astro-ph0210072/ms.tex
1: % Standard AAS-TeX format
2: \documentclass[12pt,preprint]{aastex}
3: % Single column preprint format
4: %\documentclass[preprint]{aastex}
5: % Double column preprint format
6: %\documentclass[preprint2]{aastex}
7: %\usepackage[twocolumn]{emulateapj5}
8: 
9: 
10: %\received{}
11: %\revised{}
12: %\accepted{}
13: %\journalid{\?}{\?}
14: %\articleid{\?}{\?}
15: %\cpright{\?}{\?}
16: %\ccc{\?}
17: 
18: % Running heads go here
19: \shorttitle{X-ray Light Curve of V1494 Aql}
20: \shortauthors{Drake et al.}
21: 
22: % begin document
23: 
24: \begin{document}
25: 
26: % Give title between brackets
27: \title{The Extraordinary X-ray Light Curve of the Classical Nova
28: V1494 Aquilae (1999 \#2) in Outburst: The Discovery of Pulsations
29: and a ``Burst''}
30: 
31: % Give authors and affiliations here.  Use \altaffilmark{n} appended to
32: % author's name to indicate
33: % special footnoted items such as Visiting Astronomer, mailing address, etc.
34: 
35: 
36: % author list
37: 
38: \author{Jeremy J.~Drake\altaffilmark{1}, R. Mark Wagner\altaffilmark{2}, 
39: Sumner Starrfield\altaffilmark{3}, Yousaf Butt\altaffilmark{1}, 
40: Joachim Krautter\altaffilmark{4}, H. E. Bond\altaffilmark{5}, M. 
41: Della Valle\altaffilmark{6}, R. D. Gehrz\altaffilmark{7}, Charles E. 
42: Woodward\altaffilmark{7}, A. Evans\altaffilmark{8}, M. 
43: Orio\altaffilmark{9}, P. Hauschildt\altaffilmark{10}, M. 
44: Hernanz\altaffilmark{11}, K. Mukai\altaffilmark{12}, \& J. W. 
45: Truran\altaffilmark{13}}
46: 
47: \altaffiltext{1}{Smithsonian Astrophysical Observatory, 60 
48: Garden Street, Cambridge, MA 02138; (jdrake,ybutt)@cfa.harvard.edu}
49: \altaffiltext{2}{Large Binocular Telescope Observatory, 933 North Cherry Avenue, 
50: Tucson, Arizona 85721; rmw@as.arizona.edu}
51: \altaffiltext{3}{Department of Physics \& Astronomy, Arizona State 
52: University, Tempe, Arizona 85287; sumner.starrfield@asu.edu}
53: \altaffiltext{4}{Heidelberg-Koenigstuhl Landessternwarte, Heidelberg 
54: 1, D-69121 Germany}
55: \altaffiltext{5}{Space Telescope Science Institute, 3700 San Martin 
56: Drive, Baltimore, MD 21218; bond@stsci.edu}
57: \altaffiltext{6}{Dipartamento di Astronomia e Scienza dello Spazio e 
58: Osservatorio Astronomico di Arcetri, Largo Enrico Fermi, 5, Firenzi, 
59: I-50125 Italy}
60: \altaffiltext{7}{Department of Astronomy, University of Minnesota, 
61: 116 Church Street SE, Minneapolis, MN 55455; 
62: (gehrz,chelsea)@astro.umn.edu}
63: \altaffiltext{8}{Department of Physics, Keele University, Keele, 
64: Staffordshire, ST5 5BG}
65: \altaffiltext{9}{Oss. Astronomico di Torino, Strada Osservatorio 20, 
66: Pino Torinese I-10025 Italy}
67: \altaffiltext{10}{Hamburger Sternwarte, Gojenbergsweg 112, 
68: 21029 Hamburg, Germany; phauschildt@hs.uni-hamburg.de}
69: %\altaffiltext{10}{Deparment of Physics \& Astronomy, University of 
70: %Georgia, Athens, GA 30602-2451; yeti@hal.physast.uga.edu}
71: \altaffiltext{11}{Instituto de Ciencias del Espacio (CSIC), \& 
72: Institut d'Estudis Espacials de Catalunya, Edifici Nexus-201, C/ Gran 
73: Capit\`{a} 2-4, E-08034 Barcelona, Spain}
74: \altaffiltext{12}{NASA Goddard Space Flight Center, Code 662, 
75: Laboratory for High Energy Astrophysics, Greenbelt, MD 20771; 
76: mukai@milkyway.gsfc.nasa.gov}
77: \altaffiltext{13}{Department of Astronomy \& Astrophysics, University 
78: of Chicago, 5640 Ellis Avenue, Chicago, IL 60637; 
79: truran@nova.uchicago.edu}
80: 
81: 
82: 
83: \begin{abstract}
84: V1494 Aql (Nova Aql 1999 No. 2) was discovered on 2
85: December  1999.  We obtained {\it Chandra} ACIS-I spectra on 15 April
86: and 7 June 
87: 2000 which appear to show only emission lines.  Our third
88: observation, on 6 August, showed that its spectrum had evolved
89: to that characteristic of a Super Soft X-ray Source. 
90: We then obtained {\it Chandra} LETG+HRC-S spectra on 28
91: September (8 ksec) and 1 October (17  ksec).
92: We analyzed the X-ray light curve of our grating
93: observations and found both a short time scale ``burst'' and
94: oscillations.  Neither of these phenomena have previously been seen in
95: the light curve of a nova in outburst.  The ``burst'' was a factor
96: of $\sim$10 rise in X-ray counts near the middle of the second
97: observation, and which lasted about 1000 sec; it exhibited at least
98: two peaks, in addition to other structure.  Our time series analysis of the
99: combined 25 ksec observation shows a peak at $\sim$2500~s which
100: is present in independent analyses of both the zeroth order
101: image and the dispersed spectrum and is not present in similar
102: analyses of grating data for HZ 43 and Sirius B.
103: Further analyses of the V1494 Aql data find other periods
104: present which implies that we are observing non-radial g$^+$ modes
105: from the pulsating, rekindled white dwarf.
106: \end{abstract}
107: 
108: 
109: % Give list of keywords 
110: 
111: \keywords{stars:  individual (V1494 Aquilae) --- stars:
112: novae, cataclysmic variables --- stars:  oscillations --- stars:  white dwarfs
113: --- X-rays:  binaries --- X-rays:  bursts --- X-rays:  individual (V1494
114: Aquilae) --- X-rays:  stars}
115: 
116: 
117: \section{Introduction}
118: 
119: Classical novae (CN) are the third most violent stellar
120: explosions after $\gamma$-ray bursts and supernovae.   At
121: maximum light, their bolometric luminosities exceed
122: $10^5$L$_\odot$, and they eject more than
123: $10^{-4}$M$_\odot$ of material into space, with strongly
124: non-solar compositions.
125: Previous X-ray (XR) studies of novae {\it in outburst} were
126: performed by EXOSAT, ROSAT, and most recently ASCA,
127: {\it Bepposax} and {\it Chandra}.  EXOSAT detected QU Vul, PW
128: Vul, RS Oph, and GQ Mus in the constant bolometric
129: luminosity phase (\"Ogelman et al. 1987).  ROSAT detected
130: V838 Her, V351 Pup, V1974 Cyg, GQ Mus, and LMC 1995
131: (Krautter et al. 1996, K96; Orio et al. 2001a). {\it The analyses of these
132: data show that no other wavelength region provides
133: unambiguous information on the evolution of the thermonuclear runaway on the
134: white dwarf} (WD).  The ROSAT studies identified two components of the
135: XR emission from a nova in outburst.  The first is a soft
136: component that places novae at, or near, the bright end of the
137: class of Super Soft X-ray Sources (hereafter SSS; at low
138: resolution the spectrum resembles an ``emission'' line with a peak $\sim$
139: 0.5 keV).  There is also a ``hard'' component, $\sim$ 1 keV, that was
140: seen in 3 of the 5 novae studied with ROSAT (K96, Orio et al. 2001a;
141: and in V382 Vel studied
142: with {\it Bepposax} 
143: and {\it Chandra} (Orio et al. 2001a; Starrfield et al. 2000).
144: It was the ROSAT studies of V1974 Cyg, however, that established
145: the connection between the SSS and CN in outburst (K96).
146: It stayed bright in XR for more than
147: 18 months and reached a peak of 76 cts s$^{-1}$ before
148: beginning its decline to quiescence (K96; Balman et al. 1998,
149: B98).
150: 
151: 
152: V1494 Aquilae (Nova Aql 1999 \#2) was discovered in the
153: optical at m$_{v}\sim 6$ on 1999 December 1.785 UT by
154: Pereira (1999). 
155: This nova then reached a peak magnitude
156: exceeding 4 within a few days of discovery.
157: A report on early optical spectroscopic
158: observations can be found in Kiss \& Thomson (2000) who
159: described it as a member of the Fe II class as defined by
160: Williams (1992).  This suggests that the nova ejected an
161: optically thick cloud of material.
162: 
163: 
164: We activated our {\it Chandra} Target - of Opportunity (ToO)
165: Program in January 2000 and obtained three observations with
166: the ACIS - I instrument followed by two observations with the
167: LETG+HRC-S instrument (separated by 4 days because of a
168: $\gamma$-ray burst ToO).  Our first ACIS-I spectrum was
169: obtained on 15 April 2000.  Both it and a second ACIS-I
170: spectrum (7 June) showed an emission line spectrum with no
171: obvious soft component (Starrfield et al. 2000; 2001).  Our third
172: ACIS-I spectrum on 6 August, however, showed that the nova
173: had evolved into a bright SSS 
174: with a few weak emission lines at higher energies
175: (Starrfield et al. 2000).  We obtained an LETG+HRC-S
176: spectrum at the end of September and this observation showed
177: the high dispersion spectrum of a SSS.  While it resembles the
178: XMM spectrum of the SSS CAL 83 (Paerels et al. 2001), it
179: appears somewhat hotter and the ``emission features'' are at
180: different wavelengths. The evolution from a hard source to that
181: of a SSS was also seen in the ROSAT observations of V1974
182: Cyg (K96) and combined ASCA-Bepposax observations of
183: V382 Vul (Orio et al. 2001b, 2002; Mukai \& Ishida 2001).
184: 
185: 
186: We will report on the spectral development and analyses of the
187: spectra elsewhere. Because of its spectacular nature, in this 
188: paper we concentrate only on the light
189: curve of the XR data obtained with {\it Chandra}.  In the next
190: section we describe the observations and the method of
191: obtaining the light curve.  In section 3 we describe
192: our time-series analyses, section 4 contains a discussion of
193: our results, and the last section is the
194: summary.
195: 
196: \section{Observations and Analysis}
197: 
198: V1494 Aql was observed in two segments using the
199: LETG+HRC-S.  The first observation (ID 2308) was carried out
200: on 2000 September 28 between UT~06:51:13 and 09:36:56, and
201: the second (ID 72) three days later on 2000 October 1, between
202: UT~10:08:56 and 15:39:30.  Initial reduction of satellite
203: telemetry was performed by the {\it Chandra} X-ray Center
204: Standard Data Processing software version R4CU5UPD9.  The
205: analysis described here is based on the Level 2 products of this
206: processing, together with instrument dead time corrections that
207: are products of Level 1 processing.  Dead time corrections
208: generally need to be taken into account in the determination of
209: exposure times when the total detector count rate in a particular
210: interval exceeds the telemetry limit of 183 count~s$^{-1}$. The
211: computed corrections are a multiplicative factor that, when
212: averaged and applied to the elapsed exposure time for a
213: particular time interval, yield the effective exposure time for that
214: interval.  Both of our observations experienced episodes of
215: telemetry saturation and we have included dead time effects in
216: all the subsequent analyses.  However, these episodes were
217: sufficiently short and infrequent so as to have no significant
218: affect on the derived light curves.  Total effective exposure
219: times for the first and second observation were 8146 and
220: 18232s, respectively.
221: 
222: 
223: We computed light curves from the bright zeroth order of the
224: LETG+HRC-S spectrum as seen in the Level 2 photon event list
225: using the PERL program LCURVE (P.~Ratzlaff, unpublished).
226: A circular source region with a diameter of 10 arcsec was
227: employed, together with a surrounding annular region for
228: background estimation, sized so as to be a factor ten larger in
229: surface area than the source region.  The source region was
230: sufficiently large so as to include all significant signal in the
231: LETG coarse support structure diffraction wings, and the
232: background region sufficiently small so as not to include any
233: significant amount of the first order dispersed spectrum.  
234: 
235: 
236: \section{Results}
237: 
238: \subsection{Light Curves}
239: 
240: Our light curves are illustrated in Figures~1 and 2.  The first of
241: these shows the entire observations of first (ID 72) and second (ID
242: 2308) segments using 25~s temporal bins.  Note that all three plots
243: cover the same size time interval (20~ks).  The upper two panels of
244: this figure cover the same range in count rate, while the lower panel
245: shows the full range of count rate variation scaled by the flare
246: observed near the middle of the second segment.  Readily apparent in
247: both segments is stochastic variability on time scales of minutes.  In
248: addition, a slower modulation with a period $\sim$2000-3000~s appears
249: to be present. In the third panel, the flare corresponds to a
250: $\sim$10-fold increase in intensity over the quiescent background
251: signal and lasts for $\sim$820~s.
252: 
253: In Figure~2, we show the region around the time of the outburst in the
254: second segment with a temporal bin size of 25~s.  It is clear that the
255: flare is not a single isolated event, but exhibits two main large
256: flares and perhaps a precursor and trailer.  Examining Figure~2, a
257: precursor to the main flare appears to begin at t $\simeq$ 8850 with a
258: 3-fold increase in intensity and lasts for $\sim$270~s.  This event is
259: rapidly followed by an increase of $\sim$2.5 over the next 130~s to a
260: well-defind narrow peak which lasts $\sim$60~s.  A local minimum is
261: reached at t $\simeq$ 9340 but the intensity increases rapidly again
262: peaking at t $\simeq$ 9390, a factor of $\sim$10 above the quiescent
263: background intensity.  The duration of the second flare appears
264: comparable to the first.  Approximately 110~s later, the intensity has
265: declined to nearly the quiescent level, but appears to be quickly
266: followed by a small trailing flare event which lasts for $\sim$170~s.
267: By t $\simeq$ 9670, the burst is over.
268: 
269: \subsection{Timing Analysis}
270: 
271: In order to better understand the nature of the temporal variations
272: seen in the light curves and discussed above, we performed four
273: different time series analyses, each with increasing sophistication.
274: These included: (1) a simple periodogram formed by fitting a cosine
275: function over an array of trial periods; (2) phase dispersion
276: minimization (PDM: Stellingwerf 1978);  (3) a Fourier power
277: spectrum (Deeming 1975); and (4) a "cleaned" Fourier power spectrum (Roberts,
278: et al. 1987) in which the sampling or spectral window function is
279: removed from the raw or dirty power spectrum.  This latter robust algorithm
280: is particularly well suited to the analysis of a multiperiodic time
281: series of unequally spaced data.
282: 
283: In preparation for the 4 techniques discussed above, the flare
284: detected in the second segment was excised from the 
285: HRC-S photon event lists.  This removal created three
286: separate segments, each with a duration of 8000-9000~s: a segment
287: consisting of the entire first observation (ID 2308) followed by two
288: segments from the second observation (ID 72).  These data are sampled
289: in 25~s temporal bins, and are sensitive to periodic modulations with
290: periods ranging from twice the bin width (50 s) to one half of the
291: total time span covered by the observations (13,000 s).  
292: 
293: We found that the time series spectra derived from a simple cosine
294: periodogram, a PDM, and a Deeming power spectrum, are characterised by
295: a significant periodicity at 2500~s as well as other possibly
296: significant features.
297: %All these techniques are designed to find a
298: %single periodicity, but there is evidence from the temporal spectra
299: %that the X-ray light curve of V1494 Aql is probably multiperiodic or
300: %perhaps quasiperiodic.
301: This interpretation is complicated by the fact that strong sidelobes
302: of the sampling or window function produce spurious features
303: throughout the power spectrum and thus affect the shape and power of
304: significant features.  Many attempts have been made to recognize these
305: spurious features (Deeming 1975) but they remain problematic.
306: 
307: Recently, Roberts et al. (1987) have developed a one--dimensional
308: variation of the two dimensional {\sc{CLEAN}} (H\"{o}gbom 1974)
309: algorithm which is well suited to the temporal analysis of unequally
310: spaced finite data samples as we consider here.  We have applied this
311: version of the {\sc CLEAN} algorithm to our X-ray time series data of
312: V1494 Aql, binned this time at 75s intervals, 
313: and the results are shown in Figure~3.  Using this algorithm,
314: the spectral window function has been iteratively subtracted in the
315: Fourier domain from the raw or dirty power spectrum shown in
316: Figure~3 (top panel).  We assumed a gain of 0.2 and stopped after 10,000
317: iterations since no further improvement in the residual spectrum was
318: apparent.  Small gain values prevent the accumulation of errors in the
319: subtraction of the sidelobes at the expense of computation time, and
320: are preferred over gains close to unity and fewer iterations.  The
321: data were averaged in 75~s bins for this analysis.  A frequency
322: resolution of 1.0 $\times$ 10$^{-5}$ Hz was chosen.
323: 
324: The cleaned power spectrum is shown in the bottom panel of Figure~3.
325: It is difficult to attach an overall significance level to features
326: appearing in the power spectrum because it depends on both the
327: sampling and the signal-to-noise ratio of the data.  To help assess
328: the reality of features appearing in the power spectrum, therefore, we
329: added a tracer periodic signal to the time series data with a
330: frequency of 0.0011 Hz and an amplitude of 5\% of the mean signal
331: level.  The location of this signal is indicated in Figure~3 and we
332: note that the signal is easily detected suggesting that significant
333: signals in this frequency range might have amplitudes of a few
334: percent.  The strongest peak is at $\nu$ = 0.0004002 Hz (P = 2498.8~s)
335: and has an amplitude of 15\% (0.11 counts/s) of the mean count rate
336: level.  The strength and detectability of the artificial signal
337: suggests that other comparably strong features are real as well and
338: thus the X-ray variations might be multiperiodic.  These include a
339: double--peaked feature at 0.00029 Hz (3461~s), 0.00059 Hz (1695~s),
340: 0.00079 Hz (1266~s), 0.00097 Hz (1031~s), 0.00124 Hz (809~s), and
341: perhaps 0.0019 Hz (526~s).
342: 
343: In addition to our main analysis of the 75~s binned data, we also
344: examined the cleaned power spectra of these data averaged in 25~s and
345: 150~s temporal bins to further assess their reality.  The features
346: listed above all appear in these additional power spectra at
347: significant levels as defined by the tracer signal that we
348: introduced into the data.   
349: 
350: After further investigation, we found that the 2500~s period is close
351: to the beat period of 2623.98~s formed from the two orthogonal
352: spacecraft dither motions (with periods of 768.57~s and 1086.96~s in
353: pitch and yaw, respectively).  However, 
354: neither the pitch or yaw periods are present in the power spectra so
355: it is difficult to understand how the beat period could be present in
356: isolation.  Nevertheless, in order to determine whether or not the
357: 2500~s periodicity might be related to the spacecraft motion we
358: performed identical analyses on light curves derived from calibration
359: observations of the hot white dwarf stars HZ~43 and Sirius~B (ObsID's
360: 59 and 1452).  These light curves were expected to be featureless, and
361: indeed their periodograms were mostly flat over the interval from 300
362: to 4000~s.  The power spectrum of Sirius B shows a broad ($\Delta$t
363: $\simeq$ 500~s FWZI) and a weak feature at 2500~s but it is significant
364: at less than the 2$\sigma$ level and thus not characteristic of the
365: precise spacecraft dither motions.
366: 
367: We also used the {\it Chandra}
368: Interactive Analysis of Observations (CIAO) software to isolate the
369: regions around the dispersed spectra and
370: obtained the light curve for these regions which were sampled in 50~s
371: bins.  We then performed the same cleaned power spectrum analyses on
372: the light curve of the dispersed spectra.  This analysis thus serves
373: as an independent check on the results described above.  We find the
374: same major period at $\sim$2500~s is present in the dispersed light as
375: well as the additional periods listed above.  These results, are {\it
376: obtained completely independently of the zeroth order light curve},
377: and strengthen our conclusion that the observed periods are not
378: instrumental.  
379: 
380: We can safely exlude instrumental background as being responsible for
381: any or part of the time-variable signals we have detected.  In the
382: case of the 0th order light curve, for example, the background rate
383: scaled to the source region size amounts to 0.5-$1\,
384: 10^{-2}$~count~s$^{-1}$, as compared to the quiescent source rate of
385: $\sim 0.6$~count~s$^{-1}$: the background is completely negligible.
386: We also examined the Chandra on-board particle monitor and found that
387: no unusual episodes occurred during the interval of our V1494~Aql
388: observation.
389: 
390: Another interesting feature is present in the cleaned power spectrum.
391: A strong signal appears at a frequency of $9.03 \times 10^{-5}$ Hz
392: which is consistent with the observed optical photometric orbital
393: period of V1494 Aql (0.13467~d; Retter et al. 2000) within the errors.
394: 
395: \section{Discussion}
396: 
397: There are two obvious and distinct features in these light curves, the
398: complex rise and fall in the observed count rate, which lasts about
399: 1000 sec, and the periodic signal.  We currently have no explanation
400: for the burst. The spectrum obtained during the burst is slightly
401: different from the spectrum of the rest of the observation (Starrfield
402: et al. 2001).  Once the orbital phase of the burst is known, it may
403: provide more insight as to the cause.
404: 
405: 
406: The periodic signal demands more discussion.   We have done a
407: period analysis of both  observations and find a 2500 s period in
408: the light curve.
409: We also find additional periods present in both the zeroth order
410: and the dispersed data suggesting that we are not observing the
411: rotation of the underlying WD. {\it We interpret this
412: result as the discovery of non-radial g$^+$-mode pulsations in
413: the hot, rekindled WD.}  At this stage in its evolution
414: from explosion to quiescence, the hot, luminous WD has a
415: structure which resembles that of the central stars of planetary
416: nebulae.  It contains a white dwarf core with a composition of
417: carbon and oxygen and a thin layer of accreted plus core
418: material on the surface in which nuclear burning is occurring at
419: or near the bottom. Both this star and the central stars of
420: planetary nebulae have effective temperatures exceeding
421: $10^{5}$K and, when analyzed, the spectrum will show
422: evidence for hydrogen depletion and a high abundance of carbon
423: and oxygen.  This star, therefore, resembles the planetary nebulae
424: nuclei showing O~VI in emission and absorption in optical-UV spectra,
425: some of which have been shown to pulsate at periods ranging
426: from $\sim$1000 s to $\sim$5000 s (Ciardullo \& Bond 1996).
427: The 2500 s period that we have found is close to periods found
428: in observations of NGC  2371-2, RX J2117+3412, or NGC 1501
429: for example (Ciardullo \& Bond  1996, and references therein).
430: 
431: 
432: Because of the similarity in structure we speculate that pulsation
433: driving in this star is analogous to that of either the pulsating
434: Planetary Nebulae Nuclei (PNN) or the GW Vir stars (see
435: Dreizler et al. 1995 and references therein).  For these stars it
436: has been proposed that pulsations are driven by the
437: $\kappa$/$\gamma$ effects---the instability caused by compression-induced 
438: increases in opacity and ionization---in the partial ionization zones of
439: carbon and oxygen near the surface (Starrfield et al. 1984,
440: 1985).  In support of the possibility that this mechanism could
441: be acting in V1494 Aql,  we point out that the nova mechanism
442: requires CO nuclei to have been mixed up into the accreted
443: layers and this WD should have enriched CO near the surface.
444: However, studies at that time suggested that a significant
445: amount of hydrogen could poison pulsation driving.  Since the
446: WD is still luminous and presumably still burning hydrogen, it
447: should still have hydrogen near the surface which would argue
448: against the $\kappa$/$\gamma$ mechanism.  Recent
449: observational studies, however, have found pulsations in
450: members of the GW Vir class which do have hydrogen present
451: at the surface (Dreizler et al. 1996) and, therefore, the
452: $\kappa$/$\gamma$ mechanism remains a viable possibility.
453: 
454: 
455: If theoretical studies of this star show that the
456: $\kappa$/$\gamma$ mechanism cannot be responsible for the
457: oscillations, then another possibility is the $\epsilon$
458: mechanism.  In this case, pulsation driving is caused by the
459: ongoing nuclear energy generation near the surface.
460: The pulsations would then last as long
461: as there was  nuclear burning in the envelope of the white dwarf.  
462: However, while
463: existing studies of this phenomenon either in nova models
464: (Sastri \& Simon 1973),  high luminosity degenerate stars (Vila
465: \& Sion 1976), or in hydrogen shell burning PNN (Kawaler
466: 1988) have found instabilities, they are at periods far shorter than we
467: have observed in V1494 Aql. 
468: 
469: 
470: 
471: \section {Summary}
472: 
473: We have obtained LETG+HRC-S spectra of V1494 Aql about
474: 10 months after discovery.  The X-ray light curve shows a
475: double peaked, short timescale, rise of $\sim$10 in
476: counts per second.  We have no explanation for this rise.
477: 
478: 
479: Power spectrum analyses of the zeroth order image
480: shows a dominant period at 2498.8 s.  There are
481: other periods present our analyses which implies that we are
482: observing pulsations and not rotation.
483: 
484: A period of $\sim$2500 s falls in the range of those measured
485: for pulsating PNN such as NGC 1501 or K1-16.  This is to be
486: expected, if we are observing pulsation, since the structure of the
487: WD at the time of our observations resembles that of a PNN.
488: 
489: 
490: 
491: \acknowledgments
492: 
493: We are
494: grateful to E. M. Sion for valuable scientific discussions and to the
495: referee who made valuable suggestions for improving the text.  We thank
496: P.~Ratzlaff for developing the LCURVE software used in this work.  JJD
497: and YB were supported by NASA contract NAS8-39073 to the {\em Chandra
498: X-ray Center}.  SS, RMW, PHH, JWT, RDG, and CEW were partially
499: supported by grants from the Chandra X-ray Center to their various
500: institutions.  Finally, 
501: we would like to thank the mission planners and staff of the CXC for
502: flawless scheduling and execution of this ToO program.  
503: 
504: 
505: 
506: \begin{thebibliography}{}
507: 
508: 
509: \bibitem[Balman, Krautter, \"Ogelman(1998)]{1998ApJ...499..395B} Balman,
510: S., Krautter, J., \& \"Ogelman, H.\ 1998, \apj, 499, 395
511: 
512: \bibitem[ciardullobond]{1996AJ....111.2332C} Ciardullo, R., \& Bond, H. E.,
513:  1996, AJ, 111, 2332
514: 
515: 
516: 
517: \bibitem[Deeming (1975)]{ApSS....36..137} Deeming, T. J. 1975, \apss, 36, 137
518: 
519: 
520: 
521:  
522: \bibitem[dreizleretal1995]{bbb} Dreizler, S., Werner, K., Heber, U., 1995, in
523:  Proceedings of the 9th European Workshop on White Dwarfs Lecture 
524: Notes in Physics, Vol. 443, ed. D. Koester and K. Werner. Springer-Verlag, New 
525: York, 160
526: 
527: \bibitem[dreizleretal1996]{1996A&A...309..820D} Dreizler, S., Werner,
528:  K., Heber, U., \& 
529:  Engels, D., 1996, A\&A, 309, 820
530: 
531: 
532: \bibitem[H{\" o}gbom(1974)]{1974A&AS...15..417H} H{\" o}gbom, J.~A.\ 1974, 
533: \aaps, 15, 417  
534: 
535: 
536: \bibitem[kaw88]{1988ApJ...334..220K} Kawaler, S. D.  1988, ApJ, 334, 220
537: 
538: 
539: \bibitem[Kiss \& Thomson(2000)]{2000A&A...355L...9K}
540: Kiss, L.\ L., \&  Thomson, J.\ R.\ 2000, \aap, 355, L9
541: 
542: 
543: 
544: \bibitem[Krautter et al.(1996)]{1996ApJ...456..788K} Krautter,
545: J.,  \"Ogelman, H., Starrfield, S., Wichmann, R., \& Pfeffermann, E.\
546: 1996, \apj,  456, 788
547: 
548: 
549: \bibitem[mukaiishida2001]{2001ApJ...551.1024M} Mukai, K., \& Ishida, M. 
550:  \  2001, ApJ, 551, 1024  
551: 
552: 
553: \bibitem[OKB(1987)]{123}
554: \"Ogelman, H., Krautter, J., Beuermann, K.\ 1987, \aap, 177, 110
555: 
556: 
557: 
558: \bibitem[orioetal01a]{2001A&A...373..542O}Orio, M., Covington, J., \& \"Ogelman, 
559: H.
560:  2001a, A\&A, 373, 542
561: 
562: \bibitem[orioetal01b]{2001MNRAS.326L..13O} Orio, M., Parmar, A.,
563:  Benjamin, R., Amati, L., 
564:  Frontera, F., Greiner, J., \"Ogelman, H., Mineo, T.,  Starrfield, S.,
565: \& Trussoni, E. \ 2001b, MNRAS, 326, L13
566: 
567: 
568: \bibitem[orioetal02]{2002MNRAS.333L..11O} Orio, M., Parmar, A. N., Greiner, J.,
569:  \"Ogelman, H., Starrfield, S., Trussoni, E.\ 2002, MNRAS, 333, L11
570: 
571: 
572: \bibitem[Paerels et al.(2001)]{2001AA...365L.308P} Paerels, F., Rasmussen,
573: A.\ P., Hartmann, H.\ W., Heise, J., Brinkman, A.\ C., de Vries,
574: C.\ P.,  den Herder, J.\ W.\ 2001, \aap, 365, L308
575: 
576: 
577: 
578: \bibitem[Pereira(1999)]{1999IAUC.7323} Pereira, A., 1999, \iaucirc,
579: 7323, 1
580: 
581: \bibitem[Retter et al.(2000)]{2000IAUC.7537....2R} Retter, A., Cook, L., 
582: Novak, R., Saxton, J.~M., Jensen, L.~T., Korcakova, D., \& Janik, J.\ 2000, 
583: \iaucirc, 7537, 2 
584: 
585: \bibitem[Roberts et al. (1987)]{1987AJ...93..968} Roberts, D. H., Leh\'ar, J., 
586: \& Dreher, J. W. 1987, \aj, 93, 968
587: 
588: 
589: \bibitem[sastrisimon]{ 1973ApJ...186..997S} Sastri, V.\ K., \& Simon, N. 1973,
590:  ApJ, 186, 997
591: 
592: 
593: \bibitem[starrfield1984]{1984ApJ...281..800S} Starrfield, S., Cox,
594:  A. N., Kidman, R. B., \& 
595:  Pesnell, W. D. 1984, ApJ, 281, 800
596: 
597: 
598: \bibitem[starrfield1985]{1985ApJ...293L..23S} Starrfield, S., Cox,
599:  A. N., Kidman, R. B., \& 
600:  Pesnell, W. D. 1985, ApJ, 293, L23
601: 
602: 
603: 
604: \bibitem[Starrfield et al.(2000)]{2000HEAD...32.4103S}
605: Starrfield, S.\ et  al.\ 2000, AAS/High Energy Astrophysics Division, 32, 4103
606: 
607: \bibitem[starrfield2001]{2001bulaas}
608: Starrfield, S. \  et al. 2001, Bull. AAS, 198, 11.09.
609: 
610: 
611: \bibitem[Stellingwerf (1978)]{1978ApJ,,,224..953} Stellingwerf, R. F. 1978, 
612: \apj, 224, 953
613: 
614: 
615: \bibitem[vilasion76]{1976ApJ...207..820V} Vila, S. C., \& Sion, E. M. 1976,
616:  ApJ, 207, 820 
617: 
618: 
619: \bibitem[Williams(1992)]{1992AJ....104..725W} Williams, R.\
620: E.\ 1992, \aj,  104, 725
621: 
622: 
623: \end{thebibliography}
624: 
625: % Skip to the next page.  Place figures and figure captions after
626: % this point.  Uncomment pageskip when needed.
627: 
628: %\clearpage
629: 
630: % Place figure captions here on a separate page for journal submission
631: %\figcaption[epsfile]{Put figure caption here. \label{figure label 
632: %like Fig. 1}}
633: 
634: % Place figures and figure captions here if embedded in paper
635: 
636: \begin{figure}
637: \epsscale{0.9}
638: \plotone{f1.eps}
639: \caption{{\it Chandra} LETG+HRC-S X-ray light curves of V1494 Aql.
640: Top: Light curve obtained on 2000 September 28 (ObsID 02308).  Middle:
641: Light curve obtained on 2000 October 1 (ObsID 00072).  Bottom: Same as
642: the middle panel but showing the data scaled by the full range of the
643: count rate to highlight the burst.}
644: \end{figure}
645: 
646: \begin{figure}
647: \plotone{f2.eps}
648: \caption{Enlargement of the region around the time of the X-ray burst
649: on 2000 October 1 (ObsID 00072) sampled in 25~s bins.  Note the
650: presence of two large flares as well as precursor and trailer events.}
651: \end{figure}
652: 
653: \begin{figure}
654: \epsscale{0.9}
655: \plotone{f3.eps}
656: \caption{Temporal spectra of the X-ray light curves.  Top: Raw or
657: dirty power spectrum.  Middle: Spectral window function and an
658: enlargement of the window core (inset).  Bottom: Cleaned power
659: spectrum.  The significant peaks, the tracer signal with a frequency
660: of 0.0011~Hz, and a possible orbital (0.13467~d) X-ray modulation are
661: indicated in both the top and bottom panels.  A period at
662: $\simeq$2500~s is readily apparent in both dirty and cleaned power
663: spectra.}
664: \end{figure}
665: 
666: \end{document}
667: 
668: