1: \documentstyle{pasj06}

2: %\draft

3: \begin{document}

4: 

5: \SetRunningHead{A.Imada et al.}{}

6: 

7: \title{The 2006 November outburst of EG Aquarii: the SU UMa nature revealed}

8: 

9: \author{

10: Akira \textsc{Imada}\altaffilmark{1},

11: Rod \textsc{Stubbings}\altaffilmark{2},

12: Taichi \textsc{Kato}\altaffilmark{1},

13: Makoto \textsc{Uemura}\altaffilmark{3},

14: Thomas \textsc{Krajci}\altaffilmark{4}, \\

15: Ken'ichi \textsc{Torii}\altaffilmark{5}, 

16: Kei \textsc{Sugiyasu}\altaffilmark{1},

17: Kaori \textsc{Kubota}\altaffilmark{1},

18: Yuuki \textsc{Moritani}\altaffilmark{1}

19: Ryoko \textsc{Ishioka}\altaffilmark{6}, \\

20: Gianluca \textsc{Masi}\altaffilmark{7}, 

21: Seiichiro \textsc{Kiyota}\altaffilmark{8}, 

22: L.A.G. \textsc{Monard}\altaffilmark{9},

23: Hiroyuki \textsc{Maehara}\altaffilmark{10},\\

24: Kazuhiro \textsc{Nakajima}\altaffilmark{11}, 

25: Akira \textsc{Arai}\altaffilmark{12},

26: Takashi \textsc{Ohsugi}\altaffilmark{3}, 

27: Takuya \textsc{Yamashita}\altaffilmark{3},\\

28: Koji S. \textsc{Kawabata}\altaffilmark{3}, 

29: Osamu \textsc{Nagae}\altaffilmark{12}, 

30: Shingo \textsc{Chiyonobu}\altaffilmark{12},

31: Yasushi \textsc{Fukazawa}\altaffilmark{12}, \\ 

32: Tsunefumi \textsc{Mizuno}\altaffilmark{12},

33: Hideaki \textsc{Katagiri}\altaffilmark{12},

34: Hiromitsu \textsc{Takahashi}\altaffilmark{12}, 

35: Atsushi \textsc{Ueda}\altaffilmark{12}, \\

36: Takehiro \textsc{Hayashi}\altaffilmark{13}, 

37: Kiichi \textsc{Okita}\altaffilmark{14},

38: Michitoshi \textsc{Yoshida}\altaffilmark{14},

39: Kenshi \textsc{Yanagisawa}\altaffilmark{14}, \\ 

40: Shuji \textsc{Sato}\altaffilmark{15}, 

41: Masaru \textsc{Kino}\altaffilmark{15}, 

42: Masahiro \textsc{Kitagawa}\altaffilmark{15}, 

43: Kozo \textsc{Sadakane}\altaffilmark{16} \\

44: and Daisaku \textsc{Nogami}\altaffilmark{17}

45: }

46: 

47: \affil{$^1$Department of Astronomy,Faculty of Science, Kyoto University,

48:        Sakyo-ku, Kyoto 606-8502}

49: \affil{$^2$Tetoora Observatory, Tetoora Road, Victoria, Australia}

50: \affil{$^3$Astrophysical Science Center, Hiroshima University, Kagamiyama

51: 1-3-1, \\ Higashi-Hiroshima 739-8526}

52: \affil{$^4$PO Box 1351 Cloudcroft, New Mexico 83117, USA}

53: \affil{$^5$Department of Earth and Space Science, Graduate School of

54: Science, \\ Osaka University, 1-1 Machikaneyama-cho, Toyonaka, Osaka

55:        560-0043}

56: \affil{$^6$Subaru Telescope, National Astronomical Observatory of Japan, 650

57: North A'ohoku Place,\\ Hilo, HI 96720, USA}

58: \affil{$^7$The Virtual Telescope Project, Via Madonna del Loco 47, 03023

59:        Ceccano (FR), Italy}

60: \affil{$^8$VSOLJ, Center for Balcony Astrophysics, 1-401-810 Azuma, Tsukuba,

61: Ibaraki 305-0031}

62: \affil{$^9$Bronberg Observatory, CBA Pretoria, PO Box 11426, Tiegerpoort

63:        0056, South Africa}

64: \affil{$^{10}$Kwasan and Hida Observatories, Kyoto University,

65:        Yamashina-ku, Kyoto 607-8471}

66: \affil{$^{11}$VSOLJ, 124 Isatotyo Teradani, Kumano, Mie 519-4673}

67: \altaffiltext{12}{Department of Physical Science, Hiroshima University,

68: Kagamiyama 1-3-1, \\ Higashi-Hiroshima 739-8526}

69: \altaffiltext{13}{Department of Education, Hiroshima University,

70: Kagamiyama 1-1-1, \\ Higashi-Hiroshima 739-8524}

71: \altaffiltext{14}{Okayama Astrophysical Observatory, National

72: Astronomical Observatory of Japan, \\ Kamogata Okayama 719-0232}

73: \altaffiltext{15}{Department of Physics, Nagoya University, Furo-cho,

74: Chikusa-ku, Nagoya 464-8602}

75: \altaffiltext{16}{Astronomical Institute, Osaka Kyoiku University,

76: Asahigaoka, Kashiwara, Osaka 582-8582}

77: \affil{$^{17}$Hida Observatory, Kyoto University, Kamitakara, Gifu

78:        506-1314}

79: \email{a\_imada@kusastro.kyoto-u.ac.jp}

80: 

81: \KeyWords{

82:           accretion, accretion disks

83:           --- stars: dwarf novae

84:           --- stars: individual (EG Aquarii)

85:           --- stars: novae, cataclysmic variables

86:           --- stars: oscillations

87: }

88: 

89: \maketitle

90: 

91: \begin{abstract}

92: 

93: We report time-resolved CCD photometry of the cataclysmic variable EG

94:  Aquarii during the 2006 November outburst During the outburst, superhumps

95:  were unambiguously detected with a mean period of 0.078828(6) days,

96:  firstly classifying the object as an SU UMa-type dwarf nova. It also

97:  turned out that the outburst contained a precursor. At the end of the

98:  precursor, immature profiles of humps were observed. By a phase

99:  analysis of these humps, we interpreted the features as superhumps. This

100:  is the second example that the superhumps were shown during a

101:  precursor. Near the maximum stage of the outburst, we discovered an

102:  abrupt shift of the superhump period by ${\sim}$ 0.002 days. After the

103:  supermaximum, the superhump period decreased at the rate of

104:  $\dot{P}/P$=$-$8.2${\times}$10$^{-5}$, which is typical for SU UMa-type

105:  dwarf novae. Although the outburst light curve was characteristic of SU

106:  UMa-type dwarf novae, long-term monitoring of the variable shows no

107:  outbursts over the past decade. We note on the basic properties of long

108:  period and inactive SU UMa-type dwarf novae.

109: 

110: \end{abstract}

111: 

112: \section{Introduction}

113: 

114: Dwarf novae are a subclass of cataclysmic variables (CVs) that consist

115: of a white dwarf primary and a late-type secondary. The secondary fills

116: its Roche lobe and transfers gases into the primary through the inner

117: Lagragian point (L1). The transferred matter forms an accretion disk

118: around the white dwarf. The accretion disk causes instabilities, which

119: are observed as an outburst (for a review, see \cite{war95book};

120: \cite{osa96review}; \cite{hel01book}; \cite{smi07review}).

121: 

122: SU UMa-type stars are a subclass of dwarf novae. The majority have

123: orbital periods below 2 hours. This

124: subclass exhibits two types of outbursts. One is a normal outburst,

125: lasting a few days. The other is a superoutburst lasting about 2

126: weeks. The most characteristic feature during the superoutburst is that

127: the light curve always shows modulations termed superhump. The observed

128: period of superhumps is a few percent longer than the

129: orbital period of the system. This is well explained by the precessing

130: eccentric disk, which is deformed at the 3:1 resonance radius

131: (\cite{whi88tidal}; \cite{osa89suuma}).

132: 

133: EG Aquarii (hereafter EG Aqr) was introduced as a candidate for dwarf

134: novae by \citet{1959PASP...71..469L} and later by

135: \citet{vog82atlas}. The variability of EG Aqr was reported in

136: \citet{har60egaqr}, in which the variable

137: was at 14.0 mag on 1958 November 5. There are no other positive and

138: negative observations except this around this outburst, so that we cannot

139: lead to the conclusion of the nature of the

140: outburst. \citet{szk92CVspec} performed optical spectroscopy of the

141: variable. The result was, however, that the spectrum showed no emission

142: in Balmer series. On the contrary, it resembles that of a K-type

143: star. There is a possibility that the star was simply misidentified by

144: \citet{szk92CVspec}. Our astrometric estimation of the variable from the

145: outbursting images yielded 

146: RA $23^{\rm h} 25^{\rm m} 19^{\rm s}.09$ and 

147: Dec $-08^{\circ} 18' 18''.5$, respectively. This means that the variable

148: is identical with USNO B1.0 0816-0716959 

149: (RA $23^{\rm h} 25^{\rm m} 19^{\rm s}.17$, Dec $-08^{\circ} 18' 18''.9$, 

150: $B1$=18.760 $R1$=18.260).\footnote{A. Henden gives $V$=18.651 and

151: $B-V$=-0.062.}. Concerning these results, the infrared counterpart may

152: be 2MASS J23251917-0818190 ($J$=17.432, $H$=16.309,

153: $K$=15.681.\footnote{\cite{spr96CVabsmag} gave $J$= 17.10 and $K$=16.26,

154: respectively. see also \citet{hoa02CV2MASS}).} As for the X-ray range,

155: the variable is below the detection limit of the ROSAT faint source

156: catalog \citep{1999A&A...349..389V}.

157: 

158: On 2006 November 8, one of the authors (RS) discovered the eruption

159: of EG Aqr at a visual magnitude of 12.5 ([vsnet-alert 7217]). This is

160: the first recorded outburst since the report by \citet{har60egaqr}. Here

161: we report on time-resolved CCD photometry of EG Aqr during the 2006

162: November superoutburst.

163: 

164: \section{Observations}

165: 

166: \begin{longtable}{cccccc}

167: \caption{Journal of observations.}

168: \hline\hline

169: 2006 Date & HJD-start$^*$ & HJD-end$^*$ & Exp(s)$^{\dagger}$

170: & N$^{\ddagger}$ & ID$^{\S}$ \\

171: \hline

172: \endhead

173: \hline

174: \endfoot

175:  November 8 & 48.2507 & 48.2719 & 30 & 57 & BM \\

176:  & 48.3433 & 48.4689 & 60 & 161 & GM \\

177:  November 9 & 48.5502 & 48.7896 & 30 & 559 & KTC\\

178:  & 48.8877 & 48.9843 & 30 & 140 & Njh \\

179:  & 48.9031 & 49.0906 & 30 & 285 & Kyoto \\

180:  & 48.9345 & 49.0642 & 30 & 341 & Kis \\

181:  & 49.0311 & 49.1369 & 30 & 400 & Mhh \\

182:  & 49.2279 & 49.3960 & 60 & 212 & GM \\

183:  & 49.3272 & 49.4042 & 30 & 219 & BM \\

184:  November 10 & 49.5484 & 49.7925 & 30 & 570 & KTC\\

185:  & 50.2290 & 50.4390 & 60 & 233 & GM \\

186:  November 11 & 50.5497 & 50.8028 & 30 & 591 & KTC\\

187:  & 50.9310 & 50.9632 & 30 & 39 & Njh \\

188:  November 12 & 51.8683 & 52.0512 & 30 & 390 & Kyoto \\

189:  & 51.8824 & 52.0413 & 30 & 280 & Njh \\ 

190:  & 51.8851 & 51.9495 & 30 & 142 & Kis \\

191:  & 51.9244 & 52.1404 & 30 & 749 & Mhh \\

192:  November 13 & 52.5921 & 52.8031 & 30 & 493 & KTC\\

193:  & 52.9314 & 52.9988 & 30 & 109 & Njh \\

194:  & 53.0267 & 53.0671 & 30 &  71 & Kyoto \\

195:  Novemver 14 & 53.5503 & 53.8013 & 30 & 586 & KTC\\

196:  & 53.9061 & 54.0310 & 30 & 410 & Mhh \\

197:  November 15 & 54.8652 & 55.0377 & 30 & 227 & Kyoto \\

198:  & 54.9020 & 55.0546 & 30 & 281 & Njh \\

199:  & 55.2578 & 55.4017 & 60 & 177 & GM \\

200:  November 16 & 55.5533 & 55.8009 & 30 & 578 & KTC\\

201:  & 55.8599 & 56.0782 & 30 & 338 & Kyoto \\

202:  & 55.8835 & 56.0688 & 30 & 312 & Njh \\

203: November 17 & 56.6015 & 56.7992 & 30 & 462 & KTC \\

204:  & 56.8768 & 57.1001 & 30 & 350 & Kyoto \\

205:  & 56.8834 & 56.9822 & 30 & 180 & Njh \\

206: November 18 & 57.5465 & 57.7572 & 30 & 351 & KTC \\

207:  November 20 & 59.8828 & 60.0813 & 30 & 265 & Kyoto \\

208: November 21 & 61.0241 & 61.1100 & 63 & 81 & Kanata \\

209: November 24 & 63.9465 & 64.0325 & 10 & 606 & Kanata \\

210: December 5 & 74.9825 & 75.0276 & 63 & 48 & Kanata \\

211: December 6 & 76.0116 & 76.0383 & 63 & 26 & Kanata \\

212: \hline

213: \multicolumn{6}{l}{$^*$HJD-2454000. $^{\dagger}$Exposure

214:  time. $^{\ddagger}$Number of exposure. $^{\S}$Observer's ID.} \\

215: \multicolumn{6}{l}{BM: L.A.G. Monard, South Africa. GM: G. Masi, Italy.} \\

216: \multicolumn{6}{l}{KTC: T. Krajci, USA. Njh: K. Nakajima, Japan.} \\

217: \multicolumn{6}{l}{Kyoto: A. Imada et al., Japan. Kis: S. Kiyota, Japan.} \\

218: \multicolumn{6}{l}{Mhh: H. Maehara, Japan. Kanata: M. Uemura et al., Japan.} \\

219: \label{logofobs}

220: \end{longtable}

221: 

222: 

223: Time resolved CCD photometry was performed on 15 nights using 25-150cm

224: telescopes between 2006 November 8 and 2006 December 6 at 8 sites, and a

225: log of observations is summarized in table 1. No filter was used

226: except for Higashi-Hiroshima site and the obtained data were close to

227: those of

228: $R_{\rm c}-$system. At Higashi-Hiroshima site, the data were obtained

229: with the

230: Triple Range Imager and Spectrograph (TRISPEC,

231: \cite{2005PASP..117..870W}) using a $V$ filter on 2006 November 21, December

232: 5, and December 6. The exposure time was 10-63 seconds, with a few second

233: readout time of the CCDs. The total datapoint amounts to 10506, which is

234: sufficient to investigate the variability of the system.

235: 

236: After debiasing and flat fielding, the images were processed by aperture

237: photometry packages. Since there existed an effect of the atmospheric

238: extinction due to the color difference between the variable and the

239: comparison star, we made a correction for the data. After correcting the

240: obtained magnitudes between sites, the magnitude was adjusted to that of the

241: Kyoto site, where we used a Java-based aperture photometry package

242: developed by one of the authors (TK). The differential magnitude of the

243: variables were determined using No. 12 of the Henden Catalog ($V$=11.910,

244: $B-V$=0.632), whose constancy was checked using nearby stars in the same

245: images.\footnote{ftp://ftp.aavso.org/public/calib/} The accuracy of the

246: calibration was dependent on the condition of the skies. Although some

247: data were contaminated by such as clowds or light pollutions, the

248: expected error was achieved as small as 0.03 mag for good data, which is

249: sufficient for the purpose of our observations. Heliocentric correction

250: was made for each data set before the following analyses.

251: 

252: \section{Results}

253: 

254: \begin{figure*}

255: \begin{center}

256: \resizebox{160mm}{!}{\includegraphics{lc070613.eps}}

257: \end{center}

258: \caption{Light curve of EG Aqr during the 2006 November

259:  superoutburst. The abscissa and the ordinate denote HJD and relative

260:  magnitude, respectively. The comparison star is at a magnitude of 11.91

261:  in $V$ band. The filled and open circles indicate an averaged light curves

262:  obtained by our run and the AAVSO database, respectively. The bottom

263:  triangles denotes negative observations reported to the AAVSO and

264:  VSNET. A precursor was detected on HJD 2454048 (see also figure 2). The

265:  maximum magnitude was reached around HJD 2454049.7, after that the

266:  variable declined at the rate of 0.14(1) mag d$^{-1}$. No evidence for

267:  a rebrightening was shown during our whole run.}

268: \label{lc}

269: \end{figure*}

270: 

271: 

272: 

273: \begin{figure*}

274: \begin{center}

275: \resizebox{160mm}{!}{\includegraphics{lc_early2.eps}}

276: \end{center}

277: \caption{Enlarged light curves around the supermaximum. For the purpose

278:  of description, data obtained from HJD 2454048.8877 to HJD 2454049.1369

279:  were binned by 8 frames. Although the data were sparse before HJD

280:  2454048.5, the light curves provided evidence for a precursor around

281:  the phase, where a hint of modulations was visible. As can be seen in

282:  this figure, the supermaximum was around HJD 2454049.7.}

283: \label{elc}

284: \end{figure*}

285: 

286: 

287: 

288: 

289: \subsection{light curves}

290: 

291: Figure \ref{lc} illustrates the overall light curves of EG Aqr during

292: the 2006 outburst. By virtue of the prompt reports of the outburst, the

293: data covered the early stages before the outburst maximum. Enlarged

294: light curves around the maximum are displayed in figure

295: \ref{elc}. Judging from the trend of the light curve, we may have caught

296: the end of the precursor and rising stage of the main outburst on HJD

297: 2454048. At the end of the precursor, some superhumps were visible. The

298: magnitude reached the maximum on HJD 2454049, after which the variable

299: declined at the rate of 0.14(1) mag d$^{-1}$. Based on its quiescent and

300: maximum magnitude, the superoutburst amplitude exceeded ${\sim}$5.8

301: mag. From this, we can categorize EG Aqr as large-amplitude dwarf novae

302: (TOADs, \cite{how95TOAD}). The duration of the plateau stage was 8-11

303: days. No rebrightenings were observed during our run.

304: 

305: \subsection{superhump}

306: 

307: To clarify EG Aqr as a new member of SU UMa-type dwarf novae, we

308: examined the light curves obtained at each site. Except for some data

309: obtained under bad conditions, we can unambiguously detect superhumps

310: having a rapid rise and slow decline of the profiles. This is the

311: first time that EG Aqr showed superhumps.

312: 

313: After subtracting trends for each light curve, we applied the phase

314: dispersion minimization method (PDM, \cite{ste78pdm}) to the

315: residual light curves from HJD 2454049.5 to HJD 2454057, corresponding to

316: the plateau stage. The resultant theta diagram is displayed in figure

317: \ref{pdmp}, in which we determined 0.078828(6) days as the best

318: estimated period of the mean superhump. The error of the period is

319: calculated using the Lafler-Kinman class of methods, as applied by

320: \citet{fer89error}. The same method was applied for the data before HJD

321: 2454049.5. from which we determined $P$=0.08076(11) days as being the

322: best estimated period. Figure \ref{pdme} shows the resultant theta

323: diagram. These results indicate

324: the mean superhump period around the maximum was about 0.002 days longer

325: than that during the main plateau stage. In figure \ref{mhp}, we demonstrate

326: the averaged profiles of the superhumps folded by 0.08076 days from HJD

327: 2454049 and 0.078828 days for HJD 2454050 through HJD 2454057, where one can

328: see a characteristic feature of superhumps. Although the amplitude of

329: the superhumps decreased as the superoutburst proceeded, the shape of

330: the profiles kept almost constant over the course of the plateau

331: stage. The apparent absence of eclipses indicates that the inclination

332: of the system is not too high.

333: 

334: 

335: 

336: \subsection{superhump period change}

337: 

338: We calibrated the maximum timings of superhumps mainly by eye. The

339: error is an order of 0.001 days. In table 2, we tabulate the

340: maximum timings of superhumps. A linear regression to the obtained

341: values in table 2 yielded,

342: 

343: \begin{equation}

344: HJD(max) = 2454048.7101(18) + 0.078935(29) \times E,

345: \label{eqlinear}

346: \end{equation}

347: where $E$ is the cycle count. Using the equation, we can

348: draw an $O - C$ diagram described in figure \ref{oc}. It should be noted

349: that the change of the period occurred for 8${\leq}E{\leq}$11, when we missed

350: observations unfortunately. Nevertheless, we firstly succeeded in

351: detecting the superhump period change at the early stage of

352: superoutburst for long-period SU UMa-type dwarf novae, and confirmed the

353: period change did occur. For -4$<E<$8, the best fit equation is given by

354: 

355: \begin{equation}

356: O - C = -1.40(0.09)\times10^{-2} + 2.35(0.18)\times10^{-3} E. 

357: \label{eqearly}

358: \end{equation}

359: On the other hand, the best fit quadratic between $E$=11 and $E$=114 is given by

360: \begin{eqnarray}

361: O - C =& 2.37(0.90)\times10^{-3} + 2.52(0.33)\times10^{-4} E \nonumber \\

362:        & -3.23(0.26)\times 10^{-6} E^{2}. 

363: \label{eqplateau}

364: \end{eqnarray}

365: The equation \ref{eqplateau} indicates $P_{\rm dot}$ = $\dot{P}$/$P$ =

366: $-$8.2(7)$\times$10$^{-5}$. This is a normal value for SU UMa-type dwarf

367: novae \citep{2006PASJ...58..383I}.

368: 

369: \subsection{post outburst stage}

370: 

371: Some SU UMa-type dwarf novae exhibit rebrightenings and/or late superhumps

372: after the main plateau stage. In order to search for these phenomena, we

373: mainly used 1.5m Kanata telescope. Figure \ref{postlc} shows the

374: representative light curves after the main plateau stage obtained by the

375: telescope. One can see modulations with an amplitude larger than 0.3

376: mag, which is even larger than that of superhumps. On November 24, the

377: light curve appears to be doubly-peaked indicating the light curve had

378: remained unchanged. However, due to the lack of observations, it is not

379: clear whether the observed light curves are late

380: superhumps. \footnote{We calibrated the maximum timing of the hump on

381: November 24. However, we cannot draw firm conclusion whether a phase

382: shift occurred, since the interval between the two successive humps was

383: too long (see table 2).}

384: 

385: In order to explore whether a rebrightening showed, we examined the

386: VSNET and AAVSO databases. As can be seen in figure \ref{lc}, there is

387: no evidence for a rebrightening before HJD 2454065. Although most of

388: rebrightening are observed within 5 days after the end of the plateau

389: stage (e.g., \cite{how96alcom}; \cite{ish01rzleo};

390: \cite{tem06asas0025}), we cannot rule out the possibility that a

391: rebrightening occurred between HJD 2454066 and HJD 2454070.

392: 

393: \begin{figure}

394: \begin{center}

395: \resizebox{80mm}{!}{\includegraphics{pdm.eps}}

396: \end{center}

397: \caption{Theta diagram of period analysis for the plateau stage between

398:  HJD 2454049.55 and HJD 2454056.98. The lowest minimum is definitely

399:  seen at $P_{\rm sh}$=0.078828(5) days.}

400: \label{pdmp}

401: \end{figure}

402: 

403: \begin{figure}

404: \begin{center}

405: \resizebox{80mm}{!}{\includegraphics{pdm_early.eps}}

406: \end{center}

407: \caption{Same as figure \ref{pdmp} but for the data between HJD

408:  2454048.25 and HJD 2454049.40. The best estimated period during the

409:  phase yields $P_{\rm sh}$=0.08076(12) days.}

410: \label{pdme}

411: \end{figure}

412: 

413: \begin{figure}

414: \begin{center}

415: \resizebox{80mm}{!}{\includegraphics{mhp_daily.eps}}

416: \end{center}

417: \caption{Nightly averaged profiles of superhumps. The numbers in this

418:  figure denote the days since HJD 2454000. Based on the above estimated

419:  period, the profile on HJD 2454049 was folded by 0.08076 days, while

420:  folded by 0.078828 days after HJD 2454050. Although the amplitude of

421:  the superhumps decreased gradually, the feature of a rapid rise and

422:  slow decline were visible for all the respective nights.}

423: \label{mhp}

424: \end{figure}

425: 

426: \begin{figure}

427: \begin{center}

428: \resizebox{80mm}{!}{\includegraphics{o-c.eps}}

429: \end{center}

430: \caption{$O-C$ diagram during the superoutburst. The horizontal and

431:  vertical axes mean $O-C$ and the cycle count, respectively. The maximum

432:  timings were listed in table 2. A linear regression to the

433:  observed times gives equation (\ref{eqlinear}). The solid and dashed lines

434:  indicate the best fit equation for -4$<E<$8 and 11$<E<$110,

435:  respectively. The former is given by equation (\ref{eqearly}), while

436:  the latter is given by equation (\ref{eqplateau}). This figure means

437:  that the period kept almost constant in -4$<E<$8. However, an abrupt

438:  period change occurred in 8$<E<$11. After $E>$11, the superhump

439:  period decreased gradually at the rate of $\dot{P}$/$P$ =

440:  $-$8.2$\times$10$^{-5}$.}

441: \label{oc}

442: \end{figure}

443: 

444: \begin{figure}

445: \begin{center}

446: \resizebox{80mm}{!}{\includegraphics{post.eps}}

447: \end{center}

448: \caption{Representative light curves during the post-outburst

449:  phase. For a visual purpose, these light curves were arbitrarily

450:  shifted. The numbers in the figure denote the days from 2454000. The

451:  datapoints are binned by 4 frames on HJD 2454063 (November 24) after

452:  precluding the bad data. There exist large modulations on both light

453:  curves.}

454: \label{postlc}

455: \end{figure}

456: 

457: 

458: \begin{table*}

459: \caption{Timings of superhump maxima.}

460: \begin{center}

461: \begin{tabular}{cccccccc}

462: \hline\hline

463: $E^*$ & HJD$^{\dagger}$ & err$^{\ddagger}$ & ID & $E^*$ &

464:  HJD$^{\dagger}$ & err$^{\ddagger}$ & ID \\

465: \hline

466: -4 & 48.3731 & 0.003 & GM & 51 & 52.7400 & 0.005 & KTC\\

467: -3 & 48.4498 & 0.007 & GM & 54 & 52.9823 & 0.001 & Njh \\

468: -1 & 48.6112 & 0.005 & KTC & 62 & 53.6106 & 0.002 & KTC\\ 

469: 0 & 48.6979 & 0.001 & KTC & 63 & 53.6891 & 0.005 & KTC\\

470: 1 & 48.7763 & 0.003 & KTC & 64 & 53.7696 & 0.002 & KTC\\ 

471: 3 & 48.9428 & 0.002 & Kyoto & 66 & 53.9246 & 0.008 & Mhh \\ 

472: 4 & 49.0257 & 0.004 & Kis & 67 & 54.0058 & 0.004 & Mhh \\ 

473: 5 & 49.1018 & 0.004 & Mhh & 79 & 54.9487 & 0.002 & Kyoto \\

474: 7 & 49.2623 & 0.003 & GM & 79 & 54.9498 & 0.003 & Njh \\

475: 8 & 49.3448 & 0.001 & AAVSO &  80 & 55.0256 & 0.003 & Njh \\

476: 8 & 49.3465 & 0.002 & GM &  83 & 55.2631 & 0.005 & GM \\

477: 8 & 49.3471 & 0.002 & BM &  84 & 55.3421 & 0.003 & GM \\

478: 11 & 49.5827 & 0.002 & KTC&  87 & 55.5774 & 0.002 & KTC\\

479: 12 & 49.6628 & 0.003 & KTC& 88 & 55.6556 & 0.002 & KTC\\

480: 13 & 49.7414 & 0.002 & KTC& 89 & 55.7334 & 0.002 & KTC\\

481: 20 & 50.2950 & 0.001 & GM & 91 & 55.8915 & 0.003 & Kyoto \\

482: 21 & 50.3746 & 0.001 & GM & 92 & 55.9683 & 0.002 & Kyoto \\

483: 24 & 50.6118 & 0.001 & KTC& 92 & 55.9705 & 0.005 & Njh \\

484: 25 & 50.6893 & 0.002 & KTC& 93 & 56.0482 & 0.002 & Njh \\

485: 26 & 50.7710 & 0.001 & KTC& 101 & 56.6770 & 0.004 & KTC \\

486: 41 & 51.9526 & 0.002 & Njh & 102 & 56.7557 & 0.004 & KTC \\

487: 41 & 51.9530 & 0.002 & Mhh & 104 & 56.9133 & 0.003 & Kyoto \\

488: 41 & 51.9535 & 0.002 & Kyoto & 104 & 56.9156 & 0.005 & Njh \\

489: 42 & 52.0311 & 0.004 & Mhh & 105 & 56.9891 & 0.003 & Kyoto \\

490: 42 & 52.0322 & 0.003 & Njh & 113 & 57.6194 & 0.001 & KTC \\

491: 43 & 52.1104 & 0.003 & Mhh & 114 & 57.6985 & 0.002 & KTC \\                          

492: 50 & 52.6610 & 0.002 & KTC & 194: & 63.9761 & 0.004 & Kanata \\

493: \hline

494: \multicolumn{3}{l}{$^*$ Cycle count.} \\

495: \multicolumn{3}{l}{$^{\dagger}$ HJD-2454000} \\

496: \multicolumn{3}{l}{$^{\ddagger}$ In the unit of day.} \\

497: \end{tabular}

498: \end{center}

499: \label{hjdmax}

500: \end{table*}

501: 

502: 

503: 

504: 

505: \section{Discussion}

506: 

507: \subsection{long-term activity}

508: 

509: One of the most unique properties in EG Aqr is the inactivity of the

510: system. As mentioned before, there is only one recorded outburst since

511: the discovery of EG Aqr until the current one. No

512: outbursts were detected by the ASAS-3 over the past 5 years, during

513: which the variable was below the detection limit of $V$=14.5. We also

514: explored visual observations reported to the AAVSO and VSNET since

515: 1999. EG Aqr was especially monitored in 1999, 2000, and 2004 with a

516: mean interval of observations being a few days. These intensive

517: observation may detect not only superoutburst but also normal outburst

518: if they really occur. However, no outbursts were reported until 2006.

519: Although we should await an investigation on archival plates to draw a

520: firm conclusion as to whether EG Aqr is indeed inactive, the current

521: available information on EG Aqr may suggest its inactive nature. The

522: non-detection of a normal outburst in EG Aqr may also support its long

523: recurrence time, since the frequency of normal outburst is related to

524: the length of the supercycle \citep{osa95wzsge}.

525: 

526: \subsection{outburst light curve}

527: 

528: As noted above, extensive CCD photometry revealed the SU UMa nature of

529: EG Aqr. Superhumps were clearly visible during the course of the

530: superoutburst for the first time in EG Aqr. As for the light curve

531: itself, we should note on the presence of a precursor and the presence

532: of superhumps at the end of the precursor. In the previous

533: section, we have shown the $O-C$ diagram, in which we measured the

534: maximum timings at the end of the precursor, corresponding to $E$=-4 and

535: -3. The obtained $O-C$ diagram strongly suggests that the light source

536: of the variations in -4$<E<$8 comes from the same origin, since a

537: phase shift or a change of the period were not seen. This means that the

538: modulations were coherent for -4$<E<$8. As can be seen in figure

539: \ref{elc} the modulations that we observed around $E$=8 were genuine

540: superhumps. Therefore, it is likely that superhumps had already grown at

541: the end of the precursor. If this is the case, EG Aqr is the second

542: example that superhumps were seen even during a precursor, following the

543: 1993 superoutburst of T Leo \citep{kat97tleo}.

544: 

545: Besides the presence of the superhumps during the precursor, the overall

546: light curve is typical of SU UMa-type dwarf novae, in terms of

547: the duration of the plateau phase, as well as the decline rate of the

548: magnitude. The presence of a precursor also supports its SU UMa nature

549: of EG Aqr, since no precursor has been reported for WZ Sge-type dwarf

550: novae \citep{kat04egcnc}.

551: 

552: \subsection{superhump period change}

553: 

554: As is well known, most of SU UMa-type dwarf novae decrease their

555: superhump period during the course of the plateau stage. This is due to

556: shrinkage of the disk, or a natural consequence of the depletion of the

557: mass \citep{osa85SHexcess}. Recent studies, however, have shown that some

558: systems increase the superhump period during the plateau phase

559: \citep{nog97alcom}. So far, the exact mechanism that causes the

560: superhump period change is not known.

561: 

562: Recently, \citet{uem05tvcrv} has suggested that the superhump period change

563: may be related to the presence or absence of a precursor. The authors

564: studied a short period SU UMa-type dwarf nova TV Crv for the 2002 and

565: 2004 superoutburst, and discovered that the superhump period increased

566: during the 2002 superoutburst without a precursor, while hardly changed

567: during the 2004 superoutburst with a precursor. Theoretically,

568: \citet{osa03DNoutburst} proposed the refined thermal-tidal instability

569: model, in

570: which a superoutburst without a precursor have a larger disk radius

571: than that with a precursor. When the accretion disk reaches the tidal

572: truncation radius as a result of the mass accretion onto the white

573: dwarf, the stored matter at the outer edge may prevent the inner

574: propagation of the

575: cooling wave, so that the disk keeps the hot state. On the other hand, if

576: the accretion disk does not reach the tidal truncation radius, the

577: outermost region of the disk allows the cooling wave to propagate

578: inward. As a consequence, the outburst is quenched like a normal

579: one. If the eccentricity of the accretion disk grows sufficiently,

580: the tidal dissipation at the rim of the disk leads to the orientation of

581: the heating wave. This process is observed as a superoutburst with a

582: precursor.

583: 

584: In the case of EG Aqr, the superoutburst was accompanied by a precursor,

585: indicating that the radius of the accretion disk did not reach the tidal

586: truncation radius. This, however, seems to be peculiar when taking into

587: account the long quiescent time of the object. The long quiescence means

588: a large amount of mass in the accretion disk, like WZ Sge-type dwarf

589: novae. However, the duration of the plateau stage, as well as the

590: presence of a precursor are indicative of the small disk of the object

591: compared to WZ Sge-type dwarf novae.

592: 

593: During the main plateau stage, the best fitting quadratic formula

594: yielded $P_{\rm dot}$ = $\dot{P}$/$P$ = $-$8.2$\times$10$^{-5}$, which

595: is a normal value for SU UMa-type dwarf novae. The value is quite different

596: from that observed in WZ Sge-type stars, which show

597: positive period derivatives, or almost constant value of the superhump

598: periods \citep{ish03hvvir}. Although quiescence of EG Aqr behaves as WZ

599: Sge-type dwarf novae, the overall feature of the superhump period

600: changes and the light curve of the superoutburst bear strong resemblance

601: to those of usual SU UMa-type dwarf novae.

602: 

603: Regarding the early phase of the observations, it again should be noted

604: that the superhump period unchanged and a sort of period change occurred

605: near the bright maximum. Although evidence of such period changes has

606: been provided in other SU UMa-type dwarf novae (e.g., figure 5 of

607: \citet{uem05tvcrv}), this is the first case that we specify when the

608: transition occurred. In future, we should investigate whether the period

609: change is a ubiquitous property for SU UMa-type dwarf novae by observing

610: other objects, which shed lights on understanding the origin of the

611: period shift.

612: 

613: \subsection{inactive SU UMa-type dwarf novae with long superhump periods}

614: 

615: So far, as much as 200 objects have been confirmed to be SU UMa-type

616: dwarf novae, for which we may divide them into two categories

617: according to their superhump periods and their tendency of the period

618: derivatives. There exists a rough cutoff around $P_{\rm sh}$=0.063 days

619: below which the superhump period tends to increase during the superoutburst

620: \citep{ima05gocom}. On the other hand, for systems above $P_{\rm

621: sh}$=0.063 days, where the majority of SU UMa-type dwarf novae exists,

622: the superhump period tends to decrease or keep almost constant. The latter

623: systems are sometimes introduced as ``textbook'' SU UMa-type dwarf novae

624: \citep{kat03v877arakktelpucma}.

625: 

626: However, when we focus on the long-term behavior of the respective

627: system, we notice the diversity of the nature. For example, a prototype

628: SU UMa-type star VW Hyi shows frequent normal outbursts between two

629: successive superoutbursts with a mean recurrence time of the superoutburst

630: being 150 days \citep{osa89suuma}. Such systems are well reproduced by the

631: thermal-tidal instability model (\cite{osa89suuma};

632: \cite{osa03DNoutburst}). Some long-period systems, on the other hand, hardly

633: show not only a normal outburst, but also a superoutburst. These systems

634: include EF Peg (\cite{how93efpeg}; \cite{kat02efpeg};

635: \cite{how02llandefpegHST}),

636: V725 Aql (\cite{1996JAVSO..24...14H}; \cite{uem01v725aql}), QY

637: Per\footnote{http://www.kusastro.kyoto-u.ac.jp/vsnet/DNe/qyper.html},

638: and EG Aqr. All of these systems have a long recurrence time of

639: years, indicating that the mass transfer rate from the secondary is

640: supposed to be small. This seems to be curious because the secular

641: evolution of dwarf novae suggests the existence of relatively high mass

642: secondaries compared to those in short period dwarf

643: novae unless the binaries are so called ``period bouncers''

644: (\cite{kin88binaryevolution}; \cite{kol99CVperiodminimum};

645: \cite{pat05re1255}). It is unlikely that these systems had passed

646: the period minimum of dwarf novae, since the evolutional timescale for

647: ``period bouncers'' may be as long as ${\sim}$ 10 Gyrs

648: \citep{pat98evolution}. Optical and infrared spectroscopy show mid

649: M-type secondaries in EF Peg \citep{how02llandefpegHST}. This indicates

650: that this system does not reach the period minimum, because the ``period

651: bouncers'' are believed to contain a brown dwarf secondary

652: \citep{how97periodminimum}.

653: 

654: At present, we cannot specify the reason why such inactive systems do

655: exist in long period systems. In fact, these systems are less studied,

656: especially during quiescence. Actually, we have no information on the

657: orbital period of the above four systems.\footnote{\citet{how93efpeg}

658: determined a likely orbital period of EF Peg as 2.05 hours based on the

659: observed superhump period. It should be noted that this is not a direct

660: measurement of the orbital period.} In future, quiescent spectroscopy is

661: imperative in order to determine the orbital period and the spectral

662: type of the secondary of the system. This would shed some light

663: on the evolutional status of the above mentioned systems.

664: 

665: \section{Conclusion}

666: 

667: We established the SU UMa nature of EG Aqr for the first time by the

668: detection of the superhumps with a mean period of 0.078828(6)

669: days. The observed superoutburst showed a precursor, during which a hint

670: of superhumps was seen at the last stage of the precursor. Extensive

671: observations enabled us to examine detailed changes of the superhump

672: period. It turned out that the superhump period kept almost constant

673: near the bright maximum, after which the superhump period decreased

674: normally at the rate of $P_{\rm dot}$ = $\dot{P}$/$P$ =

675: $-$8.2$\times$10$^{-5}$ all over the plateau stage. Although the origin

676: is unknown, we detected a change of the period near the bright

677: maximum. Concerning the observed period change and the presence of the

678: precursor, the maximum radius of the accretion disk was not large during

679: the superoutburst. The obtained light curves were typical of those of SU

680: UMa-type dwarf novae. This is definitely peculiar when taking into

681: account that EG Aqr showed only one recorded eruption until the 2006

682: superoutburst. Despite the inactivity of the variable, we conclude that

683: EG Aqr is a new member of SU UMa-type dwarf novae. In future, quiescent

684: studies both from photometry and spectroscopy are imperative in order to

685: further understand the enigmatic object.

686: 

687: \vskip 3mm

688: 

689: We would like to thank Dr. Steve B. Howell for helpful comments on the

690: paper. We acknowledge with thanks the variable star observations

691: from the AAVSO and VSNET International Database contributed by observers

692: worldwide and used in this research. We would also express our gratitude

693: to G. Pojmanski for providing invaluable data of ASAS-3

694: observations. This work is supported by a Grant-in-Aid for the 21st

695: Century COE ``Center for Diversity and Universality in Physics'' from

696: the Ministry of Education, Culture, Sports, Science and Technology

697: (MEXT). GM acknowledges the support of the Planetary Society and

698: Software Bisque. This work is partly supported by a grant-in aid from

699: the Ministry of Education, Culture, Sports, Science and Technology

700: (No. 14079206, 16340057, 17684004, 17340054, 17740105, 18740153,

701: 18840032). Part of this work is supported by a Research Fellowship of

702: the Japan Society for the Promotion of Science for Young Scientists (RI,

703: AI, KK, ON).

704: 

705: 

706: 

707: 

708: 

709: 

710: \begin{thebibliography}{}

711: 

712: \bibitem[{Connon Smith}(2007)]{smi07review}

713:   {Connon Smith}, R.\ 2007, ArXiv Astrophysics e-prints

714: 

715: \bibitem[Fernie(1989)]{fer89error}

716:   Fernie, J.~D.\ 1989, \pasp, 101, 225

717: 

718: \bibitem[Haro, Chavira(1960)]{har60egaqr}

719:   Haro, G., \& Chavira, E.\ 1960, the Boletin de los Observatorios Tonantzintla

720:   y Tacubaya, volume 2, number 19, 11

721: 

722: \bibitem[{Hazen}(1996)]{1996JAVSO..24...14H}

723:   {Hazen}, M.~L.\ 1996, Journal of the American Association of Variable Star

724:   Observers (JAAVSO), 24, 14

725: 

726: \bibitem[Hellier(2001)]{hel01book}

727:   Hellier, C.\ 2001, Cataclysmic Variable Stars: how and why they vary

728:   (\PublisherSpringer)

729: 

730: \bibitem[Hoard et~al.(2002)]{hoa02CV2MASS}

731:   Hoard, D.~W., Wachter, S., Clark, L.~L., \& Bowers, T.~P.\ 2002, \apj, 565,

732:   511

733: 

734: \bibitem[Howell et~al.(1996)]{how96alcom}

735:   Howell, S.~B., DeYoung, J.~A., Mattei, J.~A., Foster, G., Szkody, P.,

736:   Cannizzo, J.~K., Walker, G., \& Fierce, E.\ 1996, \aj, 111, 2367

737: 

738: \bibitem[Howell et~al.(2002)]{how02llandefpegHST}

739:   Howell, S.~B., Gaensicke, B.~T., Szkody, P., \& Sion, E.~M.\ 2002,

740:   \apj, 575, 419

741: 

742: \bibitem[Howell et~al.(1997)]{how97periodminimum}

743:   Howell, S.~B., Rappaport, S., \& Politano, M.\ 1997, \mnras, 287, 929

744: 

745: \bibitem[Howell et~al.(1993)]{how93efpeg}

746:   Howell, S.~B., Schmidt, R., DeYoung, J.~A., Fried, R., Schmeer, P., \& Gritz,

747:   L.\ 1993, \pasp, 105, 579

748: 

749: \bibitem[Howell et~al.(1995)]{how95TOAD}

750:   Howell, S.~B., Szkody, P., \& Cannizzo, J.~K.\ 1995, \apj, 439, 337

751: 

752: \bibitem[{Imada} et~al.(2006)]{2006PASJ...58..383I}

753:   {Imada}, A., {Kato}, T., {Monard}, L.~A.~G., {Retter}, A., {Liu}, A., \&

754:   {Nogami}, D.\ 2006, \pasj, 58, 383

755: 

756: \bibitem[Imada et~al.(2005)]{ima05gocom}

757:   Imada, A., {et~al.}\ 2005, \pasj, 57, 193

758: 

759: \bibitem[Ishioka et~al.(2001)]{ish01rzleo}

760:   Ishioka, R., {et~al.}\ 2001, \pasj, 53, 905

761: 

762: \bibitem[Ishioka et~al.(2003)]{ish03hvvir}

763:   Ishioka, R., {et~al.}\ 2003, \pasj, 55, 683

764: 

765: \bibitem[Kato(1997)]{kat97tleo}

766:   Kato, T.\ 1997, \pasj, 49, 583

767: 

768: \bibitem[Kato(2002)]{kat02efpeg}

769:   Kato, T.\ 2002, \pasj, 54, 87

770: 

771: \bibitem[Kato et~al.(2004)]{kat04egcnc}

772:   Kato, T., Nogami, D., Matsumoto, K., \& Baba, H.\ 2004, \pasj, 56, S109

773: 

774: \bibitem[Kato et~al.(2003)]{kat03v877arakktelpucma}

775:   Kato, T., {et~al.}\ 2003, \mnras, 339, 861

776: 

777: \bibitem[King(1988)]{kin88binaryevolution}

778:   King, A.~R.\ 1988, \qjras, 29, 1

779: 

780: \bibitem[Kolb, Baraffe(1999)]{kol99CVperiodminimum}

781:   Kolb, U., \& Baraffe, I.\ 1999, \mnras, 309, 1034

782: 

783: \bibitem[{Luyten}, {Haro}(1959)]{1959PASP...71..469L}

784:   {Luyten}, W.~J., \& {Haro}, G.\ 1959, \pasp, 71, 469

785: 

786: \bibitem[Nogami et~al.(1997)]{nog97alcom}

787:   Nogami, D., Kato, T., Baba, H., Matsumoto, K., Arimoto, J., Tanabe, K., \&

788:   Ishikawa, K.\ 1997, \apj, 490, 840

789: 

790: \bibitem[Osaki(1985)]{osa85SHexcess}

791:   Osaki, Y.\ 1985, \aap, 144, 369

792: 

793: \bibitem[Osaki(1989)]{osa89suuma}

794:   Osaki, Y.\ 1989, \pasj, 41, 1005

795: 

796: \bibitem[Osaki(1995)]{osa95wzsge}

797:   Osaki, Y.\ 1995, \pasj, 47, 47

798: 

799: \bibitem[Osaki(1996)]{osa96review}

800:   Osaki, Y.\ 1996, \pasp, 108, 39

801: 

802: \bibitem[Osaki, Meyer(2003)]{osa03DNoutburst}

803:   Osaki, Y., \& Meyer, F.\ 2003, \aap, 401, 325

804: 

805: \bibitem[Patterson(1998)]{pat98evolution}

806:   Patterson, J.\ 1998, \pasp, 110, 1132

807: 

808: \bibitem[Patterson et~al.(2005)]{pat05re1255}

809:   Patterson, J., Thorstensen, J., \& Kemp, J.\ 2005, \pasp, 117, 427

810: 

811: \bibitem[Sproats et~al.(1996)]{spr96CVabsmag}

812:   Sproats, L.~N., Howell, S.~B., \& Mason, K.~O.\ 1996, \mnras, 282, 1211

813: 

814: \bibitem[Stellingwerf(1978)]{ste78pdm}

815:   Stellingwerf, R.~F.\ 1978, \apj, 224, 953

816: 

817: \bibitem[Szkody, Howell(1992)]{szk92CVspec}

818:   Szkody, P., \& Howell, S.~B.\ 1992, \apjs, 78, 537

819: 

820: \bibitem[Templeton et~al.(2006)]{tem06asas0025}

821:   Templeton, M.R., {et~al.}\ 2006, \pasp, 118, 226

822: 

823: \bibitem[Uemura et~al.(2001)]{uem01v725aql}

824:   Uemura, M., Kato, T., Pavlenko, E., Baklanov, A., \& Pietz, J.\ 2001, \pasj,

825:   53, 539

826: 

827: \bibitem[Uemura et~al.(2005)]{uem05tvcrv}

828:   Uemura, M., {et~al.}\ 2005, \aap, 432, 261

829: 

830: \bibitem[{Voges} et~al.(1999)]{1999A&A...349..389V}

831:   {Voges}, W., {et~al.}\ 1999, \aap, 349, 389

832: 

833: \bibitem[Vogt, Bateson(1982)]{vog82atlas}

834:   Vogt, N., \& Bateson, F.~M.\ 1982, \aaps, 48, 383

835: 

836: \bibitem[Warner(1995)]{war95book}

837:   Warner, B.\ 1995, Cataclysmic Variable Stars (\PublisherCambridge)

838: 

839: \bibitem[{Watanabe} et~al.(2005)]{2005PASP..117..870W}

840:   {Watanabe}, M., {et~al.}\ 2005, \pasp, 117, 870

841: 

842: \bibitem[Whitehurst(1988)]{whi88tidal}

843:   Whitehurst, R.\ 1988, \mnras, 232, 35

844: 

845: \end{thebibliography}

846: 

847: \end{document}

848: