Stellar X-ray sources in COSMOS

Stellar X-ray sources in the Chandra COSMOS survey

Abstract

We present an analysis of the X-ray properties of a sample of solar- and late-type field stars identified in the Chandra Cosmic Evolution Survey (COSMOS), a deep (160 ks) and wide (0.9 deg) extragalactic survey. The sample of 60 sources was identified using both morphological and photometric star/galaxy separation methods. We determine X-ray count rates, extract spectra and light curves and perform spectral fits to determine fluxes and plasma temperatures. Complementary optical and near-IR photometry is also presented and combined with spectroscopy for 48 of the sources to determine spectral types and distances for the sample. We find distances ranging from 30 pc to 12 kpc, including a number of the most distant and highly active stellar X-ray sources ever detected. This stellar sample extends the known coverage of the -distance plane to greater distances and higher luminosities, but we do not detect as many intrinsically faint X-ray sources compared to previous surveys. Overall the sample is typically more luminous than the active Sun, representing the high-luminosity end of the disk and halo X-ray luminosity functions. The halo population appears to include both low-activity spectrally hard sources that may be emitting through thermal bremsstrahlung, as well as a number of highly active sources in close binaries.

Subject headings:
stars: activity – stars: coronae – stars: late-type – X-rays: stars – binaries: close – Galaxy: halo

1. Introduction

Nearly all types of star are known to emit X-rays through a range of different emission mechanisms that include shocks in the radiatively-driven winds of massive stars and emission from high-temperature stellar coronae in later-type stars (Vaiana et al., 1981). Across the stellar sequence the level of X-ray emission varies from , but can also vary over several orders of magnitude within each spectral class. Close or interacting late-type binary systems, as well as active young single stars may even emit at much higher levels for short periods of time through flare events. Only evolved late-type giants and main sequence B and A-type stars do not appear to emit X-rays at these levels and their X-ray emission properties, if they emit X-rays at all, are still unknown (e.g. Schmitt, 1997).

Solar- and late-type stars such as our Sun emit X-rays from a magnetically-confined plasma at typical temperatures of one to several million Kelvin known as a corona. The corona is thought to be heated mostly by magnetic reconnection events, powered by the stellar dynamo, which itself is thought to be generated – at least in the Sun – by differential rotation between the star’s radiative and convective layers (e.g. Skumanich, 1972; Pallavicini et al., 1981; Noyes et al., 1984). The observed decrease in stellar X-ray luminosity of several orders of magnitude between the zero age main sequence (e.g. Feigelson et al., 2002; Flaccomio et al., 2006; Wright et al., 2010) and solar age (e.g. Peres et al., 2000) has therefore been attributed to the rotational spin-down of the star, though a consistent picture has yet to emerge. Güdel et al. (1997) studied a sample of nearby solar-type stars aged  Gyr and found that the X-ray luminosities decayed as , while Micela (2002) could find no evidence for a clear decay law over a similar age range, and Feigelson et al. (2004) estimated a decay law of from a sample of faint high Galactic latitude main-sequence stars.

The origin of these discrepancies could lie with the small but diverse samples used to study coronal X-ray emission. Wide-field X-ray surveys from the Einstein and ROSAT observatories (e.g. Gioia et al., 1984; Voges et al., 1999) have resulted in large samples of stellar X-ray sources (e.g. Hünsch et al., 1999a; Schmitt & Liefke, 2004; Agüeros et al., 2009), but which are biased toward bright and nearby ( pc) thin-disk stars. The detection of X-rays from more distant, low-metallicity stars in the Galactic halo is important because it allows us to study how X-ray activity behaves at low metallicity as well as probe the close binary population of the early Galaxy. Deep surveys with the Chandra and XMM-Newton observatories are necessary for detecting these distant sources (e.g. Feigelson et al., 2004; Covey et al., 2008). With this aim in mind, we are mining multiple Chandra datasets to build large samples of X-ray sources that fully populate the stellar X-ray luminosity function. This will be necessary if we are to develop a full understanding of the influences of age, spectral type, and metallicity on stellar X-ray emission.

The Cosmic Evolution Survey (COSMOS, Scoville et al., 2007b) is a deep and wide extragalactic survey designed to probe the medium redshift galaxy and active galactic nuclei (AGN) populations. The COSMOS field has been observed at nearly all wavelengths with both ground- and space-based facilities (e.g. Capak et al., 2007; Scoville et al., 2007a) providing a large multi-wavelength catalog for studies of galaxy evolution. The survey also samples a long sight-line through the Galactic disk and halo that may be used to study the stellar populations in these regions (e.g. Robin et al., 2007) and their properties at different wavelengths.

In this paper we present a sample of stellar X-ray sources identified in the Chandra survey of the COSMOS field (Elvis et al., 2009), including an analysis of their properties. In Section 2 we describe the identifications of the stellar X-ray sample and present complementary optical spectroscopy to confirm the stellar nature of the majority of sources, determine spectral types and distances and compare our sample with other stellar X-ray surveys. In Section 3 we analyze the coronal properties of these sources in the context of other stellar coronal studies.

2. The Chandra COSMOS stellar sample

The X-ray observations presented here are from the Chandra COSMOS Survey (Elvis et al., 2009) that has imaged an area of 0.9 deg of the COSMOS field using the ACIS1 imager (Garmire et al., 2003) on board the Chandra X-ray Observatory (Weisskopf et al., 2002). The survey uses a grid of 36 overlapping pointings to give a highly uniform exposure of 160 ks over the central 0.5 deg and 80 ks over an outer region of 0.4 deg. A detailed source detection procedure (Puccetti et al., 2009) resulted in a catalog of 1761 sources detected in one or more X-ray bands, with well-defined sensitivities and completeness fractions as a function of both X-ray band and survey area.

Civano et al. (2010) used optical and near-IR observations of the COSMOS field to make identifications for 1750 of the 1761 Chandra sources, including 61 stars. Identifications were made using likelihood ratio tests and by comparing optical and near-IR images with the morphology of the X-ray source. 27 of the stars were identified morphologically or through positional alignment of an X-ray source with a bright star. A further 21 stellar identifications were made by fitting multi-wavelength photometry to the spectral energy distributions (SEDs) of templates taken from Salvato et al. (2009), while the remaining 13 were identified spectroscopically from dedicated COSMOS spectroscopic campaigns (see Section 2.2 for more details). In many cases these sources were identified as stars by multiple methods (e.g. photometrically identified and then confirmed spectroscopically). The majority of the remaining Chandra-COSMOS sources were identified as galaxies based on the above methods, with only 11 sources remaining unidentified by Civano et al. (2010): 2 of these have no identifiable counterpart and 9 either have multiple possible counterparts or are either associated with faint optical sources in close angular proximity to bright stars or galaxies, such that their properties cannot be studied. Statistically, based on the 1750 identifications made, of which only 3.5% are stars, the 11 unidentified sources are likely to be galaxies.

2.1. X-ray photon extraction and spectral fitting

Since the extraction and characterization of Chandra COSMOS sources performed by the survey collaboration has been conducted on the basis that they are extragalactic sources (which have different morphological and spectral properties to stellar sources) we have re-analyzed the observations of the stellar sources assuming that they are stellar. This was done using CIAO2 4.2 (Fruscione et al., 2006), CALDB 4.2.2, and the ACIS Extract3 code (AE, Broos et al., 2002) using the method outlined in Wright & Drake (2009). To summarize, AE uses a variety of point spread functions (PSFs) appropriate for the off-axis angle of each observation of each source to extract photons in a set fraction of the PSF (typically 90%). The background is estimated from a region surrounding this PSF that excludes the PSFs of other sources. From these extractions, AE calculates a source significance and the Poisson probability, , that the source counts are a superposition of background photons. At this point all sources were inspected visually and compared to the position of their designated optical counterpart to confirm their association. We then applied a cut to the sample, discarding any sources that had a higher probability of being a false source than of being a real source (i.e. ), which resulted in one source being discarded and reduced our sample to 60 sources. This rather liberal cut level was chosen to maintain a high level of completeness with respect to the existing Chandra COSMOS catalog.

Spectral fitting was performed for the 27 sources with net counts using xspec4 version 12.6.0 (Arnaud, 1996). The spectra were compared to apec (Smith et al., 2001) spectra corresponding to single-temperature thermal plasma models (Raymond & Smith, 1977) in collisional ionization equilibrium and absorbed by a hydrogen column density using the tbabs5 xspec model (Balucinska-Church & McCammon, 1992). Due to the low Galactic extinction in the COSMOS sight-line the hydrogen column density was allowed to vary only up to the maximum value for the field of  cm (Kalberla et al., 2005), while the thermal plasma temperature was allowed to vary freely. A grid of initial thermal plasma temperatures covering was used to prevent fitting local minima and the model with the lowest C-statistic (Cash, 1979) was then used for each source. Two-temperature thermal plasma models were also tested for these sources, but only the brightest source, CID6 546, had sufficient counts to produce a noticeably better fit using a two-component model. For the 33 sources with less than 20 net counts we used the method outlined in Getman et al. (2010) to calculate X-ray fluxes from count rates and median photon energies. Since the hydrogen column density in our field of view (FoV) is negligible, we use apparent X-ray fluxes as intrinsic fluxes. Uncertainties on these fluxes are not specified individually, but were determined statistically by Getman et al. (2010). They are an approximate function of the net counts of the source and range from 30% for sources with 20 net counts to % for sources with net counts.

The X-ray properties of the 60 retained sources are listed in Table 1. With the exception of a single very bright source (CID 546), the majority of sources have count rates of , appropriate for a sample based on observations of 100-200 ks and a source detection procedure that has extracted sources down to 3 net counts.

2.2. Optical and near-IR photometry and spectroscopy

Optical and near-IR photometry was taken from the COSMOS optical catalogs7 that include data from the Sloan Digital Sky Survey (SDSS, York et al., 2000), the Subaru photometric catalog (Capak et al., 2007), and the CFHT8/Megacam catalog (McCracken et al., 2010). For the brightest sources that saturate in the deep COSMOS catalogs we complemented this data with near-IR photometry from the Two Micron All Sky Survey (2MASS, Cutri et al., 2003). The majority of sources in our stellar sample have either 2MASS or SDSS photometry and therefore for the faintest sources we used photometry in other photometric systems translated to the SDSS or 2MASS systems using the conversions listed in Capak et al. (2007). All photometry is listed in Table 2.

Spectra for objects in our sample were compiled from a number of sources and used to determine spectral types and confirm the stellar nature of the sources. Where available, spectroscopy was taken from the accumulated spectroscopic catalogs available for the COSMOS field, including data from IMACS9/Magellan (Trump et al., 2007, 2009) and VIMOS10/VLT11 (Lilly et al., 2007, 2009). This amounted to 13 stellar spectra.

An additional 35 spectra (for sources with ) were obtained using the FAST spectrograph (Fabricant et al., 1998) on the 1.5-m Tillinghurst telescope at the Fred Lawrence Whipple Observatory, Mount Hopkins, Arizona. The spectrograph was equipped with a 300 gpm grating, resulting in a resolution of 3 Å and a wavelength coverage of 3480–7400 Å. Identification spectra were obtained from the raw data following standard data reduction procedures including bias-subtraction, flat-fielding, cosmic-ray removal and wavelength calibration. The exposure time per source ranged from 1 to 30 minutes, and the seeing was typically 1–2. Figure 1 shows examples of these spectra.

Figure 1.— Optical identification spectra of three sources observed with the FAST spectrograph on the 1.5-m Tillinghurst telescope. Spectra are not flux calibrated and are shown in normalized flux units.

Spectral classifications for the 48 sources with spectroscopy were obtained by visual comparison with low-resolution spectra of MK standards, assuming that all stars lie on the main sequence12. For solar-type stars (F/G/K-type) we used spectra from Gray & Corbally (2009) obtained at the Dark Sky Observatory and available online13, while for cooler stars (late K & M-type) we used the atlas of late-type spectra presented by Kirkpatrick et al. (1991). The spectral types determined range from F5 to M7 (see Figure 1 for examples), with some of the M-type stars showing H in emission, suggesting chromospheric activity, which is a good tracer of coronal activity in late-type stars.

To estimate the uncertainty on our spectral types we classified each star independently 5 times. The mean spectral type (assigning integer values to each subtype) was taken as the final spectral type and the standard deviation was calculated for each star as an indicator of the uncertainty. For solar type stars we found the uncertainty to be approximately subtypes, but only subtype for late-type stars (except for objects where the spectral quality was poor and the uncertainty is subtypes). All of the objects with spectra were confirmed as stellar sources, supporting the accuracy of the SED fitting technique used to separate stars from galaxies.

For the 12 sources without spectra we determined photometric spectral types based on their colors and the empirical colors from Covey et al. (2007). A comparison of the photometric and spectroscopically-determined spectral types for those stars with both spectra and SDSS photometry shows an excellent agreement with a standard deviation of only 1.1 subtypes (a similar method using photometry and empirical colors from Kenyon & Hartmann (1995), Covey et al. (2007), or Kraus & Hillenbrand (2007) is less accurate with a standard deviation of 3 subtypes). Based on this test we assign an uncertainty of subtypes for photometrically-determined spectral types. The final distribution of spectral types determined from spectroscopy (photometry) is 6 (1) F-type stars, 7 (0) G-type stars, 19 (1) K-type stars, and 16 (10) M-type stars. The accumulated photometry and spectral types are listed in Table 2.

Finally, we searched through the SIMBAD14 Astronomical data base for previously identified objects in our sample. Three sources have previously determined spectral types: CID 590 is listed as K0 (we find it to be a K2 type), CID 1560 is listed as M-type by Welsh et al. (2007) (we find M3 type) and CID 3381 is listed by Heintz (1992) as a K0-type multiple star, while we find it to be G8 type. All of these differences are within our classification uncertainties. We also identified two matches (CIDs 537 and 1560) between our catalog and the catalog of flaring M-dwarf stars from the Galactic Evolution Explorer (GALEX) all-sky survey (Welsh et al., 2007), a result that is not surprising since objects that flare in the ultraviolet are also likely to exhibit pronounced X-ray emission because of their flaring and therefore make them easier to detect in a flux-limited sample such as our own.

2.3. Distances and X-ray luminosities

Distances were determined for all sources based on the spectral types estimated above and using absolute magnitudes for main-sequence stars from Kraus & Hillenbrand (2007). To accurately sample the most luminous part of the SED we use -band photometry for solar-type stars and -band photometry for late-type stars (where both photometry exists we find that the mean difference between distances determined using these two methods is 18%). We calculated uncertainties for these distances based on the uncertainties of the underlying spectral types and derive uncertainties of % for solar-type stars and % and % for late-type stars with spectral types uncertain to and subtypes, respectively.

These distances are listed in Table 2 and vary from 30 pc to over 10 kpc (see discussion of the most distant sources in Section 3.2) with the majority of sources at distances of 100 – 1000 pc. At the Galactic latitude of the COSMOS field, +42, a Galactic star counts simulation (Girardi et al., 2005) shows that the dominant stellar population switches from the disk to the halo at a distance of 400 pc. While this value may be slightly greater for an X-ray selected sample (because the halo population is older and will therefore have typically lower X-ray luminosities), and it is impossible to separate disk and halo stars individually based on their distances alone, this does suggest that a number of sources in our sample probably belong to the Galactic halo, though radial velocities or metallicity measurements will be necessary to be certain. Based on our comparison with other samples, some of these sources appear to be the most distant X-ray emitting late-type stars currently known.

Distances were used to calculate X-ray luminosities for each source, which have uncertainties that scale as the square of the distance uncertainties, i.e. % for solar-type stars, and % and % for the late-type stars. Figure 2 shows the distribution of our sample in –distance space compared to other recent surveys of stellar X-ray sources from Chandra and ROSAT. Also shown are the range of distances and X-ray luminosities spanned by each sample, as well as their logarithmic means and standard deviations. We also show estimates of the catalog sensitivity limits using the formula provided by Feigelson et al. (2005) for the three Chandra surveys, which shows that increases in survey depth should allow fainter and more distant sources to be detected. However, the transition from the shallow ROSAT surveys through the different Chandra surveys has been most apparent in an increase in source distance, with an actual shift toward higher X-ray luminosity. This bias is caused by the selection of fields for extragalactic observations (in the case of the CDF-N and COSMOS surveys) that avoid nearby bright stars and the exclusion of fields with bright sources in the Chandra Multiwavelength Project (ChaMP) survey (as noted by Covey et al., 2008, in their version of this figure).

2.4. Comparison of Chandra surveys

Limiting our comparison of X-ray luminosities to only Chandra surveys (see Covey et al., 2008, for a discussion of the limits of the ROSAT stellar samples), we note that our sample is similar to that of Covey et al. (2008), but both are typically more X-ray luminous than the sample of Feigelson et al. (2004), where all but one source have log . To determine if these three samples are drawn from the same distribution we performed two-sample Kolmogorov-Smirnov (K-S) tests on each pair of surveys. Comparing our sample with that of Covey et al. (2008) we find a 10% probability that they are drawn from the same distribution, a discrepancy caused by the lack of objects in our sample with log . This may be partly due to size-of-sample effects, but it could also be caused by a change in the main population being sampled as we go from the thin-disk to the halo. The sources ‘missing’ from our sample would be expected to lie at distances of  pc, a range dominated by halo stars that, due to their age, might not reach the levels of X-ray luminosity seen in younger, more local samples. The presence of the distant and highly luminous X-ray sources in our sample would then not be considered part of the single-star X-ray luminosity function but potentially due to rarer, active binary systems.

Figure 2.— X-ray luminosity as a function of distance for sources detected in this study (red crosses) and compared to recent samples. Samples shown are those of Hünsch et al. (1999b) (black dots), Schmitt & Liefke (2004) (red dots), Feigelson et al. (2004) (blue crosses), Agüeros et al. (2009) (green dots), and Covey et al. (2008) (black crosses). Sources from Chandra studies are shown as crosses while those from ROSAT studies are shown as dots. For reference we also show dashed lines illustrating the estimated Chandra sensitivity limits using the formula provided by Feigelson et al. (2005) for the three Chandra surveys in the same color as their respective symbols. We also show the range of distances (above the figure) and X-ray luminosities (to the right) spanned by each sample as illustrated by dashed (for ROSAT) and full lines (for Chandra) in the same color as their respective symbols. Marked on each of these lines are the mean logarithmic values (shown with a filled circle) and the standard deviations (shown with two marks either side of the mean).

Performing the K-S test on the sample of Feigelson et al. (2004) we find a probability of 0.1% that it is drawn from the same distribution of either of the other two Chandra samples. In the sample of Feigelson et al. (2004), the first and third most X-ray faint sources were identified only in the full 2 Ms of Chandra Deep Field North (C-DFN) observations and not in the 1 Ms sample of 11 stars that the authors identify as the more uniform and complete subset of stellar emitters. Removing these and repeating the K-S test gives a probability of 5% that the sample is drawn from the same distribution as ours. This discrepancy, which may simply be due to non-uniformities in the stellar distributions, is problematic because it opens up the question of whether surveys such as these can be used to study stellar X-ray emission. However, there are a number of differences in the depths, areas, and sensitivities of the three surveys that could present possible explanations for these differences. For example, the lack of any distant or highly X-ray luminous sources in the C-DFN sample could be attributed to a size-of-sample effect that might limit the most distant source detected in the smaller field of view (FoV) of the C-DFN survey compared to that of COSMOS.

Our lack of intrinsically faint X-ray sources is more difficult to understand, but we suggest this could be explained by an exposure time bias when calculating X-ray luminosities from time-averaged X-ray fluxes, as was done in both works. If a source can only be detected during a short-duration flare event and only flares once during the observation (which is not unlikely, since Feigelson et al., 2004, detect one flare every 2.4 Ms in the C-DFN observations) then its X-ray luminosity will scale inversely with the total exposure time. This could explain why the sample of Feigelson et al. (2004) contains sources 10 times fainter in X-rays than our sample, because the typical observation time of the C-DFN stars is 10 times greater than those in the COSMOS survey. However, we note from studying the light curves shown by Feigelson et al. (2004) that many of the flaring sources detected by them appear to have quiescent X-ray luminosities sufficient for them to be detected even if they had not flared. This would therefore make this explanation unlikely.

As it is we are unable to identify the exact causes of the difference in X-ray luminosities between the two samples. We have searched the COSMOS field for X-ray emission from the nearest stellar sources in our FoV and find no evidence for X-ray emission from them. Additionally, we do not believe that any distant and highly luminous X-ray sources in the C-DFN were missed by Feigelson et al. (2004) since the optical photometry available to them was comparable in depth to ours, and the majority of C-DFN sources were observed spectroscopically, which would have allowed them to identify any faint stars. To achieve a K-S probability of 20% that the two samples are drawn from the same distribution would require either high luminosity sources to have been missed from the C-DFN sample, or for existing sources to have notably higher X-ray luminosities. We can therefore only suggest that the differences in the two populations are a combination of the above reasons and the effects of Poisson statistics on such small samples.

3. Discussion of source properties

We now consider the X-ray properties of our sample of 60 stellar X-ray sources and explore correlations between their X-ray and optical properties. We also compare their properties to the recent Chandra surveys from Feigelson et al. (2004) and Covey et al. (2008). As shown in Figure 2 the former survey has sampled moderately distant (50-500 pc) intrinsically faint X-ray emitters, while the latter survey has probed more distant, but more luminous sources. Our sample represents an extension of the Covey et al. (2008) survey, including sources with a similar range of intrinsic X-ray luminosities but at greater distances and therefore likely including a number of halo sources.

3.1. X-ray properties as a function of spectral type

We first consider the X-ray luminosities of our sample as a function of their spectral type, for which we use the proxy of color for comparison with previous works (using the table of color as a function of spectral type presented by Kenyon & Hartmann, 1995), shown in Figure 3. Our sample includes sources with X-ray luminosities in the range log  erg s, independent of spectral type. This differs from the finding of Zickgraf et al. (2005) who found a weak trend of increasing toward earlier spectral types, though their nearby sample could differ intrinsically from our more distant sample, and may include fewer highly active stars. The X-ray luminosities are typically higher than that of the contemporary Sun (Peres et al., 2000, adjusted to match Chandra’s spectral bands), with a small number of G-type stars exhibiting X-ray luminosities similar to the active Sun. Figure 3 also shows the X-ray to bolometric luminosity ratio as a function of spectral type, using bolometric luminosities determined from the table of main-sequence bolometric magnitudes presented by Kraus & Hillenbrand (2007). Our sample and that of Covey et al. (2008) are in agreement with the observed trend of increasing toward later spectral types (e.g. Fleming et al., 1995).

Figure 3.— X-ray luminosities (top panel) and X-ray to bolometric luminosity ratios (bottom panel) as a function of color for sources in our sample (red dots), the sample of Feigelson et al. (2004, blue dots), and of Covey et al. (2008, grey dots). The range of X-ray luminosities exhibited by the Sun (Peres et al., 2000), and adjusted to match the Chandra bands, is shown as a black line.

3.2. X-ray properties of the most distant members of the sample

Our sample includes a number of sources with distances  kpc that are likely to be members of the Galactic halo. Partly because of our sensitivity limits, the majority of these are highly luminous with log  erg s. Since the Galactic halo is 10 Gyrs old and declines with age through magnetic braking, these are almost certainly close binaries kept active through tidal interaction and tapping of orbital angular momentum to sustain strong dynamo activity. Ottmann et al. (1997) found that Population ii binaries typically have lower X-ray luminosities than more metal-rich systems, but do exhibit a high-luminosity tail with log  erg s. We are unable to definitively identify halo members amongst our sample of moderately bright (log  erg s) sources, but our detection of two distant and highly-luminous sources with  kpc and log  erg s suggests that a high-luminosity tail for the halo binary distribution does exist and at a higher X-ray luminosity than found by Ottmann et al. (1997). If our sample is representative, a simple extrapolation suggests that the Galactic halo contains binaries with log  erg s.

Our most distant and highly luminous source, CID 1600, is a relatively faint detection with only 5 net counts and a probability of being a background event of 0.026, though the Chandra COSMOS catalog lists it with 12 net counts (Puccetti et al., 2009). Optical and near-IR photometry suggests it is a K1-type star (though no spectroscopy exists) that would put it at a distance of 11.7 kpc and give it an X-ray luminosity of  erg s. While the X-ray source is 3 from the optical counterpart, the source is 12 off-axis, and its PSF is therefore similarly-sized. There are no other suitable optical counterparts in either the deep Hubble Space Telescope observations of the field (that extend down to , Scoville et al., 2007a) or longer wavelength mid-IR observations. Considering this we are left with the choice that the X-ray source is either associated with the optical source, or is a background event. We note that the chance of an X-ray source being within 3 of a source with (as the optical source does) is 0.027, making the probability that a random background fluctuation would be found in such a location to be . Considering that the Chandra COSMOS catalog contains 1700 sources we might expect at least one spurious event such as this and this could be a possible candidate. If the source were real it is likely to be in an active binary that was observed to flare during the observations, an interpretation supported by the high median photon energy of the source ( keV) and the fact that all but one of the detected photons came in the second of two similar-length exposures. Deep spectroscopy will be necessary to confirm the stellar nature of the source and detect evidence of its binary nature.

The second most distant and luminous source in our sample, CID 3205, is a more reliable detection with 12 net counts and a close association with an optical source whose photometry suggests it is either an F8-type star at a distance of 8 kpc or a much closer white dwarf (WD) or cataclysmic variable (since the colors of these sources overlap in the SDSS system, Fan, 1999). If the source were an F8 star, then since the main sequence lifetime of such a star is 5 Gyrs it could not be a member of the Galactic halo, but is more likely to be a member of the thin disk that has been forced onto a highly elliptical orbit through some sort of close encounter. While the X-ray luminosity of the source is high, it is not unfeasible if the source is relatively young, though it could also be evidence for the source being an active binary. However, if the source were a WD it would likely be much closer and therefore have a lower X-ray luminosity. The USNO-B catalog (Monet et al., 2003) lists a small proper motion for the source of 0.26/yr, which would be consistent either with a nearby WD or with a distant main sequence star on an elliptical orbit and at a high velocity. We note that the color of 1.01 is redder than would be expected for a WD (Smolcic et al., 2004), but at the blue end of the colors of F-type main-sequence stars (Covey et al., 2007). Again, spectroscopy will be necessary to confirm the spectral type of the source and detect evidence for binarity.

3.3. Coronal plasma temperatures and variability

The median photon energy of X-ray source events provides a simple characterization of the X-ray emission properties that is particularly useful for faint sources lacking plasma temperatures derived from spectral fits. The distribution of median energies seen in Figure 4 is highly clustered around 1 keV and in agreement with the distribution of plasma temperatures from spectral fits,  keV (6–12 MK), typical for stars of moderate to high activity levels and for active regions and flares on the Sun (Peres et al., 2000). This indicates that our sample includes a large fraction of active stars, as might be expected for a luminosity-limited sample. We also observe no correlation between either the plasma temperature determined from spectral fitting or the median photon energy of a source with its spectral type.

Figure 4.— X-ray to bolometric luminosity ratios as a function of median photon energies for all sources. Sources with identifiable flaring events are shown in red.

Figure 4 shows the X-ray to bolometric luminosity ratios as a function of median photon energy for all our sources. For sources with  keV we note a trend of increasing luminosity ratio with median photon energy. This is similar to the relationship between the luminosity ratio and plasma temperature commonly seen in late-type stars and found by Schrijver et al. (1984) and Schmitt et al. (1990). It is thought to result from the increasing size and intensity of active regions, and the growth of flaring activity as active regions fill larger fractions of the stellar surface (e.g. Drake et al., 2000). In order to look for the influence of flares, we compiled X-ray light curves from the ACIS event lists and tested for variability. We used a one-sample Kolmogorov-Smirnov test to compare the distribution of photon arrival times with that expected for a constant source (the null hypothesis) and then derived the probability of accepting the null hypothesis, , as listed in Table 1. We studied the light curves for 17 sources with and identified six flaring events with durations of 2-5 hours, three of which are shown in Figure 5. These six sources are also indicated in Figure 4 and can clearly be seen at the high end of the trend mentioned above. We then studied the light curves of the other sources with high and high , but could find no evidence for bright flares.

We find no trend of median photon energy with luminosity ratio for stars with higher median energies, corresponding to plasma temperatures of  MK. The luminosity ratios for these stars range from . The majority of these sources are distant with 60% of sources with  keV found at distances  kpc, compared to a fraction of 25% for the entire sample. One explanation for the spectral hardness of sources with high X-ray luminosity ratios is that these sources were observed during particularly long and bright flares. However this cannot be the case for all the hard sources because they do not appear to be significantly more variable than the soft sources: 30% of the hard sources have , compared to 28% for the entire sample, while all the clearly identified sources with flares are in the soft sample. There will be a bias in this analysis because variability is easier to identify in sources with more counts, which are more likely to be included in the nearby soft sample, but there appear to be multiple hard X-ray stellar sources that cannot be explained by variability. It is more likely that some of these sources, particularly those with low to moderate luminosity ratios (log ), are members of the halo population that have extremely metal-poor coronae. Indeed, a lack of a correlation between X-ray luminosity and plasma temperature was noted by Ottmann et al. (1997) based on a survey of nearby Population ii close binaries. These authors also found the Pop. ii stars to have harder spectra than their Pop. i counterparts, and attributed this to the lower radiative efficiency of metal-poor plasma. Our sample of stars appears to support this, with what must be halo stars appearing to have very hot coronae. In such coronae, it seems that the lack of plasma radiative cooling through metal lines is compensated for by much higher plasma temperatures and that radiative cooling occurs predominantly through the bremsstrahlung continuum.

Figure 5.— Segments of the X-ray light curves for sources CID 847, 989, and 3650 (black lines) showing flaring events. Coarser bins are overlaid in red with Poisson uncertainties. The black dashed lines show the mean count rates over the total observations for each source.

Finally we note that CID 546 is the only source with sufficient counts to make a reasonable two-temperature thermal plasma fit. The difference between single and double-component thermal plasma fits is a decrease in the Cash statistic (Cash, 1979) from 843 to 752 and a factor three drop in the maximum fit residuals from 0.015 to 0.005. The two-temperature thermal plasma fit consists of plasma at temperatures of 0.73 and 2.1 keV at a flux ratio of 1.3:1 (e.g. López-Santiago et al., 2007). Extraction of the light curve of CID 546 does not reveal any large flaring events, and the median photon energy remained relatively constant at 1.1 keV throughout the observations.

4. Conclusions

In this paper we have studied the stellar content of the Chandra-COSMOS survey and identified a sample of 60 stellar sources for which we present X-ray properties, as well as optical and near-IR photometry. In addition we have obtained spectroscopic classifications for 48 of the sources, confirming their stellar nature and allowing us to derive spectral types and distances. In the -distance plane the sample extends the recent survey of Covey et al. (2008) to more distant sources, with the majority of sources lying at several hundred parsecs. The most distant sources are highly likely to be members of the Galactic halo, with the two most distant sources at 8–12 kpc being the most distant late-type stellar X-ray sources known.

The X-ray luminosity distribution of our sample is in approximate agreement with that of Covey et al. (2008), but is significantly more luminous that that of Feigelson et al. (2004) and we consider a number of possible explanations for this, including an exposure-time bias and size-of-sample effects. Differences between high Galactic latitude stellar X-ray samples is potentially problematic because it raises the issue for how much can be learnt about stellar X-ray emission and dynamo activity from studies of individual sight lines. In a future paper we will attempt to model these samples and therefore explore the possibility that these differences are due to either size-of-sample effects or differences in the populations sampled that would be caused by different survey depths. Further studies in other equally deep Chandra surveys will also be useful to probe this discrepancy as well as increasing the overall sample size.

A comparison of X-ray and optical properties reveals no major differences between this and previous samples, though we note a large number of sources with high plasma temperatures and we suggest these are a combination of low-metallicity halo stars emitting through thermal bremsstrahlung at high temperatures, and a population of flaring close binaries in the Galactic halo. This reveals the excellent opportunity presented by this sample and other deep Chandra surveys to understand X-ray emission at low metallicities and probe the close binary population of the early Galaxy. High-resolution optical spectra will be necessary to measure metallicities and identify binaries.

A future paper will use this catalog to test models of the decay of X-ray activity in solar- and late-type stars in the Galaxy.

We would like to thank the referee, Kevin Covey, and the scientific editor, Eric Feigelson, for useful comments that improved the work presented here. We are grateful to the staff at the Fred Lawrence Whipple Observatory, particularly Perry Berlind and Michael Calkins, for FAST spectroscopy with the 1.5-m on Mt Hopkins, and to Susan Tokarz and Nathalie Marthinbeau for data reduction. This research has made use of data from the Chandra X-ray Observatory and software provided by the Chandra X-ray Center (CXC) in the application packages CIAO and Sherpa, and from Penn State for the ACIS Extract software package. This work has made use of data from the Chandra COSMOS Survey, which is supported in part by NASA Chandra grant number GO7-8136A. We thank the C-COSMOS team for their work on this survey, assistance with this research, and careful reading of this manuscript. Special thanks are given to Tom Aldcroft, Marcella Brusa, Martin Elvis, Michael Rich, and Gianni Zamorani. This research has also made use of NASA’s Astrophysics Data System and the Simbad and VizieR databases, operated at CDS, Strasbourg, France. This work was funded by Chandra grant AR9-0003X. JJD was supported by NASA contract NAS8-39073 to the Chandra X-ray Center. Facilities: CXO, FLWO, 2MASS
RA Dec CID Cnts Sig. log() Exp. log() Model fit X-ray fluxes (erg cm s)
(J2000) (J2000) (net) Upper Lower () (ks) (keV) (keV) log log log
10:00:48.44 2:07:34.8 48 144.67 13.3 12.2 10.9 -6.00 187.88 1.00 -0.20 1T -15.34 -14.17
10:00:49.51 2:07:14.6 49 37.77 7.6 6.5 5.0 -6.00 187.88 0.93 -0.02 1T -17.09 -13.75
10:00:20.96 1:44:32.3 268 30.96 6.9 5.8 4.5 -6.00 96.19 1.03 -0.75 1T -15.76 -14.12
9:58:23.06 2:13:11.9 321 87.36 11.3 10.3 7.7 -6.00 91.46 0.91 -0.87 1T -15.53 -13.96
10:00:57.46 1:55:48.7 367 157.32 14.0 13.0 11.2 -6.00 184.36 0.84 -0.40 1T -16.05 -14.04
10:01:42.18 1:53:19.7 397 16.63 5.6 4.4 3.0 -6.00 93.25 0.77 -0.14
9:59:54.70 2:17:06.0 444 342.87 20.2 19.2 17.0 -6.00 232.80 0.98 -1.72 1T -15.13 -13.82
10:00:22.21 2:10:19.9 462 38.63 7.6 6.5 5.1 -6.00 188.71 1.70 -0.00 1T -14.57 -14.40
9:58:56.03 2:30:40.9 516 41.33 8.2 7.1 5.0 -6.00 91.61 1.01 -1.13 1T -15.38 -14.24
10:00:09.81 2:23:49.9 527 250.56 17.5 16.4 14.3 -6.00 231.76 0.91 -1.30 1T -15.89 -13.86
10:01:41.57 2:07:59.4 537 321.81 19.4 18.4 16.6 -6.00 137.65 1.00 -10 1T -14.98 -13.65
10:01:52.18 2:11:58.4 546 1700.06 42.3 41.3 40.2 -6.00 90.37 1.06 -10 1T -13.64 -12.79
9:59:15.68 2:32:25.0 578 9.35 4.4 3.3 2.1 -5.47 93.49 0.96 -0.35
10:01:43.18 2:17:28.4 590 392.42 21.1 20.1 18.6 -6.00 181.79 0.93 -0.31 1T -15.19 -13.69
10:00:05.61 2:07:00.9 742 12.95 6.1 5.0 2.1 -3.03 188.61 1.28 -0.84
10:01:30.74 2:06:45.9 766 4.05 5.8 4.7 0.7 -0.69 186.62 4.14 -0.46
10:00:46.69 2:02:33.4 843 39.30 8.6 7.5 4.6 -6.00 233.25 1.36 -4.35 1T -15.47 -14.34
10:00:52.92 1:57:14.1 847 69.69 10.7 9.6 6.5 -6.00 232.32 1.04 -3.37 1T -16.02 -14.46
9:59:55.23 2:08:44.7 870 4.75 7.2 6.1 0.7 -0.64 234.43 1.07 -0.71
9:59:30.81 2:32:39.7 904 15.23 6.5 5.4 2.4 -3.51 185.82 1.00 -0.78
10:01:09.03 2:13:51.1 939 15.00 6.2 5.1 2.4 -3.93 234.84 1.82 -0.39
10:00:37.03 2:26:14.8 989 42.55 8.1 7.0 5.3 -6.00 187.36 1.09 -4.49 1T -15.91 -14.39
9:59:01.12 1:57:38.9 998 10.16 5.7 4.6 1.8 -2.32 189.19 2.04 -0.56
9:59:00.98 2:08:30.6 1056 73.97 10.1 9.1 7.3 -6.00 185.90 1.10 -4.23 1T -15.09 -14.43
9:59:29.44 2:05:13.5 1073 17.31 6.3 5.2 2.8 -5.21 189.34 1.00 -0.33
10:00:55.18 1:59:37.6 1103 24.52 7.9 6.8 3.1 -5.33 232.32 1.06 -0.64 1T -17.11 -13.55
9:59:41.82 2:08:59.6 1137 11.51 6.5 5.4 1.8 -2.08 234.31 3.26 -2.80
10:00:33.51 2:05:43.6 1173 29.82 8.4 7.4 3.5 -6.00 234.75 1.06 -0.09 1T -16.19 -14.61
10:02:07.84 2:22:34.9 1560 29.23 6.8 5.7 4.3 -6.00 89.31 0.90 -0.64 1T -16.50 -14.30
10:01:43.23 2:32:52.8 1592 81.57 10.9 9.9 7.5 -6.00 89.30 0.97 -0.73 1T -15.21 -13.92
9:59:11.16 2:42:24.0 1600 5.26 4.3 3.1 1.2 -1.58 46.67 3.79 -0.47
9:59:18.33 2:43:05.2 1604 32.37 7.1 6.0 4.6 -6.00 46.67 1.22 -2.33 1T -14.77 -14.16
9:58:04.42 1:52:16.8 1688 41.96 8.6 7.5 4.9 -6.00 48.26 1.22 -1.58 1T -14.53 -14.02
9:59:08.27 1:57:32.9 1710 2.65 6.7 5.6 0.4 -0.47 285.04 0.81 -2.22
10:02:21.95 2:20:41.9 1768 4.39 5.0 3.9 0.9 -0.90 89.79 0.78 -0.25
10:01:16.77 2:17:13.9 2061 38.36 10.7 9.7 3.6 -5.71 369.12 1.61 -4.44 1T -15.00 -14.78
9:59:06.13 2:34:11.1 2216 16.13 6.0 4.9 2.7 -5.30 93.49 1.41 -1.00
10:02:01.70 2:03:55.5 2331 9.24 5.6 4.5 1.6 -2.06 92.44 0.97 -0.06
9:59:10.23 2:23:34.8 2524 6.19 4.0 2.8 1.6 -3.06 91.61 0.78 -0.02
9:59:17.54 2:22:06.7 2539 20.98 7.2 6.1 2.9 -5.09 182.73 0.91 -1.85 1T -17.19 -14.34
9:59:02.31 2:15:20.3 2881 6.95 6.0 4.9 1.1 -1.16 133.87 1.06 -3.34
9:58:08.76 2:00:01.1 3205 11.58 6.7 5.6 1.7 -1.97 95.61 1.63 -0.06
9:58:39.08 2:09:05.8 3232 13.82 5.2 4.1 2.6 -6.00 91.46 0.88 -0.20
9:58:51.21 2:02:26.8 3243 12.59 5.7 4.6 2.2 -3.59 189.19 0.91 -0.33
10:00:45.93 1:48:19.9 3353 5.08 5.3 4.1 1.0 -0.99 186.67 0.85 -0.30
10:00:03.59 1:50:44.9 3381 23.60 8.0 6.9 3.0 -4.76 237.81 0.87 -3.26 1T -16.79 -14.91
9:59:20.91 1:52:03.6 3425 6.96 4.0 2.8 1.8 -4.53 93.67 1.00 -0.66
9:59:39.21 1:53:49.8 3452 12.88 7.6 6.5 1.7 -1.79 237.30 1.04 -2.30
9:59:12.91 2:00:58.4 3517 17.05 8.0 6.9 2.1 -2.55 236.78 2.20 -2.19
10:00:40.34 2:36:56.2 3650 77.04 10.6 9.5 7.3 -6.00 94.46 1.12 -2.40 1T -14.53 -14.04
10:00:55.31 2:33:30.4 3664 19.90 5.8 4.7 3.4 -6.00 92.79 1.06 -0.17
10:00:36.92 2:23:57.5 3683 8.78 5.2 4.1 1.7 -2.24 187.36 0.96 -0.03
10:01:18.22 2:05:52.4 3782 3.25 4.8 3.6 0.7 -0.72 186.62 2.23 -0.47
9:59:50.63 2:23:15.9 3811 50.30 9.9 8.9 5.1 -6.00 277.19 1.26 -5.22 1T -16.35 -13.23
9:59:10.21 1:53:14.2 10552 3.86 3.8 2.6 1.0 -1.34 93.67 0.96 -0.08
10:00:11.46 2:28:34.0 10742 16.36 5.8 4.7 2.8 -6.00 186.93 0.96 -0.42
10:01:35.76 2:03:34.7 11145 11.00 5.2 4.1 2.1 -3.60 140.82 0.81 -0.33
10:00:54.50 2:16:05.1 11537 3.48 5.0 3.9 0.7 -0.71 189.69 1.03 -1.56
10:01:28.50 1:59:32.4 11905 4.15 5.0 3.9 0.8 -0.85 184.20 0.87 -1.07
10:01:02.45 2:22:29.7 12635 8.00 5.4 4.2 1.5 -1.79 184.79 1.00 -3.29
Table 1X-ray properties of stellar sources detected in the Chandra COSMOS survey.

Notes. Columns 1-2: Source position (from optical images). Column 3: Chandra COSMOS ID. Column 4: Net counts in the full (0.5-8.0 keV) band. Columns 5-6: Upper and lower 1 errors on the number of net counts. Column 7: Source detection significance. Column 8: logarithm of the Poisson probability that the source is a chance coincidence of background events. Values below -6.0 are listed as -6.0. Column 9: Full exposure time for each source derived from the mono-energetic exposure maps for the combined observations. Column 10: Background corrected median energy of all source photons in the full (0.5-8.0 keV) band. Column 11: Logarithm of the Kolmogorov-Smirnov probability that the source is not variable. Column 12: X-ray spectral model fit type: single-temperature thermal plasma model (1T) or no model fit (-). Column 13: Thermal plasma temperature of model fit with upper and lower 90% confidence intervals (uncertainties missing when they are so large that the parameter is effectively unconstrained). Column 14: Logarithm of the extinction-corrected X-ray flux in the full (0.5-8.0 keV) band from model fit or derived from the number of net counts for unfit sources as described in Section 2. Upper and lower 1 errors are shown or left blank when the upper or lower bounds are unconstrained. For sources without model fits the flux errors are not specificed individually. Columns 15-16: Logarithm of extinction-corrected soft (0.5-2.0 keV) and hard (2.0-8.0 keV) band fluxes.

RA Dec CID Optical photometry Near-IR photometry Spectral information
(J2000) (J2000) Type Orig. d (kpc)
10:00:48.44 2:07:34.8 48 11.850 0.023 11.517 0.023 11.447 0.023 G7 F 0.35
10:00:49.51 2:07:14.6 49 22.1241 0.1166 20.099 0.011 18.698 0.006 17.096 0.003 16.249 0.004 14.762 0.038 14.255 0.033 13.864 0.058 M5 V 0.14
10:00:20.96 1:44:32.3 268 23.1336 0.2585 19.871 0.010 18.285 0.004 16.655 0.002 15.785 0.002 14.251 0.035 13.756 0.034 13.495 0.047 M5 VI 0.12
9:58:23.06 2:13:11.9 321 10.397 0.024 10.162 0.027 10.092 0.023 F8 F 0.26
10:00:57.46 1:55:48.7 367 10.755 0.023 10.566 0.027 10.519 0.023 F5 F 0.38
10:01:42.18 1:53:19.7 397 9.946 0.024 9.548 0.025 9.435 0.024 G9 F 0.13
9:59:54.70 2:17:06.0 444 18.2721 0.004 15.48 0.001 14.069 0.001 14.928 0.001 13.014 0.000 11.681 0.024 11.020 0.022 10.845 0.020 M1 F 0.12
10:00:22.21 2:10:19.9 462 22.8733 0.2368 20.404 0.015 19.002 0.007 18.23 0.005 17.79 0.017 16.642 0.002 15.853 0.003 M1 p 1.19
9:58:56.03 2:30:40.9 516 20.7987 0.0234 17.857 0.003 16.502 0.002 15.918 0.001 15.58 0.002 13.702 0.033 13.085 0.033 12.884 0.034 K7 F 0.43
10:00:09.81 2:23:49.9 527 8.766 0.025 8.215 0.031 8.159 0.031 K2 F 0.06
10:01:41.57 2:07:59.4 537 20.3912 0.0239 17.637 0.003 16.152 0.001 14.826 0.001 14.118 0.001 12.693 0.026 12.106 0.025 11.827 0.026 M3e F 0.12
10:01:52.18 2:11:58.4 546 11.069 0.026 10.451 0.024 10.318 0.021 K7e F 0.13
9:59:15.68 2:32:25.0 578 20.6441 0.0265 17.754 0.003 16.254 0.001 15.329 0.001 14.717 0.001 13.545 0.027 12.909 0.026 12.692 0.030 M2 F 0.23
10:01:43.18 2:17:28.4 590 7.038 0.017 6.567 0.029 6.461 0.024 G7 F 0.04
10:00:05.61 2:07:00.9 742 25.133 0.118 23.447 0.037 21.741 0.108 20.45 0.012 18.782 0.005 17.992 0.009 M6 p 0.58
10:01:30.74 2:06:45.9 766 17.1379 0.002 15.754 0.001 15.428 0.001 15.388 0.001 15.326 0.002 14.468 0.038 14.065 0.048 14.011 0.071 F7 F 1.85
10:00:46.69 2:02:33.4 843 23.2358 0.1812 21.395 0.031 19.945 0.015 18.252 0.005 17.316 0.007 15.806 0.001 15.168 0.095 15.052 0.002 M6 V 0.15
10:00:52.92 1:57:14.1 847 24.1586 0.4215 22.598 0.024 20.894 0.032 18.637 0.007 17.382 0.007 15.705 0.001 15.207 0.107 14.773 0.001 M7 V 0.10
9:59:55.23 2:08:44.7 870 18.7864 0.0067 15.796 0.001 14.428 0.001 13.897 0.000 13.538 0.001 12.223 0.033 11.562 0.035 11.397 0.029 M0 p 0.17
9:59:30.81 2:32:39.7 904 20.6471 0.0308 17.768 0.003 16.273 0.001 15.453 0.001 15.034 0.001 13.821 0.024 13.139 0.022 12.931 0.034 M1 F 0.31
10:01:09.03 2:13:51.1 939 21.964 0.1007 19.218 0.006 18.223 0.004 17.837 0.004 17.699 0.014 16.677 0.002 15.908 0.003 K7 V 1.74
10:00:37.03 2:26:14.8 989 20.3733 0.0216 17.639 0.003 16.481 0.002 16.042 0.001 15.795 0.002 14.690 0.041 13.974 0.038 13.910 0.061 K7 F 0.69
9:59:01.12 1:57:38.9 998 11.418 0.023 11.139 0.021 11.059 0.023 F8 F 0.41
9:59:00.98 2:08:30.6 1056 22.0476 0.1239 19.421 0.007 17.995 0.004 17.079 0.003 16.559 0.005 15.414 0.062 14.755 0.060 14.496 0.095 M1e VI 0.64
9:59:29.44 2:05:13.5 1073 12.861 0.024 12.551 0.026 12.503 0.021 F6 F 0.90
10:00:55.18 1:59:37.6 1103 23.0389 0.1515 20.635 0.017 19.216 0.008 17.851 0.004 17.115 0.006 15.746 0.001 15.147 0.097 14.824 0.001 M4e V 0.33
9:59:41.82 2:08:59.6 1137 23.4928 0.4421 21.884 0.017 20.454 0.022 19.404 0.014 18.889 0.040 17.649 0.003 16.808 0.004 M2 p 1.55
10:00:33.51 2:05:43.6 1173 22.8809 0.2385 19.637 0.009 18.162 0.004 16.792 0.002 16.048 0.004 14.655 0.035 14.012 0.042 13.638 0.048 M3 I 0.28
10:02:07.84 2:22:34.9 1560 19.796 0.0111 16.975 0.002 15.523 0.001 15.628 0.001 13.463 0.000 12.083 0.024 11.481 0.024 11.220 0.026 M3 F 0.09
10:01:43.23 2:32:52.8 1592 19.9122 0.0153 17.121 0.002 15.824 0.001 15.138 0.001 14.757 0.001 13.306 0.026 12.639 0.022 12.458 0.024 K7e I 0.36
9:59:11.16 2:42:24.0 1600 23.8312 0.3835 22.058 0.019 21.356 0.013 21.193 0.071 20.864 0.196 19.916 0.019 19.467 0.040 K1 p 11.73
9:59:18.33 2:43:05.2 1604 21.2441 0.0368 18.305 0.004 17.003 0.002 16.262 0.002 15.774 0.002 14.435 0.033 13.834 0.027 13.714 0.047 M0e I 0.51
9:58:04.42 1:52:16.8 1688 19.8183 0.0097 17.209 0.002 16.072 0.001 15.571 0.001 15.276 0.001 13.938 0.023 13.308 0.029 13.125 0.034 K7 F 0.48
9:59:08.27 1:57:32.9 1710 10.542 0.026 10.138 0.029 10.097 0.026 G7 F 0.19
10:02:21.95 2:20:41.9 1768 23.3565 0.2846 21.463 0.033 20.147 0.015 18.794 0.009 18.087 0.018 16.678 0.002 15.829 0.002 M3 p 0.76
10:01:16.77 2:17:13.9 2061 19.6953 0.0103 17.405 0.002 16.484 0.002 16.102 0.001 15.917 0.002 14.667 0.060 14.168 0.065 13.982 0.089 K3 F 0.85
9:59:06.13 2:34:11.1 2216 17.7816 0.0028 15.499 0.001 15.134 0.001 14.339 0.001 14.156 0.001 12.874 0.026 12.270 0.026 12.142 0.026 G8 F 0.84
10:02:01.70 2:03:55.5 2331 11.926 0.024 11.319 0.023 11.226 0.023 K5 F 0.22
9:59:10.23 2:23:34.8 2524 17.1923 0.0017 15.122 0.001 14.49 0.001 15.224 0.001 14.155 0.001 12.982 0.027 12.550 0.023 12.559 0.021 K2 F 0.46
9:59:17.54 2:22:06.7 2539 10.882 0.024 10.381 0.021 10.267 0.021 K3 F 0.15
9:59:02.31 2:15:20.3 2881 24.2583 0.4817 23.331 0.034 21.543 0.013 19.956 0.016 18.808 0.025 17.335 0.002 16.509 0.004 M6 p 0.29
9:58:08.76 2:00:01.1 3205 20.5412 0.0174 19.219 0.006 18.81 0.006 18.684 0.007 18.656 0.023 17.767 0.004 17.311 0.006 F8 p 8.02
9:58:39.08 2:09:05.8 3232 10.398 0.022 9.785 0.026 9.658 0.023 K7 F 0.10
9:58:51.21 2:02:26.8 3243 12.138 0.023 11.636 0.023 11.513 0.023 K3 F 0.27
10:00:45.93 1:48:19.9 3353 21.6116 0.0824 18.509 0.004 17.117 0.002 16.301 0.002 15.897 0.002 14.579 0.037 13.973 0.042 13.708 0.054 M1e F 0.44
10:00:03.59 1:50:44.9 3381 9.219 0.035 8.927 0.049 8.721 0.033 G8 F 0.09
9:59:20.91 1:52:03.6 3425 24.5451 0.6176 22.055 0.018 20.574 0.022 19.501 0.014 18.927 0.026 17.640 0.003 16.844 0.004 M3 p 1.20
9:59:39.21 1:53:49.8 3452 24.1914 0.4467 22.671 0.025 21.456 0.048 19.551 0.015 18.705 0.022 17.095 0.003 16.268 0.003 M5 p 0.42
9:59:12.91 2:00:58.4 3517 21.2677 0.0238 18.551 0.004 17.372 0.002 16.905 0.002 16.653 0.003 15.556 0.001 14.804 0.055 14.726 0.001 K5 F 1.10
10:00:40.34 2:36:56.2 3650 22.5847 0.1697 19.864 0.010 18.411 0.005 16.833 0.002 15.979 0.002 14.563 0.036 13.903 0.043 13.668 0.054 M5 VI 0.13
10:00:55.31 2:33:30.4 3664 23.3307 0.3363 21.114 0.027 19.628 0.012 17.785 0.004 16.804 0.005 15.240 0.048 14.646 0.069 14.341 0.084 M6e I 0.11
10:00:36.92 2:23:57.5 3683 11.264 0.021 10.833 0.024 10.776 0.025 G9 F 0.23
10:01:18.22 2:05:52.4 3782 11.617 0.025 11.181 0.026 11.092 0.027 K2 F 0.23
9:59:50.63 2:23:15.9 3811 20.8651 0.033 18.614 0.004 17.734 0.003 17.328 0.003 17.071 0.007 16.003 0.001 15.546 0.082 15.261 0.002 K7 I 1.29
9:59:10.21 1:53:14.2 10552 19.6035 0.0082 17.688 0.003 17.029 0.002 16.762 0.002 16.655 0.004 15.655 0.001 15.002 0.068 15.028 0.002 K3 F 1.38
10:00:11.46 2:28:34.0 10742 21.5494 0.0461 18.618 0.004 17.302 0.002 16.676 0.002 16.33 0.003 15.175 0.061 14.616 0.056 14.354 0.100 K7 F 0.85
10:01:35.76 2:03:34.7 11145 16.4887 0.0012 14.076 0.001 14.547 0.001 14.641 0.001 12.775 0.000 11.828 0.023 11.285 0.023 11.207 0.025 K3 F 0.24
10:00:54.50 2:16:05.1 11537 23.4609 0.2639 21.355 0.030 19.888 0.010 18.279 0.005 17.394 0.007 15.964 0.001 15.292 0.104 15.058 0.002 M5 p 0.24
10:01:28.50 1:59:32.4 11905 21.8892 0.0538 19.987 0.010 18.725 0.006 17.597 0.003 16.858 0.004 15.501 0.001 14.906 0.076 14.655 0.001 M2 p 0.57
10:01:02.45 2:22:29.7 12635 17.1883 0.0018 15.824 0.001 15.41 0.001 15.245 0.001 15.206 0.001 14.322 0.032 13.953 0.040 13.981 0.063 F8 F 1.68
Table 2Optical and near-IR properties of stellar sources detected in the Chandra COSMOS survey.

Notes. Columns 1-2: Source position (from optical images). Column 3: Chandra COSMOS ID. Columns 4-8: Optical photometry with errors. Columns 9-11: Near-IR photometry with errors. Column 12: Spectral type (sources with an asterisk have low-quality spectra and are accurate to subtypes). Column 13: Source of spectral classification: FAST (F), VIMOS (V), IMACS (I), VIMOS+IMACS (VI) or photometric (p). Column 13: Estimated distance based on spectral type.

Footnotes

  1. Advanced CCD Imaging Spectrometer
  2. Chandra Interactive Analysis of Observations, http://cxc.harvard.edu/ciao
  3. http://www.astro.psu.edu/xray/docs/TARA/ae_users_guide.html
  4. http://heasarc.nasa.gov/docs/xanadu/xspec
  5. http://astro.uni-tuebingen.de/nh
  6. Chandra COSMOS ID number.
  7. Website: http://cosmos.astro.caltech.edu/data/index.html
  8. Canada-France-Hawaii Telescope
  9. Inamori Magellan Areal Camera and Spectrograph
  10. VIsible MultiObject Spectrograph
  11. Very Large Telescope
  12. We have assumed that all these stars are on the main sequence since there are no known star forming regions along the line of sight, and evolved, late-type giants are known to be very weak X-ray emitters (e.g. Linsky & Haisch, 1979; Ayres et al., 1981) and therefore unlikely to be detected in our sample. It is possible that a giant might have an active secondary, and Covey et al. (2008) identify a number of confirmed and potential giant stars in their sample and find that 2-10% of their sample are likely to be giants. However, because giants are more luminous than dwarfs they are typically detected at greater distances, and therefore the X-ray emission from an active secondary would have to be particularly high to be detectable at such a distance. Therefore, while it should be noted that a small fraction of our sources could be giants, for simplicity we will assume that they all lie on the main-sequence.
  13. http://stellar.phys.appstate.edu/Standards/stdindex.html
  14. Website: http://simbad.u-strasbg.fr/simbad/

References

  1. Agüeros, M. A., et al.. 2009, ApJS, 181, 444
  2. Arnaud, K. A. 1996, in Astronomical Society of the Pacific Conference Series, Vol. 101, Astronomical Data Analysis Software and Systems V, ed. G. H. Jacoby & J. Barnes, 17–+
  3. Ayres, T. R., Linsky, J. L., Vaiana, G. S., Golub, L., & Rosner, R. 1981, ApJ, 250, 293
  4. Balucinska-Church, M., & McCammon, D. 1992, ApJ, 400, 699
  5. Broos, P., Townsley, L., Getman, K. V., & Bauer, F. 2002, ACIS Extract, An ACIS Point Source Extraction Package, Pennsylvania State University
  6. Capak, P., et al., 2007, ApJS, 172, 99
  7. Cash, W. 1979, ApJ, 228, 939
  8. Civano, F., et al., 2010, ApJS, submitted
  9. Covey, K. R., et al., 2008, ApJS, 178, 339
  10. Covey, K. R., et al., 2007, AJ, 134, 2398
  11. Cutri, R. M., et al., 2003, 2MASS All Sky Catalog of point sources. (The IRSA 2MASS All-Sky Point Source Catalog, NASA/IPAC Infrared Science Archive. http://irsa.ipac.caltech.edu/applications/Gator/)
  12. Drake, J. J., Peres, G., Orlando, S., Laming, J. M., & Maggio, A. 2000, ApJ, 545, 1074
  13. Elvis, M., et al., 2009, ApJS, 184, 158
  14. Fabricant, D., Cheimets, P., Caldwell, N., & Geary, J. 1998, PASP, 110, 79
  15. Fan, X. 1999, AJ, 117, 2528
  16. Feigelson, E. D., et al., 2002, ApJ, 574, 258
  17. Feigelson, E. D., et al., 2005, ApJS, 160, 379
  18. Feigelson, E. D., et al., 2004, ApJ, 611, 1107
  19. Flaccomio, E., Micela, G., & Sciortino, S. 2006, A&A, 455, 903
  20. Fleming, T. A., Schmitt, J. H. M. M., & Giampapa, M. S. 1995, ApJ, 450, 401
  21. Fruscione, A., et al., 2006, in Presented at the Society of Photo-Optical Instrumentation Engineers (SPIE) Conference, Vol. 6270, Observatory Operations: Strategies, Processes, and Systems. Edited by Silva, David R.; Doxsey, Rodger E.. Proceedings of the SPIE, Volume 6270, pp. 62701V (2006).
  22. Garmire, G. P., Bautz, M. W., Ford, P. G., Nousek, J. A., & Ricker, Jr., G. R. 2003, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 4851, Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, ed. J. E. Truemper & H. D. Tananbaum, 28–44
  23. Getman, K. V., Feigelson, E. D., Broos, P. S., Townsley, L. K., & Garmire, G. P. 2010, ApJ, 708, 1760
  24. Gioia, I. M., et al., 1984, in Bulletin of the American Astronomical Society, Vol. 16, Bulletin of the American Astronomical Society, 515
  25. Girardi, L., Groenewegen, M. A. T., Hatziminaoglou, E., & da Costa, L. 2005, A&A, 436, 895
  26. Gray, R. O., & Corbally, J., C. 2009, Stellar Spectral Classification (Princeton University Press)
  27. Güdel, M., Guinan, E. F., & Skinner, S. L. 1997, ApJ, 483, 947
  28. Heintz, W. D. 1992, ApJS, 83, 351
  29. Hünsch, M., Schmitt, J. H. M. M., Sterzik, M. F., & Voges, W. 1999a, A&AS, 135, 319
  30. —. 1999b, A&AS, 135, 319
  31. Kalberla, P. M. W., Burton, W. B., Hartmann, D., Arnal, E. M., Bajaja, E., Morras, R., & Pöppel, W. G. L. 2005, A&A, 440, 775
  32. Kenyon, S. J., & Hartmann, L. 1995, ApJS, 101, 117
  33. Kirkpatrick, J. D., Henry, T. J., & McCarthy, Jr., D. W. 1991, ApJS, 77, 417
  34. Kraus, A. L., & Hillenbrand, L. A. 2007, AJ, 134, 2340
  35. Lilly, S. J., et al., 2009, ApJS, 184, 218
  36. Lilly, S. J., et al., 2007, ApJS, 172, 70
  37. Linsky, J. L., & Haisch, B. M. 1979, ApJ, 229, L27
  38. López-Santiago, J., Micela, G., Sciortino, S., Favata, F., Caccianiga, A., Della Ceca, R., Severgnini, P., & Braito, V. 2007, A&A, 463, 165
  39. McCracken, H. J., et al., 2010, ApJ, 708, 202
  40. Micela, G. 2002, in Astronomical Society of the Pacific Conference Series, Vol. 277, Stellar Coronae in the Chandra and XMM-NEWTON Era, ed. F. Favata & J. J. Drake, 263
  41. Monet, D. G., et al., 2003, AJ, 125, 984
  42. Noyes, R. W., Hartmann, L. W., Baliunas, S. L., Duncan, D. K., & Vaughan, A. H. 1984, ApJ, 279, 763
  43. Ottmann, R., Fleming, T. A., & Pasquini, L. 1997, A&A, 322, 785
  44. Pallavicini, R., Golub, L., Rosner, R., Vaiana, G. S., Ayres, T., & Linsky, J. L. 1981, ApJ, 248, 279
  45. Peres, G., Orlando, S., Reale, F., Rosner, R., & Hudson, H. 2000, ApJ, 528, 537
  46. Puccetti, S., et al., 2009, ApJS, 185, 586
  47. Raymond, J. C., & Smith, B. W. 1977, ApJS, 35, 419
  48. Robin, A. C., et al., 2007, ApJS, 172, 545
  49. Salvato, M., et al., 2009, ApJ, 690, 1250
  50. Schmitt, J. H. M. M. 1997, A&A, 318, 215
  51. Schmitt, J. H. M. M., Collura, A., Sciortino, S., Vaiana, G. S., Harnden, Jr., F. R., & Rosner, R. 1990, ApJ, 365, 704
  52. Schmitt, J. H. M. M., & Liefke, C. 2004, A&A, 417, 651
  53. Schrijver, C. J., Mewe, R., & Walter, F. M. 1984, A&A, 138, 258
  54. Scoville, N., et al., 2007a, ApJS, 172, 38
  55. Scoville, N., et al., 2007b, ApJS, 172, 1
  56. Skumanich, A. 1972, ApJ, 171, 565
  57. Smith, R. K., Brickhouse, N. S., Liedahl, D. A., & Raymond, J. C. 2001, ApJ, 556, L91
  58. Smolcic, V., et al., 2004, ApJ, 615, L141
  59. Trump, J. R., et al., 2009, ApJ, 696, 1195
  60. Trump, J. R., et al., 2007, ApJS, 172, 383
  61. Vaiana, G. S., et al., 1981, ApJ, 245, 163
  62. Voges, W., et al., 1999, A&A, 349, 389
  63. Weisskopf, M. C., Brinkman, B., Canizares, C., Garmire, G., Murray, S., & Van Speybroeck, L. P. 2002, PASP, 114, 1
  64. Welsh, B. Y., et al., 2007, ApJS, 173, 673
  65. Wright, N. J., & Drake, J. J. 2009, ApJS, 184, 84
  66. Wright, N. J., Drake, J. J., Drew, J. E., & Vink, J. S. 2010, ApJ, 713, 871
  67. York, D. G., et al., 2000, AJ, 120, 1579
  68. Zickgraf, F., Krautter, J., Reffert, S., Alcalá, J. M., Mujica, R., Covino, E., & Sterzik, M. F. 2005, A&A, 433, 151
100209
This is a comment super asjknd jkasnjk adsnkj
""
The feedback cannot be empty
Submit
Cancel
Comments 0
""
The feedback cannot be empty
   
Add comment
Cancel

You’re adding your first comment!
How to quickly get a good reply:
  • Offer a constructive comment on the author work.
  • Add helpful links to code implementation or project page.