1: \documentclass[preprint]{aastex}
2:
3: \usepackage{amsmath}
4: \usepackage{graphicx}
5:
6: \shorttitle{AO O Star Companion Survey}
7: \shortauthors{Turner et~al.}
8:
9: \begin{document}
10:
11: \input{title}
12:
13: \section{INTRODUCTION}
14:
15: Binary star systems are frequently the only source for the fundamental
16: determination of the most basic property of a star, its mass.
17: The masses of a sample of stars, when determined with sufficient accuracy,
18: serve as crucial tests of our theoretical understanding of stellar formation,
19: structure, and evolution.
20: By combining information from the angular orbit projected on the sky (or from
21: a light curve in the case of an eclipsing system) with the spectroscopic orbit,
22: one can determine the distance and masses of the individual stars.
23: In the cases of the most massive stars, accurate masses are even more
24: important, not only because massive stars are relatively rare, but because
25: these stars (initial masses $>20 M_\odot$) play critical roles in galaxies.
26: Massive stars serve as signposts of star formation in galaxies
27: \citep{1995AJ....110.2715M}.
28: Current theory suggests that high mass stars are born in loosely bound star
29: clusters which disperse after a few Myr.
30: These birth conditions suggest that high mass stars might often have
31: companions, as for example, found around $\gamma^{2}$ Velorum
32: \citep{2000MNRAS.313L..23P}.
33: Some of these could be very close and gravitationally bound to the O star,
34: which is suggested by the prevalence of binary and triple systems among young
35: O stars \citep{MasonEtAl}.
36:
37: \cite{MasonEtAl} performed a speckle interferometry survey of Galactic
38: O-type stars for close companions, specifically looking for differences in
39: the multiplicity frequencies amongst the cluster, field, and runaway O-type
40: star populations, as well as the distribution of orbital periods.
41: They did their survey in the $V$-band using speckle interferometry, which is
42: sensitive to the detection of a companion if the projected separation is in
43: the range $0\farcs035 < \rho < 1\farcs5$ and the magnitude difference
44: $\triangle m_{V} < 3$.
45: Our survey is in the $I$-band (which will slightly emphasize redder companions)
46: and extends the dynamic range of their work.
47: Due to the availability of only one photometric band for this survey, we are
48: unable to determine physical associations for the new companions with any
49: certainty, and we can only address the likelihood that perhaps some of the
50: companions are gravitationally bound.
51: The results presented here invite further investigation into the O stars and
52: their potential to shed light on star formation processes.
53:
54: \section{OBSERVATIONS}
55:
56: The data were taken using the adaptive optics (AO) system and Visible Imager
57: (VisIm) camera of the Advanced Electro-Optical System (AEOS) 3.6-m telescope
58: at the Maui Space Surveillance System, located on Haleakala.
59: The data were collected during four separate observing runs between 2001
60: February and 2002 September.
61: In addition to the dedicated observing runs, some observations were taken as
62: part of a queue-scheduled observation program between 2001 May and 2005
63: November.
64:
65: \subsection{AEOS Telescope and Adaptive Optics System}
66:
67: The AEOS telescope is an altitude-azimuth (alt-az) design.
68: While the telescope is very flexible and supports several different optical
69: configurations \citep{RobertsNeyman}, of interest here is the Cassegrain
70: configuration which feeds a 726-m focal length, $f/200$ beam (with a
71: field of view of $61\farcs9$) through a Nasmyth port coud\'{e} path to a fold
72: flat below the telescope which directs the beam onto a custom optical table
73: that supports the main components of the AO system.
74: For this project, the short-wavelength light (400 -- 700 nm) is sent to the
75: wavefront sensor (WFS) and tracker camera beam trains, while the 700 nm to
76: 5 $\mu$m light is sent to the VisIm, which is used as the science detector.
77:
78: The heart of the AO system is a Shack-Hartmann WFS driving a 941-actuator
79: Xinetics deformable mirror (DM) with a stroke of $\pm4 \mu$m.
80: Optics inside the WFS produce a pupil image on a Hartmann lens array.
81: Each generated Hartmann spot is then imaged onto a 4$\times$4 group of pixels
82: on the WFS CCD.
83: These pixel values can be used directly or binned (for higher throughput) to
84: generate the wavefront slopes.
85: The WFS CCD is a Lincoln Labs frame-transfer device with 16 output ports,
86: capable of frame rates of 0.2 to 2.5 kHz.
87: The wavefront slopes are calculated from the Hartmann spot values by means of
88: many digital signal processors working in parallel, feeding these data to a
89: system which reconstructs the wavefront, taking into account alt-az induced
90: image rotation and WFS-to-DM-actuator mapping artifacts.
91:
92: The VisIm camera is described in detail in \cite{RobertsNeyman}.
93: In short, it operates from 700 to 1050 nm; is equipped with an atmospheric
94: dispersion corrector; has a two-mode image derotator (zenith at a fixed
95: position in the image or celestial north at a fixed position in the image);
96: and, for this project, has a $10''$ field of view ($0\farcs020$ pixel$^{-1}$).
97: The detector is a 512$\times$512 pixel EEV CCD with a dark current of
98: 22 $e^{-}$ pixel$^{-1}$ s$^{-1}$ (at $-$40$^{\circ}$ C) and a read noise of
99: 12 $e^{-}$ rms.
100: The camera output is digitized to 12 bits with 10 $e^{-}$ per digital number.
101:
102: \subsection{Object Selection}
103:
104: The object list started out as all the components in the 228 systems listed
105: in \cite{MasonEtAl}.
106: They were then culled for the effective magnitude limit of the AEOS AO system
107: (about $m_{V} = 8$), and for the declination limit of acceptable AO
108: correction (objects that at some time during the year get above $30^{\circ}$
109: elevation at Haleakala, i.e., a declination greater than about $-45^{\circ}$).
110: This reduced the list to 164 objects.
111: We added seven more targets from the Galactic O Star
112: Catalog\footnote{{\tt http://www-int.stsci.edu/$\sim$jmaiz/research/GOS/GOSmain.html}}
113: \citep{GOS} that are within our adopted magnitude and declination limits.
114: We actually observed, at least once, 116 of the 171 objects in our list.
115: These observations are fairly consistently distributed among the three O-star
116: populations, sampling 63\% of the cluster membership stars, 78\% of the field
117: stars, and 88\% of the runaway stars in our list of potential targets.
118:
119: %Numbers:
120: %Cluster -- 81 observed of 128
121: %Field -- 21 observed of 27
122: %Runaway -- 14 observed of 16
123:
124: \subsection{Data Collection and Reduction}
125:
126: Since VisIm is only a 12-bit camera, a stellar image can overflow the
127: digitizer in a rather short exposure time.
128: If the frame saturates, then it is omitted from the final image.
129: Frequently, this saturated frame will represent a period of particularly good
130: seeing.
131: In order to keep as many of these ``good seeing'' frames as possible, we set
132: the exposure time such that the average peak value was about 75\% of the
133: full-well depth.
134: We then built up the signal-to-noise ratio by taking many frames, then
135: weighting and summing them.
136: Figure~\ref{fig:dynamic} shows the effect of this summing on the detectable
137: magnitude difference for 1, 10, 50, 100, 250, and 1000 frame(s).
138: %
139: \input{dynamic_figure}
140: %
141: On the basis of this plot, we aimed to take at least 1000 frames of each
142: object, although time constraints, weather, and object brightness occasionally
143: limited us to a lower frame count.
144:
145: The VisIm is a frame transfer camera and is therefore unable to take a bias
146: frame.
147: To compensate, a bias frame was created by taking many dark exposures at a
148: variety of exposure times ranging from 10 ms to several minutes.
149: A linear fit was made to the signal from each pixel as a function of exposure
150: time; the $y$-intercepts became the bias frame while the slopes became the dark
151: frame.
152: Prior to 2005, flat-field frames were created using the twilight sky to
153: evenly illuminate the VisIm detector.
154: \cite{amos2001} found that sky flat-field frames varied at about the 1\% level,
155: so after 2005, flat-field frames were generated using an internal calibration
156: sphere; this lowered the frame-to-frame variation below the 0.01\% level.
157: From this point, all data frames were debiased, dark-subtracted and
158: flat-fielded in the conventional way.
159:
160: For a given sequence of data frames, a weighted shift-and-add algorithm was
161: used to create the final image.
162: The weighted shift-and-add algorithm \citep{spie3353-391} is a modification of
163: the traditional image-stacking algorithm that takes seeing conditions in each
164: individual frame into account.
165: Frames with higher peak values (which represent better seeing) influence the
166: final image (by means of weighting) more than frames with lower peak values.
167:
168: The fitting algorithm is described in detail in \cite{diffmag1,diffmag2}.
169: In short, the point-spread function (PSF) used to represent the system
170: performance is that of the primary star in the image, with a few modifications;
171: near the primary, the PSF is a pixel-for-pixel table of numbers, while farther
172: out the PSF is a radially-symmetric series of values.
173: This PSF is fitted to the primary and secondary components and iterated until
174: the intensity ratio converges.
175:
176: \subsection{Detection Sensitivity}
177:
178: Figure~\ref{fig:dynamic} only gives a qualitative sense of the companion
179: detection sensitivity of the system.
180: Cross comparisons of the reduced images, especially of the same object taken on
181: different nights, show different detection limits.
182: This is primarily due to variations in AO performance which is primarily
183: attributed to variations in the atmospheric turbulence profile.
184: In order to quantify the sensitivity of the reduced images, we have created a
185: variation of the ``dynamic range map'' technique described in
186: \cite{2007ApJ...654..633H} which defines the dynamic range of a given position
187: in a 2-D image as the faintest companion detectable at that position to the
188: $5\sigma$ level.
189: In our version, we construct a map the same size as each reduced image.
190: The intensity level of each pixel in the map is set to five times the RMS
191: intensity variation across a patch centered on the corresponding pixel
192: in the original image.
193: The patch is a square with lengths equal to the FWHM of the original image.
194: This produces the dynamic range map in intensity terms, which are then
195: converted to magnitudes.
196:
197: Figure~\ref{fig:dyn_rang_maps} shows an example of this technique applied
198: to one of the images generated in this survey.
199: %
200: \input{hd_034656_2003_269_bw_figure}
201: %
202: It is apparent from the sample figure that the detection threshold for a
203: companion is highly spatially variant.
204: The dynamic range increases with increasing radius from the central star.
205: There are several artifacts in this image which lower the dynamic range.
206: The largest is the vertical line, which is an artifact of the shutter-less
207: frame transfer process.
208: There are also four waffle spots caused by the Fried geometry of the wavefront
209: sensor \citep{2005PASP..117..831M}.
210: There are also hints of the diffraction pattern caused by the secondary mirror
211: support spiders.
212: As the AO system performance decreases, the FWHM of the PSF will increase,
213: which increases the area over which the central PSF causes confusion.
214: In addition, it also widens the companion's PSF, lowering the contrast between
215: the two and decreasing the detection rate.
216: After 2003, the waffle spots were greatly reduced when new reconstructors were
217: put in place to filter out the unsensed waffle modes.
218:
219: These maps have several purposes.
220: For those objects with multiple observations where companions are seen in some
221: but not all of the data, changes in the dynamic range can help explain the
222: reasons for these discrepancies.
223: They also illustrate the limitations of the data, showing where it is almost
224: impossible to find faint companions.
225: This is often useful when trying to detect known long-period spectroscopic
226: companions.
227:
228: \section{RESULTS}
229:
230: \subsection{New Pairs and New Measures}
231: \label{ssec:pairs-measures}
232:
233: Table~\ref{tbl:merge} lists the 40 new and 24 known pairs detected in the
234: survey.
235: Four of the new companions (HD 17505, HD 37468, HD 193793, and HD 217086) are
236: in systems with previously known components in the AEOS field of view.
237: The remainder are new companions in systems not previously known to be
238: multiple (at least in the AEOS field of view).
239: The table has nine columns which summarize the information gleaned from the
240: measurements: Column (1) lists the Washington Double Star (WDS) identifier,
241: columns (2) and (3) identify the object by HD and Hipparcos number, and
242: column (4) gives the discoverer designation (TRN for new pairs).
243: Circumstances of the observation are given in columns 5--8.
244: Column (5) gives the Besselian epoch of the start of the observation, columns
245: (6) and (7) give the position angle (in degrees, where north is 0$^{\circ}$
246: and east is 90$^{\circ}$) and separation (in seconds of arc) between the
247: primary (the survey star -- the brightest star in the image) and the detected
248: component.
249: Notice that for many of the objects, the position angle is missing or
250: ambiguous.
251: The VisIm derotator is usually in one of two modes, one which keeps zenith up
252: in the image, and one which keeps north up in the image.
253: In the observations with more than one or a missing position angle, the state
254: of the derotator was unknown, leaving the orientation of the image on the sky
255: unknown.
256: However, in some cases, position of a known pair resolves the ambiguity while
257: for the rest, new observations will be required.
258: Column (8) gives the magnitude difference at $I$-band between the primary and
259: the detected component.
260: Finally, column (9) lists the spectroscopic status.
261: Given the separations detected here, none of the short period spectroscopic
262: pairs are expected to be detected.
263: However, listing the spectroscopic binaries helps to form the complete
264: multiplicity picture.
265: The indicator in this table is {\bf V} (velocity variable) if one or more close
266: spectroscopic companions is indicated or {\bf C} (velocity constant) if radial
267: velocity work indicates no close companion.
268: The spectroscopic reference is generally given in \cite{MasonEtAl}.
269:
270: The differential magnitude errors in column~8 of Table~\ref{tbl:merge} were
271: generated from simulated binary data that we analyzed using the same fitting
272: algorithm applied to the measured binary data.
273: To generate the simulated binaries, we started by collecting all the single
274: star images from this survey (see Table~\ref{tbl:sing}) and those of a
275: duplicity survey B stars carried out contemporaneously
276: \citep[see][Table~2]{2007AJ....133..545R}.
277: We omitted single star images where there were fewer than 250 frames
278: used to make the final image.
279: This exercise gave us 167 unique images with PSFs ranging from $0\farcs083$ to
280: $0\farcs445$, with a decided emphasis on values less than $0\farcs2$.
281: For each entry in Table~\ref{tbl:merge}, 167 simulated binaries were
282: constructed using the magnitude difference in column~8 and the pixel positions
283: of the primary and secondary components in the main image.
284: We then ran the fitting algorithm on each of these 167 simulated images for
285: each table entry.
286: We calculated the magnitude variance in two passes, the first to filter out
287: outliers on the basis of the variance of the fitted pixel separation of the
288: simulated binary (namely, any deviation greater than $1\sigma$), and the second
289: to use this reduced list of simulated fits to generate the final magnitude
290: variance.
291: In general, the number of outliers per table entry was less than 10.
292: The errors quoted in column~8 are the standard errors of these variances.
293: Errors are not quoted for the separation (column~7), but in all cases, the
294: standard error is less than $0\farcs04$, usually less than $0\farcs005$.
295: For errors in position angle (column~6), we adopt the values from
296: \cite{2007AJ....133..545R}, $\pm2^{\circ}$ for separations less than $1''$,
297: and $\pm1^{\circ}$ for separations greater than $1''$.
298: Two effects lead to this uncertainty, the clocking calibration of the dove
299: prism in the derotator, and the slight natural image rotation during the frame
300: collection sequence when the derotator is in the fixed zenith position mode.
301:
302: \subsubsection{Notes to Pairs}
303:
304: Here we discuss the various binary systems in Table~\ref{tbl:merge} by
305: WDS identifier and discovery designation.
306: In the cases of the stars classified as runaways, it is unlikely that the
307: detected companion is physical (see \S~\ref{sec:discussion}).
308:
309: \noindent {\bf 02229$+$4129 = TRN\phn\phn10}:
310: This is a runaway star with a close, low-luminosity companion
311: \citep{2005ApJ...621..978B}.
312:
313: \noindent {\bf 02407+6117 = TRN\phn\phn12AD}:
314: The A component is a spectroscopic triple \citep{2003ApJ...595.1124M}.
315:
316: \noindent {\bf 02511$+$6025 = TRN\phn\phn13AH}:
317: The A component of HD~17505 is a spectroscopic triple
318: \citep{2006ApJ...639.1069H}.
319: It is uncertain if the new component (AH, $\rho \sim$ 4\farcs6) and the known
320: pair of Table~\ref{tbl:merge} (STF\phn306AB) are both gravitationally bound.
321: If STF\phn306AB is indeed gravitationally bound, \cite{2006ApJ...639.1069H}
322: estimate it to have a period of 27,000 years.
323:
324: \noindent {\bf 03556$+$5238 = HDS\phn494}:
325: The $\triangle$m was probably a little too large for \cite{MasonEtAl} to detect
326: this pair, but it was detected in later speckle observations by
327: \cite{2001AJ....121.3224M}.
328:
329: \noindent {\bf 03590$+$3548 = TRN\phn\phn16}:
330: Classified as a runaway star in \cite{1986ApJS...61..419G}.
331:
332: \noindent {\bf 05163$+$3419 = TRN\phn\phn17Aa}:
333: Classified as a runaway star in \cite{1986ApJS...61..419G}. SEI\phn136AB
334: ($\rho \sim$ 8\farcs8, $\triangle$m $\sim$ 3.3, just outside the field of view)
335: is a likely optical pair, with the B component not physically associated with
336: this close double.
337:
338: \noindent {\bf 05207$+$3726 = TRN\phn\phn18}:
339: One new close pair and three wide pairs are found here. The close one is
340: designated Aa while the wider pairs are designated AB, AC, and AD. The known
341: pair SEI\phn201 ($\rho \sim$ 24\farcs7, $\triangle$m $\sim$ 5.9) is outside
342: the field of view and has been designated AE.
343:
344: \noindent {\bf 05297$+$3523 = HU\phn\phn217}:
345: This known, relatively close, pair has been measured since 1900.
346: The position angle has only changed by $4^{\circ}$ over this time span.
347:
348: \noindent {\bf 05353$-$0523 = STF\phn748Ca,F}:
349: The primary is $\theta^1$~Ori~C, the most massive star in the Orion Trapezium.
350: It has a closer companion discovered by speckle methods
351: \citep{2007AaA...466..649K}.
352:
353: \noindent {\bf 05354$-$0525 = CHR\phn249Aa}:
354: HIPPARCOS \citep{1997hity.book.....P} measured a $\triangle H_{p}$ of
355: 3.23.
356: \cite{2000AaA...356..141F} found $\triangle B_{T} = 4.37 \pm 0.02$
357: and $\triangle V_{T} = 3.21 \pm 0.01$.
358:
359: \noindent {\bf 05387$-$0236 = TRN\phn\phn19AF}:
360: The new component (AF) is at a separation nicely nestled between BU\phn1032AB
361: and STF\phn762AC, at a value which would be consistent with a hierarchical
362: arrangement.
363: The new TRN~19AF component is much brighter in the mid-infrared where it was
364: first discovered \citep{2003AaA...405L..33V} and it is probably a low mass
365: K-star surrounded by a circumstellar disk \citep{2007AaA...466..917C}.
366: See \S~\ref{orbit_discussion} below for a discussion of the orbit of BU~1032AB.
367:
368: \noindent {\bf 05407$-$0157 = STF\phn774Aa-B}:
369: This pair has an orbit \citep{1967MiWie..13...49H} with residuals of:
370: O$-$C = $-5^{\circ}$ and $-$0\farcs03 in position angle and separation.
371: However, the orbit is classified as ``indeterminate'' (grade 5) in the
372: {\it 6th Catalog of Orbits of Visual Binary
373: Stars}\footnote{{\tt http://ad.usno.navy.mil/wds/orb6.html}}, so
374: these residuals are not indicative of measurement quality.
375: While the total number of measures has increased by about 50\% since this orbit
376: was published, the initial elements are of such low quality that even at this
377: point any sort of improvement above ``indeterminate'' subjective quality is not
378: yet possible.
379: Some of the scatter in the residuals may be due to the presence of a closer
380: companion to $\zeta$~Ori~A found by interferometry \citep{2000ApJ...540L..91H}.
381:
382: \noindent {\bf 06319$+$0457 = GAN\phn\phn\phn3AB}:
383: The measure agrees with earlier published data quite well and indicates that
384: the measure made in 1928 \citep{1931AN....241...33S} is probably erroneous.
385:
386: \noindent {\bf 06410$+$0954 = STF\phn950Aa-B}:
387: This pair has only changed by $\sim$9$^{\circ}$ and 0\farcs2 since it was first
388: resolved by Struve in 1825 \citep{1837AN.....14..249S}.
389:
390: \noindent {\bf 16550$-$4109 = I\phn\phn\phn576}:
391: I\phm{888}576 has not been measured since 1934 \citep{Wallenquist1934}, so
392: identification of this pair with that of Innes is uncertain.
393:
394: \noindent {\bf 17065$-$3527 = B\phn\phn\phn894}:
395: This binary, unconfirmed since its first resolution by \cite{vandenBos1928},
396: has closed slightly.
397:
398: \noindent {\bf 17158$-$3344 = SEE\phn322}:
399: This pair has shown noticeable Keplerian motion, completing about 70$^{\circ}$
400: of orbital revolution since discovery \citep{1898AJ.....18..181S}.
401: See \S~\ref{orbit_discussion} below for a discussion of the orbit.
402:
403: \noindent {\bf 17347$-$3235 = HDS2480Ab}:
404: The pairs, ISO\phn\phn\phn5 and HJ\phn4962 should have been seen as well.
405: ISO\phn\phn\phn5 was probably missed due to the corrected FWHM.
406: HJ\phn4962 was probably just out of the field.
407:
408: \noindent {\bf 18024$-$2302 = TRN\phn\phn26AH}:
409: While this pair may be physical, the known double H\phm{8}N\phm{88}40 almost
410: certainly is not.
411:
412: \noindent {\bf 18026$-$2415 = RST3149AB}:
413: This pair has been seen only twice since 1935.
414: The occultation pair of \cite{1975AJ.....80..689A} was not seen.
415:
416: \noindent {\bf 18061$-$2412 = B\phn\phn\phn376}:
417: Unconfirmed since its discovery in 1927 \citep{vandenBos1928}.
418: The A component is an eclipsing binary with a 4.6 d period
419: \citep{2007OEJV...72....1O}.
420:
421: \noindent {\bf 20035$+$3601 = STF2624Aa-B}:
422: The close pair, MCA\phn\phn59Aa, which is notoriously difficult to detect, was
423: not seen here.
424:
425: \noindent {\bf 20181$+$4044 = STF2666Aa-B}:
426: The close speckle pair CHR\phn\phn96Aa was not recovered.
427:
428: \noindent {\bf 20205$+$4351 = TRN\phn\phn29AC}:
429: It is unclear whether there are two new components or a nearby optical pair
430: with a large relative proper motion detected at two different positions.
431:
432: \noindent {\bf 21390$+$5729 = BU\phn1143AB}:
433: The close Aa pair (MIU\phn\phn\phn2) was not recovered, although it may have
434: been too close.
435:
436: \noindent {\bf 21449$+$6228 = TRN\phn\phn33AC}:
437: The visual pair, MLR\phn\phn16, would have been much too wide (17$''$\llap.6)
438: to detect here.
439:
440: \noindent {\bf 22393$+$3903 = TRN\phn\phn36AC}:
441: The visual binary, S\phn\phn\phn813, at $\sim$ 1 arcminute of separation is
442: much too wide to detect here.
443:
444: \noindent {\bf 22568$+$6244 = TRN\phn\phn37AC}:
445: It is hard to imagine a geometry where both this pair and MLR\phn266 would be
446: dynamically stable. More than likely one (or both) are optical.
447:
448: \subsubsection{Comparison of $V$- and $I$-band Magnitude Differences}
449:
450: Among the O star primaries listed in the WDS, we may estimate the companion
451: spectral type from the magnitude difference $\triangle V$ for those with small
452: magnitude differences.
453: As the magnitude difference gets larger, the uncertainty in the companion
454: spectral type increases.
455: However, the situation improves with the availability of magnitude differences
456: for two filters.
457: We compare the $V$-band magnitude differences from the WDS with the $I$-band
458: values we determined by AO.
459: The result is shown in Figure~\ref{fig:dvdi}.
460: %
461: \input{dvdi_figure}
462: %
463: Error bars for the visual data were based on typical Tycho-2 photometric errors
464: when available (smaller-$\triangle$m pairs), and on scatter among individual
465: visual estimates for those (larger-$\triangle$m) systems with more than one
466: such estimate.
467: A rough average value of this scatter (0.7 mag) was used for systems with only
468: one visual $\triangle m_{V}$ estimate.
469: Mean $\triangle m_I$ values and error bars are given for systems with
470: multiple observations.
471: For systems with no $I$-band error estimate a default value of 0.5 magnitudes
472: was adopted.
473:
474: This figure shows that $\triangle V \approx \triangle I$ for small magnitude
475: differences, but $\triangle V$ increases more rapidly at larger values of
476: $\triangle I$.
477: This is as expected for our set of O-star targets: the components of
478: small-$\triangle$m systems are of similar spectral type, but the secondaries
479: become later in spectral type --- hence redder --- as $\triangle$m increases.
480:
481: To determine the expected extent of this trend, absolute $V$ and $I$ magnitudes
482: for main sequence spectral types were extracted from
483: {\it Allen's Astrophysical Quantities} \citep{APQ2001}.
484: Values of $\triangle V$ and $\triangle I$ were then determined, assuming
485: primaries of spectral types O3, O5, O8, and B0 and secondaries covering a wide
486: range of spectral types.
487: The results, plotted as dotted lines in Figure~\ref{fig:dvdi}, appear to be in
488: good agreement with the trend seen in these AO observations and are consistent
489: with the assumption that the companions are physical.
490:
491: \subsubsection{Orbits of BU~1032 ($\sigma$~Ori~AB) and SEE~322 (HD~155889)}
492: \label{orbit_discussion}
493:
494: \cite{1990ebua.conf..419W} provides guidelines as to when a collection of
495: binary star measures merits the publication of a new orbit.
496: Following these criteria, we determined that two systems, BU~1032 and SEE~322,
497: deserved new orbital solutions.
498: All available measures for these pairs were extracted from the WDS
499: \citep{WDS2001} and individual measures were weighted following the precepts of
500: \cite{2001AJ....122.3472H}.
501: The orbital elements were determined with an iterative three-dimensional
502: grid-search algorithm \citep{2002AJ....123.1023S}.
503: For these two pairs, orbital elements are provided in Table~\ref{tbl:orb_el}
504: and future ephemerides in Table~\ref{tbl:ephem}.
505: Elements from the \cite{1996AJ....111..370H} orbit for BU~1032 are provided for
506: comparison.
507: Due to the preliminary nature of the SEE 322 orbit, errors are not quoted.
508: These orbits are illustrated in Figures~\ref{fig:wds05387-0236} and
509: \ref{fig:wds17158-3344}.
510:
511: Since the publication of the orbital elements of BU\phn1032 by
512: \cite{1996AJ....111..370H}, the number of measures has increased by about 35\%,
513: and while the prior orbit adequately fits these data, a new calculation at this
514: time makes a significant improvement in the orbit quality, as illustrated by
515: the decrease in formal errors for each of the elements.
516: The distance and mass of the $\sigma$~Ori~AB system are discussed in a recent
517: paper by \cite{2008MNRAS.383..750C}.
518:
519: There is no previous determination of orbital elements for SEE~322.
520: While the errors and residuals are large, the orbital solution looks promising,
521: although it is still preliminary since only a limited portion of the orbit is
522: covered by existing data.
523: The total mass associated with this solution is very large ($260~M_\odot$ for a
524: distance of 1.2~kpc; \citealt{MasonEtAl}), which may indicate that revisions are
525: required in the assumed elements and distance, or that the system contains more
526: than two stars.
527:
528: \input{wds05387-0236_figure}
529:
530: \input{wds17158-3344_figure}
531:
532: \subsection{Single Star Detections}
533:
534: In addition to discovering new companions and measuring some parameters of
535: previously known stars, we have a number of stars determined to be single under
536: the specific observing conditions, AO system performance, and field of view.
537: Table~\ref{tbl:sing} lists these stars.
538: In this table, the respective columns give the HD number, Hipparcos number,
539: epoch of observation, full-width at half-maximum of the corrected stellar PSF,
540: and the V/C code for the spectroscopic binary status
541: (see \S~\ref{ssec:pairs-measures}).
542: A larger FWHM is indicative of poorer seeing, a dimmer primary, non-optimal AO
543: system tuning, or some combination of these factors.
544:
545: \subsubsection{Notes to Unresolved Systems}
546: \label{unresolved_notes}
547:
548: \noindent {\bf HD \phn15137}:
549: This runaway star is a probable, single-lined spectroscopic binary
550: \citep{2005ApJ...621..978B}.
551:
552: \noindent {\bf HD \phn25638}:
553: The pair ES\phn2603 ($\rho \sim$ 6\farcs5, $\triangle m \sim$ 6.0) was
554: not detected.
555: It was probably just out of the field of view.
556:
557: \noindent {\bf HD \phn30614}:
558: Classified as a runaway star in \cite{1986ApJS...61..419G}.
559:
560: \noindent {\bf HD \phn36879}:
561: This star is a runaway object according to its proper motion
562: \citep{2005AaA...431L...1M}.
563:
564: \noindent {\bf HD \phn37042}:
565: The known components to this star are all far too wide for detection here.
566: The star $\theta^2$~Ori~B is radial velocity variable according
567: to \cite{1991ApJS...75..965M}.
568:
569: \noindent {\bf HD \phn37043}:
570: Iota Ori is a close speckle pair of great interest as, despite its very close
571: separation, it was deemed optical in \cite{2004MNRAS.350..615G}.
572: The speckle pair was last measured very close to the FWHM value, so it is
573: possible that the component was just barely too close to be detected.
574:
575: \noindent {\bf HD \phn37366}:
576: This star is a double-lined spectroscopic binary \citep{2007ApJ...664.1121B}.
577:
578: \noindent {\bf HD \phn39680}:
579: Classified as a runaway star in \cite{1986ApJS...61..419G}.
580: The known pair, S\phm{888}502, is much too wide (46\farcs1) for detection here.
581:
582: \noindent {\bf HD \phn41161}:
583: The known pair, ES\phn1234AB, was too wide (10\farcs3) for detection here.
584: Whether this pair is bound or not is not yet certain.
585:
586: \noindent {\bf HD \phn45314}:
587: CHR\phn251 was last measured closer (54 milliarcsec) than the FWHM of the PSF
588: of our observation.
589: The A component is an Oe star and a probable velocity variable
590: \citep{2007PASP..119..742B}.
591:
592: \noindent {\bf HD \phn52266}:
593: This star is a probable single-lined spectroscopic binary
594: \citep{2007ApJ...655..473M}.
595:
596: \noindent {\bf HD \phn54662}:
597: This star is a double-lined spectroscopic binary \citep{2007ApJ...664.1121B}.
598:
599: \noindent {\bf HD \phn60848}:
600: This Oe star shows emission line variations but no evidence of significant
601: velocity variability \citep{2007PASP..119..742B}.
602:
603: \noindent {\bf HD \phn69648}:
604: Not included in \cite{MasonEtAl} survey.
605:
606: \noindent {\bf HD 163892}:
607: These data were taken using an 800-900\,nm filter.
608:
609: \noindent {\bf HD 164794}:
610: This star is a probable long-period spectroscopic binary
611: \citep{2002AaA...394..993R}.
612:
613: \noindent {\bf HD 167771}:
614: The known companion, RST3170, last measured in 1940 at 8\farcs3
615: \citep{1955POMic..11....1R}, would have been outside the field of view.
616:
617: \noindent {\bf HD 175876}:
618: The known, probably optical, pair, HO\phm{88}271, would have been too wide to
619: detect.
620:
621: \noindent {\bf HD 191978}:
622: Not included in \cite{MasonEtAl} survey.
623: The star shows no evidence of velocity variability \citep{1972AJ.....77..138A}.
624:
625: \noindent {\bf HD 193443}:
626: The close pair, A\phm{88}1425AB ($\rho\sim$ 0\farcs1) may have been too close
627: to resolve, while the wider AC pair ($\rho\sim$ 9\farcs1) may have been outside
628: the field of view.
629:
630: \noindent {\bf HD 195592}:
631: This is a probable single-lined spectroscopic binary
632: \citep{2007ApJ...655..473M}.
633:
634: \noindent {\bf HD 216898}:
635: Not included in \cite{MasonEtAl} survey.
636:
637: \subsection{2MASS Data Mining Confirmations}
638:
639: Searches were made for 2MASS \citep{2MASS} companions to all stars on our
640: observing list, using Vizier and Aladin, as well as the `data mining'
641: technique of \cite{2006AJ....132...50W}.
642: Results are given in Table~\ref{tbl:2mass} and discussed below.
643:
644: It should first be noted that the separation/$\triangle$m regimes covered by
645: these two techniques have rather limited overlap.
646: The ``plate scale'' of the AO detector nominally limits us to companions within
647: $\sim5''$ of the primary, although this outer limit may extend to as wide as
648: 7\farcs5, depending on orientation of the pair relative to the detector, as
649: well as slight off-center placement of the primary.
650:
651: In order to get an idea of the separation/$\triangle$m limits of 2MASS, the
652: point-source catalog was searched for sources in the magnitude range 5.5
653: $\le J \le$ 8.0, corresponding to the approximate 2MASS $J$-magnitude range
654: for the AO targets in this project.
655: This yielded 99,656 sources.
656: All sources within 10$''$ of these ``primaries'' were then extracted from the
657: catalog, yielding 9,657 companions.
658: A plot of separation versus $\triangle J$ is shown in Figure~\ref{fig:jplot2}.
659: As seen in this figure, very few close pairs are detected (only 17 pairs with
660: separations in the range $1'' < \rho < 2''$ and an additional 70 in
661: the range $2'' < \rho < 3''$).
662: For the main body of companions, maximum $\triangle J$ increases to perhaps
663: 6-6.5 mag at 3$''$, 8-8.8 mag at 4$''$, and 9.5-10 mag at 7\farcs5.
664:
665: \input{jplot2_figure}
666:
667: \noindent {\it AO Single Stars}:
668: A Vizier search of the regions around each of the AO single stars found only
669: six 2MASS companions within 7\farcs5 of their respective primaries: HD~25638
670: (6\farcs2), HD~52266 (7\farcs1), HD~151515 (6\farcs2), HD~152219 (7\farcs2),
671: HD~163892 (6\farcs4), and HD~210809 (5\farcs6).
672: A further check of their orientations showed that all of these companions fell
673: either outside the observing window or so close to the edge of the field as to
674: not be measurable.
675: The companion to HD~25638 is known; see \S~\ref{unresolved_notes}.
676:
677: \noindent {\it Known Pairs}:
678: A search of the regions around these primaries found six 2MASS companions
679: within 7\farcs5.
680: Two of these (HD~48279 and 152408) correspond to known companions also measured
681: by AO and are listed in Table~\ref{tbl:2mass}, which lists relative astrometry
682: and differential photometry in the 2MASS $J$, $H$, and $K_\text{S}$ bands.
683: The remaining four, including one known companion and three objects much wider
684: than the known secondaries, all fall outside the AO observing window.
685: Note that the companion to HD~152408 is very red
686: ($\triangle m_{I} > \triangle m_{K_\text{S}}$), perhaps indicating that the
687: companion still possesses a luminous disk (like the case of the IR-bright
688: companion of $\sigma$~Ori~AB).
689:
690: \noindent {\it New Pairs}:
691: The search around these objects found five 2MASS companions within
692: 7$''$\llap.5.
693: Two of these (HD~156212 and 201345) correspond to the newly discovered AO
694: companions and are listed in Table~\ref{tbl:2mass}.
695: The other three --- two of which are known companions --- are all wide.
696:
697: \section{DISCUSSION}
698: \label{sec:discussion}
699:
700: Our AO survey of the O-stars has revealed a large number of new and generally
701: faint companions.
702: We show in Figure~\ref{fig:hdi} the distribution of numbers of companions
703: as a function of magnitude difference $\triangle I$ (in 1 mag bins), and the
704: distribution peaks near $\triangle I = 10$~mag.
705: It is certainly possible that some of these faint companions are low mass,
706: gravitationally bound stars.
707: For example, \citet{1997AJ....113.1733H} made a deep $I$-band survey of the
708: Orion cluster centered on the Orion Trapezium, and she found examples of still
709: embedded pre-main sequence stars with magnitude differences as large as
710: $\triangle I = 15$~mag compared to the bright O-stars in the Trapezium.
711: On the other hand, her survey also demonstrates that there will be many cluster
712: stars that appear projected on the sky at positions near O-stars but are not
713: orbitally bound to them.
714:
715: \input{hdi_figure}
716:
717: We used the $I$-band star counts from \citet{1997AJ....113.1733H} to model the
718: possible confusion limit caused by nearby cluster stars that we might expect to
719: find in our observations.
720: The cumulative number distribution of stars brighter than $I$ in her sample
721: (1551 stars with measured $I$ magnitudes) is well matched by a power law,
722: %
723: \begin{equation}
724: N(<I) = 0.063~ 10^{\gamma I},
725: \end{equation}
726: %
727: where the exponent is $\gamma = 0.27$ (valid down to $I=16$ mag).
728: The total surface density of stars in the central part of cluster where the
729: O-stars reside is $10^{5.4}$ stars per square degree (see Figure~4a in
730: \citealt{1997AJ....113.1733H}) or 1.9 stars per $10\arcsec \times 10\arcsec$
731: field of view (= AEOS FOV).
732: Thus, the cumulative distribution of cluster background stars per AEOS FOV is
733: approximately given by
734: %
735: \begin{equation}
736: N(<I) = 7.8\times 10^{-5} ~ 10^{\gamma I}.
737: \end{equation}
738:
739: If we suppose that all O-stars have a color index of $V-I=-0.39$
740: \citep{1994MNRAS.270..229W}, the extinction to Orion is approximately
741: $A_I=0.25$ \citep{1997AJ....113.1733H,1999PASP..111...63F}, and the distance
742: to Orion is 470~pc \citep{1997AJ....113.1733H}, then the numbers per AEOS FOV
743: can be recast in terms of an O-star absolute magnitude $M_{V}$ and
744: magnitude difference $\triangle I$ as
745: %
746: \begin{equation}
747: N(<I) = 0.023~ 10^{\gamma (\triangle I + M_{V})}.
748: \end{equation}
749: %
750: If we further assume that the same relationship holds for other O-stars in
751: more distant clusters, then the number of cluster stars appearing in the FOV
752: will scale as distance $d$ squared.
753: Let $\eta$ represent the predicted number of cluster background stars per AEOS
754: FOV.
755: The relation between the limiting magnitude difference $\triangle I$
756: associated with $\eta$ is then
757: %
758: \begin{equation}
759: \triangle I = 3.6 -{2\over\gamma}\log d - M_{V}+{1\over\gamma}\log \eta.
760: \label{eq:eta}
761: \end{equation}
762: %
763: Thus, we expect that background confusion with cluster stars will increase
764: (smaller $\triangle I$) with increasing distance and with lower luminosity
765: O-stars.
766:
767: We applied this relation to estimate a limiting $\triangle I$ for each cluster
768: and association star in our sample in order to separate the probable physical
769: companions from the faint optical companions caused by background confusion.
770: We used distances from \citet{MasonEtAl} and the spectral type -- absolute
771: magnitude calibration from \citet{2005AaA...436.1049M}.
772: The Orion cluster richness, extinction, and age may not be representative of
773: the full sample of cluster O-stars in our survey, and so the normalization of
774: the $\triangle I$ relation is somewhat uncertain.
775: We note that large numbers of companions begin to appear near
776: $\triangle I = 8$ mag (see Figure~\ref{fig:hdi}), and for typical O-star
777: distances ($d=1$ to 2 kpc) and magnitudes ($M_{V}= -4$ to $-6$) this
778: corresponds to a background density in the Orion model of $\eta = 1$ star per
779: AEOS FOV.
780: Thus, we adopted this value of $\eta$ in equation~\ref{eq:eta} to identify the
781: probable physical companions among the cluster and association stars.
782:
783: The background limits for the field and runaway stars depend instead upon on
784: the stellar number density of Galactic background stars as a function of
785: magnitude and the direction of a given target.
786: Here we adopted the cumulative star counts as a function of apparent $I$
787: magnitude and Galactic coordinates from the Besan\c{c}on model of the Galaxy
788: \citep{2003AaA...409..523R}.
789: We created a cumulative star counts versus $\triangle I$ relation for each of
790: these stars and found a limiting magnitude for a specified value of 0.01 stars
791: per AEOS FOV.
792: We set this conservative limit based on numerical experiments for the set of
793: runaway stars with observed faint companions.
794: None of these companions can be physical companions since any such wide binary
795: would be disrupted at the time the runaway star was ejected (by close
796: gravitational encounters or a supernova in a binary).
797:
798: The results of this exercise to discern the probable physical companions are
799: summarized in Table \ref{tbl:sum}.
800: The targets were placed in categories of cluster and association stars, field
801: stars, and runaway stars according to the criteria adopted by \citet{MasonEtAl}
802: (with some revisions from the subsequent work on field O-stars by
803: \citealt{2004AaA...425..937D, 2005AaA...437..247D}).
804: We see that after rejection of the possible background stars there is still a
805: high frequency ($37\%$) of detected companions among the cluster and
806: association stars in our survey.
807: In sharp contrast, there are very few such companions among the field and
808: runaway stars.
809: The only field star with a probable companion is HD~48279, and
810: \citet{2004AaA...425..937D} show that this star is $28\arcsec$ from a cluster
811: identified in 2MASS images, which suggests that this O-star may be a member of
812: a young, emerging cluster.
813: The only runaway star with a candidate companion is HD~34078 = AE~Aur, a star
814: ejected from the Orion association \citep{2004MNRAS.350..615G}.
815: There is another optical companion, SEI~136, at a separation of $8\farcs4$
816: \citep{MasonEtAl}, and we suspect that both the companion we found and SEI~136
817: form a visual system that appears by a chance alignment with AE~Aur at this
818: time in its trajectory across the sky.
819:
820: Our results confirm the trends found by \citet{MasonEtAl} that binary O-stars
821: are common among O-stars found near their birthplaces and are rare among
822: O-stars ejected from clusters.
823: These trends are consistent with current models for the ejection of runaways
824: and the formation of massive stars \citep{2007ARAaA..45..481Z}.
825: High velocity O-stars were probably ejected by close encounters between
826: binaries in dense clusters and by supernova explosions in close binaries
827: \citep{2001AaA...365...49H}, and their ejection velocities generally exceed the
828: escape velocity binding any wide, orbiting companion.
829: Many massive stars are probably born in high density stellar environments where
830: binaries may be formed through three-body interactions
831: \citep{2002MNRAS.336..705B} and through the interactions with large
832: proto-stellar accretion disks \citep{2007ApJ...661L.183M}.
833: Even in lower density environments, binaries may form through disk
834: fragmentation and subsequent gas accretion \citep{2007arXiv0709.4252K}.
835: Our results add to growing body of evidence that the formation of binaries is
836: closely linked to the formation of massive stars and the deposition of the
837: angular momentum reservoir of the natal cloud.
838:
839: \acknowledgements
840:
841: The United States Air Force provided the telescope time, on-site support,
842: and 80\% of the research funds for this Air Force Office of Scientific
843: Research and National Science Foundation (NSF) jointly sponsored research
844: under NSF grant number AST-0088498.
845: Additional support was provided by NSF grants AST-0506573 and AST-0606861.
846: LCR was funded by AFRL/DE (Contract Number F29601-00-D-0204).
847: We thank the numerous staff members of the Maui Space Surveillance System who
848: helped make this data possible.
849: Thanks as well to Gary Wycoff of USNO, who `mined' the 2MASS Catalog for
850: additional photometry and astrometry of these systems.
851: This research has made use of the {\it Washington Double Star Catalog},
852: maintained at the U.S. Naval Observatory, and of Aladin, Simbad, and Vizier,
853: operated at CDS, Strasbourg, France.
854: This publication also makes use of data products from the Two Micron All Sky
855: Survey, which is a joint project of the University of Massachusetts and
856: the Infrared Processing and Analysis Center/California Institute of Technology,
857: funded by the National Aeronautics and Space Administration and the National
858: Science Foundation.
859:
860: \input{merged_measures_table}
861:
862: \input{orbital_elements_table}
863:
864: \input{ephemerides_table}
865:
866: \input{single_stars_table}
867:
868: \input{data_mining_results_table}
869:
870: \input{summary_table}
871:
872: \newcommand{\SortNoop}[1]{}
873: \begin{thebibliography}{}
874:
875: \bibitem[Abt et~al., 1972]{1972AJ.....77..138A}
876: Abt, H.~A., Levy, S.~G., \& Gandet, T.~L. 1972,
877: \newblock \aj, 77, 138
878:
879: \bibitem[Africano et~al., 1975]{1975AJ.....80..689A}
880: Africano, J.~L., Cobb, C.~L., Dunham, D.~W., Evans, D.~S., Fekel, F.~C., \&
881: Vogt, S.~S. 1975,
882: \newblock \aj, 80, 689
883:
884: \bibitem[Bate et~al., 2002]{2002MNRAS.336..705B}
885: Bate, M.~R., Bonnell, I.~A., \& Bromm, V. 2002,
886: \newblock \mnras, 336, 705
887:
888: \bibitem[Bonnell \& Bate, 2005]{2005MNRAS.362..915B}
889: Bonnell, I.~A. \& Bate, M.~R. 2005,
890: \newblock \mnras, 362, 915
891:
892: \bibitem[{\SortNoop{Bos}}{van den Bos}, 1928]{vandenBos1928}
893: {\SortNoop{Bos}}{van den Bos}, W.~H. 1928,
894: \newblock Ann. Leiden Obs., 14,
895: \newblock pt. 4
896:
897: \bibitem[Boyajian et~al., 2005]{2005ApJ...621..978B}
898: Boyajian, T.~S. et~al. 2005,
899: \newblock \apj, 621, 978
900:
901: \bibitem[Boyajian et~al., 2007a]{2007PASP..119..742B}
902: Boyajian, T.~S. et~al. 2007a,
903: \newblock \pasp, 119, 742
904:
905: \bibitem[Boyajian et~al., 2007b]{2007ApJ...664.1121B}
906: Boyajian, T.~S. et~al. 2007b,
907: \newblock \apj, 664, 1121
908:
909: \bibitem[{\SortNoop{Brummelaar}}{ten Brummelaar} et~al., 1998]{spie3353-391}
910: {\SortNoop{Brummelaar}}{ten Brummelaar}, T.~A., Hartkopf, W.~I., McAlister,
911: H.~A., Mason, B.~D., Roberts, Jr., L.~C., \& Turner, N.~H. 1998,
912: \newblock in \procspie, Adaptive Optical System Technologies, ed. D.~Bonaccini
913: \& R.~K. Tyson (Bellingham, WA: SPIE), 3353, 391
914:
915: \bibitem[{\SortNoop{Brummelaar}}{ten Brummelaar} et~al., 1996]{diffmag1}
916: {\SortNoop{Brummelaar}}{ten Brummelaar}, T.~A., Mason, B.~D., Bagnuolo, Jr.,
917: W.~G., Hartkopf, W.~I., McAlister, H.~A., \& Turner, N.~H. 1996,
918: \newblock \aj, 112, 1180
919:
920: \bibitem[{\SortNoop{Brummelaar}}{ten Brummelaar} et~al., 2000]{diffmag2}
921: {\SortNoop{Brummelaar}}{ten Brummelaar}, T.~A., Mason, B.~D., McAlister, H.~A.,
922: Roberts, Jr., L.~C., Turner, N.~H., Hartkopf, W.~I., \& Bagnuolo, Jr., W.~G.
923: 2000,
924: \newblock \aj, 119, 2403
925:
926: \bibitem[Caballero, 2007]{2007AaA...466..917C}
927: Caballero, J.~A. 2007,
928: \newblock \aap, 466, 917
929:
930: \bibitem[Caballero, 2008]{2008MNRAS.383..750C}
931: Caballero, J.~A. 2008,
932: \newblock \mnras, 383, 750
933:
934: \bibitem[Cox, 2001]{APQ2001}
935: Cox, A.~N. 2001,
936: \newblock Allen's {A}strophysical {Q}uantities (4th ed.; New York, NY:
937: Springer)
938:
939: \bibitem[Cutri et~al., 2003]{2MASS}
940: Cutri, R.~M. et~al. 2003,
941: \newblock 2{MASS} {A}ll {S}ky {C}atalog of {P}oint {S}ources (Pasadena, CA:
942: NASA/IPAC)
943:
944: \bibitem[Fabricius \& Makarov, 2000]{2000AaA...356..141F}
945: Fabricius, C. \& Makarov, V.~V. 2000,
946: \newblock \aap, 356, 141
947:
948: \bibitem[Fitzpatrick, 1999]{1999PASP..111...63F}
949: Fitzpatrick, E.~L. 1999,
950: \newblock \pasp, 111, 63
951:
952: \bibitem[Gies \& Bolton, 1986]{1986ApJS...61..419G}
953: Gies, D.~R. \& Bolton, C.~T. 1986,
954: \newblock \apjs, 61, 419
955:
956: \bibitem[Gualandris et~al., 2004]{2004MNRAS.350..615G}
957: Gualandris, A., {Portegies Zwart}, S., \& Eggleton, P.~P. 2004,
958: \newblock \mnras, 350, 615
959:
960: \bibitem[Hartkopf et~al., 1996]{1996AJ....111..370H}
961: Hartkopf, W.~I., Mason, B.~D., \& McAlister, H.~A. 1996,
962: \newblock \aj, 111, 370
963:
964: \bibitem[Hartkopf et~al., 2001]{2001AJ....122.3472H}
965: Hartkopf, W.~I., Mason, B.~D., \& Worley, C.~E. 2001,
966: \newblock \aj, 122, 3472
967:
968: \bibitem[Hillenbrand, 1997]{1997AJ....113.1733H}
969: Hillenbrand, L.~A. 1997,
970: \newblock \aj, 113, 1733
971:
972: \bibitem[Hillwig et~al., 2006]{2006ApJ...639.1069H}
973: Hillwig, T.~C., Gies, D.~R., Bagnuolo, Jr., W.~G., Huang, W., McSwain, M.~V.,
974: \& Wingert, D.~W. 2006,
975: \newblock \apj, 639, 1069
976:
977: \bibitem[Hinkley et~al., 2007]{2007ApJ...654..633H}
978: Hinkley, S. et~al. 2007,
979: \newblock \apj, 654, 633
980:
981: \bibitem[Hoogerwerf et~al., 2001]{2001AaA...365...49H}
982: Hoogerwerf, R., {\SortNoop{Bruijne}}{de Bruijne}, J. H.~J., \&
983: {\SortNoop{Zeeuw}}{de Zeeuw}, P.~T. 2001,
984: \newblock \aap, 365, 49
985:
986: \bibitem[Hopmann, 1967]{1967MiWie..13...49H}
987: Hopmann, J. 1967,
988: \newblock Mitteilungen der Universitaets-Sternwarte Wien, 13, 49
989:
990: \bibitem[Hummel et~al., 2000]{2000ApJ...540L..91H}
991: Hummel, C.~A., White, N.~M., Elias, II, N.~M., Hajian, A.~R., \& Nordgren,
992: T.~E. 2000,
993: \newblock \apjl, 540, L91
994:
995: \bibitem[Kratter et~al., 2007]{2007arXiv0709.4252K}
996: Kratter, K.~M., Matzner, C.~D., \& Krumholz, M.~R. 2007,
997: \newblock ArXiv e-prints, 0709.4252
998:
999: \bibitem[Kraus et~al., 2007]{2007AaA...466..649K}
1000: Kraus, S. et~al. 2007,
1001: \newblock \aap, 466, 649
1002:
1003: \bibitem[{\SortNoop{Loon}}{van Loon} \& Oliveira, 2003]{2003AaA...405L..33V}
1004: {\SortNoop{Loon}}{van Loon}, J.~T. \& Oliveira, J.~M. 2003,
1005: \newblock \aap, 405, L33
1006:
1007: \bibitem[Ma\'{\i}z-Apell\'{a}niz et~al., 2004]{GOS}
1008: Ma\'{\i}z-Apell\'{a}niz, J., Walborn, N.~R., Galu\'{e}, H.~A., \& Wei, L. 2004,
1009: \newblock \apjs, 151, 103
1010:
1011: \bibitem[Makidon et~al., 2005]{2005PASP..117..831M}
1012: Makidon, R.~B., Sivaramakrishnan, A., Perrin, M.~D., Roberts~Jr., L.~C.,
1013: Oppenheimer, B.~R., Soummer, R., \& Graham, J.~R. 2005,
1014: \newblock \pasp, 117, 831
1015:
1016: \bibitem[Martins et~al., 2005]{2005AaA...436.1049M}
1017: Martins, F., Schaerer, D., \& Hillier, D.~J. 2005,
1018: \newblock \aap, 436, 1049
1019:
1020: \bibitem[Mason et~al., 1998]{MasonEtAl}
1021: Mason, B.~D., Hartkopf, W.~I., Gies, D.~R., Bagnuolo, Jr., W.~G.,
1022: {\SortNoop{Brummelaar}}{ten Brummelaar}, T.~A., \& McAlister, H.~A. 1998,
1023: \newblock \aj, 115, 821
1024:
1025: \bibitem[Mason et~al., 2001a]{2001AJ....121.3224M}
1026: Mason, B.~D., Hartkopf, W.~I., Holdenried, E.~R., \& Rafferty, T.~J. 2001a,
1027: \newblock \aj, 121, 3224
1028:
1029: \bibitem[Mason et~al., 2004]{2004AJ....127..539M}
1030: Mason, B.~D. et~al. 2004,
1031: \newblock \aj, 127, 539
1032:
1033: \bibitem[Mason et~al., 2001b]{WDS2001}
1034: Mason, B.~D., Wycoff, G.~L., Hartkopf, W.~I., Douglass, G.~G., \& Worley, C.~E.
1035: 2001b,
1036: \newblock \aj, 122, 3466
1037:
1038: \bibitem[Massey et~al., 1995]{1995AJ....110.2715M}
1039: Massey, P., Armandroff, T.~E., Pyke, R., Patel, K., \& Wilson, C.~D. 1995,
1040: \newblock \aj, 110, 2715
1041:
1042: \bibitem[McSwain, 2003]{2003ApJ...595.1124M}
1043: McSwain, M.~V. 2003,
1044: \newblock \apj, 595, 1124
1045:
1046: \bibitem[McSwain et~al., 2007]{2007ApJ...655..473M}
1047: McSwain, M.~V., Boyajian, T.~S., Grundstrom, E.~D., \& Gies, D.~R. 2007,
1048: \newblock \apj, 655, 473
1049:
1050: \bibitem[Mdzinarishvili \& Chargeishvili, 2005]{2005AaA...431L...1M}
1051: Mdzinarishvili, T.~G. \& Chargeishvili, K.~B. 2005,
1052: \newblock \aap, 431, L1
1053:
1054: \bibitem[Moeckel \& Bally, 2007]{2007ApJ...661L.183M}
1055: Moeckel, N. \& Bally, J. 2007,
1056: \newblock \apjl, 661, L183
1057:
1058: \bibitem[Morrell \& Levato, 1991]{1991ApJS...75..965M}
1059: Morrell, N. \& Levato, H. 1991,
1060: \newblock \apjs, 75, 965
1061:
1062: \bibitem[Otero, 2007]{2007OEJV...72....1O}
1063: Otero, S.~A. 2007,
1064: \newblock Open European Journal on Variable Stars, 72, 1
1065:
1066: \bibitem[Perryman \& ESA, 1997]{1997hity.book.....P}
1067: Perryman, M. A.~C. \& ESA 1997,
1068: \newblock The {HIPPARCOS} and {TYCHO} {C}atalogues. {A}strometric and
1069: {P}hotometric {S}tar {C}atalogues {D}erived from the {ESA} {HIPPARCOS}
1070: {S}pace {A}strometry {M}ission, Vol. 1200 of ESA SP Series (Noordwijk, The
1071: Netherlands: ESA Pub. Div.)
1072:
1073: \bibitem[Pozzo et~al., 2000]{2000MNRAS.313L..23P}
1074: Pozzo, M., Jeffries, R.~D., Naylor, T., Totten, E.~J., Harmer, S., \& Kenyon,
1075: M. 2000,
1076: \newblock \mnras, 313, L23
1077:
1078: \bibitem[Rauw et~al., 2002]{2002AaA...394..993R}
1079: Rauw, G. et~al. 2002,
1080: \newblock \aap, 394, 993
1081:
1082: \bibitem[Roberts, 2001]{amos2001}
1083: Roberts, Jr., L.~C. 2001,
1084: \newblock in Proc. 2001 AMOS Technical Conference, ed. P.~Kervin, L.~Bragg, \&
1085: S.~Ryan (Maui, HI: Maui Econ. Devel. Board), 326
1086:
1087: \bibitem[Roberts \& Neyman, 2002]{RobertsNeyman}
1088: Roberts, Jr., L.~C. \& Neyman, C.~R. 2002,
1089: \newblock \pasp, 114, 1260
1090:
1091: \bibitem[Roberts et~al., 2007]{2007AJ....133..545R}
1092: Roberts, Jr., L.~C., Turner, N.~H., \& {\SortNoop{Brummelaar}}{ten Brummelaar},
1093: T.~A. 2007,
1094: \newblock \aj, 133, 545
1095:
1096: \bibitem[Robin et~al., 2003]{2003AaA...409..523R}
1097: Robin, A.~C., Reyl{\'e}, C., Derri{\`e}re, S., \& Picaud, S. 2003,
1098: \newblock \aap, 409, 523
1099:
1100: \bibitem[Rossiter, 1955]{1955POMic..11....1R}
1101: Rossiter, R.~A. 1955,
1102: \newblock Publications of Michigan Observatory, 11, 1
1103:
1104: \bibitem[See, 1898]{1898AJ.....18..181S}
1105: See, T. J.~J. 1898,
1106: \newblock \aj, 18, 181
1107:
1108: \bibitem[Seymour et~al., 2002]{2002AJ....123.1023S}
1109: Seymour, D.~M., Mason, B.~D., Hartkopf, W.~I., \& Wycoff, G.~L. 2002,
1110: \newblock \aj, 123, 1023
1111:
1112: \bibitem[Silbernagel, 1931]{1931AN....241...33S}
1113: Silbernagel, E. 1931,
1114: \newblock Astronomische Nachrichten, 241, 33
1115:
1116: \bibitem[Struve, 1837]{1837AN.....14..249S}
1117: Struve, F. G.~W. 1837,
1118: \newblock Astronomische Nachrichten, 14, 249
1119:
1120: \bibitem[Wallenquist, 1934]{Wallenquist1934}
1121: Wallenquist, A. 1934,
1122: \newblock Ann. Bosscha Obs. Lembang, 6,
1123: \newblock pt. 2
1124:
1125: \bibitem[Wegner, 1994]{1994MNRAS.270..229W}
1126: Wegner, W. 1994,
1127: \newblock \mnras, 270, 229
1128:
1129: \bibitem[{\SortNoop{Wit}}{de Wit} et~al., 2004]{2004AaA...425..937D}
1130: {\SortNoop{Wit}}{de Wit}, W.~J., Testi, L., Palla, F., Vanzi, L., \& Zinnecker,
1131: H. 2004,
1132: \newblock \aap, 425, 937
1133:
1134: \bibitem[{\SortNoop{Wit}}{de Wit} et~al., 2005]{2005AaA...437..247D}
1135: {\SortNoop{Wit}}{de Wit}, W.~J., Testi, L., Palla, F., \& Zinnecker, H. 2005,
1136: \newblock \aap, 437, 247
1137:
1138: \bibitem[Worley, 1990]{1990ebua.conf..419W}
1139: Worley, C.~E. 1990,
1140: \newblock in Errors, Bias and Uncertainties in Astronomy, ed. C.~Jaschek \&
1141: F.~Murtagh (Cambridge: Cambridge Univ. Press), 419
1142:
1143: \bibitem[Wycoff et~al., 2006]{2006AJ....132...50W}
1144: Wycoff, G.~L., Mason, B.~D., \& Urban, S.~E. 2006,
1145: \newblock \aj, 132, 50
1146:
1147: \bibitem[Zinnecker \& Yorke, 2007]{2007ARAaA..45..481Z}
1148: Zinnecker, H. \& Yorke, H.~W. 2007,
1149: \newblock \araa, 45, 481
1150:
1151: \end{thebibliography}
1152:
1153: \end{document}
1154: