1: %VERSION: OW 2007-aug-02
2: \documentclass[12pt,preprint]{aastex}
3:
4: \shorttitle{X-Atlas}
5: \shortauthors{Westbrook et al.}
6:
7: \begin{document}
8: \title{X-Atlas: An Online Archive of Chandra's Stellar High Energy
9: Transmission Gratings Observations}
10:
11: \author{Owen W. Westbrook\altaffilmark{1}, Nancy Remage Evans, Scott
12: J. Wolk, Vinay L. Kashyap, Joy S. Nichols, Peter
13: J. Mendygral\altaffilmark{2}, Jonathan D. Slavin, Bradley Spitzbart}
14: \affil{Harvard-Smithsonian Center for Astrophysics, 60 Garden Street,
15: Cambridge, MA 02138}
16: \email{nevans@cfa.harvard.edu}
17:
18: \and
19:
20: \author{Wayne L. Waldron}
21: \affil{Eureka Scientific, Inc., 2452 Delmer Street Suite 100, Oakland,
22: CA 94602-3017}
23: \altaffiltext{1}{present address: Massachusetts Institute of
24: Technology, Cambridge, MA}
25: \altaffiltext{2}{present address: University of Minnesota, Twin
26: Cities, MN}
27:
28: \begin{abstract}
29: \begin{center}
30: \today
31: \end{center}
32: The high-resolution X-ray spectroscopy made
33: possible by the 1999 deployment of the {\it Chandra X-ray Observatory}
34: has revolutionized our understanding of stellar X-ray emission. Many
35: puzzles remain, though, particularly regarding the mechanisms of
36: X-ray emission from OB stars. Although numerous individual stars have
37: been observed in high-resolution, realizing the full
38: scientific potential of these observations will necessitate studying
39: the high-resolution $Chandra$ dataset as a whole. To facilitate the
40: rapid comparison and characterization of stellar spectra, we have
41: compiled a uniformly processed database of
42: all stars observed with the $Chandra$ High Energy Transmission Grating
43: (HETG). This database, known as X-Atlas, is accessible through a web
44: interface with searching, data retrieval, and interactive plotting
45: capabilities. For each target, X-Atlas also features predictions of
46: the low-resolution ACIS spectra convolved from the HETG
47: data for comparison with stellar sources in archival ACIS images.
48: Preliminary analyses of the hardness ratios, quantiles, and
49: spectral fits derived from the predicted ACIS spectra reveal
50: systematic differences between the high-mass and low-mass stars in the
51: atlas and offer evidence for at least two distinct classes of
52: high-mass stars. A high degree of X-ray variability is also seen in
53: both high and low-mass stars, including Capella, long thought to
54: exhibit minimal variability. X-Atlas contains over 130 observations of
55: approximately 25 high-mass stars and 40 low-mass stars and will be
56: updated as additional stellar HETG observations become public. The
57: atlas has recently expanded to non-stellar point sources,
58: and Low Energy Transmission Grating (LETG) observations are currently
59: being added as well.
60: \end{abstract}
61:
62: \keywords{stars: galactic --- stars: early-type --- stars: late-type
63: --- X-rays: stars --- {\it Chandra X-ray Observatory} --- transmission gratings
64: --- databases}
65:
66: \section{INTRODUCTION\label{s:intro}}
67:
68: $Chandra$ has made high-resolution X-ray spectra of bright
69: objects readily available. Over 130 observations of stars have been
70: made in high resolution and analyzed, resulting in both insight
71: and further questions about X-ray production, magnetic fields, stellar
72: winds, and line shapes. However, the extracted gratings spectra
73: are not currently expected to be part of the level 3 data products
74: slated to be released by the {\it Chandra X-ray Center} in 2008. We have therefore compiled a database, known as X-Atlas,\footnote{
75: {\tt http://cxc.harvard.edu/XATLAS/}} of all stars observed with the
76: $Chandra$ high and medium energy gratings (HETG) and extracted the
77: spectra in a uniform manner.
78:
79: Additionally, although over 60 stars have been observed in high
80: spectral resolution, thousands of stellar sources have been imaged
81: with the $Chandra$ ACIS CCDs, which have sufficient energy resolution
82: to produce an extractable low-resolution spectrum for each source.
83: To enable comparisons of the analyses made at low-resolution with
84: results from the high-resolution spectra, we have used the HETG
85: grating observations in X-Atlas to predict the low-resolution ACIS
86: spectra by convolving the HETG spectra with the response functions of
87: the ACIS detectors. These predicted low-resolution spectra, as well
88: as hardness ratios and quantiles calculated from them, are a valuable
89: tool for interpreting the physical properties of the large sample of
90: stars observed only at low spectral resolution. For further
91: comparison, we have also extracted the low-resolution zeroth-order
92: spectrum from the HETG observations.
93:
94: As a demonstration of the capabilities of the database, we
95: present some early science results from X-Atlas. We use the
96: high-resolution spectra to develop line ratio metrics for quick
97: characterization of the target. Analyses of the
98: predicted hardness ratios and quantiles of the stars in the atlas
99: indicate that high-mass and low-mass stars exhibit distinct
100: coronal temperature regimes and suggest that at least two classes of
101: high-mass OB stars exist. Variability analyses performed on each star
102: reveal that a high proportion of the stars in X-Atlas exhibit
103: variability during their $Chandra$ observation period. Even stars
104: generally considered minimally-variable, such as Capella, may undergo
105: count rate fluctuations between observations while remaining stable
106: over the time scale of a typical observation.
107:
108: Both the high-resolution and predicted low-resolution spectra are made
109: available in FITS and ASCII text formats through a dedicated web
110: interface. This web interface allows users to search the database and create
111: customized spectral plots of up to three sources, selecting either
112: individual lines or spectral regions for intercomparison. We note
113: that a similar spectral browsing utility has been created by
114: \citet{gon07} for the plotting of Reflection Grating Spectrometer
115: (RGS) observations from {\it XMM-Newton}, though without the data retrieval
116: and low-resolution predictions offered by X-Atlas. In addition to offering interactive plotting capabilities, we also present on the
117: X-Atlas web page the spectra in montages as functions of spectral type
118: and object class for high-mass OB stars, Wolf-Rayet stars, and low-mass stars.
119: Hardness ratios and quantiles also available for plotting and retrieval on
120: the web page.
121:
122: The X-Atlas spectral database will be updated as additional stellar
123: HETG observations are carried out and archived. Further, we have
124: applied the reduction and analysis methods used for normal
125: stars to HETG observations of other point sources, including neutron stars,
126: galactic novae, active galatic nuclei, and other exotic objects.
127: All LETG observations of both stellar and non-stellar
128: point sources will also eventually be included in the database.
129:
130: We describe the high-resolution spectral atlas in \S\ref{s:hires},
131: and the low-resolution predictions in \S\ref{s:lores}. We discuss
132: the characteristics of the sources in X-Atlas in \S\ref{s:discuss},
133: and summarize in \S\ref{s:summary}.
134:
135: \section{HIGH-RESOLUTION ATLAS\label{s:hires}}
136:
137: The initial motivation for creating X-Atlas was to make available for
138: comparison all of the {\it Chandra} high-resolution gratings
139: observations of OB stars. X-Atlas has since been expanded to include
140: low-mass stars as well. The database currently features over 130
141: observations\footnote{as of July 2007} of roughly 25 high-mass and 40
142: low-mass stars (Section 2.5), with additional observations being added as
143: they are made public.
144:
145: \subsection{HETG Observations and Data Reduction}
146:
147: The HETG consists of two grating arms: the High Energy Grating (HEG),
148: optimized for a spectral range of approximately 1.5 to 15 \AA, and the
149: Medium Energy Grating (MEG), sensitive up to 30 \AA\ \citep{wei02}. The
150: dispersed spectra produced by these grating arms cross at their respective
151: zeroth orders, creating the elongated 'x' for which X-Atlas is named.
152:
153: We have reprocessed each observation in the X-Atlas database uniformly,
154: eliminating the need for weighing different analysis techniques and
155: assumptions from observation to observation and enabling quick
156: comparisons of X-ray sources in the atlas. The {\it Chandra} team has
157: maintained a robust software package, {\it CIAO}, for the reduction and analysis of $Chandra$
158: data \citep{fru06}. Using the most recent version (3.4) of {\it
159: CIAO}, we have created a Perl pipeline script of the appropriate {\it
160: CIAO} tools and `threads'\footnote{{\tt
161: http://asc.harvard.edu/ciao/threads/}} needed to reduce and analyze
162: each observation and have added our own IDL routines and S-Lang
163: scripts for further analysis. The calibration files used for this
164: processing are those available with the $Chandra$ Calibration Database
165: (CALDB) 3.3.0. After retrieving the necessary files from the {\it Chandra} Data Archive
166: (CDA), this pipeline carries out a number of standard X-ray reduction
167: steps, removing various instrumental distortions, locating the
168: $0^{th}$-order centroid and the dispersed spectra of each gratings arm,
169: classifying the dispersed photons by spectral order, and coadding the
170: positive and negative grating spectral orders of the HEG and MEG
171: spectra before proceeding to processing and analyses developed
172: specifically for X-Atlas.
173:
174: Data preparation for a given observation begins with the level 1 event
175: file, a list of the arrival times and energies of incoming photons
176: that is available from the Chandra data archive. Most {\it Chandra}
177: archival data has undergone Standard Data Processing (SDP) during the
178: most recent reprocessing run, which began in February 2006, but for
179: those observations that have not yet been reprocessed, additional
180: corrections are necessary.
181:
182: Some portion of the X-ray events detected by {\it Chandra} are
183: caused by cosmic ray interactions in the CCD, and if not removed,
184: these `afterglow' events can be interpreted as faint sources.
185: Possible cosmic ray events are flagged during SDP, and are filtered
186: out in the level 2 event file. Note that the particular method used
187: in SDP versions prior to 7.4.0 has been shown to reject 3-5\% of valid
188: source photons\footnote{http://cxc.harvard.edu/ciao/threads/acisdetectafterglow/}.
189: This issue has been corrected in more recent processing versions, but
190: for archival data that has not yet been reprocessed appropriately, we
191: reset the afterglow flags on the level 1 event file and re-identify
192: afterglows and bad pixels with the updated reduction methods.
193:
194: In addition to afterglow detection, SDP also randomizes the
195: coordinates of events within the pixel where they are detected to
196: avoid aliasing effects and remove the 'gridded' appearance of the
197: data.\footnote{http://cxc.harvard.edu/ciao/threads/acispixrand/} For
198: X-Atlas reductions, we removed this randomization. Observations were
199: also corrected for the contamination of the optical/UV filter on
200: ACIS, a time-dependent drift in the detector gain values, and the
201: charge transfer inefficiency (CTI) of the detectors as charge is moved
202: from pixel to pixel during readout.
203:
204: The dispersed photons are extracted after the data are prepped as
205: described above. The {\it CIAO} wrapper script {\it tgdetect} was
206: used to locate the centroid of the $0^{th}$-order image. While {\it
207: tgdetect} is very accurate for the majority of HETG observations, if
208: the intended target is bright enough to cause the $0^{th}$-order image
209: to pile up, {\it tgdetect} may fail to properly detect the
210: centroid.\footnote{http://space.mit.edu/ASC/docs/memo\_fzero\_1.3.ps}
211: However, some of the most severe pileup effects, such as pulse saturation,
212: where the energy of a piled event is greater than the instrumental
213: threshold, are generally absent from stellar HETG observations in the
214: atlas. In cases where {\it tgdetect} does not accurately identify the
215: position of the zeroth order, the contributed {\it CIAO} tool {\it
216: findzero} will usually obtain a correct value. In addition, {\it
217: CXC's} current reprocessing effort of the entire archive correctly
218: determines the zeroth-order position for piled-up sources. In the interest of
219: applying uniform processing to each observation, we did not implement
220: more sophisticated centroid detection schemes. It is expected that
221: any in-depth scientific analysis of the stars in the atlas will
222: require separate reprocessing of the appropriate observations.
223:
224: After centroid detection, various {\it CIAO} tools determine the
225: locations of the grating arms and assign events from each arm to the
226: appropriate spectral orders. At this point, a level 2 event file is
227: created by filtering out bad grades previously marked as
228: afterglows and hot pixels, applying corrections for the effects of
229: excess streaking on chip 8, and selecting only those events falling
230: within the good time intervals determined by SDP.
231:
232: The grating spectra are then extracted from the destreaked level 2
233: event file. In addition, we create grating Redistribution Matrix
234: Files (gRMFs) to map between detector pulse height and energy space and
235: Ancillary Response Files (gARFs) to measure the instrumental area and
236: quantum efficiency of the HETG/ACIS-S configuration. The positive and
237: negative spectral orders of each arm are then coadded and stored in
238: six pulse height amplitude (PHA type I) spectral files, each of which
239: corresponds to either the HEG or MEG arm of HETG and to one of the
240: coadded $\pm1^{st}$, $\pm2^{nd}$, or $\pm3^{rd}$ orders. With the
241: high-resolution spectrum extracted, data reduction continues with
242: processing routines developed specifically for X-Atlas, starting with the
243: creation of the predicted low-resolution ACIS spectra, described in
244: section \S\ref{s:lores}.
245:
246: This data reduction procedure, including the convolution of the
247: predicted low-resolution ACIS spectra, was developed for stellar
248: observations but is applicable to any X-ray point source.
249: The current release of X-Atlas includes all point sources observed
250: with HETG, including cataclysmic variables, X-ray binaries, and active
251: galactic nuclei. The same methods are being adapted for use with observations
252: of point sources using the ACIS-S/LETG, HRC-S/LETG, and
253: HRC-I/LETG configurations as well. Because of the difficulties
254: involved in the spectroscopy of extended sources, only point sources
255: will be included in X-Atlas. There are no standard {\it CIAO} tools
256: for the analysis of grating spectra from extended sources. The
257: emission lines in such spectra will take on the shape of the
258: zeroth-order spatial structure convolved with the instrumental Line
259: Spread Function (LSF). Grating analysis of such sources therefore
260: requires spatial/spectral analysis because there is not a unique
261: mapping of wavelength to diffraction angle for extended sources.
262:
263: \subsection{Spectral Line Metrics}
264:
265: Generally, it is difficult to summarize the information contained
266: in a high-resolution spectrum in a few variables that capture the
267: relevant details about the behavior of the source. This is because
268: a coronal source spectrum is dominated by a large number of lines
269: from many elements at different ionization stages, in addition to
270: having an underlying continuum that is determined by the metallicity
271: of the plasma. We have identified a set of line and continuum
272: fluxes that together encompass the major parameters that describe
273: coronal plasmas: temperature, composition, and density. We measure
274: the counts in typically strong lines, such as the H-like and He-like
275: lines of O, Ne, Mg, Si, and S, as well as the prominent Fe\,XVII lines
276: at 15\AA\ and 17\AA. The counts due to the continuum are also measured
277: over different wavelength ranges ($<12.1$\AA, $>12.1$\AA) to characterize
278: the general shape of the spectra. The various ratios formed by these
279: fluxes serve as proxy for a complete description of the spectrum, but
280: in lower dimensional space. These metrics are listed in Table~\ref{tblmetrics},
281: and are provided in addition to the nominal hardness ratios and quantile
282: widths that are calculated after converting the grating spectra to the
283: lower resolution spectra corresponding to the imaging CCDs.
284:
285: \subsection{Coadding Multiple Observations of the Same Target}
286:
287: For observations of a single target that have been split over multiple
288: observing periods, additional processing is necessary to combine the
289: various exposures into a single dataset. We accomplish this
290: by adding the PHA spectra from each observation together after
291: ensuring that the source or instrument did not experience major
292: variability between observations.
293:
294: To determine whether a set of observations should be coadded, we
295: developed a number of simple tests. First, we check to see if the
296: actual source appeared to be variable from one observation to the
297: next by examining the count rates in both the $0^{th}$-order image
298: spectrum and the HEG and MEG grating spectra. The differences in the
299: total count rates between observations should not be significantly
300: greater than the combined error in the count rates. If either the
301: $0^{th}$-order or gratings spectra count rates have shifted more than
302: two-sigma between observations, then either the source or the
303: instrumental effective area has likely changed between observations.
304:
305: To roughly determine the degree to which changes in the effective area
306: could have contributed to a shift in the observed count rates from one
307: observation to the next, we compare the average fractional changes in
308: the effective area to the previously calculated fractional changes in
309: the count rate. As long as the fractional shifts in effective area
310: are statistically small compared to the fluctuations in the count
311: rates, changes in the HETG/ACIS-S instrument between each observation
312: can be deemed insignificant.
313:
314: After confirming that the source did not vary between observations,
315: the actual coadding process consists of summing together the
316: $\pm$$1^{st}$ order PHA datasets from each observation in their
317: original binning. The wavelength bins for each observation should
318: coincide; if not, the observations must be rebinned to the same
319: wavelength grid. Methods for testing the accuracy of the coaddition are
320: described in section 3.3. There are currently four X-Atlas targets
321: for which the coaddition of multiple observations has been performed
322: (Table \ref{tblcoadd}).
323:
324: \subsection{Extracting Spectra from Crowded Star Fields}
325:
326: A significant proportion of the stars featured in the atlas are
327: proximate to other X-ray bright sources, presenting both a scientific
328: bonus and a technical challenge. Since relatively few stars have been
329: observed at high-resolution, HETG observations of relatively crowded
330: star fields such as the Trapezium or Cyg OB2 present valuable
331: opportunities to enhance the science return of a single observation.
332: Actually performing multiple spectral extractions often proves
333: difficult, particularly if the spectral arms or zeroth-order image of a source
334: fall on the dispersed spectra of another. In such cases, all sources
335: must be extracted simultaneously, starting with the tool {\it
336: tg\_create\_mask}.\footnote{http://cxc.harvard.edu/ciao/ahelp/tg\_create\_mask.html}
337: Otherwise, artifacts may be present in the extracted spectra. In
338: general, we recommend that X-Atlas users extract any spectra of
339: interest themselves before undertaking detailed spectral analysis,
340: especially for stars situated in crowded fields.
341:
342: For the initial release of X-Atlas, we have extracted only the primary
343: target of the observation, usually the brightest star close to the ACIS-S
344: aimpoint. We are beginning to extract secondary and tertiary sources
345: from observations of clusters such as the Trapezium or Cyg OB2, using
346: the {\it SIMBAD} database to identify bright sources near the
347: aimpoint. We plan to limit the extractions to sources of more than
348: 300 $0^{th}$-order counts located no more than 4' off-axis. Beyond
349: 4', the ACIS point-spread function broadens dramatically and the 90\%
350: encircled energy radius rises well above 1'' \citep{cxc05}.
351:
352: \subsection{Observation Summary}
353:
354: Although the X-Atlas archive will eventually contain all point
355: sources observed with HETG or LETG, this article is intended to only
356: cover stellar sources currently in the atlas. At the time of
357: publication, X-Atlas consisted of all archived stellar HETG
358: observations, including the more than 130 observations shown in Table
359: \ref{tbl1}, with the exception of observations that suffered
360: instrumental malfunctions.
361:
362: \subsection{X-Atlas Website Interface}
363:
364: X-Atlas was conceived of as a public tool for research, and a primary
365: goal of the project was to make the data and analyses described here
366: readily available through a user-friendly web interface. We sought to
367: present the data in a form intelligible to people unfamiliar with X-ray
368: spectra while still offering useful capabilities to the expert X-ray
369: astronomer.
370:
371: The X-Atlas website is fully documented, with the main page containing
372: links to descriptions of the data processing procedures and threads
373: for navigating the website. Links to various {\it Chandra X-ray
374: Center} pages, other stellar catalogs, and X-ray emission line lists
375: are also provided. For an overview of the atlas's contents, a listing
376: of all observations, ordered by spectral class, can be pulled up, and
377: the proposal abstract and any publications associated with a
378: particular observation can be viewed.
379:
380: The central purpose of the website is to provide an interface for
381: searching, plotting, and retrieving the spectral data products
382: contained in the atlas (Figure \ref{figxweb}). The search interface
383: is modeled after the query pages of the $Chandra$ Data Archive.
384: The database may be queried by observation or target parameters such
385: as observation ID, target name, spectral type, object type, science
386: category, and observation date, or through a cone search of the target
387: RA and declination. Results may be sorted by spectral type,
388: observation ID, target name, observation date, RA, or declination.
389:
390: For those observations retrieved in the search, the instrumental
391: configuration of the observation, hardness and quantile data, known
392: physical parameters of the target, spectral fits of the $0^{th}$-order
393: and predicted low-resolution spectra, light curves and variability
394: analyses of the observation, the $0^{th}$-order image of the extracted
395: target, and preview plots of the HEG and MEG spectra, are readily
396: viewable. The spectral data products for the observation are also
397: made available for download. Observations returned by the search may
398: be marked for interactive, customized spectral plotting and data retrieval.
399:
400: Three sample uses of the X-Atlas website are outlined below. Note
401: that the procedures for generating the data products in the atlas are
402: discussed below (see \S\ref{s:lores} and \S\ref{s:discuss} ). Although these tasks are by no means a
403: comprehensive list of the atlas's capabilities, we believe that three
404: particularly useful applications of the website will be to view target
405: preview information for individual stars, create spectral plots for
406: the comparison of multiple targets, and retrieve and compare spectral
407: metrics such as hardness ratios and quantiles for a large sample of the
408: observations in the database.
409:
410: \subsubsection{Retrieving Target Information}
411:
412: For the first task, we will retrieve target preview information,
413: including spectral plots, hardness ratios, quantiles, spectral fits,
414: and lightcurves of a star of interest, using FK Com as an example. We
415: start at the X-Atlas search page, entering 'FK Com' as the target
416: name, and run the search. We may also search the atlas without any
417: search filters to retrieve all observations in the database and
418: select FK Com from the list of observations returned.
419:
420: In either case, after executing the search, we will find an entry for
421: FK Com, Obsid 614, among the observations retrieved. Clicking on the
422: Observation ID will call up $Chandra's$ instrumental configuration during the
423: observation. Clicking instead on the target name (FK Com) will bring
424: up a preview of the target extracted from of the observation. Under
425: some basic target information is a snapshot of the
426: $0^{th}$-order image along with the $0^{th}$-order count rate and a rough
427: estimation of the degree of pileup for a gauge of the reliability of
428: the $0^{th}$-order data. Below is a listing of predicted hardness
429: ratios and quantiles (see Section 4.1) based on the predicted ACIS-S
430: and ACIS-I spectra of FK Com (see Section 3), followed by plots of the spectral
431: fits of the $0^{th}$-order and predicted ACIS-S and
432: ACIS-I spectra (see Section 3.4). The relevant fit parameters are
433: listed as well, including N$_{H}$, two values of kT, metal abundances,
434: total flux, the flux emission ratio of the two kT components, and $\chi^{2}$ values of
435: the fit. Below the spectral fits is a light curve constructed from
436: the HETG HEG and MEG data and binned at intervals of 1200 and 4000 s.
437: Variability results from GLvary (described in section 4.2) accompany
438: the plot. Following the light curve are preview plots of the $1^{st}$
439: order HEG and MEG spectra in units of counts/s/\AA\ vs. \AA. Those
440: wishing to create their own plots of the spectra can simply click on
441: the 'Customize Spectral Plots' button provided on the page. Finally,
442: at the bottom of the preview page, various
443: spectral data products are offered for download, including PHA files
444: for the HEG and MEG spectra, FITS and text files of the reduced HEG
445: and MEG spectra, the predicted low-resolution ACIS and $0^{th}$-order spectra, and the
446: actual $0^{th}$-order spectrum, and the accompanying gARFs and aimpoint ARFs and RMFs.
447:
448: \subsubsection{Customizing Spectral Plots}
449:
450: A second, more powerful use of the atlas is to create comparison plots
451: of up to three targets at once. For the following example, we will compare
452: the MEG spectra and the Fe XVII emission lines for three T Tauri stars:
453: SU Aur, Speedy Mic, and TV Crt. To do so,
454: we return to the main X-Atlas search page, enter 'T Tauri' as
455: the object type and 'Normal stars and WD' as the science category, and
456: execute the search. This particular search will return over ten
457: observations of T Tauri stars, ordered by spectral type by default.
458: As we scroll through the list, we can click on the checkboxes in the
459: leftmost column for all observations of SU Aur (ObsID 3755), Speedy
460: Mic (ObsIDs 2536 and 3491), and TV Crt (ObsID 3728). After selecting
461: the appropriate observations, we click on 'Select Observation(s) for
462: Plotting' to call up the plotting interface.
463:
464: The plotting interface allows for the creation of customized spectral
465: plots for up to three targets at once. Three plotting options
466: are offered. The first, marked as 'Option A,' is to view pre-produced
467: plots of the high-resolution HEG or MEG spectra over one of a number
468: of pre-determined spectral ranges in units of counts and wavelength
469: (counts/s/\AA\ vs. \AA). Alternatively, the output plots may be customized through a variety of
470: plotting options. The most flexible option, 'Option B,' allows the
471: plotting of multiple datasets of differing spectral resolution and the
472: choice of plot units and spectral range of the output. The available
473: datasets include the high-resolution HETG/ACIS-S HEG and MEG spectra,
474: the predicted low-resolution ACIS-S or ACIS-I Aimpoint data, and both
475: the extracted and predicted low-resolution HETG/ACIS-S $0^{th}$-order
476: spectra. These data may be plotted in units of counts/s/\AA\ or
477: counts/s/bin vs. \AA, in counts/s/keV or counts/s/bin vs. keV, or, for the HETG
478: spectra only, in flux units of ph/s/\AA/cm$^{2}$ vs. \AA. Optional parameters
479: may be set to scale the low-resolution spectra, boxcar smooth the HETG
480: spectra, define upper bounds for the y-axis, or control the pixel
481: resolution of the output plots. These plots are generated with
482: identifications of common spectral lines provided.
483:
484: To plot the MEG spectra of SU Aur, Speedy Mic, and TV Crt from 5 \AA\ to 25 \AA, we
485: begin at 'Option B' on the webpage. First, up to three
486: desired targets are selected, and for each, the grating arm
487: (HEG or MEG) and order (1, 2, or 3) of interest. For this example, we
488: will select the $1^{st}$ order MEG spectra of SU Aur (ObsID 3755),
489: Speedy Mic (ObsID 2536), and TV Crt (ObsID 3728). The high-resolution
490: HETG/ACIS-S spectrum is already selected as the default dataset for step 2 and
491: counts/s/\AA\ vs. \AA\ is already selected as the default plot units
492: for step 3. In step 4, we enter '5' as the low bound and '25' as the
493: high bound and click on the button to 'Generate Plots.' To instead
494: view the predicted ACIS-S and ACIS-I spectra and plot in energy space,
495: we deselect the high-resolution HETG spectrum and select the predicted
496: ACIS spectra instead, choosing counts/s/keV vs. keV as the plot units.
497: The plot range may be either specified or left blank to use the
498: default values. Any of the low-resolution datasets may also be scaled
499: to the HETG spectrum for viewing on the same plot.
500:
501: \subsubsection{Emission Line Plots}
502:
503: Besides viewing pre-produced plots or specifying the dataset, plot
504: units, and plot range, a final plotting option is offered for the more
505: specific task of making comparisons of emission line profiles of the
506: HEG or MEG spectra. Up to six common X-ray emission lines can be
507: chosen from the drop-down boxes provided to create line profile montages in units of
508: wavelength (counts/s/\AA\ vs. \AA or counts/s/bin vs. \AA), energy
509: (counts/s/keV vs. keV), or Doppler velocity (counts/s/\AA\ vs. km/s).
510: For instance, to view the Fe XVII emission lines for SU Aur, Speedy
511: Mic, and TV Crt in wavelength units (counts/s/\AA\ vs. \AA), we
512: leave all of the fields under 'Option B' blank and instead proceed to
513: 'Option C.' Using any five of the six drop-down boxes, we
514: select the Fe XVII lines at 15.01, 15.26, 16.78, 17.05, and 17.10
515: \AA. We then make sure that the default plot units, counts/s/\AA\
516: vs. \AA, are selected. We may specify the distance from line center
517: of each line profile plot and choose between viewing all lines of a
518: given star on the same scale or plotting each line completely
519: independently. When we are satisfied with the settings, we may click
520: 'Generate Plot.'
521:
522: The interactively customized plots take only a few seconds to create.
523: The plots are then displayed on the screen along with options to
524: download the plots in GIF, JPEF, and/or postscript form. However, we intend the
525: plotting interface to be used mainly as a browsing tool rather than an
526: end in itself. Therefore, the datasets used to generate the plots are
527: provided for download in FITS files and ASCII text tables, and the HEG
528: and MEG PHA spectral files for the target are offered as well, along
529: with the appropriate grating ARFs and the ACIS-S and ACIS-I aimpoint
530: ARFs and RMFs.
531:
532: \subsubsection{Retrieving X-ray Properties of the X-Atlas Dataset}
533:
534: Besides the capabilities for viewing the observation data and analysis products for
535: individual observations or creating customized spectral plots, the
536: X-Atlas website also provides the means to make quick comparisons of
537: the X-ray properties of the entire X-Atlas dataset. For any subset of
538: observations in the database, hardness ratios and energy quantiles
539: derived from the predicted ACIS-S and ACIS-I spectra and spectral
540: metrics calculated from the HETG data may be retrieved by clicking on
541: the respective links for 'X-Atlas Hardness Ratios and Quantiles' and
542: 'X-Atlas Spectral Metrics' on the X-Atlas home page, the observation
543: search page, or any of the target preview pages. These links will
544: call up tables of these metrics for every X-Atlas observation that can
545: be filtered by the values of each metric and sorted by a variety of
546: target and observation parameters as well as by any of
547: the individual metrics themselves. A similar page is provided for the spectral
548: fits of the predicted ACIS spectra.
549:
550: \subsection{Sample High-Resolution Spectra: Spectral Montages}
551:
552: X-Atlas was conceived as a means of enabling comparisons of stellar
553: spectra across spectral type, hardness, or other characteristics
554: of interest. For this purpose, we have constructed montages of the
555: HETG spectra of both high-mass and low-mass stars to accompany the
556: dynamic plotting capabilities of the website. Only MEG spectra
557: are displayed below, but both HEG and MEG montages will be available on
558: the website, along with montages of the convolved low-resolution
559: spectra shown and described in Section 3. The spectra in the montages
560: are displayed in counts s$^{-1}$ \AA$^{-1}$ vs. \AA\ and are normalized for the
561: purpose of comparison.
562:
563: Normal OB stars are divided into one montage of supergiants and giants
564: (Figure \ref{figobsgh}) and another of giants, sub-giants, and dwarfs
565: (Figure \ref{figobmsh}), both ordered by spectral type. More exotic
566: high-mass stars are displayed as well. A separate montage is offered
567: of two stars, $\theta^{1}$ Ori C and $\tau$ Sco, in which a magnetic
568: field clearly plays a role in X-ray emission (Gagne, et al. 2005;
569: Donati, et al. 2007; Figure \ref{figobmagh}).
570: Several peculiar OB stars with unconfirmed X-ray emission mechanisms
571: are shown as well in Figure \ref{figobpech}. Two Wolf-Rayet stars are
572: shown in Figure \ref{figwrh}, including $\gamma$ Vel and five
573: observations of WR 140. A series of spectra of $\eta$ Car passing
574: through periastron was obtained by \citet{cor05} and is displayed in
575: Figure \ref{figetah}.
576:
577: The discovery of X-ray emission from high-mass stars in the Carina
578: nebula \citep{sew79,sew82} and in the Cyg OB2 association
579: \citep{har79} with the {\sl Einstein} satellite in the Carina nebula
580: \citep{sew79,sew82} and in Cyg OB2 \citep{har79} came as a surprise.
581: Since then, studies of the energy distributions of massive stars,
582: their spectra, and their line profiles have continued to enlighten and
583: puzzle us. Displaying normal OB stars in montages as a function of optical spectral type
584: reveals a clear shift of mean X-ray flux to lower energies at lower
585: photospheric temperatures for both the supergiants and giants (Figure \ref{figobsgh}) and giants,
586: sub-giants, and dwarfs (Figure \ref{figobmsh}). These findings were also
587: noted by \citet{walb06}. As he discusses, the constant ratio between
588: bolometric and X-ray luminosities implies such a spectral progression
589: since bolometric luminosity and temperature are related. Stellar
590: winds correlated with luminosity are likely to be the connection with
591: X-ray luminosity.
592:
593: Because of the diversity of astrophysical mechanisms at work in
594: low-mass stars, spectral montages of these stars are organized by
595: stellar type. In particular, separate montages have been made of
596: normal or rapidly rotating low-mass stars (Figure
597: \ref{figcoolnormh}), W UMa-type stars (Figure \ref{figwumah}), RS CVn
598: stars (Figure \ref{figrscvnh}), pre-main sequence T Tauri-type stars
599: (Figure \ref{figttsh}), and active or flaring low-mass stars (Figure
600: \ref{figflareh}). Two unique objects, Algol and FK Com, are also
601: shown (Figure \ref{figcoolpech}).
602:
603: \subsection{Sample High-Resolution Spectra: Line Profiles}
604:
605: Line profiles have been a key diagnostic and continuing puzzle for
606: both high-mass and low-mass stars. In addition to spectral montages, X-Atlas also
607: offers comparison plots of selected targets and emission lines to
608: illustrate some of their variety (Figures \ref{figlinprof}, \ref{figlprof2}).
609: As described previously, the interactive plotting interface allows the
610: creation of customized line profile comparisons for up to three
611: observations at a time.
612:
613: \section{LOW-RESOLUTION ATLAS\label{s:lores}}
614:
615: The high-resolution spectra can be used to predict the low-resolution
616: ACIS-S or ACIS-I image spectra that a source would have produced had
617: the HETG not been in place. This low-resolution prediction can then
618: be analyzed with spectral fitting tools such as {\it Sherpa} and
619: compared with actual low-resolution ACIS image spectra. By doing so,
620: we potentially gain greater insight into the spectral properties
621: of sources in existing ACIS images for which no high-resolution
622: grating observations currently exist.
623:
624: \subsection{Procedure}
625:
626: Convolving high-resolution HETG spectra to ACIS low-resolution imaging
627: mode spectra involves correcting for the differences in the effective
628: area response with and without the gratings in place. The effective
629: area is a product of the physical light collecting area of the
630: telescope, and the efficiencies of the mirror reflection, filters and
631: gratings, and the detector quantum efficiency. The extracted
632: high-resolution HEG and MEG spectra ($C_{HEG}$, $C_{MEG}$) and the
633: relevant effective areas ($gA_{HEG}$, $gA_{MEG}$) are first used to
634: project how the target object's high-resolution spectrum would appear
635: at the $0^{th}$-order ($f^{\rm 0th}$), or the ACIS-S or ACIS-I
636: aimpoints ($f^{\rm aimS}$, $f^{\rm aimI}$) with no grating in place.
637: The resulting fluxed spectra are then multiplied by the appropriate
638: $0^{th}$-order or aimpoint ARFs and convolved with the corresponding
639: Response Matrix File (RMF) to predict the low-resolution spectra of
640: each detector configuration ($C^{\rm 0th}$, $C^{\rm aimS}$, $C^{\rm aimI}$).
641:
642: The extraction of the high-resolution HEG and MEG spectra, described
643: in section 2.1, produces coadded $\pm1^{st}$, $\pm2^{nd}$, and $\pm3^{rd}$ order HEG
644: and MEG source and background spectra in units of counts per
645: wavelength bin. An adaptive binning scheme is then applied to the
646: source and background data
647: using a custom-built IDL routine from PINTofALE \citep{kas00}
648: such that every bin in the spectrum meets a minimum signal-to-noise ratio
649: while overall flux is conserved.\footnote{
650: {\tt http://hea-www.harvard.edu/PINTofALE/doc/PoA.html\#SMOOTHIE}
651: }
652: The appropriate signal-to-noise ratio for {\it smoothie} is
653: determined independently for the source and background data and is
654: generally set to be $\approx 3-4$. After
655: the spectra are binned, the source and background errors on each
656: adapted wavelength bin are calculated as the square root of the counts
657: in that bin. The smoothed background is then subtracted from the
658: smoothed source spectrum for both HEG and MEG, with all resulting
659: instances of negative count levels set to zero. Errors on the
660: background-subtracted counts/bin spectrum are calculated by
661: propagating the errors from the separate unsubtracted source and
662: background spectra.
663:
664: Before the high-resolution HEG and MEG spectra can be converted to
665: low-resolution, we first must determine how the HETG spectrum would
666: appear in high-resolution if observed with the effective area of ACIS-S or ACIS-I
667: rather than the full HETG/ACIS-S configuration. The first step in
668: this correction is to divide the HETG counts/bin spectrum by the
669: effective area of the HETG configuration. The effective area of the
670: gratings, in units of counts cm$^{-2}$ photons$^{-1}$, is contained in
671: a gratings ARF (gARF), one of which is produced for each grating arm
672: and spectral order during the extraction of the high-resolution
673: spectrum. The HEG and MEG gARF effective areas are interpolated onto
674: the spectral range of the appropriate high-resolution data sets, and the
675: adaptively binned high-resolution counts/bin spectrum is converted to a flux spectrum
676: (photons cm$^{-2}$ bin$^{-1}$) by dividing by the interpolated gARF
677: effective area where the gARF is nonzero. Where the gARF effective
678: area is zero, the high-resolution flux is set to zero. The errors on the flux
679: bins are calculated simply by dividing the counts/bin errors by
680: the gARF effective area and by setting the flux error equal to an arbitrary
681: large value where the effective area is zero.
682:
683: In order to predict the low-resolution spectrum of the ACIS-S or
684: ACIS-I Aimpoint, the high-resolution HEG and MEG spectra must first be
685: combined. Although the MEG sensitivity extends above 30 \AA, the HEG
686: spectra do not quite reach 20 \AA. Therefore, for the combined
687: spectrum, we define a new common HEG and MEG wavelength grid from 0
688: to 30 \AA \ at the spectral resolution of the HEG data (0.0025 \AA\ per
689: bin). The MEG wavelength grid offers slightly lower resolution
690: than the HEG (0.005 \AA\ per bin), so the MEG flux spectrum is
691: rebinned onto the common wavelength grid at the higher HEG resolution
692: using another {\it PINTofALE} \citep{kas00} function.\footnote{
693: {\tt http://hea-www.harvard.edu/PINTofALE/doc/PoA.html\#REBINW}
694: } The MEG flux errors are rebinned in similar fashion.
695:
696: Next, we calculate the combined HEG flux ($f_{HEG}$) and MEG flux
697: ($f_{MEG}$) over the common wavelength grid. For HEG and MEG fluxes
698: within the HEG wavelength range, the flux level in each wavelength bin
699: is computed by weighting the HEG and MEG fluxes by their respective
700: errors. The combined HEG+MEG flux,
701: \begin{equation}
702: f_{tot} =\ \frac{w_{HEG}f_{HEG}+w_{MEG}f_{MEG}}{w_{HEG}+w_{MEG}}\ ,
703: \end{equation}
704: where
705: \begin{equation}
706: w_{HEG} = \frac{1}{\sigma_{HEG}} ~,~ w_{MEG} = \frac{1}{\sigma_{MEG}} \,,
707: \end{equation}
708: where $\sigma_X$ are the bin-wise propagated errors in each spectrum.
709: For data beyond the HEG wavelength range, $w_{HEG}$ was set to 0.
710:
711: To predict the high-resolution spectrum of either the ACIS-S
712: aimpoint, the ACIS-I aimpoint, or the HETG/ACIS-S $0^{th}$-order, the combined HEG
713: and MEG spectrum is then multiplied by the corresponding effective area
714: of the aimpoint of interest, information contained in standardized
715: ARFs available from the $Chandra$ Data Archive,
716: \begin{eqnarray}
717: f^{\rm 0th} &=& A^{\rm 0th} f_{tot} \nonumber \\
718: f^{\rm aimS} &=& A^{\rm aimS} f_{tot} \nonumber \\
719: f^{\rm aimI} &=& A^{\rm aimI} f_{tot} \,.
720: \end{eqnarray}
721: Finally, the resulting prediction of the high-resolution counts/bin spectrum at the desired
722: aimpoint is convolved with the appropriate RMF, producing a
723: low-resolution prediction of the ACIS-S aimpoint, ACIS-I aimpoint, or
724: $0^{th}$-order spectrum:
725: \begin{eqnarray}
726: C^{\rm 0th} &=& Conv(RMF^{\rm 0th}, f^{\rm 0th}) \nonumber \\
727: C^{\rm aimS} &=& Conv(RMF^{\rm aimS}, f^{\rm aimS}) \nonumber \\
728: C^{\rm aimI} &=& Conv(RMF^{\rm aimI}, f^{\rm aimI}) \,.
729: \end{eqnarray}
730: The ACIS aimpoint ARFs and RMFs are updated every $Chandra$ cycle by
731: the $Chandra$ Data Archive. For consistency, we use the
732: Cycle 06 calibration products to generate the low-resolution ACIS
733: predictions for each observation in X-Atlas.
734:
735: Since the HEG and MEG coaddition process is entirely linear, the
736: combined HEG and MEG flux levels will fall between those found in the
737: separate HEG and MEG spectra. However, contrary to intuition, the
738: convolved low-resolution spectra do not follow this rule, i.e.,
739: it is possible that the low-resolution spectrum derived by
740: combining both HEG and MEG data could occassionally be either
741: greater or less than each of the component low-resolution
742: spectra. This effect can occur unless one of the conditions
743: \begin{eqnarray}
744: (a) && f_{HEG} \leq f_{tot} \leq f_{MEG} \nonumber \\
745: (b) && f_{HEG} \geq f_{tot} \geq f_{MEG} \,
746: \end{eqnarray}
747: holds exclusively for all wavelength bins (i.e., the averaged flux is greater
748: than the MEG flux and less than the HEG flux for every bin -- or the
749: reverse) then the following conditions will not hold:
750: \begin{eqnarray}
751: (a) && C^{\rm X}_{HEG} \leq C^{\rm X}_{tot} \leq C^{\rm X}_{MEG} \nonumber \\
752: (b) && C^{\rm X}_{HEG} \geq C^{\rm X}_{tot} \geq C^{\rm X}_{MEG} \,.
753: \end{eqnarray}
754:
755: \subsection{Test Results: Predicted Zeroth-Order Spectrum}
756:
757: To assess the accuracy of the convolution to low-resolution, we should
758: ideally compare our predictions with actual ACIS image spectra of a
759: given target. Unfortunately, ACIS imaging mode data are not always available for sources
760: observed with the HETG. But every grating observation does contain, in
761: addition to the high-resolution dispersed spectrum, a $0^{th}$-order
762: image from which a low-resolution spectrum may be extracted.\footnote{
763: {\tt http://cxc.harvard.edu/ciao/threads/pieces/index.html}}
764: Using the $0^{th}$-order ARF and RMF in place of the ACIS aimpoint ARF and RMF
765: in the high resolution to low resolution conversion process described above (\S\ref{s:lores}),
766: we can predict the $0^{th}$-order spectrum and compare
767: it with the actual extracted $0^{th}$-order spectrum to determine
768: whether the results of the convolution produce an accurate
769: low-resolution spectrum.
770:
771: This approach is unfortunately of limited use when the
772: $0^{th}$-order image has suffered significant pileup. Pileup occurs
773: when multiple photons are detected as a single event, causing both a
774: distortion the energy spectrum, as the apparent energy of a piled
775: event is actually the sum of two or more photons' energies, and a decrease in
776: the apparent count rate, as multiple photons are counted only once
777: \citep{cxc05}. For evaluation of the low-resolution convolution
778: procedure, we therefore selected several observations with average
779: count rates below 0.05 cts/s within a fifteen pixel radius of the
780: $0^{th}$-order, indicating relatively low to mild pileup (Table
781: \ref{tblpa_zero}). For these observations, we calculated the
782: $\chi^{2}$ values for the comparison of the observed and predicted
783: $0^{th}$-order spectra. Errors used to calculate the $\chi^{2}$
784: values include an extra 10\% systematic error attributable to
785: effective area uncertainties \citep{dra06}.
786:
787: For these observations, the predicted low-resolution $0^{th}$-order
788: counts spectrum exhibits a close correspondence to the
789: actual $0^{th}$-order spectrum. $\xi$ Per, from
790: observation ID 4512, is shown in Figure \ref{figzeroth}. From the success of the $0^{th}$-order
791: convolution, we have confidence that our ACIS-S and ACIS-I aimpoint
792: predictions of low-resolution spectra are indeed accurate.
793:
794: \subsection{Test Results: Coadded Observations}
795:
796: For spectra that have been obtained over multiple observations (see
797: \S{2.3}), the principles and methods described in the previous
798: section for predicting the low-resolution spectrum can be applied to
799: confirm the accuracy with which the observations were coadded.
800: Specifically, comparing the low-resolution prediction of the
801: $0^{th}$-order spectrum of the coadded observation to the actual coadded
802: $0^{th}$-order spectrum can give a measure of the success of both the
803: coaddition and the convolution to low-resolution.
804:
805: A low-resolution prediction of the $0^{th}$-order spectrum can be
806: generated from the HEG and MEG $1^{st}$ order spectra of the combined
807: observation using the same steps described in \S{3.1}. The actual
808: extracted $0^{th}$-order spectrum of the combined observation is created
809: simply by adding together the $0^{th}$-order Pulse Invariant counts
810: spectra of the individual observations and rebinning if necessary.
811:
812: The comparison of the extracted and predicted $0^{th}$-order spectra is
813: shown for the combined observation of HD 93250 in Figure
814: \ref{figcoadd}, composed of Observation IDs 5399, 5400, 7189, 7341,
815: and 7342. $\chi^{2}$ values are listed in Table \ref{tblcoadd} and
816: include additional 10\% errors in the observed spectra to approximate
817: calibration errors in the effective area.
818:
819: \subsection{Test Results: Spectral Fitting}
820:
821: Spectral fitting offers a powerful means to extract stellar
822: information from low-resolution ACIS spectra, including absorption
823: ($N_{H}$), temperature (kT), and metallicity. We have performed
824: spectral fits on our predicted low-resolution ACIS spectra both to
825: extract such information and to further evaluate the convolution
826: methods that produced the predicted spectra. Spectral fitting is art as well as science, and developing automated fitting routines for
827: bright sources such as are found in X-Atlas has proven difficult. In
828: addition, we have been forced to use techniques developed for non-dispersed
829: spectra on spectra that have been first dispersed by HETG and then
830: un-dispersed with the methods described above. However, by comparing
831: the results of our spectral fits with previously published values, we
832: have been able to verify the quality of our data and the potential for
833: extracting for accurate and useful scientific information from the
834: predicted ACIS spectra.
835:
836: The predicted ACIS spectra cannot be fit directly, since the
837: statistical errors of the original observation are scrambled during
838: the convolution process, and the errors on the predicted counts in
839: the different channels are no longer independent. In order to
840: obtain a realistic error on these spectra, we first adjust the
841: nominal exposure time such that a total of 1500 counts are predicted,
842: as in a well-exposed low-resolution ACIS observation. We then add Poisson
843: deviates corresponding to the counts intensity in each bin to simulate
844: the statistical error of an actual observation. (Note that this
845: process is similar to that adopted by the {\tt fakeit} commands
846: in {\sl Sherpa} and {\sl XSPEC}.) We then use {\it Sherpa} to perform
847: two-temperature fits of the predicted ACIS-S spectra. Where
848: $E(B-V)$ values are known (as is the case for a sample of high-mass stars),
849: we estimate the foreground H column density ($N_{\rm H}$),
850: \begin{equation}
851: N_{\rm H} = 5.9 \cdot 10^{21}\times E(B-V) ~~{\rm cm}^{-2}~~\,,
852: \end{equation}
853: and use this estimate to fix the column density during the spectral
854: fit. For hot stars, we estimated $E(B-V)$ from the observed $(B-V)$
855: and the intrinsic $(B-V)_{0}$ for O and early B stars ($(B-V)_{0}$ =
856: -0.32 \citep{sk82}).
857: Where $E(B-V)$ values are not known, we choose $N_{\rm H}=10^{20}$~cm$^{-2}$
858: as the starting value. We then employ {\it Sherpa} to fit the spectrum with
859: a 2-component {\tt mekal} (Mewe-Kaastra-Liedahl) thermal plasma model
860: \citep{mew85,mew86,kaa92,lie95}, using the {\tt cstat} statistic and the
861: Powell optimization method, extracting temperatures, metallicities, and
862: fluxes from each source. In order to ensure that the fitting is robust,
863: we carry out the fitting process in multiple stages. We first fit the
864: two temperatures, constrained to lie in disjoint regions ([0.01-1] and
865: [1-10] keV) for a fixed metallicity of 0.3 solar. The temperatures
866: found in this way are used as starting values to refit the grouped
867: spectrum, this time using the $\chi^{2}$ model-variance statistic and the
868: Levenberg-Marquardt optimization method, and allowing for an overlap
869: in the temperature ranges ([0.01-2] and [1-10] keV). The grouped
870: spectrum is then fit again, with $N_{\rm H}$ also being fit if
871: $E(B-V)$ is not known. Next, the temperatures and $N_{\rm H}$ are
872: fixed at their best-fit values and the metallicity is fit. In many
873: cases, especially when the $N_{\rm H}$ is known, this series of steps
874: yields an excellent fit.
875: For those cases which could not be well fitted with four free
876: parameters ($N_{\rm H}$, two temperatures and metalicity)
877: as noted by temperatures up against range limits and/or $\chi^2$/
878: d.o.f. $>$ 2, the
879: data are refit in a modified manner. The temperatures
880: (and $N_{\rm H}$, if necessary) are fit again, freezing
881: the metallicity to the best-fit value, before finally fitting the
882: spectra with three free parameters ($N_{\rm H}$ and
883: two temperatures). We obtain error bars using the {\sl Sherpa}
884: command {\tt projection}.
885:
886: A selection of the fit results is displayed in Table \ref{tblfit}.
887: The fit values compare favorably with published results
888: for $\theta^{1}$ Ori C. \citet{sch00}
889: report temperatures ranging from 5 to 61 $\times 10^{6}$ K, or
890: 0.4 to 5.3 keV, whereas our analyses finds
891: high-temperature components of about 10 to $>$15 keV and low-temperature
892: components of 1.1 to 1.4 keV. For $\theta^{2}$ Ori A,
893: \citet{sch06} measure temperatures of above 25 $\times 10^{6}$ K (2.2
894: keV) in the low activity phase and temperatures from 3 to 100 $\times 10^{6}$
895: K (0.3 to 9 keV) in the high activity phase. We find temperatures of
896: 0.79 and 0.34 keV in the low phase and 3.52 and 1.14 keV for the high phase,
897: in reasonable agreement for a variable target.
898:
899: For $\gamma$ Cas, \citet{smi04} report that the X-ray emission is
900: dominated by a heavily absorbed hot component at 12 keV, with cooler components near
901: 0.1, 0.4, and 3 keV exhibiting much lower absorption. We found only a
902: hot component at $>$15 keV with a relatively high absorption. The
903: inability of our methods to differentiate the absorptions of the
904: different emission components mostly likely explains why the cooler
905: components were not detected despite comprising about 14\%
906: of the total emission \citep{smi04}.
907:
908: Our methods produced highly consistent results over fifteen
909: observations of Capella. Each observation
910: returned a primary temperature component between 0.56 and 0.61 keV,
911: with three observations also returning hotter components ranging from
912: 4 to $>$10 keV. \citet{can00} examined plasma diagnostics in the
913: high-resolution HETG spectra and found a broad range in temperatures from
914: log $T$ = 6.3 to 7.2 ($\sim$0.1 to 1.2 keV), while \citet{bri00} report a peak emission
915: measure near log $T$ = 6.8 (0.4 keV) from extreme ultraviolet
916: measurements.
917:
918: While spectral fits of the Capella observations usually returned only
919: one temperature, all eight observations of the RS CVn star IM Peg returned two
920: emission components when fit. The high temperature component ranged from 1.3
921: to 3.4 keV, while the low temperature component fell between 0.3 and
922: 1.0 keV. These values agree reasonably well with \citet{hue01}, who
923: found emission peaks at log $T$ = 7.2 ($\sim$1.2 keV) in the non-flaring state and
924: log $T$ = 7.6 ($\sim$2.5 keV) in the flaring state.
925:
926: On the whole, our results are similar to those published by more
927: painstaking high-resolution analyses. Our goal is not to supplant these more accurate
928: methods but to establish the credibility of our predicted
929: low-resolution ACIS data. The variations between our results and those published emphasize the
930: oversimplification that is necessarily involved in reducing the
931: coronal temperature to two distinct values.
932:
933: \subsection{Sample Predicted Low-Resolution Spectra: Spectral Montages}
934:
935: For every high-resolution spectral montage we have produced (see section
936: 2.4), we have also created a montage of the predicted low-resolution
937: spectra of the same targets. Only predicted ACIS-S spectra are
938: displayed here, but montages of the predicted ACIS-I spectra will also
939: be made available online. As with the HETG montages, all spectra
940: shown here are normalized for viewing purposes (Figures \ref{figobsgl}
941: $-$ \ref{figflarel}).
942:
943: \section{DISCUSSION\label{s:discuss}}
944:
945: We proceed with a discussion of the properties of the ensemble of
946: stellar spectra, continuing from the summary of
947: high-resolution spectra in Section 2.7. We discuss the hardness ratios and quantiles
948: derived from the predicted low-resolution spectra in section 4.1 and
949: the X-ray variability of individual observations in Section 4.2.
950:
951: \subsection{The X-ray Color of Stars: Quantiles and Hardness ratios}
952:
953: While high-resolution spectral studies by $Chandra$ have proven
954: uniquely powerful in their ability to uncover the detailed
955: astrophysics in the source, the time required by these observations
956: limits them to no more than about 20 stellar observations per year. A database such as
957: X-Atlas enhances the high-resolution dataset by providing
958: capabilities for comparison of the low-resolution spectra of the
959: vast numbers of stars in ACIS images in the $Chandra$ archive with the
960: relatively few well-studied examples. Two techniques for reducing
961: spectral data to X-ray "color" are hardness ratios and quantiles.
962:
963: Hardness ratios are comparisons of the total number of counts observed
964: in different X-ray passbands. Using a Bayesian estimation approach
965: developed by \citet{par06}, we calculated projected hardness ratios
966: for each star in the atlas, using the predicted ACIS-S and ACIS-I
967: low-resolution spectra derived from the original HETG observations.
968: For each target, the predicted ACIS counts were divided into soft,
969: medium, and hard passbands, with energies of 0.3-0.9 keV, 0.9-1.5 keV,
970: and 1.5-8.0 keV respectively.
971:
972: Then, for each star, two hardness ratios were calculated (see Table \ref{tblqh}):
973: \begin{equation}
974: HR_{1}=\frac{H-M}{H+M}
975: \end{equation}
976: and
977: \begin{equation}
978: HR_{2}=\frac{M-S}{M+S}
979: \end{equation}
980: S, M, and H are the total number of predicted counts in the
981: soft, medium, and hard energy passbands, respectively. Error bars
982: were also calculated using the methods described in \citet{par06}.
983: For most observations, the number of predicted counts was high enough
984: that the error bars were insignificant compared to the
985: scatter of the hardness plots. Capella, for instance, exhibits relatively
986: consistent hardness over twelve observations spanning 1999 to 2006.
987: However, the dispersion in the hardness ratios among observations
988: exceeds the 1-sigma error bars, implying that even a relatively
989: unvarying star will undergo statistically significant shifts in hardness over
990: time. Capella's variability is discussed further in Section 4.3.
991:
992: When the values of each star are plotted with $HR_{1}$ vs. $HR_{2}$, a
993: clear trend emerges (Figure \ref{fighr}). Both high-mass
994: and low-mass stars appear to follow a common, confined locus arcing
995: from softer emission near $\sim$ (-0.6,-0.8) through (0.0,-0.7) to
996: harder emission at (0.9, 0.6). The high-mass stars actually occupy the
997: extrema of the locus and the cool stars occupy a smaller range
998: extending from $\sim$ (0.0,-0.2) to (0.5, 0.0). The outliers, which
999: include a $\sim$ 25 ks observation of the Wolf-Rayet star WR 140
1000: and several observations of $\eta$ Carinae, will be discussed further below.
1001:
1002: The observed hardness of a source is a function of temperature and
1003: interstellar absorption. To model this dependence, we used a Mekal
1004: emissivity model \citep{mew85,mew86} to predict the expected relationship
1005: between $HR_{1}$ and $HR_{2}$. Model photon fluxes in each hardness
1006: band were calculated from the integrated Mekal emissivities for a
1007: range of temperatures and $N_{H}$ values, assuming interstellar medium
1008: (ISM) absorption cross-sections. Normalization of the emissivities
1009: was not necessary since hardness ratios depend solely on the ratio of the
1010: defined energy bands. We then used the Cycle 06 ACIS-S and ACIS-I
1011: response functions to predict the X-ray counts in each passband,
1012: yielding hardness ratios derived from the Mekal fluxes for targets
1013: observed at the ACIS-S and ACIS-I aimpoints.
1014:
1015: In Figures \ref{fighrmodel1} and \ref{fighrmodel01}, we show the
1016: results of modeling the predicted dependence of $HR_{1}$ on
1017: $HR_{2}$ for ACIS-I count spectra. Using Mekal emissivities for
1018: typical coronal temperatures and absorptions, we overlay lines of constant
1019: temperature and absorption on separate plots for two abundance
1020: values: solar abundance in Figure \ref{fighrmodel1} and 0.1 times
1021: solar abundance (for all elements except H) in Figure
1022: \ref{fighrmodel01}. The lines of constant temperature range from 2.5
1023: to 50 MK (0.16 to 3 keV) for $N_{H}$ values between 0.01 and 2.5$\times$ 10$^{22}$/cm$^{2}$, with absorption increasing from left to right
1024: along each contour. Similarly, the lines of constant $N_{H}$ include
1025: values between 0.01 and 2.5 $\times$ 10$^{22}$/cm$^{2}$, with
1026: temperature increasing from 2.5 to 50 MK left to right on each
1027: contour. According to this model, sources with low absorption and
1028: temperature will appear in the lower left of the hardness plot, and
1029: hotter, more highly absorbed stars will fall at the harder end of the
1030: plot towards the upper right. In addition, at the soft end of the
1031: hardness plot, both the temperature and $N_{H}$ contours exhibit
1032: degeneracy, indicating that we cannot uniquely determine
1033: a given star's temperature and $N_{H}$ from its hardness ratios with
1034: this model. However, since the majority of the stars observed
1035: with HETG have low absorption, we expect most of the stars in X-Atlas to
1036: fall close to the leftmost endpoints of the appropriate temperature
1037: contour. Under this assumption of low absorption, we can use the
1038: $HR_{2}$ value of a star as a rough measure of temperature up to
1039: around (0.5, -0.6), or $\sim$15 MK ($\sim$1 keV), where the constant
1040: temperature contours fall perpendicular rather than parallel to lines
1041: of constant $N_{H}$.
1042:
1043: As noted previously, the high-mass stars show in Figure \ref{fighr} occupy two
1044: distinct ranges at the endpoints of the locus sweeping from $\sim$
1045: (-0.6,-0.8) to (0.9, 0.6), with the majority at the soft end of the
1046: plot between approximately (-0.5,-0.8) and (0.0,-0.7) and the
1047: remainder falling at the hard end from around (0.4, -0.3) to
1048: (0.9, 0.5). The low-mass stars lie in the phase space between these
1049: two groups on the hardness plot, with some overlap between the groups.
1050: Assuming low absorption, we can identify the group of
1051: high-mass stars in the lower left of the plot as corresponding to
1052: single-temperature fits of below $\sim$7.5 MK ($\sim$0.5 keV), while
1053: the low-mass stars reside in a warmer region between roughly 7.5 and 15 MK
1054: (0.5 to 1 keV). The second grouping of high-mass stars is indicative
1055: of harder emission and therefore some combination of higher temperatures and $N_{H}$ values than are seen in
1056: the first group of high-mass stars and the low-mass stars. While the first, cooler
1057: high-mass star grouping consists largely of normal OB stars, the
1058: second, harder grouping of high-mass stars includes several
1059: unusual objects, including the magnetic O star $\theta^{1}$ Ori C, the
1060: Wolf-Rayet stars WR 140 and HD 68273, and peculiar stars Cyg OB2 8A
1061: and Gamma Cas. $\tau$ Sco, another known magnetic O star, appears in
1062: the cooler group of high-mass stars, though at the harder end of that
1063: subset. Intriguingly, one star, $\theta^{2}$ Ori A, has moved
1064: from the first high-mass star group to the second in two observations
1065: separated by 20 days. This shift between two X-ray emission states is
1066: further corroborated by differences in the HETG spectra, quantile
1067: measurements and variability tests and is explored further below. For
1068: a more thorough discussion of $\theta^{2}$ Ori A's variability see \citet{sch06}.
1069:
1070: Both high-mass star groups fall roughly within the contour lines for
1071: both solar and 0.1 times solar abundance. This correspondence
1072: suggests that hardness ratios do not provide a useful means of
1073: distinguishing high-mass coronal abundances. The low-mass stars,
1074: however, do not show good agreement with the Mekal model calculated
1075: with solar abundance (Figure \ref{fighrmodel1}) but the fit improves
1076: dramatically when the non-H abundances are lowered by a factor of
1077: 0.1. To some extent, this relatively low abundance value may
1078: accurately reflect low coronal abundances in low-mass stars.
1079: \citet{sce07}, for example, found that the coronal abundances of 20
1080: X-ray bright pre-main sequence stars were significantly depleted in
1081: Fe, with an average of less than 0.2 of the solar photosphere. In
1082: addition, spectral fits of the predicted ACIS spectra of low-mass
1083: stars generally found abundances in the 0.1 to 0.4 times solar range. However, lowering the abundance may
1084: treat the symptoms of deviations between the model and the
1085: observations without telling the whole story. For OB stars,
1086: \citet{wal94} has shown that if a point on the hardness ratio plot
1087: does not fall within the predicted range using ISM opacities, then
1088: stellar wind absorption, not a variation in abundance, is most likely
1089: responsible for the discrepancy. The presence of multiple X-ray
1090: emission components could similarly skew the calculated hardness
1091: ratios and imitate a decrease in abundance. In addition, our hardness
1092: model depends on the Mekal model's treatment of abundances in calculating
1093: emissivities. Therefore, although the coronae of low-mass stars might
1094: conceivably have depleted abundances relative to solar, our model
1095: lacks the sophistication to differentiate between actual variations in
1096: abundance and other effects.
1097:
1098: The models shown in Figures \ref{fighrmodel1} and \ref{fighrmodel01}
1099: do account for the outlying observations of WR 140 at approximately
1100: (0.42, 0.95) and of $\eta$ Car between $\sim$(0.0, 0.95) and
1101: $\sim$(0.6,0.95). WR 140 was observed pre-- and post--periastron with
1102: observations separated by five months. The hardness model suggests that some
1103: combination of a large shift in column density and a rise in
1104: temperature could be responsible for the observed change in hardness.
1105: The model does not enable us to characterize the change in
1106: temperature, since thermal information is lost in the hardness plots
1107: at high absorption. We can confirm, however, that the HETG
1108: spectrum of the post-periastron observation appears strongly absorbed
1109: between 1-2 keV in a way not present in the first observation. This
1110: discrepancy is most likely due to absorption by the cool WC material
1111: along the line of sight between $Chandra$ and the region of wind
1112: interaction. This observation, then, is the only one that appears
1113: strongly affected by absorption besides the six observations of $\eta$
1114: Car. $\eta$ Car is a massive, highly-luminous star thought to produce
1115: X-rays through wind collisions with a massive companion
1116: \citep{ham07}. The spectra from series of {\it XMM-Newton} and {\it Chandra}
1117: observations have been fit by \citet{ham07}, who report kT values
1118: ranging from 2.9 to 5.4 keV (33 to 63 MK) and $N_{H}$ between 4.8 and
1119: 42 $\times$ 10$^{22}$/cm$^{2}$. The combination of high temperature
1120: and high absorption values places $\eta$ Car at the extreme hard end
1121: of the temperature contours in Figures \ref{fighrmodel1} and
1122: \ref{fighrmodel01}. The appropriate $N_{H}$ contours for $\eta$ Car
1123: are not shown on the plots.
1124:
1125: Both ACIS-I and ACIS-S versions of the hardness plots are available on
1126: the X-Atlas web page for each observation,
1127: although only the ACIS-I plot is shown here.
1128: To see the plots, use the main search page to find an observation,
1129: and then click on ``Object Spectral Type".
1130: The web page offers the additional capability of quickly identifying
1131: and highlighting individual stars of interest for both ACIS-I and
1132: ACIS-S plots.
1133:
1134: Hardness ratio plots for ACIS-I and ACIS-S are very similar.
1135: Any systematic shifts between the ACIS-I and ACIS-S
1136: versions of the hardness plots are due to the increased sensitivity at
1137: low energy of the back--illuminated S3 chip. This characteristic
1138: leads to more counts for soft sources in the "S" passband and hence
1139: lower values of $HR_{2}$ for the predicted ACIS-S data. While S3
1140: provides a somewhat broader range in $HR_{2}$, the majority of stars
1141: observed without benefit of the gratings have been observed on
1142: front-illuminated CCDs.
1143:
1144: Hardness ratio calculations are most useful if the spectra in question have
1145: multiple counts in each passband. While that criterion is certainly
1146: met for any of the predicted ACIS spectra in X-Atlas, actual ACIS
1147: observations can contain thousands of detectable sources with of order
1148: one hundred counts. When a source has few counts or is absorbed,
1149: there is a strong likelihood of low counts in one of the three
1150: pre-determined passbands. These circumstances can lead to very large
1151: and non-uniform errors in hardness calculations due to the location or
1152: type of the source.
1153:
1154: For such sources, quantile analysis provides an alternative technique
1155: to hardness ratios for classifying X-ray sources through X-ray color
1156: measurements \citep{hon04}. Since the method uses the quantile (in
1157: this case quartile) of the source's own photons, the statistics are
1158: uniform across all bands. We have applied the quantile analysis
1159: methods to X-Atlas to enable comparisons between faint X-ray sources
1160: found in ACIS images and the bright stars observed with HETG.
1161:
1162: Like the hardness ratio calculations, quantile analysis is performed on the
1163: predicted ACIS-S or ACIS-I aimpoint spectra. Using the methods
1164: described in \citet{hon04}, we take the total predicted counts within a
1165: 0.3-8.0 keV range and determine the energies that divide those
1166: predicted counts into fractions of 25\%, 50\%, and
1167: 75\%. Upon finding $E_{x\%}$, the energy below which x\% of the total
1168: counts fall, we calculate normalized quantiles $Q_{x}$:
1169: \begin{equation}
1170: Q_{x}=\frac{E_{x\%}-E_{lo}}{E_{hi}-E_{lo}},
1171: \end{equation}
1172: where $E_{lo}$ = 0.3 keV and $E_{hi}$ = 8.0 keV.
1173:
1174: The typical quantile based color-color diagram (QCCD) plots
1175: 3*$Q_{25}$/$Q_{75}$, a measure of the emission from the middle half of
1176: the spectrum, against the median ($m$=$Q_{50}$), a measure of the
1177: overall hardness of the source, in the form $log_{10}$[$m$/(1-$m$)].
1178: The particular form of the x-axis median expression is chosen to avoid
1179: compressing the regions of phase space corresponding to harder X-ray
1180: emission \citep{hon04}. The ratio of the $75^{th}$ percentile photon
1181: to the $25^{th}$ percentile photon is a function of both the temperature
1182: and the absorption of the source. This dependence means that as the absorption
1183: varies, the quantile ratio ($Q_{75}/Q_{25}$) for a given temperature
1184: may be different. In order to interpret the resulting QCCD, we
1185: overlay a model grid of hydrogen column depth ($N_{H}$) and
1186: temperature (kT). This grid is determined from the ACIS-S or ACIS-I
1187: aimpoint ARF and combined interstellar absorption and Raymond-Smith
1188: (1977) thermal models (Figures \ref{fighquant}, \ref{figcquant}).
1189: We used Raymond--Smith plasma (Raymond \& Smith 1977) for general
1190: purpose fitting of
1191: low resolution coronal spectrum. While more recent models including
1192: MeKa (Mewe et al.\ 1985)
1193: and APEC (Smith et al.\ 2001) have more lines included,
1194: Raymond-Smith models were found by
1195: Wolk et al.\ (2006) to have the lowest residual $\chi^2$
1196: (degree~of~freedom) and/or less outliers
1197: in a direct comparison of fitting about 80 moderately bright stellar
1198: sources using Sherpa.
1199: The same study found that all three models gave similar results for
1200: kT and nH for stars with between 100 and 1000 counts.
1201: On the quantile plots Figures \ref{fighquant} and \ref{figcquant}
1202: one can distinguish changes in temperature from extinction and
1203: even thermal and non-thermal changes. Further, the grids shown in
1204: Figures \ref{fighquant} and \ref{figcquant} are sensitive to
1205: metallicity.
1206:
1207: Two things become immediately clear from these plots. First, as noted
1208: previously, the vast majority of stars observed with HETG had
1209: relatively low absorption. No low-mass star exceeds $\sim N_{H}\ 5
1210: \times 10^{21}$ cm$^{-2}$ and only about five high-mass stars exceed
1211: that value. Second, while low-mass stars are smoothly distributed
1212: along the temperaturea axis in this plot, with temperatures ranging from
1213: about 1 to 3 keV, the high-mass stars are clearly congregated at the
1214: edges. The appropriate one-temperature representations appear to be
1215: either below about 1 keV or above about 3 keV with few high-mass stars
1216: occupying the middle ground dominated by the low-mass stars. This
1217: distribution has two clear implications. First, the X-ray generation
1218: mechanism is clearly different between the high and low-mass stars.
1219: This result is not unexpected, as many other lines of argument
1220: have led to the development of various dynamo-driven models for the
1221: generation of X-rays in low-mass stars. More surprising is the bimodal distribution of
1222: high-mass stars, which suggests that two different physical states are
1223: at work. A similar bimodality was noted in the hardness ratios of the
1224: high-mass stars (Figure \ref{fighr}), and as expected, the same
1225: stars that appear at the warmer, harder end of the hardness ratio
1226: plot generally lie in the high-temperature group on the quantiles
1227: plots. Once again, $\theta^{2}$ Ori A has moved from one group to the
1228: other between the two observations.
1229:
1230: $Chandra$ grating spectra have provided material for deliberation about
1231: complexities and puzzles in high-mass star interpretation. There are a
1232: number of possible explanations for high-mass stellar X-ray emission relevant
1233: to X-Atlas. Colliding winds in binaries provide high temperature
1234: plasma in some cases (e.g. \citet{ham07}), while magnetic activity
1235: plays a role in others. Quantile measurements of stars in the atlas
1236: reveal that about 50\% of the high-mass sample is cool
1237: with one-temperature component fits below $\sim$ 1 keV. This result
1238: is consistent with the standard model of X-rays originating in shocked
1239: material in line-driven spherically symmetric winds \citep{mac91}. Typical temperatures in these models are about 10MK ($\sim$ 900 eV). Hot
1240: X-ray emission from OB stars has been the subject of recent investigations.
1241: \citet{sch00,sch01,sch03} reported temperatures of about 5 keV from
1242: $\theta^{1}$ Ori C. Schulz et al. conclude that the component of the
1243: emission at this temperature must be formed ``near the terminal
1244: velocity of the wind,'' at about seven stellar radii from the
1245: photosphere. \citet{gag05} show that such emission can be fitted by a
1246: two-dimensional MHD magnetospheric wind model due to \citet{udd02}.
1247: Several other examples of X-ray hot O and B stars have since been
1248: found \citep{lop06,rak06}. \citet{mul06} have proposed a
1249: two-component scenario by which the cooler X-ray emission is generated
1250: by line-driven, spherically-symmetric winds, and a second component of
1251: the wind emerges from magnetically active regions in polar caps that
1252: may extend as low as 45$^{\rm o}$. In such a scenario, whether an O
1253: star is found to be a relatively hot or soft source may simply be a
1254: function of inclination.
1255:
1256: We note that in the star-forming region RCW 108, about 75\% of the OB
1257: star candidates in the field are found to have X-ray temperatures
1258: above 3 keV (Wolk et al. 2007). The fraction of high-temperature stars was
1259: greatest in the RCW 108--IR embedded cluster, which is still actively
1260: forming stars. The inner group indicates an evolutionary aspect to
1261: the observed temperature, in contradiction of models that propose that
1262: the observed temperature may be a function of the observation's line--of--sight
1263: unless the extent of the polar cap decreases in time. X-Atlas
1264: contains a sample of relatively old high-mass stars in which the
1265: cooler variety are in the majority. This characteristic may indicate
1266: that X-ray temperatures in high-mass stars decline with time, just as
1267: they do in cool stars, although the X-ray generation mechanisms are different.
1268:
1269: The various types of low-mass/cool stars studied with $Chandra$ seem to be
1270: very tightly correlated in Figure \ref{figcquant}. For example, the bulk of the F and G
1271: stars lie between y--axis values of 1.75 and 1.95, while the general
1272: RS CVn and T Tauri star populations have y--axis values between 1.4 and 1.7.
1273: We ran model grids at several metallicities ranging from 0.001 to 5 times solar.
1274: While no particular metallicity seemed to improve the relation between
1275: the positions of the high-mass/hot stars and the model grids, we found
1276: that for low-mass/cool stars with cool coronae at $<$ 1 keV, a best fit
1277: to the data was a metallicity of about 0.3. For the warmer coronae,
1278: the best fit was a metallicity of about 0.1 (See Figure
1279: \ref{figcquant}). The fact that the stars are still tightly
1280: correlated despite the high metal sensitivity of the model grids is
1281: indicative of very uniform metallicity among the observed coronae.
1282: This leads us to believe that the handling of the metallicity by the
1283: models as a function of temperature, not a change in the metallicity, is the root of the
1284: apparent metallicity discontinuity. Indeed, given the sensitivity of
1285: quantiles to metallicity shown here, the metallicity must be very
1286: similar in all the observed low-mass/cool star coronae.
1287:
1288: \subsection{Variability Analysis}
1289:
1290: Variability studies allow us to assess the plausibility of different X-ray
1291: generation mechanisms. These can be constrained by the timescales and
1292: flux changes observed in the variability. Specifically,
1293: for low mass stars, the cooling
1294: time depends on the loop length, plasma density, conductivity and
1295: emissivity (See \citet{rea02} for a review). \citet{tes04} undertook
1296: one such
1297: study based on 26 observations of 22 low-mass stars observed with
1298: HETG during the first
1299: three years of the Chandra mission. They concur with the findings of
1300: \citet{hue01} who found that while the continuum emission changes
1301: significantly during flares, the
1302: line emission does not. However, this is not uniformly the case.
1303: Audard et al.\ (2001) found that in the case of the
1304: RS~CVn HR 1099 Ne stays constant at the quiescent level during a
1305: flare but the detected a flux increase in the
1306: Fe XXIV lines during a flare. G{\"u}del et al. (2001) also found
1307: enhancment in the Fe XXIV lines during a flare of the rapidly rotating
1308: ZAMS star AB~Dor.
1309:
1310: A large proportion of the stars in X-Atlas appear to be variable. To test for
1311: variability in the flux level over the course of an observation, we constructed
1312: light curves and performed variability analysis on the observations
1313: using two methods. The first method was simple binning. {\it ACIS
1314: Grating Light Curve
1315: (aglc)}\footnote{http://space.mit.edu/cxc/analysis/aglc/index.html}
1316: was used to create light curves from the HEG and MEG HETG data. Light
1317: curves with bin sizes of 4000 and 1200 seconds were extracted.
1318:
1319: In a more quantitative approach, all observations were analyzed using the
1320: algorithms developed by \citet{gl92} as adapted specifically for
1321: $Chandra$ data \citep{rots06}. This algorithm uses maximum-likelihood
1322: statistics and evaluates a large number of possible break points
1323: against the prediction of constancy.
1324:
1325: An overview of the Gregory--Loredo analysis is presented in Table
1326: \ref{tblvar}. At the time of this analysis there were 131 observations
1327: of 64 sources (24 hot stars and 40 cool stars) in the database. The
1328: criteria for variability was the probability returned by the
1329: Gregory--Loredo test. If the probability exceeded 99\% the star was
1330: considered "variable." If the probability was between 90\% and 99\%,
1331: the star was deemed "probably variable." Stars with variability
1332: probability below 90\% were considered "not variable." A source was
1333: considered variable if any observation of that source was noted to be
1334: variable. Epoch to epoch comparisons were not undertaken in a
1335: systematic way for this study.
1336:
1337: Table \ref{tblvar} is designed only to give an example of the content
1338: of the database. The high-mass star classification combines single OB
1339: stars, multiple OB stars and Wolf-Rayet stars. We find a variability
1340: rate of $\sim$ 40\% among the high-mass stars. This is a remarkably
1341: high rate since the X-ray production mechanisms of high-mass stars are
1342: expected to be wind driven and hence more stable than dynamo-driven
1343: low-mass stars, which are subject to loop reconnection at all scales.
1344: A closer examination of the data reveals that of the variables, several
1345: are unique objects, including Algol, $\gamma$ Cas, $\theta^{1}$ Ori C,
1346: and $\eta$ Carina.
1347:
1348: Although it contains a B8 star, Algol also contains a K1 subgiant.
1349: The latter tends to be active. The flare seen from Algol {\em is of
1350: particular interest}, though, as it appears to have induced a change
1351: in the apparent characteristic flux level. HD 150136, 9 Sgr, $\tau$ Sco,
1352: $\tau$ CMa and HD 206267 are generally well-behaved OB stars that all
1353: exhibited small deviations on short time scales (10-20 ks). $\gamma$
1354: Cas is a unique hot star that shows fluctuations of about 25\% on
1355: short time scales. The variability of $\theta^{1}$ Ori C has been
1356: previously reported \citep{sch01,gag05} and is thought to
1357: be related to its relatively high temperature and strong variable
1358: magnetic field. Support for this hypothesis comes from the
1359: observations of $\theta^{2}$ Ori A (discussed in further detail in
1360: \citet{sch06}). This star shows variability in only one of its two
1361: HETG observations but is also seen to shift from a cool to a hot state
1362: as evidenced by the two-temperature fit in which the kT value jumps
1363: from a single temperature of around 0.7 keV in epoch one to two
1364: temperatures of approximately 17 and 2 keV in epoch two
1365: (Table \ref{tblfit}). The flux is seen to drop steadily and by 25\%
1366: over the course of the hotter, but generally brighter, epoch.
1367:
1368: About 85\% of the low-mass stars are found to be variable. This
1369: proportion is well in excess of the typical fraction reported in star
1370: forming regions, which is about 30\% in a 100 ks observation
1371: (Cf. \citet{get02,wol06}). This apparent discrepancy is undoubtedly
1372: due to counting statistics and
1373: selection effects. Wolk et al.\ (2007) report that in a 90~ks
1374: observation of RCW 108, 66\% of stars with over 200 counts are
1375: detected as variable but only 21\% of stars with between 100 and 200
1376: net counts. In the X-Atlas database almost all the stars are bright
1377: and have over 1000 $0^{th}$-order counts. The results for low-mass stars
1378: here are consistent with the results of the COUP survey \citep{get05}
1379: which found the majority of the $\sim$1600 stars monitored for $\sim$10
1380: days to be variable at some level.
1381:
1382: With the exception of the composite system Algol, no high-mass star
1383: was seen to flare. Among the low-mass stars, TZ CrB, CC Eri, TW Hya,
1384: and DoAr 21 showed the strongest flares, more than doubling their flux
1385: in less than one hour with cooling times in excess of 10 ks. The
1386: cooling time of stars with disks, such as TW Hya, is of particular
1387: interest since the cooling time is related to the loop length. Such
1388: systems can provide direct evidence of star--disk coupling via the
1389: magnetic field. While these are the longest cooling times in the
1390: group, 10-30 ks is not unusually long (cf. \citet{fav05}). The
1391: X-Atlas database is a very biased sample for flare studies. Many
1392: stars were targeted precisely because they were known flare stars
1393: (e.g. AD Leo, EV Lac, and CC Eri) or rotationally variable (e.g. HD
1394: 223460, FK Com and AB Dor).
1395:
1396: \subsection{Variability of Capella}
1397:
1398: Over twelve observations from 1999 to 2006, Capella's X-ray hardness
1399: ratios are clustered at $(HR_1,HR_2)=(-0.75\pm0.01,-0.058\pm0.025)$
1400: (see Figure~\ref{fighr}). These are significantly overdispersed compared to the typical
1401: errors in the individual hardness ratios, by factors ranging
1402: from 5x to 12x. Over this same period, the overall X-ray count
1403: rate undergoes variations of $\approx17$\%. However, the hardness
1404: ratios are uncorrelated with the count rates: we find correlations
1405: of Spearman's $\rho=-0.033\ (p=0.9)$ for $HR_1$ versus rate, and
1406: $\rho=0.11\ (p=0.7)$ for $HR_2$ versus rate.
1407:
1408: Within each observation, there is no evidence for any variability,
1409: as is confirmed by an application of the Gregory-Loredo test
1410: \citep{gl92}: the odds that the star varies during any of the
1411: individual observation is $<90$\%. In contrast, combining 15 of the
1412: 16 available observations (excepting ObsID 1199, a 2 ks observation)
1413: to construct a concatenated event list spanning over 350~ks, and
1414: applying the same test to this concatenated events list, we find
1415: that the odds of Capella being variable are 10:1, indicating a probability
1416: of over 0.99 that Capella is variable on timescales of years.
1417:
1418: Capella, as the strongest coronal source, has been a common X-ray
1419: calibration target for many instruments. It has remained stable
1420: in its total X-ray output even as the detailed emission structure
1421: has shown considerable change (see e.g., \citet{you01}). For
1422: instance, \citet{dup96} have shown based on EUVE data that the high-temperature component is
1423: variable, a finding confirmed by \citet{raa07} based on
1424: LETG data. There is also considerable evidence for the
1425: dominant emission to change between the G1\,III primary and the
1426: G8\,III secondary (e.g., \citet{lin98,joh02,ish06}).
1427:
1428: We confirm the primary conclusions of \citet{raa07}, who
1429: found no evidence of changes in intensity over timescales of $<100$~ks,
1430: but a large timescale variation that resulted in an increased flux
1431: early 2006. We find no evidence for variations at timescales $<50$~ks,
1432: and conclusively find that the source is variable over larger
1433: timescales.
1434:
1435: \section{SUMMARY\label{s:summary}}
1436:
1437: By developing a uniformly processed stellar database,
1438: we intend to provide a powerful yet accessible tool for the study of
1439: stellar X-ray emission. As the early science results from X-Atlas
1440: demonstrate, the ability to compare the high-mass and low-mass stars
1441: observed with HETG opens up intriguing avenues of investigation. The
1442: spectral montages and line-profile plots created from the
1443: high-resolution HETG data offer a detailed glimpse at the physical
1444: mechanisms for X-ray generation among stellar types. In addition, by
1445: convolving the high-resolution spectra to low-resolution, we can
1446: compare the limited number of stars studied in high-resolution to the
1447: thousands of sources in $Chandra$ images. Spectral fits and X-ray
1448: color analyses performed on the predicted low-resolution spectra
1449: reveal differences in the X-ray temperature distributions of high-mass
1450: and low-mass stars and provide evidence for at least two distinct
1451: classes of X-ray emission in high-mass stars. Variability studies
1452: help to further constrain the causes of stellar X-ray emission and
1453: suggest that the proportion of high-mass stars exhibiting X-ray variability may
1454: be greater than previously suspected. We also definitively establish
1455: the X-ray variability of Capella over timescales of years. We have made
1456: available the means to conduct systematic studies of the stellar HETG
1457: dataset, and therefore hope that these early results will represent only the
1458: beginnings of the scientific return of the database. Finally, we have
1459: applied the reduction and analysis methods described above to
1460: all non-stellar point sources observed with HETG, including neutron
1461: stars, X-ray binaries, and active galatic nuclei. Future releases of
1462: X-Atlas will adapt these methods to LETG point source observations as
1463: well, extending the atlas to all $Chandra$ gratings observation of point
1464: sources.
1465:
1466: \acknowledgments
1467:
1468: We are happy to thank the referee John Pye for a careful reading
1469: and a number of improvements.
1470: Funding for this work was provided by $Chandra$ Grants GO5-6006E and
1471: GO5-6007A, Chandra X-ray Center NASA Contract NAS8-39073, and by
1472: ANCHORS (AN archive of CHandra Observations of Regions of Star
1473: formation), $Chandra$ archival project AR5-6002A. Wayne L. Waldron
1474: was supported in part by $Chandra$ grant GO5-6006A.
1475:
1476: We would like to thank N. S. Schulz, D. P. Huenemoerder,
1477: and P. Testa for feedback during the testing stages of the project.
1478:
1479: \clearpage
1480:
1481: \begin{thebibliography}{}
1482: \bibitem[Audard et al.(2001)]{2001A&A...365L.318A} Audard, M., G{\"u}del,
1483: M., \& Mewe, R.\ 2001, \aap, 365, L318
1484: \bibitem[Brickhouse et al. (2000)]{bri00} Brickhouse, N. S., Dupree,
1485: A. K., Edgar, R. J., Liedahl, D. A., Drake, S. A., White, N. E., \&
1486: Singh, K. P., 2000, \apj, 530, 387
1487: \bibitem[Canizares et al. (2000)]{can00} Canizares et al., 2000, \apj,
1488: 539, L41
1489: \bibitem[Canizares et al. (2005)]{can05} Canizares, C. R., Davis,
1490: J. E., Dewey, D., Flanagan, K. A., Galton, E. B., Huenemoerder, D. P.,
1491: Ishibashi, K., Markert, T. H., Marshall, H. L., McGuirk, M.,
1492: Schattenburg, M. L., Schulz, N. S., Smith, H. I., Wise, M., 2005,
1493: \pasp, 117, 1144
1494: %\bibitem[Cash (1979)]{cas79} Cash, W., 1979, \apj, 228, 939
1495: \bibitem[CIAO (2006)]{ciao06} Chandra Interactive Analysis of Observations (CIAO), http://cxc.harvard.edu/ciao/
1496: \bibitem[Proposers' Observatory Guide (2005)]{cxc05} Chandra X-ray Center, 'The
1497: Chandra Proposers' Observatory Guide,' Vers. 8.0, Cambridge, MA:
1498: Dec. 2005
1499: \bibitem[Corcoran et al. (2005)]{cor05} Corcoran, M. F., Hamaguchi,
1500: K., Henley, D. B., Ishibashi, K., Gull, T., Nielsen, K., Pittard,
1501: J. M., 2005, \baas, 37, 1347
1502: \bibitem[De Becker et al. (2004)]{deb04} De Becker, M., Rauw, G., \&
1503: Swings, J. P., 2004, \apss, 297, 291
1504: \bibitem[De Marco \& Schmutz (1999)]{dem99} De Marco, O., \& Schmutz,
1505: W., 1999, \aap, 345, 163
1506: \bibitem[Donati, et al. (2006)]{don06}Donati, J.-F., Howarth, I. D.,
1507: Jardine, M. M., Petit, P., Catala, C., Landstreet, J. D., Bouret, J.-C.,
1508: Alecian, E., Barnes, J. R., Forveille, T., Paletou, F., and Manset, N.
1509: 2006, \mnras, 370, 629
1510: %\bibitem[Dougherty et al. (2005)]{dou05} Dougherty, S. M., Beasley,
1511: %A. J., Claussen, M. J., Zauderer, B. A., \& Bolingbroke, N. J., 2005,
1512: %\apj, 623, 447
1513: \bibitem[Drake et al. (2006)]{dra06} Drake, J. J., Ratzlaff, P.,
1514: Kashyap, V., Edgar, R., Izem, R., Jerius, D., Siemiginowska, A.,
1515: Vikhlinin, A., 2006, \procspie, 6270, 62701I
1516: \bibitem[Dupree \& Brickhouse (1996)]{dup96} Dupree, A.K., \&
1517: Brickhouse, N.S., 1996, in Poster Proc., IAU Symp.\ 176: Stellar
1518: Surface Structure (Wien: Institut f\"{u}r Astronomie), 184
1519: \bibitem[Evans et al. (2003)]{eva03} Evans, N. R., Seward, F. D.,
1520: Krauss, M. I., Isobe, T., Nichols, J., Schlegel, E. M., Wolk, S. J.,
1521: 2003, \apj, 589, 509
1522: \bibitem[Favata et al. (2005)]{fav05} Favata, F., Flaccomio, E.,
1523: Reale, F., Micela, G., Sciortino, S., Shang, H., Stassum, K. G.,
1524: Feigelson, E. D., 2005, \apjs, 160, 469
1525: \bibitem[Fruscione et al. (2006)]{fru06} Fruscione, A., 2006,
1526: \procspie, 6270, 62701V, D. R. Silvia \& R. E. Doxsey, eds.
1527: \bibitem[Gagn\'{e} et al. (2005)]{gag05} Gagn\'{e}, M., Oksala, M. E.,
1528: Cohen, D. H., Tonnesen, S. K., ud-Doula, A., Owocki, S. P., Townsend,
1529: R. H. D., \& MacFarlane, J. J., 2005, \apj, 628, 986
1530: \bibitem[Getman et al. (2002)]{get02} Getman, K. V., Feigelson, E. D.,
1531: Townsley, L., Bally, J., Lada, C. J., Reipurth, B., 2002, \apj, 575, 354
1532: \bibitem[Getman et al. (2005)]{get05} Getman, K. V., et al., 2005,
1533: \apjs, 160, 319
1534: \bibitem[Gies et al. (1993)]{gie93} Gies, D. R., et al., 1993, \aj,
1535: 106, 2072
1536: \bibitem[Gonz\'{a}lez-Riestra \& Rodriguez-Pascual (2007)]{gon07}
1537: Gonz\'{a}lez-Riestra, R. \& Rodriguez-Pascual, P. M., 2007, ``BiRD: A
1538: Browsing Utility for RGS Spectra,'' X-ray Spectroscopy Workshop, July 11-13
1539: \bibitem[Gregory \& Loredo (1992)]{gl92} Gregory, P. C. \& Loredo,
1540: T. J., 1992, \apj, 398, 146
1541: \bibitem[G{\"u}del et al.(2001)]{2001A&A...365L.336G} G{\"u}del, M., et
1542: al.\ 2001, \aap, 365, L336
1543: \bibitem[Hamaguchi et al. (2007)]{ham07} Hamaguchi, K., et al., 2007,
1544: \apj, 663, 522
1545: \bibitem[Harnden et al. (1979)]{har79} Harnden, F. R., Jr.,
1546: Branduardi, G., Gorenstein, P., Grindlay, J., Rosner, R., Topka, K.,
1547: Elvis, M., Pye, J. P., \& Vaiana, G. S., 1979, \apjl, 234, L51
1548: \bibitem[Hong et al. (2004)]{hon04} Hong, J., Schlegel, E. M. \&
1549: Grindlay, J. E., 2004, \apj, 614, 508
1550: \bibitem[Houck \& Denicola (2000)]{hou00} Houck, J. C., \& Denicola,
1551: L. A., 2000, in ASP Conf. Ser. 216: Astronomical Data Analysis
1552: Software and Systems IX, Vol. 9, 591
1553: \bibitem[Huenemoerder et al. (1993)]{hue93} Huenemoerder, D. P.,
1554: Ramsey, L. W., Buzasi, D. L., Nations, H. L., 1993, \apj, 404, 316
1555: \bibitem[Huenemoerder et al. (2001)]{hue01} Huenemoerder, D. P.,
1556: Canizares, C. R., \& Schulz, N. S., 2001, \apj, 559, 1135
1557: \bibitem[Ishibashi et al. (2006)]{ish06} Ishibashi, K., Dewey, D.,
1558: Huenemoerder, D. P., \& Testa, P., 2006, \apj, 644, 117
1559: \bibitem[Johnson et al. (2002)]{joh02} Johnson, O., Drake, J. J.,
1560: Kashyap, V., Brickhouse, N. S., Dupree, A. K., Freeman, P., Young, P. R.,
1561: \& Kriss, G.A., 2002, \apj, 565, 97
1562: \bibitem[Kaastra (1992)]{kaa92} Kaastra, J. S., 1992, 'An X-Ray
1563: Spectral Code for Optically Thin Plasmas, Internal SRON-Leiden Report,
1564: version 2.0
1565: \bibitem[Kashyap \& Drake(2000)]{kas00} Kashyap, V. L. \& Drake,
1566: J. J., 2000, BASI, 28, 475
1567: \bibitem[Liedahl et al. (1995)]{lie95} Liedahl, D. A., Osterheld,
1568: A. L., \& Goldstein, W. H., 1995, \apjl, 438, 115
1569: \bibitem[Linsky et al. (1998)]{lin98} Linsky, J. L., Wood, B. E.,
1570: Brown, A., \& Osten, R. A., 1998, \apj, 492, 767
1571: \bibitem[Lopes de Oliveira et al. (2006)]{lop06} Lopes de Oliveira, R.,
1572: Motch, C., Haberl, F., Negueruela, I., \& Janot-Pacheco, E., 2006,
1573: \aap, 454, 265
1574: \bibitem[MacFarlane et al. (1991)]{mac91} MacFarlane, J. J.,
1575: Cassinelli, J. P., Welsh, B. Y., Vedder, P. W., Vallerga, J. V., \&
1576: Waldron, W. L., 1991, \apj, 380, 564
1577: \bibitem[Ma\'{i}z-Apell\'{a}niz et al. (2004)]{mai04}
1578: Ma\'{i}z-Apell\'{a}niz, J., Walborn, N. R., Galu\'{e}, H. \'{A}., Wei, L. H., 2004, \apjs, 151, 103
1579: \bibitem[Mewe et al. (1985)]{mew85} Mewe, R., Gronenschild,
1580: E. H. B. M., \& van den Oord, G. H. J., 1985, \aap, 62, 197
1581: \bibitem[Mewe et al. (1986)]{mew86} Mewe, R., Lemen, J. R., \& van den
1582: Oord, G. H. J., 1986, \aap, 65, 511
1583: \bibitem[Millour et al. (2007)]{mil07} Millour, F., et al., 2007, \aap, 464, 107
1584: %Petrov, R. G.,
1585: %Chesneau, O., Bonneau, D., Dessart, L., Bechet, C., Tallon-Bosc, I.,
1586: %Tallon, M., Thiébaut, E., Vakili, F., Malbet, F., Mourard, D.,
1587: %Antonelli, P., Beckmann, U., Bresson, Y., Chelli, A., Dugué, M.,
1588: %Duvert, G., Gennari, S., Glück, L., Kern, P., Lagarde, S., Le Coarer,
1589: %E., Lisi, F., Perraut, K., Puget, P., Rantakyrö, F., Robbe-Dubois, S.,
1590: %Roussel, A., Tatulli, E., Weigelt, G., Zins, G., Accardo, M., Acke,
1591: %B., Agabi, K., Altariba, E., Arezki, B., Aristidi, E., Baffa, C.,
1592: %Behrend, J., Blöcker, T., Bonhomme, S., Busoni, S., Cassaing, F.,
1593: %Clausse, J.-M., Colin, J., Connot, C., Delboulbé, A., Domiciano de
1594: %Souza, A., Driebe, T., Feautrier, P., Ferruzzi, D., Forveille, T.,
1595: %Fossat, E., Foy, R., Fraix-Burnet, D., Gallardo, A., Giani, E., Gil,
1596: %C., Glentzlin, A., Heiden, M., Heininger, M., Hernandez Utrera, O.,
1597: %Hofmann, K.-H., Kamm, D., Kiekebusch, M., Kraus, S., Le Contel, D., Le
1598: %Contel, J.-M., Lesourd, T., Lopez, B., Lopez, M., Magnard, Y.,
1599: %Marconi, A., Mars, G., Martinot-Lagarde, G., Mathias, P., Mège, P.,
1600: %Monin, J.-L., Mouillet, D., Nussbaum, E., Ohnaka, K., Pacheco, J.,
1601: %Perrier, C., Rabbia, Y., Rebattu, S., Reynaud, F., Richichi, A.,
1602: %Robini, A., Sacchettini, M., Schertl, D., Schöller, M., Solscheid, W.,
1603: %Spang, A., Stee, P., Stefanini, P., Tasso, D., Testi, L., von der
1604: %Lühe, O., Valtier, J.-C., Vannier, M., Ventura, N., 2007, \aap, 464, 107
1605: \bibitem[Morrison \& McCammon (1983)]{mor83} Morrison, R. \& McCammon,
1606: D., \apj, 270, 119
1607: \bibitem[Mullan \& Waldron (2006)]{mul06} Mullan, D. J. \& Waldron,
1608: W. L., \apj, 637, 506
1609: \bibitem[Park et al. (2006)]{par06} Park, T., Kashyap, V. L.,
1610: Siemiginowska, A., van Dyk, D. A., Zezas, A., Heinke, C., Wargelin,
1611: B. J., 2006, \apj, 652, 610
1612: \bibitem[Preibisch (1999)]{pre99} Preibisch, T., 1999, \aap, 345, 583
1613: \bibitem[Raassen \& Kaastra (2007)]{raa07} Raassen, A. J. J., Kaastra,
1614: J. S., \aap, 461, 679
1615: \bibitem[Rakowski et al. (2006)]{rak06} Rakowski, C. E., Schulz,
1616: N. S., Wolk, S. J., \& Testa, P., \apj, 649, L111
1617: \bibitem[Raymond \& Smith (1977)]{ray77} Raymond, J. C. \& Smith, B. W.,
1618: 1977, \apjs, 35, 419
1619: \bibitem[Reale (2002)]{rea02} Reale, F., 2002, in ASP Conf. Ser.,
1620: Vol. 277, 103
1621: \bibitem[Rots (2006)]{rots06} Rots, A. H., 2006, in ASP
1622: Conf. Ser. 351: Astronomical Data Analysis Software and Systems XV,
1623: Vol. 351, 73
1624: \bibitem[Schulz et al. (2000)]{sch00} Schulz, N. S., Canizares, C. R.,
1625: Huenemoerder, D., \& Lee, J. C., 2000, \apjl, 545, 135
1626: \bibitem[Schulz et al. (2001)]{sch01} Schulz, N. S., Canizares, C.,
1627: Huenemoerder, D., Kastner, J. H., Taylor, S. C., \& Bergstrom, E. J.,
1628: 2001, \apj, 549, 441
1629: \bibitem[Schulz et al. (2003)]{sch03} Schulz, N. S., Canizares, C.,
1630: Huenemoerder, D., Tibbets, K., 2003, \apj, 595, 365
1631: \bibitem[Schulz et al. (2006)]{sch06} Schulz, N. S., Testa, P.,
1632: Huenemoerder, D. P., Ishibashi, K., \& Canizares, C. R., 2006, \apj, 653,
1633: 636
1634: \bibitem[Scelsi et al. (2007)]{sce07} Scelsi, L., Maggio, A., Micela,
1635: G., Briggs, K., \& Gudel, M., preprint (astro-ph 0707.2857)
1636: \bibitem[Schmidt-Kaler (1982)]{sk82} Schmidt-Kaler, T., 1982,
1637: Landolt-B\"ornstein, Group VI, Vol 2b, eds. K. Schaifers and
1638: H. H. Voigt (New York: Springer Verlag), 14
1639: \bibitem[Seward et al. (1979)]{sew79} Seward, F. D., Forman, W. R.,
1640: Giacconi, R., Griffiths, R. E., Harnden, F. R., Jr., Jones, C., \& Pye,
1641: J. P., 1979, \apjl, 234, L55
1642: \bibitem[Seward \& Chlebowski (1982)]{sew82} Seward, F. D. \&
1643: Chlebowski, T., 1982, \apj, 256, 530
1644: \bibitem[SIMBAD (2006)]{sim06} SIMBAD Astronomical Database, Strausborg, France
1645: \bibitem[Smith et al.(2001)]{2001AJ.556..91L} Smith, R.~K.,
1646: Brickhouse, N.~S., Liedahl, D.~A., Raymond, J~ C. 2001, \apj, 556, L91
1647: \bibitem[Smith et al. (2004)]{smi04} Smith, M. A., Cohen, D. H.,
1648: Gu. M. F., Robinson, R. D., Evans, N. R., Schran, P. G., 2004, \apj,
1649: 600, 972
1650: \bibitem[Testa et al. (2004)]{tes04} Testa, P., Drake, J. J., \& Peres, G.,
1651: 2004, \apj, 617, 508
1652: \bibitem[ud-Doula \& Owocki (2002)]{udd02} ud-Doula, A. \& Owocki,
1653: S. P., 2002, \apj, 576, 413
1654: \bibitem[van der Hucht (2001)]{huc01} van der Hucht, K. A., 2001,
1655: \nar, 45, 135
1656: \bibitem[Walborn (2006)]{walb06} Walborn, N. R., {\it The UV Universe:
1657: Stars from Birth to Death}, 2006, $26^{th}$ Meeting of the IAU, Joint
1658: Discussion 4, Prague, Czech Republic, 1
1659: \bibitem[Waldron (1994)]{wal94} Waldron, W. L., 1994, AIP
1660: Conf. Proceed., 313, 279
1661: \bibitem[Weisskopf et al. (2002)]{wei02} Weisskopf, M. C., Brinkman,
1662: B., Canizares, C., Garmire, G., Murray, S., Van Speybroeck, L. P., 2002
1663: \pasp, 114, 1
1664: \bibitem[White et al. (1980)]{whi80} White, N. E., Holt, S. S., Boldt,
1665: E. A., \& Serlemitsos, P. J., 1980, \apjl, 239, 69
1666: \bibitem[Wolk et al. (2006)]{wol06} Wolk, S. J., Spitzbart, B. D.,
1667: Bourke, T. L., \& Alves, J., 2006, \apj, 132, 1100
1668: \bibitem[Young et al. (2001)]{you01} Young, P. R., Dupree, A. K.,
1669: Wood, B. E., Redfield, S., Linsky, J. L., Ake, T. B., \& Moos, H. W., 2001, \apj, 555, L121
1670: \end{thebibliography}
1671:
1672: \clearpage
1673:
1674: \begin{figure}
1675: \includegraphics[width=3.1in]{f1a.eps}
1676: \includegraphics[width=3.1in]{f1b.eps}
1677: \newline
1678: \includegraphics[width=3.1in]{f1c.eps}
1679: \includegraphics[width=3.1in]{f1d.eps}
1680: \caption{\scriptsize The X-Atlas web interface. The database can be
1681: searched on a variety of observation and target parameters ({\it top
1682: left}) and the basic astrophysical parameters, spectra, X-ray color
1683: data, and variability of sources in the observations can be previewed
1684: ({\it top right}). After selecting observations of interest,
1685: customized spectral plots of up to three sources of interest can be created ({\it bottom left}) and both the output plots and
1686: spectral data for those sources can be retrieved, including the high-resolution HEG and
1687: MEG spectra and the predicted low-resolution ACIS-S and ACIS-I spectra.
1688: \label{figxweb}}
1689: \end{figure}
1690:
1691: \clearpage
1692:
1693: \begin{figure}
1694: \includegraphics[width=6.2in]{f2.ps}
1695: \caption{\scriptsize Montage of the high-resolution MEG spectra ([counts/s/\AA]) of
1696: normal OB supergiants and giants, ordered by spectral class.
1697: The locations of some strong lines are marked.
1698: The spectra from different sources are offset for clarity. \label{figobsgh}}
1699: \end{figure}
1700:
1701: \begin{figure}
1702: \includegraphics[width=6.2in]{f3.ps}
1703: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1704: normal OB giants, sub-giants and dwarfs.\label{figobmsh}}
1705: \end{figure}
1706:
1707: \begin{figure}
1708: \includegraphics[width=6.2in]{f4.ps}
1709: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1710: four observations $\theta^{1}$ Ori C and two observations
1711: of $\tau$ Sco, two high-mass stars thought to possess magnetic fields
1712: responsible for generating X-rays. \label{figobmagh}}
1713: \end{figure}
1714:
1715: \begin{figure}
1716: \includegraphics[width=6.2in]{f5.ps}
1717: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1718: three peculiar OB stars, $\gamma$ Cas, Cyg OB2 8A, and $\theta^{2}$
1719: Ori A, approximately ordered by X-ray hardness, descending from
1720: hardest to softest. \label{figobpech}}
1721: \end{figure}
1722:
1723: \begin{figure}
1724: \includegraphics[width=6.2in]{f6.ps}
1725: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1726: two Wolf-Rayet stars observed thus far with HETG. \label{figwrh}}
1727: \end{figure}
1728: \begin{figure}
1729: \includegraphics[width=6.2in]{f7.ps}
1730: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1731: six HETG observations of $\eta$ Carinae. $\eta$ Carinae underwent an
1732: X-ray eclipse in mid-2003: the 2003-05-02 observation was made close to
1733: X-ray maximum, and the 2003-07-02 observation was taken near X-ray
1734: minimum \citep{cor05}. \label{figetah}}
1735: \end{figure}
1736:
1737: \clearpage
1738:
1739: \begin{figure}
1740: \includegraphics[width=6.2in]{f8a.ps}
1741: \end{figure}
1742: \begin{figure}
1743: \includegraphics[width=6.2in]{f8b.ps}
1744: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1745: normal and rapidly rotating low-mass stars in the
1746: atlas. \label{figcoolnormh}}
1747: \end{figure}
1748:
1749: \begin{figure}
1750: \includegraphics[width=6.2in]{f9.ps}
1751: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1752: two unique stellar systems, Algol ({\it top}) and FK Com ({\it
1753: bottom}). Algol is a three-star system with a semi-detached B8 V/K1 IV
1754: eclipsing binary in orbit with an A V star. The X-ray emission is
1755: believed to originate from the K1 star \citep{whi80}. FK Com is an
1756: active, variable G-type giant rotating near breakup velocity,
1757: possibly a recently coalesced binary \citep{hue93}.\label{figcoolpech}}
1758: \end{figure}
1759:
1760: \begin{figure}
1761: \includegraphics[width=6.2in]{f10.ps}
1762: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1763: two W UMa--type stars, 44 Boo and VW Cep. \label{figwumah}}
1764: \end{figure}
1765:
1766: \begin{figure}
1767: \includegraphics[width=6.2in]{f11a.ps}
1768: \end{figure}
1769: \begin{figure}
1770: \includegraphics[width=6.2in]{f11b.ps}
1771: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1772: RS CVn stars in the atlas. \label{figrscvnh}}
1773: \end{figure}
1774:
1775: \begin{figure}
1776: \includegraphics[width=6.2in]{f12a.ps}
1777: \end{figure}
1778: \begin{figure}
1779: \includegraphics[width=6.2in]{f12b.ps}
1780: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1781: T Tauri-type stars in the atlas. \label{figttsh}}
1782: \end{figure}
1783:
1784: \begin{figure}
1785: \includegraphics[width=6.2in]{f13.ps}
1786: \caption{\scriptsize As in Figure~\ref{figobsgh}, for
1787: active and flaring low-mass stars. \label{figflareh}}
1788: \end{figure}
1789:
1790: \clearpage
1791:
1792: \begin{figure}
1793: \includegraphics[width=6.2in]{f14.ps}
1794: \caption{\scriptsize Line profiles of selected lines in MEG spectra.
1795: Columns from left to right: Mg\,XII~$\lambda$8.421,
1796: Ne\,X~$\lambda$12.134, Fe\,XVII~$\lambda$15.014, Fe\,XVII~$\lambda$16.780,
1797: O\,VIII~$\lambda$18.969, and N\,VII~$\lambda$24.781.
1798: Descending from top row: $\theta^{1}$~Ori~C, $\zeta$~Oph, and $\tau$~Sco.
1799: \label{figlinprof}.}
1800: \end{figure}
1801: \begin{figure}
1802: \includegraphics[width=6.2in]{f15.ps}
1803: \caption{\scriptsize As in Figure~\ref{figlinprof}, for the sources,
1804: descending from top row: $\zeta$~Pup,
1805: $\zeta$~Ori, $\iota$~Ori. \label{figlprof2}}
1806: \end{figure}
1807:
1808: \begin{figure}
1809: \includegraphics[width=6.2in]{f16.ps}
1810: \caption{\scriptsize Predicted (solid line) and actual (dashed line)
1811: $0^{th}$-order spectra for $\xi$~Per (ObsID 4512). Bin sizes vary
1812: across the wavelength range, and are determined by the PHA binning
1813: in the RMF. For this observation, the predicted and observed spectra
1814: agreed to a $\chi^{2}$ value of 381.7, with 340 non-zero bins. A
1815: 10\% calibration error in the effective area was included in the
1816: observed spectrum for the $\chi^{2}$ calculation. \label{figzeroth}}
1817: \end{figure}
1818:
1819: \begin{figure}
1820: \includegraphics[width=6.2in]{f17.ps}
1821: \caption{\scriptsize Predicted (solid line) and actual (dashed line)
1822: $0^{th}$-order spectra for HD 93250, produced by combining ObsIDs 5399, 5400,
1823: 7189, 7341, and 7342. PHA binning is determined as in
1824: Figure~\ref{figzeroth}. For this observation, the predicted and
1825: observed spectra agreed to a $\chi^{2}$ value of 865.5, with 630
1826: non-zero bins. A 10\% calibration error in the effective area was
1827: included in the observed spectrum for the $\chi^{2}$
1828: calculation. \label{figcoadd}}
1829: \end{figure}
1830:
1831: \clearpage
1832:
1833: \begin{figure}
1834: \includegraphics[width=6.2in]{f18.ps}
1835: \caption{\scriptsize The predicted low-resolution ACIS-S spectra
1836: ([counts/s/keV]) of OB
1837: supergiants and giants, ordered by spectral class. The spectrum of each
1838: source is offset for clarity.
1839: The diminishing
1840: strength of S, Si, Mg, and Ne lines and the growth of O lines in later
1841: spectral classes is still evident in low-resolution. \label{figobsgl}}
1842: \end{figure}
1843:
1844: \begin{figure}
1845: \includegraphics[width=6.2in]{f19.ps}
1846: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1847: OB giants, sub-giants and dwarfs, ordered by spectral class. \label{figobmsl}}
1848: \end{figure}
1849:
1850: \begin{figure}
1851: \includegraphics[width=6.2in]{f20.ps}
1852: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1853: the two OB stars, $\theta^{1}$ Ori C and $\tau$ Sco, thought to have
1854: strong magnetic fields responsible for their X-ray
1855: emissions. \label{figobmagl}}
1856: \end{figure}
1857:
1858: \begin{figure}
1859: \includegraphics[width=6.2in]{f21.ps}
1860: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1861: the three peculiar OB stars: Cyg OB2 8A, $\theta^{2}$ Ori A, and $\gamma$
1862: Cas. \label{figobpecl}}
1863: \end{figure}
1864:
1865: \begin{figure}
1866: \includegraphics[width=6.2in]{f22.ps}
1867: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1868: the two Wolf-Rayet stars observed with HETG. \label{figwrl}}
1869: \end{figure}
1870:
1871: \begin{figure}
1872: \includegraphics[width=6.2in]{f23.ps}
1873: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1874: the six HETG observations of Eta Carinae. $\eta$ Carinae underwent an
1875: X-ray eclipse in mid-2003: the 2003-05-02 observation was made close to
1876: X-ray maximum, and the 2003-07-02 observation was taken near X-ray
1877: minimum \citep{cor05}. \label{figetal}}
1878: \end{figure}
1879:
1880: \clearpage
1881:
1882: \begin{figure}
1883: \includegraphics[width=6.2in]{f24a.ps}
1884: \end{figure}
1885:
1886: \begin{figure}
1887: \includegraphics[width=6.2in]{f24b.ps}
1888: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1889: normal low-mass stars. Capella is
1890: not included here but is featured in a separate montage available on
1891: the X-Atlas website. \label{figcoolnorml}}
1892: \end{figure}
1893:
1894: \begin{figure}
1895: \includegraphics[width=6.2in]{f25.ps}
1896: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1897: two unique systems, Algol and FK Com. \label{figcoolpecl}}
1898: \end{figure}
1899:
1900: \begin{figure}
1901: \includegraphics[width=6.2in]{f26.ps}
1902: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1903: W UMa--type variables in the atlas. \label{figwumal}}
1904: \end{figure}
1905:
1906: \begin{figure}
1907: \includegraphics[width=6.2in]{f27a.ps}
1908: \end{figure}
1909: \begin{figure}
1910: \includegraphics[width=6.2in]{f27b.ps}
1911: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1912: eight RS CVn stars. AR Lac and IM
1913: Peg are not included here but are featured in separate montages on the
1914: X-Atlas website. \label{figrscvnl}}
1915: \end{figure}
1916:
1917: \begin{figure}
1918: \includegraphics[width=6.2in]{f28a.ps}
1919: \end{figure}
1920: \begin{figure}
1921: \includegraphics[width=6.2in]{f28b.ps}
1922: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1923: nine T Tauri stars. \label{figttsl}}
1924: \end{figure}
1925:
1926: \begin{figure}
1927: \includegraphics[width=6.2in]{f29.ps}
1928: \caption{\scriptsize As in Figure~\ref{figobsgl}, for
1929: four low-mass flare stars. \label{figflarel}}
1930: \end{figure}
1931:
1932: \clearpage
1933:
1934: \begin{figure}
1935: \includegraphics[width=6.2in]{f30.ps}
1936: \caption{\scriptsize A comparison of predicted ACIS-I hardness ratios for each stellar
1937: observation in X-Atlas. $HR_{1}$ =($\frac{H-M}{H+M}$) and $HR_{2}$=
1938: ($\frac{M-S}{M+S}$) were calculated from the predicted ACIS-I aimpoint
1939: spectra (ACIS-S version, not shown here, is similar) of all stars in
1940: X-Atlas. H, M, and S are the predicted number of counts from the
1941: predicted ACIS-I spectra in the
1942: passbands designated on the plots. The softest X-ray emitters fall in
1943: the lower left, with hardness increasing towards the upper right.
1944: Errors were calculated using a Bayesian estimation technique (Park et al.\
1945: 2006) and are
1946: typically smaller than the symbols. $\times$ designates an observation
1947: of a hot star; $\Diamond$ designates a cool star; $\sq$ designates an
1948: observation of $\eta$ Car. The outliers include six HETG observations
1949: of $\eta$ Car and one, at approximately (0.42, 0.95), of WR 140, observed at high extinction. The lines on the plot connects that
1950: observation to another observation of WR 140 taken five months earlier
1951: at low extinction, indicating the direction of the reddening vector in
1952: the hardness ratio plot. Three observations of WR 140 made about five
1953: years later are also included, clustered around (0.75, 0.1).
1954: \label{fighr}}
1955: \end{figure}
1956:
1957: \begin{figure}
1958: \includegraphics[width=6.2in]{f31.eps}
1959: \caption{\scriptsize The predicted relation of $HR_{1}$ to $HR_{2}$
1960: for ACIS-I count spectra (Cycle 06) generated with a Mekal emissivity
1961: model using solar abundances. The left plot shows lines of constant
1962: $N_{H}$ in units of 10$^{21}$/cm$^{2}$ for temperatures ranging from
1963: Log $T$ = 6 to Log $T$ = 8 starting from the bottom left along each contour. The
1964: right plot displays lines of constant $T_{x}$ from 2.5 to 50 MK, with
1965: absorption increasing from 0.01 to 2.5$\times$ 10$^{22}$/cm$^{2}$ starting from
1966: the bottom left along each contour. While this model is adequate to
1967: explain the high-mass star distribution (indicated by $\times$), it
1968: fails to account for the low-mass stars (indicated by $\Diamond$). \label{fighrmodel1}}
1969:
1970: \end{figure}
1971: \begin{figure}
1972: \includegraphics[width=6.2in]{f32.eps}
1973: \caption{\scriptsize The predicted relation of $HR_{1}$ to $HR_{2}$
1974: for ACIS-I count spectra (Cycle 06) generated with a Mekal emissivity
1975: model using an abundance of 0.1 of solar for elements except H.
1976: Contour lines are shown as in Figure \ref{fighrmodel1}, with lines of
1977: constant absorption in the left plot and constant temperature in the
1978: right plot. Both the high-mass and the low-mass stars fall on the
1979: model contours when the abundance is lowered to 0.1 of solar. \label{fighrmodel01}}
1980: \end{figure}
1981:
1982: \clearpage
1983:
1984: \begin{figure}
1985: \includegraphics[width=6.2in]{f33.ps}
1986: \caption{\scriptsize Quantile-based color-color diagram (see \citet{hon04}) for the
1987: predicted ACIS-I aimpoint spectra of the observations of high-mass
1988: stars in X-Atlas. The quantile grids were generated assuming solar
1989: metallicity.
1990: $m$ = $Q_{50}$ = ($E_{50}$ - $E_{lo}$)/($E_{hi}$ - $E_{lo}$); $E_{lo}$ = 0.3
1991: keV; $E_{hi}$ = 8.0 keV). The upper scale on the plots is the median
1992: energy in keV. The six HETG observations of the unique
1993: object $\eta$ Car are not included in the plot. \label{fighquant}}
1994: \end{figure}
1995:
1996: \begin{figure}
1997: \includegraphics[width=6.2in]{f34a.ps}
1998: \includegraphics[width=6.2in]{f34b.ps}
1999: \caption{\scriptsize Quantile-based color-color diagrams for the
2000: predicted ACIS-I aimpoint spectra of the low-mass stars in X-Atlas.
2001: The quantile grids were calculated using metallicities of 0.3 times
2002: solar ({\it top}) and 0.1 times solar ({\it bottom}). The symbols on
2003: the plot correspond to stellar types as follows:
2004: F = normal F star; G = normal G star; K = normal K star; M = normal M
2005: star; R = RS CVn star; T = T Tauri star; W = W UMa star; A = Algol; *
2006: = FK Com. $m$ = $Q_{50}$ = ($E_{50}$ - $E_{lo}$)/($E_{hi}$ - $E_{lo}$); $E_{lo}$ = 0.3
2007: keV; $E_{hi}$ = 8.0 keV. The upper scale on the plots is the median
2008: energy in keV. \label{figcquant}}
2009: \end{figure}
2010:
2011: \clearpage
2012:
2013: \begin{figure}
2014: \includegraphics[width=3.1in]{f35a.ps}
2015: \includegraphics[width=3.1in]{f35b.ps}
2016: \newline
2017: \newline
2018: \includegraphics[width=3.1in]{f35c.ps}
2019: \includegraphics[width=3.1in]{f35d.ps}
2020: \newline
2021: \newline
2022: \includegraphics[width=3.1in]{f35e.ps}
2023: \includegraphics[width=3.1in]{f35f.ps}
2024: \newline
2025: \newline
2026: \includegraphics[width=6.2in]{f35g.ps}
2027: \end{figure}
2028: \begin{figure}
2029: \includegraphics[width=6.2in]{f35h.ps}
2030: \newline
2031: \newline
2032: \includegraphics[width=3.1in]{f35i.ps}
2033: \includegraphics[width=3.1in]{f35j.ps}
2034: \newline
2035: \newline
2036: \includegraphics[width=3.1in]{f35k.ps}
2037: \includegraphics[width=3.1in]{f35l.ps}
2038: \caption{\scriptsize Light curves obtained from the combined
2039: dispersed counts in the HEG and MEG arms
2040: of selected HETG observations of
2041: high-mass stars. The light curves were binned at time intervals of 1200 s {\it (dashed line)} and 4000 s {\it
2042: (solid line)}:
2043: Row 1-2: Four observations of $\theta^{1}$ Ori C. Row 3: Two
2044: observations of $\theta^{2}$ Ori A. Row 4: $\gamma$ Cas. Row 5: HD 150136. Row 6: Two observations of 9 Sgr. Row 7:
2045: Two observations of $\zeta$ Oph.\label{figlchm}}
2046: \end{figure}
2047:
2048: \begin{figure}
2049: \includegraphics[width=6.2in]{f36a.ps}
2050: \newline
2051: \newline
2052: \includegraphics[width=6.2in]{f36b.ps}
2053: \newline
2054: \newline
2055: \includegraphics[width=6.2in]{f36c.ps}
2056: \newline
2057: \newline
2058: \includegraphics[width=6.2in]{f36d.ps}
2059: \newline
2060: \newline
2061: \end{figure}
2062: \begin{figure}
2063: \includegraphics[width=3.1in]{f36e.ps}
2064: \includegraphics[width=3.1in]{f36f.ps}
2065: \newline
2066: \newline
2067: \includegraphics[width=6.2in]{f36g.ps}
2068: \newline
2069: \newline
2070: \includegraphics[width=2.05in]{f36h.ps}
2071: \includegraphics[width=2.05in]{f36i.ps}
2072: \includegraphics[width=2.05in]{f36j.ps}
2073: \caption{\scriptsize As in Figure~\ref{figlchm}, for
2074: selected low-mass stars.
2075: Row 1: Algol. Row 2: TW Hya. Row 3: DoAr 21. Row 4: Proxima
2076: Centauri. Row 5: Two observations of CC Eri. Row 6: Hen 3-600. Row
2077: 7: Three observations of HD 245059. \label{figlclm}}
2078: \end{figure}
2079: \begin{figure}
2080: \includegraphics[width=2.05in]{f37a.ps}
2081: \includegraphics[width=2.05in]{f37b.ps}
2082: \includegraphics[width=2.05in]{f37c.ps}
2083: \newline
2084: \newline
2085: \includegraphics[width=2.05in]{f37d.ps}
2086: \includegraphics[width=2.05in]{f37e.ps}
2087: \includegraphics[width=2.05in]{f37f.ps}
2088: \newline
2089: \newline
2090: \includegraphics[width=2.05in]{f37g.ps}
2091: \includegraphics[width=2.05in]{f37h.ps}
2092: \includegraphics[width=2.05in]{f37i.ps}
2093: \newline
2094: \newline
2095: \includegraphics[width=2.05in]{f37j.ps}
2096: \includegraphics[width=2.05in]{f37k.ps}
2097: \includegraphics[width=2.05in]{f37l.ps}
2098: \newline
2099: \newline
2100: \includegraphics[width=2.05in]{f37m.ps}
2101: \includegraphics[width=2.05in]{f37n.ps}
2102: \includegraphics[width=2.05in]{f37o.ps}
2103: \caption{\scriptsize As in Figure~\ref{figlchm}, for
2104: a sequence of twelve light curves of Capella.
2105: \label{figlccap}}
2106: \end{figure}
2107:
2108: \clearpage
2109:
2110: \begin{deluxetable}{lll}
2111: \tabletypesize{\scriptsize}
2112: \tablewidth{0pt}
2113: \tablecaption{Spectral Line Metrics\label{tblmetrics}}
2114: \tablehead{\colhead{Metric} & \colhead{Flux ratio} & \colhead{Comment} \tabletypesize{5pt} \tablecolumns{3} }
2115: \startdata
2116: \tableline
2117: NeO & Ne\,X~$\lambda$12.13 / O\,VIII~$\lambda$18.97 & tracks Ne abundance \\
2118: Oif & O\,VII(i)~$\lambda$21.8 / O\,VII(f)~$\lambda$22.1 & O\,VII is density sensitive \\
2119: O78 & O\,VII(r)~$\lambda$21.6 / O\,VIII~$\lambda$18.97 & tracks temperature \\
2120: OFe & O\,VIII~$\lambda$18.97 / Fe\,XVII~$\lambda\lambda$15.01,17.1 & tracks high FIP element abundances \\
2121: MgFe & Mg\,XII$\lambda$8.419 / Fe\,XVII~$\lambda\lambda$15.01,17.1 & high-FIP elements, possibly tracks fractionation \\
2122: SiFe & Si\,XIV$\lambda$6.18 / Fe\,XVII~$\lambda\lambda$15.01,17.1 & high-FIP elements, possibly tracks fractionation \\
2123: SFe & S\,XV~$\lambda$5.1 / Fe\,XVII~$\lambda\lambda$15.01,17.1 & S is a high FIP element \\
2124: short line/cont & $f_{\rm line}(\lambda<12.13)$ / $f_{\rm continuum}$ & sensitive to high temperatures and metallicities \\
2125: long line/cont & $f_{\rm line}(\lambda>12.13)$ / $f_{\rm continuum}$ & sensitive to low temperatures \\
2126: line/cont & $f_{\rm line}$ / $f_{\rm continuum}$ & tracks metallicity \\
2127: cont long/short & $f_{\rm continuum}(13<\lambda<25)$ / $f_{\rm continuum}(3<\lambda<12)$ & categorizes continuum shape \\
2128: cont long/medium & $f_{\rm continuum}(13<\lambda<25)$ / $f_{\rm continuum}(6<\lambda<12)$ & categorizes continuum shape \\
2129: cont medium/short & $f_{\rm continuum}(6<\lambda<12)$ / $f_{\rm continuum}(3<\lambda<6)$ & categorizes continuum shape \\
2130: \enddata
2131: \end{deluxetable}
2132:
2133: \clearpage
2134:
2135: \begin{deluxetable}{lllll}
2136: \tablewidth{0pt}
2137: \tablecaption{Coadded observations in X-Atlas. 'Num. Obs.' indicates
2138: the number of observations that were combined. The $\chi^{2}$ values
2139: listed are derived from the comparison of the observed combined $0^{th}$-order
2140: spectrum and the predicted combined $0^{th}$-order, with an extra 10\%
2141: systematic error added to the observed spectrum to account for
2142: uncertainties in the effective area. 'Num. Bins' is the
2143: number of non-zero bins in the observed $0^{th}$-order spectrum used
2144: to calculate the $chi^{2}$ value. \label{tblcoadd}}
2145: \tablehead{
2146: \colhead{Star} & \colhead{Num. Obs.} &
2147: \colhead{Observation IDs} & \colhead{$\chi^{2}$} & \colhead{Num. Bins}
2148: \tabletypesize{\scriptsize}
2149: \tablecolumns{3}
2150: }
2151: \startdata
2152: \tableline
2153: HD 93129AB & 7 & 5397, 7201, 7202, 7203, 7204, 7228, 7229 & 471.4 & 298\\
2154: HD 93250 & 5 & 5399, 5400, 7189, 7341, 7342 & 865.5 & 630 \\
2155: 9 Sgr & 2 & 5398, 6285 & 401.8 & 461 \\
2156: 15 Mon & 3 & 5401, 6247, 6248 & 445.8 & 236 \\
2157: \tableline
2158: \enddata
2159: \end{deluxetable}
2160:
2161: \clearpage
2162:
2163: \begin{deluxetable}{lllllllll}
2164: \tabletypesize{\scriptsize}
2165: \tablewidth{0pt}
2166: \tablecaption{Stars in the X-Atlas database, ordered approximately by
2167: spectral type. The 'Star' column lists
2168: the common name (if applicable) of the primary target of the observation. 'Alt. Name'
2169: gives a catalog listing for that star, usually the HD number. 'ObsID'
2170: is the $Chandra$ Observation ID. 'Obs. Date' is the date the
2171: observation was made. '$t_{exp}$' is the exposure time of the
2172: observation, in ks. 'Spec. Type' is the spectral type of the
2173: star. 'B$-$V' is the B$-$V color of the star. '$d$' is the
2174: distance in parsecs from the star to Earth. 'Figure (h/l)' indicates the
2175: spectral montage figures in which the spectra from the
2176: observation can be found: the first number refers to the figure with
2177: the high-resolution MEG spectrum, and the second number refers to the
2178: figure with the projected low-resolution ACIS-S spectrum. Spectral types, B$-$V
2179: values, and distances were retrieved from the {\it SIMBAD} database,
2180: operated in Strasbourg, France, unless otherwise noted. \label{tbl1}}
2181: \tablehead{
2182: \colhead{Star} & \colhead{Alt. Name} &
2183: \colhead{ObsID} & \colhead{Obs. Date} & \colhead{$t_{exp}$ (ks)} & \colhead{Spec. Type} &
2184: \colhead{$\bv$} & \colhead{$d$ (pc)} & \colhead{Figure (h/l)}
2185: \tablecolumns{9}
2186: }
2187: \startdata
2188: \tableline
2189: HD 93129 AB\tablenotemark{a} & ... & 5397 & 2005-11-29 & 10.2 & (2x)O2If*+O3.5V((f+))\tablenotemark{e} & 0.206\tablenotemark{e} & 2500\tablenotemark{n} & \ref{figobsgh}/\ref{figobsgl} \\
2190: & & 7201 & 2005-11-08 & 18.2 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2191: & & 7202 & 2005-11-10 & 9.5 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2192: & & 7203 & 2005-11-12 & 27.3 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2193: & & 7204 & 2005-11-09 & 33.2 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2194: & & 7228 & 2005-12-01 & 19.7 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2195: & & 7229 & 2005-12-02 & 19.6 & & & & \ref{figobsgh}/\ref{figobsgl} \\
2196: HD 93250\tablenotemark{b} & ... & 5399 & 2005-10-22 & 39.7 & O3.5V((f+))\tablenotemark{f} & 0.165\tablenotemark{e} & 2500\tablenotemark{n} & \ref{figobmsh}/\ref{figobmsl} \\
2197: & & 5400 & 2005-09-05 & 33.7 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2198: & & 7189 & 2006-06-23 & 17.7 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2199: & & 7341 & 2006-06-19 & 53.3 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2200: & & 7342 & 2006-06-22 & 49.3 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2201: 9 Sgr & HD 164794 & 5398 & 2005-05-11 & 101.2 & O4V((f))\tablenotemark{f} & 0.017\tablenotemark{e} & 1500 & \ref{figobmsh}/\ref{figobmsl} \\
2202: & & 6285 & 2005-06-09 & 44.6 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2203: $\zeta$ Pup & HD 66811 & 640 & 2000-03-28 & 67.6 & O4I(n)f\tablenotemark{f} & -0.276 & 710\tablenotemark{e} & \ref{figobsgh}/\ref{figobsgl} \\
2204: HD 150136 & ... & 2569 & 2002-06-27 & 90.3 & O5III:n(f)\tablenotemark{e} & 0.161\tablenotemark{e} & \nodata & \ref{figobsgh}/\ref{figobsgl} \\
2205: $\theta^{1}$\ Ori\ C & HD 37022 & 3 & 1999-10-31 & 49.5 &
2206: O5.5V\tablenotemark{g} & 0.021\tablenotemark{e} & \nodata & \ref{figobmagh}/\ref{figobmagl} \\
2207: & & 4 & 1999-11-24 & 30.9 & & & & \ref{figobmagh}/\ref{figobmagl} \\
2208: & & 2567 & 2001-12-28 & 46.4 & & & & \ref{figobmagh}/\ref{figobmagl} \\
2209: & & 2568 & 2002-02-19 & 46.3 & & & & \ref{figobmagh}/\ref{figobmagl} \\
2210: Cyg OB2 8A & BD +40 4227 & 2572 & 2002-07-31 & 65.1 & O6I(f)+O5.5III(f)\tablenotemark{h} & 1.09 & \nodata & \ref{figobpech}/\ref{figobpecl} \\
2211: HD 206267 & ... & 1888 & 2001-02-18 & 34.1 & O6.5V((f))\tablenotemark{e} & 0.205\tablenotemark{e} & 591\tablenotemark{e} & \\
2212: & & 1889 & 2001-02-09 & 39.5 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2213: 15 Mon & HD 47839 & 5401 & 2005-11-20 & 54.8 & O7V((f))\tablenotemark{e}+O9.5Vn\tablenotemark{i} & -0.248 & 447\tablenotemark{e} & \ref{figobmsh}/\ref{figobmsl} \\
2214: & & 6247 & 2006-01-31 & 37.8 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2215: & & 6248 & 2005-12-13 & 7.6 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2216: $\xi$ Per & HD 24912 & 4512 & 2004-03-22 & 158.8 & O7.5III(n)((f))\tablenotemark{f} & 0.014\tablenotemark{e} & 2130\tablenotemark{e} & \ref{figobsgh}/\ref{figobsgl} \\
2217: $\iota$ Ori & HD 37043 & 599 & 2001-02-07 & 37.2 & O9III+?\tablenotemark{f} & -0.244\tablenotemark{e} & 1110\tablenotemark{e} & \ref{figobmsh}/\ref{figobmsl} \\
2218: & & 2420 & 2001-02-08 & 12.8 & & & & \\
2219: $\tau$ CMa & HD 57061 & 2525 & 2002-03-28 & 43.2 & O9II\tablenotemark{e} & -0.155\tablenotemark{e} & 980. & \\
2220: & & 2526 & 2002-04-24 & 43.9 & & & & \\
2221: $\zeta$ Oph & HD 149757 & 2571 & 2002-09-04 & 35.4 & O9.5Vnn\tablenotemark{e} & 0.019 & 146\tablenotemark{e} & \\
2222: & & 4367 & 2002-09-05 & 48.3 & & & & \ref{figobmsh}/\ref{figobmsl} \\
2223: $\theta^{2}$\ Ori\ A & \nodata & 4473 & 2004-11-03 & 49.1 & O9.5Vpe\tablenotemark{e} & -0.084\tablenotemark{e} & 581. & \ref{figobpech}/\ref{figobpecl} \\
2224: & & 4474 & 2004-11-23 & 50.6 & & & & \ref{figobpech}/\ref{figobpecl} \\
2225: $\sigma$ Ori & HD 37468 & 3738 & 2003-08-12 & 91.0 & O9.5V+B0.5V\tablenotemark{f} & -0.240\tablenotemark{e} & 381\tablenotemark{e} & \ref{figobmsh}/\ref{figobmsl} \\
2226: $\delta$ Ori & HD 36486 & 639 & 2000-01-13 & 49.0 & O9.5II+B0.5III\tablenotemark{f} & -0.219\tablenotemark{e} & 310\tablenotemark{e} & \ref{figobsgh}/\ref{figobsgl} \\
2227: $\zeta$ Ori & HD 37742J & 610 & 2000-04-08 & 59.7 &
2228: O9.7Ib\tablenotemark{f} & -0.208\tablenotemark{e} &
2229: 278\tablenotemark{e} & \ref{figobsgh}/\ref{figobsgl} \\
2230: & & 1524 & 2000-04-09 & 13.8 & & & & \\
2231: $\gamma$ Cas & HD 5394 & 1895 & 2001-08-10 & 51.2 & B0IVpe & -0.10 & 188 & \ref{figobpech}/\ref{figobpecl} \\
2232: $\epsilon$ Ori & HD 37128 & 3753 & 2003-12-12 & 91.6 & B0Ia & -0.19 & 412 & \ref{figobsgh}/\ref{figobsgl} \\
2233: $\tau$ Sco & HD 149438 & 638 & 2000-09-17 & 59.2 & B0.2V & -0.201 & 132 & \ref{figobmagh}/\ref{figobmagl} \\
2234: & & 2305 & 2000-09-18 & 13.0 & & & & \ref{figobmagh}/\ref{figobmagl} \\
2235: $\beta$ Cru & HD 111123 & 2575 & 2002-02-01 & 74.4 & B0.5III\tablenotemark{f} & -0.1520 & 108 & \ref{figobmsh}/\ref{figobmsl} \\
2236: WR 140 & HD 193793 & 2337 & 2000-12-29 & 45.5 & WC7pd+O4-5\tablenotemark{j} & 0.27\tablenotemark{j} & 1100\tablenotemark{j} & \ref{figwrh}/\ref{figwrl} \\
2237: & & 2338 & 2001-05-08 & 24.5 & & & & \ref{figwrh}/\ref{figwrl} \\
2238: & & 5419 & 2006-03-31 & 45.1 & & & & \ref{figwrh}/\ref{figwrl} \\
2239: & & 6286 & 2006-03-23 & 46.9 & & & & \ref{figwrh}/\ref{figwrl} \\
2240: & & 6287 & 2006-04-07 & 49.3 & & & & \ref{figwrh}/\ref{figwrl} \\
2241: $\gamma$ Vel & HD 68273 & 629 & 2000-03-15 & 64.8 & WC8+O7.5III-V\tablenotemark{j} & -0.32\tablenotemark{j} & 368\tablenotemark{k} & \ref{figwrh}/\ref{figwrl} \\
2242: $\eta$ Car & HD 93308 & 632 & 2000-11-19 & 89.5 & ?p & 0.61 & 2500\tablenotemark{n} & \ref{figetah}/\ref{figetal} \\
2243: & & 3745 & 2003-05-02 & 94.5 & & & & \ref{figetah}/\ref{figetal} \\
2244: & & 3746 & 2003-07-20 & 90.3 & & & & \ref{figetah}/\ref{figetal} \\
2245: & & 3747 & 2003-09-26 & 70.1 & & & & \ref{figetah}/\ref{figetal} \\
2246: & & 3748 & 2003-06-16 & 97.2 & & & & \ref{figetah}/\ref{figetal} \\
2247: & & 3749 & 2002-10-16 & 91.3 & & & & \ref{figetah}/\ref{figetal} \\
2248: Algol & HD 19356 & 604 & 2000-04-01 & 51.7 & B8V+K1IV\tablenotemark{l} & -0.05 & 28.5 & \ref{figcoolpech}/\ref{figcoolpecl} \\
2249: Canopus & HD 45348 & 636 & 2000-07-21 & 94.6 & F0II & 0.15 & 95.9 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2250: $\upsilon$ Peg & HD 220657 & 3731 & 2003-08-26 & 106.8 & F8IV & 0.61 & 53.1 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2251: TZ CrB & HD 146361 & 15 & 2000-06-18 & 83.7 & G0V+G0V\tablenotemark{l} & 0.51 & 21.7 & \ref{figrscvnh}/\ref{figrscvnl} \\
2252: 44 Boo & HD 133640 & 14 & 2000-04-25 & 59.1 & G1V+G2V\tablenotemark{l} & 0.65 & 12.8 & \ref{figwumah}/\ref{figwumal} \\
2253: ER Vul & HD 200391 & 1887 & 2001-03-29 & 112.0 & G0V+G5V\tablenotemark{l} & 0.57 & 49.9 & \ref{figrscvnh}/\ref{figrscvnl} \\
2254: $\xi$ UMa & HD 98230J & 1894 & 2000-12-28 & 70.9 & G0V+G0V\tablenotemark{l} & 0.59 & 7.7\tablenotemark{l} & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2255: LS Com & HD 111812 & 1891 & 2001-03-12 & 130.2 & G0IIIp & 0.627 & 94.2 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2256: OU And & HD 223460 & 1892 & 2001-08-11 & 95.7 & G1IIIe & 0.79 & 135 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2257: AR Lac & HD 210334 & 6 & 2000-09-11 & 32.1 &
2258: G2IV+K0IV\tablenotemark{l} & 0.72 & 42.0 & \\
2259: & & 7 & 2000-09-16 & 7.4 & & & & \\
2260: & & 8 & 2000-09-15 & 7.5 & & & & \\
2261: & & 9 & 2000-09-17 & 32.2 & & & & \\
2262: & & 10 & 2000-09-20 & 7.2 & & & & \\
2263: & & 11 & 2000-09-19 & 7.2 & & & & \\
2264: V987 Tau & HD 283572 & 3756 & 2003-10-20 & 100.5 & G2III & 0.77 & 128 & \ref{figttsh}/\ref{figttsl} \\
2265: SU Aur & HD 282624 & 3755 & 2003-11-27 & 98.5 & G2IIIevar & 0.80 & 152 & \ref{figttsh}/\ref{figttsl} \\
2266: 24 Ursae Majoris & HD 82210 & 2564 & 2002-03-26 & 44.4 & G4III-IV & 0.753 & 32.4 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2267: & & 3471 & 2002-03-29 & 44.5 & & & & \\
2268: TY Pyx & HD 77137 & 601 & 2001-01-03 & 49.1 & G5IV+G5IV & 0.664 & 56 & \ref{figrscvnh}/\ref{figrscvnl} \\
2269: UX Ari & HD 21242 & 605 & 2000-01-12 & 48.5 & G5IV & 0.9 & 50.2 & \ref{figrscvnh}/\ref{figrscvnl} \\
2270: Capella & HD 34029 & 57 & 2000-03-03 & 28.8 & G5IIIe+... & 0.8 & 13.2 & \\
2271: & & 1010 & 2001-02-11 & 29.5 & & & & \\
2272: & & 1099 & 1999-08-28 & 14.6 & & & & \\
2273: & & 1100 & 1999-08-28 & 14.6 & & & & \\
2274: & & 1101 & 1999-08-29 & 14.6 & & & & \\
2275: & & 1103 & 1999-09-24 & 40.5 & & & & \\
2276: & & 1199 & 1999-08-30 & 2.0 & & & & \\
2277: & & 1235 & 1999-08-28 & 14.6 & & & & \\
2278: & & 1236 & 1999-08-28 & 14.6 & & & & \\
2279: & & 1237 & 1999-08-29 & 14.6 & & & & \\
2280: & & 1318 & 1999-09-25 & 26.7 & & & & \\
2281: & & 2583 & 2002-04-29 & 27.6 & & & & \\
2282: & & 3674 & 2003-09-27 & 28.7 & & & & \\
2283: & & 5040 & 2004-09-10 & 28.7 & & & & \\
2284: & & 5955 & 2005-03-28 & 28.7 & & & & \\
2285: & & 6471 & 2006-04-18 & 29.6 & & & & \\
2286: $\mu$ Vel & HD 93497 & 1890 & 2001-09-24 & 19.7 & G5IIIa & 0.914 & 35.5 & \\
2287: & & 3403\tablenotemark{c} & 2001-10-29 & 57.4 & & & & \\
2288: & & 3410 & 2001-12-18 & 57.0 & & & & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2289: FK Com & HD 117555 & 614 & 2000-03-27 & 41.4 & G5II & 0.87 & 234 & \ref{figcoolpech}/\ref{figcoolpecl} \\
2290: $\lambda$ And & HD 222107 & 609 & 1999-12-17 & 81.9 & G8III & 1.08 & 25.8 & \ref{figrscvnh}/\ref{figrscvnl} \\
2291: V2246 Oph & DoAr 21 & 3761 & 2003-05-05 & 91.1 & K0\tablenotemark{m} & 0.33 & \nodata & \ref{figttsh}/\ref{figttsl} \\
2292: II Peg & HD 224085 & 12\tablenotemark{d} & 1999-10-17 & 0.073 & K0V... & 1.01 & 42.3 & \\
2293: & & 1451 & 1999-10-17 & 42.7 & & & & \ref{figrscvnh}/\ref{figrscvnl} \\
2294: Speedy Mic & HD 197890 & 2536 & 2002-04-26 & 34.4 & K0V & 0.94 & 44.4 & \ref{figttsh}/\ref{figttsl} \\
2295: & & 3491 & 2002-04-27 & 34.6 & & & & \\
2296: HIP 92680 & HD 174429 & 3729 & 2003-06-07 & 73.9 & K0Vp & 0.77 & 49.7 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2297: VW Cep & HD 197433 & 3766 & 2003-08-29 & 116.6 & K0Vvar & 0.75 & 27.7 & \ref{figwumah}/\ref{figwumal} \\
2298: $\beta$ Ceti & HD 4128 & 974 & 2001-06-29 & 86.1 & K0III & 1.02 & 29.4 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2299: $\lambda$ Ori X-1 & HD 245059 & 5420 & 2006-01-13 & 54.8 & K1 & 0.7 & \nodata & \ref{figttsh}/\ref{figttsl} \\
2300: & & 6241 & 2005-12-30 & 17.7 & & & & \\
2301: & & 7253 & 2006-01-07 & 20.9 & & & & \\
2302: V824 Ara & HD 155555 & 2538 & 2002-07-10 & 94.2 & K1Vp & 0.78 & 31.4 & \ref{figrscvnh}/\ref{figrscvnl} \\
2303: AB Dor & HD 36705 & 16 & 1999-10-09 & 52.3 & K1IIIp... & 0.8 & 14.9 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2304: YY Men & HD 32918 & 2557 & 2002-02-01 & 74.2 & K1IIIp & 1.01 & 292 & \ref{figcoolnormh}/\ref{figcoolnorml} \\
2305: $\sigma$ Gem & HD 62044 & 5422 & 2005-05-16 & 62.8 & K1III & 1.12 & 37.5 & \ref{figrscvnh}/\ref{figrscvnl} \\
2306: & & 6282 & 2005-05-17 & 57.9 & & & & \\
2307: IM Peg & HD 216489 & 2527 & 2002-07-01 & 24.6 & K1.5II-IIIe & 1.134 & 96.8 & \\
2308: & & 2528 & 2002-07-08 & 24.8 & & & & \\
2309: & & 2529 & 2002-07-13 & 24.8 & & & & \\
2310: & & 2530 & 2002-07-18 & 23.9 & & & & \\
2311: & & 2531 & 2002-07-25 & 23.9 & & & & \\
2312: & & 2532 & 2002-07-31 & 22.5 & & & & \\
2313: & & 2533 & 2002-08-08 & 23.7 & & & & \\
2314: & & 2534 & 2002-08-15 & 23.9 & & & & \\
2315: TV Crt & HD 98800 & 3728 & 2003-03-07 & 98.9 & K5Ve & 1.16 & 46.7 & \ref{figttsh}/\ref{figttsl} \\
2316: V4046 Sgr & HD 319139 & 5423 & 2005-08-06 & 97.7 & K5 & 0.9 & \nodata & \ref{figttsh}/\ref{figttsl} \\
2317: & & 6265 & 2005-08-09 & 45.9 & & & & \\
2318: TW Hya & HIP 53911 & 5 & 2000-07-18 & 47.7 & K8Ve & 0.7 & 56.4 & \ref{figttsh}/\ref{figttsl} \\
2319: CC Eri & HD 16157 & 4513 & 2004-10-01 & 89.5 & M0 & 1.36 & 11.5 & \ref{figflareh}/\ref{figflarel} \\
2320: & & 6132 & 2004-10-01 & 30.6 & & & & \\
2321: AU Mic & HD 197481 & 17 & 2000-11-12 & 58.8 & M1Ve & 1.44 & 9.94 & \ref{figflareh}/\ref{figflarel} \\
2322: Hen 3-600 & GSC 07726 & 4502 & 2004-03-16 & 99.4 & M3 & 1.52 & \nodata & \ref{figttsh}/\ref{figttsl} \\
2323: AD Leo & SAO 81292 & 2570 & 2002-06-01 & 45.2 & M3.5V & 1.54 & 4.70 & \\
2324: EV Lac & HIP 112460 & 1885 & 2001-09-19 & 100.0 & M3.5e & 1.36 & 5.05 & \ref{figflareh}/\ref{figflarel} \\
2325: Proxima Centauri & V645 Cen & 2388 & 2001-09-13 & 42.4 & M5.5Ve & 1.97 & 1.30 & \ref{figflareh}/\ref{figflarel} \\
2326: \tableline
2327: \enddata
2328: \tablenotetext{a}{HD 93129AB was observed with HD 93129A at the
2329: aimpoint. $Chandra$ has sufficient spatial resolution to
2330: differentiate HD 93129A from HD 93129B. However, components A and B
2331: are about 3 arcsec apart along the dispersion direction, and the ACIS
2332: CCD energy resolution is not high enough to separate the events from
2333: these two sources. A third X-ray source, Trumpler 14 MJ 176, also lies in the dispersion
2334: direction $\sim$5 arcsec away from HD 93129A. The extracted spectrum
2335: includes events from all three sources.}
2336: \tablenotetext{b}{$\eta$ Car lies directly on the spectral arm in all
2337: spectra of HD 93250. Some events from $\eta$ Car may be included in
2338: the extracted spectra, and the background counts level will be
2339: anomalously high in this region.}
2340: \tablenotetext{c}{ObsID 3403, a $\sim$60 ks observation of HD 93497,
2341: is not included in X-Atlas. During the observation, the ACIS CCD was
2342: suffering from a problem known as threshold-plane (T-plane) lockup in
2343: chips S4 and S2, causing the science run to be stopped prematurely and
2344: replanned as ObsID 3410, which was carried out successfully.}
2345: \tablenotetext{d}{ObsID 12, a $\sim$73 s observation of II Peg, is not
2346: included in X-Atlas. Like ObsID 3403, ObsID 12 also suffered ACIS
2347: T-plane lockup, rendering the data useless. The observation was
2348: repeated successfully as ObsID 1451.}
2349: \tablenotetext{e}{\citet{mai04}}
2350: \tablenotetext{f}{\citet{walb06}}
2351: \tablenotetext{g}{\citet{sch06}}
2352: \tablenotetext{h}{\citet{deb04}}
2353: \tablenotetext{i}{\citet{gie93}}
2354: \tablenotetext{j}{\citet{huc01}}
2355: \tablenotetext{k}{\citet{mil07}}
2356: \tablenotetext{l}{\citet{tes04}}
2357: \tablenotetext{m}{\citet{pre99}}
2358: \tablenotetext{n}{\citet{eva03}}
2359: \end{deluxetable}
2360:
2361: \clearpage
2362:
2363: \begin{deluxetable}{lclcc}
2364: \tablecaption{$\chi^{2}$ values for the comparison of the observed and
2365: predicted $0^{th}$-order spectra for selected X-Atlas targets with low
2366: count rates. An extra 10\% systematic error was added to the observed
2367: $0^{th}$-order spectra to account for uncertainties in the effective area.
2368: The rightmost column lists the number of non-zero bins in the observed
2369: $0^{th}$-order spectrum. \label{tblpa_zero}}
2370: \tablewidth{0pt}
2371: \tablehead{
2372: \colhead{Target} & \colhead{Observation ID} &
2373: \colhead{Count Rate} & \colhead{$\chi^{2}$} & \colhead{Num. Bins}
2374: \tabletypesize{5pt}
2375: \tablecolumns{5}
2376: }
2377: \startdata
2378: \tableline
2379: HD 93250 & 7189 & 0.0337 & 174.8 & 257 \\
2380: $\xi$ Per & 4512 & 0.0418 & 381.7 & 340 \\
2381: HD 37468 & 3738 & 0.348 & 323.1 & 153 \\
2382: 15 Mon & 5401 & 0.0260 & 280.7 & 173 \\
2383: HD 206267 & 1888 & 0.0252 & 203.8 & 152 \\
2384: \tableline
2385: \enddata
2386: \end{deluxetable}
2387:
2388: \clearpage
2389:
2390: \begin{deluxetable}{llllllll}
2391: \tabletypesize{\small}
2392: \tablewidth{0pt}
2393: \tablecaption{Selected spectral fit results from the predicted ACIS-S aimpoint
2394: spectra. Fits for $\theta^{1}$ Ori C, $\theta^{2}$ Ori A, and
2395: $\gamma$ Cas were calculated using a known, fixed $N_{H}$, while for
2396: IM Peg and Capella, $N_{H}$ fit values were allowed to float. All
2397: stars were fit with a two-temperature Mewe-Kaastra-Liedahl thermal
2398: plasma model. Fits that have only one temperature listed had a flux
2399: emission ratio (FER) of 0, meaning that all of the flux was calculated to
2400: have originated from the $kT_{1}$ component of the plasma. All
2401: abundances are with respect to solar. $\chi^{2}$ in the final column is
2402: reduced $\chi^{2}$.\label{tblfit}}
2403: \tablehead{
2404: \colhead{Star} & \colhead{ObsID} &
2405: \colhead{$N_{H}\times 10^{22}/cm^{2}$} &
2406: \colhead{$kT_{1}$ (keV)} & \colhead{$kT_{2}$ (keV)} &
2407: \colhead{Abun.} &
2408: \colhead{F.E.R. ($kT_{2}$/$kT_{1}$)} &
2409: \colhead{$\chi^{2}$}
2410: \tablecolumns{8}
2411: }
2412: \startdata
2413: \tableline
2414: $\theta^{1}$ Ori C & 3 & 0.207 & $>$15 & 1.140 & 0.12 & 0.89 & 1.67\tablenotemark{a} \\
2415: & 4 & 0.207 & 10.329 & 1.286 & 0.14 & 0.97 & 1.06\tablenotemark{a} \\
2416: & 2567 & 0.207 & 9.928 & 1.426 & 0.23 & 1.61 & 1.26\tablenotemark{a} \\
2417: & 2568 & 0.207 & 11.852 & 1.140 & 0.07 & 0.78 & 1.34\tablenotemark{a} \\
2418: $\theta^{2}$ Ori A & 4473 & 0.148 & 0.794 & 0.335 & 0.52 & 1.21 & 1.29\tablenotemark{b} \\
2419: & 4474 & 0.148 & 3.533 & 1.142 & 0.05 & 0.05 & 1.28\tablenotemark{b} \\
2420: $\gamma$ Cas & 1895 & 0.136 & $>$15 & ... & 1.25 & 0 & 1.19\tablenotemark{a} \\
2421: Capella & 57 & 0.014 & 0.602 & ... & 0.31 & 0 & 1.80\tablenotemark{b} \\
2422: & 1010 & 0.057 & 0.568 & ... & 0.31 & 0 & 1.71\tablenotemark{b} \\
2423: & 1099 & 0.022 & 0.568 & ... & 0.30 & 0 & 1.08\tablenotemark{b} \\
2424: & 1100 & 0.034 & 0.561 & ... & 0.30 & 0 & 1.14\tablenotemark{b} \\
2425: & 1101 & 0.01 & 0.597 & ... & 0.27 & 0 & 1.75\tablenotemark{b} \\
2426: & 1103 & 0.031 & 4.098 & 0.561 & 0.32 & 9.79 & 1.42\tablenotemark{b} \\
2427: & 1235 & 0.047 & $>$10 & 0.565 & 0.35 & 20.85 & 1.33\tablenotemark{b} \\
2428: & 1236 & 0.033 & 0.574 & ... & 0.33 & 0 & 1.89\tablenotemark{b} \\
2429: & 1237 & 0.01 & 0.606 & ... & 0.53 & 0 & 1.14\tablenotemark{b} \\
2430: & 1318 & 0.038 & 0.561 & ... & 0.51 & 0 & 1.49\tablenotemark{b} \\
2431: & 2583 & 0.003 & $>$10 & 0.576 & 0.52 & 5.77 & 0.99\tablenotemark{b} \\
2432: & 3674 & 0.076 & 0.561 & ... & 0.51 & 0 & 1.25\tablenotemark{b} \\
2433: & 5040 & 0.066 & 0.577 & ... & 0.44 & 0 & 1.73\tablenotemark{b} \\
2434: & 5955 & 0.081 & 0.592 & ... & 0.40 & 0 & 1.30\tablenotemark{b} \\
2435: & 6471 & 0.081 & 0.590 & ... & 0.31 & 0 & 1.26\tablenotemark{b} \\
2436: IM Peg & 2527 & 0.029 & 3.028 & 0.989 & 0.37 & 0.54 & 0.99\tablenotemark{a} \\
2437: & 2528 & 0.026 & 3.350 & 1.035 & 0.27 & 1.43 & 1.29\tablenotemark{a} \\
2438: & 2529 & 0.048 & 1.753 & 0.728 & 0.29 & 0.52 & 1.26\tablenotemark{a} \\
2439: & 2530 & 0.023 & 2.324 & 0.981 & 0.28 & 1.02 & 1.01\tablenotemark{a} \\
2440: & 2531 & 0.011 & 2.983 & 0.984 & 0.30 & 1.09 & 1.01\tablenotemark{a} \\
2441: & 2532 & 0.027 & 1.987 & 0.833 & 0.25 & 0.51 & 0.93\tablenotemark{a} \\
2442: & 2533 & 0.085 & 1.293 & 0.288 & 0.26 & 0.23 & 1.23\tablenotemark{a} \\
2443: & 2534 & 0.007 & 1.690 & 0.834 & 0.30 & 0.43 & 1.09\tablenotemark{a} \\
2444: \tableline
2445: \enddata
2446: \tablenotetext{a}{Spectral fits on these observations were carried out
2447: without letting all parameters float in the final fitting step.}
2448: \tablenotetext{b}{Spectral fits on these observations were conducted
2449: by letting all parameters (except $N_{H}$, if known) float in the
2450: final fitting step.}
2451: \end{deluxetable}
2452:
2453: \clearpage
2454:
2455: \begin{deluxetable}{llrrrrrrr}
2456: \tablewidth{0pt}
2457: \tablecaption{Hardness ratios and normalized quantiles for the stars
2458: in X-Atlas, derived from the predicted ACIS-I aimpoint spectra.
2459: Hardness ratios and quantiles calculated from the predicted ACIS-S
2460: aimpoint spectra are available on the X-Atlas website. \label{tblqh}}
2461: \tablehead{
2462: \colhead{Star} & \colhead{ObsID} &
2463: \colhead{$HR_{1}$} & \colhead{$HR_{2}$} &
2464: \colhead{$Q_{25}$} & \colhead{$Q_{50}$} &
2465: \colhead{$Q_{75}$} & \colhead{$log_{10}(\frac{m}{1-m})$} & \colhead{$3 \frac{Q_{25}}{Q_{75}}$}
2466: \tabletypesize{5pt}
2467: \tablecolumns{9}
2468: }
2469: \startdata
2470: \tableline
2471: HD 93129AB & 5397 & -0.150 & 0.304 & 0.081 & 0.119 & 0.180 & -0.868 & 1.357 \\
2472: & 7201 & -0.369 & 0.521 & 0.085 & 0.114 & 0.159 & -0.892 & 1.606 \\
2473: & 7202 & -0.221 & 0.373 & 0.083 & 0.117 & 0.172 & -0.876 & 1.441 \\
2474: & 7203 & -0.282 & 0.448 & 0.085 & 0.121 & 0.171 & -0.860 & 1.499 \\
2475: & 7204 & -0.244 & 0.503 & 0.089 & 0.125 & 0.182 & -0.845 & 1.468 \\
2476: & 7228 & -0.269 & 0.566 & 0.091 & 0.127 & 0.172 & -0.837 & 1.582 \\
2477: & 7229 & -0.271 & 0.355 & 0.081 & 0.116 & 0.163 & -0.884 & 1.499 \\
2478: & Combined & -0.291 & 0.528 & 0.089 & 0.123 & 0.169 & -0.852 & 1.584 \\
2479: HD 93250 & 5399 & -0.178 & 0.434 & 0.087 & 0.121 & 0.184 & -0.860 & 1.422 \\
2480: & 5400 & -0.151 & 0.421 & 0.087 & 0.125 & 0.188 & -0.845 & 1.393 \\
2481: & 7189 & -0.262 & 0.514 & 0.089 & 0.123 & 0.174 & -0.852 & 1.532 \\
2482: & 7341 & -0.180 & 0.407 & 0.085 & 0.121 & 0.182 & -0.860 & 1.405 \\
2483: & 7342 & -0.185 & 0.489 & 0.089 & 0.123 & 0.191 & -0.852 & 1.395 \\
2484: & Combined & -0.205 & 0.466 & 0.087 & 0.123 & 0.180 & -0.852 & 1.452 \\
2485: HD 150136 & 2569 & -0.421 & 0.321 & 0.078 & 0.104 & 0.146 & -0.934 & 1.597 \\
2486: 9 Sgr & 5398 & -0.591 & 0.038 & 0.064 & 0.085 & 0.119 & -1.031 & 1.618 \\
2487: & 6285 & -0.591 & 0.008 & 0.064 & 0.085 & 0.119 & -1.031 & 1.618 \\
2488: & Combined & -0.613 & 0.077 & 0.066 & 0.087 & 0.119 & -1.020 & 1.666 \\
2489: $\zeta$ Pup & 640 & -0.602 & 0.155 & 0.070 & 0.091 & 0.127 & -1.000 & 1.656 \\
2490: $\theta^{1}$ Ori C & 3 & 0.056 & 0.646 & 0.106 & 0.152 & 0.226 & -0.748 & 1.411 \\
2491: & 4 & 0.064 & 0.631 & 0.104 & 0.153 & 0.226 & -0.742 & 1.386 \\
2492: & 2567 & 0.009 & 0.659 & 0.104 & 0.148 & 0.216 & -0.761 & 1.447 \\
2493: & 2568 & 0.068 & 0.614 & 0.104 & 0.152 & 0.231 & -0.748 & 1.352 \\
2494: Cyg OB2 8A & 2572 & 0.121 & 0.836 & 0.127 & 0.163 & 0.222 & -0.711 & 1.717 \\
2495: HD 206267 & 1888 & -0.709 & -0.054 & 0.061 & 0.080 & 0.112 & -1.063 & 1.626 \\
2496: & 1889 & -0.695 & -0.040 & 0.064 & 0.080 & 0.114 & -1.063 & 1.699 \\
2497: 15 Mon & 5401 & -0.767 & -0.472 & 0.047 & 0.064 & 0.081 & -1.162 & 1.743 \\
2498: & 6247 & -0.816 & -0.456 & 0.044 & 0.062 & 0.083 & -1.176 & 1.567 \\
2499: & 6248 & -0.630 & -0.342 & 0.045 & 0.066 & 0.093 & -1.149 & 1.468 \\
2500: & Combined & -0.726 & -0.338 & 0.055 & 0.070 & 0.089 & -1.123 & 1.850 \\
2501: $\xi$ Per & 4512 & -0.698 & -0.075 & 0.062 & 0.080 & 0.106 & -1.063 & 1.767 \\
2502: $\zeta$ Oph & 2571 & -0.706 & -0.067 & 0.062 & 0.080 & 0.102 & -1.063 & 1.832 \\
2503: & 4367 & -0.715 & -0.037 & 0.062 & 0.080 & 0.106 & -1.063 & 1.767 \\
2504: $\iota$ Ori & 599 & -0.797 & -0.283 & 0.053 & 0.070 & 0.089 & -1.123 & 1.786 \\
2505: & 2420 & -0.759 & -0.334 & 0.051 & 0.068 & 0.089 & -1.136 & 1.722 \\
2506: $\tau$ CMa & 2525 & -0.753 & -0.285 & 0.044 & 0.059 & 0.080 & -1.110 & 1.686 \\
2507: & 2526 & -0.771 & -0.333 & 0.040 & 0.055 & 0.077 & -1.136 & 1.561 \\
2508: $\theta^{2}$ Ori A & 4473 & -0.690 & 0.114 & 0.068 & 0.085 & 0.114 & -1.031 & 1.799 \\
2509: & 4474 & 0.115 & 0.577 & 0.108 & 0.155 & 0.241 & -0.735 & 1.346 \\
2510: $\sigma$ Ori & 3738 & -0.808 & -0.329 & 0.053 & 0.070 & 0.087 & -1.123 & 1.825 \\
2511: $\delta$ Ori & 639 & -0.723 & -0.163 & 0.059 & 0.076 & 0.099 & -1.086 & 1.787 \\
2512: $\zeta$ Ori & 610 & -0.785 & -0.225 & 0.055 & 0.072 & 0.095 & -1.110 & 1.739 \\
2513: & 1524 & -0.726 & -0.211 & 0.055 & 0.074 & 0.097 & -1.098 & 1.705 \\
2514: $\gamma$ Cas & 1895 & 0.320 & 0.594 & 0.123 & 0.186 & 0.362 & -0.642 & 1.020 \\
2515: $\epsilon$ Ori & 3753 & -0.797 & -0.254 & 0.055 & 0.072 & 0.091 & -1.110 & 1.811 \\
2516: $\tau$ Sco & 638 & -0.642 & 0.079 & 0.066 & 0.085 & 0.116 & -1.031 & 1.720 \\
2517: & 2305 & -0.656 & 0.073 & 0.066 & 0.085 & 0.114 & -1.031 & 1.749 \\
2518: $\beta$ Cru & 2575 & -0.865 & -0.525 & 0.044 & 0.061 & 0.080 & -1.191 & 1.641 \\
2519: $\gamma$ Vel & 629 & 0.510 & 0.834 & 0.153 & 0.203 & 0.294 & -0.595 & 1.567 \\
2520: WR 140 & 2337 & 0.411 & 0.796 & 0.142 & 0.203 & 0.354 & -0.595 & 1.203 \\
2521: & 2338 & 0.946 & 0.420 & 0.281 & 0.394 & 0.518 & -0.186 & 1.626 \\
2522: & 5419 & 0.094 & 0.745 & 0.114 & 0.159 & 0.243 & -0.723 & 1.406 \\
2523: & 6286 & 0.124 & 0.751 & 0.114 & 0.163 & 0.258 & -0.711 & 1.323 \\
2524: & 6287 & 0.112 & 0.742 & 0.114 & 0.161 & 0.254 & -0.717 & 1.343 \\
2525: $\eta$ Car & 632 & 0.967 & 0.652 & 0.336 & 0.453 & 0.590 & -0.082 & 1.707 \\
2526: & 3745 & 0.983 & 0.521 & 0.362 & 0.480 & 0.622 & -0.035 & 1.747 \\
2527: & 3746 & 0.751 & 0.218 & 0.193 & 0.368 & 0.557 & -0.235 & 1.041 \\
2528: & 3747 & 0.984 & 0.027 & 0.493 & 0.618 & 0.770 & 0.209 & 1.921 \\
2529: & 3748 & 0.981 & 0.349 & 0.389 & 0.508 & 0.658 & 0.014 & 1.772 \\
2530: & 3749 & 0.967 & 0.607 & 0.320 & 0.440 & 0.578 & -0.105 & 1.662 \\
2531: Algol & 604 & -0.271 & 0.395 & 0.083 & 0.112 & 0.167 & -0.900 & 1.499 \\
2532: Canopus & 636 & -0.686 & 0.105 & 0.068 & 0.085 & 0.110 & -1.031 & 1.861 \\
2533: HD 220657 & 3731 & -0.738 & 0.074 & 0.068 & 0.083 & 0.106 & -1.041 & 1.928 \\
2534: TZ CrB & 15 & -0.506 & 0.232 & 0.074 & 0.093 & 0.133 & -0.990 & 1.671 \\
2535: 44 Boo & 14 & -0.555 & 0.157 & 0.070 & 0.089 & 0.127 & -1.010 & 1.656 \\
2536: ER Vul & 1887 & -0.568 & 0.259 & 0.074 & 0.093 & 0.129 & -0.990 & 1.720 \\
2537: XI Uma & 1894 & -0.673 & 0.003 & 0.064 & 0.081 & 0.110 & -1.052 & 1.758 \\
2538: HD 111812 & 1891 & -0.625 & 0.336 & 0.076 & 0.093 & 0.125 & -0.990 & 1.817 \\
2539: HD 223460 & 1892 & -0.060 & 0.470 & 0.093 & 0.133 & 0.205 & -0.816 & 1.360 \\
2540: AR Lac & 6 & -0.481 & 0.328 & 0.076 & 0.099 & 0.140 & -0.961 & 1.621 \\
2541: & 7 & -0.203 & 0.413 & 0.085 & 0.117 & 0.178 & -0.876 & 1.435 \\
2542: & 8 & -0.282 & 0.365 & 0.081 & 0.110 & 0.163 & -0.909 & 1.499 \\
2543: & 9 & -0.457 & 0.310 & 0.076 & 0.099 & 0.142 & -0.961 & 1.599 \\
2544: & 10 & -0.455 & 0.293 & 0.076 & 0.097 & 0.140 & -0.971 & 1.621 \\
2545: & 11 & -0.422 & 0.305 & 0.076 & 0.099 & 0.146 & -0.961 & 1.558 \\
2546: V987 Tau & 3756 & -0.371 & 0.388 & 0.080 & 0.104 & 0.153 & -0.934 & 1.555 \\
2547: SU Aur & 3755 & 0.101 & 0.675 & 0.108 & 0.157 & 0.258 & -0.729 & 1.257 \\
2548: 24 Ursae Majoris & 2564 & -0.701 & -0.002 & 0.066 & 0.081 & 0.108 & -1.052 & 1.841 \\
2549: & 3471 & -0.726 & -0.029 & 0.064 & 0.080 & 0.104 & -1.063 & 1.854 \\
2550: TY Pyx & 601 & -0.478 & 0.304 & 0.076 & 0.097 & 0.140 & -0.971 & 1.621 \\
2551: UX Ari & 605 & -0.390 & 0.358 & 0.080 & 0.102 & 0.152 & -0.943 & 1.574 \\
2552: Capella & 57 & -0.749 & -0.059 & 0.064 & 0.080 & 0.102 & -1.063 & 1.888 \\
2553: & 1010 & -0.759 & -0.076 & 0.064 & 0.078 & 0.100 & -1.075 & 1.924 \\
2554: & 1099 & -0.722 & -0.077 & 0.064 & 0.078 & 0.102 & -1.075 & 1.888 \\
2555: & 1100 & -0.750 & -0.053 & 0.064 & 0.080 & 0.102 & -1.063 & 1.888 \\
2556: & 1101 & -0.741 & -0.059 & 0.064 & 0.080 & 0.102 & -1.063 & 1.888 \\
2557: & 1103 & -0.744 & -0.080 & 0.064 & 0.078 & 0.102 & -1.075 & 1.888 \\
2558: & 1199 & & & & & & & \\
2559: & 1235 & -0.750 & -0.077 & 0.064 & 0.078 & 0.102 & -1.075 & 1.888 \\
2560: & 1236 & -0.753 & -0.074 & 0.064 & 0.078 & 0.102 & -1.075 & 1.888 \\
2561: & 1237 & -0.755 & -0.063 & 0.064 & 0.080 & 0.102 & -1.063 & 1.888 \\
2562: & 1318 & -0.759 & -0.091 & 0.064 & 0.078 & 0.100 & -1.075 & 1.924 \\
2563: & 2583 & -0.725 & -0.038 & 0.064 & 0.080 & 0.104 & -1.063 & 1.854 \\
2564: & 3674 & -0.746 & -0.072 & 0.064 & 0.078 & 0.102 & -1.075 & 1.888 \\
2565: & 5040 & -0.743 & -0.023 & 0.066 & 0.080 & 0.104 & -1.063 & 1.908 \\
2566: & 5955 & -0.757 & -0.022 & 0.066 & 0.080 & 0.104 & -1.063 & 1.908 \\
2567: & 6471 & -0.747 & -0.009 & 0.066 & 0.080 & 0.104 & -1.063 & 1.908 \\
2568: HD 93497 & 1890 & -0.738 & -0.013 & 0.066 & 0.080 & 0.104 & -1.063 & 1.908 \\
2569: & 3410 & -0.710 & 0.003 & 0.066 & 0.081 & 0.106 & -1.052 & 1.874 \\
2570: FK Com & 614 & -0.031 & 0.418 & 0.091 & 0.135 & 0.207 & -0.808 & 1.320 \\
2571: $\lambda$ And & 609 & -0.485 & 0.328 & 0.076 & 0.099 & 0.142 & -0.961 & 1.599 \\
2572: DoAr 21 & 3761 & 0.567 & 0.908 & 0.161 & 0.231 & 0.402 & -0.522 & 1.202 \\
2573: Speedy Mic & 2536 & -0.459 & 0.249 & 0.074 & 0.097 & 0.138 & -0.971 & 1.602 \\
2574: & 3491 & -0.441 & 0.193 & 0.070 & 0.093 & 0.138 & -0.990 & 1.520 \\
2575: HIP 92680 & 3729 & -0.431 & 0.304 & 0.076 & 0.099 & 0.144 & -0.961 & 1.578 \\
2576: VW Cep & 3766 & -0.595 & 0.178 & 0.070 & 0.089 & 0.123 & -1.010 & 1.707 \\
2577: $\beta$ Ceti & 974 & -0.689 & 0.081 & 0.068 & 0.083 & 0.112 & -1.041 & 1.830 \\
2578: HDE 245059 & 5420 & -0.469 & 0.230 & 0.072 & 0.095 & 0.136 & -0.980 & 1.582 \\
2579: & 6241 & -0.316 & 0.229 & 0.074 & 0.100 & 0.153 & -0.952 & 1.444 \\
2580: & 7253 & -0.346 & 0.265 & 0.076 & 0.100 & 0.152 & -0.952 & 1.499 \\
2581: V824 Ara & 2538 & -0.506 & 0.257 & 0.074 & 0.095 & 0.135 & -0.980 & 1.647 \\
2582: AB Dor & 16 & -0.464 & 0.232 & 0.072 & 0.095 & 0.138 & -0.980 & 1.561 \\
2583: YY Men & 2557 & -0.069 & 0.462 & 0.093 & 0.133 & 0.199 & -0.816 & 1.399 \\
2584: $\sigma$ Gem & 5422 & -0.225 & 0.391 & 0.083 & 0.116 & 0.174 & -0.884 & 1.434 \\
2585: & 6282 & -0.285 & 0.386 & 0.081 & 0.110 & 0.165 & -0.909 & 1.482 \\
2586: IM Peg & 2527 & -0.250 & 0.392 & 0.083 & 0.112 & 0.169 & -0.900 & 1.482 \\
2587: & 2528 & -0.355 & 0.387 & 0.081 & 0.106 & 0.155 & -0.926 & 1.572 \\
2588: & 2529 & -0.363 & 0.408 & 0.081 & 0.108 & 0.155 & -0.917 & 1.572 \\
2589: & 2530 & -0.361 & 0.408 & 0.081 & 0.106 & 0.155 & -0.926 & 1.572 \\
2590: & 2531 & -0.330 & 0.392 & 0.081 & 0.108 & 0.159 & -0.917 & 1.535 \\
2591: & 2532 & -0.300 & 0.405 & 0.083 & 0.110 & 0.163 & -0.909 & 1.534 \\
2592: & 2533 & -0.315 & 0.416 & 0.083 & 0.110 & 0.161 & -0.909 & 1.552 \\
2593: & 2534 & -0.348 & 0.392 & 0.081 & 0.108 & 0.157 & -0.917 & 1.553 \\
2594: II Peg & 1451 & -0.165 & 0.423 & 0.087 & 0.121 & 0.184 & -0.860 & 1.422 \\
2595: TV Crt & 3728 & -0.448 & 0.192 & 0.072 & 0.093 & 0.135 & -0.990 & 1.605 \\
2596: V4046 Sgr & 5423 & -0.535 & 0.129 & 0.070 & 0.087 & 0.123 & -1.020 & 1.707 \\
2597: & 6265 & -0.493 & 0.134 & 0.068 & 0.089 & 0.125 & -1.010 & 1.635 \\
2598: TW Hya & 5 & -0.501 & 0.139 & 0.070 & 0.087 & 0.123 & -1.020 & 1.707 \\
2599: CC Eri & 4513 & -0.617 & 0.171 & 0.070 & 0.089 & 0.117 & -1.010 & 1.789 \\
2600: & 6132 & -0.285 & 0.369 & 0.081 & 0.108 & 0.165 & -0.917 & 1.482 \\
2601: AU Mic & 17 & -0.530 & 0.179 & 0.070 & 0.091 & 0.127 & -1.000 & 1.656 \\
2602: Hen 3-600 & 4502 & -0.458 & 0.150 & 0.070 & 0.093 & 0.133 & -0.990 & 1.585 \\
2603: AD Leo & 2570 & -0.619 & 0.073 & 0.066 & 0.085 & 0.112 & -1.031 & 1.779 \\
2604: EV Lac & 1885 & -0.504 & 0.101 & 0.066 & 0.089 & 0.127 & -1.010 & 1.566 \\
2605: Proxima Centauri & 2388 & -0.678 & -0.031 & 0.062 & 0.081 & 0.106 & -1.052 & 1.767 \\
2606: \tableline
2607: \enddata
2608: \end{deluxetable}
2609:
2610: \clearpage
2611:
2612: \begin{deluxetable}{lcccc}
2613: \tablewidth{0pt}
2614: \tablecaption{Overview of the Gregory-Loredo variability analysis
2615: results. Stars were graded by their variability index. Those with
2616: indices of 7-10 have a greater than 99\% probability of being variable. Those with
2617: indices of 6 were probably variable (greater than 90\% probability)
2618: and those scoring below 6 (less than 90\% probability) were graded not
2619: variable \citep{rots06}. For stars for which multiple observations
2620: have been made, the star's variability was judged by the observation
2621: with the highest variability index. \label{tblvar}}
2622: \tablehead{
2623: \colhead{ } & \colhead{Number of Stars} & \colhead{Variable} &
2624: \colhead{Probably Variable} & \colhead{Not Variable}
2625: \tabletypesize{5pt}
2626: \tablecolumns{5}
2627: }
2628: \startdata
2629: \tableline
2630: High-Mass Stars & 24 & 10 & 1 & 13\\
2631: Low-Mass Stars & 40 & 34 & 1 & 5\\
2632: \tableline
2633: Total & 64 & 44 & 2 & 18\\
2634: \tableline
2635: \enddata
2636: \end{deluxetable}
2637:
2638: \end{document}
2639: