CHASING THE IDENTIFICATION OF ASCA GALACTIC OBJECTS (ChIcAGO): AN X-RAY SURVEY OF UNIDENTIFIED SOURCES IN THE GALACTIC PLANE. I. SOURCE SAMPLE AND INITIAL RESULTS

The main goal of the ChIcAGO survey is to localize the positions of the unidentified AGPS sources, so that multiwavelength follow-up can be used to identify them. It is therefore necessary to design an experiment that will allow each source to be localized precisely enough to identify counterparts in the crowded Galactic plane. Chandra can provide subarcsecond localization as it has an intrinsic astrometric precision accuracy of 06 at 90% confidence within 2' of the aimpoint (Weisskopf et al. 2003).

For all but the brightest targets, Chandra's Advanced CCD Imaging Spectrometer (ACIS; Garmire et al. 2003) was used as it provides simultaneous positional, temporal, and spectroscopic information. The ACIS-S configuration was chosen as it fully encompasses the AGPS positional uncertainties of up to 3' (Sugizaki et al. 2001). (The aimpoint of the ACIS-I configuration is near a chip gap.) The High Resolution Camera (HRC; Murray et al. 2000) in the I focal plane array was used to observe those sources with a predicted ACIS count rate >0.2 counts s⁻¹ to avoid positional, spectral, and temporal degradation associated with pileup (Davis 2001).

At Chandra's high angular resolution, only a small number of X-ray counts are required to localize each source sufficiently to overcome confusion from IR field stars in the Galactic plane. We first considered the number density of such stars in the K_s band at low Galactic latitudes. Figure 24 of Kaplan et al. (2004) shows that 0.2 stars arcsec⁻² are expected with a magnitude K_s ≲ 19. For there to be a <25% chance of random alignment of the Chandra source with an infrared field star of this magnitude, a total astrometric error <07 is required. Using 2MASS as a guide, given its extremely high positional precision (01 1σ error) and Chandra's 90% absolute astrometry error of 06, a centroiding error of <04 with 95% accuracy is required for Chandra. The Chandra Interactive Analysis of Observations (CIAO)¹⁸ software tool wavdetect (Freeman et al. 2002) was chosen to detect the point sources in our fields. Equation (5) of Hong et al. (2005) provides the 95% confidence position error circle of a point source detected with wavdetect for a given number of source counts at a given off-axis angle. At the maximum off-axis angle expected for an ASCA source localization (<3'), ∼100 X-ray counts are required to ensure that a source's centroiding error is below 04.

Using the count rates and power-law spectral fits calculated by Sugizaki et al. (2001) for each AGPS source and the Chandra Proposal Planning Toolkit (PIMMS),¹⁹ the exposure time required to detect ∼100 counts with Chandra for each source was estimated. For those AGPS sources that were too faint for Sugizaki et al. (2001) to calculate spectral fit parameters, an absorbed power-law model with a photon index Γ = 2 and an absorption N_H = 10²² cm⁻² was used; these values are representative of a nonthermal X-ray source and typical Galactic plane absorption.

In order to select the AGPS source candidates to be observed with Chandra, each source was investigated individually. First, those AGPS sources that have already been conclusively identified, either by Sugizaki et al. (2001) or by other groups in the literature, were removed from the Chandra target list. On the basis of this criterion, a total of 43 AGPS sources were identified and therefore rejected for Chandra follow-up (these sources are described in Appendix A). The ASCA images of each of the remaining unidentified AGPS sources were then studied to determine if any sources appeared to be too extended for Chandra to successfully localize in a short amount of time. These sources were also rejected for Chandra follow-up. The remaining unidentified AGPS sources were then prioritized for Chandra follow-up on the basis of their absorbed X-ray flux or count rate that was listed by Sugizaki et al. (2001).

A total of 93 AGPS sources have been observed with Chandra as part of the ChIcAGO survey, of which 84 were imaged with ACIS-S and 9 were imaged with HRC-I. The ChIcAGO Chandra observations took place over a 3.5 yr period, from 2007 January to 2010 July. The Chandra exposure times ranged from ∼1 to 10 ks. All the details of these Chandra observations are listed in Table 1. The initial automated analysis of these Chandra observations was conducted using the ChIcAGO Multi-wavelength Analysis Pipeline, described in Section 2.2. We then performed a more detailed X-ray analysis and counterpart study for those 74 sources with >20 X-ray counts, as such sources are approximately within the original AGPS sources X-ray flux range (see Sections 3.2 and 3.3).

Table 1. ChIcAGO MAP X-Ray Results, Including the Position, Position Errors, Net Counts, and Quantile Analysis Values for Each ChIcAGO Source

Chandra Obsa	ChIcAGOb	wavdetect Position		Offsetc	Positional Errorsd		Net Countse			Energy Quantilesf		CSCg	2XMMi-DR3h
(ID, Date,	Source ID	R.A.	Decl.	from	wav	Total	0.3–8	0.5–2	2–8 keV	E50	Quartile	Name	Name
Inst, Exp)	ChI	(J2000)	(J2000)	ASCA (')	('')	('')				(keV)	Ratio	CXO J	2XMMi
AX J143148–6021 8239 2007-09-05 ACIS-S 2.62ks
	J143148–6021_1	14:31:50.18	−60:22:08.3	0.97	0.6	0.9	$4.0_{-1.9}^{+3.2}$	$4.0_{-1.9}^{+3.2}$	$0.0_{-0.0}^{+1.9}$
	J143148–6021_2	14:31:48.56	−60:21:44.5	0.56	0.5	0.8	$10_{-3}^{+4}$	$7.0_{-2.6}^{+3.8}$	$2.0_{-1.3}^{+2.7}$	1.1 ± 0.4	1.4 ± 0.9
	J143148–6021_3	14:31:48.06	−60:19:59.5	1.19	0.3	0.8	$146_{-12}^{+13}$	$51_{-7}^{+8}$	$95_{-10}^{+11}$	2.7 ± 0.2	1.1 ± 0.1	143148.0–601959
	J143148–6021_4	14:31:41.52	−60:19:55.2	1.56	0.6	0.9	$5.9_{-2.4}^{+3.6}$	$2.9_{-1.6}^{+2.9}$	$0.0_{-0.0}^{+1.9}$	0.6 ± 0.3	0.5 ± 0.4
	J143148–6021_5	14:32:02.80	−60:18:57.4	2.81	0.7	1.0	$20_{-4}^{+6}$	$0.0_{-0.0}^{+1.9}$	$20_{-4}^{+6}$	4.4 ± 0.5	1.8 ± 0.2	143202.8–601857
AX J144042–6001 9599 2008-05-04 HRC-I 1.18ks
	J144042–6001_1	14:40:38.40	−60:01:36.8	0.56	0.3	0.8	$326_{-18}^{+19}$					144038.4–600136

Notes. ^aDetails of the Chandra observation: the observation identification number (ID), the date of the observation (yyyy-mm-dd), the Chandra instrument used for the observation, and the exposure time of the observation in ks. ^bThe ChIcAGO source name. Those ChIcAGO sources marked with a * have a probability of false detection that is >10⁻⁵. The ChIcAGO sources marked with a † fall in the chip gap between ACIS-S CCDs 6 and 7. ^cThe offset, in arcminutes, between the ChIcAGO source and the original AGPS source position published by Sugizaki et al. (2001). ^dThe ChIcAGO source position error information. wav: The 95% position error circle of the ChIcAGO source wavdetect position. This was calculated using Equation (5) of Hong et al. (2005). Total: The total 95% position error circle of the ChIcAGO source taking into account the wavdetect position error and the absolute astrometric accuracy of Chandra. Both errors are in arcseconds. ^eThe net number of counts in the 0.3–8.0, 0.5–2.0 and 2.0–8.0 keV energy ranges detected from each ChIcAGO source. The source counts were calculated using a source region with a radius equal to 95% of the PSF at 1.5 keV. The total counts have been background subtracted and the upper and lower 1σ confidence limits were calculated using Gehrels (1986) statistics. ^fThe values used in the quantile analysis (results in Tables 3 and 4). E₅₀ is the 50% photon detected in the 0.3–8.0 keV energy band. The quartile ratio is the ratio of the E₂₅ (energy of the 25% photon detected) and E₇₅ (energy of the 75% photon detected) used in quantile analysis: 3 ×(E₂₅ − 0.3)/(E₇₅ − 0.3). See Section 2.3.1 for further details. ^gThe corresponding Chandra Source Catalog name for each ChIcAGO source (Evans et al. 2010). ^hThe corresponding 2XMMi DR3 name for each ChIcAGO source (Watson et al. 2009).

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:  DataTypeset image

It is crucial to the efficiency of the project to automate the analysis of the Chandra observations, such as the detection and extraction of sources, as well as the search for multiwavelength counterparts. We therefore created the ChIcAGO Multi-wavelength Analysis Pipeline (MAP) for this task. ChIcAGO MAP takes the ACIS-S or HRC-I Chandra observation of an AGPS source field and detects and analyzes all point sources within 3', equivalent to the largest likely position error, for the original AGPS source positions supplied by Sugizaki et al. (2001). From hereon we refer to all point sources detected in the Chandra observations of the AGPS fields as "ChIcAGO sources." The X-ray analysis component of this pipeline uses the CIAO software, version 4.3, with CALDB, version 4.5.5, and follows standard reduction recipes given in the online CIAO 4.3 science threads.²⁰

ChIcAGO MAP carries out the following steps, all of which are explained in more detail below. These steps apply to both ACIS-S and HRC-I data sets unless otherwise stated.

• 1.  
  An image of the original ASCA detection of the AGPS source is created (for example, see the left image of Figure 1).
• 2.  
  The CIAO tool chandra_repro is run to reprocess the Chandra data.
• 3.  
  The new event file is filtered to only include photons with energies in the range 0.3–8.0 keV.
• 4.  
  The CIAO tool wavdetect is used to detect all X-ray point sources (ChIcAGO sources) within 3' of the AGPS position. A Chandra image is then created for each ChIcAGO source (for example, see the right image of Figure 1). If no sources are detected, ChIcAGO MAP ends.
• 5.  
  The position, source counts, and associated errors are calculated for each ChIcAGO source detected. If the data set is an ACIS-S observation, then the total counts are obtained in the 0.3–8.0, 0.5–2.0, and 2.0–8.0 keV energy ranges, and the energy quartiles (E₂₅, E₅₀, and E₇₅), which are used in quantile analysis (see Section 2.3.1), are calculated.²¹
• 6.  
  The CIAO tool specextract is run on ACIS-S data sets to obtain the source and background spectrum files and their corresponding redistribution matrix file (RMF) and the ancillary response file (ARF) for each ChIcAGO source. These files are used in quantile analysis and spectral modeling (see Section 2.3).
• 7.  
  A timing analysis is conducted on each ChIcAGO source (detected with either the ACIS or HRC instruments) to search for short-term variability and periodicity.

  • (a)  
    A light curve with eight bins is constructed using the CIAO tool dmextract. The χ² is calculated for this light curve in order to test for short-term variability (e.g., Figure 2).
  • (b)  
    The $Z^{2}_{1}$ statistic is calculated to search for sinusoidal periodicity (Buccheri et al. 1983). This process creates a power spectrum and folded light curve (e.g., Figure 3), which predicts the most likely pulsed frequency, its corresponding power, and the probability that the power is random noise.
• 8.  
  Multiwavelength follow-up and catalog searches are conducted to identify likely optical, infrared, and radio counterparts to each ChIcAGO source.

  • (a)  
    The U.S. Naval Observatory B Catalog, version 1.0 (USNO B1; Monet et al. 2003), Two Micron All Sky Survey (2MASS) Point Source Catalog (PSC; Skrutskie et al. 2006), and Galactic Legacy Infrared Mid-Plane Survey Extraordinaire (GLIMPSE) I and II catalogs (Benjamin et al. 2003) are accessed to obtain a list of all optical and infrared sources within 4'' of the ChIcAGO source's wavdetect position.
  • (b)  
    Small-sized (6' × 6') image cutouts, centered on the ChIcAGO source's wavdetect position, are obtained from the 2nd Digitized Sky Survey (Red: DSS2R, and Blue: DSS2B; McLean et al. 2000), 2MASS, and the GLIMPSE I and II surveys. A 30' by 30' radio image cutout is obtained from the Sydney University Molonglo Sky Survey (SUMSS; Bock et al. 1999). Examples of all the image cutouts can be found in Figures 4 and 5.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

ChIcAGO MAP first generates an image of the AGPS source as originally detected by the gas imaging spectrometer (GIS; Ohashi et al. 1996) on board ASCA. The GIS has a circular field of view with a 50' diameter. The left image of Figure 1 shows the ASCA GIS detection of the AGPS source AX J144701–5919. We have chosen AX J144701–5919 as the example source for illustrating the output of ChIcAGO MAP because it was the first AGPS source observed with Chandra as part of the ChIcAGO survey, and it is also an interesting source with a bright counterpart (as demonstrated by its identification as an X-ray emitting Wolf–Rayet (WR) star in Anderson et al. 2011).

ChIcAGO MAP reprocesses all of the Chandra observations, both ACIS and HRC, using the CIAO chandra_repro script, which creates a new level = 2 event file and bad pixel file. The Chandra ACIS data are filtered to only include events with energies in the range 0.3–8.0 keV in order to avoid high-energy and cosmic ray particle backgrounds.

We have chosen to optimize ChIcAGO MAP for the detection of point sources as our short Chandra observations (<10 ks) are not very sensitive to extended sources. ChIcAGO MAP uses the CIAO wavelet detection algorithm, wavdetect, which we have set to search for all sources with wavelet scales of radii of 1, 2, 4, 8, and 16 pixels. This does, however, introduce a source selection bias as it is possible that AGPS sources that were unresolved with the ASCA point-spread function (PSF) could be extended and therefore resolvable with Chandra. Rerunning ChIcAGO MAP using wavdetect with larger scales more appropriate for extended sources or using the CIAO Voronoi tessellation and percolation source detection algorithm vtdetect (Ebeling & Wiedenmann 1993) could be conducted in the future to detect extended sources in the ChIcAGO Chandra images.

ChIcAGO MAP utilizes wavdetect to detect all sources within 3' of the original ASCA position (for example, see the Chandra detection of AX J144701–5919 in the right image of Figure 1). This search radius is based on the position accuracy and spatial resolution of ASCA and therefore is designed to ensure that the majority of contributing X-ray sources are encompassed. (However, it is also possible that there are associated X-ray sources beyond 3' from the AGPS position. This could be due to the inaccuracy of the ASCA positions of those sources that are blended or near the edge of the field of view. This is further explored in Section 3.1 and Appendix B.) In many cases more than one X-ray source could have been contributing to the total X-ray flux from an AGPS source originally detected with ASCA. (Note this is not the case for the example of AX J144701–5919.) The positional accuracy of wavdetect has been well investigated and tested in previous Chandra surveys (e.g., ChaMPlane; Hong et al. 2005).

The position of each ChIcAGO source as output by wavdetect is obtained, and Equation (5) of Hong et al. (2005) is used to calculate the 95% confidence position error circle.²² This wavdetect error is then added in quadrature to the absolute astrometry error of Chandra to obtain the total position error of each ChIcAGO source.

The source regions used to calculate the total number of source counts, which are centered on the source position as output by wavdetect, have a radius equivalent to 95% of the PSF at 1.5 keV. The background subtraction is performed using an annulus whose size is between two and five times the above 95% PSF radius, centered on the source position. Given the low number of counts detected, usually <100, the 1σ lower and upper confidence limits of the total number of counts are calculated using the Gehrels (1986) statistics. The following energy-based analysis is then performed on the ACIS-S observations. The total number of counts are calculated for the 0.3–8.0, 0.5–2.0, and 2.0–8.0 keV energy ranges. ChIcAGO MAP then uses the extracted ACIS-S counts and corresponding energies to calculate the 25%, 50%, and 75% photon fractions (E₂₅, E₅₀, and E₇₅), the energies below which 25%, 50%, and 75% of the photon energies are found, respectively. These median (E₅₀) and quartile (E₂₅ and E₇₅) values can immediately characterize the hardness of a source without using conventional hardness ratios, which are not versatile enough to account for diverse X-ray spectral types in the Galactic plane. These quartile fractions can then be used to employ quantile analysis (Hong et al. 2004), which uses a quantile based color–color diagram to classify spectral features and shapes of low-count sources (see Section 2.3.1).

The CIAO tool specextract is also run on the ACIS-S detected ChIcAGO sources to generate source and background spectrum files and their corresponding RMF and the ARF. These files are used to perform spectral interpolation with quantile analysis and to conduct spectral modeling and fitting with the CIAO package Sherpa. Further details on the spectral investigations of the ChIcAGO sources can be found in Section 2.3.

ChIcAGO MAP then performs a timing analysis on all the ChIcAGO sources detected with either the ACIS or HRC instrument to search for evidence of short-term variability and periodicity. In each case a light curve is extracted using dmextract where the counts are divided into eight bins (for example, see Figure 2). The Gehrels approximation to confidence limits for a Poisson distribution is used to estimate the errors as there are <20 X-ray counts in each bin. The χ² statistics are adopted to test for variability (for example, see Gaensler & Hunstead 2000). (It should be noted that as dmextract uses the upper (larger) Gehrels confidence limit to estimate the count rate errors, the resulting χ² output by ChIcAGO MAP may be underestimated.) For seven degrees of freedom, χ² ≳ 24.3 is required for a source to be considered variable at 99.9% confidence.

After correcting the photon arrival times to the solar system barycenter, the target source is then investigated for evidence of periodicity using the $Z_{n}^{2}$ test (Buccheri et al. 1983), equivalent to the Rayleigh statistic when n, the chosen number of harmonics, is set to 1. ChIcAGO MAP uses n = 1 for the sake of simplicity and because such a test is sensitive to sinusoidal distributions. $Z_{1}^{2}$ has a probability density function equivalent to the χ² statistic with two degrees of freedom. ChIcAGO MAP searched for periodicity down to 6.48 s (twice the frame time resolution; Weisskopf et al. 2003) for sources detected with ACIS and down to 0.01s for sources detected with HRC.²³ A power spectrum (power versus frequency) and folded light curve (counts per bin versus phase) for each source are generated, predicting the pulsed frequency of the highest $Z_{1}^{2}$ power and the probability that this power was random noise for a given number of trials (for example, see Figure 3). A 99.9% confidence was required for a source to have a significant level of periodicity, which corresponds approximately to $Z_{1}^{2} \gtrsim 26$ and $Z_{1}^{2} \gtrsim 36$ for ACIS and HRC observations, respectively.

It is also possible that many of the AGPS sources are transient or undergo long-term variability and so have changed significantly in flux since the original ASCA observations. The detailed analysis of any periodic, variable, and transient sources is beyond the scope of this paper. We only flag those sources that may fit into one of the above categories for the purpose of future investigations.

The next step in the ChIcAGO identification process is to search for multiwavelength counterparts to the ChIcAGO sources. ChIcAGO MAP accesses the USNO B1 (visual magnitude bands B, R, and I; Monet et al. 2003), the 2MASS PSC (near-infrared magnitude bands J, H, and K_s; Skrutskie et al. 2006), and the GLIMPSE I and II Spring '07 Catalogs (highly reliable) and Archives (more complete, less reliable; infrared magnitude bands 3.6, 4.5, 5.8, and 8.0 μm; Benjamin et al. 2003) to obtain a list of all the optical and infrared sources within 4'' of the wavdetect position of each ChIcAGO source. The information extracted from these surveys includes the position of the source, the offset from the Chandra position, and the magnitudes listed in the given survey or data set.²⁴

Small-sized image cutouts (6' × 6') from optical and infrared surveys, centered on the ChIcAGO source wavdetect position, are also downloaded to enable a visual inspection of likely counterparts and their surrounding environments. The B and R magnitude band images are obtained from DSS2B and DSS2R (McLean et al. 2000), and the J, H, and K magnitude band images are obtained from 2MASS. Image cutouts of infrared magnitude bands 3.6, 4.5, 5.8, and 8.0 μm are obtained from the GLIMPSE I and II, version 3.5, surveys.²⁵ A 30' by 30' image cutout of the 843 MHz radio sky is also generated from SUMSS, which is a survey conducted with the Molonglo Observatory Synthesis Telescope (MOST) and has a resolution of 43'' × 43'' cosec |dec| (Bock et al. 1999).²⁶ An example of all the image cutouts generated by ChIcAGO MAP for AX J144701–5919 can be seen in Figures 4 and 5.

Deducing the best spectral model fit to the ChIcAGO sources is difficult using standard X-ray spectral fitting techniques because of the small number of X-ray source counts detected in the Chandra observations (usually <100). We therefore implement two different techniques for predicting the best spectral parameters for the brighter (>20 X-ray counts) ChIcAGO sources. The first technique is "quantile analysis" (Hong et al. 2004), which has recently been developed to address some of the problems associated with spectral modeling of sources with low number statistics. Quantile analysis allows for the interpolation of likely spectral shapes of X-ray sources with as few as 10 counts. The second technique is utilizing the CIAO spectral fitting tool Sherpa to obtain best-fit spectral parameters using Cash (1979) statistics. These statistics are based on Poisson distributed data and are therefore ideal for modeling spectra with a limited number of source counts. Both methods are described in detail below.

2.3.1. Quantile Analysis

2.3.2. Spectral Modeling

While the X-ray morphology and spectrum can provide information on the nature of a ChIcAGO source, the key to identification is usually through an extensive multiwavelength follow-up campaign. ChIcAGO MAP identifies the possible optical and infrared counterparts in existing multiwavelength Galactic plane surveys. There are, however, many cases where the counterparts are too faint to be detected in these surveys because of the high absorption in the Galactic plane. It is also possible that the high object density in the Galactic plane could result in confusion with nearby sources. In these cases, further optical and infrared photometric observations were conducted with large telescopes to obtain detections of faint counterpart candidates and to separate likely blends. If the first photometric observing attempt was unsuccessful at detecting or separating a counterpart candidate, then deeper imaging using longer exposure times was conducted. Those X-ray sources that remain undetected at optical and infrared wavelengths will need to be further investigated in the X-ray band or at other wavelengths.

The radio wavelength band is also a useful diagnostic for identifying X-ray sources in the Galactic plane. Comparing the X-ray source positions with radio surveys can indicate if there is a likely radio counterpart or whether the X-ray source lies in a diffuse region of radio emission in the Galactic plane. Interferometric radio observations of a small subset of ChIcAGO sources were obtained in order to resolve confusing regions of radio emission and to allow for the detection or confirmation of compact radio counterparts.

2.4.1. Optical and Near-infrared Observations with Magellan

2.4.2. Radio Follow-up and ATCA Observations

We present results on 93 AGPS sources that have been observed with Chandra as part of the ChIcAGO survey. In many cases more than one source was detected within 3' of the original AGPS source positions, so as mentioned in Section 2.2, we will refer to these all as ChIcAGO sources. We therefore detected a total of 253 ChIcAGO sources in these 93 Chandra observations. The naming convention we have adopted is to call each source by the AGPS coordinate name, using ChI as the prefix, with a suffix between 1 and i, where i is equal to the number of sources detected in the 3' field. (The source name order is based on the order that wavdetect detected and output the sources in ChIcAGO MAP.) For example, two ChIcAGO sources, ChI J165646–4239_1 and ChI J165646–4239_2, were detected with Chandra in the vicinity of the AGPS source AX J165646–4239. These two ChIcAGO sources are the black dots in Figure 6. The ASCA GIS detection of AX J165646–4239 is overlaid on Figure 6 in the form of contours, demonstrating that both ChI J165646–4239_1 and ChI J165646–4239_2 may have contributed to the X-ray emission originally detected for this source in the AGPS.

Zoom In Zoom Out Reset image size

The Chandra observations of the 93 AGPS sources are summarized in Table 1. This information includes the observation identification, the date of the observation, instrument (ACIS-S or HRC-I), and the exposure time. Table 1 also lists the ChIcAGO sources detected by ChIcAGO MAP within 3' of the ASCA position and their corresponding wavdetect position, the offset of this position from the original AGPS position, the 95% wavdetect and total ChIcAGO source position errors, the background-corrected net counts in the 0.3–8.0, 0.5–2.0, and 2.0–8.0 keV energy ranges,²⁷ and the median energy quartile (E₅₀) and quartile ratio (3 × (E₂₅ − 0.3)/(E₇₅ − 0.3)) used in quantile analysis (see Section 2.3.1).

The variability analysis performed by ChIcAGO MAP detected only one ChIcAGO source, ChI J170444–4109_1, with a χ² exceeding the 99.9% short-term variability threshold (χ² = 80.6). However, this variability is not real as the ACIS-S detection of ChI J170444–4109_1 fell in the chip gap between the CCDs. This resulted in its light curve displaying the Chandra spacecraft's built-in dither as the source moved on and off the chip at regular intervals.²⁸ No other ChIcAGO source had a χ² exceeding the 99.9% confidence threshold for short-term variability or periodicity. This does not, however, rule out the possibility that some of these sources are periodic as the $Z^{2}_{1}$ period search technique is extremely limited for <200 X-ray counts.

Several faint ChIcAGO sources were detected with wavdetect in the Chandra observations that are not suspected of being significant contributors to the ASCA counts detected in the AGPS. (All these faint sources are included in Table 1.) In order to determine the detection significance of these sources, we independently calculated the probability of a false detection (Pfd) for each source based on the Poisson statistics. We expect there to be ∼10,0000 trials in a 3' radius ACIS CCD field of view if a detection cell size of a 1'' radius circle is assumed with two-dimensional Nyquist sampling (Weisskopf et al. 2007). (A generic detection cell size of 1'' for this detection significance calculation was suggested by Weisskopf et al. 2007 since the CIAO source detection algorithms usually use a detection cell size of between 40% and 80% of the PSF at a given energy.²⁹) Pfd =10⁻⁵ implies that there is one false source in the 3' radius ACIS Chandra field. The probability of a false detection was calculated for each source (see footnote 13 in Kashyap et al. 2010). Only 11 sources (out of 253) have a Pfd >10⁻⁵, but all have Pfd <7 × 10⁻⁴. Several of these sources also have a likely optical and/or infrared counterpart, increasing the significance of these detections. As this technique for calculating source significance is highly theoretical (e.g., the effective detection cell size of ACIS is likely to be slightly smaller than the 1'' radius circles), without any consideration for possible counterparts, we will include these 11 marginally significant sources in Table 1 (denoted by an asterisk in the first column).

Several of the sources detected in these observations have also been listed in the Chandra Source Catalog (CSC; Evans et al. 2010). We assume that a CSC source and a ChIcAGO source are the same if the separation between their two positions is less than the quadratic sum of their 95% error radii plus a constant term of 07, which accounts for the 95% absolute astrometry error of Chandra assuming the errors follow a Gaussian distribution³⁰ (see Equation (6) of Hong et al. 2005). We have conducted the same comparison with the fifth public release of the Second XMM-Newton Serendipitous Source Catalogue (2XMMi-DR3; Watson et al. 2009).³¹ The CSC and 2XMMi-DR3 names that correspond to any ChIcAGO sources are listed in Table 1. Further analysis of the XMM observations will be conducted in future work.

The purpose of the ChIcAGO survey is to study those X-ray sources detected in the ChIcAGO Chandra observations that fall within the flux range of the AGPS sources (F_x ∼ 10⁻¹³ to 10⁻¹² erg cm⁻² s⁻¹). In 62 out of the 93 AGPS sources observed with Chandra, ChIcAGO MAP found 74 sources with >20 X-ray counts that fall within the investigated 3' field of view that is centered on the AGPS positions. For the purposes of this study we will focus on these 74 ChIcAGO sources, listed in Table 2, as they are approximately within the original AGPS flux range. The detailed analysis of those 179 ChIcAGO sources with <20 counts will be deferred to future work. There were six AGPS sources where no ChIcAGO source was detected by Chandra: AX J165951–4209, AX J175331–2538, AX J180816–2021, AX J183607–0756, AX J185905+0333, and AX J191046+0917. At least three of these AGPS sources, AX J180816–2021, AX J185905+0333, and AX J191046+0917, may be transient given how bright they were in the original ASCA detections (Sugizaki et al. 2001). The remaining 25 AGPS fields observed with Chandra only have faint sources with <20 X-ray counts. There are also six AGPS sources, AX J144519–5949, AX J151005–5824, AX J154905–5420, AX J194310+2318, AX J194332+2323, and AX J195006+2628, where multiple ChIcAGO sources detected in each region (with ≲ 30 X-ray counts) sum to ≳60 counts, which is close to the number of X-ray counts that were expected to be detected with Chandra from each AGPS source (see Section 2.1). Many of the ChIcAGO sources in these fields also have optical and/or infrared counterparts and may therefore be members of star clusters (see Section 4.1 for further discussion).

The contribution of X-ray emission beyond the 3' search radius was also investigated using wavdetect to identify all the point sources with >20 X-ray counts in the Chandra observations that lie between 3' and 5' from the original AGPS position. Only 14 X-ray point sources were found in the 3'–5' annulus surrounding 11 AGPS sources, which demonstrates that the 3' search radius used by ChIcAGO MAP is reasonable and has likely allowed us to identify the majority of ChIcAGO sources. These 14 X-ray point sources are further discussed in Appendix B, where we explore the likelihood of whether they could be associated with their nearby AGPS source.

Quantile analysis and spectral modeling using Cash statistics were both techniques used to infer the most likely absorbed power-law and absorbed thermal bremsstrahlung spectra of the bright ChIcAGO sources detected by ACIS-S.³² The spectral parameters and the absorbed and unabsorbed X-ray fluxes for absorbed power-law and absorbed thermal bremsstrahlung models are listed in Tables 3 and 4, respectively. There are also six HRC-I detected ChIcAGO sources with >20 counts (ChI J144042–6001_1, ChI J153818–5541_1, ChI J163252–4746_2, ChI J163751–4656_1, ChI J165420–4337_1, and ChI J172642–3540_1) that are not included in Tables 3 and 4 as their X-ray observations do not contain any spectral information. The possible counterparts to these six sources are, however, explored in detail in Section 3.3.

Table 3. Quantile Analysis Power-law Interpolation and Cash Statistics Power-law Spectral Fit to ACIS Chandra Detections of the ChIcAGO Sources

ChIcAGO Source	Quantile Absorbed Power-law Interpolationa					Cash Statistics Absorbed Power-law Fitb
ChI	Γ	NH	Fx, abs	Fx, unabs	$\chi ^{2}_{{\rm red}}$	Γ	NH	Fx, abs	Fx, unabs	C
J143148–6021_3	$0.91_{-0.35}^{+0.44}$	$0.61_{-0.34}^{+0.54}$	9.12 ± 0.75	10.93 ± 0.90	0.3	$0.70_{-0.25}^{+0.26}$	$0.42_{-0.17}^{+0.20}$	$9.85_{-1.15}^{+1.35}$	$11.17_{-1.31}^{+1.29}$	0.5
J144547–5931_1						$4.98_{-0.94}^{+1.06}$	$4.53_{-1.07}^{+1.25}$	$2.38_{-0.59}^{+0.45}$	$1802.71_{-1388.60}^{+4414.31}$	1.0
J144701–5919_1						$3.39_{-0.61}^{+0.67}$	$2.70_{-0.64}^{+0.73}$	$5.55_{-1.90}^{+1.15}$	$177.46_{-139.42}^{+157.69}$	0.6
J150436–5824_1	$-0.37_{-0.00}^{+0.17}$	$0.01_{-0.00}^{+0.06}$	8.45 ± 0.71	8.47 ± 0.71	0.9	$-0.05_{-0.27}^{+0.28}$	$0.49_{-0.27}^{+0.32}$	$8.46_{-0.84}^{+1.07}$	$9.03_{-0.98}^{+1.05}$	1.2
J154122–5522_1	$2.53_{-0.49}^{+0.62}$	$0.14_{-0.11}^{+0.14}$	2.50 ± 0.27	4.46 ± 0.49	0.6	$2.35_{-0.39}^{+0.43}$	$0.15_{-0.09}^{+0.10}$	$2.86_{-0.46}^{+0.38}$	$4.82_{-1.22}^{+0.99}$	1.0
J154557–5443_1	$2.98_{-0.70}^{+1.30}$	$0.29_{-0.17}^{+0.19}$	0.17 ± 0.04	0.56 ± 0.12
J155035–5408_3	$2.73_{-0.46}^{+0.55}$	$0.67_{-0.21}^{+0.24}$	2.90 ± 0.20	11.47 ± 0.80	0.9	$2.72_{-0.27}^{+0.28}$	0.61 ± 0.11	$2.88_{-0.33}^{+0.36}$	$11.33_{-3.55}^{+3.17}$	1.8
J155831–5334_1	$1.70_{-0.63}^{+0.72}$	$0.43_{-0.34}^{+0.45}$	3.62 ± 0.50	5.45 ± 0.76	0.4	$1.65_{-0.49}^{+0.54}$	$0.38_{-0.22}^{+0.28}$	$3.31_{-0.63}^{+0.74}$	$5.99_{-2.16}^{+1.39}$	1.0
J162011–5002_1	$0.88_{-1.00}^{+1.00}$	$2.50_{-1.56}^{+2.8}$	8.17 ± 1.11	11.73 ± 1.6	0.6	$0.75_{-0.75}^{+0.83}$	$2.59_{-1.53}^{+2.02}$	$7.62_{-2.68}^{+2.09}$	$12.82_{-3.94}^{+1.50}$	1.1
J162046–4942_1	$1.42_{-0.31}^{+0.35}$	$0.12_{-0.10}^{+0.15}$	3.35 ± 0.31	3.84 ± 0.35	1.4	$1.71_{-0.29}^{+0.31}$	$0.31_{-0.11}^{+0.13}$	$3.32_{-0.53}^{+0.61}$	$5.01_{-0.85}^{+0.72}$	2.7
J165105–4403_1	$0.84_{-0.31}^{+0.37}$	$0.34_{-0.25}^{+0.36}$	12.6 ± 1.03	14.19 ± 1.15	0.8	$0.97_{-0.27}^{+0.28}$	$0.58_{-0.19}^{+0.22}$	$13.42_{-1.83}^{+1.59}$	$16.43_{-1.68}^{+2.16}$	1.1
J165217–4414_1	$1.63_{-0.49}^{+0.47}$	$0.64_{-0.34}^{+0.41}$	5.97 ± 0.57	9.32 ± 0.9	0.7	$1.29_{-0.29}^{+0.31}$	$0.38_{-0.15}^{+0.18}$	$6.48_{-1.06}^{+1.14}$	$8.66_{-1.56}^{+1.01}$	0.9
J165646–4239_1						$3.93_{-0.36}^{+0.40}$	0.22 ± 0.06	$4.99_{-0.39}^{+0.36}$	$27.34_{-13.30}^{+18.76}$	2.8
J165646–4239_2	$1.49_{-0.62}^{+0.79}$	$0.31_{-0.30}^{+0.48}$	2.27 ± 0.30	2.98 ± 0.39	0.4	$1.28_{-0.41}^{+0.44}$	$0.22_{-0.17}^{+0.21}$	$2.37_{-0.36}^{+0.49}$	$3.18_{-0.56}^{+0.35}$	0.7
J165707–4255_1	$3.91_{-1.25}^{+1.07}$	$0.22_{-0.20}^{+0.21}$	2.5 ± 0.32	13.51 ± 1.73	2.1	$3.17_{-0.60}^{+0.79}$	$0.07_{-0.07}^{+0.12}$	$3.05_{-0.59}^{+0.53}$	$7.05_{-2.67}^{+4.82}$	2.6
J170017–4220_1	$-1.89_{-0.10}^{+1.14}$	$1.00_{-0.99}^{+5.60}$	19.25 ± 2.37	19.91 ± 2.45	1.0	$0.24_{-0.98}^{+1.10}$	$14.01_{-6.85}^{+8.49}$	$11.36_{-7.08}^{+5.90}$	$29.98_{-8.94}^{+6.53}$	1.0
J170052–4210_1						$4.89_{-1.45}^{+1.98}$	$0.37_{-0.23}^{+0.31}$	$0.81_{-0.22}^{+0.23}$	$32.64_{-26.84}^{+152.63}$	3.6
J170444–4109_1	$3.38_{-0.65}^{+0.86}$	$0.16_{-0.13}^{+0.18}$	9.01 ± 0.73	27.22 ± 2.21	1.7	$3.17_{-0.35}^{+0.38}$	0.23 ± 0.09	$8.35_{-1.16}^{+1.04}$	$28.45_{-8.65}^{+14.82}$	2.1
J170444–4109_2	$0.31_{-1.37}^{+1.60}$	$1.70_{-1.69}^{+3.10}$	2.10 ± 0.50	2.51 ± 0.59
J170536–4038_1	$2.88_{-0.63}^{+0.79}$	$0.26_{-0.15}^{+0.19}$	6.54 ± 0.61	19.03 ± 1.79	1.0	$2.09_{-0.31}^{+0.33}$	0.14 ± 0.08	$8.50_{-1.15}^{+1.05}$	$12.81_{-1.93}^{+2.24}$	1.7
J171910–3652_2						$3.25_{-0.44}^{+0.49}$	$0.14_{-0.07}^{+0.08}$	$2.64_{-0.29}^{+0.38}$	$6.78_{-2.33}^{+3.77}$	3.3
J171922–3703_1	$-0.75_{-0.55}^{+0.49}$	$1.15_{-1.03}^{+1.85}$	6.83 ± 0.56	7.27 ± 0.60	0.6	$-0.26_{-0.39}^{+0.41}$	$2.19_{-0.92}^{+1.12}$	$5.79_{-1.32}^{+1.10}$	$7.09_{-1.65}^{+0.84}$	1.4
J172050–3710_1						$4.70_{-0.37}^{+0.39}$	$0.54_{-0.07}^{+0.08}$	$2.07_{-0.18}^{+0.11}$	$67.55_{-32.05}^{+49.73}$	1.8
J172550–3533_1	$0.14_{-1.40}^{+1.93}$	$6.60_{-5.3}^{+11.9}$	2.8 ± 0.43	4.11 ± 0.63	0.4
J172550–3533_2	$3.28_{-1.00}^{+1.49}$	$0.38_{-0.25}^{+0.38}$	0.44 ± 0.06	2.28 ± 0.34	0.6
J172550–3533_3	$3.17_{-0.79}^{+1.11}$	$0.28_{-0.22}^{+0.30}$	0.55 ± 0.09	2.11 ± 0.33
J172623–3516_1	$2.94_{-1.28}^{+2.04}$	$0.46_{-0.40}^{+0.84}$	1.07 ± 0.16	4.44 ± 0.68	0.6	$1.63_{-0.47}^{+0.52}$	$0.13_{-0.13}^{+0.17}$	$1.64_{-0.35}^{+0.61}$	$2.49_{-0.54}^{+0.41}$	1.4
J172642–3504_1	$0.84_{-0.70}^{+0.75}$	$2.30_{-1.25}^{+2.00}$	2.86 ± 0.31	4.01 ± 0.43	0.2	$0.64_{-0.51}^{+0.55}$	$2.40_{-0.98}^{+1.18}$	$3.02_{-1.03}^{+0.57}$	$4.28_{-1.61}^{+0.52}$	0.6
J173548–3207_1	$2.04_{-0.47}^{+0.67}$	$0.20_{-0.15}^{+0.22}$	1.65 ± 0.24	2.54 ± 0.37	3.5	$2.17_{-0.55}^{+0.60}$	$0.32_{-0.19}^{+0.22}$	$1.64_{-0.48}^{+0.35}$	$3.80_{-1.69}^{+1.86}$	3.3
J180857–2004_1	$2.84_{-1.28}^{+1.86}$	$5.40_{-2.80}^{+4.60}$	1.16 ± 0.17	14.45 ± 2.18
J180857–2004_2	$3.45_{-0.93}^{+1.28}$	$6.20_{-2.00}^{+3.00}$	1.94 ± 0.22	80.67 ± 9.02	0.7	$2.71_{-0.67}^{+0.72}$	$4.35_{-1.15}^{+1.32}$	$1.96_{-0.59}^{+0.40}$	$23.39_{-13.91}^{+17.88}$	1.7
J181116–1828_2	$2.00_{-1.09}^{+1.38}$	$2.20_{-1.50}^{+2.20}$	2.77 ± 0.39	7.66 ± 1.07	1.6	$1.52_{-0.61}^{+0.68}$	$1.13_{-0.58}^{+0.75}$	$2.24_{-0.88}^{+0.58}$	$5.14_{-2.33}^{+1.53}$	2.3
J181213–1842_7	$-0.16_{-0.52}^{+0.55}$	$0.52_{-0.50}^{+0.73}$	10.49 ± 0.98	11.03 ± 1.03	0.5	$0.37_{-0.39}^{+0.42}$	$1.62_{-0.65}^{+0.77}$	$10.12_{-1.96}^{+1.67}$	$12.47_{-1.92}^{+1.70}$	0.5
J181213–1842_9	$3.63_{-1.07}^{+1.35}$	$0.20_{-0.19}^{+0.31}$	0.35 ± 0.09	1.48 ± 0.36
J181852–1559_2	$0.77_{-0.82}^{+1.11}$	$2.50_{-1.59}^{+3.20}$	6.00 ± 0.94	8.34 ± 1.30	1.1	$-0.18_{-0.77}^{+0.88}$	$1.24_{-1.24}^{+2.19}$	$5.22_{-3.62}^{+3.02}$	$6.48_{-3.06}^{+3.14}$	0.7
J181915–1601_2	$1.32_{-0.37}^{+0.49}$	$0.70_{-0.33}^{+0.50}$	3.79 ± 0.52	5.20 ± 0.71	1.5	$2.14_{-0.63}^{+0.69}$	$1.46_{-0.58}^{+0.71}$	$2.97_{-0.94}^{+0.89}$	$9.40_{-4.37}^{+7.92}$	2.0
J181915–1601_3	$1.21_{-0.34}^{+0.46}$	$0.67_{-0.31}^{+0.48}$	5.35 ± 0.62	7.03 ± 0.81	1.2	$1.61_{-0.48}^{+0.52}$	$1.38_{-0.50}^{+0.59}$	$5.25_{-1.29}^{+1.04}$	$10.99_{-4.17}^{+2.77}$	1.6
J182435–1311_1						$4.77_{-0.70}^{+0.78}$	$1.44_{-0.30}^{+0.34}$	$0.45_{-0.13}^{+0.09}$	$60.30_{-51.18}^{+99.63}$	3.4
J182509–1253_1						$3.40_{-0.52}^{+0.57}$	$0.44_{-0.13}^{+0.14}$	$0.57_{-0.09}^{+0.12}$	$4.49_{-2.67}^{+3.81}$	2.2
J182509–1253_3	$-0.05_{-0.84}^{+1.65}$	$3.00_{-2.66}^{+7.50}$	1.96 ± 0.28	2.40 ± 0.34	1.9	$-0.08_{-0.99}^{+1.15}$	$2.98_{-2.55}^{+4.05}$	$1.25_{-0.96}^{+0.79}$	$1.84_{-0.69}^{+0.74}$	2.5
J182530–1144_2	$0.14_{-0.37}^{+0.38}$	$0.55_{-0.31}^{+0.33}$	9.44 ± 0.61	10.13 ± 0.66	1.0	$0.43_{-0.23}^{+0.24}$	$1.01_{-0.28}^{+0.32}$	$9.37_{-1.22}^{+0.71}$	$10.83_{-1.16}^{+1.04}$	1.7
J182651–1206_4	$1.80_{-0.70}^{+0.89}$	$0.28_{-0.27}^{+0.42}$	0.34 ± 0.08	0.49 ± 0.12
J183116–1008_1	$2.28_{-0.55}^{+0.70}$	$4.80_{-1.30}^{+1.80}$	5.66 ± 0.61	28.9 ± 3.13	1.4	$3.28_{-0.79}^{+0.86}$	$8.15_{-1.97}^{+2.26}$	$5.33_{-2.67}^{+1.11}$	$373.72_{-273.72}^{+629.26}$	1.7
J183206–0938_1	$0.91_{-0.46}^{+0.47}$	$1.50_{-0.71}^{+0.90}$	29.85 ± 2.40	39.86 ± 3.20	0.7	$0.86_{-0.35}^{+0.36}$	$1.81_{-0.51}^{+0.58}$	$33.04_{-7.94}^{+3.69}$	$47.45_{-11.41}^{+5.54}$	1.4
J183345–0828_1	$2.63_{-1.03}^{+1.40}$	$2.60_{-1.35}^{+2.00}$	0.65 ± 0.10	4.04 ± 0.64
J183356–0822_2	$2.38_{-1.75}^{+2.60}$	$18.0_{-10.20}^{+17.00}$	2.06 ± 0.25	26.08 ± 3.14	1.0	$0.82_{-1.13}^{+1.30}$	$10.46_{-5.41}^{+7.15}$	$1.65_{-1.21}^{+0.90}$	$5.80_{-1.38}^{+0.95}$	1.2
J183356–0822_3	$1.28_{-1.11}^{+1.28}$	$1.80_{-1.61}^{+2.40}$	0.43 ± 0.11	0.69 ± 0.17
J184652–0240_1	$2.88_{-0.70}^{+0.96}$	$1.50_{-0.56}^{+0.80}$	4.14 ± 0.45	28.0 ± 3.02	1.0	$2.92_{-0.51}^{+0.55}$	$1.51_{-0.38}^{+0.44}$	$3.93_{-0.71}^{+0.79}$	$35.73_{-19.70}^{+24.49}$	1.8
J184738–0156_1	$2.35_{-1.14}^{+1.79}$	$12.0_{-5.20}^{+9.00}$	19.56 ± 1.76	178.36 ± 16.08	1.2	$0.53_{-0.53}^{+0.57}$	$5.13_{-1.75}^{+2.14}$	$21.11_{-8.53}^{+4.10}$	$35.70_{-12.23}^{+3.87}$	2.2
J184741–0219_3	$-1.62_{-0.38}^{+1.86}$	$1.00_{-0.99}^{+12.00}$	2.57 ± 0.46	2.67 ± 0.47
J185608+0218_1						$3.83_{-0.40}^{+0.44}$	0.33 ± 0.08	$6.97_{-0.74}^{+0.66}$	$50.56_{-22.77}^{+47.69}$	1.8
J185643+0220_2	$0.00_{-0.44}^{+0.63}$	$0.97_{-0.81}^{+1.53}$	3.27 ± 0.39	3.59 ± 0.43	1.3	$0.75_{-0.64}^{+0.70}$	$3.22_{-1.47}^{+1.83}$	$3.03_{-1.35}^{+0.53}$	$4.97_{-0.80}^{+0.63}$	1.6
J185750+0240_1	$0.25_{-0.58}^{+0.74}$	$0.32_{-0.31}^{+1.08}$	4.20 ± 0.63	4.43 ± 0.66	0.5	$0.31_{-0.52}^{+0.60}$	$0.30_{-0.30}^{+0.61}$	$4.01_{-1.45}^{+0.78}$	$4.28_{-0.97}^{+1.01}$	0.4
J190534+0659_1	$2.28_{-0.96}^{+1.35}$	$0.52_{-0.43}^{+0.68}$	0.73 ± 0.19	1.70 ± 0.45
J190749+0803_1	$1.74_{-0.75}^{+0.93}$	$4.80_{-2.00}^{+3.00}$	6.12 ± 0.64	17.21 ± 1.80	0.2	$0.75_{-0.51}^{+0.55}$	$2.73_{-1.04}^{+1.24}$	$6.79_{-1.89}^{+0.99}$	$10.52_{-2.06}^{+1.11}$	0.6
J190814+0832_2	$2.66_{-0.56}^{+0.62}$	$0.73_{-0.27}^{+0.37}$	0.9 ± 0.14	3.41 ± 0.53	1.0	$3.22_{-0.74}^{+0.86}$	$0.74_{-0.29}^{+0.36}$	$0.73_{-0.21}^{+0.18}$	$5.83_{-3.54}^{+6.28}$	0.7
J190818+0745_1	$1.32_{-0.25}^{+0.28}$	$0.20_{-0.12}^{+0.15}$	10.45 ± 0.89	12.38 ± 1.05	0.7	$1.35_{-0.25}^{+0.26}$	$0.22_{-0.10}^{+0.11}$	$10.35_{-1.38}^{+1.30}$	$12.91_{-1.30}^{+1.55}$	1.2
J194152+2251_2	$0.38_{-0.69}^{+0.56}$	$0.82_{-0.63}^{+0.58}$	6.07 ± 0.72	6.84 ± 0.81	0.3	$0.36_{-0.46}^{+0.50}$	$0.91_{-0.55}^{+0.72}$	$5.96_{-2.02}^{+1.36}$	$7.02_{-1.88}^{+1.39}$	0.3
J194939+2631_1	$0.67_{-0.49}^{+0.69}$	$2.30_{-1.15}^{+1.90}$	16.71 ± 1.57	22.28 ± 2.09	0.9	$-0.10_{-0.35}^{+0.38}$	$0.67_{-0.43}^{+0.57}$	$18.24_{-4.08}^{+2.62}$	$20.06_{-3.87}^{+2.18}$	1.5

Notes. The quantile analysis spectral interpolation and Cash statistics spectral modeling are performed on those ChIcAGO sources with >20 and >50 X-ray counts, respectively. ^aThe quantile analysis interpolated absorbed power law model parameters including the spectral index, Γ, and absorption column density N_H (10²² cm⁻²), as well as the absorbed and unabsorbed X-ray fluxes, F_{x, abs} and F_{x, unabs} (10⁻¹³ erg cm⁻² s⁻¹), respectively, in the 0.3–8.0 keV energy range. The quoted $\chi ^{2}_{{\rm red}}$ value is what is expected if these interpolated spectral parameters are used to fit the X-ray spectrum of the source. ^bThe absorbed power law fit model parameters using Cash (1979) statistics including the spectral index, Γ, and absorption column density N_H (10²² cm⁻²), as well as the absorbed and unabsorbed X-ray fluxes, F_{x, abs} and F_{x, unabs} (10⁻¹³ erg cm⁻² s⁻¹), respectively, in the 0.3–8.0 keV energy range. The quoted C value is the reduced Cash statistic, which is the observed statistic divided by the number of degrees of freedom. A good fit is indicated by a C value of the order of one.

Download table as:  ASCIITypeset images: 1 2

Table 4. Quantile Analysis Bremsstrahlung Spectral Interpolation and Cash Statistic Bremsstrahlung Spectral Fit to ACIS Chandra Detections of the ChIcAGO Sources

ChIcAGO Source	Absorbed Bremsstrahlung Interpolationa					Cash Statistics Absorbed Bremsstrahlung Fitb
ChI	kT	NH	Fx, abs	Fx, unabs	$\chi ^{2}_{{\rm red}}$	kT	NH	Fx, abs	Fx, unabs	C
J144519–5949_2	$1.38_{-0.65}^{+2.15}$	$4.80_{-2.2}^{+3.60}$	2.17 ± 0.44	21.59 ± 4.37
J144547–5931_1	$0.84_{-0.21}^{+0.30}$	$4.10_{-0.9}^{+1.20}$	2.56 ± 0.31	57.74 ± 6.95	0.9	$1.01_{-0.23}^{+0.37}$	$3.27_{-0.74}^{+0.86}$	$1.93_{-1.24}^{+0.68}$	$37.28_{-22.51}^{+9.98}$	0.9
J144701–5919_1	$1.00_{-0.39}^{+0.70}$	$3.50_{-1.2}^{+1.90}$	5.49 ± 0.60	74.01 ± 8.12	0.6	$2.04_{-0.54}^{+0.95}$	$1.93_{-0.43}^{+0.49}$	$5.78_{-2.03}^{+1.16}$	$23.52_{-7.51}^{+4.04}$	0.9
J154122–5522_1	$1.88_{-0.70}^{+0.96}$	$0.04_{-0.03}^{+0.09}$	2.35 ± 0.26	2.72 ± 0.3	0.6	$2.62_{-0.80}^{+1.68}$	$0.05_{-0.05}^{+0.06}$	$2.46_{-0.47}^{+0.45}$	$3.08_{-0.70}^{+0.47}$	1.0
J154557–5443_1	$1.38_{-0.73}^{+1.32}$	$0.14_{-0.1}^{+0.15}$	0.16 ± 0.03	0.25 ± 0.05
J154557–5443_2	$1.77_{-1.33}^{+96.43}$	$24.0_{-15.8}^{+46.0}$	0.63 ± 0.16	17.65 ± 4.49
J154557–5443_3	$0.10_{-0.00}^{+0.30}$	$0.97_{-0.88}^{+1.23}$	0.11 ± 0.03	224.95 ± 51.46
J155035–5408_1	$0.35_{-0.18}^{+0.18}$	$0.14_{-0.14}^{+0.38}$	0.18 ± 0.05	0.56 ± 0.14
J155035–5408_3	$2.14_{-0.72}^{+0.91}$	$0.43_{-0.1}^{+0.21}$	2.73 ± 0.19	5.29 ± 0.37	0.9	$2.17_{-0.38}^{+0.55}$	$0.42_{-0.07}^{+0.08}$	$2.81_{-0.37}^{+0.28}$	$5.43_{-0.85}^{+0.45}$	1.8
J155331–5347_1	$0.31_{-0.16}^{+0.31}$	$0.70_{-0.39}^{+0.85}$	1.25 ± 0.15	30.92 ± 3.67	2.2
J155831–5334_1	$9.04_{-5.99}^{+89.16}$	$0.34_{-0.22}^{+0.30}$	3.54 ± 0.49	4.73 ± 0.66	0.4
J162046–4942_1						$8.37_{-3.50}^{+16.53}$	$0.24_{-0.08}^{+0.09}$	$3.48_{-0.61}^{+0.50}$	$4.48_{-0.89}^{+0.56}$	2.7
J165646–4239_1	$0.35_{-0.10}^{+0.13}$	$0.25_{-0.10}^{+0.15}$	4.06 ± 0.25	20.59 ± 1.26	1.8	$0.78_{-0.11}^{+0.14}$	0.03 ± 0.03	$5.06_{-0.64}^{+0.53}$	$6.62_{-0.99}^{+0.49}$	2.8
J165707–4255_1	$0.71_{-0.31}^{+0.21}$	$0.05_{-0.04}^{+0.16}$	2.53 ± 0.32	3.39 ± 0.43	2.1	$0.87_{-0.16}^{+0.24}$	<0.02	$2.69_{-0.53}^{+0.24}$	$2.78_{-0.57}^{+0.24}$	3.3
J170052–4210_1	$0.31_{-0.16}^{+0.35}$	$0.25_{-0.22}^{+0.45}$	0.84 ± 0.14	4.93 ± 0.82		$0.51_{-0.19}^{+0.42}$	$0.13_{-0.13}^{+0.17}$	$0.84_{-0.27}^{+0.18}$	$1.98_{-0.80}^{+0.69}$	3.9
J170444–4109_1	$0.97_{-0.29}^{+0.28}$	$0.01_{0.00}^{+0.10}$	9.28 ± 0.75	9.85 ± 0.80	1.8	$1.54_{-0.30}^{+0.42}$	$0.03_{-0.03}^{+0.06}$	$8.51_{-0.88}^{+0.66}$	$10.48_{-1.02}^{+0.92}$	2.4
J170536–4038_1	$1.46_{-0.55}^{+1.11}$	$0.13_{-0.10}^{+0.14}$	6.09 ± 0.57	9.33 ± 0.88	1.1	$3.95_{-1.29}^{+2.79}$	$0.05_{-0.05}^{+0.06}$	$8.65_{-1.65}^{+0.81}$	$9.68_{-1.40}^{+1.46}$	1.8
J171910–3652_2	$0.17_{-0.06}^{+0.07}$	$0.61_{-0.24}^{+0.39}$	1.81 ± 0.17	136.22 ± 12.59	1.1	$1.23_{-0.20}^{+0.25}$	<0.03	$2.59_{-0.26}^{+0.30}$	$2.71_{-0.20}^{+0.25}$	4.00
J172050–3710_1	$0.38_{-0.06}^{+0.08}$	$0.37_{-0.09}^{+0.09}$	1.77 ± 0.10	12.78 ± 0.72	1.1	$0.67_{-0.08}^{+0.09}$	$0.24_{-0.04}^{+0.05}$	$2.03_{-0.28}^{+0.13}$	$6.06_{-0.98}^{+0.75}$	2.1
J172550–3533_2	$1.14_{-0.55}^{+1.35}$	$0.20_{-0.16}^{+0.25}$	0.41 ± 0.06	0.81 ± 0.12	0.6
J172550–3533_3	$1.28_{-0.56}^{+1.14}$	$0.10_{-0.09}^{+0.21}$	0.55 ± 0.09	0.79 ± 0.12
J172623–3516_1	$1.56_{-1.10}^{+6.01}$	$0.28_{-0.26}^{+0.69}$	1.0 ± 0.15	1.93 ± 0.29	0.6
J173548–3207_1	$4.00_{-2.30}^{+6.00}$	$0.12_{-0.09}^{+0.21}$	1.58 ± 0.23	1.95 ± 0.29	3.6	$3.67_{-1.59}^{+5.98}$	$0.21_{-0.13}^{+0.15}$	$1.57_{-0.70}^{+0.42}$	$2.13_{-0.59}^{+0.68}$	3.3
J175404–2553_3	$2.56_{-1.85}^{+95.64}$	$33.00_{-18.00}^{+40.00}$	1.34 ± 0.27	28.13 ± 5.69
J180857–2004_1	$3.19_{-1.77}^{+19.41}$	$4.40_{-2.00}^{+3.20}$	1.12 ± 0.17	4.19 ± 0.63
J180857–2004_2	$2.21_{-0.82}^{+1.68}$	$4.90_{-1.40}^{+1.90}$	1.88 ± 0.21	10.23 ± 1.14	0.6	$4.03_{-1.50}^{+4.75}$	$3.37_{-0.78}^{+0.88}$	$1.78_{-1.13}^{+0.55}$	$5.52_{-3.43}^{+1.58}$	1.9
J181116–1828_2	$6.52_{-4.45}^{+91.68}$	$1.85_{-0.85}^{+1.55}$	2.7 ± 0.38	5.36 ± 0.75	1.5
J181213–1842_9	$0.84_{-0.49}^{+0.79}$	$0.04_{-0.03}^{+0.27}$	0.35 ± 0.09	0.45 ± 0.11
J181915–1601_2						$5.44_{-2.61}^{+17.68}$	$1.15_{-0.40}^{+0.49}$	$3.25_{-0.99}^{+0.55}$	$5.92_{-1.37}^{+1.18}$	2.1
J182435–1311_1	$0.84_{-0.17}^{+0.19}$	$0.82_{-0.15}^{+0.18}$	0.47 ± 0.05	2.54 ± 0.27	1.7	$0.74_{-0.15}^{+0.21}$	$0.99_{-0.21}^{+0.24}$	$0.44_{-0.21}^{+0.08}$	$3.90_{-2.11}^{+0.91}$	3.3
J182509–1253_1	$0.76_{-0.23}^{+0.38}$	$0.26_{-0.12}^{+0.17}$	0.49 ± 0.05	1.4 ± 0.15	1.6	$1.37_{-0.34}^{+0.56}$	$0.20_{-0.08}^{+0.09}$	$0.54_{-0.19}^{+0.09}$	$1.03_{-0.22}^{+0.21}$	2.7
J182538–1214_1	$0.38_{-0.26}^{+0.90}$	$0.37_{-0.36}^{+1.13}$	0.52 ± 0.10	3.86 ± 0.72
J182651–1206_4	$6.31_{-4.17}^{+91.89}$	$0.20_{-0.19}^{+0.28}$	0.32 ± 0.08	0.42 ± 0.10
J183116–1008_1	$5.26_{-2.28}^{+7.89}$	$4.10_{-1.00}^{+1.30}$	5.55 ± 0.60	15.35 ± 1.67	1.5	$2.96_{-0.96}^{+2.26}$	$6.37_{-1.35}^{+1.54}$	$5.18_{-2.86}^{+1.52}$	$26.82_{-13.83}^{+7.42}$	1.9
J183345–0828_1	$3.26_{-1.77}^{+13.04}$	$2.00_{-0.90}^{+1.40}$	0.62 ± 0.10	1.64 ± 0.26
J183356–0822_2	$6.31_{-4.64}^{+91.89}$	$16.0_{-6.6}^{+13.0}$	2.05 ± 0.25	10.44 ± 1.26	1.0
J184652–0240_1	$2.21_{-0.86}^{+2.00}$	$1.10_{-0.37}^{+0.55}$	3.9 ± 0.42	10.29 ± 1.11	0.9	$2.21_{-0.59}^{+1.06}$	$1.14_{-0.27}^{+0.31}$	$3.66_{-1.46}^{+0.79}$	$10.87_{-3.90}^{+1.81}$	1.8
J184738–0156_1	$5.47_{-3.33}^{+92.73}$	$11.0_{-4.20}^{+6.00}$	19.34 ± 1.74	83.9 ± 7.56	1.2
J185608+0218_1	$0.42_{-0.11}^{+0.17}$	$0.31_{-0.12}^{+0.15}$	5.66 ± 0.41	30.15 ± 2.16	1.3	$0.96_{-0.16}^{+0.21}$	$0.10_{-0.04}^{+0.05}$	$6.82_{-0.83}^{+0.72}$	$11.02_{-2.17}^{+1.91}$	2.0
J190534+0659_1	$3.12_{-1.91}^{+41.53}$	$0.37_{-0.30}^{+0.45}$	0.69 ± 0.18	1.12 ± 0.30
J190814+0832_2	$2.21_{-0.72}^{+1.75}$	$0.52_{-0.19}^{+0.27}$	0.84 ± 0.13	1.71 ± 0.26	0.8	$1.46_{-0.48}^{+1.00}$	$0.50_{-0.21}^{+0.25}$	0.77		0.4
J194310+2318_5	$0.55_{-0.25}^{+0.29}$	$0.12_{-0.1}^{+0.22}$	0.69 ± 0.13	1.42 ± 0.26
J195006+2628_2	$0.48_{-0.32}^{+1.01}$	$0.29_{-0.28}^{+0.81}$	0.22 ± 0.04	0.97 ± 0.20

Notes. The quantile analysis spectral interpolation and Cash statistics spectral modeling are performed on those ChIcAGO sources with >20 and >50 X-ray counts, respectively. The CIAO task sample_flux failed for ChI J190814+0832_2 and so the quoted flux is taken from the best fit model parameters. ^aThe quantile analysis interpolated absorbed thermal bremsstrahlung model parameters including the temperature, kT (keV), and absorption column density N_H (10²² cm⁻²), as well as the absorbed and unabsorbed X-ray fluxes, F_{x, abs} and F_{x, unabs} (10⁻¹³ erg cm⁻² s⁻¹), respectively, in the 0.3–8.0 keV energy range. The quoted $\chi ^{2}_{{\rm red}}$ value is what is expected if these interpolated spectral parameters are used to fit the X-ray spectrum of the source. ^bThe absorbed thermal bremsstrahlung fit model parameters using Cash (1979) statistics including the temperature, kT (keV), and absorption column density N_H (10²² cm⁻²), as well as the absorbed and unabsorbed X-ray fluxes, F_{x, abs} and F_{x, unabs} (10⁻¹³ erg cm⁻² s⁻¹), respectively, in the 0.3–8.0 keV energy range. The quoted C value is the reduced Cash statistic, which is the observed statistic divided by the number of degrees of freedom. A good fit is indicated by a C value of the order of one.

Download table as:  ASCIITypeset image

There is no measure of goodness for the quantile-analysis-derived absorbed power-law and absorbed bremsstrahlung spectral parameters listed for the ChIcAGO sources in Tables 3 and 4. Instead, a spectral interpolation is classified as unreasonable if the resulting parameters are outside −2 < Γ < 4 for a power-law model and 0.1 < kT < 10 keV for a bremsstrahlung model. These parameter limits can allow for the grouping of the ChIcAGO sources into existing categories that are based on the physical understanding of X-ray sources that emit in the Chandra energy range (0.3–8 keV). While these selection criteria explicitly exclude the identification of new types of X-ray sources with unusual spectra, such an investigation is beyond the scope of this work. In all cases, if one of the quantile analysis spectral interpolations was disregarded for a given ChIcAGO source, the other is reasonable when considering the above criteria.

As a way to further explore the goodness of the spectral interpolations, the CIAO spectral fitting tool Sherpa was used to fit the X-ray spectrum of each ChIcAGO source with spectral parameters derived from quantile analysis in Tables 3 and 4. The $\chi ^{2}_{{\rm red}}$ was calculated for each ChIcAGO source where >40 X-ray counts were detected with ACIS-S. In many cases we found $\chi ^{2}_{{\rm red}} < 1$, which is not unexpected given the low number of X-ray counts detected, indicating that any reasonable model is a decent fit. However, there are also a few cases where $\chi ^{2}_{{\rm red}} > 2$, indicating that the spectral parameters do not adequately describe the spectrum. The values of $\chi ^{2}_{{\rm red}}$ corresponding to the quantile interpolated spectral parameters are included in Tables 3 and 4.

The parameters and fluxes resulting from the absorbed power-law and absorbed thermal bremsstrahlung spectral modeling of the ChIcAGO sources have also been included in Tables 3 and 4 so that they can be directly compared to the quantile analysis results. Best-fit absorbed power-law parameters were obtained for all the ACIS-S detected ChIcAGO sources with >50 X-ray counts (see Table 3). However, there were several cases where the absorbed thermal bremsstrahlung fitting was unsuccessful as the Sherpa modeling algorithm hit the hard maximum limit on the kT parameter (10 keV). These unsuccessful fits were not included in Table 4.

All of the parameters and absorbed fluxes derived from the Cash statistics spectral fitting agree within 3σ of those derived from the quantile analysis spectral interpolation, the majority of which agree within 1σ errors. The only exception is the kT parameter value for ChI J171910–3652_2. However, the unabsorbed fluxes were far less agreeable between the two techniques as there were several ChIcAGO sources for which this value differed by >3σ. These include ChI J183356–0822_2 and ChI J184738–0156_1 from the power-law spectral analysis and ChI J144547–5931_1, ChI J144701–5919_1, ChI J165646–4239_1, ChI J171910–3652_2, ChI J172050–3710_1, and ChI J185608+0218_1 from the bremsstrahlung spectral analysis. Both techniques therefore appear to be successful in constraining the spectral parameters and absorbed fluxes for each of the investigated ChIcAGO sources but less successful in constraining the unabsorbed fluxes. It should also be noted that the majority of the reduced Cash statistics from the spectral modeling were systematically higher than the corresponding $\chi ^{2}_{{\rm red}}$ derived from the spectral parameters interpolated through quantile analysis.

Infrared and optical follow-up were primarily performed on those ChIcAGO sources with >20 X-ray counts (see Table 2). In order to determine which optical and infrared sources are counterparts to ChIcAGO sources, we used a technique similar to that described by Zhao et al. (2005), using their Equation (11). If the separation between a ChIcAGO source's wavdetect position and its possible counterpart is less than the quadratic sum of their 3σ position errors and the 3σ Chandra pointing error,³³ then the X-ray and optical (or infrared) sources are likely to be associated. The 1σ position errors for all sources in 2MASS PSC and the GLIMPSE³⁴ catalogs are 01 (Skrutskie et al. 2006) and 03, respectively. USNO B has an astrometric accuracy of <025 (Monet et al. 2003). We have assumed that the error distributions of the Chandra observations, Chandra pointing, and USNO B Catalog are all Gaussian for the purposes of identifying possible counterparts to the ChIcAGO sources. While this assumption is not necessarily correct in every case, the full examination required to obtain the Gaussian errors would involve a very complicated approach. However, other Chandra Galactic plane X-ray surveys that search for multiwavelength counterparts assuming Gaussian errors for cross-correlation purposes have had successful results (e.g., ChaMPlane; Zhao et al. 2005). On the basis of these results we feel that Gaussian errors are an acceptable assumption for the purpose of identifying optical and infrared counterparts to the ChIcAGO sources.

The infrared properties, such as the names and magnitudes of any likely 2MASS or GLIMPSE counterparts, together with the NIR magnitudes (J, H, and K) obtained from Magellan PANIC observations, are listed in Table 2 along with the date of each observation. We assume that those ChIcAGO sources with no listed 2MASS (or PANIC) counterpart have 2MASS PSC limiting magnitudes J > 15.8, H > 15.1, and K > 14.3 (Skrutskie et al. 2006). In Table 5, optical magnitudes have also been provided for those 44 ChIcAGO sources with >20 X-ray counts that have optical counterparts in the USNO B1 Catalog or in one of the IMACS or MagIC Magellan observations. Two other ChIcAGO sources (ChI J170017–4220_1 and ChI J181213–1842_7) with magnitude limits obtained with either IMACS or MagIC are also included in Table 5.

Of the 74 ChIcAGO sources with >20 X-ray counts listed in Table 2, 59 have a NIR counterpart, 44 of which are 2MASS sources and 15 of which were detected in PANIC observations. Looking into the mid-infrared wavelength bands, we find that 41 of these 2MASS sources and 3 of the PANIC sources also have GLIMPSE counterparts. NIR magnitude limits were obtained for 4 other PANIC-observed ChIcAGO sources since any possible counterparts were too faint to be detected or, in the case of ChI J181116–1828_2 and ChI J185643+0220_2, the counterpart appeared to be blended. (ChI J181116–1828_2 does, however, have a unique GLIMPSE counterpart.) If we include ChI J181116–1828_2 and ChI J185643+0220_2, for which we have detected a counterpart but the magnitudes are only an upper limit because of blending, then 89% of our PANIC observations of ChIcAGO sources have yielded a detection in one of more of the J, H, or K filter bands. There are also a few ChIcAGO sources where the 2MASS counterpart magnitudes are listed as 95% confidence upper limits due to a nondetection or inconsistent deblending. We were therefore able to use PANIC to obtain more accurate magnitudes for four ChIcAGO sources that have limited 2MASS magnitude information in one or more bands. (These four ChIcAGO sources can be identified as those with three letters listed in the "Data" column of Table 2.)

All 46 ChIcAGO sources listed in Table 5 (44 ChIcAGO sources with optical counterparts and 2 with limiting magnitudes obtained with Magellan instruments) have NIR counterparts detected with either 2MASS or PANIC. Of the 44 with optical counterparts, 41 have USNO B1 counterparts, and for 4 of those, extra magnitude measurements were obtained with one of the two Magellan optical imagers. A further 3 ChIcAGO sources, not cataloged in USNO B1, were also detected in the optical with these Magellan instruments. Of the 74 ChIcAGO sources with >20 X-ray counts, 14 do not have a detected optical, NIR, or IR counterpart.

We conducted an experiment similar to that outlined in Kaplan et al. (2004) to quantify the probability that the optical and infrared survey counterparts quoted in Tables 2 and 5 are a random chance association with the >20 X-ray count ChIcAGO source with which they are coincident. We searched for all 2MASS, GLIMPSE, and USNO B1 sources brighter than the possible counterpart listed in Tables 2 and 5 within 10' of each ChIcAGO source position. (We only searched for survey sources that have a brighter K_s, 3.6 μm, and second epoch R band magnitude for the 2MASS, GLIMPSE, and USNO B1 surveys, respectively.) We then used the resulting statistics to determine the number of survey sources brighter than the listed counterpart that are likely to be detected within a region the same size as the ChIcAGO source's 95% position error circle. We did this for each ChIcAGO source individually as the density of sources can vary dramatically across the Galactic plane. In most cases the resulting chance of a random association is very low (<0.01). Those ChIcAGO sources that have a random chance of association >0.01 in either 2MASS or GLIMPSE are listed in Table 6. For each ChIcAGO source the chance of a random association with a USNO B1 is <0.01.

We refer to the ChIcAGO sources that have <20 counts as "secondary" sources. In Table 7 we list the names of any USNO B1, 2MASS, and GLIMPSE sources that appear to be coincident with a secondary ChIcAGO source on the basis of our position agreement criteria outlined in Section 2.4.1. Table 7 also includes the offset in arcseconds between the secondary ChIcAGO source's wavdetect position and the position of the coincident survey source. (Only those secondary ChIcAGO sources that have a coincident survey source have been included in Table 7.) A summary of the fraction of ChIcAGO sources with a coincident source in the 2MASS, GLIMPSE, and USNO B1 catalogs can be found in Table 8, which includes the fraction of the total number of ChIcAGO sources, as well the fraction of just the >20 X-ray count ChIcAGO sources, with a coincident survey source. (Note that we assume the survey sources coincident with the >20 X-ray count ChIcAGO sources are counterparts based on the very low chance of random associations, as demonstrated by Table 6.)

As mentioned in Section 2.4.2, the position of each ChIcAGO source (above and below 20 X-ray counts) was visually inspected for any possibly associated radio emission in the MGPS, MAGPIS, VGPS, and the 90 cm Multi-configuration Very Large Array Survey of the Galactic Plane. The results of this inspection are listed in Table 9. ChIcAGO source position comparisons were also made with the Green (2009) catalog of Galactic SNRs. Any possible counterparts that are known objects, such as SNRs, H ii regions, infrared dark clouds (IRDCs; Peretto & Fuller 2009), CWBs, or massive stars, are listed in Table 9 in the "Type" column. However, if the radio sources are uncataloged, they have instead been flagged as being possibly compact, diffuse, or arc/shell structured diffuse emission. Two ChIcAGO sources, ChI J181116–1828_5 and ChI J184741–0219_3, appear to be previously unidentified AGNs as their coincident radio sources show core–lobe morphologies in the MAGPIS 1.4 GHz survey images.

Each of the ChIcAGO sources with a possible radio association is then listed in Table 9 as being either coincident with, adjacent to, or on the limb of the radio source (such as on the limb of a SNR, diffuse emission, or H ii region). If a ChIcAGO source is listed as either coincident with or on the limb of a SNR, then this means it is within the extent of the SNR based on the SNR's size quoted in the Green (2009) catalog. In the cases of the two candidate AGNs, these ChIcAGO sources appear to be directly coincident with the core of the AGN. The name of each radio source and the corresponding reference are listed in Table 9.

The 10 ChIcAGO sources observed with the ATCA are also included in Table 9. The ATCA-detected compact radio counterparts to both ChI J144701–5919_1 and ChI J163252–4746_2 aided in their identification as X-ray emitting massive stars or CWBs (see Anderson et al. 2011). No radio counterparts were detected for the other 8 ChIcAGO sources observed with the ATCA.

In summary, Table 9 shows there are 16 ChIcAGO sources from the Chandra observations of 8 different AGPS source regions that are coincident with or on the limb of nine SNRs. There are 54 ChIcAGO sources from 7 different AGPS source regions that fall within the extent of 6 H ii regions. There are also four massive stars, all of which are confirmed or candidate CWBs (see Anderson et al. 2011; Motch et al. 2010), with radio counterparts. Only two ChIcAGO sources, both toward the same single AGPS target, are coincident with an IRDC.

Several ChIcAGO sources also fall within regions of uncataloged extended radio emission. These include 15 ChIcAGO sources from 5 different AGPS source regions falling within the extent of 5 regions of diffuse radio emission. Of these 15 ChIcAGO sources, there are 6 (from 3 different AGPS source regions) that are coincident with uncataloged diffuse emission with an arc or shell structure. Excluding the two AGN candidates, there is only one other ChIcAGO source coincident with an unidentified compact radio source.

There are 74 ChIcAGO sources (14 sources with >20 X-ray counts and 60 with <20 X-ray counts), out of the 253 detected, with no optical or infrared counterparts, making them possible compact object candidates and therefore potentially detectable in the radio. We therefore searched for any possible pulsar counterparts in the Australia Telescope National Facility Pulsar Catalogue (version 1.44³⁵; Manchester et al. 2005), but no known pulsars exist within 06 of the wavdetect position of any of the 74 sources.

Table 9 lists 16 ChIcAGO sources that fall within the extent of 9 SNRs. X-ray point sources within SNRs could be associated compact objects. Identification of an optical or infrared counterpart discounts such a possibility since the optical/IR counterparts to NSs and other compact objects are expected to be very faint (for the details on this approach, see Kaplan et al. 2004). Those SNR coincident ChIcAGO sources with a random chance of association >0.01 in the USNO B1 and GLIMPSE catalogs are listed in Table 10. (The chance of a random association between one of these 16 ChIcAGO sources and a 2MASS catalog source is <0.01.)

Of the 16 ChIcAGO sources inside SNRs, only 5 have no optical or infrared counterparts. These five X-ray sources are all very faint, with <8 X-ray counts detected for each in the ChIcAGO Chandra observations. These are ChI J145732–5901_2 in SNR G318.2+0.1 (Whiteoak & Green 1996), ChI J182435–1311_2,3,4 in SNR G18.1–0.1 (Helfand et al. 2006; Brogan et al. 2006), and ChI J184447–0305_1 in SNR G29.3667+0.1000 (Helfand et al. 2006). Bocchino et al. (2001) has already reported on three X-ray sources within SNR G318.2+0.1 but not at the position of ChI J145732–5901_2. Both SNR G18.1–0.1 and SNR G29.3667+0.1000 are newly discovered SNRs, so little X-ray analysis has been done on these objects. Further investigation is required to determine if any of these five ChIcAGO sources are compact objects and if they are associated with the surrounding SNRs.

There are 54 ChIcAGO sources coincident with 6 different H ii regions that were detected in the ChIcAGO Chandra observations of 7 AGPS sources (see Table 9). On the basis of the results from X-ray observations of other H ii regions (for example, see Broos et al. 2007) it is possible that many of these 54 ChIcAGO sources could be pre-main-sequence (PMS) stars, massive OB and WR stars, and CWBs. Simpson & Cotera (2004) also found optically obscured star clusters in 2MASS images within 1' of three of these AGPS sources; AX J144519–5949, AX J151005–5824, and AX J162208–5005, which supports a H ii region and young and massive star interpretation for these ASCA sources. Of these 54 ChIcAGO sources, 43 have optical and/or infrared counterparts, supporting a possible stellar origin (see Tables 2, 5, and 7). In most cases the chance of random association with a field source is low (<0.01). Those H ii region coincident ChIcAGO sources for which the chance of random association with a 2MASS or GLIMPSE source is >0.01 are listed in Table 11.

Table 11. Quantile Spectral Interpolations and Luminosity Estimates for All ChIcAGO Sources within H ii Regions That Have ≳5 X-Ray Counts

ChIcAGO Source	Absorbed Mekal Interpolationa				Distanceb		Luminositiesc		False Associations (× 10−2)d
ChI	kT	NH	Fx, abs	Fx, unabs	(kpc)	Ref	log Lx, abs	log Lx, unabs	2MASS N(< Ks)	GLIMPSE N(< M3.6)
J144519–5949_2	$1.52_{-0.73}^{+1.38}$	$5.00_{-2.10}^{+3.60}$	23.70 ± 4.80	218.44 ± 44.20	2.8	1	32.35	33.31
J144519–5949_3	$0.78_{-0.57}^{+2.69}$	$22.00_{-15.00}^{+51.00}$	13.05 ± 4.73	7207.60 ± 2610.31	2.8	1	32.09	34.83
J144519–5949_5	>1.94	$0.14_{-0.14}^{+0.68}$	3.37 ± 2.29	3.98 ± 2.71	2.8	1	31.50	31.57
J151005–5824_1	$0.90_{-0.77}^{+1.80}$	$7.60_{-4.60}^{+27.40}$	0.74 ± 0.45	47.76 ± 28.64	4.7	2	31.29	33.10
J151005–5824_2	82.45	$2.80_{-2.07}^{+0.00}$	1.72 ± 0.85	2.92 ± 1.45	4.7	2	31.66	31.89
J151005–5824_3	$0.21_{-0.11}^{+1.42}$	$1.60_{-1.35}^{+1.30}$	0.18 ± 0.12	65.24 ± 44.33	4.7	2	30.68	33.24
J151005–5824_4	82.45	$2.20_{-1.53}^{+-0.10}$	1.21 ± 0.73	1.96 ± 1.18	4.7	2	31.51	31.71
J151005–5824_5	82.45	$1.80_{-1.50}^{+0.00}$	1.14 ± 0.68	1.76 ± 1.06	4.7	2	31.48	31.67
J151005–5824_6	>0.1	$2.50_{-2.49}^{+10.50}$	0.44 ± 0.18	452.64 ± 182.20	4.7	2	31.07	34.08
J151005–5824_7	>2.21	$21.00_{-17.00}^{+11.00}$	6.46 ± 1.94	36.97 ± 11.10	4.7	2	32.23	32.99
J151005–5824_8	$0.17_{-0.07}^{+0.33}$	$33.00_{-21.00}^{+28.00}$	0.96 ± 0.48	218762449.44 ± 108233831.76	4.7	2	31.41	39.76
J151005–5824_9	>2.49	$0.01_{0.00}^{+0.25}$	0.41 ± 0.28	0.42 ± 0.28	4.7	2	31.04	31.04		3.15
J151005–5824_10	>5.26	$1.20_{-1.16}^{+1.00}$	1.39 ± 0.69	2.00 ± 0.99	4.7	2	31.56	31.72		1.71
J151005–5824_11	>0.1	$0.38_{-0.38}^{+0.87}$	0.64 ± 0.43	0.79 ± 0.54	4.7	2	31.23	31.32
J154905–5420_1	$0.61_{-0.33}^{+0.81}$	$11.50_{-5.60}^{+12.50}$	1.49 ± 0.68	743.78 ± 340.69	3.7	3	31.39	34.09
J154905–5420_2	>2.63	$2.30_{-0.80}^{+3.30}$	4.07 ± 1.35	6.63 ± 2.20	3.7	3	31.82	32.04
J154905–5420_3	$0.74_{-0.56}^{+1.54}$	$15.50_{-9.50}^{+39.50}$	0.99 ± 0.67	397.68 ± 270.25	3.7	3	31.21	33.81	2.40	2.21
J154905–5420_4	$1.80_{-1.69}^{+3.04}$	$0.01_{0.00}^{+1.29}$	0.40 ± 0.24	0.41 ± 0.25	3.7	3	30.82	30.83	1.62	1.88
J154905–5420_5	>0.1	$0.01_{0.00}^{+0.42}$	1.66 ± 0.71	1.69 ± 0.72	3.7	3	31.44	31.44
J154905–5420_7	>16.3	$1.05_{-1.04}^{+0.35}$	1.89 ± 0.87	2.66 ± 1.22	3.7	3	31.49	31.64
J154905–5420_8	>4.63	$1.20_{-1.19}^{+1.10}$	1.09 ± 0.74	1.57 ± 1.07	3.7	3	31.25	31.41		3.02
J154905–5420_9	>0.1	$0.01_{0.00}^{+0.60}$	0.70 ± 0.38	0.71 ± 0.38	3.7	3	31.06	31.07
J154905–5420_10	>0.1	$0.01_{0.00}^{+0.66}$	0.46 ± 0.31	0.47 ± 0.32	3.7	3	30.88	30.89	3.25	5.29
J154905–5420_11	>0.63	$0.46_{-0.45}^{+1.39}$	1.24 ± 0.47	2.05 ± 0.78	3.7	3	31.31	31.53
J162208–5005_1	>1.91	$2.40_{-1.10}^{+3.60}$	23.73 ± 6.53	42.69 ± 11.74	3.1	2	32.44	32.69
J162208–5005_3	>6.52	$2.60_{-1.40}^{+2.30}$	6.75 ± 4.59	11.30 ± 7.68	3.1	2	31.89	32.11
J194310+2318_1	$2.28_{-1.86}^{+6.97}$	$0.01_{0.00}^{+1.04}$	5.31 ± 1.46	5.47 ± 1.50	2.6	2	31.63	31.65
J194310+2318_4	$0.12_{-0.02}^{+19.33}$	$0.67_{-0.66}^{+0.00}$	1.06 ± 0.72	253.21 ± 172.07	2.6	2	30.93	33.31		1.06
J194310+2318_5	$0.14_{-0.03}^{+3.33}$	$0.76_{-0.75}^{+-0.03}$	5.97 ± 1.08	1446.16 ± 262.87	2.6	2	31.68	34.07
J194310+2318_6	$0.76_{-0.62}^{+1.38}$	$1.75_{-1.05}^{+3.85}$	5.33 ± 1.41	94.55 ± 24.98	2.6	2	31.63	32.88
J194310+2318_7	>0.19	$1.60_{-1.59}^{+10.90}$	1.70 ± 0.92	909.74 ± 491.76	2.6	2	31.14	33.87
J194310+2318_8	82.45	$0.94_{-0.93}^{+0.26}$	3.77 ± 2.56	5.22 ± 3.55	2.6	2	31.48	31.63
J194310+2318_9	$0.10_{0.00}^{+1.28}$	$1.70_{-1.69}^{+1.40}$	0.99 ± 0.59	18914.51 ± 11342.64	2.6	2	30.90	35.18
J194332+2323_1	$2.91_{-2.56}^{+7.09}$	$0.13_{-0.12}^{+1.52}$	1.65 ± 0.55	2.10 ± 0.69	2.6	2	31.13	31.23
J194332+2323_3	$0.22_{-0.11}^{+0.33}$	$3.00_{-1.30}^{+4.60}$	0.64 ± 0.35	1053.73 ± 569.60	2.6	2	30.72	33.93
J194332+2323_5	>0.1	$1.15_{-1.14}^{+10.35}$	3.81 ± 1.32	24.79 ± 8.57	2.6	2	31.49	32.30
J194332+2323_6	>0.1	$0.01_{0.00}^{+2.49}$	0.70 ± 0.38	0.73 ± 0.39	2.6	2	30.76	30.77	1.35	1.38
J194332+2323_7	$0.15_{-0.05}^{+0.55}$	$0.58_{-0.57}^{+.15}$	0.64 ± 0.29	51.52 ± 23.60	2.6	2	30.72	32.62
J195006+2628_1	$0.10_{0.00}^{+0.04}$	$1.25_{-0.73}^{+2.35}$	0.70 ± 0.38	3815.10 ± 2062.27	2.3	4	30.65	34.38
J195006+2628_2	$1.74_{-1.62}^{+2.02}$	$0.01_{0.00}^{+1.49}$	2.74 ± 0.55	2.83 ± 0.57	2.3	4	31.24	31.25
J195006+2628_3	$2.70_{-2.59}^{+16.75}$	$0.23_{-0.22}^{+1.87}$	0.61 ± 0.42	.88 ±.60	2.3	4	30.59	30.75
J195006+2628_4	$0.40_{-0.29}^{+1.23}$	$2.10_{-1.37}^{+3.50}$	0.63 ± 0.34	58.70 ± 31.73	2.3	4	30.60	32.57
J195006+2628_5	$0.11_{-0.01}^{+1.76}$	$1.55_{-1.54}^{+1.35}$	0.74 ± 0.28	5807.00 ± 2211.85	2.3	4	30.67	34.57

Notes. ^aThe quantile analysis interpolated absorbed Mekal model parameters including the temperature, kT, and absorption column density N_H (10²² cm⁻²), as well as the absorbed and unabsorbed X-ray fluxes, F_{x, abs} and F_{x, unabs} (10⁻¹⁴ erg cm⁻² s⁻¹), respectively, in the 0.3–8.0 keV energy range. (If kT = 82.45 then this indicates that the spectral interpolation of this ChIcAGO source has hit the hard limit for this parameter.) ^bThe distances used to estimate the absorbed and unabsorbed luminosities. The corresponding reference is also listed. ^cThe log of the absorbed and unabsorbed luminosities, L_{x, abs} and L_{x, unabs} (erg s^−s), respectively, in the 0.3–8.0 keV energy range. ^dChance of finding a 2MASS (or GLIMPSE) source with a K_s (or 3.6 μm) magnitude brighter than the counterpart listed in Tables 2 or 7, within the ChIcAGO source's 95% position error circle. This value is only quoted for the ChIcAGO sources where the chance of random alignment with a survey source is >0.01. References. (1) Busfield et al. 2006; (2) Russeil 2003; (3) McClure-Griffiths et al. 2001; (4) Fich & Blitz 1984.

Download table as:  ASCIITypeset image

The possible nature of these H ii region coincident ChIcAGO sources could be further investigated by comparing their luminosities to those of PMS stars, massive OB and WR stars, and CWBs. A spectral analysis of the 54 ChIcAGO sources is extremely difficult given that they all have ⩽32 X-ray counts. However, the primary goal is to obtain a wide-band flux that can then be converted into a luminosity. Quantile analysis was therefore performed to obtain absorbed Mewe–Kaastra–Liedahl (Mekal; Mewe et al. 1985, 1986; Kaastra 1992; Liedahl et al. 1995) spectral interpolations of all the H ii region coincident ChIcAGO sources with ⩾5 X-ray counts. A Mekal model was chosen as thin thermal plasma emission is expected from hot X-ray emitting stars (for example, see Wolk et al. 2005; Sana et al. 2006). The absorbed Mekal spectral interpolations can be found in Table 11.

The luminosities of these ChIcAGO sources were calculated using kinematic distance estimates to the H ii regions with which they are coincident. Kinematic distances calculated by Russeil (2003) were used to calculate luminosities for the ChIcAGO sources coincident with G320.3–0.3, G333.6–0.2, and G59.5–0.2. (Note that there is a more distant kinematic distance estimate of 6.3 kpc to G59.5–0.2 calculated by Kuchar & Bania 1994, but we have decided to use the more recent estimate from Russeil 2003.) The kinematic distances used in the luminosity calculations for G326.96+0.03 and G62.9+0.1 were obtained from McClure-Griffiths et al. (2001) and Fich & Blitz (1984), respectively. The kinematic distance to the massive young stellar object G316.8112–00.0566 (Busfield et al. 2006), likely embedded within GAL 316.8–00.1, was used in the luminosity calculations for those ChIcAGO sources coincident with this H ii region. (The distance from Busfield et al. 2006 agrees reasonably well with the near kinematic distance to GAL 316.8–00.1 calculated by Caswell & Haynes 1987 when revised for a modern Galactic center distance of 8.5 kpc.) Table 11 lists the kinematic distance and corresponding absorbed and unabsorbed luminosities calculated for each H ii region coincident ChIcAGO source.

The range of absorbed luminosities calculated for all the ChIcAGO sources coincident with the 6 H ii regions span the range 30.6 erg s⁻¹ < log L_{x, abs} < 32.4 erg s⁻¹ (0.3–8 keV). This absorbed luminosity range is similar to that observed from the H ii region M17 (29.3 erg s⁻¹ < log L_{x, abs} < 32.8 erg s⁻¹ (0.5–8 keV); Broos et al. 2007). With the exception of ChI J151005–5824_8, the H ii region coincident ChIcAGO sources in Table 11 have unabsorbed luminosities between L_{x, unab} ∼ 10³¹ and 10³⁵ erg s⁻¹. The unabsorbed luminosities of these ChIcAGO sources cover the ranges of what has been observed from flaring PMS stars (L_{x, unab} ∼ 10³⁰ to 10³³ erg s⁻¹; Favata et al. 2005; Wolk et al. 2005), single and binary massive O-type stars (L_{x, unab} ∼ 10³¹ to 10³³ erg s⁻¹; Oskinova 2005; Sana et al. 2006), WR stars (L_{x, unab} ∼ 10³¹ to 10³⁴ erg s⁻¹; Oskinova 2005; Mauerhan et al. 2010), and CWBs (L_{x, unab} ∼ 10³² to 10³⁴ erg s⁻¹; Oskinova 2005; Mauerhan et al. 2010). In fact, we were able to identify ChI J194310+2318_5, which is within G59.5–0.2, as the O7V((f))-type star HD 344784 (Walborn 1973) using the SIMBAD Astronomical Database. It is therefore likely that the AGPS sources AX J144519–5949, AX J151005–5824, AX J154905–5420, AX J162208–5005, AX J194310+2318, AX J194332+2323, and AX J195006+2628 are young and massive stars within H ii regions. Deeper X-ray, radio, and IR observations are required to determine the precise nature of the individual ChIcAGO sources.

The X-ray and infrared population statistics performed in this section are conducted using just those ChIcAGO sources with >20 X-ray counts, therefore concentrating on the persistent populations that fall within the AGPS flux range (F_x ∼ 10⁻¹³ to 10⁻¹¹ erg cm⁻² s⁻¹). However, discarding the ChIcAGO sources with <20 X-ray counts limits our analysis as we are excluding the populations of sources that exhibit long-term variability or transient behavior. Sources with <20 X-ray counts will need to be investigated in future work using archival X-ray observations at different epochs.

4.2.1. X-ray and Infrared Populations Statistics

Using the X-ray and infrared properties of the unidentified ChIcAGO sources, it is possible to classify some of the sources detected in the AGPS into possible populations. We chose to focus on the IR counterparts as this wave band is less affected by interstellar extinction when compared to the optical band. This group of sources therefore makes up a larger subset of the unidentified ChIcAGO sources than those with optical counterparts (see Section 3.3). For those unidentified ChIcAGO sources observed with the Chandra HRC instrument (for which there is no spectral information), we generated fake spectra, using XSpec and the CIAO spectral fitting tool Sherpa. These spectra are based on the absorbed power-law fits reported in Sugizaki et al. (2001), allowing the absorbed X-ray flux and median energy (E₅₀) of the unidentified ChIcAGO source in question to be calculated. These values are used in the statistical plots described below.

To help identify possible distinct populations in the statistical plots, we have also included both archival sources (those AGPS source that were identified by Sugizaki et al. 2001 or in the literature and so were not observed with Chandra as part of the ChIcAGO survey) and the previously identified ChIcAGO sources (for example, those investigated by Anderson et al. 2011, 2012). Fake spectra were generated using Sherpa and XSpec for these archival sources, using spectral fits in the literature, to determine their absorbed X-ray fluxes and E₅₀ values for the energy ranges investigated. All archival AGPS sources are summarized and tabulated in Section 4.5 (see Table 12) and are individually described in Appendix A.

Table 12. Confirmed and Tentative Identifications of the AGPS Sources, Including Both Literature and ChIcAGO Survey Identifications

AGPS ID	ChIcAGO IDa	Typeb	IDc	Referenced	Flage
AX J143148–6021	ChI J143148–6021_3	U		1	n,R,n
AX J143416–6024		RS CVn	HD 127535	2,3*	n,n,n
AX J144042–6001	ChI J144042–6001_1	PMS	HD 128696	2*	n,n,n
AX J144519–5949	ChI J144519–5949_2(1,3-6)	WR and H ii	GAL 316.8–00.1	1	T,T,N
AX J144547–5931	ChI J144547–5931_1	OIf+		4,4	I,I,n

Notes. ^aThe equivalent ChIcAGO name of the AGPS source. All ChIcAGO sources with >20 X-ray counts are listed. In a few cases there is more than one ChIcAGO source listed for a given AGPS source. For a given identified H ii region, all of the ChIcAGO sources contributed to the X-ray emission originally detected with ASCA in the AGPS. In these cases the ChIcAGO source name suffixes are listed in parentheses. If a suffix is listed before the parentheses then this ChIcAGO source corresponds to an X-ray source whose stellar type is listed in column three, before the H ii classification. ^bThe source type abbreviations are ASC: active stellar corona, AGN: active galactic nucleus, CV: cataclysmic variable, CWB: colliding wind binary, H ii: young and massive stars in the H ii region named in the "ID" column; HMXB: high-mass X-ray binary, LBV: luminous blue variable, LMXB: low-mass X-ray binary, Magnetar: magnetar, MS: massive star, MS-O: massive O-type star, ND: no ChIcAGO sources detected in the Chandra observation (no detection), PMS: pre-main sequence star, PSR: X-ray emission from a rotation-powered pulsar, PWN: pulsar wind nebula, SNR: supernova remnant, SyXB: symbiotic X-ray binary, U: unknown, and WR: Wolf–Rayet star. The other abbreviations are the spectral type of the stellar counterpart. ^cThe most commonly used name ^dThe * and ** symbols indicate those AGPS sources that were correctly and incorrectly identified by Sugizaki et al. (2001), respectively. Those archival AGPS sources and identified ChIcAGO sources used in the statistical analysis in Section 4.2 have two references in this Table, the first being the paper describing an X-ray fit from which we derived an absorbed X-ray flux, and the second being the paper from which the NIR counterpart (J, H, and K_s magnitudes) information was obtained. All the longer-wavelength IR counterpart information (3.6, 4.5, 5.8, and 8.0 μm magnitudes) was gathered from the GLIMPSE catalogs and archives. ^eFlags that correspond to the identification properties of the AGPS and ChIcAGO sources. One flag is listed for each of the first three columns. I: identified sources through work in the ChIcAGO survey; T: tentatively identified sources using the X-ray and infrared population statistics in Section 4.2; F: AGPS sources for which only ChIcAGO sources with ⩽20 X-ray counts were detected. These include AGPS sources where no sources were detected in the corresponding ChIcAGO Chandra observations but excludes those AGPS sources that have been identified as H ii regions. R: Figure 8 Region iv sources; N: unconfirmed classification type; n: no flag. References. (1) This paper; (2) Sugizaki et al. 2001; (3) Skrutskie et al. 2006; (4) Anderson et al. 2011; (5) Degenaar et al. 2012; (6) Bernardini et al. 2011; (7) Israel et al. 2009; (8) Torres et al. 2006; (9) Tomsick et al. 2006; (10) Anderson et al. 2012; (11) Rodriguez et al. 2006; (12) Walter et al. 2006; (13) Nazé et al. 2008; (14) Combi et al. 2006; (15) Bodaghee et al. 2006; (16) Funk et al. 2007; (17) Hamaguchi et al. 2005; (18) Kaur et al. 2010; (19) Markwardt et al. 2010; (20) Chakrabarty et al. 2002; (21) Combi et al. 2010; (22) Lazendic et al. 2005; (23) Gaensler et al. 2008; (24) Giacani et al. 2009; (25) De Becker et al. 2004; (26) Yamauchi et al. 2008; (27) Rho et al. 2004; (28) Mereghetti et al. 2005; (29) Israel et al. 2005; (30) Brogan et al. 2006; (31) Kargaltsev & Pavlov 2007; (32) Gotthelf & Halpern 2007; (33) Israel et al. 2004; (34) Mereghetti et al. 2012; (35) Helfand et al. 2003a; (36) Bassani et al. 2009; (37) Motch et al. 2010; (38) Kargaltsev et al. 2012; (39) Kaplan et al. 2007; (40) Gotthelf & Halpern 2008; (41) Paron et al. 2012; (42) Morii et al. 2003; (43) Bamba et al. 2001; (44) Helfand et al. 2003b; (45) Pooley et al. 2007; (46) Chen et al. 2004; (47) Yamaguchi et al. 2004; (48) Gotthelf & Halpern 2005; (49) Petre et al. 2002; (50) Yamauchi et al. 2011; (51) Safi-Harb et al. 2005; (52) Pavan et al. 2011; (53) Hwang et al. 2000; (54) Kohoutek & Wehmeyer 1997; (55) Zolotukhin & Chilingarian 2011.

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:  DataTypeset image

The identified sources were divided into the following categories: AGNs, CVs, CWBs, high-mass X-ray binaries (HMXBs), magnetars, massive stars, and stars. The "AGN" category includes ChI J184741–0219_3, which we identified by positional comparison with MAGPIS radio data (for the radio identification and further details on this source, see Sections 3.4 and 4.3.15, respectively). The "CWB" category includes AX J163252–4746 and AX J184738–0156, which were identified in Anderson et al. (2011, listed as ChI J163252–4746_2 and ChI J184738–0156_1 in Table 1, respectively), and AX J183116–1008 and AX J183206–0938, which were identified in the XGPS (Motch et al. 2010, listed as ChI J183116–1008_1 and ChI J183206–0938_1 in Table 1, respectively). The "HMXB" category includes the archival AGPS sources that are supergiant HMXBs (McClintock & Remillard 2006), supergiant fast X-ray transients (SFXTs; Sguera et al. 2006), and SyXBs (Masetti et al. 2007). The infrared and X-ray fluxes from magnetars are variable and correlated (Durant & van Kerkwijk 2005), so the fluxes we used in the "magnetar" category are from infrared and X-ray observations that occurred close together in time. PSR J1622–4950, which was determined through the ChIcAGO Chandra observation to be the main contributor to AX J162246–4946, has also been included as an identified magnetar (see Anderson et al. 2012 and Section 4.3.6 for further details). The sources included in the "massive star" category are massive stars that are WR, luminous blue variable (LBV) stars, and massive O-type stars, which emit X-rays through instability-driven wind shocks (Lucy & White 1980; Lucy 1982) and possibly through colliding winds in a CWB. These include AX J144547–5931 and AX J144701–5919, which were identified by Anderson et al. (2011, listed as ChI J144547–5931_1 and ChI J144701–5919_1 in Table 1, respectively). All other nondegenerate stars are in the "star" category and most likely correspond to ASCs or PMS stars.

We first investigated the relationship between the X-ray and infrared flux of the ChIcAGO sources by comparing these properties to those of known stars and AGNs. Figure 5 of Gelfand & Gaensler (2007) shows the X-ray versus K_s-band flux of sources from the Chandra Orion Ultradeep Project (COUP; Getman et al. 2005) and XBootes Survey (Jannuzi et al. 2004; Kenter et al. 2005), which are stars (predominantly in the PMS) and AGNs, respectively. In Figure 7 we create a similar plot that includes the unidentified ChIcAGO sources (U ChIcAGO; red data points) along with the COUP stars (blue crosses) and the XBootes Survey AGNs (magenta diamonds; for further details on the data from these surveys see Gelfand & Gaensler 2007, and references therein). The X-ray flux (F_{x, 2–7 keV}) is over the 2.0–7.0 keV energy range, and the K_s band flux (F_Ks = λF_{λ, Ks}) is derived from F_{λ, Ks} (erg cm⁻² s⁻¹μm⁻¹) where the effective wavelength is λ = 2.159 μm. The archival sources and the identified ChIcAGO sources with K-band counterparts have also been included in Figure 7 in order to further distinguish between possible X-ray populations.

Zoom In Zoom Out Reset image size

Two groups and two outliers are apparent in Figure 7 among the unidentified ChIcAGO sources: group 1, those coincident with the COUP stars, and group 2, which sits to the right of the AGN detected in the XBootes survey. The two outliers are ChI J154557–5443_3, on the very left of Figure 7 with a hard X-ray flux limit of F_{x, 2–7 keV} ≲ 3 × 10⁻¹⁹ erg cm⁻² s⁻¹,³⁶ and the source on the bottom right, ChI J153818–5541_1, with an X-ray flux of F_{x, 2–7 keV} ∼ 1 × 10⁻¹¹ erg cm⁻² s⁻¹, located near the identified magnetars.

Group 1 is distributed similar to the COUP stars, following a "track" indicating an increase in K-band flux with X-ray flux. This comparison demonstrates that a large number of the ChIcAGO stellar population could be PMS stars. While the overall Galactic X-ray population is dominated by field stars, we would expect that the ChIcAGO survey (and the AGPS) is biased toward PMS stars as such objects are brighter and harder X-ray emitters (for example, see Wolk et al. 2005). The identified ChIcAGO and archival sources that have been categorized as massive stars (light pink dots) and CWBs (green data points) congregate near the top right of group 1, beyond most of the COUP stars. These massive stars and CWBs are composed of massive late-type WR stars in the nitrogen sequence that are hydrogen rich (WNH; Smith et al. 2008) or their massive O star progenitors (Of; Crowther et al. 1995), all of which are expected to be bright in the infrared and potentially harder in X-rays than other types of X-ray emitting stars (yellow data points). (See Anderson et al. 2011 for further details on these WR and massive O stars.) It is therefore possible that the unidentified ChIcAGO sources located near these massive stars and CWBs in Figure 7 could be similar objects. It should also be noted that the identified AGPS HMXBs sit to the right of group 1, with no unidentified ChIcAGO sources near their positions with which to draw a comparison.

Group 2 sits at a similar K_s-band flux but higher X-ray flux than the AGN from the XBootes survey. The identified ChIcAGO and archival AGNs are coincident with the unidentified ChIcAGO sources in group 2, indicating that at least part of this group could also be AGNs. Such AGNs would have to be very X-ray bright in order to be detected through the high foreground column density in the Galactic plane. The identified archival CVs sit adjacent to group 2, at a slightly higher X-ray flux, indicating another possible population identification.

The unidentified ChIcAGO source ChI J154557–5443_3 has a K-band flux similar to the COUP stars but has an extremely faint hard X-ray flux. No identified ChIcAGO or archival sources are located near ChI J154557–5443_3 in Figure 7 that suggest a clue to its nature. ChI J154557–5443_3 is discussed further in Section 4.3.4.

The unidentified ChIcAGO source that sits at the bottom right of Figure 7, ChI J153818–5541_1, is very faint in the near-infrared but bright in high energy X-rays. It is therefore very similar in both X-ray and K_s-band fluxes to the identified archival magnetars, suggesting a similar identification that warrants further investigation. ChI J153818–5541_1 is further discussed in Section 4.3.3.

In order to further identify the possible populations that make up the above-described groups, we created a statistical plot that also takes into account the hardness of each of the unidentified ChIcAGO sources. Figure 8 plots the X-ray to K_s-band flux ratio (F_{x, 0.3–8.0 keV}/F_Ks) versus the median energy (E₅₀ keV), in the 0.3–8.0 keV energy band, of each unidentified ChIcAGO source ("U ChIcAGO"; red data points). We further separated the ChIcAGO sources into the categories "low," "medium," and "high" on the basis of their ACIS-S and HRC-I X-ray count rates, using symbol sizes to distinguish these categories. The smallest data points (low) have count rates of <15 counts s⁻¹, the medium-sized data points (medium) have count rates between 15 and 40 counts s⁻¹, and the largest data points (high) have count rates of >40 counts s⁻¹. The identified ChIcAGO and archival sources, also divided into categories on the basis of the X-ray count rate expected from a Chandra ACIS-S observation, have also been included in Figure 8. As the X-ray fluxes and median energy errors are fundamentally based on the number of X-ray counts detected in the Chandra observations, we show two representative error bars for the low-count ChIcAGO sources (20–60 counts) and the high-count ChIcAGO sources (80–120 counts). The error associated with the K_s-band flux is greater for PANIC magnitudes than for 2MASS magnitudes. We therefore adopt the PANIC K_s-band flux errors so that the vertical error bars represent the maximum possible error in the F_x/F_Ks ratio for low- and high-count unidentified ChIcAGO sources.

Zoom In Zoom Out Reset image size

In Figure 8, group 1 and source ChI J154557–5443_3 from Figure 7 have flux ratios F_x/F_Ks < 0.1 but have a wide range of median energies. Group 2 is harder than group 1, with E₅₀ > 1.5 keV but with a flux ratio 0.1 < F_x/F_Ks < 20. The unidentified ChIcAGO source ChI J153818–5541_1 from Figure 7 is quite hard (E₅₀ > 3 keV), with F_x/F_Ks ≈ 300. Using Figure 7 as a guide, as well as the relative positions between the unidentified ChIcAGO sources and the identified ChIcAGO and archival sources in Figure 8, we divided the groups and isolated sources from Figure 7 into six different population regions. These regions are marked by dashed lines and labeled with Roman numerals in Figure 8. The region boundaries in this plot are based on the observed X-ray properties and K_s band flux. If these values were extinction and absorption corrected, both the F_x/F_Ks and E₅₀ region boundaries would lower.

Region i (E₅₀ < 1.5 keV and F_x/F_Ks < 0.1) in Figure 8 contains unidentified ChIcAGO sources with low, medium, and high count rates. The majority of the low-count-rate unidentified ChIcAGO sources included in Figure 8 fall into this region. These low-count-rate sources could be very nearby objects that were only detected because of their close proximity to the solar system, or they are simply a more distant version of the medium- and high-count-rate sources in Region i. All unidentified ChIcAGO sources in this region are unlikely to be extragalactic given their softer X-ray spectra and bright K_s-band counterparts compared to the XBootes AGNs in Figure 7. Within this region are three of the archival X-ray stars (yellow data points), of which there is an RS CVn star, a PMS star, and a multiple system. The unidentified ChIcAGO sources in Region i are therefore likely to be soft X-ray stars with ASCs or PMS stars, similar to the three archival stars but at a variety of distances in the Galaxy.

The sources in Region ii (E₅₀ > 1.5 keV and F_x/F_Ks < 0.1) of Figure 8 have similar F_x/F_Ks ratios but slightly harder X-ray emission compared to those in Region i. This region also has fewer unidentified ChIcAGO sources than Region i, implying that it may contain a slightly rarer X-ray source population. The majority of the unidentified ChIcAGO sources in Region ii have low or medium count rates, with only one in the high-count-rate category. The identified ChIcAGO and archival sources in this region are CWBs (green data points) and massive stars (pink dots), all of which sit at the top of the stellar track in Figure 7. We defined E₅₀ > 1.5 keV as the lower-energy cutoff of Region ii by looking at the unidentified ChIcAGO sources closest to the CWBs and massive stars in Figure 7. The CWBs and massive stars are WNH and Of stars, which produce X-rays through instability-driven wind shocks but can also, in the case of CWBs, produce hard X-rays due to colliding winds (for example, see Anderson et al. 2011). It is therefore likely that many of the unidentified ChIcAGO sources in Region ii are WR and Of stars, some of which may also be CWBs.

Region iii (E₅₀ > 4.0 keV and 1 × 10⁻³ < F_x/F_Ks < 20) in Figure 8 encompasses the archival HMXBs (gray dots) and the archival and identified ChIcAGO AGNs (cyan dots). The only two unidentified ChIcAGO sources in this region are ChI J170017–4220_1 (high) and ChI J172550–3533_1 (low) and are therefore quite hard X-ray sources, making HMXB or AGN identifications a strong possibility.

Region iv (1.5 keV < E₅₀ < 4.0 keV and 0.1 < F_x/F_Ks < 10) in Figure 8 contains 12 medium- and high-count-rate unidentified ChIcAGO sources. However, there are no identified ChIcAGO or archival sources in this region that could indicate any likely source populations. The best clue comes from Figure 7, which shows that these unidentified sources are in the same region of the plot as the identified ChIcAGO and archival AGNs. In the Galactic plane, the log N–log S relation of X-ray sources in the 2.0–10.0 keV energy range (see Figure 15 of Hands et al. 2004) demonstrates that within the X-ray flux range 1 × 10⁻¹³ < F_x < 2 × 10⁻¹² of Region iv, between 0.06 and 8 extragalactic sources are expected per square degree. These number densities are consistent with more recent log N–log S modeling conducted by Mateos et al. (2008), who used 1129 XMM observations at |b| > 20° to demonstrate that sources in the 2–10 keV energy range at high Galactic latitudes agree with AGN models to better than 10%. As the ChIcAGO sources in Region iv have a number density <8 deg⁻², it is not unreasonable to speculate that many of the unidentified ChIcAGO sources in this region could be AGNs. However, we compared the N_H values of the Region iv ChIcAGO sources calculated from the power-law quantile analysis and spectral fits in Table 3 to the Galactic column densities in their direction from surveys conducted by Kalberla et al. (2005) and Dickey & Lockman (1990).³⁷ In all but four cases the ChIcAGO sources have N_H values an order of a magnitude lower than the Galactic N_H, suggesting a possible Galactic origin. ChIcAGO sources ChI J181116–1828_2, ChI J181213–1842_7, ChI J190749+0803_1, and ChI J194152+2251_2 have N_H values of the same order as the Galactic column density ($N_{{\rm H}} = 1.2 \times 10^{22}, 1.3 \times 10^{22}, 1.5\times 10^{22}, \rm { and }\ 1.1\times 10^{22}\, \rm { cm}^{-2}$, respectively; Kalberla et al. 2005), which is more indicative of an extragalacitc origin and therefore an AGN identification. XMM detections of ChI J181116–1828_2 and ChI J181213–1842_7 were also investigated by Cackett et al. (2006, who referred to these sources by their AGPS names AX J1811.2–1828 and AX J1812.2–1842). They derived absorbed power law spectral fits from these data and obtained parameters that agree within the 1σ errors that we derived from quantile analysis and Cash statistics spectral modeling to the Chandra data. However, their bestfit N_H values are lower than ours and therefore lower than the Galactic value. Cackett et al. (2006) therefore suggested a Galactic origin for these two sources proposing a possible HMXB, low-mass X-ray binary (LMXB) or CV identification for AX J1811.2–1828 (ChI J181116–1828_2) and a possible CV origin for AX J1812.2–1842 (ChI J181213–1842_7). Further multi-wavelength investigations are required to confirm the nature of the Region iv population.

Region v (1.5 keV < E₅₀ < 4.0 keV and 10 < F_x/F_Ks < 50) of Figure 8 encompasses the two identified archival CVs and the lower F_x/F_Ks ratio limit of unidentified ChIcAGO source ChI J181852–1559_2. As its flux ratio is only a lower limit (corresponding to an upper limit on the K-band flux), it is possible that ChI J181852–1559_2 may instead be a member of Region vi (described below).

Region vi (1.3 keV < E₅₀ < 4.0 keV and F_x/F_Ks > 1 × 10²) contains four identified archival magnetars. The only unidentified ChIcAGO source that sits within this region is ChI J153818–5541_1. On the basis of its proximity to these magnetars, it is possible that ChI J153818–5541_1 could also be a magnetar; however, its X-ray emission is much harder in comparison. Regardless, the position of ChI J153818–5541_1 in Figure 8 indicates that this source is definitely worthy of further study.

4.2.2. Infrared Population Statistics

In order to further refine the possible populations within Figure 8, we investigated the infrared colors of the unidentified ChIcAGO sources, once again drawing comparisons with the identified ChIcAGO and archival sources. As mentioned above, there is strong evidence in Figure 8 that some of the unidentified ChIcAGO sources in Region ii could be massive stars such as WR and Of stars and perhaps even CWBs. It is also possible that many of the ChIcAGO sources, particularly in Region i, are PMS stars given their X-ray to infrared flux ratio is similar to that of the COUP stars in Figure 8. However, for the purposes of this paper we have chosen to concentrate on the selection criteria for the hard X-ray emitting massive stars. We therefore leave the investigation of the PMS star population in the ChIcAGO survey for future work (for infrared selection criteria for PMS stars, see Favata et al. 2005; Maercker & Burton 2005; Maercker et al. 2006).

Hadfield et al. (2007) created a selection criterion for WR stars using GLIMPSE and 2MASS magnitudes that was further refined by Mauerhan et al. (2011). Figures 9(a) and (b) are recreations of the Hadfield et al. (2007) [3.6]–[4.5] versus [3.6]–[8.0] and J − K_s versus K_s − [8.0] color–color plots, showing the unidentified ChIcAGO sources. (The numbers in brackets correspond to the effective wavelength in microns of the GLIMPSE magnitude bands.) The dashed lines in Figures 9(a) and (b) indicate the color space used by Mauerhan et al. (2011) to select WR candidates. These color spaces indicate where WR stars are expected to fall in comparison to field stars because of the infrared excess of WR stars resulting from free-free emission generated in their strong, dense stellar winds (see Mauerhan et al. 2011, and references therein).

Zoom In Zoom Out Reset image size

In Figures 9(a) and (b), the unidentified ChIcAGO sources (UChI) have been separated into different colored symbols based on whether they are in Region i (red), Region ii (blue), or Region iii (black). In both plots, the majority of unidentified ChIcAGO sources, particularly from Region i, do not have an infrared excess, so they cluster together near the origin where Hadfield et al. (2007) state the general stellar locus is located. These unidentified ChIcAGO sources are therefore unlikely to have strong stellar winds, so they are more likely to have ASCs. There are, however, several Region ii sources and three Region i sources that have more unusual colors, falling within, very close to, or below the indicated WR color spaces. (It should be noted that many of these unusually colored unidentified ChIcAGO sources only have magnitude lower limits in one or more of the filter bands, making their positions in these color–color diagrams uncertain. See Figures 9(a) and (b) for details on these magnitude-limited sources.)

Rather than keeping the original identified source categories, we have instead separated the identified ChIcAGO and archival massive stars and CWBs into categories based on their dominant stellar components. These categories are WR and LBVs (green data points), massive O-type stars (pink data points), and all other nondegenerate stars (yellow data points). The HMXBs are still indicated by gray data points.

The identified WR stars fall within one or both of the Mauerhan et al. (2011) WR color spaces in Figures 9(a) and (b). On the basis of their positions within one or both WR color spaces, it is possible that unidentified Region ii ChIcAGO sources ChI J181915–1601_2, ChI J180857–2004_2, and ChI J183345–0828_1 may also be WR stars. In fact, using this color space technique, Hadfield et al. (2007) identified the 2MASS counterpart of ChI J181915–1601_2, 2MASS J18192219-1603123, as a WR star and further classified it as a WN7o star using spectroscopy. There are also three Region i unidentified ChIcAGO sources, ChI J154557–5443_1, ChI J154557–5443_3, and ChI J172550–3533_2, whose magnitude limits fall within the WR color space of Figure 9(b) and below the WR color space of Figure 9(a), which may also be worth further investigation.

The identified ChIcAGO source AX J144547–5931 (an Of-type star, which is also shown as a pink dot in Figures 7 and 8; see Anderson et al. 2011) sits to the left of both WR color spaces in Figures 9(a) and (b). Unidentified Region ii ChIcAGO sources ChI J182435–1311_1, ChI J182651–1206_4, ChI J183356–0822_3, and ChI J184652–0240_1 are also located in the same vicinity as AX J144547–5931 in one or both plots (within 0.2 < [3.6] − [8.0] < 0.5 in Figure 9(a) and within 0.5 < K_s − [8.0] < 1.2 in Figure 9(b)). While these four unidentified ChIcAGO sources have unusual infrared colors, they are not as extreme as those of typical WR stars. Given their proximity in Figure 9 to the Of star AX J144547–5931, it is possible that these unidentified ChIcAGO sources could be similar massive O-type stars. The remaining Region ii ChIcAGO sources lie within the general stellar locus of Figures 9(a) and (b), so they could be massive stars, ASCs, or PMS stars.

The archival HMXBs also have unusual infrared colors as they fall within or above the top edge of the WR color spaces in Figures 9(a) and (b). These infrared colors may be due to intrinsic absorption (for example, see Rodriguez et al. 2006). Several unidentified Region ii ChIcAGO sources also lie close to the HMXBs in these plots, but as they have much softer median X-ray energies (see Figure 8), such an identification is unlikely. ChI J170017–4220_1 is the only Region iii source with sufficient infrared magnitude information to be displayed in Figures 9(a) and (b). This unidentified ChIcAGO source falls within the vicinity of the HMXBs in Figure 8 but has much more extreme infrared colors and lies above the WR color space in both color–color plots. Further investigation is required to confirm the X-ray binary nature of ChI J170017–4220_1.

None of the 11 unidentified Region iv ChIcAGO sources have been detected at 8 μm, with only 3, ChI J163751–4656_1, ChI J165420–4337_1, and ChI J190818+0745_1, detected with GLIMPSE at 3.6 and 4.5 μm, all of which are likely to be Galactic in origin (see Section 4.2.1). In Figure 9(b) we have plotted these three unidentified ChIcAGO sources using their 4.5 μm magnitude in place of 8 μm magnitude. Only ChI J165420–4337_1 falls within the WR color space. Given the extreme uncertainties associated with the K_s − [8.0] color of ChI J165420–4337_1, no conclusions can be drawn about its possible nature.

We have flagged several of the unidentified ChIcAGO sources as interesting and worthy of future follow-up on the basis of their X-ray and infrared properties explored in Section 4.2. These population statistics have also allowed us to make some tentative identifications. The details of these sources are described below. There are several unidentified ChIcAGO sources that are not detailed in this section but are listed, along with their tentative identification based on the population statistics, in Table 12.

4.3.1. ChI J144519–5949_2

4.3.2. ChI J150436–5824_1

4.3.3. ChI J153818–5541_1

4.3.4. ChI J154557–5443_3

4.3.5. ChI J162011–5002_1

4.3.6. AX J162246–4946

4.3.7. AX J165951–4209

4.3.8. ChI J170017–4220_1

4.3.9. ChI J172050–3710_1

4.3.10. ChI J180857–2004_2

4.3.11. ChI J181116–1828_5

4.3.12. ChI J181852–1559_2

4.3.13. ChI J181915–1601_2

4.3.14. ChI J183345–0828_1

4.3.15. ChI J184741–0219_3

4.3.16. AX J185905+0333

4.3.17. AX J191046+0917

4.3.18. ChI J194939+2631_1

We were also able to identify four of the bright (>20 X-ray counts) ChIcAGO sources using the SIMBAD Astronomical Database. These four ChIcAGO sources are described below. We also discuss the secondary ChIcAGO source ChI J165420–4337_2.

4.4.1. ChI J144042–6001_1

4.4.2. ChI J165420–4337_2

4.4.3. ChI J172550–3533_2

4.4.4. ChI J172642–3540_1

4.4.5. ChI J194310+2318_5

Table 12 reproduces the original list of the 163 AGPS sources from Sugizaki et al. (2001), now including the corresponding ChIcAGO sources with >20 X-ray counts in the second column. The confirmed or tentative identifications of these ChIcAGO sources are listed in the third column, where the abbreviations are given in the table notes. Those AGPS sources and corresponding ChIcAGO sources with confirmed identifications either were obtained from the literature, and are therefore called "archival AGPS" sources, or were obtained through the ChIcAGO survey's Chandra observations (i.e., this paper and Anderson et al. 2011, 2012). The tentative identifications were made through the multiwavelength follow-up and population statistics conducted on the ChIcAGO sources in Section 4.2. The fourth column gives the most common name for the AGPS and/or ChIcAGO source, while the fifth column lists the references from which the X-ray and infrared properties of a given source were obtained for use in the statistical plots (Figures 7, 8, and 9).

The tentative identifications of the ChIcAGO sources are based on the statistical plots in Section 4.2. For example, if a ChIcAGO source falls within Region i of Figure 8, then its type in Table 12 has been listed as being either an ASC or PMS star (ASC/PMS). If the ChIcAGO source falls within Region ii of Figure 8 and also within one or both of the WR color spaces in Figures 9(a) and (b), then its type has been classified as a WR star. Those ChIcAGO sources that fall within Region ii of Figure 8 and near AX J144547–5931 in Figures 9(a) and (b) (see Section 4.2.2) have been classified as massive O-type stars (MS-O). All Region ii stars that fall within the general stellar loci of Figures 9(a) and (b) could be either massive stars (MS) or ASC and so are listed as MS/ASC. (A PMS star interpretation is also possible for these Region ii sources.)

Those Region ii ChIcAGO sources listed as WR, MS, or MS-O still need to be properly investigated for evidence of X-ray emission emanating from colliding winds in a CWB. Such an identification requires further X-ray spectroscopic follow-up to identify plasma temperatures between 1 and 10 keV (Usov 1992). There are also seven AGPS sources that are listed as H ii regions. Those AGPS sources identified as "H ii" are actually made up of many X-ray point sources that were detected in our ChIcAGO Chandra observations (see Section 4.1) and are therefore young and massive stars in the H ii region listed. If there is an identified X-ray star that is a significant contributor to the X-ray emission, the stellar type is listed before the H ii abbreviation in the third column, along with its ChIcAGO source name in the second column. The suffixes of the other ChIcAGO sources coincident with the H ii region are listed in parentheses. All other AGPS and ChIcAGO sources for which there is no identification information are classified as unknown (U). If no sources were detected in a ChIcAGO Chandra observation, then the AGPS source falls in the no detection (ND) category. A detailed description of some of the AGPS sources identified through the ChIcAGO survey and of the archival AGPS sources (i.e., those sources identified in the literature and therefore not observed with Chandra as part of the ChIcAGO survey) can be found in Section 4.3 and Appendix A, respectively.

The final column in Table 12 contains a flag for the first and second columns that indicates whether an AGPS and/or ChIcAGO source has a confirmed identification (I) obtained through work in the ChIcAGO survey or a tentative identification (T) using the population statistics in Section 4.2. Those AGPS sources for which no ChIcAGO sources were detected within 3' of the ASCA position or only faint (<20 X-ray counts) ChIcAGO sources were detected (F) are also indicated. (This flag excludes those AGPS sources that have been identified as H ii regions.) The unidentified population of ChIcAGO sources that fall in Region iv of Figure 8 (R) is also included. There is also a flag for the third column that indicates whether the type identification for a ChIcAGO source is unconfirmed (N) because it is based on its tentative statistical identification in Section 4.2.