THE BLAST VIEW OF THE STAR-FORMING REGION IN AQUILA (ℓ = 45°, b = 0°)

GRSMC 45.122+0.132 (G45.12; IRAS 19111+1048; A16 in the present work) contains a cometary UCHiiR (WC89) classified as a massive young stellar object (MYSO) by Chan et al. (1996). It shows extended submillimeter emission toward the north (upper left of Figure 6), observed in the MIR as a bright rimmed H ii region. This structure is likely produced by the activity of the young massive population embedded in the innermost regions of the clump, as detected by Vig et al. (2006) using the Giant Metrewave Radio Telescope (GMRT) at 1200 and 610 MHz. The 19 embedded OB stars observed by these authors are located within and surrounding the peak of our BLAST clump, mainly within the H ii region. Their brightest source agrees well with our submillimeter peak (≲5''), and although it remained unresolved in their data sets, they suspected it to be a cluster responsible for the extended morphology in the radio. A possible cluster powering this region has also been suggested through [Ne ii] observations by Takahashi et al. (2000), and later by Zhu et al. (2008), who detected three additional peaks (G45.12N, G45.12SE, and G45.12SW) nearby the main UCHiiR from WC89.

Zoom In Zoom Out Reset image size

Our analysis of the 4.8 GHz CORNISH data confirms such suspicions, with the detection of 28 radio sources in a region ∼0.8 pc in radius centered on (and including) the central brightest member (A16−C24; Table 3 and Figure 7). These objects are believed to be forming the unresolved source observed by Vig et al. (2006; S14 in their sample). None have a radio ZAMS spectral type later than B0 (with the exception of A16−C18, which was not initially detected by our extraction routine).

Zoom In Zoom Out Reset image size

We find that 22 of these radio sources lie within ∼0.5 pc of A16−C24. This object has a bolometric luminosity equivalent to an O6 ZAMS star, which is in good agreement with previous results for a distance of ∼6 kpc (e.g., Kraemer et al. 2003). The position of A16−C24 also agrees with the UCHiiR of WC89 and Testi et al. (1999). The two Two Micron All Sky Survey (2MASS) IR sources detected by Vig et al. (2006) within their S14 source (IR4 and IR5) agree well with A16−C24 and A16−C26 and the two peaks (KJK1 and KJK2) detected by Kraemer et al. (2003). The G45.12SE peak from Zhu et al. (2008) traces the region of A16−C26; G45.12SW is formed by the conglomeration of radio sources toward A16−C13, and G45.12N is centered on the arc-shaped radio emission around A16−C21. The estimated spectral types of all radio sources within ∼1' of the BLAST coordinates have been included in Table 3 for an electron temperature similar to that estimated by Vig et al. (2006; ∼8000 K, for their source containing the cluster S14).

Taking into account the 3σ positional error of our CORNISH source A16−C24 (∼05) and that expected for the OH maser detection of Baart & Cohen (1985), ∼10 mas, the separation of ∼3'' between the two may be physically significant. Indeed, our projected separation of ∼0.1 pc positions the maser source at the edge of the UCHiiR, and agrees well with the displacement measured by Baart & Cohen (1985), of the order of ∼0.03 pc (with respect to the coordinates from Matthews et al. 1977). This distance, as pointed out by Baart & Cohen (1985), is well within the upper limit for the radius at which OH masers are produced in a shell-like front surrounding a massive young star. This suggests that these may be dense fragments of the expanding shell. The blueshift for the maser clusters detected by these same authors (∼6 km s⁻¹) is also similar to that observed for the dense gas traced with ammonia by Hofner et al. (1999), whose absorption peak coordinates are also offset from our radio peak in the same direction as the maser source. This supports the real physical origin of the observed displacement.

Although further investigation is required, our analysis of the CORNISH data and the evidence in the mid-infrared (MIR) and submillimeter suggest that A16−C24 may be the initial trigger and power source of G45.12. The massive outflow detected by Hunter et al. (1997) appears to follow the overall direction of the distribution of CORNISH sources well, which could indicate that A16−C24 has induced the formation of its lower mass companions. In addition, the distribution of CORNISH sources appears to have a slightly curved, bow-shock-like morphology pointing toward the submillimeter comma-like structure extending downward in Figure 8. This structure also coincides with the strongest emission in ¹³CO and CS. Considering its central position, A16−C24 appears as a suitable candidate for the main triggering/compressing source.

Zoom In Zoom Out Reset image size

Just like its neighbor, IRAS 19110+1045 (A12; GRSMC 45.073+0.129 [G45.07]) also shows submillimeter, IR, and radio structure signaling the presence of massive star formation, albeit significantly less pronounced than in G45.12. Only nine possible CORNISH detections are found within ∼1' of the BLAST position, with radio ZAMS/class V spectral types ranging between B0.5 and O9 for an electron temperature of 8000 K (Table 3). The lower number of stars and their later apparent radio spectral type is consistent with the more compact size of the resulting ionized region compared to G45.12.

Two CORNISH sources, A12−C2 and A12−C3 (Figure 9), match (≲35) those detected by Vig et al. (2006) and the two embedded IR sources (KJK1, KJK3) from Kraemer et al. (2003). Despite the positional accuracy of our CORNISH measurements (<15 difference between our two coordinates and those given by Testi et al. 1999 and WC89), we find no indication of the third IR source (KJK2) detected by Kraemer et al. (2003). This source is located ∼2'' from A12−C3 and should have been identifiable at the resolution limit of the CORNISH data.

Zoom In Zoom Out Reset image size

As with G45.12, G45.07 also contains a UCHiiR (spherical/unresolved UCHiiR; WC89) with a CO outflow (Hunter et al. 1997) and masers. Our brightest radio source lies at <15 from the coordinates given by WC89. The masers show a more significant displacement (∼2'') and appear to extend along the overall outflow direction. In addition to hydroxyl masers, G45.07 also shows several water maser sources (e.g., Hofner & Churchwell 1996) and 6.7 GHz methanol emission (Pandian et al. 2007), with distances ranging between ∼1''–35 and ∼105 from our brightest CORNISH radio peak, respectively.

Most water masers align well with our radio source A12−C3 and the overall outflow direction. This, together with the clear separation from our radio peak, again supports the scenario where they are energized by the bipolar outflow (e.g., Hofner & Churchwell 1996; De Buizer et al. 2005). The presence of these water masers also implies a less advanced evolutionary stage than that of its neighbor G45.12 (in agreement with previous studies: e.g., Vig et al. 2006). This is supported by the simpler overall morphology, the lower stellar content, and the much earlier spectral type inferred from the IR/submillimeter for A12−C3 and for the total clump with respect to those derived from the radio, suggesting a much more deeply embedded stage with significant radio optical depth.

While there appears to be extended MIR emission in the GLIMPSE images, consistent with the outflow orientation, there is also considerable emission in an almost orthogonal direction to its main axis (western upper right corner in Figure 10). The peak of the 21 cm radio continuum is displaced toward this region. The overall morphology appears to be similar to the extended H ii in G45.12, albeit in a smaller scale, which could be due to its younger age and/or the lower stellar content in the innermost central region.

Zoom In Zoom Out Reset image size

Our analysis of the SEDs yields total system masses for G45.12 and G45.07 of ∼3550 M_☉ and ∼1400 M_☉, respectively, for a distance of 7 kpc (Table 6). This differs from the estimates given by Hunter et al. (1997), who obtained (with temperature, dust emissivity index, and optical depth as free parameters) masses ∼7 and ∼4 times larger than our estimated masses, respectively (when scaled to their distance of 8.3 kpc). Vig et al. (2006) obtained, through SED fitting and radiative transfer modeling (assuming a spherically symmetric homogeneous cloud of gas and dust and a distance of 6 kpc), masses of 5000 M_☉ (∼20 M_☉ of dust and a larger gas-to-dust ratio of 250) and 13,000 M_☉ (for a dust mass of 30 M_☉ and gas-to-dust ratio of 450) for G45.12 and G45.07, respectively. Scaling to the same distance, we obtain M_d ∼ 26 and M_d ∼ 10.5 M_☉. While our dust mass estimate for G45.12 is in good agreement with their estimate from the SCUBA maps, the mass for G45.07 is in better agreement with the mass these authors estimated from their 130 μm map (obtained with the TIFR 1 m balloon-borne telescope; Ghosh et al. 1988) of ∼13 M_☉.

We find temperatures for G45.12 and G45.07 of ∼40 K, and clump luminosities equivalent to an ∼O6 ZAMS spectral type star (∼O5 class V star). This is in good agreement with those estimated from the IR (e.g., Kraemer et al. 2003) for the brightest sources within the clumps. This suggests that A16−C24 and A12−C3 are the main sources responsible for the dust heating in G45.12 and G45.07, respectively. We note, however, that the significantly later spectral type derived from the radio for A12−C3 appears to suggest a much younger and highly embedded stage than A16−C24.

A29 (G45.47+0.05) is the most prominent of all submillimeter peaks in this complex, and yet lacks significant emission in the MIR (e.g., Figures 12 and 13). The submillimeter peak is located in the neighborhood of 14 GR sources (five of which are associated with bright IR sources; Figure 12). All these sources are located within ∼30'' of the BLAST peak. The appearance in the MIR may suggest that the submillimeter emission is tracing a particularly dense part of the molecular clump. This material could be confining the expansion of the photodissociation region (PDR) as seen in IRAC, which also appears to be in a very young and embedded stage of star formation. Pillar-like structures seen in extinction and extending from A29 into G45.45 (Figure 12), and the main five IR detections embedded within the BLAST source (as suggested by their flux increasing with increasing wavelength) support this scenario.

The brightest CORNISH source in the neighborhood of the submillimeter peak, A16−C0 (Figure 14), is a known UCHiiR (WC89; Testi et al. 1999 center agreement <1''). The flux from our radio analysis is equivalent to an O9.5 radio ZAMS star at 7 kpc, which agrees with the spectral type we derived using the 6 cm flux from Urquhart et al. (2009). This source is also one of the most complex of the clump, and contains H₂O (e.g., Forster & Caswell 1989), OH (e.g., Forster & Caswell 1989; Argon et al. 2000), and methanol (e.g., Menten 1991) masers. The most prominent OH sources are located closer (<05) to the radio peak than the methanol and H₂O masers (∼1''), and they all lie preferentially toward the northern (upper left direction; Figure 14) rim of the radio emission and toward the extended MIR region.

Zoom In Zoom Out Reset image size

In the IRAC maps, the GR source (G045.4661+00.0457; G3 in Figure 14) is at the peak of the IR emission, at the opposite corner of the MIR elongation to the CORNISH source (at ∼3''). Other GR sources are also present nearby the radio peak. Such astrometry differences between the radio and IR detections have been previously observed (De Buizer et al. 2005), which confirms this to be a true physical separation. The MIR elongation may be tracing the outflow morphology detected by Hunter et al. (1997) and the NH₃ emission (e.g., Hofner et al. 1999). Most maser sources are indeed located between the MIR peak (G3) and the radio peak, along the elongation, with a few OH masers located between the radio source and G2 (Figure 14). At the northern extreme edge of the elongation and closer to the main MIR peak, there are more detections of methanol maser signatures (Kurtz et al. 2004), as well as SiO and NH₃ emission peaks (∼10''–15'' from the radio peak; Hofner et al. 1999). The presence of shocked gas along the northern MIR emission, the outflow, and the location of the UCHiiR opposite the main MIR peak may suggest the scenario where the outflow/jet originating at the UCHiiR interacts with its local environment, producing the observed structure in the MIR. An alternative interpretation claims this elongation may arise from accretion (De Buizer et al. 2005) or collapse (e.g., Cesaroni et al. 1992; Hofner et al. 1999) onto the UCHiiR. This could explain the emission and redshifted absorption observed in NH₃. More recent studies have not confirmed the presence of infall signatures (e.g., Wilner et al. 1996; Klaassen & Wilson 2007). Further analysis of the innermost regions is needed to solve this controversy.

The maser emission (especially methanol and H₂O) suggests that this source is one of the youngest in this complex. It has also been classified as an Extended Green Object (EGO) and a likely MYSO outflow candidate (Cyganowski et al. 2008). This supports the HCO analysis of Wilner et al. (1996), who suggested that this source is in the early stages of forming an OB star cluster.

A28 is close to the most active region in the cloud as observed in the MIR (G45.45+0.06 [G45.45]), although from its position and extended morphology the submillimeter emission may also be tracing the compressed and extended molecular gas. G45.45 (also a MYSO candidate in the RMS Survey; e.g., Mottram et al. 2010) has been observed to lie in the border of a larger and fainter H ii region ∼3' in size (G45L; Paron et al. 2009). Although this structure is also traced by the BLAST images, it has not been included in our photometry/radio analysis.

The IRAC images show a complex morphology, with the strongest emission in the MIR and 21 cm coincident with the main nebula-like radio structure observed in the CORNISH data (e.g., Figures 12 and 13). In Figure 13, the MIR emission appears to follow a horseshoe shape around a cavity-like structure (∼1 pc in size at 7 kpc), within which no stars are detected in the IRAC images. Although this structure can be observed at 1.3 cm in the analysis of Mooney et al. (1995) at ∼12'' resolution, the details and morphology of the dust in this region remain unresolved in our BLAST maps.

In the radio, the CORNISH data reveal a similarly complex substructure for G45.45 (Figure 15). There are 18 radio peaks detected within and surrounding the radio "nebula," coincident with the brightest area in the GLIMPSE images. Within this extended radio emission, called "The Orion nebula's younger brother" by Feldt et al. (1998), numerous emission peaks are noticeable. While we cannot easily distinguish or fit these peaks, we detect at least eight within an estimated size of 0.35 × 0.2 pc at 7 kpc (Figure 16). This supports previous studies that classify this source as a cluster of H ii regions (e.g., Garay et al. 1993; Testi et al. 1999). Aperture photometry on the whole structure (including extended emission) yields a radio flux density comparable to that of an ∼O5.5 ZAMS spectral type (∼O5 class V) star, which is in agreement with previous estimates (e.g., Garay et al. 1993). Using the 6 cm flux from Urquhart et al. (2009) for a source major axis of ∼8'', we obtain an estimate of ∼O6.5. This is comparable to the results we obtain using the flux estimates from Testi et al. (1999; Table 4).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Regarding the stellar content, Feldt et al. (1998) detected up to 14 near-infrared (NIR) sources (from a to o) possibly associated with this complex. They also detected additional MIR sources outside the main radio nebula (Figure 16). The roman numbers indicate other point sources in their NIR images outside their VLA map. We confirm their observations that the majority of these objects lie on the sharp and distorted upper edge of the radio emission (Figure 16). This may support the scenario proposed by these authors where a central cluster of OB stars, deeply embedded within the radio complex, produces a front that has triggered further star formation. In their analysis, they claim that IR sources l, m, n, o (Figure 16) may be the source of the UCHiiR. Although these objects do lie close to two bright radio peaks within the cluster (A28−C13, A28−C11; Table 5), their source b is the closest to the brightest peak in the radio (A28−C15 in Figure 16). This CORNISH source lies at less than ∼1'' from both the WC89 coordinates and source b.

At a resolution of 04, WC89 defined this source as a cometary UCHiiR. Indeed, two parallel extensions appear to emerge from the central source A28−C15. These structures form a tail-like morphology extending toward the right in Figure 16, within which A28−C13, A28−C14, and A28−C11 are detected. In total, we find that eight of these IR sources have CORNISH counterparts within ∼1''. Sometimes one radio peak is found between several in the NIR (Table 5 and Figure 16).

Blum & McGregor (2008) classified the main radio structure as a group of MYSOs with late and early B stars surrounding the central source, probably similar to the central cluster powering G45.12 within A16. The offset between the NIR and radio peaks could be explained if the radio detections are dense clumps ionized by the surrounding massive stars (Feldt et al. 1998). Alternatively, we do not discard the possibility that the NIR stars may have been produced by the central and embedded star/cluster, whose expanding H ii region could have triggered the formation of these sources by compression of the dense material traced in the submillimeter. This last scenario is supported by the overall morphology in the CORNISH images; the observed emission supports A28−C15, the brightest radio source, as being responsible for the origin of l, m, n, o, rather than the latter sources being the actual "ground zero" for the triggering mechanism (as suggested by Feldt et al. 1998). Further investigation at high resolution is required to corroborate the triggering hypothesis and to fully distinguish between the different cases.

The candidate counterpart to MIR1 (A28−C10) is clearly visible in the CORNISH data, albeit relatively weak compared to the most central sources. This weak emission is in contrast with the strong IR luminosity of this source. In combination with the nearby OH maser (Argon et al. 2000), this may again suggest that at least some structures are actual stars still embedded within their natal cocoons (e.g., Kraemer et al. 2003).

The remaining possible radio detections within ∼1' and not lying within the main complex have been included in Table 4. Due to the complexity of the radio structures around this complex, we do not discard the presence of additional sources in this region (e.g., Figure 15).

AA30 (G45.48; north of AA29 in Figure 11) also shows complex structure in the submillimeter and IR. Our two submillimeter peaks lie at opposite sides of the main emission visible in the IRAC images. As in previous cases, they may be tracing relatively cold, dense gas that is confining the expansion of the H ii region.

WC89 classified this UCHiiR as irregular, but suggested that the elongated form could be due to an imaging artifact. Testi et al. (1999) identified no compact sources, only extended emission. The elongation of this source, as well as several embedded peaks, are evident at 4.8 GHz (Figure 17), and the structure is identified in the MIR images as the equatorial ridge of the bipolar-like MIR structure (Figure 18).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Just like in G45.45, the MIR and main radio emission appear to wrap around a cavity-shaped structure, within which no IRAC sources are visible. The BLAST emission includes this "cavity," but any such structure would be unresolved in our images. The adjacent radio/IR ridge could be compatible with illumination from the south, although the main source powering the region could in fact be located more toward the north, closer to source A30−C0 (Figure 18), the northern radio source detected by Garay et al. (1993), and the diffuse emission in the IR (KJK1) of Kraemer et al. (2003). Only one GC was found within ∼5'' of A30−C0 (G045.4790+00.1365). KJK1 does appear to be coincident with an IRAC emission peak, even though there is no counterpart in the GLIMPSE I catalogs. The irregular radio structures, and the lack of a significant radio counterpart for KJK1, could imply the presence of an embedded star/cluster around this position, which also coincides with an opacity peak and is estimated to be much colder than the southern counterpart (Kraemer et al. 2003).

The multi-peaked radio morphology of the elongated radio structure (coincident with the southernmost source detected by Garay et al. 1993 and the IR source KJK2 from Kraemer et al. 2003) clearly indicates the presence of embedded sources. Our Gaussian routine detects at least four main peaks in a filamentary structure about ∼0.2 pc in length (Figure 17). Aperture photometry on the entire filament yields a total radio flux equivalent to an O7.5 radio ZAMS (O8 class V) star. This is significantly later than the IR estimate (Table 4), perhaps an indication of self-absorption and a young embedded stage of evolution, despite our earlier spectral type estimate obtained using the 3.6 cm flux from Testi et al. (1999). This structure is also coincident with methanol masers (Menten 1991), recognized as a signature of highly embedded stages of massive star formation. In addition, we observe (weaker) non-Gaussian, filamentary structures to the right of the main structure in Figure 17, also likely associated with methanol emission. A prominent cluster of methanol, OH, and water masers has been detected about ∼15''–20'' southwest from A30−C0 and A30−C1. The presence of a GC/GR counterpart (G045.4725+00.1335) for this source and the lack of an equivalent CORNISH detection again argues in favor of the young stage of the stellar population in this region.

Away from the main complex, A35 and A36 (Figure 19) also show some signatures of stellar activity, as suggested by their morphology.

Zoom In Zoom Out Reset image size

Although our BLAST peaks generally overlap well with the emission observed in IRAC, IRAS 100 μm shows poor correlation. Only at 60 μm do we observe a reasonable match with A36, the only one of these two sources with a PSC IRAS match within ∼30''. This IR source, IRAS 19124+1106, was estimated to be at a distance of 6.5 kpc, and has been suggested to be one of the impact sites of the jets emanating from the microquasar and superluminal source GRS 1915+105 (Kaiser et al. 2004). The other impact site was identified as our BLAST source A19, a candidate MYSO from the RMS Survey (Mottram et al. 2010). GRS 1915+105 is believed to be a binary system consisting of an early-type K giant, accreting via a Roche lobe overflow onto a black hole (e.g., Greiner 2001). The MIR structure of A36, if associated with this IRAS source, does not resemble what one would expect to see if caused by a jet originating at the position of GRS 1915+105. Furthermore, the new distance estimate for GRS 1915+105 of 11 kpc (Zdziarski et al. 2005) is now much larger than the kinematic distance to this IR/submillimeter emission.

Our analysis of the CORNISH images reveals that, with the exception of some weak structures near the emission peak of A36 (likely tracing the ionized region as observed in the VGPS 21 cm maps), there are no major radio sources near these objects. This may suggest very early or very late/isolated low-mass star-forming activity.

Estimated BLAST parameters for the main objects in this complex have been included in Table 6. The temperatures are warm, characteristic of active star-forming clumps with ongoing stellar activity. The total mass of AA29 is ∼4800 M_☉ for a temperature of ∼35 K, which contrasts with the mass estimate of 26,000 M_☉ obtained by Mooney et al. (1995) for a distance of 8.1 kpc and a much lower fixed dust temperature of 20 K. Scaling to this distance we obtain a total mass of ∼6300 M_☉, and ∼34,000 M_☉ for a fixed temperature of 20 K and free β (best-fit β = 2.2). This shows the importance of actually determining the temperature using the BLAST observations.

Our estimated IR+submillimeter single-star spectral type for AA29 of ∼O5.5 is in excellent agreement with the radio spectral type of the radio nebula within G45.45 (Complex 1 in Table 4), the most luminous system within the BLAST emission formed by A28 and A29 (AA29).

The IR+submillimeter spectral type of A30+A30a is in agreement with the IR spectral type estimate of Complex 2 (the elongated radio emission; Table 4), but the CORNISH radio spectral type of this substructure is considerably later.

Although the radio flux estimates from Testi et al. (1999) at 3.6 cm and the uncertainties in the SED bolometric luminosities could be compatible with a relatively optically thin state for this source, the CORNISH measurements suggest a more embedded (and younger) stage for AA30 than for AA29. Just like the clumps within GRSMC G045.14+00.14, the BLAST sources in GRSMC G045.49+00.04 appear to be powered by the central OB young stellar clusters, which could also be responsible for the additional star formation detected within their parent clumps.

In the submillimeter, A50 (IRAS 19145+1116) is largely structureless and symmetric, possibly with some weak emission surrounding the central peak. In the MIR, this H ii region reveals a clumpy structure and numerous MIR peaks clustered within the main submillimeter emission. There are seven GC sources found at less than ∼20'' from our BLAST peak. Among the closest of these we find a string of three sources, G045.9352−00.4017 (the central one), G045.9362−00.4017 (the brightest), and G045.9344−00.4016, which constitute the most prominent structure in the GLIMPSE images.

Overall, the extended structure in the MIR and at 21 cm is rather circular, although emission arcs are visible within the region in the MIR. We detect only one weak 4.8 GHz source, found at less than ∼3'' from the GLIMPSE source G045.9344−00.4016, with a flux corresponding to a B0.5−B0 ZAMS star. The IRAS clump was reported to satisfy the color criteria of WC89 for UCHiiRs (e.g., Bronfman et al. 1996), and was reported by Codella et al. (1995) to be possibly associated with water masers. This would argue against the more evolved state of the system indicated by its "disturbed" morphology, although from the coordinates provided by these authors we suspect that this maser emission could also be associated with a later generation of star formation within an older clump. The lack of a CORNISH or IR counterpart for this maser may also be indicating that the structure is due to a low-mass star.

While relatively structureless in the submillimeter, the GLIMPSE images of A46 (IRAS 19141+1110) show a bow-shock cometary-like structure, with the front oriented toward the south (lower right of Figure 20). Some possibly associated sources toward the south and southeast of the main structure are also detected. Despite the presence of numerous GC/GR sources, the bow shock itself may be produced by a highly embedded source, closest to G045.8062−00.3518 (Figure 20).

Even though A46 also has IRAS, MSX, and IRAC counterparts, there appears to be negligible 21 cm emission and no NVSS, 4.8 GHz emission, H₂O or OH masers, or any other tracers that may be indicative of ongoing star formation. An exception is the methanol maser emission ∼25'' eastward (lower left) from the BLAST peak detected by Pandian et al. (2007). This is the 5₁ − 6₀A⁺ line of methanol at 6.7 GHz, the strongest of class II methanol masers, seen exclusively near massive stars. The authors suggest a very early stage of evolution, in which methanol activity has just turned on, before the onset of additional signs of star formation. Class II maser sources are radiatively pumped but are believed to be quenched by the formation of an UCHiiR, which is consistent with the absence of 4.8 GHz CORNISH sources and the "young" age of the system (e.g., Minier et al. 2005; Ellingsen et al. 2007). This scenario would also be in agreement with Cyganowski et al. (2008), who classified this source as a "possible" MYSO. The observed IRAC structures and methanol emission are all within the prominent BLAST submillimeter emission, which also suggests the presence of local high density molecular material.

A47 (IRAS 19139+1113) is a prominent MYSO (e.g., Chan et al. 1996) showing clear signatures of active star formation. A prominent carved structure (∼2 pc in size at 7 kpc) is observed toward the north (Figure 20). This H ii region is the most extended feature of this cloud, and is detectable in the IR, submillimeter, millimeter, and 21 cm images. The submillimeter and radio peaks appear to trace a region of diffuse emission within this structure, close to three bright IRAC sources (G045.8263−00.2820, G045.8212−00.2846, and G045.8291−00.2858). The location of these IR objects, close to the center of curvature of some of the ionization fronts forming the edge of the ionized region in the molecular cloud, suggests that the combined contribution from these objects results in the sinusoidal boundary of the overall region.

We find three likely sources at 4.8 GHz within ∼1' of our BLAST source, all with GC counterparts within ∼5'', and indicative of ongoing massive star formation. The extended morphology and the advanced state of stellar activity within the main H ii would support the "older" age suggested by the lack of maser emission (e.g., van der Walt et al. 1995; Szymczak et al. 2000).

From their measured fluxes, we estimate ZAMS spectral types for the radio peaks in these clumps of later than ∼B0 stars. Our SED fits show that, overall, none of the host clumps have an equivalent single-star ZAMS spectral type earlier than an ∼O8 star. In the case of A47, it is possible that additional sources are contributing to the total luminosity, in addition to those traced in the radio.

The detection of 4.8 GHz emission confirms previous studies that classified A47 and A50 as UCHiiRs from the WC89 color criteria (Bronfman et al. 1996). Although the radio peaks are not coincident with the brightest IRAC sources, they clearly indicate that massive stars are indeed being formed in this cloud, albeit at a lower level than that observed in the two main clouds described above.