1: \documentclass[12pt,preprint]{aastex}
2: %\usepackage{epsfig}
3: %\usepackage{wasysym}
4: \def\ascnode {\mbox{\wasyfamily\char19}}
5: \def\descnode {\mbox{\wasyfamily\char20}}
6: \def\saturnring1 {\mbox{\wasyfamily\char21}}
7: \def\vernal {\mbox{\wasyfamily\char23}}
8: \def\venus {\mbox{\wasyfamily\char25}}
9: \def\mars {\mbox{\wasyfamily\char26}}
10: \def\viewangle {\mbox{\wasyfamily\char30}}
11: \def\saturnring2 {\mbox{\wasyfamily\char31}}
12: \def\newmoon {\mbox{\wasyfamily\char32}}
13: \def\leftmoon {\mbox{\wasyfamily\char36}}
14: \def\rightmoon {\mbox{\wasyfamily\char37}}
15: \def\earth {\mbox{\wasyfamily\char38}}
16: \def\mercury {\mbox{\wasyfamily\char39}}
17: \def\jupiter {\mbox{\wasyfamily\char40}}
18: \def\saturn {\mbox{\wasyfamily\char41}}
19: \def\uranus {\mbox{\wasyfamily\char42}}
20: \def\neptune {\mbox{\wasyfamily\char43}}
21: \def\pluto {\mbox{\wasyfamily\char48}}
22: \def\leftblackmoon {\mbox{\wasyfamily\char71}}
23: \def\rightblackmoon {\mbox{\wasyfamily\char72}}
24: \def\leftwhitemoon {\mbox{\wasyfamily\char73}}
25: \def\rightwhitemoon {\mbox{\wasyfamily\char74}}
26:
27: \shorttitle{FUSE Spectroscopy of BB Dor}
28: \shortauthors{Godon et al.}
29: \begin{document}
30: \bibliographystyle{apj}
31:
32: \title{Far Ultraviolet Spectroscopic Explorer Spectroscopy of the
33: Nova-like BB Doradus
34: \altaffilmark{1}}
35:
36: \author{Patrick Godon\altaffilmark{2}, Edward M. Sion}
37: \affil{Department of Astronomy and Astrophysics,
38: Villanova University,
39: Villanova, PA 19085}
40: \email{patrick.godon@villanova.edu}
41: \email{edward.sion@villanova.edu}
42:
43: \author{Paul E. Barrett}
44: \affil{United States Naval Observatory,
45: Washington, DC 20392}
46: \email{barrett.paul@usno.navy.mil}
47:
48: \author{Paula Szkody}
49: \affil{Department of Astronomy,
50: University of Washington,
51: Seattle, WA 98195}
52: \email{szkody@astro.washington.edu}
53:
54: \and
55:
56: \author{Eric M. Schlegel}
57: \affil{Department of Physics and Astronomy,
58: University of Texas,
59: San Antonio, TX 78249}
60: \email{eric.schlegel@utsa.edu}
61:
62: \altaffiltext{1}
63: {Based on observations made with the
64: NASA-CNES-CSA Far Ultraviolet Spectroscopic
65: Explorer. {\it{FUSE}} was operated for NASA
66: by the Johns Hopkins University under
67: NASA contract NAS5-32985}
68: \altaffiltext{2}
69: {Visiting at the Space Telescope Science Institute, Baltimore, MD 21218,
70: godon@stsci.edu}
71:
72:
73: \begin{abstract}
74:
75: We present an analysis of the Far Ultraviolet Spectroscopic Explorer
76: ({\it{FUSE}}) spectra of the little-known southern nova-like BB
77: Doradus. The spectrum was obtained as part of our Cycle 8 {\it FUSE}
78: survey of high declination nova-like stars.
79:
80: The {\it FUSE} spectrum of BB Dor, observed in a high state, is
81: modeled with an accretion disk with a very low inclination (possibly
82: lower than 10$^{\circ}$). Assuming an average WD mass of
83: $0.8M_{\odot}$ leads to a mass accretion rate of $10^{-9}
84: M_{\odot}$/yr and a distance of the order of $\sim$650 pc, consistent
85: with the extremely low galactic reddening in its direction. The
86: spectrum presents some broad and deep silicon and sulfur absorption
87: lines, indicating that these elements are over-abundant by 3 and 20
88: times solar, respectively.
89:
90: \end{abstract}
91:
92: \keywords{
93: accretion, accretion disks
94: ---
95: novae, cataclysmic variables
96: ---
97: stars: abundances
98: ---
99: stars: individual (BB Doradus)
100: ---
101: stars: nova-likes
102: ---
103: stars: white dwarfs
104: }
105:
106: \section{Introduction}
107:
108: \subsection{Cataclysmic Variable Nova-Likes}
109:
110: Cataclysmic variables (CVs) are close binaries in which the primary, a
111: white dwarf (WD), accretes matter and angular momentum from the
112: secondary, a main sequence star, filling its Roche-lobe. The matter
113: is transferred by means of either an accretion disk around the WD, or
114: an accretion column, when the WD has a strong ($\sim 10$ MG) magnetic
115: field. Ongoing accretion at a low rate (quiescence) is interrupted
116: every few weeks to months by intense accretion (outburst) of days to
117: weeks (a dwarf nova accretion event), and every few thousand years by
118: a thermonuclear explosion (TNR; the classical nova event). CV systems
119: are divided into sub-classes according to the duration, occurrence and
120: amplitude of their outburst: dwarf novae (DN) are found mostly in the
121: low state; while nova-likes (NL) exhibit the spectroscopic and
122: photometric characteristics of novae between or after outbursts but
123: have never had a recorded outburst. NLs form a less homogeneous class
124: and are divided into sub-types depending on certain properties, such
125: eclipsing systems, magnetic systems, or systems that go into
126: unexpected low states \citep{war95}. Non-magnetic NLs can be divided
127: into UX UMa systems which remain in a high optical brightness state or
128: "permanent outburst state", and the VY Scl systems (anti-dwarf novae)
129: which experience unexpected low states when the optical brightness
130: plummets.
131:
132: CVs are believed to evolve from long period ($\sim $2 day) toward
133: short period ($\sim$1 hour) as the mass ratio $q$ decreases, crossing
134: the 2--3 hour 'period gap', where few system are found. Theory
135: predicts that the period of those systems having a sufficiently low
136: $q$ will increase as $q$ continues to decrease, the so-called 'period
137: bouncers' \citep{pat05b}. The driving mechanism behind the mass
138: transfer ($\dot{M}$) and hence, the evolution of CVs is believed to be
139: angular momentum loss, dominated by magnetic stellar wind braking at
140: periods above 3 hours and by gravitational wave emission below this
141: period. The secondaries are stripped to $< 0.08 M_{\odot}$ on a
142: timescale of only 1-4 Gyr after they form as close binaries above the
143: gap. Consequently, in the lifetime of the galaxy, the vast majority
144: of CVs should have evolved to a period minimum near 80 min and now
145: have degenerate brown dwarf-like secondaries \citep{how01}. However,
146: recent models \citep{and03} suggest (i) much lower angular momentum
147: loss rates, so it takes 10-12 Gyr to even reach a 3 hr period, and
148: (ii) multiple evolutionary tracks yielding different populations of
149: CVs above and below the 3 hour orbital period.
150:
151: The accretion disks and underlying accreting WDs in CVs can provide
152: crucial clues and constraints on these evolutionary scenarios. It is
153: important to derive $\dot{M}$ from observations, as the mass transfer
154: rate are tied to the angular momentum loss. As to the WDs, they are
155: the central engines of the observed outbursts, either as potential
156: wells for the release of accretional energy or as the sites of
157: explosive TNR shell burning. In addition, the differences between the
158: temperatures, rotation, and chemical abundances of CV WDs and single
159: isolated WDs provide clues as to the effects of accretion, diffusion,
160: and long term heating and evolution \citep{sio95}.
161:
162: With a wavelength range from 905--1187\AA\AA, {\it FUSE} covers that
163: part of the far-ultraviolet (FUV) spectrum where the hotter CV
164: component is dominant. For nova-likes, the dominant component in the
165: {\it FUSE} range is usually the accretion disk. Additional possible
166: components are the hot white dwarf, the boundary layer or accretion
167: belt, and other hot regions on or close to the WD. Many CV FUV
168: spectra are modeled with 2 components, usually a WD plus a disk. For
169: each system, one usually derives from the {\it FUSE} spectrum the
170: temperature of the WD ($T_{eff}$), the gravity, the rotational
171: velocity ($V_{rot} \sin{i}$), the chemical abundances, the mass
172: accretion rate, the inclination, and the distance to the system.
173:
174: It has therefore been very important to try and obtain {\it FUSE} data
175: for more CV systems in order to populate the $T_{eff}$-Period and the
176: $\dot{M}$-Period parameter spaces. There is a need to enlarge the
177: sample to obtain an accurate global picture of the systems above and
178: below the period gap. For that reason, we proposed to observe with
179: {\it FUSE} of a set of 16 high declination dwarf novae (our Cycle 7
180: survey of DNe) and 16 high declination nova-likes (our Cycle 8 survey
181: of NLs - all chosen from the on-line CV catalog of Downes {\it et
182: al.}). Unfortunately, our {\it FUSE} NL survey was cut short due to
183: the fatal failure of the reaction wheel of the telescope. Of the 16
184: targets, only two were observed: BB Dor and P831-57. The analysis of
185: the {\it FUSE} spectrum of P831-57 is presented elsewhere
186: \citep{bar08}. In this work we present the result of the {\it FUSE}
187: spectroscopic analysis of BB Dor.
188:
189: \subsection{The Nova-like BB Doradus}
190:
191: BB Dor (also known as EC 05287-5847, an object from the
192: Edingburgh-Cape Survey \citep{che01}) was spectroscopically identified
193: as a CV in December 1987 and was seen fainter than $V \sim 16.5$
194: until November 1992 \citep{che01}. Since that time it seems to be in
195: bright state ($V\sim 14.6-13.6$). A first (extremely uncertain)
196: estimate of its period by \citet{che01} put it right in the middle of
197: the period gap with a period of 0.107 day or 2.57 h. However, a 45
198: day coverage \citep{pat02,pat05a} corrected this value to 0.14923
199: d or 3.58 h, as expected for VY Scl stars which usually have periods
200: between $\sim 3$ and $\sim 4$ h. BB Dor exhibits superhumps with a
201: period excess of $\varepsilon=$9.4\% \citep{pat02} --- it has the largest
202: known fractional superhump of all systems
203: and therefore provides a calibrating point
204: at large $\varepsilon$ \citep{pat05a}.
205: BB Dor is a VY Scl nova-like and so far no magnetic activity has been found.
206: The system also exhibits quasi-periodic oscillations \citep{che01}.
207: \citet{che01} also noted that due to its very narrow Balmer lines
208: (with wings extending $\sim 650$km$~$s$^{-1}$ from the line center) BB
209: Dor is most likely a low-inclination system. From the Austral
210: Variable Star Observer Network (AVSON), it appears that BB Dor has
211: been around $V \approx 13.6-14.6$ mag (with an error $<0.1$ mag) for
212: the last couple of years and can vary as much as 0.4 mag within one
213: hour. The reddening value E(B-V) towards BB Dor is not known.
214: However, the Galactic reddening value in the direction
215: of BB Dor as inferred from the 100$\micron$ dust emission map \citep{sch98}
216: is very small: E(B-V)=$\sim$0.03. Since the dust map
217: gives the maximum dust extinction, we deduce that BB Dor must have a
218: negligible reddening ($E(B-V) \sim 0.01$ at most) which should not
219: affect the results; consequently we
220: do not deredden its {\it FUSE} spectrum.
221:
222: \section{Observations}
223:
224:
225: \subsection{The {\it FUSE} Spectrum of BB Dor}
226:
227: BB Doradus was observed with {\it FUSE} on 09 July 2007 at 02h51m09s
228: (UT), and 3 exposures were obtained (H9030301 - exposures 001, 002,
229: 003). The {\it{FUSE}} spectrum was obtained through the low resolution
230: (LWRS; 30"x30") aperture in TIME TAG mode. The data were processed
231: with the last and final version of CalFUSE v3.2.0 \citep{dix07}.
232:
233: A look at the count rate graph (generated by CALFUSE) indicates that
234: the data for each exposure were collected mainly during the first 1000
235: s, followed by a jitter lasting about 1,500 s. Toward the end of each
236: exposure, the count rate dropped to zero. Therefore, while the
237: co-added exposures for the 8 individual segments after processing by
238: CalFUSE give a good exposure time ({\it after screening}) of between
239: about 8750 s (segments 1) and 9750 s (segments 2), the count rate plot
240: indicates that the total {\it good collection time} ranges between
241: 3350 s (segments 1) and 3650 s (segments 2). We therefore weighted
242: these segments accordingly, multiplying the flux (of the order of $1
243: \times 10^{-13}$ergs$~$s$^{-1}$cm$^{-2}$\AA$^{-1}$ in the on-line {\it
244: preview}) by a factor of 2.61 (segments 1) and 2.67 (segments 2).
245: After that we processed the data as usual (see below).
246:
247: The spectral regions covered by the spectral channels overlap, and
248: these overlap regions are then used to scale the spectra in the SiC1,
249: SiC2, and LiF2 channels to the flux in the LiF1 channel. The low
250: sensitivity portions (usually the edges) of each channel are
251: discarded. In the present case the SiC2b spectral segment was very
252: noisy and we discarded it too. We also carried out a visual
253: inspection of the {\it{FUSE}} channels to locate the worm and we
254: {\it{manually}} discarded these portions of the spectrum affected by
255: the worm. We combine the individual exposures and channels to create
256: a time-averaged spectrum weighting the flux in each output datum by
257: the exposure time and sensitivity of the input exposure and channel of
258: origin. The final product is a spectrum that covers the full
259: {\it{FUSE}} wavelength range $905--1187$\AA\AA. Because we
260: disregarded the right edge of both the SiC1a and SiC2b segments, there
261: is a gap between the LiF1a and LiF2b segments ($\sim 1082$ -- $\sim
262: 1087$\AA\AA).
263:
264: The {\it FUSE} spectrum of BB Dor is shown in Figure 1. Note
265: that the {\it FUSE} exposures of BB Dor
266: consist entirely of observations carried out during the DAY. The small
267: NIGHT exposure was ignored because of jitters and other problems. We
268: suspect that most of the emission lines are due to helio- and
269: geo-coronal emissions.
270:
271: \subsection{The {\it{FUSE}} Lines}
272:
273: The {\it{FUSE}} spectra of CVs exhibit mainly broad absorption lines
274: from the accretion disk and WD, as well as sharp absorption lines from
275: circumstellar (or circumbinary) material and the ISM. Emission lines
276: from the hotter regions are usually broadened, depending on the
277: inclination angle $i$, due to the large Keplerian velocity in the
278: inner disk. Sharp emission lines from air glow (geo- and
279: helio-coronal in origin) are also present, due mainly to the
280: reverberation of sunlight inside the {\it FUSE} telescope during DAY
281: time observations. The present {\it FUSE} spectrum of BB Dor presents
282: such a complexity, and all the lines are listed in Table 1.
283:
284: \paragraph{Absorption Lines.}
285: The main characteristic of the {\it{FUSE}} spectrum of BB Dor is the
286: broad Ly$\beta$ absorption feature due to either the exposed WD, the
287: disk at low inclination angle $i$, or possibly both (see the Results
288: section). Absorption features from the higher orders of the Lyman
289: series are also clearly visible, indicating a temperature
290: T$\sim35,000$K. The other main absorption features expected at this
291: temperature are the C\,{\sc iii} (1175 \AA), Si\,{\sc iii}
292: ($\approx$1108--1114\AA\AA\ and $\approx$1140--1144\AA\AA), Si\,{\sc
293: iv} ($\approx$1120--1130\AA\AA) and the C\,{\sc ii} (1010\AA). The
294: spectrum, is however, also marked by some high-order ionization
295: absorption lines such as S\,{\sc iv} (1063 \& 1073\AA), Si\,{\sc iv}
296: (1066.6\AA), and the O\,{\sc vi} doublet. The S\,{\sc iv} (1006\AA),
297: C\,{\sc ii} (1010\AA), and S\,{\sc iv} (1099\AA) absorption multiplets
298: are all unresolved and shallow. However, we have indicated them on
299: the figures, as they appear deeper in the modeling. On the other side
300: the S\,{\sc iv} (1063\AA\ \& 1073\AA) and Si\,{\sc iii}
301: ($\sim1010$\AA) absorption lines appear deeper in the observed
302: spectrum than in the modeling. The presence of the high-order
303: ionization species (such as the oxygen doublet) indicates the presence
304: of a hotter absorbing component above the main FUV emitting region.
305: The P\,{\sc ii} line ($\sim$961\AA ) has been marked in the spectrum,
306: though it is most probably a feature due to the noise.
307: All the the lines are listed in Table 1 with their wavelengths.
308: Most of the broad absorption lines and features associated with
309: the source are redshifted by about 0.6-0.9\AA , corresponding to
310: a receding velocity of $200 \pm 40$km$~$s$^{-1}$. It seems very
311: likely that this redshif is due to the orbital motion of the WD,
312: however, we do not have enough time tag data to verify this assumption.
313:
314:
315: \paragraph{Emission Lines.}
316: The spectrum of BB Dor exhibits sharp emission lines including all the
317: orders of the Lyman series, N\,{\sc iv} ($\sim$923\AA), S\,{\sc vi}
318: (933.5\AA\ \& 944.5\AA), C\,{\sc iii} (977\AA), He\,{\sc ii} (992\AA),
319: and He\,{\sc i} (1068\AA).
320: Some of these lines seems to be due to
321: helio-coronal emission (sunlight reflected inside the telescope) that
322: contaminates the SiC channels, in particular the H\,{\sc i} lines, the
323: He\,{\sc i} \& He\,{\sc ii} lines, the C\,{\sc iii} (977\AA) line, and
324: the O\,{\sc vi} doublet.
325: The segments of the spectrum from the LiF
326: channels overlapping the SiC Channels down to about 1000\AA\ do not
327: show any oxygen lines, and we therefore did not include these spectral
328: regions of the SiC channels. However, other emission lines may be
329: solely from the source. In many NLs, the emission lines from the disk
330: are broadened by the Keplerian velocity and are consequently easily
331: identified. However, \citet{che01} have detected very narrow Balmer
332: lines and concluded that BB Dor is most likely a low-inclination
333: system. Because of this, the identification of the emission lines
334: from the source is not trivial. This issue is further complicated by
335: the lack of NIGHT exposure and the very low signal-to-noise ratio,
336: which makes it difficult to analyze the {\it two-dimensional image} of
337: the spectrum for the effect of scattered sunlight on the emission
338: lines.
339:
340: %
341: Since the system is a NL in a high state, we expect at first the N\,{\sc iv}
342: ($\approx$923\AA), S\,{\sc vi} (933.5 \& 944.5 \AA), and C\,{\sc iii}
343: (977\AA) emission lines to be from the source with some possible
344: helio-coronal contamination mostly affecting the carbon emission
345: lines. The N\,{\sc iv} (923.06\AA) emission line is contaminated with
346: the H\,{\sc i} (923.15\AA) line; the N\,{\sc iv} (922.52\AA\ +
347: 924.28\AA) lines are not.
348: However, the complete absence of narrow emission lines from the
349: oxygen doublet and the C\,{\sc iii} (1175) seems to indicate
350: otherwise --- i.e. all the sharp emission lines might be helio-coronal
351: in origin. To further check this possibility we measured the narrow
352: emission line relative intensities below 1000\AA and compared
353: with those of the solar spectrum \citep{cur01}. We
354: found a good agreement with the solar-disk average quiet Sun data,
355: except for
356: the SVI 944.5 which has a higher relative intensity in the spectrum
357: of BB Dor. This could be due sun spot activity during the FUSE observation
358: of BB Dor, as this line intensity increases by a factor of 30 inside
359: sun spots while the other lines increase at most by a factor 3.
360: We therefore conclude that all the sharp emission lines in the {\it FUSE}
361: spectrum of BB Dor are due to sun light reflected inside the telescope.
362:
363:
364: \section{Spectral Modeling}
365:
366: We created a grid of models of synthetic spectra of WDs and accretion
367: disks for different values of the WD temperature $T_{eff}$, gravity
368: $Log(g)$, projected rotational velocity $V_{rot} \sin(i)$, inclination
369: $i$, mass accretion rate $\dot{M}$, and abundances. We then ran a
370: $\chi^2$ fitting program to find the best fit for (i) a single WD
371: component, (ii) a single accretion disk component, and (iii) a
372: combined WD + accretion disk. We describe how we generate these
373: synthetic spectra and how we perform the fitting.
374:
375: We create the synthetic model spectra for high-gravity stellar
376: atmospheres using codes TLUSTY and SYNSPEC\footnote{
377: http://nova.astro.umd.edu; TLUSTY version 200, SYNSPEC version 48}
378: \citep{hub88,hub95}. Atmospheric structure is computed (using TLUSTY)
379: assuming a H-He LTE atmosphere; the other species are then added in
380: the spectrum synthesis stage using SYNSPEC. We generate photospheric
381: models with effective temperatures ranging from $\sim$20,000--50,000 K
382: in increments of 1,000 K. We chose values of $Log(g)$ ranging between
383: 7.5 and 9.0. We also varied the stellar rotational velocity $V_{rot}
384: sin(i)$ from $100$km$~$s$^{-1}$ to $1000$km$~$s$^{-1}$ in steps of
385: $100$km$~$s$^{-1}$ (or smaller if needed). The WD rotation ($V_{rot}
386: sin(i)$) rate is determined by fitting the WD model to the spectrum
387: while paying careful attention to the line profiles in the {\it{FUSE}}
388: spectrum. We do not carry out separate fits to individual lines but
389: rather try to fit the lines and continuum simultaneously while paying
390: careful attention to the absorption lines. It is important to note
391: that the depth of the absorption features depends not only on the
392: abundances but also on the rotational velocity. Increasing the
393: rotational velocity reduces the central depths of the absoption
394: features, thus reducing the abundance. However, the widths of the
395: absorption features also increase with increasing rotational velocity.
396: As a consequence, abundances and rotational velocity effects
397: are interwined and cannot always be easily separated without
398: additional information on either parameter.
399: For any WD mass there is a corresponding radius, or
400: equivalently one single value of $Log(g)$ (e.g. see the mass radius
401: relation from \citet{ham61}, or see \citet{woo90,pan00} for different
402: composition and non-zero temperature WDs). Therefore, by scaling the
403: theoretical spectrum to the observed one, we obtain the distance to
404: the system.
405:
406: We model accretion disk spectra by first assuming that the disk is
407: made of a collection of annuli, where each annulus has a temperature
408: $T(r)$ and gravity $Log(g(r))$ given by the standard disk model
409: \citep{sha73,pri81}, for a given central mass $M_{WD}$ and accretion
410: rate $\dot{M}$. A variant of the code TLUSTY is then used, TLDISK,
411: that generates an atmosphere model for each annuli that is then used
412: as input for the code SYNSPEC. The contribution of all the annuli are
413: then combined using the code DISKSYN, and a final spectrum is obtained
414: for any given inclination angle. A detailed explanation of the
415: procedure is given in \citet{wad98}. In the present work we do not
416: use the grid of synthetic accretion disk spectra tabulated by
417: \citet{wad98}, instead we generate them. This allows us to compute
418: disk spectra assuming non-solar abundances and for any inclination
419: angle (the disk spectra of \citet{wad98} have solar abundances and
420: have been generated for given value of the inclination $i$).
421:
422: Before carrying out a synthetic spectral fit of the spectra, we masked
423: portions of the spectra with strong emission lines, strong ISM
424: absorption lines, detector noise, and air glow. The regions excluded
425: from the fit are in blue in Figures 1 and 2.
426:
427: After having generated grids of models for the {\it FUSE} spectrum of
428: BB Dor, we use FIT \citep{numrec}, a $\chi^2$ minimization routine, to
429: compute the reduced $\chi^{2}_{\nu}$ ($\chi^2$ per number of degrees
430: of freedom) and scale factor (which gives the distance) for each model
431: fit. While we use a $\chi^2$ minimization technique, we do not
432: blindly select the least $\chi^2$ models, but we also examine the
433: models that best fit some of the features such as absorption lines
434: and, when possible, the slope of the wings of the broad Lyman
435: absorption features.
436:
437: Initially we generate solar abundances models, and when a good fit is
438: found, we start varying the chemical abundances of C, N, S and Si, to
439: fit the absorption features of the spectrum. In particular, the
440: carbon abundance was set using the C\,{\sc ii} (1010\AA) and C\,{\sc
441: iii} (1175\AA) multiplets; the sulfur abundance was set using the
442: S\,{\sc iv} (1063\AA\ \& 1073\AA) lines; and the silicon was set using
443: the Si\,{\sc iv} (1067\AA\ , 1023\AA\ , 1028\AA) and Si\,{\sc iii}
444: ($\sim$1110\AA\ , $\approx$1138-1146\AA) lines.
445:
446: \section{Results}
447:
448: The data obtained by the AVSON implies that BB Dor has been in a high
449: state around $V \approx 13.6-14.6$ for the last couple of years. We
450: therefore expect the {\it FUSE} spectrum to be dominated by flux from
451: the accretion disk with a relatively high $\dot{M}$. However, in our
452: modeling we follow a systematic approach which consists in fitting the
453: following: (i) a single WD; (ii) a single accretion disk; and (iii) a
454: WD + accretion disk composite.
455:
456: \subsection{White Dwarf.}
457: Since we do not have any information on the mass of the WD, we look
458: for all the best-fit models in the $Log(g)$ versus $T_{eff}$ plane.
459: Namely, for each assumed value of $Log(g)$ we vary the temperature to
460: find the best-fit model. As expected we find that the temperature is
461: somewhere between 34,000K (for $Log(g)=7.5$) and 40,000K (for
462: $Log(g)=9.0$), with a distance between 377 pc and 135 pc, respectively
463: (see Table 1). The least $\chi^2_{\nu}$ is obtained for
464: $Log(g)=8.0-8.65$. We chose the intermediate value $Log(g)=8.3$
465: model, with T=37,000K and d=247 pc and solar abundances, to illustrate
466: our results in Figure 1. We then further improve the fit by varying
467: the abundances. We find that in order to better fit the sulfur and
468: silicon lines, we have to set the sulfur to 20 times its solar
469: abundance, and the silicon to 3 times, while all the other species are
470: kept at solar abundances. The $\chi^2_{\nu}$ value decreases from
471: 0.3348 to 0.3129 (by about 7\%). While some of the lines are better
472: fitted (such as S\,{\sc iv} 1063\AA\ \& 1073\AA; Si\,{\sc iii}
473: $\sim$1010\AA\ \& Si\,{\sc iv} $\sim$1025\AA), the S\,{\sc iv}
474: (1006\AA\ \& 1100\AA) are far too deep (see Figure 2).
475:
476: \subsection{Accretion Disk.}
477: Next, we fit the solar abundances disk models. We find that a low
478: inclination is needed in order to match the absorption lines, and in
479: our models we initially set $i$=5, 8, 12, and 18 degrees. Again,
480: because we have no information on the WD mass, for each value of
481: $Log(g)$ (ranging between 7.5 and 9.0) we vary the mass accretion rate
482: between $10^{-10.5}M_{\odot}$yr$^{-1}$ and
483: $10^{-8.0}M_{\odot}$yr$^{-1}$ to find the best-fit models. Again we
484: find that the best-fit models are around $Log(g) = 8.3$, but the
485: difference between the $\chi^2_{\nu}$ values is not significant, and
486: the improvement over the single WD models is also only marginal (of
487: the order of $\sim$1\% in the $\chi^2_{\nu}$ value). All these models
488: are presented in Table 2. Next, we decide to improve the disk model
489: by varying the abundances, though we do not expect the fit to improve
490: much because of the Keplerian velocity broadening of the disk.
491: However, we find that setting the silicon and sulfur to 3 and 20 times
492: solar, resp., actually reduced the $\chi^2_{\nu}$ value by 10\%, more
493: than for the single WD model. This is because the the S\,{\sc iv}
494: (1006\AA\ \& 1100\AA) lines have better fits (Figure 3). This
495: best-fit single disk model has $i=8^{\circ}$, $\dot{M} = 10^{-9}
496: M_{\odot}$yr$^{-1}$, a distance of 665 pc and $\chi^2_{\nu}=0.3000$.
497:
498: \subsection{Composite Model: WD + Accretion Disk.}
499:
500: Last, we fit composite WD+Disk models. Since the number of models
501: increases exponentially when we change $Log(g)$, $T_{eff}$ and
502: $\dot{M}$, we restrict our search using our best fit for $Log(g)$,
503: namely 8.3. We also set the sulfur to 20 times solar and silicon 3
504: times solar. This procedure reduces the number of free parameters to
505: 3 ($\dot{M},T_{eff},i$).
506:
507: \paragraph{Low Inclination.} Since the
508: system is believed to have a low inclination angle \citep{che01}, we
509: generate low-inclination models ($i=5^{\circ},~8^{\circ},~
510: 12^{\circ},$ \& $18^{\circ}$). Not surprisingly, the best fit model
511: has $i=8^{\circ}$ and $\dot{M}=10^{-9}M_{\odot}$yr$^{-1}$, but this
512: time the WD has a temperature of 32,000K and provides only 10\% of the
513: flux, while the disk provides the remaining 90\%. Such a model again
514: brings an insignificant improvement in the value of $\chi^2_{\nu}$ and
515: it is clear (Figure 4) that it is barely distinguishable from the
516: best-fit single disk model. From the point of view of the physics,
517: this model is preferred because, if the system has a low-inclination
518: angle, then the emission from the WD must contribute to the flux.
519: Models with a WD temperature $T<30,000$K had a WD contribution of only
520: a few percent of the total FUV flux, and could not be distinguished
521: from the single disk models. If the WD has a temperature
522: $T_{eff}<$30,000K it will not be detected while the system is in a
523: high state (with $\dot{M}=10^{-9}M_{\odot}$yr$^{-1}$). These results
524: implies that the contribution from the WD is not very large and that
525: the temperature of the WD must be $\leq 32,000$ K.
526:
527: \paragraph{High Inclination.}
528: From the sharp Balmer emission lines, \citet{che01} suggest the system
529: has a low inclination. However, we cannot confirm the inclination
530: directly from the emission lines of the {\it FUSE} spectrum, as {\it
531: all} the sharp emission lines in the {\it FUSE} spectrum are of
532: helio-coronal origin. Our low-inclination angle single disk models
533: provide a slightly better fit than the single WD models (and a much
534: better fit than the high-inclination single disk models), however, for
535: completeness we include here the results from the composite WD + disk
536: model fits when the assumption on the inclination angle is relaxed.
537:
538: We search for the best-fit WD+disk models (assuming $Log(g)=8.3$) in
539: the parameter space $T_{eff}$ versus $\dot{M}$ including {\it all}
540: inclination angles, and find that the models with an intermediate
541: inclination ($i=18^{\circ}$, $41^{\circ}$, $60^{\circ}$, \&
542: $75^{\circ}$) do not provide the best fit. The best-fit model has
543: $i=80^{\circ}$, and reflects a situation in which the WD is dominant
544: with $T=37,000$K and contributes 3/4 of the total flux, while the disk
545: has $\dot{M}=10^{-9}M_{\odot}$yr$^{-1}$ and contributes only 1/4 of
546: the flux. The distance obtained from this model is 246 pc and
547: $\chi^2_{\nu}=0.2873$. This is the least $\chi^2$ value we obtained
548: from all our models. This model is presented in figure 5. Similar
549: results were obtained when assuming $Log(g)=8.65$, but with a slightly
550: lower mass accretion rate (Table 2).
551:
552: \section{Discussion and Conclusion}
553:
554: BB Dor is a little-known southern NL, and, consequently, both the
555: distance and the mass of the WD are unknown, which implies a larger
556: uncertainty in the results. In addition the {\it FUSE} spectrum is
557: definitely of poor quality. In theory, a fine tuning of the
558: temperature (say to an accuracy of about $\pm$50 K) and mass accretion
559: rate can be carried out by fitting the flux levels such that the
560: distance to the system (when known) is obtained accurately. However,
561: the fitting to the distance depends strongly on the radius (and
562: therefore the mass) of the WD. We discuss some additional
563: restrictions that we use to constrain the properties of the system.
564:
565: On the basis of the least $\chi^2_{\nu}$, the best model is the high
566: inclination WD+disk composite. However, (see Figure 5), because of
567: the high inclination, the disk contributes a rather flat component all
568: the way into the shorter wavelengths. At such a high inclination, one
569: would not expect the WD to dominate the flux, but rather the WD would
570: be almost completely masked by the swollen disk; actually the
571: system would likely be observed to undergo eclipses --- but none
572: have been observed. Also, since BB Dor
573: can vary as much 0.4 mag in one hr, this means that the light cannot
574: be dominated by the WD, and the disk must be contributing at least
575: 40\% of the light. In other words, while the least $\chi^2$ indicates
576: that the best fit is a WD with a rather flat disk component, there are
577: other indications that this cannot be correct. Actually, the WD+disk
578: high inclination model strikingly resembles the second component
579: observed in the {\it FUSE} spectra of some dwarf novae during
580: quiescence (e.g. VW Hyi, \citet{god04}). This is likely a caveat in
581: the {\it state-of-the-art} spectral modeling, rather than an
582: indication of a physical link between the the spectrum of BB Dor and
583: that of a DN in quiescence. The need for improved modeling also stems
584: from the difficulty in producing a model that fits low and high-order
585: ionization lines at the same time. For BB Dor it is possible that the
586: S\,{\sc iv} 1063\AA\ \& 1073\AA\ absorption lines form in the same
587: hotter region/layer where the O\,{\sc vi} doublet form; while the
588: C\,{\sc iii} and Si\,{\sc iii} lines form in a cooler region/layer
589: where the Lyman series (and continuum) form. This is similar to the
590: HST/STIS spectrum of TT Crt \citep{sio08} which exhibits a rich
591: variety of absorption lines from different ionization stages,
592: suggesting line formation in (at least) two different temperature
593: regions.
594:
595: In order to reduce the size of the domain for which we have best-fit
596: models in the parameter space, we use the infrared magnitudes J, H, \&
597: K from the {\it Two Micron All Sky Survey} (2MASS) to assess the
598: distance to BB Dor as prescribed by \citet{kni06} for systems with
599: $P<6$ h. The IR data were collected on 9 November 1999, at a time
600: when BB Dor was in a high state with a visual red magnitude R=14.60
601: and a blue magnitude B=13.90 (whereas B$\sim$16.5 in the low state).
602: The {\it 2MASS} IR apparent magnitudes are J=14.322, H=14.089, and
603: K=14.053. For a primary star $M_{wd}=0.75 M_{\odot}$ and period of
604: 3.559 h (corresponding to BB Dor), the donor star mass is
605: $M_2=0.25 M_{\odot}$
606: (\citet{pat05a} suggest $M_2=0.256 M_{\odot}$)
607: and the IR absolute magnitude estimates are $M_J=7.47$,
608: $M_H=6.90$ and $M_K=6.63$. Inserting these into eq.(15) of
609: \citet{kni06} gives a distance of 235, 274 \& 305 pc, respectively.
610: These distances are typically underestimated by factors of 2.05, 1.86
611: and 1.75 for the J, H and \& K bands giving distances of 482, 510, and
612: 534 pc, respectively, assuming the donor star contributes only $\sim
613: 1/4 - 1/3$ of the total IR flux \citep{kni06}. The distance to BB Dor
614: is therefore certainly larger than 300 pc, and most likely in the
615: range of $\sim 500$ pc. Since BB Dor was observed in a high state
616: (with 2MASS), it is likely that the infrared flux of the donor star is
617: on the lower side, namely even less than 1/4, in which case the 500 pc
618: itself is only a lower limit. Using this restriction on the distance,
619: we see that we can safely reject the single WD models (with
620: $d<400$pc), the high inclination models (with $d<250$pc), and the
621: large mass $Log(g)=9.0$ models (with $d=361$pc and less).
622:
623: On the basis of the distance alone, the low inclination disk and
624: WD+disk models are the best fits. These are also the best fit on the
625: basis of the optical spectrum (low inclination) and visual magnitude
626: (high state/disk).
627:
628: We are now able to summarize the basic characteristics of the system
629: and its WD. We are confident in the fact that the system must have a
630: WD with $Log(g) = 8.3\pm0.3$, $T_{eff}=32,000$K {\it or lower}
631: (disk+WD composite models), a mass accretion rate of the order of
632: $\dot{M}=10^{-9}M_{\odot}$yr$^{-1}$ (disk and disk+WD models) and a
633: distance somewhere between 500 and 700 pc (choosing the least $\chi^2$
634: models agreeing with the distance). From the best-fit single disk
635: models alone we find that the inclination angle must be very small
636: ($\sim8^{\circ}$) to fit the absorption lines. There is an indication
637: that the sulfur and silicon are over abundant and it is also possible
638: that some of the high-order ionization lines form in a layer in front
639: of the disk and the WD.
640:
641: Although the {\it FUSE} spectrum of BB Dor is rather poor and
642: multiwavelength data are limited, we have narrowed down the region in
643: the parameter space ($\dot{M}$, $P$, $T_{eff}$) for this system. BB
644: Dor is now the sixth VY Scl NL variable with a temperature estimate
645: for its WD. With a temperature of $\leq 32,000$ K, the WD of BB Dor
646: marks the lower end of the temperature distribution for WDs of VY Scl
647: NL variables, which, so far, for the 5 other systems, ranged between
648: 40,000K and 47,000K \citep{ham08}. The {\it FUSE} spectrum of BB Dor
649: clearly shows the drop in flux in the shorter wavelengths, the
650: signature of a moderate temperature when compared to the {\it FUSE}
651: spectra of these other VY Scl systems (see e.g. V794 Aql,
652: \citep{god07}). Even if we consider the single WD best-fit model, the
653: WD temperature of BB Dor only reaches 37,000K (and the 40,000K
654: $Log(g)=9.0$ model gives an unrealistic close distance of 135pc in
655: disagreement with the expected IR emission from the secondary). In
656: that respect, the data point in the ($T_{eff},P$) plane for BB Dor
657: falls {\it under} the region of the VY Scl NL variables, seemingly
658: {\it into} the region of the parameter space populated with dwarf
659: novae \citep{sio08}. However, as shown in Figure 6, the data point
660: for BB Dor does not seem to stand apart, and the separation between VY
661: Scl systems and DN can be made easily by drawing a diagonal line. In
662: that respect, it is actually V794 Aql that divides the two regions of
663: the graph as indicated by the slanted line.
664:
665: \acknowledgments
666: We thank the anonymous referee for a prompt, concise
667: and pertinent report which helped improve the analysis of the
668: lines.
669: Special thanks to the Austral Variable Star Observer Network (AVSON),
670: and in particular to Berto Monard for providing us with the visual
671: magnitude of BB Dor between 02 Sep., 2005 and 30 Dec., 2007, and to
672: Peter Nelson for a quick look V band photometry at BB Dor on 28 Jan.,
673: 2008. PG wishes to thank Mario Livio, for his kind hospitality at the
674: Space Telescope Science Institute, where part of this work was done.
675: The infra-red J, H, \& K magnitudes of BB Dor were retrieved from the
676: online archival data of the {\it 2MASS} at IPAC.
677: Part of this publication makes use of the data products from the Two Micron
678: All Sky Survey, which is a joint project of the University of
679: Massachusetts and the Infrared Processing and Analysis Center/California
680: Institute of Technology, funded by the National Aeronautics and
681: Space Administration and the National Science Foundation.
682: This research was
683: based on observations made with the NASA-CNES-CSA Far Ultraviolet
684: Spectroscopic Explorer. {\it{FUSE}} is operated for NASA by the Johns
685: Hopkins University under NASA contract NAS5-32985. This research was
686: supported by NASA {\it FUSE} grant NNX07AU50G to Villanova University
687: (P. Godon). This work was also supported in part by NSF grant
688: AST05-07514 to Villanova (E.M. Sion).
689:
690: \begin{thebibliography}{}
691:
692:
693: \bibitem[Andronov et al. (2003)]{and03}
694: Andronov, N., Pinsonneault, M., \& Sills, A. 2003, \apj, 582, 358
695:
696: \bibitem[Barrett et al. (2008)]{bar08}
697: Barrett, P., et al., 2008, \apj, in preparation
698:
699: \bibitem[Chen et al. (2001)]{che01}
700: Chen, A., O'Donoghue, D., Stobie, R.S., Kilkenny, D., \& Warner, B. 2001,
701: \mnras, 325, 89
702:
703: \bibitem[Curdt et al. (2001)]{cur01}
704: Curdt, W., Brekke, P., Feldman, U., Wilhelm, K., Dwivedi, B.N.,
705: Sch\"ule, U., \& Lemaire, P. 2001, \aap, 375, 591
706:
707: \bibitem[Dixon et al. (2007)]{dix07}
708: Dixon, W.V., et al. 2007, \pasp, 119, 527
709:
710: \bibitem[Godon et al. (2004)]{god04}
711: Godon, P., Sion, E.M., Cheng, F.H., Szkody, P., Long, K.S., \& Froning, C.S.
712: 2004, \apj, 612, 429
713:
714: \bibitem[Godon et al. (2007)]{god07}
715: Godon, P., Sion, E.M., Barrett, P.E., \& Szkody, P. 2007,
716: \apj, 656, 1092
717:
718: \bibitem[Godon et al. (2008)]{god08}
719: Godon, P., Sion, E.M., Barrett, P.E., Hubeny, I., Linnell, A.P.,
720: \& Szkody, P. 2007, \apj, 679, 1447
721:
722: \bibitem[Hamada \& Salpeter (1961)]{ham61}
723: Hamada, T., \& Salpeter, E.E. 1961, \apj, 134, 683
724:
725: \bibitem[Hamilton \& Sion (2008)]{ham08}
726: Hamilton, R.T., \& Sion, E.M. 2008, \pasp, 120, 165
727:
728: \bibitem[Howell et al. (2001)]{how01}
729: Howell, S., Nelson, L, \& Rappaport, S. 2001, \apj, 550, 897
730:
731: \bibitem[Howk (2008)]{how08}
732: Howk, J. 2008, private communication
733:
734: \bibitem[Hubeny (1988)]{hub88}
735: Hubeny, I. 1988, Comput. Phys. Commun., 52, 103
736:
737: \bibitem[Hubeny \& Lanz (1995)]{hub95}
738: Hubeny, I.,\& Lanz, T. 1995, \apj, 439, 875
739:
740: \bibitem[Knigge (2006)]{kni06}
741: Knigge, C. 2006, \mnras, 373, 484
742:
743: \bibitem[Koester et al. (1985)]{koe85}
744: Koester, D., Weidemann, V., Zeidler-K.T., E.M., \& Vauclair, G. 1985,
745: A\&A, 142, L5
746:
747: \bibitem[Nelan \& Wegner (1985)]{nel85}
748: Nelan, E.P. \& Wegner, G. 1985, \apj, 289, L31
749:
750: \bibitem[Panei et al. (2000)]{pan00}
751: Panei, J.A., Althaus, L.G., \& Benvenuto, O.G. 2000, A\&A, 353, 970
752:
753: \bibitem[Patterson (2002)]{pat02}
754: Patterson, J., 2002, Center for Backyard Astrophysics Communication
755: ( http://cbastro.org/communications/news/2002/november21-a.html )
756:
757: \bibitem[Patterson et al. (2005)]{pat05a}
758: Patterson, J. et al. 2005, \pasp, 117, 1204
759:
760: \bibitem[Patterson et al. (2005)]{pat05b}
761: Patterson, J., Thorstensen, J.R., \& Kemp, J. 2005, \pasp, 117, 427
762:
763: \bibitem[Press et al.(1992)]{numrec}
764: Press, W.H., Teukolsky, S.A., Vetterling, W.T., Flannery, B.P.,
765: Numerical Recipes in Fortran 77, The Art of Scientific Computing,
766: Second Edition, 1992, Cambridge University Press.
767:
768: \bibitem[Pringle (1981)]{pri81}
769: Pringle, J.E. 1981, \araa, 19, 137
770:
771: \bibitem[Schlegel et al. (1998)]{sch98}
772: Schegel, D.J., Finkbeiner, D.P., \& Davis, M. 1998, \apj, 500, 525
773:
774: \bibitem[Shakura \& Sunyaev (1973)]{sha73}
775: Shakura, N.I., \& Sunyaev, R.A. 1973, A\&A, 24, 337
776:
777: \bibitem[Sion (1995)]{sio95}
778: Sion, E.M. 1995, \apj, 438, 876
779:
780: \bibitem[Sion et al. (2008)]{sio08}
781: Sion, E.M., B\"ansicke, B.T., Long, K.S.,
782: Szkody, P., Knigge, C., Hubeny, I.,
783: deMartino, D., \&
784: Godon, P. 2008, \apj, 679, n.1, in press
785:
786: \bibitem[Skrutskie et al. (2006)]{skr06}
787: Skrutskie, M.F. 2006, \aj, 131, 1163
788:
789: \bibitem[Wade \& Hubeny (1998)]{wad98}
790: Wade, R.A., \& Hubeny, I. 1998, \apj, 509, 350
791:
792: \bibitem[Warner (1995)]{war95}
793: Warner, B. 1995, Cataclysmic Variable Stars (Cambridge: Cambridge Univ.
794: Press)
795:
796: \bibitem[Wood (1990)]{woo90}
797: Wood, M.A. 1990, Ph.D. thesis, University of Texas at Austin
798:
799: \end{thebibliography}
800:
801: \begin{deluxetable}{cccc}
802: \tablewidth{0pc}
803: \tablecaption{ {\it FUSE} Line Identification}
804: \tablehead{
805: Ion & $\lambda_{rest}$ (\AA) & $\lambda_{obs}$ (\AA ) & Comments$^a$ }
806: \startdata
807: H\,{\sc i} & 918.13 & 918.0 & c/e \\
808: & 919.35 & 919.3 & c/e \\
809: & 920.96 & 920.9 & c/e \\
810: N\,{\sc iv} & 922.52 & 922.3 & s/e \\
811: & 923.06 & 923.0 & s,c/e \\
812: H\,{\sc i} & 923.15 & 923.2 & c/e \\
813: N\,{\sc iv} & 923.22 & 923.2 & c/e \\
814: & 924.28 & 924.2 & c/e \\
815: H\,{\sc i} & 926.23 & 926.2 & c/e \\
816: & 930.75 & 930.7 & c/e \\
817: S\,{\sc vi} & 933.50 & 933.4 & c/e \\
818: H\,{\sc i} & 937.80 & 937.8 & c/e \\
819: S\,{\sc vi} & 944.50 & 944.4 & c/e \\
820: H\,{\sc i} & 949.74 & 949.7 & c/e \\
821: N\,{\sc i} & 952.40 & 925.5 & ism/a \\
822: & 953.42 & --- & --- \\
823: & 953.66 & --- & --- \\
824: & 953.97 & --- & --- \\
825: & 954.10 & 954.2 & ism/a \\
826: P\,{\sc ii} & 961.04 & 961.2 & ism/a? \\
827: H\,{\sc i} & 972.54 & 972.5 & c/e \\
828: C\,{\sc iii} & 977.02 & 977.0 & c/e \\
829: Si\,{\sc iii}& 993.52 & 993.2 & s?/a \\
830: & 994.19 & 994.9 & s?/a \\
831: & 997.39 & 997.1 & s?/a \\
832: S\,{\sc iv} & 1006.07 & --- & s/a - unresolved \\
833: & 1006.39 & --- & s/a - unresolved \\
834: C\,{\sc ii} & 1009.85 & --- & s/a - unresolved \\
835: & 1010.08 & --- & s/a - unresolved \\
836: & 1010.37 & --- & s/a - unresolved \\
837: H\,{\sc i} & 1025.77 & 1025.7 & c/e \\
838: O\,{\sc vi} & 1031.91 & 1032.9 & s/a \\
839: C\,{\sc ii} & 1036.34 & 1036.4 & ism/a \\
840: O\,{\sc vi} & 1037.61 & 1038.3 & s/a \\
841: O\,{\sc i} & 1039.10 & 1039.3 & ism/a \\
842: Ar\,{\sc i} & 1048.20 & 1048.3 & ism/a \\
843: S\,{\sc iv} & 1062.65 & 1063.4 & s/a \\
844: Si\,{\sc iv} & 1066.60 & 1067.1 & s/a \\
845: Ar\,{\sc i} & 1066.66 & 1066.8 & ism/a \\
846: S\,{\sc iv} & 1072.97 & 1073.9 & s/a \\
847: & 1073.52 & 1073.9 & s/a \\
848: S\,{\sc iv} & 1098.36 & --- & s/a - unresolved \\
849: & 1098.93 & --- & s/a - unresolved \\
850: & 1099.48 & --- & s/a - unresolved \\
851: & 1100.53 & --- & s/a - unresolved \\
852: Si\,{\sc iii} & 1108.36 & 1108.8 & s/a \\
853: & 1109.94 & 1110.6 & s/a \\
854: & $\sim$1113.20 & 1113.8 & s/a \\
855: Si\,{\sc iv} & 1122.49 & 1123.0 & s/a \\
856: & $\sim$1128.33 & 1128.9 & s/a \\
857: N\,{\sc i} & 1134.16 & 1134.3 & c/a \\
858: & 1134.42 & 1134.3 & c/a \\
859: & 1134.98 & 1135.2 & c/a \\
860: S\,{\sc i} & 1145.10 & 1145.0 & ism/a \\
861: He\,{\sc ii} & 1168.61 & 1168.6 & c/e \\
862: C\,{\sc iii} & 1174.90 & 1176.4 & s/a - unresolved \\
863: & 1175.26 & 1176.4 & s/a - unresolved \\
864: & 1175.60 & 1176.4 & s/a - unresolved \\
865: & 1175.71 & 1176.4 & s/a - unresolved \\
866: & 1176.00 & 1176.4 & s/a - unresolved \\
867: & 1176.40 & 1176.4 & s/a - unresolved \\
868: \enddata
869: \tablenotetext{a}{The following abbreviations have been used:
870: a - for absorption; e - for emission; c - for contamination
871: (e.g. air glow; helio-coronal emission); s - for source; and
872: ism - for interstellar medium.}
873: \end{deluxetable}
874:
875: \begin{deluxetable}{cccccccccccc}
876: \tablewidth{0pc}
877: \tablecaption{Synthetic Spectra}
878: \tablehead{
879: $Log(g)$&$T_{wd}$&$V_{rot}sin{i}$&[Si] &[S] &[Z] & $i$ & $Log(\dot{M})$ &WD/Disk& d &$\chi^2_{\nu}$& Fig. \\
880: (cgs) &($10^3$K)&(km/s)&Solar & Solar & Solar & (deg) & ($M_{\odot}$/yr) &(\%) &(pc)& & \\
881: }
882: \startdata
883: 7.50 & 34.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 377 & 0.3427 & \\
884: 8.00 & 36.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 266 & 0.3361 & \\
885: 8.30 & 37.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 247 & 0.3348 & 1 \\
886: 8.30 & 37.0 & 400 & 3.0 & 20. & 1.0 & --- & --- & 100/0 & 211 & 0.3129 & 2 \\
887: 8.50 & 38.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 193 & 0.3345 & \\
888: 8.50 & 38.0 & 400 & 3.0 & 20. & 1.0 & --- & --- & 100/0 & 184 & 0.3100 & \\
889: 8.65 & 38.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 172 & 0.3360 & \\
890: 9.00 & 40.0 & 400 & 1.0 & 1.0 & 1.0 & --- & --- & 100/0 & 135 & 0.3390 & \\
891: 7.50 & --- & --- & 1.0 & 1.0 & 1.0 & 12 & -8.0 & 0/100 & 878 & 0.3346 & \\
892: 7.88 & --- & --- & 1.0 & 1.0 & 1.0 & 12 & -8.5 & 0/100 & 878 & 0.3346 & \\
893: 8.30 & --- & --- & 1.0 & 1.0 & 1.0 & 8 & -9.0 & 0/100 & 693 & 0.3297 & \\
894: 8.30 & --- & --- & 3.0 & 20. & 1.0 & 8 & -9.0 & 0/100 & 665 & 0.3000 & 3 \\
895: 8.65 & --- & --- & 1.0 & 1.0 & 1.0 & 5 & -9.5 & 0/100 & 496 & 0.3332 & \\
896: 9.00 & --- & --- & 1.0 & 1.0 & 1.0 & 5 & -10.0 & 0/100 & 361 & 0.3376 & \\
897: 8.30 & 32.0 & 400 & 3.0 & 20. & 1.0 & 08 & -9.0 & 10/90 & 700 & 0.2989 & 4 \\
898: 8.30 & 37.0 & 400 & 3.0 & 20. & 1.0 & 80 & -9.0 & 74/26 & 246 & 0.2873 & 5 \\
899: 8.65 & 38.0 & 400 & 3.0 & 20. & 1.0 & 80 & -9.5 & 75/25 & 184 & 0.2880 & \\
900: \enddata
901: \end{deluxetable}
902:
903:
904: \clearpage
905:
906: \begin{figure}
907: \plotone{f1.ps}
908: \vspace{-5.cm}
909: \figcaption{Solar abundance single WD model. The best-fit WD model
910: assuming solar abundances (solid black line) is shown together with
911: the {\it{FUSE}} spectrum of BB Dor (light grey/red line). The
912: segments of the spectrum that have been masked before the fitting are
913: in dark grey/blue. The WD has a temperature $T=37,000$K, a projected
914: rotational velocity $V_{rot} \sin{i} =400$km/s, and $Log(g)=8.3$. The
915: distance is $d=$217 pc, and $\chi^2_{\nu}=0.3348$.}
916: \end{figure}
917:
918: \clearpage
919:
920: \begin{figure}
921: \plotone{f2.ps}
922: \vspace{-5.cm}
923: \figcaption{Non-solar abundance single WD model. The best-fit WD model
924: to the {\it{FUSE}} spectrum of BB Dor is shown where the abundances of
925: sulfur and silicon have been set to: [S]=20 solar and [Si]=3 solar.
926: The abundances of all the other elements have been set to solar. The
927: WD has a temperature $T=37,000$K, a projected rotational velocity
928: $V_{rot} \sin{i} =400$km/s, $Log(g)=8.3$. The distance obtained is
929: $d=$211 pc, and $\chi^2_{\nu}=0.3129$.}
930: \end{figure}
931:
932:
933:
934: \clearpage
935:
936: \begin{figure}
937: \plotone{f3.ps}
938: \vspace{-5.cm}
939: \figcaption{Non-solar abundance single disk model. The {\it{FUSE}}
940: spectrum of BB Dor is shown together with a one of the best-fit
941: synthetic disk models. The model has $M=0.80M_{\odot}$,
942: $\dot{M}=10^{-9}M_{\odot}$/yr, $i=8^{\circ}$, and gives a distance of
943: $d=665$ pc and a $\chi^2_{\nu}=0.3000$. The sulfur and silicon
944: abundances have been set to: [S]=20 solar and [Si]=3 solar. The
945: abundances of all the other elements have been set to solar. The low
946: inclination is needed in order to fit the silicon, sulfur and carbon
947: absorption lines.}
948: \end{figure}
949:
950:
951: \clearpage
952:
953: \begin{figure}
954: \plotone{f4.ps}
955: \vspace{-5.cm}
956: \figcaption{Low-inclination non-solar abundance WD+disk model. The
957: {\it{FUSE}} spectrum of BB Dor is shown together with a one of the
958: best-fit WD+disk models (solid black line), assuming a low inclination
959: angle. The model has $M=0.80M_{\odot}$, $\dot{M}=10^{-9}M_{\odot}$/yr,
960: $i=8^{\circ}$, and gives a distance of $d=700$ pc and a
961: $\chi^2_{\nu}=0.2889$. The sulfur and silicon abundances have been set
962: to: [S]=20 solar and [Si]=3 solar. The abundances of all the other
963: elements have been set to solar. The WD (dotted line) contributes
964: 10\% of the flux while the high-inclination disk (dashed line)
965: contributes the remaining 90\%.}
966: \end{figure}
967:
968: \clearpage
969:
970: \begin{figure}
971: \plotone{f5.ps}
972: \vspace{-5.cm}
973: \figcaption{High-inclination non-solar abundance WD+disk model. The
974: {\it{FUSE}} spectrum of BB Dor is shown together with a one of the
975: best-fit WD+disk models (solid black line). The model has
976: $M=0.80M_{\odot}$, $\dot{M}=10^{-9}M_{\odot}$/yr, and gives a distance
977: of $d=246$ pc and a $\chi^2_{\nu}=0.2873$. The inclination angle of
978: the system is kept as a free parameter and the best-fit model gives
979: $i=80^{\circ}$. The sulfur and silicon abundances have been set to:
980: [S]=20 solar and [Si]=3 solar. The abundances of all the other
981: elements have been set to solar. The fit of the silicon, sulfur and
982: carbon absorption lines is obtained by having the WD component
983: dominating the spectrum. The WD (dotted line) contributes 74\% of the
984: flux while the high-inclination disk (dashed line) contributes the
985: remaining 26\%.}
986: \end{figure}
987:
988: \clearpage
989:
990: \begin{figure}
991: \plotone{f6.ps}
992: \vspace{-10.cm}
993: \figcaption{Dwarf nova systems and VY Scl nova-like variables above the
994: Period-gap. The systems have been plotted in the ($T_{eff}$,P)
995: parameter space on a logarithmic scale. The data for the nova-likes
996: are taken from \citet{ham08} and references therein; the data for the
997: dwarf nova systems are taken \citet{sio08} and the references
998: therein. For RU Peg we have used the more recent estimate from
999: \citep{god08}. The data point for V794 Aql is an estimate of the WD
1000: temperature based on an IUE spectrum while the system was in a lower
1001: state \citep{god07}. While the data point of BB Dor is lower than the
1002: other VY Scl system there is still a clear separation between the VY
1003: Scl systems and the DN systems as indicated by the slanted line.}
1004: \end{figure}
1005:
1006: \end{document}
1007:
1008:
1009: