NEW YOUNG STAR CANDIDATES IN CG4 AND Sa101

We used the IRAC data for CG4 from program 462, AORKEY⁹ 24250880; for Sa101, we used the IRAC data from program 20714, AORKEY 16805888. The IRAC data from program 462 (for CG4) were taken on 2007 December 29 with 12 s high-dynamic-range (HDR) frames, so there are two exposures at each pointing, 0.6 and 12 s, with three dithers per position, for a total integration time of 36 s (on average). The IRAC data from program 20714 (for Sa101) were taken on 2006 March 27 with 30 s HDR frames; there are also two exposures per pointing, but deeper, 1.2 and 30 s. For this observation, there are two dithers per position, for a total integration time of 60 s (on average). Because of the different integration times, we reduced the Sa101 and CG4 observations independently even though they overlap on the sky (in Figures 2–5, the jagged-edged observation is from the observation in program 462, and the smooth-edged observation on the right is from the observation in program 20714).

We note that there are additional IRAC data in this region that we did not use. IRAC data from program 202 were of a very small region centered on the head of the globule. We did not include these data in an effort to make our survey as uniform as possible over the entire surveyed region. IRAC data from program 20714 for CG4 were taken in non-HDR mode, with 30 s exposures; as a result of some very bright stars in the field of view, the instrumental effects rendered these data very difficult to work with. Our science goals near CG4 can be met with the total integration time from program 462 alone. In addition, the data from program 462 cover a larger area (∼08 × ∼1°) than from program 20714 (∼04 × ∼05). For these reasons, we did not incorporate the CG4 IRAC data from program 20714 in this analysis.

We started with the corrected basic calibrated data (CBCD) processed using SSC pipeline version 18.7. Because of the very bright stars in the field of view, and because the data from program 462 were taken with cluster targets, we could not use the pipeline-processed mosaics. Moreover, the artifact correction, which is normally done for individual cluster targets separately in the SSC pipeline processing, is much improved when using the CBCD files from program 462 all at once. We reprocessed the IRAC data from both program 462 (for CG4) and program 20714 (for Sa101), using MOPEX (Makovoz & Marleau 2005) to calculate overlap corrections and create mosaics with very much reduced instrumental artifacts compared to the pipeline mosaics. The pixel size for our mosaics was the same as the pipeline mosaics, 0.6 arcsec, half of the native pixel scale. We created separate mosaics for the long and the short exposures at each channel for photometric analysis. For display purposes, we further used MOPEX to combine the two long-frame observations into one large mosaic per channel, as seen in Figures 2–5. The component mosaics were properly weighted in terms of signal-to-noise and exposure time. The total area covered by at least one IRAC channel (as seen in Figures 2–5) is ∼0.5 deg².

To obtain photometry of sources in this region, we used the APEX-1frame module from MOPEX to perform source detection on the resultant long and short mosaics for each observation separately. We took those source lists and used the aper.pro routine in IDL to perform aperture photometry on the mosaics with an aperture of 3 native pixels (6 resampled pixels), and an annulus of 3–7 native pixels (6–14 resampled pixels). The corresponding aperture corrections are, for the four IRAC channels, 1.124, 1.127, 1.143, and 1.234, respectively, as listed in the IRAC Instrument Handbook. As a check on the photometry, the educators and students associated with this project (see the acknowledgments) used the Aperture Photometry Tool (Laher et al. 2011a, 2011b) to confirm by hand the measurements for all the targets of interest. To convert the flux densities to magnitudes, we used the zero points as provided in the IRAC Instrument Handbook: 280.9, 179.7, 115.0, and 64.13 Jy, respectively, for the four channels. (No array-dependent color corrections nor regular color corrections were applied.) We took the errors as produced by IDL to be the best possible internal error estimates; to compare to flux densities from other sources, we took a flat error estimate of 5% added in quadrature.

To obtain one source list per channel per observation, we then merged the short and the long exposures for each channel separately, and for each observation independently, because of the different exposure times as noted above. The crossover points between taking fluxes from the short and long exposures were taken from empirical studies of prior star-forming regions and were magnitudes of 9.5, 9.0, 8.0, and 7.0 for the four IRAC channels, respectively. We performed this merging via a strict by-position search, looking for the closest match within 1''. This maximum radius for matching was determined via experience with other star-forming regions (e.g., Rebull et al. 2010). The limiting magnitudes of these final source lists are the same for both observations and are [3.6] ∼ 17 mag, [4.5] ∼ 17 mag, [5.8] ∼ 15.5 mag, and [8] ∼ 14.5 mag.

There are two MIPS AORs in the CG4 region and two MIPS AORs in the Sa101 region, all four of which were obtained as part of program 20714 (AORKEYs 16805632, 16807936, 16806144, and 16808192) on 2006 May 8 or June 12; see Figures 1, 6, and 7. The AORs were obtained in MIPS photometry mode, nominally centered on 7:31:18.7, −46:57:45 for Sa101 and 7:33:48 −46:49:59.9 for CG4. One of each pair of AORs is explicitly a MIPS-24 photometry observation, and the other is a MIPS-70 photometry observation. During the 70 μm observation, the 24 μm array is still turned on and is still collecting valid data. We combined the prime 24 μm data from the MIPS-24 photometry-mode observations with the serendipitous 24 μm data from the MIPS-70 photometry-mode observations to obtain larger maps at 24 μm. The original explicitly 24 μm photometry observations are small photometry-mode maps, with 3 s integration per pointing, but a net integration time of ∼210 s over most of the resultant mosaic. The serendipitously obtained data averaged ∼350 s net integration time; where the deliberate and serendipitous data overlapped, integration times can be ∼450 s. The 70 μm photometry observations were also small photometry-mode maps, with 10 s integration per pointing; the net integration time over most of the mosaics was ∼400 s.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The data for 24 μm, like those for IRAC, were affected by the bright objects and required additional processing beyond what the online pipelines could provide. We started with S18.13 enhanced BCDs from the pipeline. We implemented a self-flat for each AOR separately, as described in the MIPS Instrument Handbook, available from the SSC or IRSA SHA Web sites. For each pair of overlapping 24 μm maps, we then ran an overlap correction using the overlap script that comes with MOPEX and then created one mosaic for CG4 and one for Sa101, again using MOPEX. Our mosaics had the same pixel size as the online mosaics, 245. In order to combine the images into one mosaic for display in Figure 6, the different overall background levels between the two observations (having an origin in the different zodiacal light levels at the times of the two observations) were problematic. The brighter of the two was artificially lowered via median subtraction to bring its dynamic range into a similar regime as the fainter; this renders photometry on this net mosaic invalid, but the morphology seen in Figure 6 is still valid. The total area covered by the net 24 μm map is only ∼0.3 deg², smaller than that of the IRAC map.

To obtain photometry at 24 μm, we ran APEX-1frame on each of the mosaics (one per observation) and performed point-response-function (PRF) fitting photometry using the SSC-provided PRF. Tests using the apex_qa module portion of MOPEX suggest that our photometry is well within expected errors. For three problematic sources, we used aperture photometry instead of the PRF-fitted photometry, as they provided a better fit in apex_qa. We used the signal-to-noise ratio (S/N) value returned by APEX-1frame as the best estimate of the internal (statistical) errors, adding a 4% flux density error in quadrature as a best estimate of the absolute uncertainty. The limiting magnitude of these observations is [24] ∼ 10.5 mag. Note that we optimized our data reduction to obtain measurements of the brighter sources and sources superimposed on the nebulosity; many sources fainter than this are apparent in the image but not included in our catalog, simply because our scientific goals are aimed at the brighter objects. For one source of interest below (073243.5-464941, which was considered and then rejected as a YSO candidate; see Section 3), an upper limit was obtained at the given position by laying down an aperture as if a source were there, and taking three times that value for the 3σ limit. To convert the flux densities to magnitudes, we used the zero point as found in the MIPS Instrument Handbook, 7.14 Jy.

At 70 μm, there are viable observations from AORKEYs 16807936 and 16808192. We downloaded data processed with pipeline version S18.12. The online pipeline does a very good job of producing mosaics; see Figure 7, where there are a handful of point sources and extended emission visible. The online pipeline produces both filtered and unfiltered mosaics; the filtering preserves the flux densities of the point sources and improves their S/N, especially for faint sources, but destroys the flux density information for the extended emission. The unfiltered mosaics are shown in Figure 7, but we performed photometry on the filtered mosaics. The pipeline mosaics have resampled 4'' pixels (as opposed to 53 native pixels), and the two observations together cover about 0.1 deg². We used APEX-1frame to do PRF fitting on the pipeline filtered mosaics for the point sources, using the SSC-provided PRF. For one problematic source, aperture photometry provided a better flux density estimate. There are only 11 objects with 70 μm detections, and there is a large variation in background levels, so quoting a limiting magnitude is difficult, but is approximately 3 mag. We assumed a conservative, flat 20% flux density error. The zero point we used again came from the MIPS Instrument Handbook, 0.775 Jy.

Where there was 70 μm coverage for the sources of interest, we placed an aperture at the expected location of the source and performed photometry as if there were a source there, taking three times that value as the 3σ limit that appears in Table 2 below.

The optical data will be discussed further in J. S. Kim et al. (2011, in preparation), but we summarize the important aspects of the data reduction here.

The BVR_cI_c photometry of the CG4+Sa101 region was obtained during 2001 March 6, 7, 9, 10, and 11 using the 2 K × 2 K CCD at the 0.9 m telescope at the Cerro Tololo Inter-American Observatory (CTIO). The images have a pixel scale of 04 in a 136 field of view.

Bias and twilight sky flat fields were taken at the beginning and at the end of each night. Long and short (300 s and 30 s) exposures were taken for object fields. During every photometric night we observed Landolt (1992) standard stars of two or three fields several times per night for photometric calibration.

We performed aperture photometry using multiple aperture sizes, and the photometry with highest S/N was chosen as final photometry for each star. For the standard stars, we used an aperture radius of 17 pixels. We used IRAF¹⁰/PHOTCAL routines to solve for the zero point, extinction, and color terms of the standard star solution.

For the target stars in CG4 and Sa101, we used a custom IDL photometry pipeline (written by W. H. Sherry), which was developed for the CTIO 0.9 m telescope. For each target, aperture photometry was performed using multiple size apertures starting from aperture size of 2 pixels to 17 pixels. The highest S/N photometry was chosen as the final photometry.

We used an aperture correction to place our photometry on the same photometry system as our standard stars. The point-spread function (PSF) of the CTIO 0.9 m telescope varies noticeably as a function of location on the CCD. This is insignificant for large apertures, but for aperture sizes of 2–3 pixels, the difference can be a few percent. We accounted for the spatial dependence of the aperture correction in each image by fitting the aperture corrections for stars with photometric errors better than 0.02 mag with a quadratic function. The uncertainty of the aperture correction is about 0.01 mag. For each aperture, the uncertainty on the instrumental magnitude is calculated including the uncertainty on the aperture correction for each aperture.

We used the imwcs program written by D. Mink¹¹ (Mink 1997) to determine the astrometric solution of each image. It fits the pixel coordinates of our targets to the known positions of USNO A2.0 stars located in each field. We then measured positions to an accuracy of ∼03 relative to the USNO A2.0 reference frame.

For each pointing, we used a 1'' matching radius to match sources within the optical BVR_cI_C filters. The coordinates are averaged over different filters, and the typical average uncertainty is 02–03. To find duplicates between adjacent pointings, we again looked for positional matches within 1'' and then took a weighted mean of the available photometry.

Basically the entire IRAC map was covered west of 1138 R.A. (07:35:12); see Figure 1. We have found no YSO candidates east of this. The completeness limits over the field are as follows: B ∼ 19 mag, V ∼ 18 mag, R_c ∼ 17.5 mag, and I_c ∼ 17 mag. However, we note that there are fewer sources per projected area detected (e.g., effectively shallower limits) in the regions where there is molecular cloud material. The zero points we used for conversion of the magnitudes to flux densities (for inclusion in the spectral energy distributions (SEDs) in Figures 14–16) were, respectively, 4000.87, 3597.28, 2746.63, and 2432.84 Jy.

In summary, to bandmerge the available data, we first merged the photometry from all four IRAC channels together with near-IR 2MASS data within each observation (CG4 and Sa101) and then merged together those source lists from each observation. We next included the MIPS data and then the optical data. We now discuss each of these steps in more detail. At the end of this section, we discuss some aggregate statistics of the bandmerged catalog.

To merge the photometry from all four IRAC channels together, we started with a source list from 2MASS. This source list includes JHK_s photometry and limits, with high-quality astrometry. We merged this source list by position to the IRAC-1 source list, using a search radius of 1'', a value empirically determined via experience with other star-forming regions (e.g., Rebull et al. 2010). Objects appearing in the IRAC-1 list but not the JHK_s list were retained as new potential sources. The master catalog was then merged, in succession, to IRAC-2, 3, and 4, again using a matching radius of 1''.

Because we are primarily interested in objects detected by Spitzer, we dropped any objects not having flux densities in at least one Spitzer band (e.g., objects off the edge of the Spitzer maps, having measurements only in 2MASS). Because the source detection algorithm we used can be fooled by instrumental artifacts, we also explicitly dropped objects seen in only one IRAC band as likely artifacts.

For the sources of interest later in the paper, most have counterparts in all three bands in 2MASS by this point in the merging. However, for two sources, this matching failed, at least in part. For one source, 073121.8-465745 (=CG-Ha4), the automatic merging found a counterpart with a measured J magnitude, but the HK_s measurements are flagged with a photometric quality (ph_qual) flag of "E," denoting that the goodness of fit quality was very poor, or that the photometry fit did not converge, or that there were insufficient individual data frames for the measurement. Since there was a good measurement at J, we assumed that there were sufficient frames at HK_s, and that something had happened to the fit. We took the values as reported in the catalog, assumed a large error bar, and used these values in the table and plots below; as will be seen in the SED for this object (Figure 15, top center), these values are probably close to what is most appropriate for this object. For source 073425.3-465409, the automatic merging fails, most likely because the 2MASS counterpart is slightly extended, and it may be slightly extended at 3.6 μm as well. The nearest match in the catalog is ∼2'' away, very large by comparison to other source matches here, but manual inspection strongly suggests that this is the appropriate counterpart to the source seen at Spitzer bands. See Appendix A.21 for more on this interesting source.

We then compared the source lists from the separate observations, CG4 and Sa101, again using a matching radius of 1''. For objects detected in more than one mosaic, we took a weighted average of the flux density at the corresponding band.

Next, we merged the 70 μm source list to the 24 μm source list. The 70 μm PSF is large compared to the positional accuracy needed, and astrophysically, each 70 μm source ought to have a counterpart at 24 μm, given the sensitivity of these observations. We individually verified that each of the 70 μm point sources had a 24 μm counterpart, and then merged these two source lists. To successfully have the computer match the sources that were clearly matched by eye, a positional accuracy of 25 was required, consistent with our experience in other star-forming regions (e.g., Rebull et al. 2010). Since the two MIPS observations do not overlap with each other, no explicit merging of the MIPS source lists from the two observations was required beyond simple concatenation.

To combine the merged MIPS source list into the merged 2MASS+IRAC catalog, we used a positional source match radius of 2'', again determined via experience with other star-forming regions (e.g., Rebull et al. 2010). The MOPEX source detection algorithm can be fooled by structure in the nebulosity in the image, and by inspection, this was the case for these data. To weed out these false "sources," we then dropped objects from the catalog that were detections only at 24 μm and no other bands. Finally, to merge the J through 70 μm catalog to the optical (BVR_cI_c) catalog, we looked for nearest neighbors within 1''.

After this entire process, there are ∼21,000 sources with IRAC-1 (3.6 μm) or IRAC-2 (4.5 μm) detections, ∼9000 sources with IRAC-3 (5.8 μm) detections, and ∼4000 sources with IRAC-4 (8 μm) detections. About 3000 (∼15%) of the IRAC sources have viable data at all four IRAC bands, nearly all of which have counterparts in 2MASS. The optical data do not cover the entire IRAC map, but about half of the four-band IRAC detections have counterparts in the optical catalog. There are only ∼500 sources at 24 μm in our catalog and just 11 sources at 70 μm; note that the MIPS-24 map covers a much smaller area than the IRAC maps, and the MIPS-70 map is smaller still (see Figures 1–7). Ten of the 11 MIPS-70 sources have counterparts at all four IRAC bands; the one that appears to have no four-band IRAC counterpart is saturated in two of the four IRAC bands. About 200 sources have all four IRAC bands plus MIPS-24.

Optical data can greatly aid in confirming or refuting YSO candidacy because they provide constraints on the Wien side of the SED. In Guieu et al. (2010), most of the IRAC-selected candidates in IC 2118 proved to be too faint, most vividly in the optical, to be likely cluster members. Just 10 of our candidate YSOs have optical data available, and they appear in Figure 11. The objects with optical data that have already been ruled out as YSOs based on their IRAC properties are all well below the Siess et al. (2000) 30 Myr isochrone scaled to 500 pc. One YSO candidate object appears below the 30 Myr isochrone; 073337.6-464246 is within the distribution of clear non-YSO points. We do not remove this object from our list, since a variety of reasons (such as scattered light) could result in a YSO appearing below the 30 Myr isochrone; for more discussion of this object, please see Appendix A.19. As noted above, the distance to this association is uncertain; if the isochrones are instead scaled to 300 pc, then one other object (073121.8-465745, a previously identified YSO) appears to be just below the 30 Myr isochrone instead of just above it.

Zoom In Zoom Out Reset image size

Deeper optical data are desirable in order to obtain magnitude estimates for the remaining YSO candidates.

Near-IR data can also aid in confirming or refuting YSO candidacy. Since we do not have spectral types for most of our sources, it is difficult to estimate the degree of reddening. Figure 12 shows J − H versus H − K_s for the sample. This plot suggests that most of our YSO candidates have an infrared excess with a moderate degree of reddening.

Zoom In Zoom Out Reset image size

Young stars that are actively accreting from their circumstellar disk can have excess ultraviolet emission in the U or B bands, or even longer bands during periods of intense accretion. However, these bands are also most sensitive to reddening. Figure 13 shows B − V versus R_c − I_c for the sample, with the main sequence indicated as a solid line. Stars with a B-band excess and relatively small values of A_V would be blue, e.g., below the line in this figure. Objects above the line on the upper right of this figure are pushed into that location by high A_V. At least four of the YSO candidates have a B excess, most likely from mass accretion; four more appear to have been pushed from the region of clear B excess by high A_V. Similar results for the same objects are obtained from the B − V versus V − I_c plot. The individual objects are listed in the Appendix.

Zoom In Zoom Out Reset image size

Coordinates and our measured magnitudes between B and 70 μm for our 22 YSO candidates appear in Table 2. Six of them (27%) are rediscoveries of the previously known YSOs in this region from Table 1. Figures 14–16 are the SEDs for the YSO candidates.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

To guide the eye, we wished to add reddened stellar models to the SEDs, but spectral types are known only for the six previously known YSOs. In order to provide a reference for the remaining objects, we assumed a spectral type of M0. For each object, a reddened model is shown, selected from the Kurucz–Lejeune model grid (Lejeune et al. 1997, 1998) and normalized to the K_s band where possible (and to the closest band otherwise). Note that this is not meant to be a robust fit to the object, but rather a representative stellar SED to guide the eye such that the infrared excesses are immediately apparent. In some cases, ultraviolet excesses may also be present. Additional spectroscopic observations are needed to better constrain these fits.

In the spirit of Wilking et al. (2001), we define the near- to mid-IR slope of the SED, α = dlog λF_λ/dlog λ, where α > 0.3 for a Class I, 0.3 to −0.3 for a flat-spectrum source, −0.3 to −1.6 for a Class II, and <−1.6 for a Class III. For each of the YSO candidate objects in our sample, we performed a simple ordinary least-squares linear fit to all available photometry (just detections, not including upper or lower limits) between 2 and 24 μm, inclusive. Note that errors on the infrared points are so small as to not affect the fitted SED slope. The precise definition of α can vary, resulting in different classifications for certain objects. Classification via this method is provided specifically to enable comparison within this paper via internally consistent means. Note that the formal classification puts no lower limit on the colors of Class III objects (thereby including those with SEDs resembling bare stellar photospheres, and allowing for other criteria to define youth). By searching for IR excesses, we are incomplete in our sample of Class III objects. The classes for the YSO (previously known and candidate) sample appear in Table 3. Out of the 22 stars, 16 (73%) are Class II.

Based on the SEDs and location in several color–color and color–magnitude diagrams, we have ranked the YSO candidates loosely into three bins: high likelihood of being YSOs (grade A), mid-grade quality (grade B), and relatively low likelihood of being YSOs (grade C). This grade also appears in Table 3. Most of them (16) are in the grade A bin, which includes the six previously identified ones. Comments on individual objects (including justifications for the grades that were given) appear in the Appendix.

Figure 17 shows the 8 μm mosaic with the positions of the YSO candidates overlaid, color coded by YSO quality. Most of the grade A and B objects are clustered near the previously known YSOs. The Sa101 region has a relatively tight clumping of most (16) of the YSOs, with a median nearest neighbor distance of ∼62''. The CG4 region has six YSOs, much less tightly clumped, with a median nearest neighbor distance nearly five times larger, ∼301''. Clustering is also very commonly found among young stars, so especially in the case of the Sa101 association, the fact that they are clustered also bolsters the case that they are legitimate YSOs.

Zoom In Zoom Out Reset image size

Because the distance to this association is uncertain, we looked at whether the relative placement of stars in the optical CMD could be used to constrain the distance to the stars. Figure 18 presents the M_V versus V − I CMD, comparing CG4+Sa101 stars to YSOs in Taurus (with data from Rebull et al. 2010; Güdel et al. 2007 and references therein). Based on morphological grounds (e.g., the degree to which the YSOs are still embedded in their natal gas), we expect that the CG4+Sa101 stars might be slightly younger than the often more physically dispersed Taurus stars. On the other hand, based on the ratio of Class I to Class II sources, the CG4+Sa101 objects might be slightly older than Taurus. In Figure 18, there are not many CG4+Sa101 objects, and the distribution is broad, but assuming a distance of 500 pc, then CG4+Sa101 appears to be quite comparable in age to Taurus at ∼3 Myr. Assuming a distance of 300 pc, the CG4+Sa101 stars are on the whole older than the Taurus stars. Figure 18 thus weakly suggests that CG4+Sa101 is farther rather than closer.

Zoom In Zoom Out Reset image size

This object, 073049.1−470209, the westernmost in the set, is also the reddest source in the set. It is the YSO candidate farthest in the upper right of the IRAC color–color diagram (Figure 8) and is the reddest and faintest YSO in [3.6] versus [3.6]−[24] (Figure 9). Based on its SED (Figure 14), it is not surprising that there are no optical or even 2MASS counterparts; it very steeply falls from 24 μm back to 3.6 μm, and it seems that very deep integrations would be needed to obtain measurements of this object at shorter wavelength bands. It is not detected at 70 μm, but it is in a high background region, so the 70 μm limit is not very constraining. This type of SED can be found in very young, very embedded, very low mass objects, but also in extragalactic contaminants. Because of its proximity to the rest of the Sa101 objects, it remains in the list, but we have given it the lowest-quality grade (C). Its steep SED means that it is one of the three Class I objects in our list. Additional follow-up data are needed to determine the true nature of this source.

This object, 073049.8−465806, is relatively faint at 3.6 μm, at least in comparison to the rest of the YSO candidates here. It appears on the top edge of the clump of likely extragalactic sources in the [3.6] versus [3.6]−[24] plot (Figure 9). It is only detected at K_s in 2MASS with limits at J and H. Based on the SED, it looks quite extinguished, e.g., there seems to be high A_V in the direction of this source. It is located in close (projected) proximity to other YSOs and YSO candidates; YSOs are also frequently found near other YSOs. It is not detected at 70 μm, but the limit is shallow enough that it does not provide a strong constraint on the SED. It has an SED characteristic of YSOs, and that plus the apparently high A_V plus its location in the cloud close to other YSOs has yielded a "B" grade. The steep rise of the SED from 4.5 to 24 μm influences the classification slope fitting such that it is one of the three Class I objects in our list. Additional follow-up data are needed.

Object 073053.6−465742 has 2MASS data, but no available optical data, and just a limit at 70 μm. The SED (Figure 14) suggests that there is high A_V toward this source. Given our overly simple modeling, there seems to be a small but significant excess at 8 μm (better modeling is required to confirm this), and there is an apparently large excess at 24 μm. Whenever an apparent excess is seen only at 24 μm, because the MIPS-24 camera has lower spatial resolution than the shorter bands, there is a risk that the flux density measured at 24 μm is contaminated by source confusion, most likely with a nearby background source, but it could also be with a low-mass companion to the young star candidate. There is only a limit at 70 μm to help constrain the SED, though it is located in close (projected) proximity to other YSOs and YSO candidates. In this case, the fact that the excess appears to be significant at at least two bands, its proximity to other nearby Sa101 sources, and the likelihood that there is high A_V toward this source suggests that it may be a legitimate YSO, so we have placed it in the highest-quality source bin (grade "A"), and it is a Class II. Spectroscopy of this target is required to determine whether or not it is a YSO.

CG-Hα2 was identified in Reipurth & Pettersson (1993) as a YSO. It has counterparts at all bands we considered here, including 70 μm. In the optical (Figure 11), it is in the locus of young stars above the 30 Myr isochrone, and it appears there to be the lowest mass object of the YSO candidates with optical data. The spectral type given in Reipurth & Pettersson (1993) is M2:, so that is the model we have used in Figure 14. It is also one of the stars with a clear B excess in Figure 13, which can also be noted in the SED itself. Such an excess is a characteristic of active accretion. Because it is a previously identified YSO, it appears as grade "A+" in Table 3; it is a Class II.

073106.5-465454 is the northernmost YSO candidate within the Sa101 association. It has a 2MASS counterpart, but the SED (Figure 14) seems somewhat "disjointed" between the 2MASS and IRAC portions. This could be from very high extinction, or intrinsic stellar variability between the epochs of observation at the NIR and MIR, or source confusion. It is detected at 24 μm but there is only a limit at 70 μm. We categorize it as a "C"-grade YSO candidate, with a "flat" SED class. Additional data are needed to confirm or refute the YSO status of this object.

This object, 073108.4-470130, has an SED (Figure 14) quite consistent with other known YSOs, although there are no optical data and just a limit at 70 μm. If it is really a young star, it has a significant excess in at least 8 and 24 μm, and probably 5.8 μm as well. We categorize it as a very high quality YSO candidate ("A"), SED Class II.

Like the prior object, 073109.9-465750 has an SED (Figure 14) quite consistent with other known YSOs, again without optical data. There is a limit at 70 μm, and it is in a high background region, so there is little constraint on the SED. Like the prior object (073108.4-470130), if it is really a young star, it has a significant excess in at least two bands. We categorize it as a very high quality YSO candidate ("A"), SED Class II.

CG-Hα3 is another YSO identified in Reipurth & Pettersson (1993) and has some photometry reported there as well. It is detected in all bands considered here, B through 70 μm. In the optical (Figure 11), it is well within the clump of YSO candidates above the 30 Myr isochrone. As seen in its SED (Figure 14), the optical and near-IR data from Reipurth & Pettersson (1993) are quite consistent with the optical and near-IR data we report here, with some weak evidence for variability in the optical. Reipurth & Pettersson (1993) report a type of K7, and that is the model we have used in Figure 14. Like CG-Hα2, the SED suggests that there may be some B-band excess (most likely due to accretion); in Figure 13, it appears as one of the objects apparently pushed above the main sequence due to reddening. Because it is a previously identified YSO, it appears as grade "A+" in Table 3; it is a Class II.

This object can be seen at all four IRAC bands and both of the MIPS bands considered here. It also has a 2MASS counterpart, but no optical data. The SED (Figure 14) suggests that there is relatively high A_V toward this object. Of the new YSO candidates that appear to have some photometric points on the photosphere, this object appears to have the highest A_V. If this object is actually a young star, it has an IR excess in four bands. We categorize this as another highest-quality (grade "A") YSO candidate, with SED class "flat." Additional data are needed, including optical photometry.

Another new YSO candidate, 073114.9-470055, has a 2MASS counterpart but no optical data and just a limit at 70 μm (Figure 15). If it is a legitimate young star, it has an IR excess in more than two bands. We categorize it as a grade "A" YSO candidate, SED Class II. Additional data are needed, including optical photometry.

CG-Hα4 is another YSO from Reipurth & Pettersson (1993) and has some photometry reported there as well as a spectral type of K7–M0. As discussed above, this object has a high-quality J measurement in the 2MASS point-source catalog, but the HK_s measurements were flagged as having low-quality photometry. We accepted the low-quality measurements with a large uncertainty and added them to the SED in Figure 15. Given the normalization as seen there, there could be a disk excess beginning as early as 3.6 μm, but with a better measurement at the near-IR and thus a better constraint on the location of the photosphere, the disk excess could start at a longer wavelength, and we suspect that it probably does. The optical data we report, as compared with the optical data from Reipurth & Pettersson (1993), suggest substantial intrinsic source variability, a common characteristic of young stars. In the optical CMD in Figure 11, it is apparently the oldest YSO of the set of YSO candidates above the 30 Myr isochrone. This object's apparent intrinsic variability could move it around in this diagram; moreover, the uncertain distance to the CG4+Sa101 cloud moves this object above or below the 30 Myr isochrone (as discussed above). This is the only previously known YSO in our set that is not detected at 70 μm. It appears in Figure 13 as having one of the smallest B-band excesses, but the photometry as seen in Figure 15 suggests that perhaps its placement in Figure 13 could be improved. Because it is a previously identified YSO, it appears as grade "A+" in Table 3; it is a Class II.

CG-Hα5 is reported as a K2–K5 in Reipurth & Pettersson (1993). It is detected at all available bands we discuss here. It appears in Figure 13 as having a B-band excess and subject to high A_V. Because it is a previously identified YSO, it appears as grade "A+" in Table 3; it is a Class II.

CG-Hα6, located very close to CG-Hα5, is reported as a K7 in Reipurth & Pettersson (1993). It is detected at all available bands we discuss here, B through 70 μm. The infrared excess appears to start around 8 μm, suggesting a possible inner disk hole; more detailed modeling is needed to be sure. It appears in Figure 13 as having a B-band excess and subject to high A_V. It is another "A+," Class II, in Table 3.

This object, 073143.8-465818, is detected at optical through 24 μm. It is a high-quality YSO candidate (grade "A"), and SED class "flat." In the optical (Figure 11), it is the lowest mass YSO candidate without a prior identification. It has a clear B excess in Figure 13. It is not detected at 70 μm. Spectroscopy of this object is required to confirm/refute its YSO status and obtain an initial guess at its mass.

073144.1-470008 is another high-quality YSO candidate (grade "A," Class II), detected in the optical through 24 μm. It too has a clear B excess in Figure 13. It is undetected at 70 μm.

073145.6-465917 is the last of the YSO candidates in this list associated with Sa101. It is a high-quality (grade "A") YSO candidate, detected at optical through 24 μm, but with only a limit at 70 μm. Based on the approximate SED fit in Figure 15, the disk excess starts at 8 μm, though additional modeling (and a spectral type) is required to be sure. In Figure 13, it appears as having a B-band excess and subject to A_V. It has an SED Class of II.

CG-Hα7 is the only previously known YSO in the CG4 region covered by our maps. Reipurth & Pettersson (1993) report a K5 spectral type; their optical data are quite consistent with the optical data we report. We detect this object at all available bands reported here, B through 70 μm. In Figure 11, it is the highest apparent mass YSO candidate, though reddening can strongly influence this placement. The disk excess (Figure 15) begins at the longer IRAC bands, suggesting a possible inner disk hole. It does not appear to have a B-band excess in Figure 13. It is located far from any nebulosity (Figure 17), as noted by Reipurth & Pettersson (1993). It is an SED Class II.

Object 073337.0-465455 has counterparts at J through 24 μm; there are no optical data available. Its placement in the [3.6] versus [3.6]−[24] diagram (Figure 9) suggests that it is at the edge of our detection limits; it appears at [3.6]−[24] ∼ 3.4, [3.6] ∼ 14. It is undetected at 70 μm. If it is a legitimate YSO, it has an excess in at least three bands. It is located just off the northern edge of the globule. We place it in the "grade B" bin, and it is SED Class II. Additional follow-up data are needed.

This object, 073337.6-464246, is detected at optical through 24 μm, but not at 70 μm. It is the candidate seen in Figure 11 as very low in the optical CMD, and closest to the IRAC photospheric locus of points in Figure 8. There is a marginal excess seen at 5.8 and 8 μm, and then a larger excess at 24 μm; as discussed above, excesses seen only (or primarily) in a single point at 24 μm can be a result of source confusion with adjacent sources. This source is also located very far from any nebulosity, near the edges of some of our IRAC maps (Figure 17). We place this marginal candidate as a grade "C"; it is an SED Class II. Additional data are needed.

073406.9-465805 is the only YSO candidate of our set located projected onto the globule/elephant trunk. It appears to also be projected onto a bright rim at 8 μm (Figure 17), so it is potentially being revealed now by the action of the ionization front. There are no optical or 70 μm counterparts, but there are counterparts at J through 24 μm. If it is a young star, there is an IR excess at more than two bands. We place it as a grade "A," Class II object, and additional data are needed.

073425.3-465409 is a very interesting source. In the [3.6] versus [3.6]−[24] diagram (Figure 9), it is the brightest, reddest object ([3.6] ∼ 11.7, [3.6]−[24] ∼ 8.2), and it is located right on the "lip" of the globule (Figure 17), in a region where YSOs might be expected to form. None of the other "fingers" of the molecular cloud have apparent associations with infrared objects. After the automatic bandmerging described in Section 2 above, this object did not appear to have a 2MASS counterpart. However, in looking at the overall shape and brightness of the SED, we suspected that it should have a match in 2MASS. Examining the images by eye, there is clearly a source at this location visible in 2MASS and the Palomar Observatory Sky Survey (POSS) images. At the POSS bands, it is distinctly fuzzy. It is also slightly resolved at J; it has been identified with 2MASX J07342550-4654106. The position given for this source in the 2MASS point-source catalog is 21 away from the position in the IRAC catalog, and the position from the 2MASS extended source catalog is 198 away from the IRAC position, both of which are very large compared to most of the rest of the catalog. If it is also slightly resolved in IRAC, this could affect the positional uncertainty. Certainly, the structure of the molecular cloud around this source is complex and could also have affected the position in the catalog. In any case, manual inspection of the images ensures that the object is really the same in the two catalogs, so the JHK_s photometry from the extended source catalog was attached to this Spitzer source. The SED (Figure 16) rises steeply at the long wavelengths, but it is unfortunately just off the edge of the 70 μm maps.

Given this object's SED and location, we have classified this as a high-quality (grade "A") YSO candidate, with SED Class I. Additional data are required to confirm or refute this object's status.

073439.9-465548, the last object in our list, has counterparts at J through only 8 μm, and as such, it has a sparse SED. (It is off the edge of both the 24 and 70 μm maps.) It is located just to the east of the end of the globule, consistent with it having been relatively recently uncovered. The shape of the IRAC portion of the SED is a little different than that for other objects in this set; despite the negative overall slope of the IRAC points, none of the IRAC points seem to be photospheric (compare to other SEDs in Figures 14–16), and there is even a very slight negative curvature (the slope between 3.6 and 4.5 μm is slightly shallower than that between 4.5 and 5.8 μm). On the basis of experience looking at many hundreds of SEDs for YSOs and contaminants (Rebull et al. 2010, 2011), we have some reservations about the shape of this SED; without additional photometric data, it is hard to give this object a high grade. We give this one a "C" grade; it is a Class II.