1: %\documentclass[11pt,preprint]{aastex} % 1-col, single-spaced
2: \documentclass[preprint2]{aastex} % dbl-col, single-spaced
3: %\documentclass[manuscript]{aastex} %! 1-col, dbl-spaced
4:
5: \usepackage{longtable}
6: \usepackage{lscape}
7:
8: \renewcommand{\arraystretch}{.6}
9:
10: \newcommand\LSIV{LS\,IV$-08^\circ 3$}
11: %
12: \newcommand\JKs{\ensuremath{J\!-\!K_S}}
13: \newcommand\HKs{\ensuremath{H\!-\!K_S}}
14: \newcommand\JH{\ensuremath{J\!-\!H}}
15: %
16: \newcommand\CIII{\ion{C}{3}}
17: \newcommand\NIII{\ion{N}{3}}
18: \newcommand\HI{\ion{H}{1}}
19: \newcommand\HeI{\ion{He}{1}}
20: \newcommand\HeII{\ion{He}{2}}
21: \newcommand\Ha{H$\alpha$}
22: \newcommand\Hb{H$\beta$}
23: \newcommand\NaI{\ion{Na}{1}}
24: \newcommand\OI{\ion{O}{1}}
25:
26: \slugcomment{to be published in Astronomical Journal}
27:
28: \shorttitle{A New Bright Emission Line Binary}
29: \shortauthors{Stark et.\ al.}
30:
31: \hyphenation{sub-dwarf sub-dwarfs}
32:
33: \begin{document}
34:
35: \title{A New, Bright, Short-Period, Emission Line Binary in
36: Ophiuchus\footnote{Based in part on observations obtained with the
37: Hobby-Eberly Telescope, which is a joint project of the University of
38: Texas at Austin, the Pennsylvania State University, Stanford
39: University, Ludwig-Maximillians-Universit\"at M\"unchen, and
40: Georg-August-Universit\"at G\"ottingen.}}
41:
42: \author{M.\ A.\ Stark\footnote{Visiting Astronomer, Kitt Peak National
43: Observatory, National Optical Astronomy Observatory, which is operated
44: by the Association of Universities for Research in Astronomy,
45: Inc. (AURA) under cooperative agreement with the National Science
46: Foundation.}}
47: \affil{Department of Physics and Astronomy, University of Wyoming,
48: Dept.\ 3905, 1000 E.\ University Ave.,\ Laramie, WY 82071}
49: \email{stark@uwyo.edu}
50:
51: \author{Richard A.\ Wade}
52: \affil{Department of Astronomy and Astrophysics, The Pennsylvania
53: State University, 525 Davey Lab, University Park, PA 16802}
54: \email{wade@astro.psu.edu}
55:
56: \author{John R.\ Thorstensen, Christopher S.\ Peters}
57: \affil{Department of Physics and Astronomy,
58: 6127 Wilder Laboratory, Dartmouth College,
59: Hanover, NH 03755-3528}
60:
61: \author{Horace A.\ Smith, Robert D.\ Miller}
62: \affil{Department of Physics and Astronomy,
63: Bio-Physical Sciences Bldg, Michigan State University,
64: East Lansing, MI 48824-1116}
65:
66: \author{E.\ M.\ Green} \affil{Steward Observatory, University of
67: Arizona, Tucson, AZ 85721-0065}
68:
69: \begin{abstract}
70: The 11th magnitude star {\LSIV} has been classified previously as an
71: OB star in the Luminous Stars survey, or alternatively as a hot
72: subdwarf. It is actually a binary star. We present spectroscopy,
73: spectroscopic orbital elements, and time series photometry, from
74: observations made at the Kitt Peak National Observatory 2.1m, Steward
75: Observatory 2.3m, MDM Observatory 1.3m and 2.4m, Hobby-Eberly 9.2m,
76: and Michigan State University 0.6m telescopes. The star exhibits
77: emission of varying strength in the cores of H and {\HeI} absorption
78: lines. Emission is also present at 4686\,\AA\ ({\HeII}) and near
79: 4640/4650\,\AA\ ({\NIII}/{\CIII}). Time-series spectroscopy collected
80: from 2005 July to 2007 June shows coherent, periodic radial velocity
81: variations of the {\Ha} line, which we interpret as orbital motion
82: with a period of 0.1952894(10)\,days. High-resolution spectra show
83: that there are two emission components, one broad and one narrow,
84: moving in antiphase, as might arise from an accretion disk and the
85: irradiated face of the mass donor star. Less coherent, low-amplitude
86: photometric variability is also present on a timescale similar to the
87: orbital period. Diffuse interstellar bands indicate considerable
88: reddening, which however is consistent with a distance of
89: $\sim$100--200 pc. The star is the likely counterpart of a weak {\em
90: ROSAT} X-ray source, whose properties are consistent with accretion in
91: a cataclysmic variable (CV) binary system. We classify {\LSIV} as a
92: new member of the UX UMa subclass of CV stars.
93: \end{abstract}
94:
95: \keywords{
96: binaries: close ---
97: novae, cataclysmic variables ---
98: stars: emission-line ---
99: stars: individual ({\LSIV}) ---
100: stars: variables: other
101: }
102:
103: %----------------------------------------------------------------------
104:
105: %Section 1
106: \section{Introduction} \label{sec:Intro}
107:
108: The star {\LSIV} has $V = 11.475$, $(\bv) = +0.190$ {\citep{RN00}},
109: and was first classified ``OB'' in the {\em Luminous Stars in the
110: Northern Milky Way} catalog {\citep{LSIV}}. Later it was reclassified
111: ``sdB'' by {\citet{KB92}} based on Reticon spectrograms and
112: Str\"{o}mgren photometry. Additional photometry of this object is
113: available in the literature {\citep{RCL,Reed,RN00}}, but no additional
114: spectroscopic observations have been reported. Astrometric and
115: photometric data for this star, collected from the literature,
116: are reported in Table~\ref{table:data}.
117:
118: We obtained spectra of {\LSIV} as part of our program on composite
119: spectrum hot subdwarf stars {\citep{SW,SW04,Thesis,SW06}}, owing to its
120: unusually red $V-K_s$ and $J-K_s$ colors as observed in the {\em Two
121: Micron All-Sky Survey} {\citep[2MASS;][]{Skrutskie06}}. We found that
122: {\LSIV} exhibits broad absorption lines with core emission, which is
123: variable in both strength and profile shape. We obtained time-series
124: photometry and additional spectroscopy in order to correctly classify
125: this star.
126:
127: %----------------------------------------------------------------------
128:
129: %section 2
130: \section{Observations} \label{sec:Observations and Findings}
131:
132: \subsection{Low-Dispersion Spectra}\label{sec:Low-disp}
133:
134: \subsubsection{Kitt Peak National Observatory 2.1m Telescope}\label{sec:GoldCam}
135:
136: {\LSIV} was observed from the Kitt Peak National Observatory (KPNO)
137: 2.1m telescope using the GoldCam spectrograph during 2003 June and
138: September. A journal of observations is given in
139: Table~\ref{table:GoldCamObs}. All exposures were between 5 and 20
140: minutes long. Two different spectrograph configurations were used,
141: which covered approximately 4600--6750\,\AA\ and 6400--8900\,\AA\ at
142: $\sim$1.3\,\AA~pixel$^{-1}$. The features observed in the GoldCam
143: spectra include broad absorption lines with emission cores in {\Ha},
144: {\Hb}, and {\HeI} 5875\,\AA; emission from {\HeI} 6678\,\AA, {\HeII}
145: 4685\,\AA, and {\NIII}/{\CIII} $\sim$4640/4650\,\AA. The emission
146: components were observed to vary in both strength and profile in as
147: little as $\sim$30 minutes (Fig.~\ref{fig:GoldCamspec}; see also
148: \S\ref{sec:HRS}).
149:
150: All GoldCam spectra were processed using standard IRAF\footnote{IRAF
151: is distributed by the National Optical Astronomy Observatories, which
152: are operated by the Association of Universities for Research in
153: Astronomy, Inc., under cooperative agreement with the National Science
154: Foundation.} routines. Unfavorable observing conditions prevented
155: reliable flux calibrations on most nights, so we discuss only
156: the normalized spectra. The continua were fitted by hand using
157: splines. The equivalent widths (EWs) of features of interest were
158: measured in each spectrum; the wavelength interval encompassed by each
159: feature was determined by eye, noting where the line deviated from the
160: continuum. For lines with broad absorption wings and core emission,
161: two EW measurements were made. The first is the total EW of the whole
162: line (absorption and emission included). The second EW refers to just
163: the emission component, measured above the point where the emission
164: emerges from the absorption wings, giving a lower limit for the
165: ``true'' emission EW. Reported in Table~\ref{table:EWLSIV} are the
166: average EWs and their formal errors (not including any systematic
167: errors or accounting for variability). Also reported are the number of
168: measurements, their standard deviation, minimum and maximum values,
169: and the difference between the maximum and minimum. (Note that EW$>0$
170: indicates absorption, while EW$<0$ indicates emission.)
171:
172: Significant reddening of this object is evidenced by strong
173: interstellar {\NaI}~D lines and several diffuse interstellar bands
174: (DIBs). {\citet*{Schlegel}} report total galactic reddening of
175: $E(\bv) \approx +0.48$ along the line of sight to {\LSIV}. The
176: reddening and distance are discussed in \S\ref{sec:reddening}.
177:
178: \subsubsection{Steward Observatory 2.3m Telescope} \label{sec:Steward}
179:
180: A flux-calibrated spectrum of {\LSIV}, covering 3620--6895\,\AA\ at
181: $\sim$3\,\AA~pixel$^{-1}$ dispersion ($\sim$9\,\AA\ resolution), was
182: obtained on UT 2002 April 3 at the Bok 2.3m telescope on Kitt Peak,
183: Arizona, using the B\&C spectrograph with a thinned, back-illuminated
184: 1200 $\times$ 800 Loral CCD. This ``pre-discovery'' spectrum is shown
185: in Figure~\ref{fig:betsyflux2002}. It extends to
186: shorter wavelengths then the GoldCam observations, including the Balmer
187: lines up to the series limit. A second spectrum was obtained with the
188: same instrument setup on UT 2007 May 8 and was flux-calibrated in a
189: homogeneous way with the first; the overall spectral energy
190: distribution is very similar, with the flux level differing by about
191: 3\%. The H emission lines were weaker in 2007, with the emission
192: reversal in H$\gamma$ virtually disappearing and the height of {\Ha}
193: emission above the continuum diminishing by a factor of 2.
194: \citet{KB92} classified {\LSIV} as ``sdB'' based on {\em uvby}
195: photometry and 2\,\AA\ resolution Reticon spectroscopy covering
196: 3900--4600\,\AA, so {\Hb} was not observed, and weak emission in the
197: absorption core of H$\gamma$ might easily have been overlooked or even
198: absent. The spectral energy distribution (SED) is quite blue, even
199: despite the reddening evidenced by the DIBs. The SED is further
200: discussed in \S\ref{sec:Discussion}.
201:
202: %------------------------------------------------------------------------
203:
204: \subsubsection{MDM Observatory 1.3m and 2.4m Telescopes} \label{sec:MDM}
205:
206: Time-series spectra of {\LSIV} were obtained at the MDM
207: Observatory\footnote{MDM Observatory is operated by a consortium
208: consisting of Dartmouth College, Columbia University, Ohio State
209: University, the University of Michigan, and Ohio University.} on Kitt
210: Peak, Arizona. Table~\ref{table:mdmjournal} gives a journal of
211: observations. Most of the MDM spectra were obtained using the Mark
212: III spectrograph and a SITe 1024$^2$ CCD detector mounted on the
213: McGraw-Hill 1.3m telescope. A 600 line mm$^{-1}$ grism gave
214: $\sim$2.3\,\AA\,pixel$^{-1}$ dispersion over the range
215: 4660--6900\,\AA, with a full width half maximum (FWHM) resolution of
216: 4.1\,\AA. All stellar exposures were bracketed with spectra of Hg,
217: Ne, and Xe comparison lamps, in order to track the spectrograph
218: flexure. (In 2004 June, an attempt was made to use the $\lambda 5577$
219: night-sky emission feature to set the wavelength zero point;
220: unfortunately, this did not work well, and those spectra are excluded
221: from the present analysis.)
222:
223: A few MDM spectra are from the Hiltner 2.4m telescope and modular
224: spectrograph, equipped with a SITe 2048$^2$ CCD; this gave 2\,\AA\
225: pixel$^{-1}$ dispersion from 4300 to 7500\,\AA, with severe vignetting
226: toward the ends of this range. The spectral resolution of this setup
227: was slightly better than that of the Mark III. Some of these
228: exposures were taken in bright twilight and bracketed with comparison
229: exposures; for those taken away from twilight, the night-sky features
230: were used to find the zero point shift. Extensive internal checks
231: show that this procedure worked well with this setup. Aliasing
232: related to daily cycle count uncertainties was suppressed by pushing
233: to large hour angles; the target's brightness made it possible to
234: obtain good signal-to-noise spectra even at large airmass.
235:
236: Flux standard stars were observed to calibrate the instrument
237: response, when the twilight sky was reasonably clear. Experience
238: suggests that the zero point of the MDM spectra should be accurate to
239: about 20\%. Exposures of bright O and B stars were also obtained and
240: used to map and correct for the telluric absorption features.
241:
242: An average flux-calibrated spectrum was constructed from the 1.3m data
243: taken in 2005 June and July. Using the IRAF {\it sbands} task and the
244: passband tabulated by \citet{bessell}, we find a synthetic $V$
245: magnitude of 11.47. The MDM 1.3m average spectrum is very similar to
246: the Steward 2.3m spectrum from 2002 April in the wavelength region
247: common to both, except the overall flux level is about 10\% higher,
248: and the emission lines are more evident (the {\Hb} emission reversal
249: protrudes slightly above the continuum; the \ion{He}{1} $\lambda$6678
250: line is in emission at about 15\% of the intensity of {\Ha}).
251:
252: \subsubsection{Radial Velocity Variations} \label{sec:RV}
253:
254: All of the MDM spectra show {\Ha} emission. To measure the radial
255: velocity (RV), we convolved the line profile with the derivative of a
256: Gaussian and took the line center to be the zero of this convolution
257: \citep{sy80}. The width of the convolving function was optimized for
258: a line with a FWHM of 11.5\,\AA, which is approximately the observed
259: line width. We searched the velocity time series for periodicities
260: using a ``residual-gram'' technique \citep{tpst}, which is especially
261: effective when the modulation is accurately sinusoidal. Because the
262: data span hundreds of days, we were careful to use a sufficiently fine
263: grid of trial frequencies. The period search yielded a single,
264: uniquely-defined frequency near 5.12 cycle d$^{-1}$; a least-squares
265: fit to the velocities of the form
266: $$v(t) = \gamma + K \sin\left[2\pi (t - T_0) \over P\right]$$ then
267: yielded a preliminary period of 0.195290 d, with a velocity
268: semiamplitude of 80~km~s$^{-1}$.
269:
270: Twenty-three additional RVs were used to extend the span of RV
271: observations and to improve the precision of the ephemeris. These
272: were measured using the same 11.5\,\AA\ FWHM Gaussian-derivative
273: technique, applied to high-resolution spectra from the Hobby-Eberly
274: Telescope (HET) obtained in 2004 and 2007 (see \S\ref{sec:HRS} below).
275: The final heliocentric {\Ha} RV ephemeris, based on 163 points, is
276: given by:
277: \begin{eqnarray*}
278: T_0 &=& \hbox{HJD } 2453628.6389 \pm 0.0016, \\
279: P &=& 0.1952894 \pm 0.0000010\ \hbox{d}, \\
280: K &=& 77 \pm 4\ {\rm km\ s^{-1}}, \ \hbox{and} \\
281: \gamma &=& -44 \pm 3\ {\rm km\ s^{-1}}. \\
282: \end{eqnarray*}
283: The rms residual between the data and the fitted curve is 17
284: km~s$^{-1}$. Figure \ref{fig:foldedRV} shows the RVs folded on the
285: period, with the best-fitting sinusoid superposed. The periodicity in
286: the {\Ha} RVs is coherent over three observing seasons, and the
287: velocities admit no possibility of a cycle count error --- the data
288: are consistent with a binary orbit.
289:
290: \subsection{High-Resolution Spectra} \label{sec:HRS}
291:
292: Follow-up observations were taken at the HET using the High Resolution
293: Spectrograph \citep[HRS;][]{Tull98}. Owing to the declination of
294: {\LSIV}, it can only be observed by the HET for about one hour per
295: night as it is transiting, from mid-April through mid-July. Queue
296: observations of {\LSIV} were made with the HRS using $2''$ fibers (one
297: object fiber and two sky fibers) and 2$\times$3 on-chip binning,
298: resulting in $R \equiv \lambda/\Delta\lambda \approx30,000$, covering
299: the wavelength region from {\Hb} to {\Ha}. (Unfortunately, the
300: $\lambda$5780 DIB falls into the gap between the ``red'' and ``blue''
301: CCDs with this setup.) Exposure durations were 10 minutes, with the
302: exception of the first observation of 20 min duration and a few
303: shortened exposures obtained as {\LSIV} approached the end of the HET
304: track. Three or four separate exposures were obtained in quick
305: succession on each of four nights in 2004 and three nights in 2007
306: (see Table~\ref{table:HRS} for a journal of observations). The typical
307: signal-to-noise ratio was 100 at 5822\,\AA\ (range 75--120). Th-Ar
308: lamps and flat field exposures were taken either before or after the
309: sequence of exposures each night.
310:
311: These spectra were processed with standard IRAF routines. At the
312: dispersion of the HRS, the absorption components of the Balmer lines
313: in the {\LSIV} spectrum are roughly as broad as a single spectral
314: order. Thus, a pseudo-continuum was fitted to the broad H absorption
315: profiles, and each spectrum was normalized using this fit. The data
316: were then rebinned in uniform heliocentric wavelength bins.
317: Figure~\ref{fig:HRS-keyphases} shows the normalized {\Ha} core
318: emission profiles at several key orbital phases. As the phase varies,
319: the profile does not simply shift back and forth in velocity, but
320: rather changes shape. To the eye, a decomposition into a broad
321: component and a narrow component moving in anti-phase is suggested.
322: This is made more evident in Figure~\ref{fig:trailedspec}, which gives
323: a trailed-spectrogram representation of all the data.
324:
325: The measured EW of the {\Ha} emission (above the pseudo-continuum) was
326: roughly constant at $\sim$3.0\,\AA\ in 2004, but was lower in 2007,
327: averaging 1.7, 2.1, and 1.8\,\AA\ on 2007 April 19, June 8, and June
328: 22 respectively. To make Figs.~\ref{fig:HRS-keyphases} and
329: \ref{fig:trailedspec}, we have ``stretched'' each spectrum to achieve
330: a uniform emission EW of 3.0\,\AA, keeping the continuum level at
331: unity. In the trailed spectrogram (Fig.\ref{fig:trailedspec}), each
332: spectrum has then been made to fill a phase range of 0.05 cycles,
333: centered on the phase bin (bin width = 0.01 cycles) nearest to the
334: orbital phase at mid-observation. Note that data from 2004 and 2007
335: have been intermingled, and different orbital phases were observed in
336: different years. The change in EW from one year to the other may
337: signal a change in the behavior of the profile {\em vs} orbital phase,
338: so we caution that quantitative conclusions about these distinct
339: narrow and broad velocity components, based on these figures, are
340: tentative.
341:
342: The phase of the broad component in the {\Ha} emission profile is
343: given by the RV ephemeris derived in \S\ref{sec:RV}, with the narrow
344: component moving in anti-phase. Thus, the Gaussian-derivative method
345: using FWHM = 11.5\,\AA\ described in \S\ref{sec:RV} likely gives a
346: ``diluted'' RV semiamplitude for the broad component. Given the
347: possible change in the profile behavior between 2004 and 2007 noted
348: above, we did not attempt to decompose each observed profile or
349: otherwise make formal two-component fits to the high-resolution data.
350: A rough estimate gives FWHM $\sim$8--10 and $\sim$3\,\AA\ for the
351: broad and narrow components. Estimates of the RV semiamplitudes are
352: $K_{\rm broad}\sim 120$ and $K_{\rm narrow}\sim 90$~km~s$^{-1}$, with
353: errors of perhaps 20~km~s$^{-1}$.
354:
355: We tentatively identify the broad emission as arising from an
356: accretion disk around a compact object and the narrow emission as
357: arising from the mass donating star in a cataclysmic binary.
358: %If the region giving rise to the broad emission is centered on the
359: %compact object and traces its dynamics, the corresponding mass
360: %function is $f(m) = PK_{\rm broad}^3/2\pi G = 0.035~{\rm M_\odot}$
361: %with large errors.
362: If the narrow emission arises on the hemisphere of the mass donor that
363: faces the disk (see \S\ref{sec:novalike}), one might expect the EW of
364: the narrow component to vary, such that the emission EW is largest
365: when the mass donor is at superior conjunction, which is orbital phase
366: 0.5 in our convention. We searched for this effect using the
367: low-resolution MDM spectra from the most extensive data set (2005
368: July), but did not find any firm indication of an EW variation with
369: orbital phase. The limit on the variation in the total EW is 15--20\%
370: full amplitude. It is not possible to decompose these low-resolution
371: spectra into separate components. The high-resolution HRS spectra are
372: unfortunately not suitable for this test, since all data collected
373: near orbital phase 0.5 come from the 2007 observations, and all data
374: colleced near orbital phase 0.0 come from 2004. While there is a
375: tantalizing difference in total EW between these two observing
376: seasons, we cannot know whether it is related to orbital phase; a more
377: concentrated observing campaign is needed. We note that there need
378: not be a large modulation of the EW of the narrow component, if the
379: orbital inclination of the binary is moderate.
380:
381: The {\Hb} emission line profile shows behavior similar to {\Ha}, with
382: the narrow component being perhaps more pronounced (slightly higher
383: intensity compared with the broad component). Similar behavior is
384: also reflected in the \ion{He}{1} $\lambda$5876 emission profile,
385: although it is weaker.
386:
387: %----------------------------------------------------------------------
388:
389: \subsection{Photometry} \label{sec:Photometry}
390:
391: \subsubsection{Michigan State University 0.6m Telescope}\label{sec:MSU}
392:
393: Differential $V$ band photometry of {\LSIV} was obtained using the
394: Michigan State University 0.6m telescope equipped with an Apogee Ap47p
395: CCD camera. Aperture photometry for {\LSIV} was obtained relative to
396: the star TYC~$5642$$-00482$$-1$, which has $B_T = 12.389$, $V_T =
397: 11.183$. (This star is significantly redder than the program star,
398: but there is no blue star of comparable brightness to the target
399: within the CCD field of view.) Observations were made on twelve
400: nights: UT 2004 July 26; 2004 August 1, 7, 9, 16, 22, 31; 2004 September
401: 18; 2005 July 15, 20, 23; and 2006 May 27. Exposure times were
402: typically 30 s, varying somewhat depending on observing conditions.
403: Observing run lengths varied from as short as about an hour to 6.3 hr
404: on 2006 May 27, this being about the maximum run length possible at
405: reasonable airmass from the latitude of the observatory. Calibrations
406: were done using standard procedures and sky flats. Since observations
407: were only taken in the $V$ filter, no transformations to the standard
408: system involving color terms could be attempted. The typical error
409: of a single observation is $\pm 0.02$~mag.
410:
411: The photometry shows {\LSIV} to be variable on both short (flickering)
412: and long timescales, with no indication of eclipses. By itself, the
413: photometry suggests no clear or consistent period of variation ---
414: folding the data on numerous trial periods showed many possible alias
415: periods. Folding the data on the period derived from the RV
416: observations (\S\ref{sec:RV}) does show weak modulations or waves,
417: perhaps multi-periodic, in the brightness of {\LSIV}
418: (Fig.~\ref{fig:foldedphot}). The rapid modulation in $\Delta V$ that
419: is seen in the 2005 Jul 23 data, of amplitude $\sim$0.03 mag and
420: lasting several cycles over the phase interval $\sim$0.2--0.6, is not
421: seen in the photometric check stars; this may be simply an instance of
422: strong flickering. The median light level is about $\sim$0.06 mag brighter
423: in 2005--2006 than in 2004. Additional time-series photometry, from a
424: more southern site to permit longer nightly runs, would be desirable
425: in characterizing the light curve and searching for persistent
426: periodic signals.
427:
428: %\notetoeditor{Fig.~\ref{fig:foldedphot} is provided in two
429: %versions: f6_BW.eps (black and white, for the paper journal)
430: %and f6_color.eps (color, on-line journal only).}
431:
432: \subsubsection{Northern Sky Variability Survey}\label{sec:NSVS}
433:
434: {\LSIV} was detected as a variable star in the Northern Sky
435: Variability Survey {\citep[NSVS; LS\,IV$-08^\circ 3$ $\equiv$
436: NSVS~16408817;][]{NSVS}}. The NSVS record shows 92 observations of
437: this star. The typical error of a single measurement is $\sim$0.02
438: mag. The data are sparse, collected on 45 nights during the interval
439: 1999 April to 2000 March, with no more than four observations per
440: night. A light curve folded on our RV period shows no phase-dependent
441: pattern. The median unfiltered optical magnitude is 11.855 with a
442: scatter of 0.052 mag. The total range of observed magnitudes is
443: 11.698$\pm$0.026 to 11.977$\pm$0.068. The data exhibit variations in
444: median brightness at the 0.05 mag level with characteristic time
445: scales of 20--40 days. Similar to the epoch-to-epoch variations seen
446: in the MSU observations, this suggests that the system has been
447: consistent in its photometric behavior over the period of 1999-2006.
448:
449: %----------------------------------------------------------------------
450:
451: %Section 3
452: \section{Discussion}\label{sec:Discussion}
453:
454: \subsection{Novalike CV Interpretation}\label{sec:novalike}
455:
456: We have already suggested (\S\ref{sec:Steward}) that {\LSIV} could
457: have been misclassified as a hot subdwarf (sdB) star, given only
458: photometry and spectroscopy confined to wavelengths blueward of {\Hb}.
459: Indeed, the dwarf nova FO Per (aliases: RL 92 and RWT 92) was classified
460: by \citet{Chromey79} as an sdB star, when, unbeknownst to him, it must
461: have been near maximum light in its outburst cycle. Two Steward
462: Observatory 2.3m spectra of FO Per, also near maximum light, obtained
463: in 2004 Dec and 2006 Dec with the identical setup as described in
464: \S\ref{sec:Steward}, show a spectrum that is almost indistinguishable
465: from a somewhat reddened sdB star, except for emission cores in
466: H$\alpha$, H$\beta$, and, very faintly, in the next few higher order
467: Balmer lines, i.e., at almost exactly the same level as seen in our
468: 2007 May spectrum of {\LSIV}.
469:
470: The SED of {\LSIV} is considerably redder than those of known single
471: sdB stars. Dereddening by the equivalent of $E(\bv) \sim 0.6$ or more
472: would be required to bring the Steward SED at optical wavelengths into
473: agreement with sdBs that have $T_{\rm eff}$ in the range 24,000 ---
474: 35,000 K. Likewise, the $E(\bv)$ needed to bring the 2MASS infrared
475: colors into the range of single sdB stars is 0.6 mag or more.
476:
477: On the other hand, {\LSIV} shares many properties with novalike
478: variables of the UX UMa subclass. These systems show persistent broad
479: Balmer absorption-line spectra. Ironically, these lines were once
480: thought to indicate pressure-broadening in the atmosphere of a hot
481: subdwarf or white dwarf --- e.g., \citet{WH54} --- and only later
482: were reinterpreted as arising from an optically thick accretion disk,
483: where they are (at least in part) kinematically broadened.
484:
485: UX UMa stars also exhibit a wide range in strengths of the emission
486: line components. Emission from combinations of {\HI}, {\HeI},
487: {\HeII}, {\NIII}, and {\CIII} have all been observed at a variety of
488: strengths in various UX UMa stars, or variable over time in a single
489: star of the class. Emission is seen in all of these lines in {\LSIV},
490: and the features are consistent with those observed in the UX UMa
491: stars RW Sex and IX Vel
492: {\citep[][respectively]{BSS92,BeuermannThomas90}}. For UX UMa stars,
493: the Balmer decrement in the emission components is steeper than in the
494: absorption lines. This can result in a situation where {\Ha} is
495: purely in emission, while higher members show progressively stronger
496: (less filled-in) absorption troughs. This effect is seen in {\LSIV}
497: (Fig.~\ref{fig:betsyflux2002}).
498:
499: Narrow Balmer emission has been found in many UX UMa systems,
500: superimposed on the broader emission from the disk. This narrow
501: component is attributed to irradiation of the atmosphere of the
502: secondary star, with reprocessing of the light to Balmer emission. As
503: discussed in \S\ref{sec:HRS}, the asymmetry and variations in the
504: emission line profile shape (Figs.~{\ref{fig:HRS-keyphases}} {\&}
505: {\ref{fig:trailedspec}}) suggest the presence of such a component,
506: moving in anti-phase to the broad component, as required for this
507: interpretation.
508:
509: The orbital period of $\sim0.195$\,d ($\sim4.7$ hours) determined for
510: {\LSIV} is comparable to other UX UMa systems.
511:
512: \subsection{Distance and Reddening}\label{sec:reddening}
513:
514: {\LSIV} lies at galactic coordinates $(l,b) = (11\fdg0, +20\fdg8)$.
515: The \citet{Schlegel} reddening maps show that the total line-of-sight
516: extinction varies on small angular scales in this direction. We
517: attempted to infer the $E(\bv)$ color excess directly from the
518: strength of the DIBs, using the linear relations, $W = a_0 E(\bv)$,
519: established by {\citet{HerbigDIB}}. The $\lambda$5780 DIB has an EW
520: of $0.32 \pm 0.03$\,\AA, suggesting $E(\bv) \approx 0.50$ mag with an
521: uncertainty of at least 10\%, if we are measuring the narrow component
522: of this feature only, or $E(\bv) \approx 0.24$ mag if we are measuring
523: the sum of the narrow and broad DIBs (the latter is more likely). The
524: $\lambda$4430 DIB has an EW of $0.26 \pm 0.04$\,\AA, corresponding to
525: $E(\bv) \approx 0.12$ with at least 15\% uncertainty. Some of the
526: uncertainty arises, depending on whether ``all-sky'' or ``Sco-Oph''
527: values of $a_0$ are adopted. Both DIBs imply reddening less than the
528: full \citet{Schlegel} column: $E(\bv) = +0.47$.
529:
530: The heliocentric RV of the interstellar \ion{Na}{1} D lines is
531: $-14.5~{\rm km~s^{-1}}$ (average of D$_1$ and D$_2$). Most of this
532: can be explained by the reflex of the Sun's space motion with respect
533: to the Local Standard of Rest, so kinematic information about the
534: distance of the absorbing material between the Sun and {\LSIV} is
535: unavailable.
536:
537: If {\LSIV} is assumed to be a UX UMa variable star, how far away is
538: it, and is this consistent with the measured reddening? We adopt an
539: absolute magnitude $M_V \approx +5$ and a representative apparent
540: magnitude $V = 11.50$ for {\LSIV}. We use the standard relation $A_V =
541: 3.1 E(\bv)$ to compute distance $d$ as a function of assumed $E(\bv)$.
542: Values between 100 and 200 pc result for $E(\bv)$ in the range 0.0 to
543: 0.5 mag, smaller distances corresponding to larger reddenings.
544: \citet{Ak07} have put forward a calibration of $M_J$ for CVs, using as
545: inputs the orbital period and the extinction-corrected 2MASS $J$
546: magnitude and $J-H$, $H-K_{s}$ colors. Again varying the assumed
547: $E(\bv)$, we find $d$ in the range 140--180 pc by this method (minimum
548: error at fixed reddening is 40\%). Comparing the 2MASS colors of
549: {\LSIV} and IX Vel directly, we find a reasonable match if $E(\bv)
550: \approx 0.10$ mag. We conclude that, if $E(\bv)$ in the range
551: 0.10--0.20 mag can be accommodated within distances of 100--200 pc,
552: the colors and magnitudes of {\LSIV} are consistent with those of a UX
553: UMa variable. Such reddening values within such distances are
554: plausible along the line of sight to {\LSIV}, according to the maps
555: of, e.g., {\citet{PerryJohnston82}}.
556:
557: \subsection{{\em ROSAT} X-ray Observations}
558:
559: A weak X-ray source detected in the {\em ROSAT} All-Sky Survey (RASS)
560: appears to be coincident with {\LSIV}
561: {\citep[$1$RXS~J$165630.2$$-083442$;][]{Voges}}. The count rate in
562: the position sensitive proportional counter (PSPC) was
563: 0.11$\pm$0.02~s$^{-1}$, yielding an estimated 0.1--2.4\,keV flux of
564: $1.5 \times 10^{-12}~{\rm erg~cm^{-2}~s^{-1}}$. Taking {\LSIV} to be
565: the optical counterpart, the X-ray/optical flux ratio is given by
566: $\log f_X/f_{\rm opt} \approx -2.0$.
567:
568: This source was also observed serendipitously by {\em ROSAT}
569: {\citep[$1$WGA~J$1656.4$$-0834$;][]{White}} during a pointed
570: observation of another target. In this $\sim$8000\, s exposure, 150
571: net counts were collected, and the count rate was
572: 0.02$\pm$0.002~s$^{-1}$ (5 times lower than in the RASS),
573: corresponding to $f_X \approx 3.0\times 10^{-13}~{\rm
574: erg~cm^{-2}~s^{-1}}$ and $\log f_X/f_{\rm opt} \approx -2.6$. A
575: best-fitting blackbody spectrum has $kT \approx 0.2$~keV with an
576: intervening hydrogen column density of $1.3\times 10^{21}$~cm$^{-2}$,
577: corresponding to $E(\bv) \approx 0.2$. The unabsorbed flux is
578: $F_{\mathrm{0.5-2~keV}} \approx
579: 2.7\times10^{-13}$~erg\,s$^{-1}$\,cm$^{-2}$. Fits using a bremsstrahlung or
580: thermal plasma emission model would result in a similar $kT$ value,
581: given the highly absorbed spectrum and the PSPC's limited energy range
582: and resolution. The inferred total X-ray flux is quite uncertain.
583:
584: X-ray emission is not associated with sdB stars. On the other hand,
585: CVs of all subclasses show X-ray emission. {\citet{VBRP97}} discuss
586: the CVs detected in the RASS. A typical luminosity in the
587: 0.5--2.5~keV band for a novalike variable is $L_X \sim 10^{31}~{\rm
588: erg~s^{-1}}$. The required distance for {\LSIV} to have this $L_X$,
589: given its modeled flux, is $d \sim 500$ pc. Using a 2 keV
590: bremsstrahlung spectrum to model the ROSAT counts, {\citet{VBRP97}}
591: find $\log F_X/F_{\rm UV+opt} < -2.5$ to be typical for novalikes.
592: Given the uncertainties in the input spectrum and absorption column,
593: we conclude that {\LSIV} has X-ray properties consistent with
594: membership in the UX UMa subclass of CVs.
595:
596: %Section 4
597: \section{Summary}\label{sec:Summary}
598:
599: Given the many similarities between our observations and other UX UMa
600: novalike variables, we propose that {\LSIV} should be classified with
601: the UX UMa stars. Evidence supporting this classification includes
602: the presence of emission reversals in the Balmer absorption series;
603: other emission lines including \ion{He}{1}, \ion{He}{2}, and
604: \ion{N}{3}/\ion{C}{3}; the variability of the emission line strengths;
605: the intrinsically blue continuum; and the mild photometric
606: variability. The key evidence in favor of a CV interpretation lies in
607: the periodically modulated Doppler shifts and profiles of the emission
608: lines, with a period of 4.7 hr, consistent with a low-mass quasi-main
609: sequence star that fills its Roche lobe and transfers mass to a
610: luminous accretion disk around a white dwarf. The inner hemisphere of
611: the donor star may be irradiated by the accretion disk. Reddening and
612: distance estimates are consistent with the CV interpretation, from the
613: standpoint of the expected luminosity of the system and the
614: distribution of interstellar dust in the direction of {\LSIV}.
615: Finally, weak X-rays are apparently associated with the system, and
616: the X-ray luminosity and X-ray/optical flux ratio are consistent with
617: observations of other novalike CVs.
618:
619: We refrain from deriving the dynamical mass ratio of the binary or
620: individual masses of the stars in {\LSIV} from any of the various RV
621: semiamplitudes that we have presented above, mainly because the HRS
622: spectra on which such estimates would rely were obtained on widely
623: separated dates, and the EW of the emission lines varied significantly
624: among the various epochs. It is thus possible that one or both sites
625: of the emission varied as well. Work aimed at determining the mass
626: ratio is best undertaken with intensive time-series spectroscopy that
627: can provide full orbital phase coverage in one or two nights, to
628: minimize such variations. Likewise, we cannot offer a definitive
629: interpretation of the photometric variability, given our present
630: limited data. {\LSIV} thus remains as an attractive and interesting
631: target for further study.
632:
633: Despite the apparent line-of-sight extinction to the source, {\LSIV}
634: is one of the brighter UX UMa stars so far discovered.
635: {\citet{Downes}} list only three ``UX Uma'' systems (of two dozen
636: catalogued) that are brighter: IX Vel, V3885 Sgr, and RW Sex. Only
637: seven CVs of any ``novalike'' sub-type (89 systems total) are brighter
638: than {\LSIV} at their maximum light. Bright, relatively nearby
639: systems are not only appropriate for follow-up studies, but contribute
640: disproportionately to studies of the space density and kinematics of
641: the various subclasses of CVs. Even though hundreds of thousands of
642: stars that are much fainter have been observed spectroscopically by,
643: e.g., the Sloan Digital Sky Survey, the sky has not been thoroughly
644: explored at $V\sim 11$.
645:
646: \acknowledgments
647:
648: M.~A.~Stark thanks G.~B.~Berriman, for his help interpreting the 2MASS
649: catalog, and K.~T.~Lewis, A.~Narayanan, and K.~Herrmann, for assisting
650: with the observations at KPNO. We are grateful to K.\ T.\ Lewis for
651: carrying out the fit to the {\em ROSAT} data, and to L.\ Townsley, who
652: provided helpful interpretation.
653: We also thank the HET resident astronomers M.\ Shetrone, S.\ C.\
654: Odewahn, and H.\ Edelmann and telescope operators F.\ Deglman, M.\
655: Soukup, M.\ Villarreal, and V.\ Riley. The Hobby-Eberly Telescope is
656: a joint project of the University of Texas at Austin, the Pennsylvania
657: State University, Stanford University,
658: Ludwig-Maximillians-Universit\"at M\"unchen, and
659: Georg-August-Universit\"at G\"ottingen. The HET is named in honor of
660: its principal benefactors, William P. Hobby and Robert E. Eberly.
661: This research has been supported by grants from NASA (including NASA
662: GSRP grant NGT5-50399), the Zaccheus Daniel Foundation for
663: Astronomical Science, and the Pennsylvania Space Grant Consortium.
664: JRT and CSP gratefully acknowledge funding from the National Science
665: Foundation through grant AST-0307413, and thank the MDM staff for
666: expert and conscientious observing support. Holly Sheets took some of
667: the MDM spectra. As always, we are grateful to the Tohono O'odham for
668: leasing us their mountain so that we may study the great universe
669: around us. HAS thanks the Center for Cosmic Evolution and the
670: National Science Foundation (grant AST-0440061) for their support.
671: EMG acknowledges support from NSF grant AST-0098699. This publication
672: makes use of data products from the Two Micron All Sky Survey, which
673: is a joint project of the University of Massachusetts and the Infrared
674: Processing and Analysis Center/California Institute of Technology,
675: funded by the National Aeronautics and Space Administration and the
676: National Science Foundation.
677:
678: {\it Facilities:} \facility{KPNO:2.1m (GoldCam), \facility{Bok (B\&C
679: spectrograph)}, \facility{Hiltner (modular spectrograph)},
680: \facility{McGraw-Hill (Mark III spectrograph)}, \facility{HET (HRS)}}
681:
682: %----------------------------------------------------------------------
683:
684: \singlespace
685:
686: \bibliographystyle{apj}
687: \bibliography{bibs}
688:
689: \clearpage
690:
691: %----------------------------------------------------------------------
692:
693: %Figures go here
694:
695: %Fig 1; single-column
696: \begin{figure}
697: \epsscale{0.45}
698: {\plotone{f1.eps}}
699: \caption{KPNO 2.1m Goldcam spectra of {\LSIV} showing the variable
700: nature of the emission components at {\Hb} ({\em top}),
701: {\HeI}~5875\,\AA\ partly blended with interstellar {\NaI}~D lines
702: ({\em middle}), and {\Ha}/{\HeI}~6678\,\AA\ ({\em bottom}). Dates and
703: times are UT 2003; times correspond to mid-exposure and are
704: heliocentric. Phases corresponding to the ephemeris in \S\ref{sec:RV}
705: are also indicated. Dates that are italicized in parentheses in the
706: bottom panel do not have corresponding data for {\Hb} and
707: {\HeI}~5875\,\AA. Dotted lines show the laboratory wavelengths of the
708: labeled transitions; the spectra themselves are plotted using observed
709: (topocentric) wavelengths. Several gaps in the spectra result from
710: cosmic ray removal.
711: \label{fig:GoldCamspec}}
712: \end{figure}
713:
714: %Fig 2; likely 2-column
715: \begin{figure}
716: \epsscale{0.40}
717: \rotatebox{-90}
718: {\plotone{f2.eps}}
719: \caption{Flux-calibrated spectrum of {\LSIV}, obtained UT 2002 April 3
720: at the Steward Observatory 2.3m telescope, including higher-order
721: Balmer lines. Absorption bandheads at $\lambda$6277 and $\lambda$6867
722: are telluric; absorption features at $\lambda$4430 (near the stronger
723: \ion{He}{1} $\lambda$4471 feature) and $\lambda$5780 are diffuse
724: interstellar bands.
725: \label{fig:betsyflux2002}}
726: \end{figure}
727:
728: \clearpage
729:
730: %Fig 3; likely single-column
731: \begin{figure}
732: \epsscale{0.55}
733: \plotone{f3.eps}
734: \caption{{\Ha} radial velocities used in the analysis, folded on the
735: best-fitting period, with the sinusoidal fit superposed. {\em
736: Crosses:} MDM velocities. {\em Filled squares:} HET-HRS velocities.
737: Two cycles are shown for continuity.
738: \label{fig:foldedRV}}
739: \end{figure}
740:
741: \clearpage
742:
743: %Fig 4, likely 2-columns
744: \begin{figure}
745: \epsscale{0.45}
746: \rotatebox{-90}{\plotone{f4.eps}}
747: \caption{HET-HRS spectra showing only the emission component of the
748: {\Ha} line, for selected orbital phases:
749: {\em left:} $\phi$=0.256 (2004 May 24);
750: {\em middle:} $\phi$=0.508 (2007 April 19);
751: {\em right:} $\phi$=0.769 (2004 May 29).
752: Wavelengths are heliocentric.
753: The spectra have been ``stretched'' to make the emission EW uniformly
754: 3.0\,\AA\ (see \S\ref{sec:HRS}).
755: The variation of the emission profile is evident.
756: The dashed vertical line in each panel indicates the laboratory
757: wavelength of {\Ha}.
758: The narrow absorption lines, which are stationary in the topocentric
759: frame, are telluric.
760: \label{fig:HRS-keyphases}}
761: \end{figure}
762:
763: %Fig 5
764: \begin{figure}
765: \epsscale{0.55}
766: \plotone{f5.eps}
767: \caption{A trailed-spectrogram representation of the variations in the
768: {\Ha} emission profile from the HET-HRS.
769: Wavelengths are heliocentric. Orbital phase increases upwards, with
770: 1.5 cycles being shown for continuity (data are repeated).
771: Intensities are in units of the continuum level and are indicated by
772: the calibration strip along the {\em right} side of the main panel.
773: The first inset ({\em upper left}) is intended as a guide to
774: identifying the two distinct emission components moving in antiphase.
775: The second inset ({\em upper right}) shows the single RV curve, with
776: reduced amplitude, resulting from the analysis of the low-resolution
777: MDM data. See text for details and cautions.
778: \label{fig:trailedspec}}
779: \end{figure}
780:
781: %Fig 6
782: \begin{figure}
783: \epsscale{0.60}
784: \plotone{f6_color.eps}
785: \caption{Folded differential light curve of {\LSIV}, based on the RV
786: ephemeris. The typical error of a single observation is $\pm
787: 0.02$~mag. Different symbols indicate different observing nights.
788: Data are plotted twice for phase continuity.
789: {\em Top:} observations from 2004 ---
790: Jul 26 filled diamond (colored red in electronic edition);
791: Aug 1 cross (green);
792: Aug 7 filled square (black);
793: Aug 9 filled triangle (blue);
794: Aug 16 open circle (red);
795: Aug 22 open square (black);
796: Aug 31 filled circle (green);
797: Sep 18 open triangle (blue).
798: {\em Bottom:} observations from 2005 ---
799: Jul 15 cross (green);
800: Jul 20 open circle (black);
801: Jul 23 filled square (blue) --- and 2006 --- May 27 open triangle (red).
802: \label{fig:foldedphot}}
803: \end{figure}
804:
805: \clearpage
806:
807: %---------------------------------------------------------------------------
808:
809: %Tables go here
810:
811: \input{tab1} % coordinates, photometry
812: \input{tab2} % GoldCam journal
813: \input{tab3} % GoldcCam EWs
814: \input{tab4} % MDM journal
815: \input{tab5} % HET-HRS journal
816:
817: \end{document}
818:
819: % =====================================================================
820: