1: %\documentclass[preprint2,11pt]{aastex}
2: %\documentclass[manuscript,11pt]{aastex}
3: \documentclass[manuscript,12pt]{aastex}
4: %\documentclass[preprint2,12pt]{aastex}
5: %\usepackage{natbib}
6: \def\simless{\mathbin{\lower 3pt\hbox
7: {$\rlap{\raise 5pt\hbox{$\char'074$}}\mathchar"7218$}}} %< or of order
8: \def\simmore{\mathbin{\lower 3pt\hbox
9: {$\rlap{\raise 5pt\hbox{$\char'076$}}\mathchar"7218$}}} %> or of order
10: \def\msun{{\rm M}_\odot} % solar masses
11: \def\4u{4U 1728--34}
12:
13: \received{}
14: \accepted{}
15: \journalid{337}{15 January 1989}
16: %\articleid{11}{14}
17: \slugcomment{Submitted to ApJ, March 21, 2007}
18:
19: \lefthead{et al.}
20: \righthead{\4u}
21:
22: \begin{document}
23:
24: \title{Fourier resolved spectroscopy of \4u: New Insights
25: into Spectral and Temporal Properties of Low-Mass X-ray Binaries}
26:
27: \author{ }
28: %
29: %
30: \author{C.R. Shrader$^{1,4}$, P. Reig$^{2,3}$ and D. Kazanas$^1$}
31:
32: \vskip 30pt
33: \affil{$^1$ Astrophysics Science Division, NASA Goddard
34: Space Flight Center, Greenbelt, MD 20771}
35: %
36: \affil{$^2$IESL, Foundation for Research and Technology, 711 10
37: Heraklion, Crete, Greece}
38: %
39: \affil{$^3$Physics Department, University of Crete, PO Box 2208, 710
40: 03 Heraklion, Crete, Greece}
41:
42: \affil{$^4$ University Space Reseach Association, 10211 Wincopin Circle,
43: Suite 620, Columbia, MD 21044}
44:
45:
46: \email{Chris.R.Shrader@gsfc.nasa.gov, pau@physics.uoc.gr,
47: Demos.Kazanas-1@nasa.gov}
48: %
49: %
50:
51: \begin{abstract}
52:
53: \baselineskip=15pt
54:
55: Using archival RXTE data we derive the 2-16 keV Fourier-resolved spectra of
56: the Atoll source \4u in a sequence of its timing states as its low QPO
57: frequency spans the range between 6 and 94 Hz. The increase in the QPO
58: frequency accompanies a spectral transition of the source from its island
59: to its banana states. The banana-states' Fourier-resolved spectra are
60: well fitted by a single blackbody component with $kT \sim 2-3$ keV depending on
61: the source position in the color -- color diagram and the Fourier frequency,
62: thus indicating that this spectral component is responsible for the source
63: variability on these timescales. This result is in
64: approximate agreement with similar behavior exhibited by the Z sources,
65: suggesting that, as in that case, the bounday layer -- the likely source
66: of the thermal component -- is supported by radiation pressure.
67: Furthermore, it is found that the iron line at $\sim$6.6 keV, clearly
68: present in the averaged spectra, not apparent within the limitations of our
69: measurements in the
70: frequency-resolved spectra irrespective of the frequency range. This would
71: indicate that this spectral component exhibits
72: little variability on time scales
73: comprising the interval $10^{-2}-10^2$ seconds.
74: In the island state the single blackbody model proved inadequate,
75: particularly notable in our lowest frequency band ($0.008-0.8$ Hz).
76: An absorbed powerlaw or an additive
77: blackbody plus hard powerlaw model was required
78: to obtain a satisfactory fit. Statisics do not allow
79: unambiguous discrimination between these possible scenarios.
80:
81: %
82:
83:
84: \end{abstract}
85:
86:
87: \keywords{accretion, accretion disks --- neutron stars --- stars:
88: individual (\4u) --- X-rays: stars}
89:
90: \baselineskip=15pt
91:
92: \section{Introduction}
93:
94:
95: Rapid X-ray variability and spectral distributions are powerful
96: probes of the physics of accretion flows onto compact objects, i.e.
97: onto neutron stars and black holes. More than two decades of
98: persistent efforts to probe and comprehend the physics involved in
99: these flows led to the accumulation of a wealth of data of both
100: their spectra and time variations. Significant progress in both
101: these directions has been made over the last decade, in particular,
102: with the wealth of data provided by the Rossi X-Ray Timing
103: Explorer (RXTE). Its large collecting area and high time resolution
104: have facilitated detailed studies on the rapid variability and the
105: spectra of a large sample of sources. However, the accompanying
106: improvement in the understanding of the dynamics and radiation
107: emission physics associated with these accretion powered sources has
108: progressed considerably slower than the accumulation of the data and {
109: their ensuing classification within semi-coherent phenomenological models.}
110:
111: Concerning the differentiation between the spectral and timing
112: characteristics of accreting neutron stars and black holes, the
113: former appeared to present higher complexity in
114: both their spectral and timing properties than the latter. This
115: additional complexity is generally attributed to the presence of the
116: neutron star surface which can complicate the dynamics of accretion
117: and as a result the ensuing spectral and timing properties. In the
118: temporal domain, the power density spectra (PDS) of neutron star
119: low-mass X-ray binaries (LMXBs) exhibit a variety of features
120: ranging from narrow quasi-periodic oscillations (QPOs) to broad
121: noise components (for a review, see Wijnands (2001); also van der
122: Klis (2006) and references therein), only some of which appear to
123: have corresponding features in the black hole case.
124: %
125: In the spectral domain, while Comptonization is presumably the predominant
126: process for producing the observed X-ray emission above $\sim 1$ keV, the
127: neutron-star LMXB spectra are quite complex, more so than those of
128: accreting black hole candidates. This is likely due to the fact that a
129: number of components contribute to the emission in this energy range and
130: that their individual spectra are hard to disentangle.
131:
132: A simplified way to classify the (apparently complex) spectra of
133: these sources has been that of the color-color (or color -
134: intensity) diagrams (Hasinger \& van der Klis 1989); these exhibit,
135: instead of a detailed spectral decomposition, the ratios of fluxes
136: in adjacent bands (colors) within the detector's range (color -
137: color diagrams) or one color as a function of the total flux of the
138: source (color - intensity diagrams). The large body of accumulated
139: data then made it clear that in response to changes in luminosity,
140: these sources cover a certain trajectory in this color - color (or
141: color - intensity) space whose shape has since been used to classify
142: them: The most luminous LMXBs, $L_x \sim 10^{38}$ erg s$^{-1}$,
143: cover a Z-shape track in color - color space, hence the name Z-sources.
144: The less luminous sources, $L_x \sim 10^{37}$ erg s$^{-1}$, tend to
145: cover a circular-like path in same space, leading to their
146: designation as atoll sources. At the same time, it was found that
147: there exists a correspondence between their spectral and timing
148: properties, with the frequencies of timing features, such as QPOs or
149: breaks, generally increasing in response to increases in the
150: X-ray flux and presumably the accretion rate.
151:
152:
153: The physics governing the spectral and timing properties of Z or
154: atoll sources and their differences remains obscure, although as
155: early as 1984, Mitsuda et al. (1984) noticed the possible
156: decomposition of their spectra into hard and soft components with
157: the hard one being the more variable of the two. A great advance in
158: the understanding of these spectra was made recently with a
159: refinement of the arguments put forward by Mistuda et al. (1984).
160: This involves the implementation of a novel technique that combines the
161: spectral and timing properties of these sources in an effort to
162: produce a coherent picture of the dynamics of these sources.
163:
164: This technique is known as Fourier Resolved Spectroscopy (FRS) (e.g.
165: Revnivtsev et al. 1999; Zycki 2003; Papadakis et al. 2005, 2006). In short,
166: instead of producing the power spectra as a function of photon
167: energy, as it is customary, this technique accumulates the
168: variability amplitudes over several well defined frequency bands for
169: each energy bin to produce the energy spectra of a source for the
170: specific frequency bands considered. This particular method of
171: presenting the data facilitates the identification of the variable
172: spectral components as well as the time scales over which the
173: variability occurs; as such, it allows certain immediate insights
174: into the dynamics responsible for the emission of the specific
175: spectral components. It was pointed out that this method is
176: particularly suited for the study of spectral features that result
177: from X-ray reprocessing where the light crossing time provides a
178: natural frequency filter; in this case FRS can provide
179: straightforward insights about the geometry of the reprocessing
180: medium. So far, this technique has been applied successfully to
181: galactic black hole candidates (Revnivtsev et al. 1999; Gilfanov et
182: al. 2001; Sobolewska \& Zycki 2006; Reig et al. 2006), to
183: neutron-star LMXBs (Gilfanov \& Revnivtsev et al 2005) and also to
184: AGN (Papadakis et al. 2005, 2006) yielding insights into
185: variability of different spectral components and infer their spatial
186: arrangements.
187:
188: Recently, the same technique has been applied to the spectro -
189: temporal data of accreting neutron stars and more specifically
190: to those of Z-sources (Gilfanov et al. 2003; Revnivtsev \& Gilfanov
191: 2006). This analysis did confirm the earlier tentative decomposition
192: of the Z-source spectra into a variable hard component and a soft
193: less variable one. The normalization of these components varies with
194: the accretion rate as do their spectra to produce the well known
195: color-color diagram of the Z-sources. In particular, the softer of
196: the two components was identified with the accretion disk and the
197: harder one with the boundary layer of these sources.
198: The temperature of the former was found
199: to change slowly in response changes in luminosity, as
200: expected. The temperature of the boundary layer component remained
201: constant, independent of luminosity and of Fourier frequency. This
202: supports the conjecture that the boundary layer is dominated by
203: radiation pressure. It was also found (Revnivtsev \& Gilfanov 2006)
204: that the disk component contributes to the spectra only at the
205: lowest Fourier frequencies. This is consistent with the findings of
206: Reig et al. (2006) pertaining to the lack of variability of the disk
207: component around a transient black hole LMXB 4U 1543-47. This
208: separation and the identification of these two components in the
209: spectra was possible only because of the combined spectral - timing
210: analysis.
211:
212: In the present paper we extend the application of the method of Fourier
213: Resolved Spectroscopy to the spectra of a source of the atoll class (i.e.
214: neutron stars accreting at rates $\sim 10$ times lower than those of
215: Z-sources). The trajectories of these sources in the color-color diagram
216: have a different shape from those of the Z-sources as discussed above. They are
217: known to produce a sequence of states of increasing luminosity known
218: as island, lower banana and upper banana states. Variations in the spectral
219: properties between these states are accompanied by corresponding changes
220: in their timing properties (van der Klis 2006). The source under study is
221: 4U 1728--34, a neutron star LMXB system that is considered to be one of
222: the proto-typical examples of an atoll source. It exhibits variability on a
223: variety timescales, from months down to milliseconds (e.g. Di Salvo et al
224: 2001) as well as exhibiting Type-I bursting behavior.
225:
226:
227: In \S 2 we outline the details of our data selection and reduction
228: procedures, while in \S 3 we describe the FRS methodology. In \S 4 we
229: present our results for \4u offering our interpretation of the variability
230: of individual spectral components. In \S 5 we present our general
231: conclusions, including insights gained into accreion geometry and overall
232: physical system configuratons associated with the separate states.
233:
234:
235:
236:
237: \section{Observations and Data Reduction}
238:
239: Data obtained with RXTE can be collected and telemetered to the
240: ground in many different ways depending on the intensity of a source
241: and the spectral and timing resolution desired. The specific
242: observational modes are selected by the observer and may change
243: during the overall observation. This project requires a temporal
244: sampling of at least twice the highest frequency band to be probed,
245: and enough spectral resolution to separate out the different
246: continuum components as well as to crudely study possible iron line
247: signatures. Since this is an archival study, we had to examine the
248: various datasets for the epochs of interest in order to identify
249: suitable data acquistion modes. Typically, we found that the
250: event-by-event data sampled at 125 microseconds over 64 onboard
251: spectral channels covering the full bandpass of the PCA, suited our
252: purposes. Practically, in order to ensure homogeneity in the
253: reduction process and for signal-to-noise considerations the
254: energy resolution was restricted to be between
255: 16--24 channels covering the energy band 2-16 keV.
256:
257: Light curves were extracted for each onboard channel range using the
258: current RXTE
259: software\footnote{http://heasarc.gsfc.nasa.gov/docs/software/lheasoft/},
260: binned at a resolution of 0.0078125 s. We then divided the data into
261: 256-s segments and, following the prescription of Reig et al (2006), we
262: obtained the Fourier resolved spectra of the source in the following broad
263: frequency bands: 0.008-0.8 Hz, 0.8-8 Hz and 8-64 Hz.
264: In total, we examined a total of approximately 54.3 ks of data from the RXTE
265: archives at 7 epochs (see Table \ref{log}).
266:
267:
268:
269: \subsection{Spectral Index -- QPO Frequency Correlation}
270:
271: It is well established that in accreting neutron star systems, the
272: X-ray luminosity and their spectral and/or timing quantities (X-ray
273: colors, QPO frequencies) tend to correlate with the spectral state
274: of a an individual source, but that these correlations are not identical
275: for the ensemble of sources (see e.g. van der Klis 2006).
276: Furthermore, for a given source, the correlations are much more
277: pronounced over short (hours to days) than on longer time scales
278: (van der Klis 2000 and references therein). Therefore, the use of
279: hardness ratios and source intensity as tracers of the source
280: spectral state can become blurred. In addition, the problem of gain
281: changes associated with the aging of the PCU detectors aboard $RXTE$
282: makes even the determination and comparison of X-ray colors at
283: different epochs difficult.
284:
285: To avoid these difficulties we have opted to use one of the QPO
286: frequencies as a proxy for the state
287: of the source. As has been previously pointed out, e.g. van der Klis
288: (2006), the frequencies of all PDS features, most notably QPOs,
289: correlate with the source intensity and also with its spectral
290: state. Furthermore, there is a noted
291: correspondence in the sequence of the spectral and timing states
292: with source intensity between accreting neutron stars and black hole
293: systems.
294:
295: Motivated by the correlation between the (low-frequency) QPO centroid
296: frequency and photon index in several black-hole XRBs (Vignarca et al.
297: 2003),
298: we have opted to use the corresponding QPO frequency of \4u as a
299: tracer of the configuration changes in the accretion flow. Although
300: as noted above, the neutron star spectra are more complicated than
301: those of the black holes, we find that the correlation between the
302: spectral and timing properties of \4u and similar systems (van der
303: Klis 2006) is sufficiently robust to allow the use of the QPO
304: frequency as an indicator of the spectral state. In fact, Titarchuk
305: \& Shaposhnikov (2005) found a correlation between the QPO frequency
306: and their spectral index parameter $\Gamma$ in their analysis of \4u.
307: However, given that its
308: spectrum is more complicated than those of galactic black holes one
309: should not consider this correlation to be universal.
310:
311: \section{Conventional Spectral and Timing Analysis}
312:
313: Fig 1 shows typical energy spectra of the source corresponding to the
314: observational periods selected. The spectra were extracted from PCA
315: {\em Standard 2} mode data, and in this case no Fourier frequency
316: decomposition has been applied.
317: The response matrix and background models were created using the standard
318: HEADAS software, version 5.3. The number of detectors (PCU) that were
319: switched on for each observation varies, and, in order to be able to
320: compare the spectra, we divided them by the respective number of PCUs.
321:
322: The spectral analysis was performed using the XSPEC analysis package
323: version 11.3.1. We have added systematic errors of 1\% to all channels and
324: have restricted our analysis to the 2-16 keV band only (to match the energy
325: band used in the case of the Fourier-resolved spectra). The results are
326: listed in Table \ref{avsptab0}. The errors quoted for
327: the best fit values correspond to the 90\% confidence limit for one
328: interesting parameter.
329:
330: %In cases that the errors are sufficiently large so that the best-fit
331: %parameter value is consistent with being zero, we simply note the best-fit
332: %value plus upper error and we accept it as the upper 90\% limit for the
333: %respective parameter.
334:
335: We analyzed data at seven different epochs as indicated in Table 1.
336: The 21 and 26
337: September 1997, 25 January 1999 and 9 February 2001,
338: observations were selected because they span a wide range in both
339: the QPO and break frequencies, $\nu_1$ and $\nu_{\rm b}$, respectively,
340: as well as the spectral parameter $\Gamma$ of \citep{tita05}. As
341: mentioned above we inferred the spectral state of \4u\
342: for a given observation based on the pair ($\Gamma$, $\nu_{\rm b}$).
343: In three cases, 7 June 2001, 18 February 1996, and 19 August, 1999
344: the source was known to be in a particlular region of the
345: color-color diagram based on results in the published literature.
346:
347: As shown in Table 2 the average spectra comprise a
348: number of continuum and atomic transition components: (a) A low temperature
349: disk blackbody of innermost temperature $T_{\rm in} \simeq 1$ keV. (b)
350: A blackbody component of $T_{\rm bb} \sim 2$ keV attributed to the
351: boundary layer. (c) A power law with index $\Gamma \simeq 2$, presumably
352: the result of
353: the Comptonization of the disk or boundary layer photons by the accretion
354: disk corona. (d) An Fe line feature at $E_{\rm c} \sim 6.6$ keV and, (e) An
355: Fe line edge at $E_{\rm edge} \sim 9.5$ keV. In figs \ref{is}a,b. we
356: present the overall spectrum as the decomposition of these three components
357: for two epochs when the source was in its island (09 Feb. 2001) and upper
358: banana (21 Sep. 1997) states. The disk blackbody component is not required
359: in the spectrum of the island state (IS) of Feb 09, 2001 but it is
360: necessary to fit the lower banana (LB) and upper banana (UB) branch spectra.
361: The blackbody temperature $T_{\rm bb}$ appears to increase as the source moves
362: from the IS to the UB while its normalization traces the overall increase
363: in the source counts along the above sequence. The progression from the IS
364: to the LB and UB states is also accompanied by an increase of the power law
365: normalization and its index $\Gamma$. The latter is usually interpreted as
366: a steepening of the Comptonization spectrum due to the cooling of the
367: Comptonizing electrons by the increased soft (blackbody) photon flux.
368:
369: We have also found that the source variability properties change
370: markedly with the spectral state. First, as shown in Table 1 the
371: break and QPO frequencies increase monotonically as the source flux
372: %and power law index $\Gamma$ increase. In fig \ref{psd} we present
373: and power law index $\Gamma$ increase. In Fig 3 we present
374: the PSD of the source in three different spectral states. The island
375: state exhibits the highest RMS variability at all frequency bands,
376: its $\nu P_{\nu}$ spectrum exhibiting a maximum at the limit of the
377: searched frequencies ($\sim 500$ Hz). In the lower banana state the
378: $\nu P_{\nu}$ spectrum rises from the low frequencies to a plateau
379: (i.e. $\nu P_{\nu}\propto \nu^0$ at $\nu \sim 20$ Hz) and the RMS
380: variability is lower at all frequencies than that of the island
381: state. Finally, in the upper banana state the shape remains
382: qualitatively similar to that of the LB but with greatly reduced
383: amplitude and the additional appearance of excess power at
384: frequencies lower than 0.3 Hz. We believe that this sequence in the
385: timing properties is suggestive of the progressive influence of the
386: effects of radiation pressure in the accretion process (negligible
387: in the IS but progressively important towards the UB state). It
388: is reasonable to consider the possibility that
389: the radiation feedback on the flow could
390: damp the short scale variability by the diffusion of photons through
391: the accreting plasma; inundating the system with soft photons
392: that could conceivably also reduce the overall variability
393: amplitude \citep{kazhua}.
394:
395: \section{Fourier Resolved Spectral Analysis}
396:
397: In this section we present the results of our Fourier resovled spectral
398: analysis for each state and compare the resulting best-fit
399: parameters to those of the previous section (i.e, those obtained from the
400: average energy spectra). As before, the { XSPEC} package was used for
401: the model fitting. In all cases we added an absorption component which
402: we fixed at $3\times10^{22}$ cm$^{-2}$ (e.g. Narita, Grindlay \& Barret, 1999).
403: A uniform systematic error of 1\%
404: was added quadratically to the statistical error of all Fourier spectra in
405: each energy channel. Errors quoted for the best-fit values correspond to
406: the 90\% confidence limit for one interesting parameter.
407:
408: The virtue of the frequency-resolved spectra is that they receive
409: significant contribution only from the spectral components that are
410: variable on the time scales sampled; the FRS do not
411: represent photon rates but rather variability amplitudes at a given
412: energy. Therefore, through Fourier-resolved spectroscopy we can
413: investigate whether different spectral components in the overall
414: spectrum of the source (e.g. blackbody, Comptonization, iron line)
415: are variable and at which frequency. In general, the interpretation
416: of the FRS is not unique (see Papadakis et al. 2005,
417: Reig et al. 2006 for some insights). For spectral features
418: that result from the reprocessing of higher energy ($E\ge 7$ keV)
419: continuum, such as the iron Fe K$\alpha$ line and the Compton
420: reflection "hump", one plausible interpretation of
421: the Fourier-resolved analysis is that the light crossing delay $R/c$
422: (where $R$
423: is the size of the reprocessing region) filters out variability of
424: these features on frequencies higher than $\sim c/R$. We would note that
425: while this interpretation may be the most straightforward one,
426: other interpretations, such as scenarios
427: involving screening geometries are also possible e.g. (\cite{revn99}).
428:
429: In the following subsections we describe the results of our
430: analysis for each of the spectral
431: states of the source. In addition to the four sets of data mentioned
432: above, we analyzed data for three additional epochs for which the spectral
433: state of the source was previously documented. Our objective is to
434: compare the results obtained above with the traditional designations
435: of island and banana states in order to characterize the spectral
436: and timing properties of each state. We searched in the literature
437: for observations for which \4u\ was clearly identified as being in
438: one of the characteristic atoll states. \4u was in the island state
439: on June 7, 2001 \citep{migl03}, upper banana state on August 19,
440: 1999 \citep{pira00} and lower banana state on February 18, 1996
441: \citep{salv01}.
442:
443: \subsection{The Island State}
444:
445: The source was in the island state on Feb. 09, 2001, and, as noted by
446: Migliari et al. (2003) \4u\ also appeared to be in the island state during
447: the June 7, 2001 observations. We obtained the 64-channel, 1-ms data sets
448: from those observations and formed the FRS as described in \S 3.
449: We found that a single-temperature blackbody model led to fits which were
450: unacceptable, particularly at our low-frequency band. In particular,
451: residuals at the low- and high-energy channels were evident.
452: Improved fits were provided with an additive combination of the blackbody
453: plus a powerlaw. The powerlaw required is rather
454: flat $\Gamma \simeq 0.5$, but is is poorly constrained, particulary
455: in the Feb. 09, 2001 observation. The reason is that in those model
456: fits the blackbody is the predominant component except in the
457: (several) highest energy energy bins.
458:
459: The blackbody component in these cases is characterized at each frequency
460: band by a temperature of about $kT=1.0-1.7$ keV without an obvious trend.
461: The spectral decomposition parameters for the island-state FRS
462: are given in Tables \ref{ISplfits} and \ref{ISbbplfits}. As summarized
463: there and shown in fig. \ref{island}, the Feb. 9, 2001 FRS, for example,
464: can be fitted well by a
465: blackbody plus powerlaw spectrum with temperature $k T \sim 1.5-1.7$ keV,
466: similar in magnitude to the temperature $T_{\rm bb}$ of the conventional
467: spectra attributed to emission by the boundary layer. { The Power Law
468: (energy) spectral indices} are quite hard ($\simeq 0.6$) although they are
469: marginally constrained. The same spectra can actually be fit by a
470: simple power law absorbed at low energies, without the presence of a
471: black body component. A summary of such fits is given in Table \ref{ISplfits},
472: for an absorbing column of $N_H = 3 \times 10^{22}$ cm$^{-2}$. Also
473: see \ref{island}. The fits
474: exhibit no particular trend with Fourier frequency.
475:
476: We also tested several other simple models.
477: A powerlaw model with low-energy cutoff was unable to
478: constrain the inflection energy value and broken powerlaw fits
479: led to unacceptable results.
480:
481: There seems to be a general trend in the island state FRS
482: towards hard or even positively sloped spectral-energy
483: distributions, i.e. a trend
484: variablity to increase with energy. This is particularly notable
485: for example in the low-frequency spectra for both island state epochs
486: (Tables \ref{ISplfits} and \ref{ISbbplfits}), and it is consistent
487: with the RMS variability versus energy analysis of \cite{revn00}.
488: This is reminiscent of Blazar variability.
489:
490:
491: While none of the FRS results require an iron line - upper limits
492: on its equivalent width range from a few hundred to a thousand
493: eV - it is clearly
494: seen in the conventional spectra for 3 of the 4 epochs
495: denoted in Table 2; specifically the EW determination indicates
496: a $\sim 10\sigma$ significance in several cases. The FRS
497: are harder than the averaged spectra, indicating again that the
498: hard emission may be the predominant component of the source variability.
499:
500: Finally, we note that although the island-state count rate was lower
501: by about a factor of 10 than the banana-state cases, the rms
502: variability in the island state is significantly higher. Thus,
503: the statistics of the island-state FRS are comparable or
504: superior to those of some banana state epochs.
505:
506:
507: \subsection{The Lower Banana state}
508:
509: On Jan. 25, 1999 as well as on Feb. 18, 1996 (also see di Salvo et
510: al. 2001) the source intensity along with the PDS break frequency at
511: about 16 Hz, indicated that \4u\ was in the the lower banana state.
512: A single blackbody model provides acceptable fits for all three of
513: our frequency ranges (our fit results are summarized in Table
514: \ref{frstab1}). Its temperature is consistent with that of the
515: time averaged spectra, indicating that it is the boundary layer
516: component which is responsible for the source variability. The
517: absence of the power law and Fe line components that are present in
518: the time averaged spectral fits, suggests that these components are
519: not variable on the time scales examined. The normalization of this
520: component is comparable in the low and high frequencies and slightly
521: lower at the mid-frequency range, indicating that all frequencies
522: contribute roughly equally to the variations of this component.
523:
524: The single blackbody fit to the FRS for various banana spectral states,
525: and for the three frequency bands have characteristic temperatures
526: that span the nominal kT$\simeq 2-3$ keV range. However, no
527: unambiguous trends
528: with temperature versus frequency are seen.
529:
530: %
531: % There is a marginal
532: % trend of temperature increase with frequency
533: % for the island states of 7 June 2001 and 9 February 2001. In
534: % the low-frequency band, it seems to be the case that the island
535: % state temperature is lower than for the banana state case, but
536: % again the statistics limit us from making a definitive conclusion.
537: % We note that if the FRS are modeled as an additive combination
538: % of thermal and non-thermal (hard) continua, a weak trend
539: % of decreasing temperature with increasing frequency was seen.
540: %
541:
542: \subsection{The Upper Banana State}
543:
544: The source was found to be in the upper banana state on Sept.
545: 26 and Sept. 21 1996, as determined by the value of the PSD break
546: frequency $\nu_b$. We also obtained data for the upper banana state
547: noted during August 19, 1999 by (Piraino et al. 2000). In this case,
548: we found that the RMS variations above $\sim$8~Hz (high-frequency)
549: were apparently small as we were unable to obtain useful
550: signal-to-noise for any of our FRS bands for these data. The overall
551: count rate on the other hand was high, $\sim 10$ times that for the
552: island state. As discussed above a single blackbody provides a good
553: fit to the FRS without the need for additional components at
554: any of the frequency bands. The amplitude of this component appears
555: to increase with decreasing frequency in the 1996 observations while
556: remained the same during the 1999 one, thus indicating that this
557: attribute of the spectra is not characteristic of the particular
558: state.
559:
560:
561: \section{Discussion}
562:
563: We have investigated the spectro-temporal characteristics of
564: the atoll X-ray binary \4u. To this
565: end we have produced the energy and power spectra of the source in
566: its different spectral states, i.e. island (IS), lower banana (LB) and
567: upper banana (UB) and also the Fourier-resolved spectra (FRS) of each
568: state for three different frequency bands.
569:
570: Fits of the time average spectra of the source require a combination of a
571: blackbody (to model the boundary layer emission), a multicolor blackbody
572: (to model emission by the accretion disk), a power law (to model coronal
573: emission), an Fe line and a low-energy galactic-absorption
574: column of $N_H = 3\times 10^{22}$ cm$^{-2}$. The results
575: for the conventional spectra
576: are given in Table \ref{avsptab0}. As the source crosses from the island to
577: the lower and upper banana states the source spectra change significantly
578: %(see fig. \ref{avsp}). Our spectral decomposition indicates for
579: (see Fig. 1). Our spectral decomposition indicates for
580: this sequence the component exhibiting the largest parameter
581: range is the multi-color disk. While its presence is not required
582: in the island state, it is necessary for acceptable fits in the other states
583: with approximately
584: constant temperature and highly variable normalization. Unfortunately, the
585: RXTE spectra do not cover the energy regime $E < 2$ keV, necessary for the
586: more accurate determination of this component. Also variable is the index
587: $\Gamma$ of the power law component; the increase of $\Gamma$ with the
588: total flux of the source is generally attributed to the decrease in the
589: temperature of the Comptonizing medium by the increase of soft thermal
590: photon flux and the resulting Compton cooling effects. The temperature of the
591: blackbody component exhibits an increase between the island and the banana
592: states (referring to the blackbody plus powelaw fits to the island state),
593: but it appears to remain constant once the banana states are
594: reached. This suggests a behavior different from that of the Z-sources.
595: Gilfanov et al. (2003) find from their FRS analysis
596: that the temperature of this component remains constant,
597: suggesting that
598: the sources' boundary layer is radiation pressure dominated. Following the
599: same argument, the increase of the thermal component temperature with
600: increasing source flux leads to the conclusion that in the atoll sources
601: the boundary layer is not radiation dominated in the island state, while
602: it appears to become so in the banana states. Finally, the EW of the Fe
603: line appeared to remain constant to within statistics throughout the source's
604: spectral sequence.
605:
606: The variability properties of the source, as manifest by the source
607: power-density spectra, change significantly with its spectral
608: state: The overall RMS fluctuations across all frequencies decrease with
609: increasing source flux; in addition the PDS shape changes as the source
610: %moves from the island to its banana states (see fig. \ref{PDS}). Finally,
611: moves from the island to its banana states (see Fig. 3). Finally,
612: in the general case, the frequency of features in the PDS such as breaks
613: and QPOs increase along the island to upper banana sequence. We suggest
614: that the decrease in variability is due to the effects of radiation
615: pressure on the accretion flow, which through diffusion damps the high
616: frequency components and the overall amplitude.
617:
618: We have elaborated further on the study of the spectro-temporal
619: properties of the source by producing its Fourier-resolved spectra
620: at three frequency bands for each of the source's spectral states,
621: namely $0.008 - 0.8$ Hz, $0.8 - 8$ Hz and $8 - 64$ Hz. While
622: interesting features in the Fourier domain are known to occur at
623: higher frequencies (i.e. kHz QPO) our data did not allow the
624: application of this technique to these frequencies with any
625: significance. The source state was determined either from prior
626: analysis gleaned from the published literature or was defined by the
627: spectral hardness--break frequency correlation. In the latter
628: cases, we mostly relied on the measurements and analysis of
629: \citep{tita05}, but in several instances we measured these
630: quantities ourselves. Several notable trends are gleaned from our
631: analysis and are identified below.
632:
633: The results of our FRS analysis can be summarized as follows:
634:
635: (i) The iron line at $\sim$6.6 keV, distinctly present in the
636: conventional spectra ($\simeq 10 \sigma$ for the banana state cases),
637: is not apparent in any of our frequency-resolved spectra.
638: For example, when we fix the line energy at 6.6 keV and the width
639: to 0.6 keV, we derive only upper limits on the equivalent widths
640: (values ranged from about 50 to 600 ev).
641: This suggests that, at leaset within the limitations of our
642: measurements, there may be no significant Fe line
643: variability on time scales { between $10^{-2} - 10^2$ s}. This
644: is in agreement with the lack of correlation between the Fe
645: K$\alpha$ equivalent width and QPO frequencies reported by
646: \citep{tita05}. These authors concluded that the size of the
647: Fe K$\alpha$ emitting region is not of the same order as that of the
648: X-ray continuum emission, the former possibly being orders of
649: magnitude larger than the latter. On the other hand, a different
650: view has been promoted by Miniutti et al. (2003) with respect to the
651: Fe line variability in AGN. In as much as the geometry and physics
652: of Fe line production in these very diverse objects is similar, such
653: an interpretation should also be considered.
654:
655: (ii) No disk component is present in the FRS, indicating as in the
656: case of the iron line, that the accretion disk emission is significantly
657: less variable at the frequencies sampled by our study, i.e. 0.08-64 Hz.
658: This result agrees with those found in other neutron-star \citep{revn06}
659: and black-hole \citep{reig06} binaries.
660:
661: (iii) We have
662: examined five data sets corresponding to the banana state
663: spectral/temporal configuration. Our FRS reveal that the temperature of the
664: thermal spectral component is consistent with being constant across
665: frequencies and states. Unfortunately, this statement cannot be made with
666: high degree of confidence for the UB states as the decrease in the variability
667: amplitude makes the determination of the Fourier resolved
668: parameters of the spectra difficult.
669:
670: %
671: % Similarly, the power law index was found consistent with being constant
672: % even though it seems to be higher in the 25
673: % January 1999 spectra. Nonetheless, the values of $\Gamma$ obtained in the
674: % FRS are, { at high level of significance, consistently } smaller than those
675: % of the average spectra, { a general feature of such analyses}.
676: %
677:
678: (iv) Two observations corresponding to the island state reveal evidence
679: for a spectral hardening at the lowest frequencies examined. The island
680: state are generally harder than the corresponding frequency averaged
681: spectra. Single blackbody model fits to the island state FRS,
682: were unsatisfactory, with a significant high-energy residual appearing,
683: particularly for our lowest frequency interval. The addition of a
684: hard powerlaw component leads to improved fits to the data, although
685: the slope is poorly constrained. { We have
686: also tried an alternative models, the most satisfactory of which was
687: a single power law with a low energy
688: neutral absorpion by a column of $N_H = 3 \times 10^{22}$ cm$^{-2}$. It was
689: found that this presents an good fit to the data with the
690: corresponding parameters as given in Table \ref{ISplfits}. In general,}
691: the FRS are notably harder than the conventional spectra for the island state,
692: indicating a substantial contribution of the high-energy emission to the
693: source variablity, at least for this spectral state, a situation not unlike that
694: found in other studies as previously referenced.
695:
696: (v) The results of our FRS analysis are in general agreement with those of
697: Gilfanov et al. (2003) of Z-sources, in that the FRS spectra are dominated
698: by a single blackbody component of constant temperature when
699: $\dot{M}\sim 0.1-1 \; \dot M_{\rm Edd}$, that is, for Z sources and atoll
700: sources in the banana state. This is attributed
701: to emission by the boundary layer which in addition is radiation pressure
702: dominated hence the independence of the temperature on the source flux (the
703: local flux is always that of Eddington). The temperature of this component
704: is consistent with that of the blackbody component in the time average
705: spectra.
706:
707: (vi) Finally, we have extended the study of the Fourier-resolved spectra of
708: neutron-star low-mass X-ray binaries to the lower luminosity end. Previous
709: studies had concentrated on { higher luminosity states of these objects},
710: i.e., Z-sources or atoll sources in
711: their soft/high spectral state \citep{gilf03,gilf05,revn06}. The X-ray
712: luminosity in the 2--16 keV range of \4u\ in the island state, assuming a
713: distance of 4.3 kpc \citep{fost86} is $\sim 3\times 10^{36}$ erg s$^{-1}$,
714: i.e. $\sim 10^{-2} L_{\rm Edd}$, about one order of magnitude lower than
715: previous studies. We found that the temperature of the blackbody component
716: of the conventional spectrum of the island state is significantly lower than
717: that of the corresponding FRS when the powerlaw plus blackbody model
718: is applied. The most straightforward
719: interpretation of this fact is
720: that the conventional island-state spectrum includes
721: contribution from a component of lower temperature and little variability.
722:
723: Future applicaton of Fourier-Resolved spectroscopy to low-mass X-ray
724: binaries, incorporating both larger data sets and additional objects of the
725: representative subclasses offers the posiblity for furher insight
726: into each of these issues.
727:
728:
729: {\noindent{\bf Acknowledgements}}
730:
731: The authors wish to acknowledge Nikolai Shaposhnikov for providing
732: the \citep{tita05} spectral-QPO frequency correlation results in
733: tabular form. We also thank the referee, Mikhail Revnivtsev for
734: a thorough reading of the manuscript and a number of useful suggestions.
735: This work has made use of data obtained through the High
736: Energy Astrophysics Science Research Center of the NASA Goddard
737: Space Flight Center. Part of this work was supported by the General
738: Sectreteriat of Research and Technology of Greece.
739:
740: \clearpage
741:
742: \begin{thebibliography}{999}
743:
744: \bibitem[di Salvo et al.(2001)]{salv01}
745: di Salvo, T., M\'endez, M., van der Klis, M., Ford, E., Robba, N.R. 2001,
746: ApJ, 546, 1107
747:
748: \bibitem[Ford \& van der Klis(1998)]{ford98}
749: Ford, E., \& van der Klis, M., 1998, ApJ, 506, L39
750:
751: \bibitem[Foster etal.(1986)]{fost86}
752: Foster, A. J., Fabian, A. C., Ross, R. R. 1986, MNRAS, 221, 409
753:
754: \bibitem[Gilfanov et al.(2000)]{gilf00}
755: Gilfanov, M., Churazov, E. \& Revnivtsev, M. 2000, MNRAS, 316, 923
756:
757: \bibitem[Gilfanov et al.(2003)]{gilf03}
758: Gilfanov, M., Revnivtsev, M., \& Molkov, S.,
759: A\&A, 410, 217
760:
761: \bibitem[Gilfanov \& Revnivtsev(2005)]{gilf05}
762: Gilfanov, M., \& Revnivtsev, M. 2005, Astr, Nachr., 326, 812
763:
764: \bibitem[Kazanas \& Hua(1999)]{kazhua}
765: Kazanas, D. \& Hua, X.-M., 1999, ApJ, , 519, 750
766:
767: \bibitem[Mendez \& van der Klis(1999)]{mend99}
768: Mendez, M., \& van der Klis, M., 1999, ApJ, 517, L51.
769:
770: \bibitem[Migliari et al.(2003)]{migl03}
771: Migliari et al., 2003, MNRAS, 342, L67
772:
773: \bibitem[Narita et al (1999)]{nar99}
774: Narita, T., Grindlay, J.E., and Barret, D., bull AAS, 31, 904
775:
776: \bibitem[Papadakis et al. (2005)]{papa05}
777: Papadakis, I. E., Kazanas, D., \& Akylas, A. 2005, ApJ, 631, 727
778:
779: \bibitem[Papadakis (2006)]{papa06}
780: Papadakis, I.E., Ioannou, Z., \& Kazanas, D.,
781: 2006, ASNA, 327, 1047
782:
783: \bibitem[Piraino et al (2000)]{pira00}
784: Piraino, S., Santangelo, A., \& Kaaret, P., 2000, A\&A, 360, L35
785:
786: \bibitem[Reig et al. (2006)]{reig06}
787: Reig, P., Papadakis, I.E., Shrader, C.R.,
788: \& Kazanas, D., 2006, ApJ, 644, 424
789:
790: \bibitem[Revnivtsev et al. (1999)]{revn99}
791: Revnivtsev, M., Gilfanov, M., \& Churazov, E. 1999, A\&A, 347, L23
792:
793: \bibitem[Revnivtsev et al. (2000)]{revn00}
794: Revnivtsev, Mikhail G.,
795: Borozdin, Konstantin N., Priedhorsky, William C.,\& Vikhlinin, Alexey
796: 2000, Ap.J., 230, 955
797:
798: \bibitem[Revnivtsev et al.(2001)]{revn01}
799: Revnivtsev, M., Gilfanov, M., \& Churazov, E. 2001, A\&A, 380, 502
800:
801: \bibitem[Revnivtsev \& Gilfanov (2006)]{revn06}
802: Revnivtsev, M. G., Gilfanov, M. R. 2006, A\&A, 453, 253
803:
804: \bibitem[Sobolewska, M. A.; Zycki, P. T. (2006)]{Sob06}
805: Sobolewska, M.A., \& Zycki, P.T., 2006, MNRAS, 370, 405
806:
807: \bibitem[Titarchuk \& Shaposhnikov (2005)]{tita05}
808: Titarchuk, L. \& Shaposhnikov, N. 2005, ApJ, 626, 298
809:
810: \bibitem[van der Klis (2000)]{klis00}
811: van der Klis, M., 2000, ARA\&A, 38, 717
812:
813: \bibitem[van der Klis (2006)]{klis06}
814: van der Klis, M., 2006,
815: in "In: Compact stellar X-ray sources", Ed. by W. Lewin \&
816: M. van der Klis. Cambridge Astrophysics Series, No. 39. Cambridge
817: University Press.
818:
819: \bibitem[Vignarca et al.(2003)]{vign03}
820: Vignarca, F., Migliari, S., Belloni, T., Psaltis, D., van der Klis, M.
821: 2003, A\&A, 397, 729
822:
823: \bibitem[Wijnands (2001)]{wijn01}
824: Wijnands, R., 2001, AdSpR, 28, 469
825:
826: \bibitem[Zycki et al (2003)]{zycki03}
827: Zycki, P., 2003, ASPC, 290, 135
828:
829: \end{thebibliography}
830:
831: %
832: % tables, Figures ...
833: %
834: \clearpage
835: %----------------------------------------------------------------------------
836: \begin{table}
837: \begin{center}
838: \caption{Log of the observations \label{log}}
839: \begin{tabular}{lccccc}
840: \tableline \tableline
841: Epoch &$\Gamma$ &$\nu_{\rm b}$ &Source &2-16 keV &Exposure\\
842: & &(Hz) &state &flux (c/s)$^*$ &time(ks) \\
843: \tableline
844: Feb 9, 2001 &1.40 &0.75 & IS+ &640 &4.40 \\
845: Jan 25, 1999 &2.07 &12.54 & LB+ &1824 &11.3 \\
846: Sep 26, 1997 &2.88 &24.2 & UB+ &1659 &12.8 \\
847: %May 5, 2000 &2.54 &22.64 & UB+ &1213 &1.28 \\
848: Sep 21, 1997 &5.54 &41.55 & UB+ &1953 &13.5 \\
849: %Feb16, 1996 &6.0 &40.43 & &2025 &3.07 \\
850: %Sep 24, 1997 & & & &2265 &3.07 \\
851: \tableline
852: Jun 7, 2001 & 1.46 & 0.9 &IS &445 &6.91 \\
853: Feb 18, 1996 &2.41 &12.21 &LB &1380 &3.33 \\
854: Aug 19, 1999 &3.35 & 24.1 &UB &2600 &2.05 \\
855: \tableline \tableline
856: \tablenotetext{*}{For 5 PCU}
857: \tablenotetext{+}{Inferred from PDS}
858: %\tablenotetext{b}{Parameter value kept fixed, as in \citet{kale05}}
859: \end{tabular}
860: \end{center}
861: \end{table}
862: %%----------------------------------------------------------------------------
863: %
864: \clearpage
865: %----------------------------------------------------------------------------
866: \begin{table}
867: \begin{center}
868: \caption{Conventional Spectral Analysis Results \label{avsptab0}}
869: \begin{tabular}{lcccc}
870: \hline \hline
871: Observation &Feb 9, 2001 &Jan 25, 1999 &Sep 26, 1997 &Sep 21, 1997 \\
872: \hline \hline
873: Flux$^1$ (erg s$^{-1}$ cm$^{-2}$) &1.4 &4.1 &3.8 &4.6 \\
874: \hline
875: \multicolumn{5}{c}{DISKBB} \\
876: \hline
877: $T_{\rm in}$ (keV) &-- &1.7$^{+1.1}_{-0.5}$ &1.06$\pm$0.03 &1.4$\pm$0.1 \\
878: norm &-- &5$^{+10}_{-4}$ &110$\pm$15 &45$^{+20}_{-10}$ \\
879: \hline
880: \multicolumn{5}{c}{BBODY} \\
881: \hline
882: $kT_{\rm bb}$ (keV) &1.27$\pm$0.05 &2.3$\pm$0.4 &2.1$\pm$0.1 &2.2$\pm$0.1 \\
883: norm ($\times 10^{-2}$)&0.34$\pm$0.08 &1.3$^{+0.5}_{-0.1}$ &2.8$\pm$0.1 &3.3$\pm$0.3 \\
884: \hline
885: \multicolumn{5}{c}{POWER LAW} \\
886: \hline
887: $\Gamma$ &1.75$\pm$0.04 &2.03$\pm$0.08 &2.1$\pm$0.2 &2.3$^{+0.2}_{-1.4}$ \\
888: norm &0.28$\pm$0.04 &1.0$\pm$0.2 &0.4$\pm$0.2 &0.42$\pm$0.05 \\
889: \hline
890: \multicolumn{5}{c}{GAUSS} \\
891: \hline
892: E$_{\rm c}$ (keV) &6.6$\pm$0.2 &6.6$\pm$0.2 &6.5$\pm$0.3 &6.6$\pm$0.3 \\
893: $\sigma$ (keV) &0.5$\pm$0.4 &0.6$\pm$0.2 &0.8$\pm$0.3 &0.7$\pm$0.3 \\
894: EW (eV) &100$\pm$60 &150$\pm$15 &180$\pm$20 &140$\pm$15 \\
895: \hline
896: \multicolumn{5}{c}{EDGE} \\
897: \hline
898: E$_{\rm edge}$ (keV)&-- &9.5$\pm$0.5 &9.6$\pm$0.6 &9.9$\pm$0.6 \\
899: $\tau$ &-- &0.04$\pm$0.02 &0.04$\pm$0.03 &0.04$\pm$0.03 \\
900: %\hline
901: %$\chi^2$ &0.75/21 &0.99/25 &2.2/25 &2.4/25 \\
902: \hline \hline
903: \multicolumn{5}{l}{All fits include a fixed absorption component with $N_{\rm H}=3\times10^{22}$ cm$^{-2}$} \\
904: \multicolumn{5}{l}{$^1$: $\times 10^{-9}$ in the energy range 2--16 keV} \\
905: \end{tabular}
906: \end{center}
907: \end{table}
908: %
909: \clearpage
910: %----------------------------------------------------------------------------
911: % insert new tables ...
912:
913: \begin{table}
914: \begin{center}
915: \caption{Results of the spectral fits (single power law) to the
916: Fourier-resolved spectra spectra of the island state. The fits include an
917: absorption compoonent with fixed $N_{\rm H}=3\times 10^{22}$ cm$^{-2}$.
918: \label{ISplfits}}
919: \begin{tabular}{lcc}
920: \hline \hline
921: Observation &09/02/01 &07/06/2001 \\
922: \hline
923: \multicolumn{3}{c}{0.008--0.8 Hz} \\
924: \hline
925: $\Gamma$ &1.7$\pm$0.1 &1.8$\pm$0.2 \\
926: norm ($\times 10^{-2}$) &2.0$\pm$0.3 &0.9$\pm$0.4 \\
927: \hline
928: \multicolumn{3}{c}{0.8--8 Hz} \\
929: \hline
930: $\Gamma$ &1.7$\pm$0.1 &1.7$\pm$0.2 \\
931: norm ($\times 10^{-2}$) &3.4$\pm$0.6 &1.6$\pm$0.5 \\
932: \hline
933: \multicolumn{3}{c}{8--64 Hz} \\
934: \hline
935: $\Gamma$ &1.6$\pm$0.3 &1.4$\pm$0.6 \\
936: norm ($\times 10^{-2}$) &3.0$\pm$1.5 &1.0$\pm$0.8 \\
937: \hline
938: $\chi^2$/dof &1.5/52 &0.91/46\\
939: \hline \hline
940: %\multicolumn{4}{l}{}
941: \end{tabular}
942: \end{center}
943: \end{table}
944:
945:
946: %----------------------------------------------------------------------------
947: \begin{table}
948: \begin{center}
949: \caption{Results of the spectral fits (blackbody plus power law) to the
950: Fourier-resolved spectra spectra of the island state \label{ISbbplfits}}
951: \begin{tabular}{lcc}
952: \hline \hline
953: Observation &09/02/01 &07/06/2001 \\
954: \hline
955: \multicolumn{3}{c}{0.008--0.8 Hz} \\
956: \hline
957: $kT_{\rm bb}$ (keV) &1.7$\pm$0.1 &1.0$\pm$0.3 \\
958: norm ($\times 10^{-3}$) &0.7$\pm$0.1 &0.19$\pm$0.04 \\
959: $\Gamma$ &0.6$\pm$0.3 &0.4$\pm$0.7 \\
960: norm ($\times 10^{-3}$) &0.55$\pm$0.06 &0.35$\pm$0.07 \\
961: \hline
962: \multicolumn{3}{c}{0.8--8 Hz} \\
963: \hline
964: $kT_{\rm bb}$ (keV) &1.5$\pm$0.1 &1.7$\pm$0.3 \\
965: norm ($\times 10^{-3}$) &1.0$\pm$0.1 &0.5$\pm$0.1 \\
966: $\Gamma$ &0.6$^f$ &0.4$^f$ \\
967: norm ($\times 10^{-3}$) &1.4$\pm$0.1 &0.32$\pm$0.06 \\
968: \hline
969: \multicolumn{3}{c}{8--64 Hz} \\
970: \hline
971: $kT_{\rm bb}$ (keV) &1.7$\pm$0.6 &1.0$^{+2.0}_{-0.5}$\\
972: norm ($\times 10^{-3}$) &0.9$^{+0.9}_{-0.4}$ &0.9$\pm$0.3 \\
973: $\Gamma$ &0.6$^f$ &0.4$^f$ \\
974: norm ($\times 10^{-3}$) &1.6$\pm$0.4 &0.8$\pm$0.2 \\
975: \hline
976: $\chi^2$/dof &0.90/51 &0.73/43\\
977: F-test prob.$^*$($\times 10^{-2}$) &$7\times 10^{-8}$ &$7\times 10^{-6}$\\
978: \hline \hline
979: \multicolumn{3}{l}{$^f$: fixed parameter} \\
980: \multicolumn{3}{l}{$^*$: probability that the addition of the power-law
981: component occurs by chance} \\
982: \end{tabular}
983: \end{center}
984: \end{table}
985:
986:
987: \clearpage
988: %----------------------------------------------------------------------------
989:
990: % %----------------------------------------------------------------------------
991: % \begin{table}
992: % \begin{center}
993: % \caption{Results of the spectral fits (blackbody) to the Fourier resolved
994: % spectra spectra ***need to remove IS column** \label{frstab1}}
995: % \begin{tabular}{lcccc}
996: % \hline \hline
997: % Observation &Feb 9, 2001 &Jan 25, 1999 &Sep 26, 1997 &Sep 21, 1997 \\
998: % \hline
999: % \multicolumn{5}{c}{0.008--0.8 Hz} \\
1000: % \hline
1001: % $kT_{\rm bb}$ (keV) &2.01$\pm$0.08 &2.27$\pm$0.04 &2.27$\pm$0.04 &2.14$\pm$0.05 \\
1002: % norm ($\times 10^{-3}$) &1.02$\pm$0.04 &4.8$\pm$0.1 &5.0$\pm$0.1 &3.7$\pm$0.1 \\
1003: % \hline
1004: % \multicolumn{5}{c}{0.8--8 Hz} \\
1005: % \hline
1006: % $kT_{\rm bb}$ (keV) &1.98$\pm$0.08 &2.02$\pm$0.05 &2.3$\pm$0.2 &2.1$\pm$0.4 \\
1007: % norm ($\times 10^{-3}$) &1.69$\pm$0.07 &3.07$\pm$0.07 &1.5$\pm$0.2 &1.4$\pm$0.3 \\
1008: % \hline
1009: % \multicolumn{5}{c}{8--64 Hz} \\
1010: % \hline
1011: % $kT_{\rm bb}$ (keV) &2.3$\pm$0.3 &2.15$\pm$0.07 &2.2$\pm$0.4 &2.2$\pm$0.3 \\
1012: % norm ($\times 10^{-3}$) &1.8$\pm$0.3 &4.6$\pm$0.1 &2.2$\pm$0.3 &2.5$\pm$0.3 \\
1013: % \hline
1014: % $\chi^2$/dof &1.98/52 &1.80/64 &1.05/53 &0.94/55 \\
1015: % \hline \hline \multicolumn{4}{l}{$^f$: fixed parameter}
1016: % \end{tabular}
1017: % \end{center}
1018: % \end{table}
1019: %
1020: %
1021: % \clearpage
1022: %
1023: % %----------------------------------------------------------------------------
1024: % \begin{table}
1025: % \begin{center}
1026: % \caption{Results of the spectral fits (blackbody) to the Fourier resolved
1027: % spectra **need to remove IS column** \label{frstab2}}
1028: % \begin{tabular}{lccc}
1029: % \hline \hline
1030: % Observation &Jun 7, 2001 &Feb 18, 1996 &Aug 19, 1999 \\
1031: % \hline
1032: % \multicolumn{4}{c}{0.008--0.8 Hz} \\
1033: % \hline
1034: % $kT_{\rm bb}$ (keV) &1.9$\pm$0.3 &2.7$\pm$0.4 &2.4$\pm$0.2 \\
1035: % norm ($\times 10^{-3}$) &0.38$\pm$0.06 &1.0$\pm$0.1 &1.8$\pm$0.1 \\
1036: % \hline
1037: % \multicolumn{4}{c}{0.8--8 Hz} \\
1038: % \hline
1039: % $kT_{\rm bb}$ (keV) &2.1$\pm$0.2 &2.10$\pm$0.08 &2.5$\pm$0.5 \\
1040: % norm ($\times 10^{-3}$) &0.78$\pm$0.06 &2.6$\pm$0.1 &2.2$\pm$0.4 \\
1041: % \hline
1042: % \multicolumn{4}{c}{8--64 Hz} \\
1043: % \hline
1044: % $kT_{\rm bb}$ (keV) &2.7$^{+1.0}_{-0.6}$&2.18$\pm$0.09 &2.8$\pm$0.6 \\
1045: % norm ($\times 10^{-3}$) &1.0$\pm$0.2 &4.2$\pm$0.2 &4.0$\pm$0.8 \\
1046: % \hline
1047: % $\chi^2$/dof &1.57/46 &0.95/37 &0.95/39 \\
1048: % \hline \hline \multicolumn{4}{l}{$^f$: Freeze to the first
1049: % appearance of the parameter}
1050: % \end{tabular}
1051: % \end{center}
1052: % \end{table}
1053: %
1054: %----------------------------------------------------------------------------
1055: \begin{table*}
1056: \begin{center}
1057: \caption{Results of the spectral fits (blackbody) to the Fourier resolved
1058: spectra spectra \label{frstab1}}
1059: \begin{tabular}{lccccc}
1060: \hline \hline
1061: Observation &Feb 18, 1996 &Jan 25, 1999 &Sep 26, 1997 &Sep 21, 1997 &August 19, 1999 \\
1062: \hline
1063: \multicolumn{6}{c}{0.008--0.8 Hz} \\
1064: \hline
1065: $kT_{\rm bb}$ (keV) &2.7$\pm$0.4 &2.27$\pm$0.04 &2.27$\pm$0.04 &2.14$\pm$0.05 &2.4$\pm$0.2 \\
1066: norm ($\times 10^{-3}$) &1.0$\pm$0.1 &4.8$\pm$0.1 &5.0$\pm$0.1 &3.7$\pm$0.1 &1.8$\pm$0.1 \\
1067: \hline
1068: \multicolumn{6}{c}{0.8--8 Hz} \\
1069: \hline
1070: $kT_{\rm bb}$ (keV) &2.10$\pm$0.08 &2.02$\pm$0.05 &2.3$\pm$0.2 &2.1$\pm$0.4 &2.5$\pm$0.5 \\
1071: norm ($\times 10^{-3}$) &2.6$\pm$0.1 &3.07$\pm$0.07 &1.5$\pm$0.2 &1.4$\pm$0.3 &2.2$\pm$0.4 \\
1072: \hline
1073: \multicolumn{6}{c}{8--64 Hz} \\
1074: \hline
1075: $kT_{\rm bb}$ (keV) &2.18$\pm$0.09 &2.15$\pm$0.07 &2.2$\pm$0.4 &2.2$\pm$0.3 &2.8$\pm$0.6 \\
1076: norm ($\times 10^{-3}$) &4.2$\pm$0.2 &4.6$\pm$0.1 &2.2$\pm$0.3 &2.5$\pm$0.3 &4.0$\pm$0.8 \\
1077: \hline
1078: $\chi^2$/dof &0.95/37 &1.80/64 &1.05/53 &0.94/55 &0.95/39 \\
1079: \end{tabular}
1080: \end{center}
1081: \end{table*}
1082:
1083: %
1084: %
1085: \clearpage
1086: %
1087: \begin{figure}
1088: \epsscale{0.5} \plotone{f1.eps} \caption{Results of our (conventional)
1089: spectral analysis for the various epochs and corresponding spectral state
1090: configurations. The spectra are satisfactorily fitted with an absorbed blackbody
1091: plus multi-color accretion disk components at low energies and a
1092: power-law component at higher energies. An Fe line plus edge were also
1093: required (see Table~\ref{avsptab0}).} \label{avsp}
1094: \end{figure}
1095: %
1096: \clearpage
1097:
1098: \begin{figure}
1099: \epsscale{1.0}
1100: %\plottwo{f2a.eps}{f2b.eps} \caption{Spectral
1101: \plotone{f2.eps} \caption{Spectral
1102: decomposition of the average source spectra for the island (09
1103: February 2001) and upper banana (21 September 1997) states. The model
1104: components are described in the text and in Table \ref{avsptab0}, and
1105: are plotted in $\nu f_{\nu}$ space. Specifically, for Feb. 9 2001 a blackbody
1106: component (green), powerlaw (brown) and a gaussian line (purple)
1107: are depicted; for Sep. 21 1997 one can also see the blackbody component (blue),
1108: the multi-color disk (green), the powerlaw (brown) and the line (purple).
1109: As the
1110: source flux increases so does the blackbody temperature, the index
1111: of the power law and the Fe line flux. } \label{is}
1112: \end{figure}
1113:
1114: \clearpage
1115:
1116: \begin{figure}
1117: \epsscale{1.} \plottwo{f3a.eps}{f3b.eps}
1118: \epsscale{0.45} \plotone{f3c.eps} \caption{
1119: Power spectral density for the representative states of \4u: island state,
1120: 7 June 2001 (upper left), lower banana state, 18 February 1996 (upper
1121: right) and upper banana, 19 August 1999 (lower).
1122: }
1123: \label{pds}
1124: \end{figure}
1125:
1126: \clearpage
1127:
1128: \begin{figure}
1129: \epsscale{0.85}
1130: \plotone{f4.eps}
1131: %\plottwo{f4a.eps}{f4b.eps}
1132: %\plottwo{f4c.eps}{f4d.eps}
1133: \caption{Fourier resolved
1134: spectra and their corresponding spectral-model fits for the days
1135: annotated in the figure.
1136: The different frequency bands are as follows: circles $0.008 - 0.8$
1137: Hz; squares $0.8 - 8$ Hz; triangles $8 - 64$ Hz. The solid,
1138: dashed, and dotted curves are the respective model fits
1139: plotted in $\nu f_{\nu}$ space. } \label{frs}
1140: \end{figure}
1141:
1142:
1143: %\clearpage
1144:
1145:
1146: \clearpage
1147:
1148: \begin{figure}
1149: \epsscale{0.85}
1150: %\plottwo{f5a.eps}{f5b.eps}
1151: %\plottwo{f5c.eps}{f5d.eps} \caption{
1152: \plotone{f5.eps} \caption{
1153: The upper figures show the low- and medium-frequency spectra for
1154: (a) the June 7, 2001 and
1155: (b) February 9, 2001 island state observations fitted to an absorbed
1156: powerlaw model (refer to \ref{ISplfits} for the parameter values).
1157: The solid curves are the model in $\nu f_\nu$ space with the
1158: FRS spectral data points overlayed.
1159: The lower figures show the spectral decomposition
1160: of the same two data
1161: sets with an additive blackbody plus powerlaw model. } \label{island}
1162: \end{figure}
1163:
1164: \clearpage
1165:
1166: %
1167: %
1168: \end{document}
1169: %
1170: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1171:
1172: