1: %revised June 4, 2007 in response to referee's comments
2: %revised May 7 2007
3: %revised april 5 2007
4: \documentclass[12pt,preprint]{aastex}
5:
6: \shorttitle{\emph{XMM-Newton} Observations of the Cataclysmic Variable GW Lib}
7: \shortauthors{Hilton et al.}
8:
9: \bibliographystyle{apj}
10:
11: \begin{document}
12:
13: \title{\emph{XMM-Newton} Observations of the Cataclysmic Variable GW Lib}
14:
15: \author{Eric J. Hilton, Paula Szkody, Anjum Mukadam}
16: \affil{Astronomy Department, Box 351580, University of Washington, Seattle WA 98115}
17: \email{hilton@astro.washington.edu}
18:
19: \author{Koji Mukai}
20: \affil{NASA/Goddard Space Flight Center, Greenbelt, MD 20771}
21: \author{Coel Hellier and Liza van Zyl}
22: \affil{Astrophysics Group, Keele University, Keele, Staffordshire ST5 5BG, UK}
23: \author{Lee Homer}
24: \affil{Liverpool CC, Liverpool, UK}
25:
26: \begin{abstract}
27: \emph{XMM-Newton} observations of the accreting, pulsating white dwarf in
28: the quiescent dwarf nova GW Librae were conducted to determine if the non-radial
29: pulsations present in previous UV and optical data affect the X-ray emission.
30: The non-radial pulsations are evident in the simultaneous Optical Monitor data but
31: are not detected in X-ray with an upper limit on the pulsation amplitude of 0.092 mags.
32: The best fits to the X-ray spectrum are with a low temperature diffuse gas
33: model or a multi-temperature
34: cooling flow model, with a strong OVIII line, similar to other short period
35: dwarf novae, but with a
36: lower temperature range than evident in normal short period dwarf novae.
37: The lack of pulsations and the spectrum likely indicate
38: that the boundary layer
39: does not extend to the surface of the white dwarf.
40: \end{abstract}
41:
42: \keywords{stars: individual -- GW Lib -- stars: dwarf novae -- X-rays: stars}
43:
44: \section {Introduction}
45:
46: GW Librae was the first cataclysmic variable in which the white
47: dwarf was found to exhibit non-radial pulsations \citep{Warner1998}.
48: Since that first discovery,
49: ten other systems have been identified (see recent summary in \citet{Mukadam2007}), but GW Lib remains the brightest and best
50: studied of the accreting, pulsating white dwarfs. It has a very
51: short orbital
52: period of 77 min \citep{Szkody2000,Thorstensen2002}
53: and a very low accretion rate, resulting in a
54: large contribution of the white dwarf to the system light in the
55: optical and ultraviolet regions of the spectrum. Parallax and proper motion
56: measurements by \citet{Thorstensen2003} give a distance of 104 pc.
57: Until
58: recently, only one outburst of GW Lib was known \citep{Maza1983} but
59: in April 2007, a second outburst began. Monitoring of the superhumps
60: that developed during the course of the outburst \citep{Kato2007}
61: revealed a period excess of the superhump period over that of the orbital
62: period of 1.1\%. The relation of period excess to mass ratio and
63: orbital period \citep{Patterson2001} then implies that the secondary in GW
64: Lib has passed the period minimum for close binary evolution and is
65: degenerate.
66:
67:
68: Photometry over
69: several years has shown the characteristics of the pulsations of GW Lib
70: \citep{vanzyl04}. There are three primary pulsation periods, although
71: these all show different amplitudes at different times and some of
72: the periods are not always visible. The most common periods are near
73: 650, 370 and 230 s with typical amplitudes of 0.15, 0.010 and 0.007 mags.
74: \citet{woudt02} also identified a long period of 2.09 hrs that
75: was present in 2001 observations but not during 1997-1998. The origin of
76: this period is unknown but these long periods are present in several
77: short orbital period disk systems.
78:
79: Analysis of HST ultraviolet data \citep{Szkody2002} showed the same
80: pulsations were present in the UV as the optical, but the amplitudes
81: were about six times larger. An unexpected result from the HST study
82: was that the best model fit to the spectrum was with a two-temperature
83: white dwarf, with a T$_{eff}$=13,300K for 63\% of the white dwarf
84: surface and 17,100K for the remaining 37\%. It was unclear whether the
85: dual temperatures were a result of the boundary layer (where the
86: fast moving layers of the inner disk meet the slower rotation of the
87: white dwarf) providing
88: accretion heating of the equatorial regions of the white
89: dwarf, or due to the pulsations. Further UV studies of accreting
90: pulsating white dwarfs \citep{Szkody2007} have not shown this
91: dual temperature structure.
92:
93: While the soft X-ray emission from the stellar photosphere of the single
94: hot white dwarf PG1159-036 is known to exhibit similar pulsations modes
95: as the optical but with 20-30 times the optical amplitudes \citep{Barstow1986},
96: the interesting question is whether the non-radial
97: pulsations affect the boundary layer where the X-rays are produced
98: in CVs.
99: In order for theoretical disk instability models to account for the
100: long interoutburst timescales between dwarf novae outbursts such as
101: in GW Lib, the
102: accretion rate has to be very low, requiring very low viscosity and
103: truncation of the inner accretion disk, possibly by coronal siphons
104: or from a strong magnetic field on the white dwarf
105: \citep{Meyer1994,Warner1996}. Fitting
106: of spectral energy distributions to models also often invokes a
107: truncation of the inner disk in order to alleviate excess UV flux
108: from the models \citep{Linnell2007}. If the inner disk of GW Lib
109: is truncated, the
110: X-ray emission should not be pulsed. However, \emph{Chandra}
111: data on the dwarf nova U Gem \citep{Szkody2002a} shows that the
112: boundary layer is close to the white dwarf and moving at low velocity.
113: If the boundary layer in GW Lib
114: extends to the white dwarf surface, the X-ray emission may be
115: modulated at the same periods evident in the UV and optical.
116: Thus, the X-ray emission from
117: GW Lib could provide some constraints on the location and characteristics
118: of the boundary layer in low accretion rate systems.
119:
120: Since GW Lib was not detected in the ROSAT All Sky Survey, nor has
121: any previous X-ray observation, we obtained time on \emph{XMM-Newton} to
122: obtain light curves and spectra to determine if GW Lib has the
123: normal hard X-ray emission that is generally present in all low
124: mass transfer rate, disk-accreting dwarf novae, and if the X-rays
125: are modulated by the non-radial pulsations evident on its white
126: dwarf.
127:
128:
129: \section{Observations and Data Reduction}
130:
131: \emph{XMM-Newton} observations of GW Lib on August 25-26, 2005 provided simultaneous optical imaging
132: from the Optical Monitor \citep[OM;][]{mason01}, and X-ray data from the EPIC pn \citep{struder01},
133: and two MOS detectors \citep{turner01}.
134: The pn has roughly twice the effective area of either MOS detector.
135: Because of a low count rate, the Reflection Grating Spectrograph data were not useful.
136: The X-ray observations lasted approximately 20ks, while the OM consisted of 5 observations of
137: approximately 4ks each.
138: The UT times, length of total observations, and average count rates are listed in Table \ref{tbl:obs}.
139:
140: The data were reduced using SAS (ver. 7.0.0) following the guidelines from the main
141: \emph{XMM-Newton} Web site (Vilspa) and from the NASA/GSFC XMM-Newton Guest Observer Facility ABC Guide (ver. 2.01).
142: Calibration files are current to August 15, 2006.
143: The SAS tools were used to create new event list files from the observation data files.
144: In order to screen out background flaring events, whole-chip light curves for each detector
145: were created in the 10-18 keV range and the data were ignored when the count rate was greater than
146: 2.0 c/s for the pn and greater than 0.6 c/s for each MOS detector.
147: These background flaring times when the count rate limits were exceeded were nearly identical for all detectors.
148: The event list files were also screened with the standard canned expressions.
149: The source aperture was taken to be circular with a radius of 360 pixels for the pn and 320 pixels for
150: both MOS detectors in order to maximize the signal-to-noise.
151: For the MOS detectors, the source-free background aperture was taken to be an annulus on the central
152: chip centered on the source, while for the pn the background was taken to be rectangular regions on
153: adjacent chips with similar \emph{Y} locations as the target.
154: Energies were restricted to the well-calibrated ranges: 0.2 - 15 keV for spectral analysis and 0.1 - 12.5 keV for light curve analysis.
155: Events were restricted for the pn to singles (pattern = 0) for the spectrum and singles and doubles (pattern $\leq$ 4) for the lightcurve.
156: For the MOS detectors, up to quadruples (pattern $\leq$ 12) were permitted for both the spectrum and the lightcurve.
157: FTOOLS\footnote{http://heasarc.gsfc.nasa.gov/ftools/} \citep{Blackburn1995} software tasks were used to
158: group the spectral bins and associate various files for spectral analysis in XSPEC, create background
159: subtracted light curves, and correct the time stamps to the solar system barycenter.
160:
161: Data from both MOS detectors and the pn were combined to construct the X-ray
162: light curve, which had an average
163: count rate of 0.042 c/s.
164: Only data when all three detectors were live and free of background flaring
165: events were kept.
166: These are called good time intervals.
167: Data were also binned to increase the signal-to-noise of this faint source.
168: The time bin size was chosen to be 150 seconds to simultaneously optimize
169: signal-to-noise with time
170: resolution.
171: Although the time bins were primarily 150 seconds, the time bins at the
172: edges of the good time intervals
173: were of different sizes to accommodate all the data.
174: The time bin size is discussed further in section 3.1.2.
175:
176: For the OM observations, the B filter was used, and the Pipeline light curves were binned at 50 seconds for
177: the analysis.
178: The average count rate for the OM is 6.2 c/s, which is equivalent to a B magnitude of 17.3.
179: %****see http://xmm.vilspa.esa.es/external/xmm_user_support/documentation/sas_usg/USG/node117.html**
180:
181:
182: \section{Results}
183:
184: \subsection{Light Curves}
185:
186: \subsubsection{Optical}
187: The optical light curve of GW Lib, shown in Figure \ref{fig:light_curve}, is dominated by the 2.09 hour period
188: that was intermittently present in the data of \citet{woudt02}.
189: The discrete Fourier transform (DFT) of the optical data shown in Figure \ref{fig:dft}
190: shows this long period as well as modulations at 671 seconds with an amplitude of 0.02 mags and 397 seconds
191: with an amplitude of 0.021 mags.
192: These modulations are consistent with the previously observed pulsation periods near 650 s (1540 $\mu$Hz)
193: and 370 s (2700 $\mu$Hz), whose periods and amplitudes are known to vary \citep{vanzyl04}.
194: Van Zyl et al. also find a pulsation near 230 s (4350 $\mu$Hz) that is not seen in the OM data.
195: However, the typical amplitude of this period is below the average noise level of this DFT, so its
196: presence cannot
197: be ruled out.
198:
199: \subsubsection{X-ray}
200:
201: The DFT of the combined X-ray data showed no significant periodicities.
202: In order to place an upper-limit on the magnitude of variability, the
203: following light-curve shuffling
204: technique was applied to empirically determine the noise in the light
205: curve.
206: A light curve consists of a series of fractional intensity values each with
207: a corresponding time value.
208: Each value of fractional intensity was randomly reassigned to one of the
209: unchanged, existing time values.
210: This random shuffling destroys any coherent frequencies in the light curve
211: but maintains the same time
212: sampling and random white noise as the original light curve.
213: The DFT of the shuffled light curve gives the amplitude of the noise at each
214: frequency up to the Nyquist frequency.
215: The original light curve was randomly shuffled 10 times and the average
216: noise was computed each time.
217: The noise of the original light curve was taken to be the mean of these 10
218: values.
219:
220: As a check on the time bin size, light curves were produced with time bins
221: of primarily 50, 75, 100,
222: 150, and 200 seconds.
223: In all cases, there were no strong signals present in the light curves and
224: there were no significant
225: differences in the average noise values.
226: Because the count rate was so low, the time bin size was chosen to maximize
227: the signal-to-noise without
228: destroying the time resolution.
229: Since the shortest period seen in the simultaneous optical observations was
230: 397 seconds, the 150 second
231: time resolution provides more than two points per cycle, which is
232: sufficient time resolution.
233: The unshuffled DFT is shown in Figure \ref{fig:dft}.
234: The average noise averaged over ten random shufflings is 0.092 mags, which
235: is taken to be the upper
236: limit of the X-ray
237: pulsations for GW Lib.
238:
239: \subsection{Spectral Analysis}
240:
241: The extracted background-subtracted spectrum from the pn detector was
242: binned at 10 counts per bin to
243: facilitate the use of $\chi^2$ statistics to find the best fit models.
244: The spectrum was restricted to the energy range 0.2-15.0 keV because the
245: calibration of the EPIC detectors
246: at the lowest energies is not certain and the count rate above 15.0 keV is too low to be useful.
247: Although the data reduction allows high energy photons, there were very few photons detected with energies
248: greater than 3 keV.
249: The spectrum has a strong \ion{O}{8} emission line at $\sim$0.65 keV and
250: an increase in emission at
251: $\sim$ 1.0 keV that is possibly a Ne-Fe emission complex.
252: Several models were used, starting with the simplest emission mechanisms
253: (bremsstrahlung), and advancing in
254: complexity to more detailed models and variable abundances.
255: All models used absorption, but since all models
256: consistently found a low value for the hydrogen column density,
257: it was subsequently fixed at $10^{20}$ cm$^{-2}$ to reduce the number of
258: parameters.
259: The redshift was fixed at $10^{-9}$ for the \textbf{mekal} and \textbf{mkcflow}
260: families of models and the hydrogen density of the gas was fixed at
261: 0.1 cm$^{-3}$ for the \textbf{mekal} family of models.
262: Parameters of the model and the goodness of fit statistics are listed in Table \ref{tbl:fits}.
263:
264: The simple absorbed bremsstrahlung (\textbf{wabs(bremss)}) model had a reduced $\chi^2 = 1.05$, but was
265: unable to fit the strong emission lines.
266: Explicitly adding a Gaussian to model the oxygen line decreased the
267: residuals, and had a reduced $\chi^2 = 0.73$, but was
268: unable to fit the lines near 1 keV.
269:
270: The model of hot diffuse gas with line emissions from several elements \textbf{wabs(mekal)} with a solar
271: abundance mixture also could not fit the emission lines (reduced $\chi^2 = 0.95$).
272: The variable abundance version of this model (\textbf{wabs(vmekal)})
273: gave a better fit to the both the $\sim$0.65 keV and the
274: $\sim$1.0 keV lines.
275: All combinations of
276: varying the oxygen, neon, and iron abundances were tried.
277: As there were no significant differences in the model fits with different
278: iron and neon abundances, these were finally left fixed at solar abundance.
279: The model with oxygen as a parameter of the fit is shown in
280: Figure \ref{fig:vmekal}, and has a reduced $\chi^2 =0.81$.
281:
282: \citet{Mukai2003} and \citet{Pandel2003} found successful fits using a cooling flow model
283: (\textbf{wabs(mkcflow)}), so this model was also tried, although it
284: did not fit the oxygen line nor fully fit the lines at $\sim$1.0 keV.
285: Adjusting the oxygen and neon abundances using (\textbf{wabs(vmcflow)}) did give a better fit to the
286: emission lines with significantly higher oxygen abundance (compared to solar) and a slightly increased
287: neon abundance.
288: Since the mekal models showed that the neon abundance was very uncertain, the cooling flow model was also
289: tried with leaving the neon fixed at solar abundance and allowing only the oxygen to be fit.
290: This model is shown in Figure \ref{fig:vmcflow}.
291:
292: There are still residuals in both the vmekal and vmcflow model fits (Figures \ref{fig:vmekal} and \ref{fig:vmcflow}) near 0.9 keV.
293: A Gaussian was added to the vmekal model at that energy but there was no significant
294: improvement in the fits.
295: Regardless of the model that was fit to the data, the temperature is generally low (1.5 - 2.5 keV)
296: compared to most dwarf novae \citep{Ramsay2001,Pandel2003,Hakala2004}.
297:
298:
299: \section{Discussion}
300:
301: The X-ray flux of GW Lib is much lower than expected for its optical magnitude and physical parameters.
302: The cataclysmic variable WZ Sge has an orbital period and long-term outburst
303: characteristics similar to GW Lib. The absolute visual magnitudes of the two systems
304: are comparable
305: (11.8 for WZ Sge and 11.9 for GW Lib) and the white dwarfs have comparable temperatures.
306: Using the 4.5 keV thermal bremsstrahlung model of WZ Sge \citep{Patterson1998} as a comparison, and correcting for
307: distance, PIMMS predicts a count rate of about 0.1-0.2 c/s for GW Lib with
308: the EPIC pn
309: detector.
310: The actual average count rate was much lower: 0.02 c/s for the pn and only 0.04 c/s after
311: combining all three X-ray detectors (see Figure \ref{fig:light_curve}).
312: \emph{XMM-Newton} observations of other relatively nearby short period
313: dwarf novae (T Leo, OY Car, VW Hyi, WX Hyi, SU UMa, TY PsA and YZ Cnc with
314: orbital periods between 85-125 min and more frequent outbursts than GW Lib and
315: WZ Sge), the pn count
316: rates were between 1-7 c/s \citep{Ramsay2001,Pandel2003,Hakala2004,Pandel2005}
317: The 0.2-10 keV fluxes for the best fit vmekal (hot diffuse gas) and vmcflow
318: (cooling flow) models for GW Lib shown in Figures \ref{fig:vmekal}
319: and \ref{fig:vmcflow} are 6.82 $\times 10^{-14}$ and 6.90 $\times 10^{-14}$ ergs cm$^{-2}$ s$^{-1}$
320: respectively. For a distance of 104 pc \citep{Thorstensen2003}, the X-ray luminosity would be
321: 9$\times 10^{28}$ ergs s$^{-1}$. This compares to L$_{x}$ of 4$\times 10^{30}$,
322: 8$\times 10^{30}$ and
323: 1.4$\times 10^{32}$ ergs s$^{-1}$ for OY Car, VW Hyi and YZ Cnc. Assuming
324: this is the boundary layer luminosity, and using the relation given in
325: \citet{Pandel2003}:
326:
327: L$_{bl}$ = 5/2 kT$_{max}$/$\mu$ m$_{p}$
328:
329: where T$_{max}$ is the maximum temperature in the cooling flow model
330: (5 keV), $\mu$ =0.6 and m$_{p}$ is the proton mass, we can estimate
331: that $\dot{M}_{bl}$ = 7$\times 10^{-14}$ M$_{\odot}$ yr$^{-1}$. This value
332: is typically 2 orders of magnitude lower than that for the other dwarf
333: novae \citep{Pandel2005}. This value is also much lower than the time-averaged
334: $\dot{M}$ of 7.3$\times 10^{-11}$ estimated by \citep{Townsley2004} from
335: their model parameters for GW Lib.
336:
337: All model fits to the spectrum of GW Lib resulted in
338: lower temperatures compared to OY Car, VW Hyi and YZ Cnc and the other systems
339: as well. Although the best fit
340: to all systems involve a range of temperatures, the
341: maximum temperature for GW Lib is around 5 keV while the kT$_{max}$ for
342: the short period objects in Pandel et al. ranges from 8-26 keV. The low temperature
343: is likely not due to an exceptionally low mass for the primary in GW Lib, as
344: the UV fits \citep{Szkody2002} and the pulsation models \citep{Townsley2004} indicate
345: a high mass white dwarf.
346: The low temperature in GW Lib suggests that the accreting gas is low density or is
347: only mildly shocked, so the X-ray cooling is very inefficient. It is likely that the
348: shock occurs high above the white dwarf surface which lowers the shock temperature.
349: The stronger oxygen line in GW Lib compared to these other systems and the
350: lack of FeK$\alpha$ at 6.4 keV are likely artifacts of the low temperature (although
351: we cannot rule out that there is some peculiar atomic physics that is not taken
352: into account in the mekal-type models).
353: It is noteworthy that FeK is also missing in WZ Sge and its temperature is
354: similar to GW Lib \citep{Patterson1998} so the lower accretion rates in
355: these systems with rare but tremendous amplitude outbursts \citep{Howell1995} likely lead to
356: similar weak boundary layers.
357:
358: The pulsations that are visible in the optical and UV are limited to
359: an X-ray amplitude of $<$ 0.09 mag. The low X-ray flux, cool temperatures
360: and absence of
361: strong X-ray pulsations all imply that the boundary layer in GW Lib
362: does not reach to
363: the white dwarf surface to create a strong shock or to be affected by
364: the surface pulsations.
365:
366: The lack of X-ray modulation at the 2.09 hr period argues against an
367: origin for this period in the inner disk of a magnetic, precessing
368: white dwarf as has been suggested for the long periods seen in FS Aur
369: and HS2331+39 \citep{Tovmassian2007}.
370:
371: \section{Conclusions}
372:
373: The \emph{XMM-Newton} observations of GW Lib have shown that the X-ray
374: emitting region of the accreting, pulsating white dwarf is not
375: strongly affected by the non-radial pulsations evident in the UV and optical.
376: The unusually weak
377: X-ray flux from this system precludes a stringent limit, but does rule
378: out pulsation amplitudes of greater than 0.09 mag, specifically at the periods where
379: significant signals are detected simultaneously in the optical band.
380: The low X-ray flux, cool maximum temperature of the X-ray spectrum,
381: combined with the lack of X-ray pulsation indicate the boundary layer
382: is not dense enough to create a strong shock at the white dwarf surface.
383: This has implications for the two-temperature model for the white
384: dwarf that was needed to explain the HST UV spectrum \citep{Szkody2002}
385: in that the origin of the hotter temperature component
386: may be related to the pulsations, and
387: not to boundary layer heating.
388:
389: \acknowledgements
390:
391: This work was supported by \emph{XMM-Newton} grant NNG05GR47G to the University
392: of Washington and is based on observations obtained with \emph{XMM-Newton},
393: an ESA science mission with instruments and contributions directly funded
394: by ESA Member States and the USA (NASA).
395:
396: %\bibliography{hilton_bib}
397: \begin{thebibliography}{28}
398: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
399:
400: \bibitem[{{Barstow} {et~al.}(1986){Barstow}, {Holberg}, {Grauer}, \&
401: {Winget}}]{Barstow1986}
402: {Barstow}, M.~A., {Holberg}, J.~B., {Grauer}, A.~D., \& {Winget}, D.~E. 1986,
403: \apjl, 306, L25
404:
405: \bibitem[{{Blackburn}(1995)}]{Blackburn1995}
406: {Blackburn}, J.~K. 1995, in ASP Conf. Ser. 77: Astronomical Data Analysis
407: Software and Systems IV, ed. R.~A. {Shaw}, H.~E. {Payne}, \& J.~J.~E.
408: {Hayes} (San Francisco: ASP), 367
409:
410: \bibitem[{{Hakala} {et~al.}(2004){Hakala}, {Ramsay}, {Wheatley}, {Harlaftis},
411: \& {Papadimitriou}}]{Hakala2004}
412: {Hakala}, P., {Ramsay}, G., {Wheatley}, P., {Harlaftis}, E.~T., \&
413: {Papadimitriou}, C. 2004, \aap, 420, 273
414:
415: \bibitem[{{Howell} {et~al.}(1995){Howell}, {Szkody}, \&
416: {Cannizzo}}]{Howell1995}
417: {Howell}, S.~B., {Szkody}, P., \& {Cannizzo}, J.~K. 1995, \apj, 439, 337
418:
419: \bibitem[{Kato(2007)}]{Kato2007}
420: Kato, T. 2007, vsnet-alert 9326
421:
422: \bibitem[{{Linnell} {et~al.}(2007){Linnell}, {Hoard}, {Szkody}, {Long},
423: {Hubeny}, {G{\"a}nsicke}, \& {Sion}}]{Linnell2007}
424: {Linnell}, A.~P., {Hoard}, D.~W., {Szkody}, P., {Long}, K.~S., {Hubeny}, I.,
425: {G{\"a}nsicke}, B., \& {Sion}, E.~M. 2007, \apj, 654, 1036
426:
427: \bibitem[{{Mason} {et~al.}(2001){Mason}, {Breeveld}, {Much}, {Carter},
428: {Cordova}, {Cropper}, {Fordham}, {Huckle}, {Ho}, {Kawakami}, {Kennea},
429: {Kennedy}, {Mittaz}, {Pandel}, {Priedhorsky}, {Sasseen}, {Shirey}, {Smith},
430: \& {Vreux}}]{mason01}
431: {Mason}, K.~O. {et~al.} 2001, \aap, 365, L36
432:
433: \bibitem[{{Maza} \& {Gonzalez}(1983)}]{Maza1983}
434: {Maza}, J., \& {Gonzalez}, L.~E. 1983, \iaucirc, 3854, 2
435:
436: \bibitem[{{Meyer} \& {Meyer-Hofmeister}(1994)}]{Meyer1994}
437: {Meyer}, F., \& {Meyer-Hofmeister}, E. 1994, \aap, 288, 175
438:
439: \bibitem[{Mukadam} {et~al.}(2007)]{Mukadam2007}
440: {Mukadam}, A. {et~al.} 2007, \apj, Submitted
441:
442: \bibitem[{{Mukai} {et~al.}(2003){Mukai}, {Kinkhabwala}, {Peterson}, {Kahn}, \&
443: {Paerels}}]{Mukai2003}
444: {Mukai}, K., {Kinkhabwala}, A., {Peterson}, J.~R., {Kahn}, S.~M., \& {Paerels},
445: F. 2003, \apjl, 586, L77
446:
447: \bibitem[{{Pandel} {et~al.}(2003){Pandel}, {C{\'o}rdova}, \&
448: {Howell}}]{Pandel2003}
449: {Pandel}, D., {C{\'o}rdova}, F.~A., \& {Howell}, S.~B. 2003, \mnras, 346, 1231
450:
451: \bibitem[{{Pandel} {et~al.}(2005){Pandel}, {C{\'o}rdova}, {Mason}, \&
452: {Priedhorsky}}]{Pandel2005}
453: {Pandel}, D., {C{\'o}rdova}, F.~A., {Mason}, K.~O., \& {Priedhorsky}, W.~C.
454: 2005, \apj, 626, 396
455:
456: \bibitem[{{Patterson} {et~al.}(1998){Patterson}, {Richman}, {Kemp}, \&
457: {Mukai}}]{Patterson1998}
458: {Patterson}, J., {Richman}, H., {Kemp}, J., \& {Mukai}, K. 1998, \pasp, 110,
459: 403
460:
461: \bibitem[{{Patterson}(2001)}]{Patterson2001}
462: {Patterson}, J. 2001, \pasp, 113, 736
463:
464: \bibitem[{{Ramsay} {et~al.}(2001){Ramsay}, {C{\'o}rdova}, {Cottam}, {Mason},
465: {Much}, {Osborne}, {Pandel}, {Poole}, \& {Wheatley}}]{Ramsay2001}
466: {Ramsay}, G. {et~al.} 2001, \aap, 365, L294
467:
468: \bibitem[{{Str{\"u}der} {et~al.}(2001){Str{\"u}der}, {Briel}, {Dennerl},
469: {Hartmann}, {Kendziorra}, {Meidinger}, {Pfeffermann}, {Reppin}, {Aschenbach},
470: {Bornemann}, {Br{\"a}uninger}, {Burkert}, {Elender}, {Freyberg}, {Haberl},
471: {Hartner}, {Heuschmann}, {Hippmann}, {Kastelic}, {Kemmer}, {Kettenring},
472: {Kink}, {Krause}, {M{\"u}ller}, {Oppitz}, {Pietsch}, {Popp}, {Predehl},
473: {Read}, {Stephan}, {St{\"o}tter}, {Tr{\"u}mper}, {Holl}, {Kemmer}, {Soltau},
474: {St{\"o}tter}, {Weber}, {Weichert}, {von Zanthier}, {Carathanassis}, {Lutz},
475: {Richter}, {Solc}, {B{\"o}ttcher}, {Kuster}, {Staubert}, {Abbey}, {Holland},
476: {Turner}, {Balasini}, {Bignami}, {La Palombara}, {Villa}, {Buttler},
477: {Gianini}, {Lain{\'e}}, {Lumb}, \& {Dhez}}]{struder01}
478: {Str{\"u}der}, L. {et~al.} 2001, \aap, 365, L18
479:
480: \bibitem[{{Szkody} {et~al.}(2000){Szkody}, {Desai}, \& {Hoard}}]{Szkody2000}
481: {Szkody}, P., {Desai}, V., \& {Hoard}, D.~W. 2000, \aj, 119, 365
482:
483: \bibitem[{{Szkody} {et~al.}(2002{\natexlab{a}}){Szkody}, {G{\"a}nsicke},
484: {Howell}, \& {Sion}}]{Szkody2002}
485: {Szkody}, P., {G{\"a}nsicke}, B.~T., {Howell}, S.~B., \& {Sion}, E.~M.
486: 2002{\natexlab{a}}, \apjl, 575, L79
487:
488: \bibitem[{{Szkody} {et~al.}(2002{\natexlab{b}}){Szkody}, {Nishikida},
489: {Raymond}, {Seth}, {Hoard}, {Long}, \& {Sion}}]{Szkody2002a}
490: {Szkody}, P., {Nishikida}, K., {Raymond}, J.~C., {Seth}, A., {Hoard}, D.~W.,
491: {Long}, K.~S., \& {Sion}, E.~M. 2002{\natexlab{b}}, \apj, 574, 942
492:
493: \bibitem[{{Szkody} {et~al.}(2007){Szkody}, {Mukadam}, {G{\"a}nsicke}, {Woudt},
494: {Solheim}, {Nitta}, {Sion}, {Warner}, {Sahu}, {Prabhu}, \&
495: {Henden}}]{Szkody2007}
496: {Szkody}, P. {et~al.} 2007, \apj, 658, 1188
497:
498: \bibitem[{{Thorstensen} {et~al.}(2002){Thorstensen}, {Patterson}, {Kemp}, \&
499: {Vennes}}]{Thorstensen2002}
500: {Thorstensen}, J.~R., {Patterson}, J., {Kemp}, J., \& {Vennes}, S. 2002, \pasp,
501: 114, 1108
502:
503: \bibitem[{{Thorstensen}(2003)}]{Thorstensen2003}
504: {Thorstensen}, J.~R. 2003, \aj, 126, 3017
505:
506: \bibitem[{{Tovmassian} {et~al.}(2007){Tovmassian}, {Zharikov}, \& {Neustroev}}]{Tovmassian2007}
507: {Tovmassian}, G.~H., {Zharikov}, S.~V. \& {Neustroev}, V.~V. 2007, \apj, 655, 466
508:
509: \bibitem[{{Townsley} {et~al.}(2004){Townsley}, {Arras}, \&
510: {Bildsten}}]{Townsley2004}
511: {Townsley}, D.~M., {Arras}, P., \& {Bildsten}, L. 2004, \apjl, 608, L105
512:
513: \bibitem[{{Turner} {et~al.}(2001){Turner}, {Abbey}, {Arnaud}, {Balasini},
514: {Barbera}, {Belsole}, {Bennie}, {Bernard}, {Bignami}, {Boer}, {Briel},
515: {Butler}, {Cara}, {Chabaud}, {Cole}, {Collura}, {Conte}, {Cros}, {Denby},
516: {Dhez}, {Di Coco}, {Dowson}, {Ferrando}, {Ghizzardi}, {Gianotti}, {Goodall},
517: {Gretton}, {Griffiths}, {Hainaut}, {Hochedez}, {Holland}, {Jourdain},
518: {Kendziorra}, {Lagostina}, {Laine}, {La Palombara}, {Lortholary}, {Lumb},
519: {Marty}, {Molendi}, {Pigot}, {Poindron}, {Pounds}, {Reeves}, {Reppin},
520: {Rothenflug}, {Salvetat}, {Sauvageot}, {Schmitt}, {Sembay}, {Short},
521: {Spragg}, {Stephen}, {Str{\"u}der}, {Tiengo}, {Trifoglio}, {Tr{\"u}mper},
522: {Vercellone}, {Vigroux}, {Villa}, {Ward}, {Whitehead}, \& {Zonca}}]{turner01}
523: {Turner}, M.~J.~L. {et~al.} 2001, \aap, 365, L27
524:
525: \bibitem[{{van Zyl} {et~al.}(2004){van Zyl}, {Warner}, {O'Donoghue}, {Hellier},
526: {Woudt}, {Sullivan}, {Pritchard}, {Kemp}, {Patterson}, {Welsh}, {Casares},
527: {Shahbaz}, {van der Hooft}, \& {Vennes}}]{vanzyl04}
528: {van Zyl}, L. {et~al.} 2004, \mnras, 350, 307
529:
530: \bibitem[{{Warner} {et~al.}(1996){Warner}, {Livio}, \& {Tout}}]{Warner1996}
531: {Warner}, B., {Livio}, M., \& {Tout}, C.~A. 1996, \mnras, 282, 735
532:
533: \bibitem[{{Warner} \& {van Zyl}(1998)}]{Warner1998}
534: {Warner}, B., \& {van Zyl}, L. 1998, in IAU Symp. 185: New Eyes to See Inside
535: the Sun and Stars, ed. F.-L. {Deubner}, J.~{Christensen-Dalsgaard}, \&
536: D.~{Kurtz} (Dordrecht: Kluwer), 321
537:
538: \bibitem[{{Woudt} \& {Warner}(2002)}]{woudt02}
539: {Woudt}, P.~A., \& {Warner}, B. 2002, \apss, 282, 433
540:
541: \end{thebibliography}
542:
543:
544:
545: \clearpage
546: \begin{figure}%lightcurves
547: \epsscale{.80}
548: \plotone{f1.eps}
549: \caption{The optical lightcurve (top, shown here binned at 200 sec for clarity) shows a 2.09 hour period.
550: The X-ray lightcurve (bottom) shows no apparent period.
551: Notice the low count rate of the X-ray data.
552: \label{fig:light_curve} }
553: \end{figure}
554:
555: \clearpage
556: \begin{figure}%dft
557: \epsscale{.80}
558: \plotone{f2.eps}
559: \caption{The DFT of the optical lightcurve binned at 50 sec (top) shows the 2.09 hour period as well as
560: the two labelled periods from \citep{vanzyl04}.
561: The DFT of X-ray light curve binned at 150 sec (bottom) shows no periods.
562: Note that with the longer time bins for the X-ray data, the Nyquist frequency is lower, and the DFT
563: does not extend to as high frequencies as the OM data.
564: See text for a discussion of the limit of variability. \label{fig:dft} }
565: \end{figure}
566:
567: \clearpage
568: \begin{figure}%vmekal spectral fit
569: \epsscale{.80}
570: \includegraphics[angle=270,scale=.60]{f3.eps}
571: %\plotone{f3.eps}
572: \caption{The hot diffuse gas model with variable oxygen abundance.
573: \label{fig:vmekal} }
574: \end{figure}
575:
576: \clearpage
577: \begin{figure}%vmcflow spectral fit
578: \epsscale{.80}
579: \includegraphics[angle=270,scale=.60]{f4.eps}
580: %\plotone{f4.eps}
581: \caption{The cooling flow model with variable oxygen abundance.
582: \label{fig:vmcflow} }
583: \end{figure}
584:
585:
586: %make tables
587: \clearpage
588: \begin{deluxetable}{llllll} % observations
589: \tabletypesize{\scriptsize}
590: \tablewidth{0pt}
591: \tablecaption{25-26 August, 2005 Observations.\label{tbl:obs}}
592: \tablehead{
593: \colhead{Instrument} & \colhead{Filter} & \colhead{Duration (s)} &\colhead{UT Start Time} &\colhead{UT Stop Time} &\colhead{Ave. count rate (C/s)\tablenotemark{a}}}
594: \startdata
595: PN & Thin1 & 19936 & 22:09:28 & 03:41:44 & (2.32$\pm$ 0.15)$\times10^{-2}$ \\
596: MOS1 & Thin1 & 21809 & 21:47:09 & 03:50:38 & (7.06$\pm$ 0.72)$\times10^{-3}$ \\
597: MOS2 & Thin1 & 21577 & 21:47:09 & 03:46:46 & (8.43$\pm$ 0.78)$\times10^{-3}$ \\
598: OM & B & 19901 & 21:55:31 & 03:48:49 & 6.2 $\pm$ 0.6 (B = 17.3)\tablenotemark{b} \\
599: \enddata
600: \tablenotetext{a}{X-ray count rates determined from spectral reductions}
601: \tablenotetext{b}{OM count rate determined from light curve and converted to standard B magnitude}
602: \end{deluxetable}
603:
604: \clearpage
605: \begin{deluxetable}{lllll} %spectral fits
606: \tabletypesize{\scriptsize}
607: \tablewidth{0pt}
608: \tablecaption{XSPEC models used to fit the X-ray spectrum.\label{tbl:fits}}
609: \tablehead{
610: \colhead{Model Name} & \colhead{Reduced $\chi^2$} & \colhead{kT} &\colhead{Normalization} &\colhead{Parameters}}
611: \startdata
612:
613: Bremss\tablenotemark{a} & 1.05 & 2.2 & 2.3$\times 10^{-5}$ & \\
614:
615:
616: Bremss\tablenotemark{b} & 0.73 & 2.11 & 2.1 $\times 10^{-5}$& LineE = 0.67 keV \\
617: + Gauss &&&&$\sigma$ = 9.9$\times 10^{-7}$ keV\\
618: &&&& norm = 4.1$\times 10^{-6}$\\
619:
620: mekal\tablenotemark{c} & 0.95 & 1.90 & 6.0$\times 10^{-5}$ & abundance = 0.19 \\
621:
622: vmekal\tablenotemark{d} & 0.81 & 2.50 & 2.9$\times10^{-5}$ & O abund. = 7.99$\times$solar \\
623:
624: mkcflow\tablenotemark{e} & 0.83 & 0.091 - 4.70 & 1.6$\times 10^{-16}$ & Abundance = 0.28 \\
625:
626: vmcflow\tablenotemark{f} & 0.70 & 0.38 - 5.52 & 1.4$\times 10^{-16}$ & O abund. = 6.3$\times$solar \\
627: &&&& Ne abund. = 1.4$\times$solar \\
628:
629: vmcflow\tablenotemark{g} & 0.67 & 0.38 - 5.45 & 1.5$\times10^{-16}$ & O abund. = 6.12$\times$solar \\
630:
631: \enddata
632: \tablenotetext{a}{Thermal Bremsstrahlung - Didn't fit the emission lines }
633: \tablenotetext{b}{Thermal Bremsstrahlung plus Gaussian - Fit the oxygen line well}
634: \tablenotetext{c}{Emission from a hot diffuse gas - model shows bump at 0.6 keV, but doesn't fit line}
635: \tablenotetext{d}{Emission from a hot diffuse gas with variable abundances, neon fixed at solar}
636: \tablenotetext{e}{Cooling flow}
637: \tablenotetext{f}{Cooling flow with variable abundances}
638: \tablenotetext{g}{Cooling flow with variable abundances, neon fixed at solar}
639:
640: \end{deluxetable}
641:
642:
643: \end{document}
644:
645:
646: