AN ALL-SKY SAMPLE OF INTERMEDIATE-MASS STAR-FORMING REGIONS

Michael J. Lundquist; Henry A. Kobulnicky; Michael J. Alexander; Charles R. Kerton; Kim Arvidsson

doi:10.1088/0004-637X/784/2/111

1. INTRODUCTION

It is well established that most stars form in groups or clusters (Larson 1982). Weidner & Kroupa (2006) suggested that an underlying physical relationship exists between the most massive star in an embedded cluster and the total stellar mass of the cluster. Low-mass stars tend to form in small, loose aggregates of 10–100 stars pc⁻³ (Testi et al. 1999), while high-mass stars are associated with clusters that contain up to 1000 stars pc⁻³ (Hillenbrand et al. 1995). Evidence suggests that there may be a turn-on mass for stellar clustering at the scale of intermediate-mass stars of 8–10 M_☉ and that this onset may be sudden (Testi et al. 1997) or gradual (Testi et al. 1999). Hence, identifying and studying clusters having intermediate-mass stars as their most massive members may yield clues about the onset of cluster formation.

Although it is reasonably well understood how isolated, low-mass stars form (McKee & Ostriker 2007), it may not be as simple as scaling this standard theory to explain high-mass star formation (Zinnecker & Yorke 2007). Scaling up low-mass star formation to higher masses (>10 M_☉) results in stars that begin to burn hydrogen while still accreting. The resulting radiation pressure should expel the accreting material, preventing more massive stars from forming, yet stars much more massive do exist (Larson & Starrfield 1971; Zinnecker & Yorke 2007; Krumholz et al. 2009). At least two competing theories have been proposed to explain this paradox. In the competitive accretion model, stars within a common gravitational potential accrete from a distributed gas reservoir. The gravitational potential funnels gas down toward the cluster center, so the stars located closer to the center of the potential accrete at higher rates than isolated stars (Bonnell et al. 1997, 2001; Bonnell & Bate 2006). In the turbulent core accretion model, molecular clumps fragment into gaseous cores that subsequently collapse to make individual stars. As gas accretes onto the protostar, gravitational energy is released that heats up the surrounding gas. This increase in gas temperature raises the Jeans mass, which suppresses fragmentation and results in fewer, but more massive, stars. This monolithic collapse would dictate a correlation of the clump mass with the stellar mass, as the mass of the core determines the mass reservoir available to form the star (McKee & Tan 2003; Krumholz et al. 2005; Krumholz & Bonnell 2007). These divergent theories have led to increased interest in studying star-forming regions that occupy the transition from the low- to high-mass regimes.

Arvidsson et al. (2010) defined and introduced a small sample of 70 intermediate-mass star-forming regions (IM SFRs)—regions producing stars up to but not exceeding ∼8 M_☉. IM SFRs populate the continuum linking the low- and high-mass star-forming regimes where the onset of cluster formation may occur and the mechanisms that differentiate high-mass from low-mass star formation may become significant. IM SFRs are sites of star formation distinct from regions of high-mass star formation. They have typical luminosities of ∼10⁴ L_☉ and typical diameters of ∼1 pc, as judged from the sizes of the surrounding photodissociation regions (PDRs). They are typically associated with molecular clumps of mass ∼10³ M_☉. The molecular material associated with IM SFRs typically has an order of magnitude lower peak-mass column density and clump mass compared to ultracompact H ii regions that are examples of high-mass star formation (Arvidsson et al. 2010).

We have begun a program to study the stellar properties and the local environments surrounding IM SFRs and identify a larger, more complete, sample of IM SFRs. Our primary goal in this article is to identify the stellar content and confirm the intermediate-mass nature of these candidate IM SFRs. In Section 2, we define an all-sky sample of IM SFRs and categorize them based on their mid-IR morphological appearance. In Section 3, we use the Two Micron All-Sky Survey (2MASS) Point Source Catalog (PSC; Cutri et al. 2003), new optical spectroscopy, and young stellar object (YSO) spectral energy distribution (SED) fitting to investigate the stellar content of a sample of four IM SFRs. In Section 4, we present the results of our stellar content analysis, and we summarize key results in Section 5.

2. SAMPLE SELECTION

Kerton (2002) used Infrared Astronomical Satellite (IRAS) mid-infrared (IR) fluxes and H i 21 cm emission maps to detect PDRs associated with embedded intermediate-mass stars. These were used to define candidate IM SFRs using a set of color selection criteria that specifies cool dust and large polycyclic aromatic hydrocarbon (PAH) contribution: $-0.18 < \rm {log}({F}_{\nu }(25)/{F}_{\nu }(12)) < 0.4, 0.96 < \rm {log}({F}_{\nu }(60)/{F}_{\nu }(25)) < 1.56,$ and 0.29 < log(F_ν(100)/F_ν(60)) < 0.69. For all candidate IM SFRs, the IRAS flux density quality flag was required to be either high (F_qual = 3) or moderate (F_qual = 2).

Arvidsson et al. (2010) adopted the Kerton (2002) criteria to define a sample totaling 984 candidate IM SFRs and used Spitzer Galactic Legacy Infrared MidPlane Survey Extraordinaire (GLIMPSE I; Benjamin et al. 2003) and MIPSGAL I (Carey et al. 2005) legacy surveys to morphologically classify 70 inner-Galaxy targets that had complementary ¹³CO spectra from the Boston University Five College Radio Astronomy Observatory (BU-FCRAO) Galactic Ring Survey (GRS; Jackson et al. 2006). They classified the candidate objects into three groups based on their mid-IR morphologies: "blobs/shells," "filamentary," and "star-like." Arvidsson et al. (2010) found that the blobs/shells are star-forming regions characterized by circular or elliptical morphologies in the 8.0 and 24 μm images. The filamentary objects appear to be small knots along larger filamentary structures including regions where filaments intersect. Star-like objects exhibit compact emission from one or more point sources in the mid-IR bands. The Arvidsson et al. (2010) pilot study classified 49 (70.0%) objects as blobs/shells, 11 (15.7%) objects as filamentary, and 10 (14.3%) as star-like.

Arvidsson et al. (2010) measured molecular column densities for 28 blobs/shells, 15 of which had distance determinations. All 15 blob/shells with distance determinations, and thus masses, lie below the 1 g cm⁻² minimum mass column density threshold for high-mass star formation posited by Krumholz & McKee (2008). They found that these regions lack or have weak thermal radio continuum, limiting the most massive star(s) to early B or later, i.e., stars capable of exciting the PAHs but not ionizing hydrogen. Arvidsson et al. (2010) also found typical luminosities of ∼10⁴ L_☉, where a luminosity of 10³–10⁴ corresponds roughly to a single ∼4–10 M_☉ YSO (Robitaille et al. 2006). On the basis of these properties they concluded that these regions host, at most, intermediate-mass stars, though the stellar content was not directly observed.

We used Wide-field Infrared Survey Explorer (WISE; Wright et al. 2010) 12 and 22 μm images to extend the morphological classifications to the entire sample of 984 candidate IM SFRs. We introduce a new category, "galaxies," to the "blobs/shells," "filamentary," and "star-like" scheme of Arvidsson et al. (2010). Figure 1 is a three-color RGB image (WISE 22 μm, WISE 12 μm, WISE 3.4 μm) that shows typical objects for each morphological category. The blobs/shells are characterized by extended emission at 12 and 22 μm, including isolated blobs, shells, and enhancements at the edges of pillars or bright-rimmed clouds. The filamentary objects appear to be long, narrow, low-surface-brightness wisps of emission at 12 and 22 μm. The star-like objects appear as isolated point sources (occasionally multiple) in all WISE bands. The galaxies exhibit extended emission that appears circular or lenticular in all WISE bands, but they also have extended emission in near-infrared 2MASS and optical DSS images, rendering their morphology unmistakable. Of the 984 candidate IM SFRs, 616 (62.6%) are classified as blobs/shells, 128 (13.0%) are classified as filamentary, 39 (4.0%) are classified as star-like, and 201 (20.4%) are classified as galaxies. Table 1 lists the IRAS source name, coordinates, and morphological classification for all 984 objects in the sample. Figure 2 shows the distribution of these regions projected on the sky in Galactic coordinates. Red triangles correspond to blobs/shells, green diamonds to filamentary, blue stars to star-like, and gray squares to galaxies. Figure 2 illustrates that galaxies dominate the sample at high Galactic latitudes while blobs/shells, filamentary structures, and star-like objects are heavily concentrated toward the Galactic Plane. There is a conspicuous overdensity of filamentary objects that constitutes nearly one-third of all filamentary objects, 41 of the 128, around ℓ = 80° corresponding to the Cygnus-X region. This overdensity may be a result of the massive star content (Odenwald & Schwartz 1993) coupled with the relative proximity of the region at ∼1.2–1.4 kpc (Kobulnicky et al. 2013). There is also an overdensity of nine objects classified as blobs/shells around ℓ = 280°, b = −32°at the location of the Large Magellanic Cloud. Another object classified with the blobs/shells (IRAS 00402+4120) is located in M31.

**Figure 1.** Three-color RGB image (*WISE* 22 μm, 12 μm, 3.4 μm) showing a typical example of objects that fulfill the *IRAS* color selection criteria for each morphological type: blobs/shells, filamentary, star-like, and galaxies.
Download figure:
Standard image High-resolution image

**Figure 2.** Aitoff projection in Galactic coordinates showing the spatial distribution of 984 candidate IM SFRs. The candidate IM SFRs are color coded by morphological classification. Red triangles correspond to blobs/shells, green diamonds to filamentary objects, blue stars to star-like objects, and grey squares to galaxies. The blobs/shells, filamentary objects, and star-like objects are concentrated in the Galactic Plane with a small grouping in the Magellanic Clouds.
Download figure:
Standard image High-resolution image

Table 1. IM SFR Source Table

Number	IRAS Name	ℓ	b	Morphological Classification
Number	IRAS Name	(°)	(°)	Morphological Classification
1	00001+6417	117.7266	2.1899	Blob/Shell
2	00002+6647	118.2070	4.6387	Filamentary
3	00025+6708	118.4940	4.9407	Blob/Shell
4	00034+6658	118.5597	4.7714	Filamentary
5	00063+6706	118.8581	4.8497	Filamentary
6	00070+6503	118.5951	2.8161	Blob/Shell
7	00080+6329	118.4449	1.2530	Blob/Shell
8	00127+6058	118.6205	−1.3256	Blob/Shell
9	00257+6601	120.6401	3.5368	Blob/Shell
10	00259+5625	119.8038	−6.0319	Blob/Shell

Only a portion of this table is shown here to demonstrate its form and content. A machine-readable version of the full table is available.

Download table as: Data Typeset image

Taken together, the distribution of the morphological classes on the sky is consistent with the blobs/shells and filaments being star-forming regions (or affiliated with star-forming regions) in the Galactic Plane. The uniform distribution of galaxies is consistent with their classification as extra-Galactic sources. The nature of the star-like objects is not known. Some of these are likely to be very compact star-forming regions, perhaps akin to the Ultra-Compact Embedded Clusters identified by Alexander & Kobulnicky (2012). Some of these may be post- or premain sequence objects with circumstellar disks, such as IRAS 00487+5118 (Number 22 in Table 1) a Vega-type excess source (Dent et al. 2005). Higher resolution mid-IR imaging and/or infrared spectroscopy would help discern the nature of these sources.

Figure 3 plots the 984 candidate IM SFRs in a log (f_ν(25)/f_ν(12)) versus log (f_ν(60)/f_ν(25)) IRAS color–color diagram (CCD). The box indicates the Kerton (2002) color selection domain, and the symbols denote candidate regions by morphological type as in Figure 2. The larger symbols with error bars indicate the mean color and error of the mean for each morphological type. The log (f_ν(60)/f_ν(25)) color is a measure of dust temperature, while the log (f_ν(25)/f_ν(12)) color is sensitive to the PAH contribution. While the galaxies are more concentrated toward the higher dust temperatures at the left of the diagram, the mean colors of the blobs/shells, filamentary, and star-like objects are indistinguishable. The density of points drops markedly toward the right edge (low dust temp) and bottom edge (high PAH-to-dust ratio) of the plot, suggesting that a more generous color selection box would not yield significantly more objects. Given the combination of our color criteria and follow-up morphological classification, it is likely that our list of 984 candidate objects contains a reasonably complete sample of intermediate-mass star forming regions, at least to the limits imposed by IRAS sensitivity and beamsize (1 farcm 5 × 4 farcm 7 at 60 μm). The sample list could be expanded by a more generous color selection box toward higher dust temperatures, but this would lead to larger contamination by galaxies and require additional visual examination. Expanding the color selection toward smaller PAH-to-dust content would pick up additional star forming regions, but many of these would be increasingly dominated by massive, ionizing stars (e.g., see van der Veen & Habing 1988; Walker et al. 1989; Zhang et al. 2004).

3. A PILOT SURVEY OF STELLAR CONTENT IN FOUR INTERMEDIATE-MASS STAR-FORMING REGIONS

Although Arvidsson et al. (2010) studied the gaseous content of a sample of inner-Galaxy IM SFRs, the stellar content of these regions remains unexplored. We have selected four outer Galaxy intermediate-mass regions classified as blobs/shells and having four or more optically visible stars for an initial study of the stellar content of IM SFRs. The regions in this initial study are isolated and at high Galactic latitudes where there is expected to be less extinction and little confusion with other Galactic star-forming regions. Table 2 lists the IRAS names, central positions, and IRAS fluxes of these sources. Figures 4 and 5 show three-color RGB images (WISE 22 μm, 12 μm, 3.4 μm) of the four IM SFRs. Cyan crosses mark the locations of optically visible stars selected for spectroscopic study (discussed in Section 3.2 below). White circles indicate 2'–3 farcm 5 diameter regions used to search for possible stellar clusters using photometry from the 2MASS PSC (Cutri et al. 2003). Red circles indicate candidate YSOs. Black diamonds indicate maser sources, and white diamonds indicate sub-mm sources.

**Figure 4.** Three-color RGB image (*WISE* 22 μm, 12 μm, 3.4 μm) of IRAS 00259+5625 (left) and IRAS 00420+5530 (right). Cyan crosses indicate individual stars selected for optical spectroscopy, and white circles indicate the regions used to search for possible stellar clusters. White diamonds mark the positions of known sub-mm sources, and black diamonds mark the positions of known masers.
Download figure:
Standard image High-resolution image

**Figure 5.** As Figure 4 for IRAS 01082+5717 (left) and IRAS 05380+2020 (right).
Download figure:
Standard image High-resolution image

Table 2. Pilot Study IM SFRs

IRAS Name	ℓ	b	F₁₂	F₂₅	F₆₀	F₁₀₀	Other ID
IRAS Name	(°)	(°)	(Jy)	(Jy)	(Jy)	(Jy)	Other ID
00259+5625	119.8038	−6.0319	0.7603	1.116	30.56	108.5	CB 3^a, LBN 594^b
00420+5530	122.0145	−7.0718	7.328	15.13	217.0	469.7	Mol 3^c
01082+5717	125.6137	−5.2090	1.138	1.089	13.56	44.78	...
05380+2020	186.7611	−5.3793	2.294	1.931	34.72	118.6	...

Notes. ^aClemens & Barvainis (1988). ^bLynds (1965). ^cMolinari et al. (1996).

Download table as: ASCII Typeset image

3.1. Infrared Photometry

We used the 2MASS PSC to generate JHK_S color–magnitude diagrams (CMDs) and CCDs of each region. Figures 6 and 7 (top row) plot the raw K_S versus J − K_S diagram for each region enclosed within the circles shown in Figures 4 and 5. As field stars can outnumber the associated main-sequence and pre-main-sequence stars in a region, they must be removed statistically from the CMD to recover the likely cluster stars. We use the cluster cleaning tool of Alexander et al. (2013), which follows the prescription of Maia et al. (2010) to evaluate and remove the field star contamination. The middle row of panels in Figures 6 and 7 show the K_S versus J − K_S diagrams for the field region. The field region is an annulus surrounding the cluster aperture that covers eight times the area of the cluster aperture. The field star estimation tool compares a target region to a field region using a 2MASS JHK_S three-dimensional color–color–magnitude diagrams (CCMD). After investigating different color and magnitude combinations for the CCMDs, we used the J, J − H, and J − K_S magnitudes, as in Maia et al. (2010) for field subtraction. The lower row of panels in Figures 6 and 7 show the field-subtracted K_S versus J − K_S cluster CMDs. The solid line is a Padova (Marigo et al. 2008; Girardi et al. 2010) zero-age main sequence (ZAMS) that has been placed at the estimated distance and reddening (discussed in Section 4 below). The dotted, dashed, and dot–dashed lines indicate Siess et al. (2000) 0.5, 2.5, and 10.0 Myr pre-main-sequence isochrones, respectively. Stars with optical spectra are indicated by pluses and labeled by spectral type.

**Figure 7.** CMDs, as in Figure 6, for IRAS 01082+5717 (left column) and IRAS 05380+2020 (right column).
Download figure:
Standard image High-resolution image

Figures 8 and 9 show the J − H versus H − K_S CCDs cleaned in the same manner as Figure 6. The top row plots the raw J − H versus H − K_S CCDs for each region enclosed within the circles shown in Figures 4 and 5. The middle row of panels plots J − H versus H − K_S for the field region. The lower row of panels in Figures 8 and 9 show the field-subtracted J − H versus H − K_S cluster CCDs. The solid line is a Padova (Marigo et al. 2008; Girardi et al. 2010) ZAMS that has been placed at the estimated reddening (discussed in Section 4 below). Stars having optical spectroscopy are indicated by pluses and labeled by spectral type.

**Figure 8.** CCDs of IRAS 00259+5625 (left column) and IRAS 00420+5530 (right column) for the defined cluster region (top row), comparison field annulus (middle row), and the cleaned cluster region (bottom row). The solid curve in the lower panel shows a Padova ZAMS isochrone with upper and lower X's indicating the locations of B0V and A0V stars, respectively. Pluses indicate stars selected for spectroscopy. The extinction is given for each cleaned cluster, and arrows indicate the direction of the reddening vectors.
Download figure:
Standard image High-resolution image

**Figure 9.** CCDs, as in Figure 8, for IRAS 01082+5717 (left column) and IRAS 05380+2020 (right column).
Download figure:
Standard image High-resolution image

We interpret these CMDs and CCDs in Section 4 to look for compact stellar populations of similar reddening and distance.

3.2. Optical Spectroscopy

We used the CMDs and CCDs described above in conjunction with optical catalogs to select sources within the four candidate IM SFRs having bright K_S-band magnitudes and V ≲ 16. This K_S-band magnitude criterion includes the stars that are likely to be the most luminous and, therefore, the most massive cluster members. We obtained optical spectra for 18 sources using the LongSlit spectrograph on the Wyoming Infrared Observatory 2.3 m telescope in 2011 December and 2012 September. Each object lies within the central 1'–2' of the four candidate IM SFRs. The data for 9 of the 18 objects were obtained with a 600 l mm⁻¹ grating giving a spectral coverage of 4000–7000 Å at a typical reciprocal dispersion of ∼1.5 Å pixel⁻¹ across the chip. The remaining 9 spectroscopic targets were observed with a 2000 l mm⁻¹ grating, giving a spectral coverage of 5400–6800 Å at a typical reciprocal dispersion of ∼0.72 Å pixel⁻¹. The typical signal-to-noise ratio at 5500 Å was greater than 40, with several stars exceeding 200. The data were reduced in the standard manner using dome lamp flat-fielding, bias frames, and CuAr comparison lamp spectra for wavelength calibration to a typical precision of 0.015 Å rms for the 2000 l mm⁻¹ grating and 0.05 Å rms for 600 l mm⁻¹.

3.3. Spectral Classification

Extinction by dust along the line of sight produces low signal-to-noise in the typical temperature diagnostic He i λ4471 Å and Mg ii λ4481 Å lines, making classifications difficult for B- and early A-type stars. In our spectra, the available lines that are useful for spectral classification of these stars are the He ii λ5411 Å, He i λ5876 Å, and Hα λ6563 Å lines. The absence of the He ii λ5411 Å line in all of our spectra limits the spectral type to B0.5 or later. The absence of the He i λ5876 Å line in some of the targets limits the spectral type to later than about A0. The EW of the Hα line is at a maximum at a spectral type of about A1, but it is double-valued, decreasing in strength for stars both earlier and later.

Spectral classification was done by comparison with the spectral atlas of Jacoby et al. (1984) and by comparison to the spectral indices of Hernández et al. (2004) and Kobulnicky et al. (2012). For B-stars, this included comparisons of the equivalent widths of Hα and He i λ5876 Å. For A- and F-stars, this included comparisons of the equivalent width of Hα. For G-stars, this included comparisons of the equivalent width of Hα and the Mg i triplet at λλ5167, 5172, 5183 Å. For K-stars, this included comparisons of the equivalent width of Hα and visual comparisons with the MgH band at λ5000–5200 Å. Spectral classifications of B-, A-, and F-type stars are typically certain to ±1 spectral subtype for a typical signal-to-noise ratio of 40 using this method. He i and Hα lack the sensitivity to differentiate B0.5–B3 stars, so the two stars that appear to fall in this range have been classified as B2. Owing to interstellar extinction and limited wavelength coverage, no stars have the combination of good signal-to-noise in the blue part of the optical range needed for luminosity classification; in the absence of constraining data, we assume that all stars are luminosity class V, but we caution that this may not always be the case.

Table 3 lists the relevant parameters for determining the spectral types and spectrophotometric distances for each star. Column 1 lists the IRAS region name and the corresponding alphabetic identifier in Figures 4 and 5 for each star with optical spectra. Column 2 lists the 2MASS ID. Columns 3–5 list the J-, H-, and K_S-band magnitudes given by the 2MASS PSC. Columns 6–9 list the equivalent widths and equivalent width uncertainties for Hα and He i, the primary spectral lines used to determine spectral types. Column 10 lists the estimated spectral type. Column 11 shows the extinction determined from the 2MASS magnitudes and the spectral type including upper and lower limits assuming an uncertainty of ±1 subtype for B-, A-, and F-type stars and ±3 subtypes for G-type and later stars. Column 12, D_spec, is the spectrophotometric distance determined from the 2MASS magnitudes and the spectral type, including upper and lower limits, where the upper/lower limits result from the random uncertainty in spectral type and the second upper-only limit accounts for systematic uncertainty resulting from the unknown binarity status of each star; if any of these stars are close binaries with equal-luminosity companions (common among massive stars; Kiminki & Kobulnicky 2012), the distances could be underestimated by as much as $\sqrt{2}$ . Column 13, D_lit, lists distances found in the literature for each region. Estimates of the distance to each region were done by using 2MASS H − K_S color excesses to derive A_V and A_K using a Cardelli et al. (1989) extinction curve. A distance modulus was calculated with A_K, the 2MASS K_S-band magnitude, and the intrinsic K_S-band magnitude from a Padova ZAMS isochrone, assuming the observed stars are dwarfs (luminosity class V). The uncertainties on the distance estimates are dominated by the uncertainties on our spectral classifications and the (unknown) binary status.

Table 3. Spectrophotometric Parameters

IRAS Name	2MASS ID	J	H	K_S	Hα 6563			He I 5876		Sp. Type	A_V	D_spec	D_lit
IRAS Name	2MASS ID	J	H	K_S	EW	σ		EW	σ	Sp. Type	(mag)	(kpc)	(kpc)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)
00259+5625													2.5^a, 2.1^b
A	J00283934+5641507	13.089^c	12.896^c	12.737	8.00	0.57		0.30	0.13	B7	2.43 $^{+0.07}_{-0.09}$	2.12 $^{+0.21+0.88}_{-0.21}$
B	J00284837+5641468	13.140	12.960	12.873	14.95	0.18		...	...	A1	1.03 $^{+0.04}_{-0.03}$	1.65 $^{+0.04+0.68}_{-0.08}$
C	J00284282+5642554	12.958	12.632	12.590	4.53	0.20		...	...	F6	0.20 $^{+0.03}_{-0.04}$	0.91 $^{+0.06+0.38}_{-0.06}$
D	J00284536+5641525	12.432	12.088	12.016	3.36	0.15		...	...	G0	0.50 $^{+0.05}_{-0.04}$	0.59 $^{+0.05+0.25}_{-0.05}$
E	J00285071+5641430	11.226	10.512	10.285	1.44	0.21		...	...	G9	2.39 $^{+0.05}_{-0.41}$	0.19 $^{+0.01+0.08}_{-0.05}$
00420+5530													5.0^d, 7.72^e, 2.17^f
A	J00444967+5546225	10.758	10.292	9.981	-8.66	0.02		0.19	0.02	B8e	...	...
B	J00445732+5547231	11.288	10.598	10.183	4.49	0.27		0.73	0.12	B2	6.33 $^{+0.17}_{-0.17}$	1.14 $^{+0.34+0.47}_{-0.25}$
C	J00450384+5546483^g	10.564	10.362	9.979	6.34	0.13		...	...	F2	4.75 $^{+0.01}_{-0.01}$	0.26 $^{+0.01+0.11}_{-0.01}$
D	J00445393+5546244	10.826	12.620	12.489	4.00	0.07		...	...	F9	1.28 $^{+0.05}_{-0.04}$	0.71 $^{+0.06+0.29}_{-0.05}$
01082+5717													...
A	J01112094+5733282	11.645	11.543	11.496	7.14	0.52		0.53	0.01	B4	1.22 $^{+0.09}_{-0.11}$	1.84 $^{+0.26+0.76}_{-0.23}$
B	J01112143+5733047a^h	11.499	11.412	11.279	7.40	0.06		0.15	0.02	B8	...	...
C	J01112143+5733047b^h	...	...	...	8.76	0.14		0.15	0.03	B8	...	...
D	J01112045+5732558	12.453	12.099	11.990	3.80	0.07		...	...	G0	0.99 $^{+0.05}_{-0.04}$	0.57 $^{+0.05+0.24}_{-0.04}$
05380+2020													...
A	J05405928+2021318	10.513	10.255	10.113	5.61	0.13		0.72	0.03	B2	2.74 $^{+0.17}_{-0.17}$	1.34 $^{+0.40+0.55}_{-0.30}$
B	J05405527+2021146	12.018	11.684	11.529	6.63	0.20		0.19	0.06	B8	2.29 $^{+0.09}_{-0.22}$	1.10 $^{+0.12+0.45}_{-0.16}$
C	J05410199+2022208	11.322	11.251	11.196	14.40	0.11		...	...	A1	0.61 $^{+0.04}_{-0.03}$	0.78 $^{+0.02+0.32}_{-0.04}$
D	J05410100+2021221	13.188	12.674	...	2.94	0.24		...	...	G4	...	...
E	J05410336+2021176	8.162	7.589	7.470	1.06	0.07		...	...	K0	0.80 $^{+0.17}_{-0.24}$	0.05 $^{+0.01+0.02}_{-0.01}$

Notes. Spectrophotometric distances and A_V include upper and lower limits resulting from the random uncertainty in spectral type and a second limit upper only accounting for systematic uncertainty resulting from the unknown binarity status of each star. ^aKinematic Distance (Launhardt & Henning 1997). ^bKinematic Distance (Yun & Clemens 1994). ^cUpper Limit ^dKinematic Distance (Molinari et al. 2002). ^eKinematic Distance (Zhang et al. 2005). ^fMaser Trigonometric Parallax (Moellenbrock et al. 2009). ^gForeground F star and background YSO both contributing flux. ^hBlended in 2MASS.

Download table as: ASCII Typeset image

3.4. Young Stellar Objects

The YSOs near each region were identified by generating SEDs using 2MASS JHK_S near-IR and WISE 3.4, 4.5, 12, and 22 μm mid-IR data with source crossband matching done from the WISE PSC, which includes 2MASS. These sources were fit using the SED fitting tool of Robitaille et al. (2007) in a similar manner to Alexander et al. (2013).

All WISE point sources within 10' of the IM SFR were selected for fitting. We required a minimum of three photometric data points per target and removed objects that did not meet this criterion. We first fit sources to Kurucz (1993) stellar models where we allowed the extinction parameter to vary from A_V = 0 to 40 magnitudes. Sources with χ²/n_data ⩽ 2 were deemed to be well fit by a reddened stellar photosphere and removed. To remove sources with bad fits due to high mid-IR sky backgrounds, the remaining sources were visually inspected at WISE bands W3 and W4 and removed if they did not appear to be point sources. The SED fitting tool was then used to fit the remaining sources to a grid of 200,000 YSO model SEDs (Robitaille et al. 2006). Table 4 lists the input parameters for this fit. Column 1 indicates which IRAS target field contains the YSO. Column 2 lists the assigned YSO ID. Column 3 gives the WISE ID. Columns 4–6 indicate the 2MASS PSC magnitudes. Columns 7–10 list the WISE PSC magnitudes. We chose model aperture sizes of 3'' for 2MASS, 6'' for WISE bands W1, W2, and W3, and 12'' for WISE band W4 for all regions, while the extinction parameter varied from A_V = 0 to 35 magnitudes for all regions. We allowed the distance parameter to vary from 1.5–2.5 kpc for IRAS 00259+5625, 2.0–2.5 kpc for IRAS 00420+5530 (best estimate maser distance), 1.0–2.0 kpc for IRAS 01082+5717, and 0.7–1.5 kpc for IRAS 05380+2020. We selected models where the fits were limited to $(\chi ^{2}-\chi ^{2}_{{\rm best}})/{n}_{{\rm data}} < 2$ and used a weighted average to estimate YSO properties in the manner of Alexander et al. (2013).

Table 4. Candidate YSOs

IRAS Name	YSO ID	WISE ID	J	H	K_S	W1	W2	W3	W4
00259+5625
	Y1	J002841.65+564152.7	17.111	14.439	12.371	10.460	9.262	6.667	3.736
	Y2	J002831.33+564419.9	11.998	11.770	11.656	11.410	11.281	7.827	5.799
	Y3	J002845.22+564307.6	17.233	14.718	12.744	11.271	10.181	8.412	6.298
	Y4	J002836.86+564245.4	16.267	14.490	13.672	12.564	11.858	7.134	4.341
00420+5530
	Y5	J004356.01+555118.6	15.169	14.265	13.951	12.255	11.651	9.595	7.661
	Y6	J004503.67+554648.9	10.564	10.362	9.979	8.693	7.980	4.345	1.532
	Y7	J004414.63+554847.0	13.934	12.641	11.900	10.456	9.030	6.640	3.959
	Y8	J004412.68+554859.7	16.755	14.996	13.506	12.152	10.694	8.052	4.513
	Y9	J004409.63+554916.3	16.867	15.276	14.362	11.828	10.353	7.660	5.233
	Y10	J004421.45+554642.6	11.269	10.754	10.135	8.731	7.924	6.036	4.667
	Y11	J004413.27+555256.8	11.020	10.901	10.835	10.727	10.628	9.180	5.931
	Y12	J004457.31+554718.1	13.024	11.160	9.947	8.094	7.204	3.432	0.147
	Y13	J004412.78+555205.1	14.162	13.286	12.830	11.836	11.253	9.409	7.701
01082+5717
	Y14	J011122.88+573315.9	14.817	13.834	13.283	11.490	10.747	5.742	4.116
	Y15	J011121.12+573328.7	11.645	11.543	11.496	10.762	10.144	6.176	4.202
	Y16	J011111.23+573648.9	12.288	12.264	12.222	11.999	11.816	8.845	6.478
	Y17	J011123.14+573150.0	13.757	13.544	13.382	12.640	12.371	7.406	5.906
	Y18	J011134.42+573632.9	13.669	13.009	12.214	11.089	10.466	8.202	5.637
05380+2020
	Y19	J054054.13+202247.7	16.930	16.224	15.180	10.255	9.325	6.580	3.967
	Y20	J054107.67+201806.9	15.775	14.932	14.502	13.873	13.092	8.897	5.182

Download table as: ASCII Typeset image

Table 5 compiles the results of the YSO fitting. This table lists the number of models well fit to each YSO SED and the χ²/n_data as a measure of the goodness of fit. It also lists best fit and average values for the extinction, stellar mass, disk mass, and disk accretion rate, in addition to the range of stellar mass fit by the models and the total YSO luminosity. These parameters were used to determine the stage of the YSO. Robitaille et al. (2006) assigns stage 0/I for heavily embedded sources with envelope accretion rates $\dot{{M}}/{M}_{*}$ > 10⁻⁶ yr⁻¹. Stage II YSOs have $\dot{{M}}/{M}_{*}$ < 10⁻⁶ yr⁻¹ and M_disk/M_* > 10⁻⁶. They have little or no mass left in the envelope, but are young enough to still have a protostellar disk. Stage III YSOs have $\dot{{M}}/{M}_{*}$ < 10⁻⁶ yr⁻¹ and M_disk/M_* < 10⁻⁶. They have little or no mass left in either an envelope or a protostellar disk. Sources with fits that do not have a total weight (probability) greater than 0.67 for any single stage are assigned as ambiguous (amb).

Table 5. Candidate YSO Fitted Parameters

IRAS Name	ID	No. of Fits	χ²/n_data	A_V		M_*(M_☉)		M_*(M_☉)	log₁₀(M_disk/M_☉)		$\mathrm{log}_{10}(\dot{{M}}/{M}_\odot \mathrm{yr}^{-1})$		log₁₀(L/L_☉)	Integer Stage		F_stage	σ_stage
IRAS Name	ID	No. of Fits	Best	Best	Avg	Best	Avg	Range	Best	Avg	Best	Avg	Avg	Best	Avg	Avg	Avg
00259+5625
	Y1	294	0.03	24.72	24.09	3.7	4.0	1.5–5.7	-3.22	−1.41	−4.94	−4.37	2.14	I	amb	1.54	0.53
	Y2	163	0.94	0.00	0.42	2.3	2.8	0.8–3.7	-1.70	−2.55	−7.12	−6.32	1.39	II	II	2.07	0.19
	Y3	530	0.87	16.38	17.56	3.0	3.0	1.0–4.5	-1.31	−1.99	0.00	−6.26	1.82	II	II	1.99	0.02
	Y4	51	0.50	4.83	4.86	1.2	0.8	0.3–3.0	-1.62	−1.84	−4.91	−4.52	1.00	I	I	1.02	0.04
00420+5530
	Y5	41	2.73	5.63	5.18	1.9	1.9	0.1–2.8	-5.07	−2.95	0.00	−7.56	1.12	II	II	1.99	0.02
	Y6	24	0.87	2.27	2.44	5.2	5.1	4.8–5.2	-5.83	−5.54	0.00	0.00	2.70	III	amb	2.66	0.45
	Y7	33	1.18	0.46	0.66	1.3	2.6	0.1–4.6	-2.67	−1.12	−5.07	−4.39	1.44	I	I	1.06	0.11
	Y8	415	0.89	12.88	14.20	1.9	1.0	0.1–4.6	-2.32	−1.97	−5.55	−4.82	0.95	I	I	1.01	0.02
	Y9	30	2.18	12.64	12.06	3.0	2.1	0.2–3.8	-2.40	−2.06	0.00	−4.43	1.65	II	amb	1.63	0.46
	Y10	102	0.58	0.69	1.17	2.3	2.8	2.2–4.0	-2.13	−2.28	0.00	0.00	1.80	II	II	2.00	0.00
	Y11	55	0.88	1.76	1.08	2.3	2.5	2.0–2.7	-7.27	−5.80	0.00	−8.64	1.48	III	III	2.87	0.23
	Y12	174	0.21	3.69	3.10	6.8	5.3	0.9–7.4	-2.53	−1.25	−4.69	−4.33	2.41	I	I	1.01	0.01
	Y13	399	0.18	3.89	3.57	1.8	2.0	0.1–3.4	-4.38	−2.21	0.00	−7.54	1.17	II	II	2.01	0.06
01082+5717
	Y14	46	6.24	5.33	9.63	3.2	3.4	2.7–3.8	-4.45	−4.10	0.00	0.00	2.11	II	II	2.02	0.04
	Y15	16	3.49	0.03	1.12	3.2	3.1	3.0–3.2	-4.64	−3.33	0.00	0.00	1.77	II	II	2.00	0.00
	Y16	40	1.21	0.76	0.86	2.5	2.5	2.4–3.5	-5.36	−5.44	0.00	0.00	1.58	II	II	2.30	0.42
	Y17	11	5.85	2.31	2.79	2.2	2.3	2.3–2.5	-2.78	−2.89	0.00	0.00	1.48	II	II	2.00	0.00
	Y18	31	0.57	0.07	2.50	2.2	2.3	0.9–3.2	-2.17	−2.23	0.00	−8.23	1.37	II	II	2.00	0.00
05380+2020
	Y19	2	14.73	3.61	3.69	3.3	3.3	3.3–3.5	-3.07	−3.08	−3.21	−3.21	1.90	I	I	1.00	0.01
	Y20	25	0.62	4.58	3.60	1.3	1.0	0.1–3.2	-1.49	−1.65	−6.16	−4.96	0.84	II	amb	1.59	0.48

Download table as: ASCII Typeset image

4. THE STELLAR CONTENT OF IM SFRs

4.1. IRAS 00259+5625

IRAS 00259+5625 was classified as an IM SFR in the Bok globule CB3 by Codella & Bachiller (1999) due to its IR bolometric luminosity of 930 L_☉ (Launhardt & Henning 1997). Launhardt & Henning (1997) found a kinematic distance of 2.5 kpc assuming that IRAS 00259+5625 was associated with the near side of the Perseus arm. Yun & Clemens (1994) obtained ¹²CO velocity distance of 2.1 kpc using the Clemens (1985) rotation curve. There is a compact submillimeter source likely consisting of an aggregate of multiple Class 0 sources 15'' west and 6'' south of the IRAS 00259+5625 source position (Huard et al. 2000) indicated by the white diamond in the left panel of 4. The black diamond in the left panel of 4 indicates an H₂O maser also present in the region (Scappini et al. 1991).

Figure 10 shows our optical spectra of the five stellar targets. Black solid lines denote our spectra, while the dotted lines show the best comparison spectra from the Jacoby et al. (1984) spectral atlas. Star A has a weak He i λ5876 Å absorption feature (EW = 0.29 ± 0.13 Å) and a strong Hα absorption line (EW = 7.99 ± 0.57 Å), indicative of a mid-B star (B7). Star B has a spectrum with strong Hα absorption (EW = 14.95 ± 0.17 Å) indicative of an early A star (A1). Star C has strong, narrow Hα absorption (EW = 4.53 ± 0.19 Å) indicative of a mid-F star (F6). Star D has weak, narrow Hα absorption and the presence of the Mg i triplet at λλ5167, 5172, 5183 Å indicative of an early G star (G0). Star E has a weaker, narrow Hα absorption and an increasing presence of the Mg i triplet at λλ5167, 5172, 5183 Å indicative of a late G star (G9). We used the spectrum of Star B and its 2MASS magnitudes and colors to derive a spectrophotometric distance of 1.65 $^{+0.04+0.68}_{-0.08}$ kpc with an extinction of A_V = 1.03 $^{+0.04}_{-0.03}$ mag for the region.

IRAS 00259+5625 contains four candidate YSOs in the vicinity of the IM SFR. These include three intermediate-mass YSOs (Y1, Y2, and Y3) at 4.0, 2.8, and 3.0 M_☉ and one lower-mass YSOs (Y4) at 0.8 M_☉. Of the three intermediate-mass YSOs, two are Stage II YSOs (Y2 and Y3), while the Y1 fit is ambiguous. The lower-mass YSO, Y4, is Stage I. The total luminosity contribution of these YSOs is 2.4 × 10² L_☉. Using the Sanders & Mirabel (1996) definition for infrared luminosity, we use the IRAS fluxes to derive an infrared luminosity of 3.1 × 10² L_☉ for the region, which is consistent with a total luminosity from the most massive YSOs identified and tabulated in Table 5.

The cleaned CMD of IRAS 00259+5625 in the left panel of Figure 6 is rather sparse, showing just a few sources near the nominal main sequence at a distance of 1.65 kpc and reddening of A_V = 1.03 mag. Star B, found to be an early A star from our optical spectroscopy, is labeled and lies near the nominal A0V star (lower X) along the main sequence. Star A (∼B7) is not drawn because the 2MASS Catalog lists only upper limits for its J and H photometry as the source is detected, but not resolved from a nearby companion, in these bands. Stars C and D (mid- to late-F) lie just to the right of the main sequence along the 10 Myr isochrone near where it joins the main sequence. These appear slightly too bright to be mid-F stars at the adopted cluster distance and reddening; they may be foreground objects at d ≃ 0.6 kpc (see Table 3), or they could be binary/multiple systems having magnitudes brighter than those of single stars of the equivalent type. It is unlikely that they are overluminous pre-main-sequence objects because our optical spectra do not show the Hα emission signatures of active accretion. Star E, found to be a late G type, lies well to the upper right of the main sequence (brighter and redder), consistent with it possibly being a field giant. A group of about eight stars lie along the 2.5 and 10 Myr isochrones, most of them near J − K_S = 1.2, at the locations where late-stage pre-main-sequence stars would be expected. Another sequence of about eight stars stretch redward from pre-main-sequence tracks, consistent with them being either enshrouded pre-main-sequence stars or more reddened background objects. Of the four YSOs in this region, three lie within the marked cluster aperture (Y1, Y3, and Y4), but only Y4 appears on Figure 6 near K = 13.67, J − K_S = 2.6. The other two lie off the plot because of their extreme colors (J − K_S ∼ 4.6). The CCD for IRAS 00259+5625 (Figure 8) reveals a loose group of stars near H − K_S = 0.4, J − H = 1.0, with a few stretching along the reddening vector to more extreme colors. This grouping of stars may consist of moderately reddened (A_V ∼ 4) intermediate-mass stars, lightly reddened low-mass stars, or some combination thereof. Taken together, the cleaned CMD, CCD, and the spectra of IRAS 00259+5625 are consistent with a poor cluster of pre-main-sequence stars in the vicinity of the mid-B, early A, and late F stars that appear to be the most massive constituents of this region. The limited depth of the 2MASS photometry precludes a more definitive characterization of the possible stellar cluster in IRAS 00259+5625, though by using the Weidner et al. (2013) most massive star–cluster mass relation we can estimate a total cluster mass of ∼20 M_☉ in ∼65 stars above 0.1 M_☉ after integrating a typical Kroupa (2001) initial mass function (IMF).

4.2. IRAS 00420+5530

Kumar et al. (2006) investigated the IRAS 00420+5530 region and found it to contain a 380 M_☉ cluster. Molinari et al. (2002) used millimeter data to determine that this region had a bolometric luminosity of L = 12, 400 L_☉ at a kinematic distance of 5.0 kpc. Based on its radio free–free flux, they also determined that a ∼3000 L_☉ B2 ZAMS star was present. Zhang et al. (2005) found a kinematic distance of 7.72 kpc. Moellenbrock et al. (2009) obtained a distance of 2.17 kpc via the trigonometric parallax of an H₂O maser. The maser is indicated by the black diamond in the right panel of Figure 4. Koenig et al. (2012) noted a small cluster of YSOs associated with this region and considered it to be distinct from the nearby massive star-forming region NGC 281, for which Sato et al. (2008) obtained an H₂O maser trigonometric parallax distance of 2.82 kpc. Rygl et al. (2010) suggested that both IRAS 00420+5530 and NGC 281 may be on the edge of an expanding superbubble.

Figure 11 shows spectra of the four stars that were obtained revealing two intermediate-mass stars. Star A has a spectrum with a weak He i λ5876 Å absorption feature (EW = 0.19 ± 0.02 Å) and a strong Hα emission feature (EW = −8.663 ± 0.02 Å). Spectral classification using the He i line results in a B8e star; however, the He i EW may be a lower limit if there is a nebular emission in that line. Therefore, the temperature class could be earlier than B8. Furthermore, this wavelength region lacks the luminosity diagnostics to distinguish a classical Be star from a B supergiant. As a result, the spectrophotometric distance for star A was not calculated. Star B has a spectrum with strong He i λ5876 Å absorption (EW = 0.73 ± 0.12 Å) and Hα absorption (EW = 4.491 ± 0.27 Å) that appears to indicate an early B star (B2). Star C has strong, narrow Hα absorption (EW = 6.339 ± 0.13 Å) indicative of an early F star (F2). Star D has a strong, narrow Hα absorption (EW = 4.00 ± 0.07 Å) indicative of a late F star (F9). We used the spectrum of Star B and its 2MASS colors to derive a spectrophotometric distance of 1.14 $^{+0.34+0.47}_{-0.25}$ kpc at an extinction of A_V = 6.33 $^{+0.17}_{-0.17}$ mag for this region. This distance is not consistent with the maser parallax distance measurement of 2.17 kpc, though the agreement is better than with the kinematic distance measurements of >5 kpc. A portion of this discrepancy can be reconciled if Star B is in a binary system with a companion of nearly equal luminosity.

IRAS 00420+5530 contains nine candidate YSOs in the vicinity of the IM SFR. These include eight intermediate-mass YSOs and one lower-mass YSO. Of the eight intermediate-mass YSOs, two are Stage I (Y7 and Y12), three are Stage II (Y5, Y10, and Y13), one is Stage III (Y11), and two fits are ambiguous (Y6 lies between Stage I and II, and Y9 lies between Stage II and III). The lower-mass YSO, Y8, is Stage I. Y6 apparently coincides spatially with Star C. The spectrum obtained from Star C is consistent with an early F temperature class and shows no indication of accretion. The SED model fits, however, are consistent with a 5.1 M_☉ YSO. This may indicate that a foreground F star (estimated distance 0.46 kpc) is projected in front of the IR emission from a intermediate-massive YSO with the IM SFR estimated to lie at 2.2 kpc. In that case, the flux contribution from the YSO and the foreground star would be ambiguous in the 2MASS bands increasing the uncertainty in the YSO fit. The calculated spectrophotometric distance and extinction listed in Table 3 is almost certainly in error because of this possible blend. The total luminosity contribution of these YSOs is 9.6 × 10² L_☉. Using the Sanders & Mirabel (1996) definition for infrared luminosity, we use the IRAS fluxes to derive an infrared luminosity of 8.8 × 10² L_☉ at 1.14 kpc or 3.2 × 10³ L_☉ at 2.17 kpc. The total YSO luminosity is consistent with the closer spectro-photometric distance, but if the cluster is at the maser parallax distance this would suggest that there are more intermediate-mass YSOs contributing to the infrared luminosity that have not been accounted for.

The cleaned CMD of IRAS 00420+5530 is rather rich, showing a distinct grouping of stars near the nominal main sequence at a distance of 1.14 kpc and reddening of A_V = 6.33 mag. Star A lies to the left of the main sequence and is possibly a foreground classical Be star. Star B, found to be an early B star from our optical spectroscopy, is labeled and lies near the nominal B0V star (upper X) along the main sequence. Stars C and D (early to late F) lie to the left of the main sequence. These appear too bright to be mid-F stars at the adopted cluster distance and reddening; they are likely foreground objects at d ≃ 0.26 kpc and d ≃ 0.71 (see Table 3). A group of about 25 stars lies near or below the nominal main sequence, most of them between J − K_S = 1.2–1.7. Another sequence of about eight stars stretch redward from pre-main-sequence tracks, consistent with them being either enshrouded pre-main-sequence stars or more reddened background objects. Two of the nine YSOs identified in the IRAS 00420+5530 field lie within the marked cluster aperture (Y6 and Y12). Star C (Stage II/III; ≡Y6?) appears on Figure 6 near K_S = 9.98, J − K_S ≃ 0.6 and is likely a foreground object along the line of sight toward the YSO, as discussed above. Y12 (Stage I/II; 5.3 M_☉) appears on Figure 6 near K_S = 9.95, J − K_S ≃ 3.1 where a young, enshrouded pre-main-sequence star would be expected. The CCD for IRAS 00420+5530 (Figure 8) reveals that the rich grouping of stars from the CMD has a large spread in color from H − K_S = 0.1 to H − K_S = 0.8. There are a few stars stretching along the reddening vector to more extreme colors, including Y12. Several of these stars appear to have H − K_S colors redder than would be expected for main-sequence objects at higher reddening. This suggests that these are possible pre-main-sequence stars with k-band excesses. The nature of the stars that lie to the left of the main sequence is less clear. The unusually blue H − K_S colors may be the result of reflection nebulosity or poor 2MASS photometry, as these stars have typical uncertainties of ∼0.2 mag in J, H, and K_S bands. Taken together, the cleaned CMD, CCD, and the spectra of IRAS 00420+5530 are consistent with a cluster of pre-main-sequence stars in the vicinity of the early B star that appear to be the most massive constituent of this region. Using the Weidner et al. (2013) most massive star–cluster mass relation, we estimate a total cluster mass of ∼78 M_☉ in ∼250 stars above 0.1 M_☉ after integrating a typical Kroupa (2001) IMF.

4.3. IRAS 01082+5717

IRAS 01082+5717 has largely been unexplored in the literature. It was identified as a site of diffuse IRAS emission with a radius of 9' by Fich & Terebey (1996) in a survey of IRAS 60 μm images.

Figure 12 shows the spectra of the four stellar targets revealing three intermediate-mass stars. Star A has a spectrum with a strong He i λ5876 Å absorption feature (EW = 0.53 ± 0.01 Å) and strong Hα absorption (EW = 7.14 ± 0.52 Å), indicative of a mid-B star (B4). Star B shows a weak He i λ5876 Å absorption feature (EW = 0.15 ± 0.02 Å) and an Hα absorption feature (EW = 7.40 ± 0.06 Å) indicating a late B star (B8). Star C exhibits a weak He i λ5876 Å absorption feature (EW = 0.15 ± 0.03 Å) and an Hα absorption feature (EW = 8.76 ± 0.14 Å) indicating a late B star (B8). Star D has a strong, narrow Hα absorption feature (EW = 3.798 ± 0.07 Å) indicative of an early G star (G0). The separation between Star B and Star C is ∼2'' leading to blending in the 2MASS images. Unfiltered guide camera images taken at the time as the spectroscopic observations indicate that Stars B and C are roughly the same visual magnitude. The 2MASS catalog photometry for Star B, which includes the flux contribution from Star C, is included in the CMD and CCD for reference, as well as in Table 3. We used the spectrum of Star A and its 2MASS photometry to derive a spectrophotometric distance of 1.84 $^{+0.26+0.76}_{-0.23}$ kpc at an extinction of A_V = 1.22 $^{+0.09}_{-0.11}$ mag for this region.

IRAS 01082+5717 contains five YSOs, all of intermediate mass and all Stage II (Y14, Y15, Y16, Y17, and Y18) in the vicinity of the IM SFR. Y16–Y18 all lie outside the nominal cluster aperture shown in Figure 5, but they are still in the field of IRAS 01082+5717. Y15 corresponds to Star A. The spectrum of Star A presents as a mid-B star (∼5 M_☉) at an A_V of 1.22 mag and shows no indication of accretion. This agrees with the average SED model fit of a 3.1 M_☉ YSO at an average A_V of 1.12 mag. The mean extinctions for the five YSOs vary from 0.86 to 9.6 mag., but this could be the result of differential extinction within the SFR coupled with the limited precision of the SED fitting technique. The total luminosity contribution of these YSOs is 2.8 × 10² L_☉. Using the Sanders & Mirabel (1996) definition for infrared luminosity, we use the IRAS fluxes to derive an infrared luminosity of 1.9 × 10² L_☉ for the region.

The cleaned CMD of IRAS 01082+5717 is rather sparse, showing just a few sources near the nominal main sequence at a distance of 1.84 kpc and reddening of A_V = 1.22 mag. Star A, found to be a mid-B star from our optical spectroscopy, is labeled and lies between the nominal B0V and A0V loci (upper and lower X's) along the main sequence. Star B (late B) lies above its expected location on the main sequence. This is likely owing to confusion with Star C (late B), which was not drawn owing to its lack of photometry in the 2MASS catalog. Star D (early G) lies just to the right of and above the main sequence consistent with it possibly being a foreground dwarf or a background giant. Accordingly, we do not compute a spectrophotometric distance for this star. A group of about seven stars lies along the 2.5 Myr isochrone at locations where pre-main-sequence stars would be expected. Another sequence of about five stars lies along the main-sequence track. There is also a group of about five stars that stretches redward from pre-main-sequence tracks, consistent with them being either enshrouded pre-main-sequence stars or more reddened background objects. Two of the five YSOs identified in this field lie within the marked cluster aperture (Y14 and Y15). Y14 appears on Figure 7 near K = 13.28, J − K_S ≃ 1.5. Y15 (Star A) lies along the main sequence at the adopted distance and reddening based upon its spectral type. The CCD (Figure 9) reveals a sparse group of stars near H − K_S = 0.3, J − H = 0.8, with a few stretching along the reddening vector to more extreme colors. Three of these stars appear to have k-band excesses, which suggest that they may be pre-main-sequence stars. Taken together, the cleaned CMD, CCD, and the spectra are consistent with the region hosting a poor cluster of pre-main-sequence stars in the vicinity of the mid-B and late B stars that appear to be the most massive constituents. The limited depth of the 2MASS photometry precludes a more definitive characterization of the possible stellar cluster in IRAS 01082+5717, though using the Weidner et al. (2013) most massive star–cluster mass relation, we can estimate a total cluster mass of ∼37 M_☉ in ∼120 stars above 0.1 M_☉ after integrating a typical Kroupa (2001) IMF.

4.4. IRAS 05380+2020

IRAS 05380+2020 is a star-forming region included in an H i survey by Lu et al. (1990), who found a galaxy with a redshift of 6735 km s⁻¹ at this position. Takata et al. (1994) argued that the infrared luminosity is very high and not compatible with the H i profile of a galaxy at such a distance, indicating that there is likely a galaxy behind this Galactic infrared source.

Figure 13 shows the five spectra that were obtained for IRAS 05380+2020, revealing three intermediate-mass stars. Star A has a strong He i λ5876 Å absorption feature (EW = 0.71 ± 0.33 Å) and Hα absorption (EW = 5.61 ± 0.13 Å). This spectrum is indicative of an early B star (B2). Star B has weak He i λ5876 Å absorption (EW = 0.19 ± 0.06 Å) and an Hα absorption feature (EW = 6.63 ± 0.20 Å) indicating a late B (B8) star. Star C has a strong Hα absorption feature (EW = 14.40 ± 0.11 Å) indicating an early A star (A1). Star D has a narrow Hα absorption (EW = 2.94 ± 0.24 Å) indicative of a mid-G (G4) star. Star E has a very narrow Hα absorption feature (EW = 1.06 ± 0.07 Å) and a strong Fe i λ6495 Å absorption feature indicative of an early K (K0) star. We used the spectrum of Star A and its 2MASS photometry to derive a spectrophotometric distance of 1.34 $^{+0.40+0.55}_{-0.30}$ kpc at an extinction of A_V = 2.74 $^{+0.17}_{-0.17}$ mag for this region.

IRAS 05380+2020 contains two candidate YSOs in the vicinity of the IM SFR. Of these, Y19 is of intermediate mass and Y20 is of lower mass. Y19 is Stage I and Y20 is ambiguously fit by the models (I/II). Y19 is poorly fit as a result of a ∼5 magnitude increase from the K_S band to the WISE W1 band. This may be owing to variability of the source in the ∼12 yr between the 2MASS and WISE observations. The total luminosity contribution of these YSOs is 8.7 × 10¹ L_☉. Using the Sanders & Mirabel (1996) definition for infrared luminosity, we use the IRAS fluxes to derive an infrared luminosity of 2.5 × 10² L_☉.

The cleaned CMD of IRAS 05380+2020 is rather rich, showing a distinct grouping of stars near the nominal main sequence at a distance of 1.34 kpc and reddening of A_V = 2.74 mag. Star A, found to be an early B star from our optical spectroscopy, is labeled and lies near the nominal B0V locus (upper X) along the main sequence. Star B, found to be a late B star, is labeled and lies between the nominal B0V and A0V loci along the main sequence. Star C lies to the left of the main sequence and above the nominal A0V locus. This appears too bright to be an early A star at the adopted cluster distance and reddening; this is likely a foreground object at d ≃ 0.78 kpc and correspondingly lower reddening, A_V = 0.61 mag (see Table 3). A group of about 28 stars scatters about the nominal 2.5 Myr isochrone, most of them between J − K_S = 1.2–1.6. Another sequence of about six stars stretches redward from the pre-main-sequence tracks, consistent with them being either enshrouded pre-main-sequence stars or more reddened background objects. One of the two YSOs (Y19) lies within the marked cluster aperture. Y19 is located on Figure 7 near K = 15.18, J − K_S = 1.75 where a young, enshrouded pre-main-sequence star would be expected. The CCD for IRAS 05380+2020 (Figure 9) reveals that the rich grouping of stars from the CMD has a small spread in J − H color, but a large spread in H − K_S color from 0.2 to 0.7. There is also a grouping along the main sequence and several stars stretching along the reddening vector to more extreme colors. Several stars appear to have k-band excesses suggesting possible pre-main-sequence stars. Taken together, cleaned CMD, CCD, and the spectra are consistent with a cluster of pre-main-sequence stars in the vicinity of the early and mid-B stars that appear to be the most massive constituents of this region. Using the Weidner et al. (2013) most massive star–cluster mass relation, we estimate a total cluster mass of ∼78 M_☉ in ∼250 stars above 0.1 M_☉ after integrating a typical Kroupa (2001) IMF.

4.5. Initial Mass Function Modeling

Given that low- and intermediate-mass clusters suffer from stochastic effects when populating the upper end of the IMF, we utilize the stellar cluster simulation software MASSCLEAN (Popescu & Hanson 2009) to model the clusters in our pilot study. MASSCLEAN allows for stochastic fluctuations in the simulations. We compute the mass distribution using a Kroupa–Salpeter IMF (Kroupa 2002; Salpeter 1955) with an upper mass limit of 150 M_☉. We randomly generate 1000 clusters for a range of total cluster masses from 3 to 1000 M_☉ in 1 M_☉ bins and tabulate the fraction that reproduce the upper end of the mass function observed in our clusters. Figure 14 shows the fraction of simulated clusters within each mass bin producing a cluster with the most massive star 2 M_☉ ⩽ M_max ⩽ 8 M_☉ (solid bold line) and the most massive star 4 M_☉ ⩽ M_max ⩽ 10 M_☉ (dotted bold line). We also explored requiring the cluster to have exactly two stars in the mass range 2 M_☉ ⩽ M ⩽ 8 M_☉ including the most massive star (solid light line) and requiring the cluster to have exactly two stars in the mass range 4 M_☉ ⩽ M ⩽ 10 M_☉ including the most massive star (dotted, light line). The lines in Figure 14 have been smoothed to three solar mass bins to reduce scatter. This simulation demonstrates that it is likely that clusters forming only two intermediate mass stars have a total mass between 20 and 80 M_☉, consistent with masses estimated from the Weidner et al. (2013) most massive star–cluster mass relation. While we cannot completely rule out that these are fairly massive clusters that have happened not to form a massive star, it is highly improbable that clusters with intermediate-mass stars as their most massive constituents have a total mass greater than 150 M_☉. While it is common for a massive cluster to produce a massive star and eject it via few-body interactions to become a runaway OB star, it is highly improbable to dynamically eject the most massive star from a cluster with <316 M_☉(Oh & Kroupa 2012).

**Figure 14.** Fraction of simulated clusters in each total cluster mass bin that meets the specified stellar population criteria: the most massive star 2 M_☉ ⩽ M_max ⩽ 8 M_☉ (solid bold line), the most massive star 4 M_☉ ⩽ M_max ⩽ 10 M_☉ (dotted bold line), exactly two stars in the mass range 2 M_☉ ⩽ M ⩽ 8 M_☉, including the most massive star (solid light line), and exactly two stars in the mass range 4 M_☉ ⩽ M ⩽ 10 M_☉, including the most massive star (dotted light line).
Download figure:
Standard image High-resolution image

5. SUMMARY AND CONCLUSIONS

In this article, we morphologically classify an all-sky sample of 984 candidate IM SFRs using WISE mid-IR images. The objects in this all-sky sample were selected from the IRAS PSC using colors that indicate large PAH contribution and cool dust found in IM SFRs. We determined that 616 (62.6%) are classified as "blobs/shells," 128 (13.0%) are classified as "filamentary," 39 (4.0%) are classified as "star-like" objects, and 201 (20.4%) are classified as "galaxies." We find that the blobs/shells, filamentary, and star-like objects are concentrated in the Galactic Plane, while galaxies appear to be uniformly distributed throughout the sky.

We conducted a pilot study of four Galactic IM SFRs that are classified as blobs/shells using optical spectroscopy and near-IR photometry. Optical spectroscopy of four to five stars per region demonstrates that each contains, within its projected boundaries, two or more intermediate-mass stars. Eight of the eighteen spectroscopically classified stars are B-type and two are A-type with no stars more massive than ∼B2 (8 M_☉) found. Near-IR CMDs using 2MASS data indicate that these stars are the most massive & luminous constituents, suggesting that there may be an upper limit to the mass of stars in IM SFRs of ∼B2. SEDs of candidate YSOs in each region also indicate that they are continuing to produce intermediate- and lower-mass stars, but not massive stars. These findings support the idea that the IRAS color selection (Kerton 2002; Arvidsson et al. 2010) does indeed select regions producing intermediate-mass stars and that these regions can be distinguished from high-mass SFRs using IRAS mid-IR colors.

Cleaned 2MASS CMDs reveal a strong excess of stars along the pre-main-sequence isochrones in IRAS 00420+5530 and IRAS 05380+2020 and a less obvious excess in IRAS 00259+5625 and IRAS 01082+5717, indicating the presence of small clusters associated with the spectroscopically-identified IM stars. The former two regions (IRAS 00420+5530 and IRAS 05380+2020) that have the strongest evidence for stellar clusters also host the earliest stars as their most massive members (∼B2), while the latter two clusters (IRAS 00259+5625 and IRAS 01082+5717) host somewhat later spectral types (B7 and B4, respectively). Thus, the most massive stars appear to be associated with richer and more massive stellar clusters. While the current small sample size precludes a definitive statement, this correlation is consistent with the idea that there may be a turn-on point for cluster formation at the cloud masses and stellar masses associated with the formation of mid-B spectral types as the most massive constituent.

We acknowledge the support of Carlos A. Vargas Alvarez, Anirban Bhattacharjee, Michael DiPompeo, and Jiajun Chen for their assistance in obtaining spectroscopy data. This work was supported by NASA through grants NAS2-97001 from the Stratospheric Observatory for Infrared Astronomy and ADAP NNX10AD55G. We thank James Weger and Jerry Bucher for their work in maintaining the Wyoming Infrared Observatory as an excellent astronomical facility. We also thank the anonymous referee for comments and suggestions that helped to improve this manuscript.

AN ALL-SKY SAMPLE OF INTERMEDIATE-MASS STAR-FORMING REGIONS

Article metrics

Permissions

Author affiliations

Dates

ABSTRACT

1. INTRODUCTION

2. SAMPLE SELECTION