1: \documentclass[referee]{cjaa}
2: \usepackage{graphicx}
3: \input{epsf.sty} %for PS/EPS graphics inclusion, old
4: \input{psfig.sty}
5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
6:
7: \begin{document}
8:
9: \title{Spectroscopic Study of SU UMa-type Dwarf Nova YZ~Cnc during its 2002 Superoutburst}
10: \author{Ying-he Zhao, Zong-yun Li\mailto{}, Xiao-an Wu and Qiu-he Peng}
11:
12: \institute{Department of Astronomy, Nanjing University, Nanjing,
13: 210093, China\\
14: \email{zyli@nju.edu.cn}}
15:
16: \date{Accepted ~, Received ~}
17:
18:
19: \abstract{ We report time-resolved spectroscopic observations of
20: the SU~Ursae Majoris dwarf nova, YZ Cnc, for 2 nights over 11 hrs
21: during its 2002 January superoutburst. The spectra only show
22: absorption-line profiles in the first day. But the lines display
23: blue and red troughs, with ``W'' profiles in the second day. The
24: radial velocity curve of the absorption troughs and emission peaks
25: of H$\beta$ has an amplitude of $49\pm10$ km s$^{-1}$ and a phase
26: offset of $-0.07\pm0.04$, which are very similar to those measured
27: in quiescence; however, the $\gamma$ velocity deviates strongly
28: from the systemic velocity measured in quiescence, showing
29: variation of the order of $\pm$60 km s$^{-1}$. And large shifts of
30: $\sim$70 km s$^{-1}$ and $\sim$0.09, for the orbital-averaged
31: velocity and phase respectively, are also found in our
32: observations. All these phenomena can be well explained with a
33: precession of an eccentric disk and we conclude that these
34: phenomena are the characteristic products of an eccentric
35: accretion disk. \keywords{ accretion, accretion disks --binaries:
36: close -- novae, cataclysmic variables -- stars: dwarf novae --
37: stars: individual (YZ Cancri)} }
38:
39: \titlerunning{Spectroscopic study of YZ Cnc during superoutburst}
40: \authorrunning{Y. H. Zhao, Z. Y. Li, X. A. Wu et al.}
41: \maketitle
42:
43:
44: \section{Introduction}
45: YZ Cnc is a member of the class of SU UMa type dwarf novae, in
46: which normal outbursts (i.e., short outbursts) are occasionally
47: interspersed by longer and brighter distinctive superoutbursts,
48: accompanying superhump phenomena. Superhumps are large amplitude
49: luminosity variations with a period usually a few percent longer
50: than the orbital period of the binary system. And this is
51: generally thought to arise from the interaction of the donor star
52: orbit with slowly progradely precessing non-axisymmetric accretion
53: disk. The eccentricity of the disk arises because a 3:1 resonance
54: occurs between the donor star orbit and motion of matter in the
55: outer disk (for a good review, see Warner 1995).
56:
57: YZ Cnc is remarkable in several respects. Photometric study showed
58: that YZ Cnc has a visual magnitude of $\sim$14.5 in quiescence and
59: $\sim$10.5 in outburst and is one of the most active cataclysmic
60: variables because of the large flickering amplitude of 0.75 mag
61: peak to peak (Moffett \& Barnes 1974). It has a very short
62: recurrence time of $\sim$11.3 days (Vorob'yeva \& Kukarkin 1961).
63: Patterson (1979) discovered superhumps in the light curve with a
64: period of 0.09204 day of YZ Cnc and defined it as an SU UMa type
65: star, and his continuous study (Patterson 1981) with high-speed
66: photometric observations of YZ Cnc found no evidence for coherent
67: oscillations either in quiescence or during eruptions. But van~
68: Paradijs et al. (1994) found that orbital variability was present
69: during quiescence.
70:
71: In X-ray band, YZ Cnc has also been studied very extensively with
72: the Einstein satellite (C\'{o}rdova \& Mason 1984, Eracleous et
73: al. 1991), with EXOSAT (van der Woerd 1987), with the ROSAT PSPC
74: during the ROSAT All Sky Survey and in subsequent pointings
75: (Verbunt et al. 1997, 1999; van Teeseling \& Verbunt 1994), and
76: with XMM-Newton (Hakala et al. 2004).
77:
78: However, spectroscopy has not been as extensive as photometry or
79: X-ray, specially when the system is undergoing superoutburst. The
80: sporadical spectroscopic observations have been performed by
81: several authors as part of general surveys of cataclysmic
82: variables (Szkody 1981, Oke \& Wade 1982, Wade 1982, Williams
83: 1983). Shafter \& Hessman (1988, named SH hereafter) presented a
84: detailed spectroscopic study of YZ Cnc when the star was in
85: quiescence and gave an orbital period of 0.0868(2) day. According
86: to this orbital period and the superhump period given by Patterson
87: (1979), a precessing period of 1.52 day can be obtained. We thus
88: did a 2-days observation to study the accretion disk of YZ Cnc
89: during its 2002 January superoutburst. In this paper we report our
90: observations and reduction of the spectroscopic data in Sect. 2.
91: In Sect. 3 we describe the main characteristics of the
92: spectroscopic results and explanations for these results. In the
93: last two sections we present a brief discussion and conclusions
94: for this work.
95:
96: \section{Observations}
97: The observations were conducted with the Optomechanics Research,
98: Inc., Cassegrain spectrograph attached to the 2.16-m telescope
99: with a TEK1024 CCD camera at Xinglong Station of the National
100: Astronomical Observatory. Total observational time was 11 hrs, 5.3
101: times of the orbital period. A 300 groove mm$^{-1}$ grating blazed
102: at 5000 \AA\ was used, and the slit width was set to 2$''$.5. Dome
103: flats were taken at the beginning and end of each night. Exposure
104: time for the star ranged from 600 to 1800, depending on weather
105: conditions. Fifteen and fourteen star spectra were collected on
106: January 21 and 22 (Beijing time), respectively. The journal of the
107: observations is listed in Table 1.
108:
109: The technique of data processing is similar to that in Wu et al.
110: (2001). After bias subtraction and flat field correction, we used
111: the $IRAF$\footnotemark, \footnotetext{IRAF is distributed by the
112: National Optical Astronomy Observatories, which is operated by
113: Associated of Universities for Research in Astronomy, Inc., under
114: cooperative agreement with the National Science Foundation.} task
115: $cosmicray$ to eliminate the cosmic rays roughly and then used
116: $imedit$ to get the cosmic rays rejected more clearly by hand. The
117: lamp spectra recorded before and after every two successive star
118: exposures were used to interpolate the coefficients of the
119: wavelength scales. We derived a spectral resolution of 12 \AA \
120: from FWHM measurement of the lamp spectra. The rms error of
121: identified lines was less than 0.2 \AA \ using a fourth-order
122: Legendre polynomial to fit the lines, corresponding to 12 km
123: s$^{-1}$ near H$\beta$. The flux was calibrated used the standard
124: star, HD109995, and had an estimate error of $\sim$10\%.
125:
126: %-------------------Table 1------------------------------
127: \begin{table}
128: \caption{Journal of observations.}
129: \begin{tabular}{ccccc}
130: \hline\hline
131: Date (UT) & HJD Start & Duration & Exposure & Plates\\
132: (Year 2002) & -2452000 & (hr) & (s) & \\
133: \hline
134: Jan 21 ........& 296.1457 & 5.22 & 900,1200 & 15\\
135: Jan 22 ........& 297.1064 & 5.78 & 1500 & 14\\
136: \hline
137: \end{tabular}
138: \end{table}
139:
140: \section{Results and Analysis}
141: \subsection{Average Spectra}
142: %
143: Fig. 1 shows the average spectra of YZ Cnc during its
144: superoutburst. The top and lower panel are the sum of all 15
145: individual spectra recorded on January 21 and the sum of all 14
146: individual spectra obtained on January 22, respectively. The
147: spectra of January 21 is characterized by broad Balmer absorption
148: and an energy distribution significantly bluer than that of
149: January 22, when the eruption was fading out. This spectra is
150: typical of an optically thick accretion disk with a high accretion
151: rate. The emission component came out and was specially stronger
152: in H$\beta$ absorption on January 22. The He~I ($\lambda
153: \lambda$4471, 4922) absorption, He~I $\lambda$5015+Fe~II
154: $\lambda$5018 and Fe~II $\lambda$5169 absorption (emission on
155: January 22) are also present (see more clearly in Fig.~2). The
156: equivalent widths of the absorption and emission lines are
157: summarized in Table 2.
158:
159: %----------------------Figure 1-------------------------
160: \begin{figure}
161: \centering
162: \includegraphics[width=13cm]{fig1.eps}
163: \caption{Spectra of YZ Cnc during superoutburst. The top spectrum,
164: obtained on January 21, is much bluer and the flux is also much
165: higher than the bottom one, obtained on January 22. \label{fig1}}
166: \end{figure}
167:
168: %--------------------Table 2-------------------------------
169: \begin{table}
170: \caption{Equivalent widths of spectral lines.}
171: \begin{tabular}{ccccc}
172: \hline\hline
173: Date & Element & EW & Element & EW\\
174: (2003) &Rest Wavelength & (\AA) & Rest Wavelength &(\AA)\\
175: \hline
176: \ &H$\zeta$ $\lambda$3889&-6.2 &H$\beta$ $\lambda$4861&-8.5\\
177: \ &H$\epsilon$ $\lambda$3970 &-6.9 &He~I $\lambda$4471 &-1.7\\
178: Jan 21 &H$\delta$ $\lambda$4101 &-11.3 &He~I $\lambda$4922& -0.9\\
179: \ &H$\gamma$ $\lambda$4340 &-8.2 &He~I $\lambda$5015+Fe~II $\lambda$5018 &-1.7\\
180: \hline
181: \ &H$\zeta$ $\lambda$3889&-4.2 &H$\beta$ $\lambda$4861 (emission)&1.3\\
182: \ &H$\epsilon$ $\lambda$3970 &-4.7 &H$\beta$ $\lambda$4861& -5.2\\
183: Jan 22 &H$\delta$ $\lambda$4101 &-8.9 &He~I $\lambda$4471 &-1.3\\
184: \ &H$\gamma$ $\lambda$4340 &-7.6 &Fe~II $\lambda$5169 (emission) & 0.18\\
185: \hline
186: \end{tabular}
187: \end{table}
188:
189: \subsection{Radial Velocity}
190:
191: In Fig.~2A we show the normalized spectrum which is the sum of all
192: 15 individual spectra obtained on January 21. And Fig.~2B shows
193: the normalized spectrum obtained on January 22. When combined, no
194: radial velocity shift was applied.
195:
196: We measured the centers of H$\beta$ absorption troughs of January
197: 21 and emission peaks of January 22 with Gaussian-fit method. We
198: used H$\beta$ line because it had good signal-to-noise ratios in
199: both nights. Fig.~3 shows the velocities folded on the orbital
200: period with the best-fit sinusoidal curve superposed. The orbital
201: phase was computed according to the ephemeris given by SH,
202: \[T_0=HJD 2,446,113.794+0.0868(2)E\]
203: where $T_0$ is the time of the $\gamma$ crossover from negative to
204: positive velocities and $E$ is a cycle number. The best-fit
205: sinusoidal shows that H$\beta$ has an amplitude, $K$, of 49$\pm$10
206: km~ s$^{-1}$ and a systemic velocity, $\gamma$, of 62$\pm$7 km
207: s$^{-1}$. The value of $K$ is very consistent with the result of
208: SH, 50$\pm$20 km s$^{-1}$, which was measured during the
209: quiescence. But the $\gamma$ velocity is somewhat larger than the
210: value of $\sim$16 km s$^{-1}$ in SH.
211:
212: It is clearly shown in Fig.~3 that there are two abnormal points
213: near phase 0.6 and 1 whose velocities are much larger than the
214: others'. This maybe occur at the accretion flow.
215:
216: %-------------------------Figure 2-----------------------------
217: \begin{figure}
218: \centering
219: \includegraphics[width=13cm]{fig2.eps}
220: \caption{Normalized average spectra of YZ Cnc during
221: superoutburst. (A):observed on January 21; there was no emission
222: component in \emph{all} spectral lines. (B): observed on January
223: 22; almost all Balmer absorptions were partially filled by
224: emission on this day. The emission component also came out in He~I
225: $\lambda$5015+Fe~II $\lambda$5018 absorption. Moreover, the Fe~II
226: $\lambda$5169 had gone into emission. These differences showed
227: that the star was going back to quiescence. \label{fig2}}
228: \end{figure}
229:
230: %----------------------Figure 3---------------------------------
231: \begin{figure}
232: \centering
233: \includegraphics[width=13cm]{fig3.eps}
234: \caption{Least-squares sinusoidal fitted for the radial velocities
235: of the centers of H$\beta$ obtained on January 21 and 22,
236: representing with filled and open triangles respectively. Note
237: that almost all velocities in the first night are larger than
238: those in the second night. And it clearly shows that there are two
239: abnormal points near phase 0.6 and 1 whose velocities are much
240: larger than the others'. This maybe occur at the accretion flow.
241: \label{fig3}}
242: \end{figure}
243:
244: \subsection{An Eccentric Disk}
245: \subsubsection{The variation of $\gamma$ and orbital-averaged velocity}
246: In Fig.~3, we show the radial velocities marked with filled and
247: open triangles, corresponding to January 21 and 22, respectively.
248: It can be seen clearly that there is a systemic discrepancy
249: between these velocities obtained in these two days. We have used
250: the sky emission line 5577\AA\ to check whether this occurred due
251: to the systemic error and the difference between these two days is
252: less than 0.2\AA. So we believe the existence of discrepancy
253: between these two $\gamma$ velocities is real. And we derived
254: these two $\gamma$ velocities from fitting sinusoidal to the data
255: shown in Fig.~3 (excluding the two abnormal points near the phase
256: of 0.6 and 1.0, also see Fig.~4) as 91 km s$^{-1}$ and 34 km
257: s$^{-1}$, for January 21 and 22 respectively.
258:
259: We also measured the centers of H$\beta$ of the average spectra of
260: these two days. And we obtained that the relative velocities,
261: which are averaged throughout the orbital period, are 97 km
262: s$^{-1}$ and 28 km s$^{-1}$, corresponding to January 21 and 22,
263: respectively. This phenomenon was also found in IY UMa (Wu et al.
264: 2001), KS UMa (Zhao et al. 2005a). These two features of our
265: radial velocities are summarized in Table 3.
266:
267: \subsubsection{The large phase shift of $\sim$0.09 between two days}
268: As shown in Fig.~4, there obviously existed a phase shift of
269: $\sim$0.09 between the two sinusoidal velocity curves of January
270: 21 and January 22, as shown by continuous and dashed line
271: respectively. This feature of the radial velocities is discovered
272: for the first time for YZ Cnc, even for SU UMa stars. Such big
273: phase shift in one day could not be due to the uncertainty of the
274: orbital period because the error of the orbital period of
275: $2\times10^{-4}$ day given by SH only gives a phase shift of
276: $\sim0.026$, which is in agreement with the phase offset on
277: January 22. So there must exist some other reasons responding for
278: this phenomenon. The feature of phase shift of our radial velocity
279: curve is also listed in Table 3.
280:
281: %---------------------------Table 3-----------------------
282: \begin{table}
283: \caption{The $\gamma$ velocity, $K$, phase offset and
284: orbital-averaged velocity}
285: \begin{tabular}{ccccc}
286: \hline\hline
287: Date (UT) & $\gamma$ &$K$ & Phase offset & $V_{average}$\\
288: (Year 2002)& (km s$^{-1}$)& (km s$^{-1}$)& &(km s$^{-1}$) \\
289: \hline
290: Jan 21 ........& 91$\pm$4 &46$\pm$5 & -0.12$\pm$0.02 & 97$\pm$12\\
291: Jan 22 ........& 34$\pm$6 &49$\pm$9& -0.03$\pm$0.03 & 28$\pm$10\\
292: \hline
293: Shift & 57$\pm$7 & ...& 0.09$\pm$0.04 &69$\pm$16\\
294: \hline
295: Jan 21 \& 22 &62$\pm7$&49$\pm$10&-0.07$\pm$04 &...\\
296: \hline
297: \end{tabular}
298: \end{table}
299:
300: %--------------------------Figure 4--------------------------
301: \begin{figure}
302: \centering
303: \includegraphics[width=13cm]{fig4.eps}
304: \caption{Phase shifted in two days. The least-squares sinusoidal
305: fitted results for the radial velocities on January 21 (filled
306: triangles) and 22 (open triangles) are present with solid line and
307: dashed line, respectively. It is obviously that there existed a
308: shift ($\sim$0.09) between the phases in these two days.
309: \label{fig4}}
310: \end{figure}
311:
312: \subsubsection{An eccentric disk}
313: The phenomena described above can be well explained with a slow
314: precession of an eccentric outer disk. The relative velocity
315: (line-of-sight component) at point $r(\theta)$ on the boundary to
316: white dwarf is (Wu et al. 2001)
317: \begin{equation}
318: V(r,\theta)=C[-e \sin (\theta_0)- \sin(\theta+\theta_0)]
319: \end{equation}
320: where $C=\sin i \sqrt{\frac{GM_1}{a(1-e^2)}}=constant$; $i$, $a$
321: and $e$ are the inclination, half of the major axis and
322: eccentricity of the accretion disk, respectively. Hence, the mean
323: velocities of the troughs of the absorption lines or the peaks of
324: the emission lines are $V=-Ce\sin(\theta_0)$. According to this
325: equation, the increase of $\theta_0$ will lead to the result that
326: the central wavelengths of spectral lines, i.e., the
327: orbital-averaged velocity, will be variable with the precessing
328: phase of the disk.
329:
330: Generally, the $\gamma$ velocity is thought to represent the
331: systemic motion of the binary. This is correct provided that the
332: accretion disk was axis-symmetry. If the material in the ring
333: surrounds the primary is not in circular orbits, i.e., the
334: accretion disk is eccentric and processing, then white dwarf will
335: be at one of the focus and the mass center of the system (nearly
336: locating at the geometry center) would change with the precession
337: period of the disk. Thus, the systemic velocity, i.e., the
338: $\gamma$ velocity, maybe also change with different precession
339: phase of the disk.
340:
341: The large phase shift of $\sim$0.09 between these two days shows
342: that the variation of the centers of the absorption or emission
343: peaks can not represent the movement of the white dwarf. And this
344: phenomenon can also been interpreted with a slow precession of an
345: eccentric disk. As described above, the position of the mass
346: center of the system will vary with the precession of the disk.
347: This would lead to the variation of the time ($T_0$) of the
348: $\gamma$ crossover from negative to positive velocities, resulting
349: in the phase shift with the disk at different precession phase.
350:
351: \subsubsection{A constraint on the eccentricity}
352: The precessing period ($P_{prec}$) of the disk is the beat period
353: of the orbital period ($P_{orb}$) and the superhump period
354: ($P_{sh}$). It can be written as
355: \begin{equation}
356: \frac{1}{P_{prec}}=\frac{1}{P_{orb}}-\frac{1}{P_{sh}}
357: \end{equation}
358: According to Eq. (2), the precessing period of YZ Cnc is 1.52
359: days, computed with $P_{orb}$=0.0868 day, given by SH and
360: $P_{sh}$=0.09204, given by Patterson (1979). Thus, $\theta_0$ will
361: increase 4.12 rad within 1 day. So we have the mean velocities of
362: the absorption troughs to be $-Ce\sin(\theta_0)$ on January 21,
363: the emission peaks to be $-Ce\sin(\theta_0+4.12)$ on January 22.
364: Comparing their difference with the measured data (the shift of
365: ``$V_{average}$'' in Table 3), we have
366: \begin{equation}
367: C e \cos(\theta_0+2.06)=60
368: \end{equation}
369: If we know the value of $C$, i.e., know $M_1$, $a$, $q$ and
370: $P_{orb}$, we can give a constraint on the eccentricity of the
371: disk.
372: \subsection{Mass and Inclination}
373: The mass and inclination of stars are very difficult to be
374: determined if the system is not eclipsing. SH used the properties
375: of the emission lines (specially the linewidth) to provide a
376: relation between the inclination and the white dwarf mass with an
377: empirical assumption that the velocity from the line center at
378: 30\% of the emission-line intensity best represents $V_d\sin i$.
379: They gave that the mass of the secondary is $\sim$0.17 $M_\odot$
380: and 0.75-0.9 $M_\odot$ of the primary.
381:
382: Substituting $P_{orb}$ with 0.0868(2) day in equation (1) in Zhao
383: et al. (2005a), we have
384: \begin{equation}
385: M_1=M_2/q, \ M_2=[0.829(1+1/q)^{1/3}Q(q)]^{15/7}
386: \end{equation}
387: We can obtain the mass ratio by using an empirical relation found
388: by Patterson (2001), $\epsilon=0.216(\pm0.018)q$, where
389: $\epsilon=(P_{sh}-P_{orb})/P_{orb}$. It gives $q=0.28\pm0.03$,
390: which is somewhat larger than that of SH.
391:
392: According to equation (4), it is only requires $q>0.09$ to meet
393: the condition that the white dwarf mass should be less than 1.44
394: $M_{\odot}$. The mass ratio of 0.28$\pm0.03$ derived above is
395: consistent with this requirement. Therefore we can obtain that
396: $M_{2}=0.13\pm0.01\ M_{\odot}$ and $M_{1}=0.46\pm0.06M_{\odot}$.
397: Using the $K_1=49\pm10$ km s${^-1}$ (see \S3.2), the mass function
398: $f(M)=(M_{2}\sin i)^3/(M_1+M_2)^2]=K_1^3P_{orb}/(2\pi
399: G)=0.00106(10)\ M_{\odot}$ gives $i=34^\circ\pm9^\circ$.
400:
401: The values of $M_1$ and $M_2$ are smaller than those of SH. This
402: is believed due to the systemic difference between different
403: methods and some unproved empirical assumptions. Hence, the masses
404: and inclination given here are rather uncertain.
405:
406: \section{Discussion}
407:
408: The variation of $\gamma$ velocity discovered in YZ~Cnc is for the
409: first time but not alone. It has been discovered in several other
410: SU~UMa stars, i.e., Z~Cha (Vogt 1982, Honey et al., 1988), KS~UMa
411: (Zhao et al. 2005a) and ER~UMa (Zhao et al. 2005b). As described
412: by these different groups, the RV curves for different CVs are
413: \emph{all} gotten by measuring spectra obtained when the stars
414: went through \emph{eruptions}.
415:
416: Vogt (1982) and Honey et al. (1988) have found that the $\gamma$
417: velocities of Z~Cha varied during its superoutburst. Vogt (1981)
418: proposed a model in which he considered the behavior of a
419: precessing, elliptical ring surrounding a circular accretion disk.
420: This gives the variation of the $\gamma$ velocity on a
421: night-to-night basis as a result of variations in the projected
422: motion of the ring material against that of the inner (circular)
423: disk. Honey et al. (1988) interpreted their observational result
424: with new non-axisymmetric disk simulations as arising in an
425: eccentric, precessing disk which is tidally distorted by the
426: secondary.
427:
428: Our results that the $\gamma$ velocity vary with time and that the
429: phase shifts between different days are based on the measurement
430: of centers of the absorption troughs of H$\beta$ and the centers
431: of the peaks of the emission cores. We can not help but do this.
432: We can't use the double-Gaussian convolution method (Shafter et
433: al. 1988) to measure RV because the wings of H$\beta$ are blended
434: with He I $\lambda$4922 and the other Balmer lines not only are
435: contaminated but also have bad signal-to-noise ratio, especially
436: on January 22. Despite of this, our observational result confirmed
437: that the $\gamma$ velocity do actually vary when the star was
438: ongoing a superoutburst.
439:
440: We find for the first time that the phase of the system would
441: change. If we believe that the orbital period and the error given
442: by SH are reliable, the phase shift between these 2 days does
443: actually exist. It is not surprised that this phenomenon can be
444: found in YZ Cnc between 2 days because the precession period of
445: YZ~Cnc is only 1.52 days. The phase shift is enough to be observed
446: between 2 days. So our observation provides more evidences to
447: convince us that the accretion disk is precessing and eccentric
448: when the binary system going through superoutburst. If the system
449: has a larger inclination, we could get more information and do
450: more detailed analysis like IY UMa (Wu et al. 2001) and KS UMa
451: (Zhao et al. 2005a).
452:
453: We can also estimate the eccentricity crudely by substituting the
454: mass of the white dwarf $M_1$, the mass ratio $q$ and the
455: inclination $i$, with the values given above (see $\S$3.4), we
456: obtained
457: \[\cos(\theta_0+2.06)=0.231/e\]
458: If we substituted these parameter with the values given by SH, we
459: would get
460: \[\cos(\theta_0+2.06)=0.184/e\]
461: So $e$ must be larger than 0.184 or 0.231, according to whose
462: parameters are more reliable.
463:
464: \section{Conclusions}
465: So far, we have shown the properties of our spectra and the radial
466: velocities of YZ Cnc obtained during its 2002 January
467: superoutburst. These properties include following three aspects,
468:
469: (1) The $\gamma$ velocity (62$\pm$7 km s$^{-1}$) obtained in these
470: two days deviated strongly from the systemic velocity (16$\pm$10
471: km s$^{-1}$) measured by SH when the binary system was in
472: quiescence. And there is a discrepancy of $\sim$60 km s$^{-1}$ of
473: the $\gamma$ velocities between these two days.
474:
475: (2) The mean velocities averaged throughout the orbital period of
476: these two days have large offset of the order of $\pm$70 km
477: s$^{-1}$.
478:
479: (3) There is large phase offset of $\sim$0.09 between these two
480: days.
481:
482: As detailedly described in $\S$3.3, we can make a conclusion that
483: these features are all ascribing to the precession of an eccentric
484: accretion disk. Therefore, we can make use of these properties to
485: confirm whether the accretion disk is eccentric or not.
486:
487: \begin{acknowledgements}
488: We would like to thank the Optical Astronomy Laboratory, Chinese
489: Academy of Sciences and Prof. Jianyan Wei of the National
490: Astronomical Observatory for scheduling the observations. We are
491: grateful for the financial support from the National Natural
492: Science Foundation of China (through grants 10173005 and
493: 10010120074).
494: \end{acknowledgements}
495:
496:
497: \begin{thebibliography}{}
498:
499: \bibitem[{C\'{o}rdova \& Mason}{1984}]{b1} C\'{o}rdova F., Mason K., 1984, MNRAS, 206, 879
500:
501: \bibitem[{Eracleous et al.}{1991}]{b2} Eracleous M., Halpern J., Patterson J., 1991, ApJ, 382,
502: 290
503:
504: \bibitem[{Hakala et al.}{2004}]{b3} Hakala P., Ramsay G., Wheatley P. et al., 2004, A\&A, 420, 273
505:
506: \bibitem[{Honey et al.}{1988}]{b4} Honey W. B., Charles P. A., Whitehurst R. et al., 1988, MNRAS, 231, 1
507:
508: \bibitem[{Moffett \& Barnes}{1974}]{b6} Moffett T. J., Barnes T. G., 1974, ApJ, 194, 141
509:
510: \bibitem[{Oke \& Wade}{1982}]{b7} Oke J. B., Wade R. A., 1982, AJ, 87, 670
511:
512: \bibitem[{Patterson}{1979}]{b8} Patterson J., 1979, AJ, 84, 804
513:
514: \bibitem[{Patterson}{1981}]{b9} Patterson J., 1981, ApJS, 45, 517
515:
516: \bibitem[{Patterson}{2001}]{b11} Patterson J., 2001, PASP, 113, 736
517:
518: \bibitem[{Shafter \& Hessman}{1988, named SH hereafter}]{b12} Shafter A. W., Hessman F. V., 1988, AJ, 95, 178
519:
520: \bibitem[{Shafter et al.}{1988}]{1988} Shafter A. W., Hessman F. V., Zhang E. H., 1988, ApJ, 327, 248
521:
522: \bibitem[{Szkody}{1981}]{b13} Szkody P., 1981, AJ, 247, 577
523:
524: \bibitem[{van der Woerd}{1987}]{b14} van der Woerd H., 1987, Ph.D. Thesis, Utrecht
525: University
526:
527: \bibitem[{van Paradijs et al.}{1994}]{b15} van Paradijs J., Charles P. A., Harlaftis E. T. et
528: al., 1994, MNRAS, 267, 465
529:
530: \bibitem[{van Teeseling \& Verbunt}{1994}]{b16} van Teeseling A., Verbunt F., 1994, A\&A, 292, 519
531:
532: \bibitem[{Verbunt et al.}{1997}]{b17} Verbunt F., Bunk W., Ritter H. et al.,
533: 1997, A\&A, 327, 602
534:
535: \bibitem[{Verbunt et al.}{1999}]{b18} Verbunt F., Wheatley P. J., Mattei J. A., 1999,
536: A\&A, 346, 146
537:
538: \bibitem[{Vogt}{1982}]{b19} Vogt N., 1982, ApJ, 252, 653
539:
540: \bibitem[{Vorob'yeva \& Kukarkin}{1961}]{b20} Vorob'yeva V. A., Kukarkin B. V., 1961, Peremenye Zvezdy,
541: 13, 428
542:
543: \bibitem[{Wade}{1982}]{b21} Wade R. A., 1982, AJ, 87, 1558
544:
545: \bibitem[{Warner}{1995}]{b22} Warner B., 1995, Cataclysmic Variable
546: Stars, Cambridge: Cambridge University Press
547:
548: \bibitem[{Williams}{1983}]{b23} Williams G., 1983, ApJS, 53, 523
549:
550: \bibitem[{Wu et al.}{2001}]{b24} Wu X. A., Li Z. Y., Gao W. H., 2001, ApJ, 549, L81
551:
552: \bibitem[{Zhao et al.}{2005a}]{b25} Zhao Y. H., Li Z. Y., Peng Q. H. et al., 2005a,
553: submitted
554:
555: \bibitem[{Zhao et al.}{2005b}]{b26} Zhao Y. H., Li Z. Y., Wu X. A. et al., 2005b,
556: submitted
557: \end{thebibliography}
558:
559: \end{document}
560: