gas density of the regions. It would seem natural that an environment with lower gas density would result in less massive circumstellar disks, and therefore a more modest mass accretion rate. To verify this effect, we compare the median mass accretion rate in LH 91 with that of the three star-forming regions at similar metallicity in the LMC, namely LH 95, SN 1987A, and 30 Dor, for which a study of accretion properties was performed. Considering targets with the same mass range (0.4−1.0 M⊙) and younger than 8 Myr, we obtained a mean Ṁacc value of 1.1 × 10−8 M⊙ yr−1 for LH 91, 4.4 × 10−8 M⊙ yr−1 for LH 95, 3.7 × 10−7 M⊙ yr−1 for SN 1987A, and 5.9 × 10−8 M⊙ yr−1 for 30 Dor. We also estimated the mean dust density of the aforementioned four regions taking into account the mass surface density map by Utomo et al. (2019).

Considering regions with a radius of 1.5 arcmin, we found values of 0.11 ± 0.01 M⊙ pc−2 for LH 91, 0.16 ± 0.01 M⊙ pc−2 for LH 95, and similar value for SN 1987A, and 0.65 ± 0.09 M⊙ pc−2 for 30 Dor. Even though this kind of analysis is only qualitative, we find some tentative indication that regions with higher dust densities also have higher mass accretion rates (and possibly higher gas density), with the exception of SN 1987. A much more detailed analysis would be necessary to address this issue in more detail, which goes beyond the scope of this work.

5.4. Spatial distribution of the PMS candidates

The left panel of Fig. 11 shows a color-composite image from WFC3 observations in the F555W (red), F814W (blue), and F656N (green) filters of LH 91. The right panel in the same figure shows the spatial distribution of the PMS stars in our sample projected onto the sky. As in Fig. 5, the sizes of the dots and squares are proportionate to the ranges of the mass accretion rate. Different colors represent different ages: older than 8 Myr (blue dots) and younger than 8 Myr (red squares). The crosses are massive stars selected from the 2MASS catalog (Cutri et al. 2003) with J − H <  0.8 and J <  15 mag (see also Biazzo et al. 2019), while the triangles are Be stars found by Gouliermis et al. (2002) (see their Table 1). To align the orientation of the two figures, the left image is rotated 30 deg. With the aim of better understanding the correspondence between the fields, we indicate some bright stars with the letters A to D. Regions with a lack of stars in the right figure correspond to those rich in gas in the left figure, shown in green. The dust associated to the gas could be obscuring the stars behind it. The PMS objects appear to be distributed more or less uniformly over the region, and are not clustered around the massive stars, unlike the younger population of LH 95 (Biazzo et al. 2019). This result is in agreement with the conclusions of Gouliermis et al. (2002), who found only a weak match between the HII region of LH 91 and the two Be stars located to the southwest side of the region. Therefore, it appears that in LH 91 there is no obvious region of higher star-formation intensity, at least currently.

thumbnail Fig. 11.

Left panel: color-composite image of LH 91 from WFC3 observations in the F555W, F814W, and F656N filters. The image is rotated 30 deg to align the two figures. Right panel: distribution of the PMS stars. Symbols are the same as in Fig. 5. The crosses represent the massive stars from the 2MASS catalog, and the triangles are Be stars from Table 1 of Gouliermis et al. (2002). North is up and east to the left.

6. Conclusions

We presented a multiwavelenght analysis of the stellar populations in LH 91, a star-forming region in the LMC, observed with the WFC3 on board the HST. We applied a photometric detection method to identify PMS candidates still actively accreting matter from their circumstellar disks. The method combines HST broad-band F555W and F814W photometry with narrow-band F656N imaging in order to identify stars with Hα excess emission and to subsequently measure their accretion luminosity Lacc and equivalent width EWHα, and to derive their mass accretion rate Ṁacc. The main results of our analysis can be summarized as follows:

  1. From the photometric catalog of 9423 well-detected stars, we identified about 180 low-mass PMS candidates on the basis of their excess Hα emission, that is, with their (m555 − m656) color exceeding that of the reference template at the same (m555 − m814) color by more than three times the combined uncertainties on their (m555 − m656) values.

  2. We measured the EWHα of the PMS stars, finding values in the range of ∼3 Å–17 Å, with a median of 9 Å. We selected stars with EWHα ≥ 10 Å, which are typical values of actively accreting PMS stars. A total of 75 objects satisfy this condition.

  3. We estimated the stellar effective temperature and luminosity thanks to the Bessell et al. (1998) relations for 3500 ≤ Teff ≤ 40 000 K, and the Pecaut & Mamajek (2013) calibrations for Teff <  3500 K.

  4. We obtained the mass and age of the PMS candidates by comparing the location of each star in the HR diagram with theoretical PMS evolutionary tracks (Bressan et al. 2012). The range of the stellar masses in our sample is between ∼0.2 M⊙ and ∼1.0 M⊙ with a median of ∼0.8 M⊙. The age of the stars is distributed between a few million years and as much as ∼60 Myr with an apparent gap between 5 Myr and 10 Myr. For this reason we divided our sample in two populations, which we call younger (t ≤ 8 Myr with median age ∼3.5 Myr) and older PMS candidates (t >  8 Myr with median age ∼35 Myr).

  5. We measured the Hα luminosity of the PMS candidates and consequently their accretion luminosity. We find a median value of ∼0.12 L⊙. The accretion luminosity increases with L⋆, while the dispersion in Lacc seems to decreases with L⋆. We also find that the accretion luminosity spans the range 0.1–1 L⋆, with a peak in the distribution at about 0.3 L⋆.

  6. Through the accretion luminosity and other physical parameters, we determined the mass accretion rate of PMS stars, finding a median value of ∼4.8 × 10−9 M⊙ yr−1, with higher values for the younger population (∼1.2 × 10−8 M⊙ yr−1), and lower values for the older candidates (∼4.7 × 10−9 M⊙ yr−1).

  7. We studied the relation between the mass accretion rate and both age and stellar mass. As expected, the mass accretion rate appears to decrease with time and to increase with stellar mass.

  8. We compared our results with other star formation regions in the Large Magellanic Cloud, in particular with LH 95, which is the closest region to LH 91 for which accretion properties of PMS candidates have been derived. LH 91 is a star-forming region that is less rich in PMS stars than LH 95, with lower stellar masses (0.2–1.0 M⊙ vs. 0.2–1.8 M⊙) but similar range in age (few Myr up to 60 Myr). The accretion luminosity and the mass accretion rate of PMS candidates in LH 91 are both slightly lower than in LH 95; in particular the median values are 0.12 L⊙ versus 0.17 L⊙, and 7.5 × 10−9 M⊙ yr−1 vs. 4.8 × 10−9 M⊙ yr−1 , respectively.

  9. We explored the possibility that the density of the environment (which we probe using dust emission) could affect the mass accretion rate. We compared the median mass accretion rate of star-forming regions with similar metallicity but different dust density, namely LH 91, LH 95, SN 1987A, and 30 Dor. We considered targets in the same mass range (0.4–1.1 M⊙) and younger than 8 Myr. From a qualitative analysis, we find that the mass accretion rate increases with dust density of the environment in which the stars are formed.

  10. Finally, we find the spatial distribution of the PMS stars to be rather uniform, without any evidence of clumps around more massive stars.

The advent of the James Webb Space Telescope will allow us to put strong constraints on accretion phenomena of members in star-forming regions with different stellar properties (such as metallicity, age, and distance). In particular, the spectroscopic observations would give us information on the density and ionization state of the material undergoing accretion as well as on its kinematics, thereby providing a clearer picture of the accretion process itself in different environmental conditions.

Acknowledgments

We are very thankful to the anonymous referee for precious comments and suggestions that have helped us to improve this paper. RC is grateful to ESA for the support during the data analysis useful for the preparation of this paper. This work was 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 Spacte Telescope European Coordinating Facility (ST-ECF/ESA) and the Canadian Astronomy Data Centre (CADC/NRC/CSA). Some of the data presented in this paper were obtained from the Mikulski Archive for Space Telescopes (MAST). STScI is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS5-26555. This research made use of the SIMBAD database, operated at the CDS (Strasbourg, France) and data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. This research has made also use of the SVO Filter Profile Service supported from the Spanish MINECO through grant AyA2014-55216.

References

  1. Beccari, G., Spezzi, L., De Marchi, G., et al. 2010, ApJ, 720, 1108 [NASA ADS] [CrossRef] [Google Scholar]
  2. Beccari, G., De Marchi, G., Panagia, N., et al. 2015, A&A, 574, A44 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  3. Bessell, M. S., Castelli, F., & Plez, B. 1998, A&A, 333, 231 [NASA ADS] [Google Scholar]
  4. Biazzo, K., Beccari, G., De Marchi, G., & Panagia, N. 2019, ApJ, 875, 51 [Google Scholar]
  5. Book, L. G., Chu, Y.-H., Gruendl, R. A., & Fukui, Y. 2009, AJ, 137, 3599 [NASA ADS] [CrossRef] [Google Scholar]
  6. Braun, J. M., Bomans, D. J., Will, J.-M., & de Boer, K. S. 1997, A&A, 328, 167 [NASA ADS] [Google Scholar]
  7. Braun, J. M., de Boer, K. S., & Altmann, M. 2000, ArXiv-eprints [arXiv:astro-ph/0006060] [Google Scholar]
  8. Bressan, A., Marigo, P., Girardi, L., et al. 2012, MNRAS, 427, 127 [Google Scholar]
  9. Calamida, A., Mack, J., Medina, J., et al. 2021, New time-dependent WFC3 UVIS inverse sensitivities, Space Telescope WFC Instrument Science Report [Google Scholar]
  10. Calvet, N., & Gullbring, E. 1998, ApJ, 509, 802 [Google Scholar]
  11. Colucci, J. E., Bernstein, R. A., Cameron, S. A., & McWilliam, A. 2012, ApJ, 746, 29 [NASA ADS] [CrossRef] [Google Scholar]
  12. Cutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003, VizieR Online Data Catalog: II/246 [Google Scholar]
  13. Da Rio, N., Gouliermis, D. A., & Henning, T. 2009, ApJ, 696, 528 [NASA ADS] [CrossRef] [Google Scholar]
  14. Dahm, S. E. 2008, AJ, 136, 521 [NASA ADS] [CrossRef] [Google Scholar]
  15. De Marchi, G., & Panagia, N. 2014, MNRAS, 445, 93 [NASA ADS] [CrossRef] [Google Scholar]
  16. De Marchi, G., Panagia, N., & Romaniello, M. 2010, ApJ, 715, 1 [NASA ADS] [CrossRef] [Google Scholar]
  17. De Marchi, G., Panagia, N., Romaniello, M., et al. 2011, ApJ, 740, 11 [NASA ADS] [CrossRef] [Google Scholar]
  18. De Marchi, G., Beccari, G., & Panagia, N. 2013, ApJ, 775, 68 [NASA ADS] [CrossRef] [Google Scholar]
  19. De Marchi, G., Panagia, N., & Beccari, G. 2017, ApJ, 846, 110 [NASA ADS] [CrossRef] [Google Scholar]
  20. Fitzpatrick, E. L., & Savage, B. D. 1984, ApJ, 279, 578 [NASA ADS] [CrossRef] [Google Scholar]
  21. Gouliermis, D., Keller, S. C., de Boer, K. S., Kontizas, M., & Kontizas, E. 2002, A&A, 381, 862 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  22. Gullbring, E., Hartmann, L., Briceño, C., & Calvet, N. 1998, ApJ, 492, 323 [Google Scholar]
  23. Hartmann, L., Herczeg, G., & Calvet, N. 2016, ARA&A, 54, 135 [Google Scholar]
  24. Koenigl, A. 1991, ApJ, 370, L39 [Google Scholar]
  25. Kontizas, M., Kontizas, E., Dapergolas, A., Argyropoulos, S., & Bellas-Velidis, Y. 1994, A&AS, 107, 77 [NASA ADS] [Google Scholar]
  26. Lucke, P. B. 1974, ApJS, 28, 73 [NASA ADS] [CrossRef] [Google Scholar]
  27. Panagia, N. 1999, in New Views of the Magellanic Clouds, eds. Y. H. Chu, N. Suntzeff, J. Hesser, & D. Bohlender, 190, 549 [NASA ADS] [Google Scholar]
  28. Panagia, N., Gilmozzi, R., Macchetto, F., Adorf, H. M., & Kirshner, R. P. 1991, ApJ, 380, L23 [NASA ADS] [CrossRef] [Google Scholar]
  29. Pecaut, M. J., & Mamajek, E. E. 2013, ApJS, 208, 9 [Google Scholar]
  30. Rigliaco, E., Natta, A., Randich, S., Testi, L., & Biazzo, K. 2011, A&A, 525, A47 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  31. Romaniello, M. 1998, PhD Thesis, European Southern Observatory [Google Scholar]
  32. Romaniello, M., Robberto, M., & Panagia, N. 2004, ApJ, 608, 220 [NASA ADS] [CrossRef] [Google Scholar]
  33. Ryon, J. E. 2018, Advanced Camera for Surveys Instrument Handbook for Cycle 26 v. 17.0 [Google Scholar]
  34. Schaerer, D., Meynet, G., Maeder, A., & Schaller, G. 1993, A&AS, 98, 523 [NASA ADS] [Google Scholar]
  35. Spezzi, L., De Marchi, G., Panagia, N., Sicilia-Aguilar, A., & Ercolano, B. 2012, MNRAS, 421, 78 [NASA ADS] [Google Scholar]
  36. Stetson, P. B. 1987, PASP, 99, 191 [Google Scholar]
  37. Utomo, D., Chiang, I.-D., Leroy, A. K., Sandstrom, K. M., & Chastenet, J. 2019, ApJ, 874, 141 [NASA ADS] [CrossRef] [Google Scholar]
  38. White, R. J., & Basri, G. 2003, ApJ, 582, 1109 [Google Scholar]
  39. Zeidler, P., Grebel, E. K., Nota, A., et al. 2016, AJ, 152, 84 [NASA ADS] [CrossRef] [Google Scholar]

All Tables

Table 1.

Logbook of the observations.

In the text

All Figures

thumbnail Fig. 1.

Color–color diagram of the selected stars in the field of LH 91. All magnitudes are already corrected for the extinction contribution of our Galaxy, mag and E(m555 − m814)MW = 0.1. The arrow shows the reddening vector of E(m555 − m814) = 0.25 and E(m555 − m656) = 0.13 for the adopted LH 91 extinction law. The dashed line represents the median photospheric (m555 − m814) color for the 254 stars with δ555, δ814, and δ656 <  0.05 (gray filled dots). The dotted line shows the model atmospheres of Bessell et al. (1998) computed for the three WFC3/UVES filters. The PMS star candidates with Hα emission excess at the 3σ level are represented with large red dots. The large green dots are the brightest PMS star candidates, with m555 <  20 mag. Error bars are also shown.
In the text
thumbnail Fig. 2.

Color–magnitude diagram of the field of LH 91. All magnitudes are already corrected for the extinction contribution of our Galaxy and LH 91. The small black dots are the targets of the whole sample (9423 stars). Small gray-filled dots are the stars with photometric uncertainties < 0.05 mag in each band. The large red dots represent the PMS star candidates with Hα excess emission at the 3σ level. Solid green and dashed light-blue lines show the theoretical isochrones from Bressan et al. (2012) for ages 8 Myr and 1.5 Gyr, respectively, metallicity Z = 0.007, and a distance modulus (mV − MV)0 = 18.55.
In the text
thumbnail Fig. 3.

Hα equivalent width of the selected low-mass PMS candidates, as a function of the de-reddened (m555 − m814) color. The red squares represent the values of the PMS stars younger than 8 Myr, the blue dots are the oldest PMS stars.
In the text
thumbnail Fig. 4.

HR diagram with the location of the low-mass young (red squares) and old (blue dots) PMS candidates. The theoretical isochrones (Bressan et al. 2012) are calculated for ages of 2, 4, 8, 16, 32, and 64 Myr (lines from right to left, respectively) and Z = 0.007. The small black dots are the targets of the sample of our catalog.
In the text
thumbnail Fig. 5.

HR diagram of our PMS candidates. Red squares and blue dots represent young and old PMS stars, respectively. The size of the symbols is proportionate to the rate of mass accretion, as in the legend. We adopted the PARSEC tracks for metallicity Z = 0.007 (Bressan et al. 2012) from 0.2 M⊙ to 1.2 M⊙ (dashed lines).
In the text
thumbnail Fig. 6.

Histograms of the stellar mass (upper panel, bin of 0.05) and age (lower panel, bin of 0.2) for the 75 low-mass PMS candidates in logarithmic scale. The red dashed and blue dotted lines represent the distribution of the young and older populations, respectively.
In the text
thumbnail Fig. 7.

Accretion luminosity as a function of stellar luminosity. Blue and red dots represent the older (age greater than 8 Myr) and younger (age smaller than 8 Myr) PMS candidates of LH 91, respectively. The gray filled dots, green diamonds, and black empty dots are the PMS of LH 95 by Biazzo et al. (2019), SN 1987A by De Marchi et al. (2010), and 30 Dor by De Marchi et al. (2017), respectively. The dashed lines show the linear Lacc − L⋆ relationship for different values of the coefficient, as indicated.
In the text
thumbnail Fig. 8.

Accretion luminosity vs. effective temperature. The blue dots and red squares are as in Fig. 7, the gray filled dots represent the PMS stars of LH 95 by Biazzo et al. (2019).
In the text
thumbnail Fig. 9.

Observed mass accretion rate versus age in LH 91. Yellow filled squares and empty black triangles represent the targets with mass greater and smaller than the median mass of the PMS candidates sample, respectively. The gray-filled dots are the PMS candidates in LH 95 (Biazzo et al. 2019). The error bars on the age and mass accretion rate are reported. When the uncertainties are smaller than the symbol size, they are not visible. The arrows represent lower limits. The dashed yellow and solid black lines represent the regression fit of the two subsamples of LH 91.
In the text
thumbnail Fig. 10.

Distribution of Ṁacc as a function of M⋆. The symbols are as in Fig. 7.
In the text
thumbnail Fig. 11.

Left panel: color-composite image of LH 91 from WFC3 observations in the F555W, F814W, and F656N filters. The image is rotated 30 deg to align the two figures. Right panel: distribution of the PMS stars. Symbols are the same as in Fig. 5. The crosses represent the massive stars from the 2MASS catalog, and the triangles are Be stars from Table 1 of Gouliermis et al. (2002). North is up and east to the left.
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.