1: %\documentclass[11pt,aasms4]{aastex}
2: %\documentclass[11pt,aaspp4]{aastex}
3: \documentstyle[11pt,aaspp4,flushrt]{article}
4: %\documentstyle[emulateapj]{article}
5:
6: % Version 1. 03/02/00
7: % Version 2. 03/07/00 comments from TY
8: % Version 3. 03/08/00 - Edited by LCH
9: % Version 4. 03/10/00 comments from LCH
10: % Version 5. 03/15/00 comments from AFP, KM, & TY
11: % Version 5.1 03/15/00 very minor changes
12: % Version 6 04/10/00 referee comments
13: % Version 6.1 04/10/00 comments from AFP
14: % Version 7 5/24/00 proofs
15:
16:
17: \renewcommand{\tabcolsep}{3pt}
18:
19: \newcommand{\NH}{{$N_{\rm H}$}}
20: \newcommand{\LX}{{$L_{\rm X}$}}
21: \newcommand{\LB}{{$L_{\rm B}$}}
22: \newcommand{\LFIR}{{$L_{\rm FIR}$}}
23: \newcommand{\LHa}{$L_{{\rm H} \alpha}$}
24: \newcommand{\LOIII}{$L_{\rm [OIII]}$}
25: \newcommand{\FX}{{$F_{\rm X}$}}
26:
27: \newcommand{\eps}{ergs s$^{-1}$}
28: \newcommand{\pcm}{cm$^{-2}$}
29: \newcommand{\ps}{s$^{-1}$}
30:
31: \newcommand{\asca}{{\it ASCA}}
32: \newcommand{\rosat}{{\it ROSAT}}
33: \newcommand{\HST}{{\it HST}}
34: \newcommand{\xmm}{{\it XMM}}
35:
36: \newcommand{\gtsima}{$\; \buildrel > \over \sim \;$}
37: \newcommand{\simgt}{\lower.5ex\hbox{\gtsima}}
38: \newcommand{\ltsima}{$\; \buildrel < \over \sim \;$}
39: \newcommand{\simlt}{\lower.5ex\hbox{\ltsima}}
40:
41: \slugcomment{To appear in {\it The Astrophysical Journal (Letters).}}
42:
43: \lefthead{Terashima et al.}
44: \righthead{Iron K Variability in NGC 4579}
45:
46:
47: \begin{document}
48:
49: \title{Iron K line Variability in the Low-Luminosity AGN NGC 4579}
50:
51: \author{Yuichi Terashima\altaffilmark{1},
52: Luis C. Ho\altaffilmark{2},
53: Andrew F. Ptak\altaffilmark{3},
54: Tahir Yaqoob\altaffilmark{1,4},
55: Hideyo Kunieda\altaffilmark{5},
56: Kazutami Misaki\altaffilmark{5},
57: and Peter J. Serlemitsos\altaffilmark{1}
58: }
59:
60: \altaffiltext{1}{NASA Goddard Space Flight Center, Code 662,
61: Greenbelt, MD 20771}
62:
63: \altaffiltext{2}{The Observatories of the Carnegie Institution of
64: Washington, 813 Santa Barbara St., Pasadena, CA 91101-1292}
65:
66: \altaffiltext{3}{Department of Physics, Carnegie Mellon University, 5000 Forbes
67: Ave., Pittsburgh, PA 15213}
68:
69: \altaffiltext{4}{Universities Space Research Association}
70:
71: \altaffiltext{5}{Institute of Space and Astronautical
72: Science, Yoshinodai 3-1-1, Sagamihara, Kanagawa 229-8510, Japan}
73:
74:
75: \begin{abstract}
76:
77: We present results of new {\asca} observations of the low-luminosity
78: AGN (LLAGN) NGC 4579 obtained on 1998 December 18 and 28, and we
79: report on detection of variability of an iron K emission line. The
80: X-ray luminosities in the 2--10 keV band for the two observations are
81: nearly identical ({\LX} $\approx$ 2$\times10^{41}$ {\eps}), but they
82: are $\sim$35\% larger than that measured in 1995 July by Terashima et
83: al. An Fe K emission line is detected at $6.39\pm0.09$ keV
84: (source rest frame) which is lower than the line energy
85: $6.73^{+0.13}_{-0.12}$ keV in the 1995 observation. If we fit the Fe
86: lines with a blend of two Gaussians centered at 6.39 keV and 6.73 keV,
87: the intensity of the 6.7 keV line decreases, while the intensity of
88: the 6.4 keV line increases, within an interval of 3.5 yr. This
89: variability rules out thermal plasmas in the host galaxy as the origin
90: of the ionized Fe line in this LLAGN. The detection and
91: variability of the 6.4 keV line indicates that cold matter subtends a
92: large solid angle viewed from the nucleus and that it is located
93: within $\sim1$ pc from the nucleus. It could be identified with an
94: optically thick standard accretion disk. If this is the case, a
95: standard accretion disk is present at the Eddington ratio of $L_{\rm
96: Bol}$/$L_{\rm Eddington} \sim 2\times10^{-3}$. A broad
97: disk-line profile is not clearly seen and the structure of the
98: innermost part of accretion disk remains unclear.
99:
100:
101: \end{abstract}
102:
103: \keywords{galaxies: active --- galaxies: individual(NGC 4579) --- galaxies:
104: nuclei --- galaxies: Seyfert --- X-rays: galaxies}
105:
106: \section{Introduction}
107:
108: Fe line emission in AGNs is an X-ray spectral feature produced through
109: reprocessing by the matter surrounding the nucleus, and it can be used
110: as a probe of the matter in the vicinity of the central black hole,
111: such as accretion disks. Broad Fe emission lines have been detected
112: in a number of Seyfert 1 galaxies, and they have been interpreted as
113: arising from a relativistic accretion disk (e.g. Tanaka et al. 1995;
114: Nandra et al. 1997). Fe K emission has also been detected in several
115: low-luminosity AGNs (LLAGNs). The origin of Fe lines in these
116: objects, however, is unclear. The line centroid energy of some LLAGNs
117: is close to 6.7 keV (Ishisaki et al. 1996; Serlemitsos et al. 1996;
118: Iyomoto et al. 1997; Terashima et al. 1998a, hereafter T98; Roberts,
119: Warwick, \& Ohashi 1999), which is higher than 6.4 keV as seen in
120: Seyfert 1s, while some objects show a line at 6.4 keV (Terashima,
121: Kunieda, \& Misaki 1999). Possible origins of the Fe line at 6.7 keV
122: include an ionized standard accretion disk, a thermal plasma in the
123: outer region of an advection-dominated accretion flow (Narayan, \&
124: Raymond 1999), an ionized absorber and emitter along the line-of-sight
125: (Pellegrini et al. 2000), or a thermal plasma in the host galaxy (Ptak
126: et al. 2000). If the higher line energy is due to ionized iron, the
127: problem arises as to how to achieve a high degree of ionization given
128: the low luminosities and low accretion rates in LLAGNs. The
129: variability of Fe lines provides information on the location of the
130: line-emitting region, and thus constraints on the origin of the lines.
131: Studies of Fe lines in LLAGNs are also of great importance to
132: understand the structure of accretion disks in low luminosity objects.
133:
134: NGC 4579 contains an LLAGN with {\LX}(2--10 keV) =
135: $1.5\times10^{41}$~{\eps} (we assume a distance of 16.8 Mpc; Tully
136: 1988). An {\asca} observation was made on 1995 June 25, and an Fe K
137: line is detected at $6.73^{+0.13}_{-0.12}$ keV (source rest frame) with an
138: equivalent width (EW) of $490^{+180}_{-190}$ eV and a width of
139: $0.17^{+0.11}_{-0.12}$ keV (T98). We observed NGC 4579 again with
140: {\asca} in 1998 to search for possible variability of the Fe line. In
141: this {\it Letter}, we report the change of the line centroid energy
142: associated with a continuum luminosity increase of about 35\%.
143:
144:
145: \section{The {\it ASCA} Data}
146:
147: We observed NGC 4579 on 1998 Dec. 18 and 28 with the {\asca}
148: satellite (Tanaka, Inoue, \& Holt 1994). The two Solid-State Imaging
149: Spectrometers (SIS0 and SIS1; Burke et al. 1994) were operated in the
150: 1CCD Faint mode. The two Gas Imaging Spectrometers (GIS2 and GIS3;
151: Ohashi et al. 1996; Makishima et al. 1996) were operated in the
152: nominal pulse height mode.
153:
154: X-ray spectra were accumulated from circular regions with radii of
155: 4$^\prime$ for the SIS and 6$^\prime$ for the GIS. Background spectra
156: were taken from a source-free region in the same field. The effective
157: exposure times, after data screening, were 18.6 ks for each SIS and
158: 19.6 ks for each GIS on December 18, and 17.6 ks for each SIS and 19.6
159: ks for each GIS on December 28. The count rates were similar for both
160: observations: 0.14 counts s$^{-1}$ for SIS and 0.13 counts s$^{-1}$
161: for GIS, after background subtraction. The spectra from SIS0 and SIS1
162: were combined, as were those from GIS2 and GIS3. We fitted the SIS0
163: plus SIS1 spectra and the GIS2 plus GIS3 spectra simultaneously. It is
164: known that SIS and GIS spectra at the low-energy band ($<$1 keV) are
165: inconsistent with each other due to accumulated radiation damage of
166: the SIS (see Yaqoob et al. 2000). In order to minimize the uncertainty
167: in the low-energy band, we used the region 1.0--10 keV for the SIS. We
168: confirmed that the SIS and GIS data within this band yield
169: statistically consistent spectral fits.
170:
171:
172: \section{Results}
173:
174: % (1) averaged spectrum
175:
176: We fitted the combined spectrum of the two observations in 1998. At
177: first we fitted the spectrum with a power law plus Gaussian model and a
178: thermal bremsstrahlung plus Gaussian model. The former model provided
179: an acceptable fit with $\chi^2=201.0$ for 227 degrees of freedom
180: (dof). The latter model resulted in a significantly worse fit with
181: $\chi^2=302.7$ for 227 dof. Spectral features of a soft thermal
182: plasma, detected in the 1995 observation, were not clearly
183: seen in the residuals. We tried to add a Raymond-Smith thermal plasma
184: component (hereinafter RS) to the power law plus Gaussian model. The
185: temperature and abundance of the RS component were fixed at the values
186: of T98 ($kT$=0.90 keV and abundance = 0.5 solar), and the normalization
187: was left to vary. The absorption column density for the RS component
188: was fixed at the Galactic value ({\NH} = $3.1\times10^{20}$ {\pcm};
189: Murphy et al. 1996). The $\chi^2$ improved only by $\Delta \chi^2$ =
190: 1.0. Thus we found no clear evidence for the presence of soft thermal
191: emission in the 1998 observation. This is probably for the
192: following reasons. First, the flux of the hard component has brightened
193: by a factor of 1.35 from 1995, and thus the soft thermal component has been
194: diluted. Second, we ignored the soft energy band below 1 keV of the
195: SIS data in the fit, where the soft component is expected to be more
196: prominent. The spectral slope of the hard component is nearly
197: unaffected by the addition of the soft thermal component, while the
198: absorption column density depends on it. We use the power law + Gaussian + RS
199: model in the following analysis.
200:
201: \placefigure{fig-1}
202:
203: \placefigure{fig-2}
204:
205: \placetable{tbl-1}
206:
207: The best-fit parameters for the continuum and X-ray fluxes in the
208: 2--10 keV band are shown in Table 1, as well as the fitting results of
209: the 1995 data taken from T98. The Fe line parameters are summarized in
210: Table 2. The SIS and GIS spectra (along with the best-fit continuum
211: model), and ratios of data to the best-fit continuum around Fe lines
212: in the 1998 observation are shown in Figure 1. Ratios of data to the
213: best-fit continuum for the 1995 observation are also shown in Figure 1
214: for comparison. The hard X-ray luminosity in the 2--10 keV band ({\LX}
215: = $2.0\times10^{41}$ {\eps}) is $\sim$35\% larger in 1998 than in 1995.
216: Although the best-fit spectral slope has slightly steepened, the
217: statistical errors overlap each other. No significant change of the
218: absorption column density is seen either, although the errors are
219: large because of the limited energy band of SIS used in the analysis.
220: An intriguing spectral change is the centroid energy of the the Fe
221: line. The line center energy varied from $6.73^{+0.13}_{-0.12}$ keV in
222: 1995 to $6.39\pm0.09$ keV in 1998. Confidence contours for the line
223: center energy versus normalization for the two observations are shown
224: in Figure 2. The change of the line center energy is significant at
225: more than 90\% confidence level for two interesting parameters. Note
226: that the calibration uncertainty of the energy scale is less than 1\%.
227: We examined the gain uncertainties by using the gold edge around 2.3
228: keV and the Cu-K line of the GIS detector background and found no
229: evidence for a significant gain shift larger than the 1\% level. We
230: also confirmed that all the four detectors provided Fe K line center
231: energies consistent with each other within 1\% using a recent
232: observation of the bright AGN NGC 5548 in 1998 June and July.
233:
234: %[$>$2.3\% gain shift is rejected by chi-square fit to the edge and
235: %Cu-K line.])
236:
237: \placetable{tbl-2}
238:
239: \placefigure{fig-3}
240:
241: % double Gaussian fit
242:
243: The change of the line center energy could be caused by a change of
244: the ionization state of the line emitter or by variability of multiple
245: emission lines with different line centroid energies. In order to
246: quantify the second possibility, we modeled the line feature with a
247: combination of two Gaussians. We fixed the line center energies and
248: widths of each line at the best-fit values in 1995 and 1998. The
249: obtained EWs and intensities are shown in Table 2 and Figure 3. The
250: intensity and EW of the 6.7 keV line decreased significantly. The
251: intensity of the 6.4 keV line significantly increased, while the EW
252: marginally increased. The continuum parameters in this fit are same
253: as in the single Gaussian model fits shown in Table 1. If we assume
254: zero width for the 6.7 keV line instead of the best-fit value in 1995,
255: the conclusion on line variability is unchanged.
256:
257:
258: % variability between two observations (10 days apart)
259:
260: Finally, we fitted the spectra of the two observations in 1998
261: separately to search for spectral variability on a time scale of 10
262: days. We used the same model in the fit of the combined spectrum. The
263: Fe line was modeled with a single Gaussian. The obtained spectral
264: parameters are also summarized in Table 1 and Table 2. The X-ray
265: luminosity in the 2--10 keV band is slightly ($\sim$15\%) higher in
266: the first observation compared to the second. We found no clear
267: spectral change between these two observations.
268:
269:
270: \section{Discussion}
271:
272: We found that the center energy of the Fe K line in NGC 4579 decreased
273: in 3.5 yrs, while the luminosity in the 2--10 keV band increased by
274: 35\%. We examined two models to represent the observed variability: a
275: single Gaussian model and a double Gaussian model. In the double
276: Gaussian model, the center energies were fixed at 6.39 keV and 6.73
277: keV. The results of these model fits indicate that the line intensity
278: of the 6.7 keV line varied within an interval of 3.5 yr. This
279: variability indicates that the emitter of the Fe line at 6.7 keV is
280: located within $\sim1$ pc from the nucleus, and thus rules out hot gas
281: in the host galaxy (starburst activity and/or ridge emission, which is
282: diffuse X-ray emission with $kT\sim5-10$ keV, as observed in our
283: Galaxy) as the origin of the 6.7 keV Fe line. On the other hand, the
284: Fe line at 6.4 keV has an origin in fluorescence in cold or slightly
285: ionized matter illuminated by a central X-ray source (Makishima 1986).
286: Thus, the observed Fe lines at 6.4 keV and 6.7 keV most likely
287: originate from the AGN itself rather than from hot gas further out in
288: the host galaxy. The relatively strong 6.4 keV line in 1998 (EW =
289: $250^{+105}_{-95}$ eV) indicates that cold matter subtends a large
290: solid angle viewed from the nucleus (e.g., Gerorge \& Fabian 1991). A
291: part of the line could come from an obscuring torus which is assumed
292: in unified models of AGNs (e.g., Antonucci 1993). If the torus is
293: present and located at further than 1 pc from the center, the Fe line
294: from it does not vary within an interval of $\sim$ 3 yr. Then, the
295: torus contribution can be at most $8.5\times10^{-6}$ photons$^{-1}$
296: {\ps} {\pcm}, which is the measured 90\% confidence upper limit on the
297: constant 6.4 keV line from the 1995 observation (see Table 2). This
298: line flux corresponds to an EW of 155 eV for the continuum level in
299: 1998. The rest of the 6.4 keV line should come from the region closer
300: to the nucleus. Such a Fe line emitter could be identified with an
301: optically thick accretion disk. If this is the case, a standard
302: accretion disk should exist even at the low Eddington ratio of NGC
303: 4579 ($L_{\rm Bol}/L_{\rm Eddigton} = 2.0\times10^{-3}$; Ho 1999). The
304: presence of an optically thick accretion disk is consistent with the
305: estimate of the transition radius from an advection dominated
306: accretion flow to a standard disk ($\sim$100 Schwarzschild radii) by
307: Quataert et al. (1999). However, the suggestion of a truncated disk
308: might not be appropriate
309: to achieve the observed large equivalent width ($\approx$250
310: eV). Therefore, cold matter beyond the accretion disk, such as an
311: obscuring torus, must contribute somewhat to the observed EW of the
312: 6.4 keV line, if the truncated disk interpretation is correct.
313:
314:
315: The origin of the 6.7 keV line seen in 1995 still remains puzzling,
316: however. If the Fe lines in 1995 and 1998 originated from the same
317: region, the decrease of ionization state responding to the small flux
318: increase of only 35\% is difficult to be explained by photoionization
319: under an assumption of a constant density. If the ionized matter is
320: located far from the nucleus, the observed behavior of the Fe line
321: could be due to a time lag between the continuum and Fe line
322: variability. The X-ray flux in the 0.5--2 keV band observed with the
323: {\rosat} PSPC in 1991 Dec. is a factor of 2.3 higher than that of the
324: first {\asca} observation in 1995 June (Ptak et al. 1999). Therefore,
325: the decrease of ionization state might be explained by a time lag
326: effect, if the Fe line emitter is located 1 pc away form the nucleus.
327: In order to explain the change of ionization state from \ion{Fe}{25}
328: (6.7 keV line) to \ion{Fe}{17} or less (6.4 keV line), the ionization
329: parameter $\xi = L/nR^2$ should change from $\log \xi \approx 3$ to 2
330: (e.g., Kallman, \& McCray 1982), where $L$, $n$, $R$ are the
331: luminosity of ionizing photons, the number density of photoionized
332: matter, and the distance from the ionizing source to the photoionized
333: matter, respectively. The observed variability, however, is not enough
334: to explain the change of the ionization parameter unless the density
335: and/or geometry of the ionized matter also changed, or large-amplitude
336: variability occurred during the unobserved span. Long-term
337: variability of a factor 10 is reported in only a few objects (M81,
338: Pellegrini et al. 2000; and possibly M51, Terashima et al. 1998b).
339:
340: Alternatively, the 6.4 keV and 6.7 keV lines may originate from
341: different regions. If most of the 6.4 keV line comes from the torus
342: and the 6.7 keV line is from different photoionized matter, the
343: disappearance of the 6.7 keV line might be understood by increasing
344: the ionization state such that the gas is almost fully ionized. In
345: this case, the 6.4 keV line could vary in response to the continuum
346: variability with some time lag which depends on the size and geometry
347: of the torus. However, the small flux change, again, is not enough to
348: change the ionization state so drastically.
349:
350: Thus, it is conceivable that the physical conditions (density and/or
351: geometry) of the ionized matter changed between the two observations
352: whether the 6.4 keV and 6.7 keV lines originate from a single region
353: or different regions. Such a change could be caused by a change of the
354: ionization balance in the surface layer of the standard accretion
355: disk, or in the boundary region between the standard disk and the
356: advection-dominated flow. It is worth noting that there are several
357: Seyfert 1 galaxies showing the change of Fe K line centroid energy
358: accompanied by small variations in the continuum luminosity (Weaver,
359: Gelbord, \& Yaqoob 2000). The cause of such variability, however, is
360: yet to be understood. We note also that a drastic change of Fe line
361: profile is observed during a flare in MCG--6--30--15 (Iwasawa et
362: al. 1999), althogh the time
363: scale they observed is much shorter than ours. The peak energy of the
364: Fe line during the flare was around 5 keV and such a profile change can
365: be interpreted as being due to a flare occuring very close to the
366: black hole (Iwasawa et al. 1999).
367:
368: The observed line does not show an obvious asymmetric profile skewed
369: toward lower energies, a signature expected from a relativistic disk
370: (e.g., Fabian et al. 1989) and one which is observed in many Seyfert 1
371: nuclei (Tanaka et al. 1995; Nandra et al. 1997). The line width is
372: evidently rather small; we place an upper limit of $\sigma<0.16$
373: keV. Our data, unfortunately, do not have sufficient signal-to-noise
374: ratio to confirm the presence or absence of disklike kinematic
375: features in NGC 4579. Future measurements of the line profile using
376: higher energy resolution and larger effective areas will be crucial to
377: investigate the origin of the Fe line and the structure of the
378: accretion disk in LLAGNs.
379:
380: \acknowledgments
381:
382: The authors are grateful to all the {\asca} team members. We also
383: thank an anonymous referee for useful comments. YT and KM thank Japan
384: Society for the Promotion of Science for support.
385:
386:
387: \begin{thebibliography}{}
388:
389: \bibitem[]{} Antonucci, R. 1993, \araa, 31, 473
390:
391: \bibitem[]{} Burke, B.~E., Mountain, R.~W., Daniels, P.~J., \& Dolat, V.~S. 1994, IEEE Trans. NS-41, 375
392:
393: \bibitem[]{} Fabian, A.C., Rees, M.J., Stella, L., \& White, N.E. 1989,
394: \mnras, 238, 729 % disk line
395:
396: \bibitem[]{} George, I.~M., \& Fabian, A.~C. 1991, \mnras, 249, 352
397:
398: \bibitem[]{} Ho, L.~C. 1999, \apj, 516, 672 % SED
399:
400: \bibitem[]{} Ishisaki, Y. et al. 1996, \pasj, 48, 237
401:
402: \bibitem[]{} Iyomoto, N., Makishima, K., Fukazawa, Y., Tashiro, M., Ishisaki, Y. 1997, \pasj, 49, 425 % N1365, 1386
403:
404: \bibitem[]{} Iwasawa, K., Fabian, A.~C., Young, A.~J., Inoue, H., \&
405: Matsumoto, C., 1999, \mnras, 306, L19 % MCG-6
406:
407: \bibitem[]{} Kallman, T.~R., \& McCray, R. 1982, \apjs, 50, 263
408:
409: \bibitem[]{} Makishima, K. 1986, in The Physics of Accretion onto Compact objects; eds K.~O. Mason, M.~G. Watson, \& N.~E. White (Berlin: Springer), 249 %Fe
410:
411: \bibitem[]{} Makishima, K. et al. 1996, \pasj, 48, 171 % GIS
412:
413: \bibitem[]{} Murphy, E.M., Lockman, F.J., Laor, A., \& Elvis, M. 1996,
414: \apjs, 105, 369 % NH
415:
416: \bibitem[]{} Nandra, K., George, I.M., Mushotzky, R.F.,
417: Turner, T.J., \& Yaqoob, T. 1997a, \apj, 477, 602 % Sy1 Fe
418:
419: \bibitem[]{} Narayan, R., \& Raymond, J. 1999, \apj, 515, L69
420:
421: \bibitem[]{} Ohashi, T., et al. 1996, \pasj, 48, 157 % GIS
422:
423: \bibitem[]{} Pellegrini, S., Cappi, M., Bassani, L., Malaguti, G., Palumbo, G.~G.~C., \& Persic, M. 2000, \aap, 353, 447 % M81
424:
425: \bibitem[]{} Ptak, A., Serlemitsos, P., Yaqoob, T., \& Mushotzky,
426: R. 1999, \apjs, 120, 179
427:
428: \bibitem[]{} Ptak, A. Mushotzky, R., Yaqoob, T., \& Serlemitsos, P. 2000, \apj, submitted
429:
430: \bibitem[]{} Quataert, E., Di Matteo, T., Narayan, R., \& Ho,
431: L.~C. 1999, \apj, 525, L89
432:
433: \bibitem[]{} Roberts, T.~P., Warwick, R.~S., \& Ohashi, T 1999, \mnras, 304, 52
434:
435: \bibitem[]{} Serlemitsos, P.J., Ptak, A.,\& Yaqoob, T. 1996 in The Physics of LINERs in View of Recent Observations, eds. Eracleous, M. et al., (San Francisco:ASP), 70
436:
437: \bibitem[]{} Tanaka, Y., Inoue, H., \& Holt, S.S. 1994, \pasj, 46, L37
438:
439: \bibitem[]{} Tanaka, Y. et al. 1995, \nat, 375, 659
440:
441: \bibitem[]{} Terashima, Y., Kunieda, H., Misaki, K., Mushotzky, R.F., Ptak, A.F., \& G.A. Reichert 1998a, \apj, 503, 212
442:
443: \bibitem[]{} Terashima, Y., Ptak, A., Fujimoto, R., Itoh, M., Kunieda, H.,
444: Makishima, K ., \& Serlemitsos, P.~J. 1998b, \apj, 496, 210 % M51
445:
446: \bibitem[]{} Terashima, Y., Kunieda, H., \& Misaki, K. 1999, \pasj, 51, 277
447:
448: \bibitem[]{} Tully, R.B. 1988, Nearby Galaxies Catalog, Cambridge University Press
449:
450: \bibitem[]{} Roberts, T.P., Warwick, R.S., \& Ohashi, T. \mnras, 304, 52, 1999
451:
452: \bibitem[]{} Weaver, K.~A., Gelbord, J., \& Yaqoob, T. 2000, \apj, submitted % Sy1 Fe lines
453:
454: \bibitem[]{} Yaqoob, T. et al. 2000, in preparation % cal
455:
456:
457: \newpage
458:
459: \end{thebibliography}
460:
461:
462: \figcaption[]{({\it a}) {\it ASCA} SIS (with circle) and GIS spectra
463: of NGC 4579. The data sets from two observations in 1998 are
464: combined. The solid histograms are the best-fit model but the
465: normalization of the Gaussian component is set to be zero. The dotted
466: and dashed histograms represent the Raymond-Smith and the power-law
467: component, respectively.
468: ({\it b}) Ratio of the data to the best-fit model around Fe
469: lines in the 1998 observation and ({\it c}) 1995 observation. The
470: normalization of the Gaussian component is set to be zero. The crosses
471: with and without filled circle are SIS and GIS data, respectively.
472: The energy scale is not redshift corrected. \label{fig-1} }
473:
474:
475: \figcaption[]{
476: Confidence contours ($\Delta \chi^2$=2.3, 4.6, and 9.1)
477: for the line energy and normalization. Dashed lines are for the 1995
478: observation and solid lines are for the combined spectrum of the two
479: 1998 observations. The fitting model is Raymond-Smith + Power-law +
480: Gaussian. The energy scale is redshift corrected.
481: \label{fig-2}
482: }
483:
484:
485: \figcaption[]{
486: The equivalent width and intensity of the 6.4 keV and 6.7 keV line for
487: the 1995 and 1998 observations. A double Gaussian model with centroid
488: energies fixed at 6.4 keV and 6.7 keV is assumed for the Fe line.
489: \label{fig-3}
490: }
491:
492:
493:
494: \footnotesize
495:
496: %\small
497:
498: \begin{table*}
499: \begin{center}
500: \begin{tabular}{ccccccccc}
501: \tableline \tableline
502: Data & \NH (galactic) & $kT$ & Abundance & \NH (hard component)
503: & $\Gamma$ & Flux (2-10 keV) & $\chi^2$/dof\\
504: & [$10^{20}$~{\pcm}] & [keV] & [solar] &
505: [$10^{20}$~{\pcm}] & & [$10^{-12}${\eps}{\pcm}] &\\
506: \tableline
507: 1995 & 3.1(f) & $0.90^{+0.11}_{-0.05}$ & 0.50(f) & $4.1\pm2.7$ & $1.72\pm0.05$ & 4.3 & 192.4/201 \\
508: 1998(1+2)& 3.1(f) & 0.90(f) & 0.50(f) & $3.8 (<13)$ & $1.81\pm0.06$ & 5.8 & 200.5/225 \\
509: 1998(1) & 3.1(f) & 0.90(f) & 0.50(f) & $14 (<28)$ & $1.86\pm0.09$ & 6.1 & 188.3/207 \\
510: 1998(2) & 3.1(f) & 0.90(f) & 0.50(f) & $3.4 (<13)$ & $1.84\pm0.08$& 5.3 & 201.1/200 \\
511: \tableline
512: \end{tabular}
513: \end{center}
514: \tablenum{1}
515: \caption{Results of spectral fitting to the SIS and GIS spectra of NGC 4579\label{tbl-1}}
516: \tablecomments{The fitting model is Raymond-Smith + Power-law + Gaussian. (f)
517: in the table denotes a frozen parameter. The absorption column for the RS component is
518: assumed to be the Galactic value. The quoted errors
519: are at the 90\% confidence level for one interesting parameter.}
520: \end{table*}
521:
522:
523:
524: %\clearpage
525:
526: \begin{table*}
527: \begin{center}
528: \begin{tabular}{ccccc}
529: \tableline \tableline
530: model & $E_L$ & $\sigma$ & EW & Intensity\\%$\chi^2$/dof\\
531: & [keV] & [keV] & [eV] & [10$^{-5}$photons s$^{-1}$ cm$^{-2}$]\\
532: \tableline
533: Single Gaussian \\
534: %1995 & $6.73^{+0.13}_{-0.12}$ & $0.17^{+0.11}_{-0.12}$ & $490^{+180}_{-190}$ & \\ %192.4/201\\
535: %1998 & $6.39\pm0.09$ & $0.0 (<0.16)$ & $260\pm100$ & \\ %216.2/232\\
536: 1995 & 6.73 (6.61--6.86) & 0.17 (0.05--0.28) & 490 (300-670) & 1.90 (1.16--2.60)\\
537: 1998(1+2)& 6.39 (6.30--6.48) & 0.0 (0--0.16) & 250 (155--355) & 1.39 (0.86--1.97)\\
538: 1998(1)& 6.37 (6.27--6.49) & 0.0 (0--0.19) & 270 (130--410) & 1.54 (0.75--2.32)\\
539: 1998(2)& 6.40 (6.24--6.62) & 0.0 (0--0.28) & 250 (110--445) & 1.22 (0.55--2.27)\\
540: \tableline
541: Double Gaussian \\
542: 1995 & 6.4(f) & 0.0 (f) & 0 (0--200) & 0 (0--0.85)\\
543: & 6.73 (f) & 0.17 (f) & 490 (300-670) & 1.90 (1.16--2.60)\\
544: 1998(1+2)& 6.4(f) & 0.0 (f) & 260 (155--360) & 1.42 (0.85--1.97)\\
545: & 6.73 (f) & 0.17 (f) & 0 (0--110) & 0 (0--0.54)\\
546: \tableline
547: \end{tabular}
548: \end{center}
549: \tablenum{2}
550: \caption{Gaussian fits to the iron K line\label{tbl-2}}
551: \tablecomments{The line energies are at the source
552: rest frame. (f) in the table denotes a frozen parameter.
553: The errors are 90\% confidence range for one interesting parameter.}
554: \end{table*}
555:
556: \end{document}
557: