1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %%% PASJ LaTeX template for draft(body)<2007/01/19>
3: %%%
4: %%% IMPORTANT NOTICE FOR AUTHORS
5: %%% 1. ``\draft'' creates single column and double spaces format.
6: %%% 2. If you comment out ``\draft'', the output will be double column
7: %%% and single space.
8: %%% 3. For cross-references, the use of \label/\ref/\cite and the
9: %%% thebibliography environment is strongly recommended
10: %%% 4. Do NOT use \def/\renewcommand.
11: %%% 5. Do NOT redifine commands provided by PASJ00.cls.
12: %%%
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14: \documentclass{pasj00}
15: \draft
16:
17:
18: \begin{document}
19: \SetRunningHead{S. Katsuda et al.}{Abundance Inhomogeneity in the NE Rim of
20: the Cygnus Loop}
21: \Received{2007/06/12}%{yyyy/mm/dd}
22: \Accepted{2007/08/07}%{yyyy/mm/dd}
23:
24: \title{Abundance Inhomogeneity in the Northeastern Rim of
25: the Cygnus Loop Revealed by Suzaku Observatory}
26:
27: %%% begin:list of authors
28: % Do NOT capitalize all letters in "textsc".
29: \author{Satoru \textsc{Katsuda}, Hiroshi \textsc{Tsunemi}, Hiroyuki
30: \textsc{Uchida}, Emi \textsc{Miyata}, and Norbert \textsc{Nemes}} %
31: % \thanks{Example: Present Address is xxxxxxxxxx}}
32: \affil{Department of Earth and Space Science, Graduate School of
33: Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka
34: 560-0043
35: }\email{katsuda@ess.sci.osaka-u.ac.jp,tsunemi@ess.sci.osaka-u.ac.jp,uchida@ess.sci.osaka-u.ac.jp,miyata@ess.sci.osaka-u.ac.jp,nnemes@ess.sci.osaka-u.ac.jp}
36:
37: \author{Eric D. \textsc{Miller}}
38: \affil{Kavli Institute for Astrophysics and Space Research, Massachusetts
39: Institute of Technology, Cambridge, MA 02139,
40: U.S.A.}\email{milleric@space.mit.edu}
41:
42: \author{Koji \textsc{Mori}}
43: \affil{Department of Applied Physics, Faculty of Engineering,
44: University of Miyazaki, 889-2192, Japan}\email{mori@astro.miyazaki-u.ac.jp}
45:
46: \and
47: \author{John. P. {\sc Hughes}}
48: \affil{Department of Physics and Astronomy, Rutgers University, 136
49: Frelinghhuysen Road, Piscataway, NJ 08854-8019,
50: U.S.A.}\email{jackph@physics.rutgers.edu}
51:
52: %\author{Motohide {\sc Kokubun}}
53: %\affil{Institute of Space and Astronautical Science, Japan Aerospace
54: %Exploration Agency, 3-1-1 Yoshinodai, Sagamihara, Kanagawa, 229-8510, Japan }\email{kokubun@astro.isas.jaxa.jp}
55:
56: %\and
57: %\author{F. Scott. {\sc Porter}}
58: %\affil{NASA Goddard Space Flight Center, Greenbelt, MD 20771,
59: %U.S.A.}\email{Frederick.S.Porter@gsfc.nasa.gov}
60:
61: %%% end:list of authors
62:
63: %%% Please use the following style in case that sorting by
64: %%% affilation is impossible.
65: %
66: % \author{%
67: % D-Firstname \textsc{D-Familyname}\altaffilmark{1}
68: % E-Firstname \textsc{E-Familyname}\altaffilmark{1,2}
69: % and
70: % F-Firstname \textsc{F-Familyname}\altaffilmark{2}}
71: % \altaffiltext{1}{Address of Institute}
72: % \email{ddddd@xxx.xxx.xx.xx}
73: % \email{eeeee@xxx.xxx.xx.xx}
74: % \altaffiltext{2}{Address of Institute}
75:
76: %% `\KeyWords{}' always has to be placed before `\maketitle'.
77: \KeyWords{ISM: abundances --- ISM: individual (Cygnus Loop) --- ISM: supernova remnants --- X-rays: ISM} %Do NOT move this preamble from here!
78:
79: \maketitle
80:
81: \begin{abstract}
82: We present the results of a spatially resolved spectral analysis from
83: four Suzaku observations covering the northeastern rim of the Cygnus
84: Loop. A two-$kT_\mathrm{e}$ non-ionization equilibrium (NEI) model fairly
85: well represents our data, which confirms the NEI condition of the
86: plasma there. The metal abundances are depleted relative
87: to the solar values almost everywhere in our field of view. We find
88: abundance inhomogeneities across the field: the northernmost region
89: (Region~A) has enhanced absolute abundances compared with other regions.
90: In addition, the relative abundances of Mg/O and Fe/O in Region~A are lower
91: than the solar values, while those in the other regions are twice higher
92: than the solar values. As far as we are concerned, neither a circumstellar
93: medium, (nor) fragments of ejecta, nor abundance inhomogeneities of the
94: local interstellar medium around the Cygnus Loop can explain the
95: relatively enhanced abundance in Region~A. This point is left as an
96: open question for future work.
97: \end{abstract}
98:
99:
100: \section{Introduction}
101:
102: The Cygnus Loop is a nearby (540\,pc: Blair et al.\ 2005) proto-typical
103: middle-aged ($\sim$10000 yr) supernova remnant (SNR) located at ($l,
104: b$)=(74$^\circ$, $-$\timeform{8.5D}).
105: The foreground neutral hydrogen column density, $N_\mathrm{H}$, is
106: estimated to be $\sim0.04\times10^{22} \mathrm{cm}^{-2}$
107: (Inoue et al.\ 1980; Kahn et al.\ 1980). The low foreground absorbing
108: material as well as the large apparent size (\timeform{2.5D}
109: $\times$ \timeform{3.5D}: Levenson et al.\ 1997; Aschenbach \& Leahy
110: 1999) and high surface brightness enable us to study the soft X-ray
111: emission from the Cygnus Loop.
112:
113: Miyata et al.\ (1994) observed the northeastern (NE) rim of the Cygnus Loop
114: with ASCA. They found non-equilibrium ionization (NEI)
115: conditions and depleted metal abundances relative to the solar
116: values. The low metal abundances led them to consider that the plasma
117: in the NE-rim of the Cygnus Loop originated from swept-up matter,
118: rather than SN ejecta. Recently, Miyata et al.\
119: (2007) (hereafter, Paper {\scshape I}) observed the same region with Suzaku
120: (Mitsuda et al.\ 2007) and performed spectral analysis from 2$'$ thick annular
121: regions. They confirmed the metal deficiency as
122: well as the NEI conditions there. Furthermore, the extended 0.2--12\,keV
123: energy range of the Suzaku X-ray CCD XIS camera (Koyama et al.\ 2007),
124: combined with its superior energy resolution, allowed them to detect
125: emission lines from highly ionized C and N for the first time from the
126: Cygnus Loop.
127:
128: Using the Suzaku satellite, we observed the NE-rim of
129: the Cygnus Loop in four pointings (NE1--4) during the science
130: working group observing time. The fields of view (FOV) are shown in
131: figure~\ref{fig:hri_image}. In Paper I, we presented the results
132: of an analysis for the NE2~region. We here present the results of the
133: analyses for all four FOV using improved response files.
134:
135: \begin{figure*}
136: \begin{center}
137: \FigureFile(80mm,80mm){fig1.eps}
138: %%% \FigureFile(width,height){filename}
139: \end{center}
140: \caption{ROSAT HRI image of the entire Cygnus Loop. The Suzaku FOV
141: from NE1 to NE4 are shown as white rectangles. }\label{fig:hri_image}
142: \end{figure*}
143:
144:
145: %\newpage
146:
147: \section{Observations and Data Screening}
148:
149: The observations were performed on 2005 November 23, 24, 29, and 30 for
150: NE1 (Obs. ID 500020010), NE2 (Obs. ID 500021010), NE3 (Obs. ID
151: 500022010), and NE4 (Obs. ID 500023010), respectively. We employed
152: revision 1.2 of the cleaned event data, and excluded the time region
153: where the attitude was unstable. Furthermore, we excluded data taken in
154: the low cut-off rigidity $<$\,6\,GV. The net exposure time was 84\,ks
155: for all four observations after screening. We subtracted a blank-sky
156: spectrum obtained from the Lockman Hole, since the observation date of
157: the Lockman Hole (2005 November 14; Obs ID 100046010) was close to that
158: of the Cygnus Loop.
159: Figure~\ref{fig:xis_image} shows the Suzaku XIS1 (back-illuminated CCD;
160: BI CCD) three-color image. For spectrum fitting, we used photons in the
161: energy range of 0.2--3.0\,keV for XIS1 and 0.4--3.0\,keV for XIS0, 2, and 3
162: (front-illuminated CCD; FI CCD).
163:
164: \begin{figure*}
165: \begin{center}
166: \FigureFile(80mm,80mm){fig2.eps}
167: %%% \FigureFile(width,height){filename}
168: \end{center}
169: \caption{Merged Suzaku XIS1 three-color image of the four pointings
170: (Red: 0.31--0.38\,keV band, i.e., C {\scshape VI} K$\alpha$, Green:
171: 0.38--0.46\,keV band, i.e., N {\scshape VI} K$\alpha$,
172: Blue: 0.60--0.69\,keV band, i.e., O {\scshape VIII} K$\alpha$). The
173: data were binned by 8 pixels and smoothed by Gaussian
174: distribution of $\sigma = 25^{\prime\prime}$. The
175: effects of exposure, vignetting, and contamination were corrected. A
176: small white rectangles are the cells where we extracted spectra. We show
177: example spectra from cells~1 and 2 in figure~\ref{fig:ex_spec}. The red
178: polygon identifies Region~A (see text).}
179: \label{fig:xis_image}
180: \end{figure*}
181:
182:
183:
184: \section{Spatially Resolved Spectral Analysis}
185:
186: We divided the entire FOV into 184 cells (shown as small rectangles in
187: Figure~\ref{fig:xis_image}), such that each cell contains
188: 2500--5000 photons for XIS0 to equalize the statistics.
189: We extracted spectra from them and performed spectral
190: analysis. We can investigate the plasma structures along the azimuthal
191: direction as well as the radial direction from this analysis.
192: Since the energy scale was not perfectly calibrated, we manually
193: adjusted the energy scale by shifting the energy within the uncertainty
194: of the calibration ($\pm5$\,eV; Koyama et al.\ 2007), so that
195: we could obtain better fits. In order to generate a response matrix
196: file (RMF) and an ancillary response file (ARF), we employed {\tt
197: xisrmfgen} (Ishisaki et al.\ 2007) and {\tt xissimarfgen}
198: (version 2006-10-26), respectively.
199: The low-energy efficiency of the XIS's shows degradation caused by
200: contaminants accumulated on the optical blocking filter
201: (Koyama et al.\ 2007). This was taken into account when generating
202: the ARF file.
203:
204: Since the analysis in Paper-{\scshape I} already revealed that at least
205: two NEI components with different $kT_\mathrm{e}$ were required to
206: represent the spectra, we applied an absorbed
207: two-$kT_\mathrm{e}$-component NEI model for all spectra (the wabs
208: (Morrison \& McCammon 1983) and VNEI model (NEI version 2.0) in XSPEC
209: v\,11.3.1; e.g., Borkowski et al.\ 2001). The free
210: parameters were $N_\mathrm{H}$; electron temperature, $kT_\mathrm{e}$;
211: ionization timescale, $\tau$; emission measure, EM (EM$=\int
212: n_\mathrm{e}n_\mathrm{H} dl$, where $n_\mathrm{e}$ and $n_\mathrm{H}$
213: are the number densities of electrons and protons, respectively and $l$ is
214: the plasma depth); abundances of C, N, O, Ne, Mg, Si, S, Fe, and
215: Ni. We set the abundance of Ni equal to that of Fe. The other
216: elemental abundances were fixed to the solar values (Anders \&
217: Grevesse 1989).
218: We individually varied $kT_\mathrm{e}$ and EM while other parameters were
219: tied in the two components. We confined the variation of $N_\mathrm{H}$
220: to be 0.01 to 0.06$\times10^{22}\mathrm{cm}^{-2}$ (Inoue et al.\ 1980;
221: Miyata et al.\ 2007). We here refer to this model as a VNEI1 model.
222: It gave us fairly good fits for all
223: spectra (reduced $\chi^2$ ranges from 0.90 to 1.27).
224: The fit statistics are dramatically improved compared to those obtained in
225: Paper {\scshape I} (maximum reduced $\chi^2$ of 2.81).
226: This is mainly due to the fact that the post-launch degradation of the
227: XIS energy resolution is now included in our spectral-response function,
228: which was not possible at the time Paper {\scshape I} was written.
229: Figure~\ref{fig:ex_spec} shows example spectra from cells~1 and 2 in
230: figure~\ref{fig:xis_image} with the best-fit models. The best-fit
231: parameters for the cells are summarized in table~\ref{tab:ex_param} (VNEI1).
232: Maps of the best-fit values are presented in figure~\ref{fig:param}.
233:
234: \begin{table*}
235: %\tabletypesize{\tiny}
236: \begin{center}
237: \caption{Spectral-fit parameters for cells~1 and 2.}\label{tab:ex_param}
238: \begin{tabular}{lccccc}
239: \hline
240: Parameter & cell~1 (VNEI1) & cell~1 (VNEI2) & cell~1 (VPSHOCK)& cell~2
241: (VNEI1) & cell~2 (VNEI2)\\
242: \hline
243: $N_\mathrm{H}$[$\times10^{22}$cm$^{-2}$]\dotfill &
244: 0.024$\pm0.001$&0.023$^{+0.02}_{-0.01}$&0.024$^{+0.003}_{-0.001}$&
245: $<0.021$ &$<0.020$ \\
246: $kT_\mathrm{e1}$[keV] \dotfill & 0.27$\pm$0.01 & 0.27$\pm$0.01&0.26$\pm$0.01&0.39$\pm0.02$&0.39$\pm$0.01\\
247: $kT_\mathrm{e2}$[keV] \dotfill & 0.09$\pm$0.01 & 0.08$^{+0.02}_{-0.04}$& 0.08$\pm$0.01&0.21$\pm0.02$&0.23$\pm$0.02\\
248: C \dotfill& 1.06$\pm$0.08 &1.03$\pm$0.08&1.2$\pm$0.1&0.24$\pm$0.04&0.20$\pm$0.04\\
249: N \dotfill& 1.03$\pm$0.06&1.04$\pm$0.06&1.08$^{+0.1}_{-0.04}$&0.09$\pm$0.02&0.10$\pm$0.02\\
250: O \dotfill& 0.53$\pm$0.02&0.53$\pm$0.02&0.54$\pm$0.02&0.131$\pm$0.004&0.131$\pm$0.004\\
251: Ne \dotfill& 0.84$\pm$0.04&0.84$\pm$0.04&0.91$^{+0.03}_{-0.05}$ &0.29$\pm$0.01&0.28$\pm$0.01\\
252: Mg \dotfill& 0.35$\pm$0.12& 0.35$\pm$0.12&0.39$^{+0.13}_{-0.16}$ &0.20$\pm$0.03&0.20$\pm$0.03\\
253: Si \dotfill& 1.9$\pm$0.2& 1.8$\pm$0.2&1.9$^{+0.14}_{-0.15}$&0.24$\pm$0.04&0.24$\pm$0.04\\
254: S \dotfill& $<$0.5&$<$0.5&$<$0.4&0.14$\pm$0.10&0.16$\pm$0.12\\
255: Fe(=Ni) \dotfill&0.52$\pm$0.04&0.52$\pm$0.04&0.60$^{+0.06}_{-0.03}$ &0.22$\pm$0.01&0.21$\pm$0.01\\
256: log$(\tau /\mathrm{cm}^{-3}\,\mathrm{s})$\dotfill
257: &10.65$\pm0.08$ &\dotfill&\dotfill&11.21$^{+0.09}_{-0.10}$&\dotfill\\
258: log$(\tau_1 /\mathrm{cm}^{-3}\,\mathrm{s})$\dotfill
259: &\dotfill &10.63$\pm$0.04&\dotfill&\dotfill&11.24$^{+0.05}_{-0.06}$\\
260: log$(\tau_2 /\mathrm{cm}^{-3}\,\mathrm{s})$\dotfill
261: &\dotfill&10.6$<$&\dotfill&\dotfill&11.0$\pm$0.1\\
262: log$(\tau_\mathrm{lower} /\mathrm{cm}^{-3}\,\mathrm{s})$\dotfill
263: &\dotfill&\dotfill&0 (fixed)&\dotfill&\dotfill\\
264: log$(\tau_\mathrm{upper} /\mathrm{cm}^{-3}\,\mathrm{s})$\dotfill
265: &\dotfill&\dotfill&11.00$\pm$0.03&\dotfill&\dotfill\\
266: EM$_1$[$\times10^{19}$ cm$^{-5}$]\dotfill& 0.48$\pm0.01$& 0.48$\pm0.01$&0.52$\pm$0.02 &1.4$\pm$0.04&1.39$\pm$0.04\\
267: EM$_2$[$\times10^{19}$ cm$^{-5}$]\dotfill& 0.9$\pm$0.2& 0.9$\pm$0.2
268: &0.5$\pm$0.2&2.4$\pm$0.1 &2.4$\pm$0.1\\
269: \hline
270: $\chi^2$/d.o.f. \dotfill & 611/508& 611/507 & 619/508 &706/699 &706/698\\
271: \hline
272: \\[-8pt]
273: \multicolumn{3}{@{}l@{}}{\hbox to 0pt{\parbox{140mm}{\footnotesize
274: \par\noindent
275: \footnotemark[$*$]Other elements are fixed to those of solar values.\\
276: The values of abundances are multiples of solar value.\\ The errors
277: are in the range $\Delta\,\chi^2\,<\,2.7$ on one parameter. \\
278: The subscript 1 denotes the high temperature component \\
279: while 2 denotes the low temperature component.
280: \par\noindent
281: %\footnotemark[$\dagger$]EM denotes the emission measure $\int
282: % n_\mathrm{e} n_\mathrm{H} dl$.
283: }\hss}}
284:
285: \end{tabular}
286: \end{center}
287: \end{table*}
288:
289:
290: \begin{figure*}
291: \begin{center}
292: \FigureFile(80mm,80mm){fig3a.eps}
293: \FigureFile(80mm,80mm){fig3b.eps}
294: %%% \FigureFile(width,height){filename}
295: \end{center}
296: \caption{Left: X-ray spectra extracted from cell~1 in
297: figure~\ref{fig:xis_image}. The best-fit curves are shown with solid
298: lines for the four XIS's. The contribution of each component is
299: shown by dotted lines only for XIS1. The dashed line represents
300: the low-temperature component while the dotted line represents the
301: high-temperature component. The lower panels show the residuals.
302: Right: Same as left, but for cell~2.}\label{fig:ex_spec}
303: \end{figure*}
304:
305:
306: \section{Results}
307:
308: \begin{figure*}
309: \begin{center}
310: \FigureFile(160mm,100mm){fig4.eps}
311: %%% \FigureFile(width,height){filename}
312: \end{center}
313: \caption{Maps of the best-fit parameters. The units are $10^{22}
314: \mathrm{cm}^{-2}$ for $N_\mathrm{H}$, keV for $kT_\mathrm{e1}$ and
315: $kT_\mathrm{e2}$, cm$^{-3}$\,s for $\tau$,
316: $10^{19} \mathrm{cm}^{-5}$ for EM$_\mathrm{1}$ and
317: EM$_\mathrm{2}$, and solar values for abundances.}
318: \label{fig:param}
319: \end{figure*}
320:
321: \subsection{Abundances}
322:
323: The abundances of O, Ne, Mg, and Fe are consistent with those in
324: Paper-{\scshape I}. On the other hand, C and N in our analysis are
325: systematically higher, while Si and S are systematically lower than those
326: in Paper {\scshape I}. There are two main reasons that can explain
327: this discrepancy. Firstly, we allowed the abundance of S to vary freely in
328: our models, while in Paper {\scshape I} it was fixed to the solar value.
329: Since strong emission lines of S L fall around the C K
330: ($\sim$0.35\,keV) and N K ($\sim$0.5\,keV) emission lines, the
331: abundances of C and N are affected by that of S. In our
332: spectral analysis, the typical abundance of S is $\sim$0.2-times the
333: solar value, resulting in higher abundances
334: of C and N than those in Paper {\scshape I}. Secondly, we
335: included data in the Si K band (1.7--1.9\,keV), while in Paper {\scshape
336: I} excluded it due to a calibration uncertainty.
337: Since the Si K line was not so strong in our data, we found that the
338: calibration uncertainty did not play an important role in our results.
339: Due to limited available atomic data for these transitions, the
340: emissivities of the Si L lines in our models are expected to contain large
341: uncertainty relative to those of the Si K lines. Therefore, we believe that
342: our results are more reliable than those in Paper {\scshape I} in
343: which the abundance of Si was determined by the emission lines of Si L.
344:
345: \begin{figure*}
346: \begin{center}
347: \FigureFile(160mm,100mm){fig5.eps}
348: %%% \FigureFile(width,height){filename}
349: \end{center}
350: \caption{Each elemental abundance or ionization timescale versus O is
351: plotted for the 184 cells. Red data points come from Region~A, while
352: black points come from the remaining cells. The dashed line from the
353: upper right to the lower left represents the solar
354: ratio.}\label{fig:correlation}
355: \end{figure*}
356:
357: We obtained abundances within 184 spatial cells, as shown in
358: figure~\ref{fig:param}. In Figures~\ref{fig:correlation}
359: (a)--\ref{fig:correlation}(g), we show plots of the correlations
360: between each elemental abundance versus the O abundance for all of these
361: regions. The absolute abundances (relative to H) for all the elements
362: vary significantly from cell to cell. The relative abundances of
363: C, Ne, Si, and S to O are similar in all the cells, while
364: those of Mg and Fe are divided into two groups; one shows lower values
365: than the solar value (red crosses in figure~\ref{fig:correlation}), while the
366: other shows about two-times higher values than the solar value (black
367: crosses in figure~\ref{fig:correlation}). We indicate the regions of red
368: crosses in a red polygon in figure~\ref{fig:xis_image}. We found that those
369: regions were concentrated north of our FOV, where the absolute
370: abundances are relatively higher than those in other regions.
371: Hereafter we refer to the region outlined by the red polygon in
372: figure~\ref{fig:xis_image} as Region A. We summarize mean abundances,
373: mean 90\% errors, and standard deviations of the abundances for both
374: Region~A and the other regions in table~\ref{tab:abund}.
375:
376: In our model, the ionization timescale is imposed to be the same
377: for the two components for simplicity of the model, just as employed in
378: Paper {\scshape I}. In this paragraph, we consider whether or not
379: the introduction of separate ionization timescales
380: for the two components changes the results. We fitted the spectra
381: in cells~1 and 2 with a two-component VNEI model whose free parameters
382: were the same as those used in the VNEI1 model, but used separate ionization
383: timescales for the two components that we call the VNEI2
384: model. The best-fit parameters and fit statistics are summarized in
385: table~\ref{tab:ex_param} (VNEI2).
386: We found that all of the best-fit parameters were consistent with
387: those obtained with the VNEI1 model. We should note that the
388: results are quite similar to those from the VNEI1 model.
389: We checked the value of $\chi^2$ as a function of $\tau_2$ over a large
390: range of values (i.e., $2\times10^{10}$--$1\times10^{12}$ cm$^{-3}$\,s),
391: and confirmed that the results were not due to secondary local minima.
392: We thus conclude that the obtained abundances are not affected by
393: introducing a separate ionization timescale for the two
394: components.
395:
396: Figure~\ref{fig:param} clearly shows that Region~A is
397: the location of a significantly different best-fit value for the
398: ionization timescale. We plot the correlation between the ionization timescale
399: and the O abundance in figure~\ref{fig:correlation} (h). It shows
400: anti-correlation, which causes us some worry about the derived abundances,
401: since the NEI models that we employ are really just very
402: simplistic approximations to the true physical conditions under which
403: these plasmas emit. Figure~\ref{fig:param} also indicates that Region~A
404: has a rapidly changing ionization state throughout: from $\tau = 0$
405: at the shock front to $\tau =~\sim10^{11}\,\mathrm{cm^{-3}\,s}$.
406: Our models, however, assume a single ionization timescale (VNEI).
407: In this context, we fitted the spectrum in cell~1 (which is within
408: Region~A) with the VPSHOCK model (e.g., Borkowski et al.\ 2001) in
409: XSPEC, which assumes a constant temperature and a distribution of the
410: ionization timescale, in which we take all ionization timescales up to a
411: fitted maximum value, starting from zero. We employed a
412: two-temperature VPSHOCK model
413: with the same number of free parameters as in the case of the VNEI model.
414: The best-fit values and fit statistics are summarized in
415: table~\ref{tab:ex_param}. We find that the abundances are almost
416: equal between the VNEI model and the VPSHOCK model, which supports the
417: idea that the abundances in Region~A are really different from those in
418: the rest of our FOV.
419:
420:
421: \begin{table*}
422: %\tabletypesize{\tiny}
423: \caption{Mean elemental abundances in Region~A and the rest of the
424: region.$^{\ast}$}\label{tab:abund}
425: \begin{center}
426: \begin{tabular}{lcc}
427: \hline
428: Parameters & Region~A & Excluding Region~A\\
429: \hline
430: C \dotfill&0.74$^{+0.06}_{-0.08}$ (0.26)& 0.36$^{+0.07}_{-0.08}$ (0.18)\\
431: N \dotfill&0.66$\pm$0.05 (0.27)& 0.17$\pm$0.02 (0.08)\\
432: O \dotfill&0.38$\pm$0.02 (0.13)& 0.16$\pm$0.01 (0.05)\\
433: Ne \dotfill&0.63$\pm$0.03 (0.22) & 0.31$\pm$0.02 (0.09)\\
434: Mg \dotfill&0.32$\pm$0.08 (0.12) & 0.24$\pm$0.04 (0.08)\\
435: Si \dotfill&0.9$\pm$0.1 (0.5) & 0.41 $\pm$0.06 (0.18)\\
436: S \dotfill& 0.2$\pm$0.2 (0.1)& 0.2$^{+0.2}_{-0.1}$ (0.2)\\
437: Fe(=Ni) \dotfill&0.34$\pm$0.02 (0.12) &0.26$\pm$0.01 (0.07)\\
438: \hline
439: & & \\[-8pt]
440: \multicolumn{3}{@{}l@{}}{\hbox to 0pt{\parbox{140mm}{\footnotesize
441: \par\noindent
442: \footnotemark[$*$]The values in brackets represent the standard
443: deviations.\\ Errors quoted are mean values for each cell.
444: \par\noindent
445: }\hss}}
446:
447: \end{tabular}
448: \end{center}
449: \end{table*}
450:
451:
452: \subsection{$kT_\mathrm{e}$, $\tau$, $N_\mathrm{H}$, and EM}
453:
454: In the NE2~region, the values of both $kT_\mathrm{e1}$ and $kT_\mathrm{e2}$
455: increase from the outermost cells toward the innermost cells.
456: This trend is consistent with previous X-ray observations of the
457: relevant regions obtained with ASCA (Miyata et al.\ 1994;
458: Miyata \& Tsunemi 1999), XMM-Newton (Katsuda \& Tsunemi 2007; Tsunemi
459: et al.\ 2007), and Suzaku (Miyata et al.\ 2007).
460: The values are also quantitatively consistent with those in
461: Paper {\scshape I}. Also, we confirmed that the ionization state
462: is far from the CIE condition in the NE2~region, and found that
463: the ionization state is in the NEI condition everywhere in the
464: NE1--4~regions.
465: The ionization states in the outermost cells in the NE3 and 4~regions
466: (which correspond to Region~A) are relatively lower than those in
467: the other regions. The EMs for the hot component in the responsible
468: regions are relatively lower than those in the other regions, suggesting
469: low electron densities there. Since almost all of the emission lines come
470: from the hot component (see, figure~\ref{fig:ex_spec} left), the
471: obtained ionization states represent those for the hot component.
472: Therefore, the relatively low ionization states are likely due
473: to the low density.
474: We found that the $N_\mathrm{H}$ for almost all regions is
475: around 0.02$\times10^{22}\mathrm{cm^{-2}}$, while we see about a
476: twice-enhanced column density in the NE4~region.
477:
478: To check the significance level of the observed variations for
479: $kT_\mathrm{e}$, $\tau$, $N_\mathrm{H}$, and EM, and also to look for
480: possible correlations among those parameters, we give plots of the
481: correlations between each parameter versus $kT_\mathrm{e1}$ in
482: figure~\ref{fig:correlation2}. We found that all the parameters were
483: sufficiently constrained to confirm the observed variations in our FOV. We
484: also found that the ionization states in Region~A are significantly
485: lower than those in the other region. We could not find any
486: significant correlations among those parameters, although the
487: distribution of $kT_\mathrm{e1}$ is clustered around
488: $\sim$0.28\,keV and $\sim$0.35\,keV. As shown in figure~\ref{fig:param},
489: the low- and high-temperature clusters are generally located in the
490: outer and inner regions of the remnant, respectively, which is at least
491: qualitatively consistent with what we expect from Sedov-phase SNRs.
492:
493: \begin{figure*}
494: \begin{center}
495: \FigureFile(160mm,100mm){fig6.eps}
496: %%% \FigureFile(width,height){filename}
497: \end{center}
498: \caption{$N_\mathrm{H}$, $kT_\mathrm{e2}$, log($\tau$), EM$_2$, and
499: EM$_1$ versus $kT_\mathrm{e1}$ are plotted for the 184 cells. Red data
500: points come from Region~A while black points come from the remaining
501: cells.}\label{fig:correlation2}
502: \end{figure*}
503:
504:
505: \section{Discussion and Conclusions}
506:
507: We observed the northeastern rim of the Cygnus Loop with the Suzaku
508: observatory in four pointings. We then divided the FOV into 184 cells
509: and performed a spatially resolved spectral analysis. Following the
510: analysis in Paper {\scshape I} in which the authors concluded that
511: there was a multi-temperature plasma along the line of sight, we
512: applied a two-component NEI model with different $kT_\mathrm{e}$
513: values for all of the spectra.
514:
515: Assuming a spherically-symmetric uniform emitting region, and that the
516: two temperature components are in pressure equilibrium, we can estimate
517: the plasma depth for the two components in each cell. We calculated the
518: total masses of the low and high-temperature components in Region~A and the
519: rest of the region, respectively, to be $\sim$0.7\,M$_\odot$ and
520: $\sim$6\,M$_\odot$. In these calculations, we assumed that
521: $n_\mathrm{e} = n_\mathrm{H}$ and the volume filling factor was unity.
522:
523: The relative abundances of C, Ne, Si, and S to O are almost constant in
524: our FOV, while those of Mg and Fe are divided into two groups: one
525: shows lower values than the solar value, while the other shows
526: two-times higher values than the solar value.
527: Regions with low relative abundances are concentrated into the very
528: northernmost cells (Region~A). The absolute abundances in
529: Region~A turned out to be relatively higher than those in the rest of the
530: region.
531:
532: The low abundances in regions other than Region~A confirm the
533: absence of SN ejecta contamination at the northeast rim, and
534: argues for a swept-up origin. Depleted abundances at the rim
535: of the Cygnus Loop have now been reported by several X-ray
536: studies. For example, an ASCA observation of the NE rim (which overlaps
537: the NE2~region here) revealed the abundance of O to be 0.2-times the
538: solar value (Miyata \& Tsunemi 1999). A Chandra observation of the
539: southwestern rim showed the O-group abundance to be 0.22-times the
540: solar value (Leahy 2004; in his spectral
541: analysis, he fixed the abundances of C and N to be relatively the same
542: as O (O-group)). Low abundances appear to be a common result of X-ray
543: spectral analysis of the rim of the Cygnus Loop, although an adequate
544: explanation for this result has yet to be proposed.
545:
546: The abundances in Region~A are higher than those in the rest of the
547: region by factors of $\sim$2.1 (C), $\sim$3.9 (N), $\sim$2.4 (O),
548: $\sim$2.1 (Ne), $\sim$1.3 (Mg), $\sim$2.4 (Si), $\sim$1.2 (S), and
549: $\sim$1.3 (Fe). Since there is evidence in many SNe that the
550: circumstellar medium (CSM) frequently shows enhanced abundance ratios of
551: N/C and N/O (Fransson et al.\ 2005; Chevalier 2005) relative to the
552: solar values as a result of CNO processing in progenitor stars, the
553: strongly enhanced abundance of N in Region~A
554: relative to the rest of the region may lead us to consider that CSM
555: contamination is evident. However, theoretical nucleosynthetic
556: calculations show that the CSM is only rich in N
557: abundance compared with the initial composition of a progenitor
558: star (e.g., \cite{Rauscher2002}). This cannot fully explain that all of
559: the metal abundances in Region~A are enhanced relative to those in the
560: rest of the region.
561: Can fragments of ejecta from the SN explosion explain the enhanced
562: metal abundances? Fragments of ejecta observed in many SNRs (e.g., Cas A:
563: Fesen et al.\ 2001, Vela: Aschenbach et al.\ 1995, Tycho:
564: Decourchelle et al.\ 2001) commonly show knotty shapes or head-tail
565: structures. There is no such structure in Region~A, showing no
566: indication of fragments of ejecta. We next consider the possibility
567: of abundance inhomogeneities of the local interstellar medium in the
568: vicinity of the Cygnus Loop. Observations of an
569: optically thin 1356\,$\AA$ resonance line and H {\scshape I} Ly$\alpha$
570: in 13 sight lines showed remarkable homogeneity of the interstellar
571: gas-phase O/H ratio at a level of $\sim$5\% within
572: about 500\,pc of the Sun (Meyer et al.\ 1998). More recent observations
573: for 36 sight lines confirmed the homogeneity of O/H abundance ratio and
574: revealed that the ratio is uniform within 800\,pc of the Sun
575: (Cartledge et al.\ 2004). Therefore, it is difficult to explain such a strong
576: variation in the abundances at these very small scales by abundance
577: inhomogeneities of the local interstellar medium. The nature of the
578: observed abundance inhomogeneity is left as an open question for future
579: work.
580:
581: %It is well known that there is metallicity
582: %gradients in our galaxy. Assuming that the Cygnus Loop is 8.5\,kpc away
583: %from the galactic center, we calculated the metal abundances of N, O,
584: %and S to be 0.5$\sim$0.8, 0.3$\sim$0.6, and 0.3$\sim$1, respectively
585: %based on table~7 in Rudolph et al.\ (2006). Then, we note that the
586: %abundances of N, O, and S in Region~A are consistent with those
587: %expected by metallicity gradients in our galaxy.
588:
589: \bigskip
590:
591: This work is partly supported by a Grant-in-Aid for Scientific
592: Research by the Ministry of Education, Culture, Sports, Science and
593: Technology (16002004). This study is also carried out as part of
594: the 21st Century COE Program, \lq{\it Towards a new basic science:
595: depth and synthesis}\rq. S. K. is supported by a JSPS Research Fellowship
596: for Young Scientists.
597:
598: %\newpage
599: %\newpage
600:
601: %\newpage
602: %\newpage
603: %\newpage
604:
605:
606: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
607:
608: %%%
609: % See the manual for the detail.
610: %%%
611: \begin{thebibliography}{}
612: % Journals(e.g. A\&A,ApJ,AJ,NMRAS,PASP ...)
613: % Authors, Year, Journal, Vol#, Page#
614: % Journal Title Abbreviation >> http://www.asj.or.jp/pasj/Jabb.html
615: %\bibitem[Aauthor et al.(2001)]{}
616: % Aauthor, A., Bauthor, B., Cauthor, C.\ 2001, PASJ, vol, page
617: % Books
618: \bibitem[Anders \& Grevesse 1989]{Anders1989}
619: Anders, E., \& Grevesse, N. 1989, Geochem. Cosmochim. Acta, 53, 197
620: \bibitem[Aschenbach et al.\ 1995]{Aschenbach1995}
621: Aschenbach, B., Egger, R., \& Trumper, J. 1995, Nature, 373, 587
622: \bibitem[Aschenbach \& Leahy 1999]{Aschenbach1999}
623: Aschenbach, B. \& Leahy, D. A. 1999, A\&A, 341, 602
624: %\bibitem[Blair et al.\ 1999]{Blair1999}
625: % Blair, W. P., Sankrit, R., \& Raymond, J. C., and
626: % Long. K. S. 1999, AJ, 118, 942
627: \bibitem[Blair et al.\ 2005]{Blair2005}
628: Blair, W. P., Sankrit, R., \& Raymond, J. C. 2005, AJ, 129, 2268
629: %\bibitem[Borkowski et al.\ 1994]{Borkowski1994}
630: % Borkowski, K. J., Sarazin, C. L., \& Blondin, J. M. 1994, ApJ,
631: % 429, 710
632: \bibitem[Borkowski et al.\ 2001]{Borkowski2001}
633: Borkowski, K. J., Lyerly, W. J., \& Reynolds, S. P. 2001,
634: ApJ, 548, 820
635: \bibitem[Cartledge et al.\ 2004]{Cartledge2004}
636: Cartledge, S. I. B., Lauroesch, J. T., Meyer, D. M., \& Sofia,
637: U. J. 2004, ApJ, 613, 1037
638: \bibitem[Chevalier 2005]{Chevalier2005}
639: Chevalier, R. A. 2005, ApJ, 619, 839
640: \bibitem[Decourchelle et al.\ 2001]{Decourchelle2001}
641: Decourchelle, A. et al.\ 2001, A\&A, 365, L218
642: %\bibitem[Dopita et al.\ 1977]{Dopita1977}
643: % Dopita, M. A., Mathewson, D. S., and Ford, V. L. 1977 ApJ, 214, 179
644: \bibitem[Fesen 2001]{Fesen2001}
645: Fesen, R. A.\ 2001, ApJS, 133, 161
646: \bibitem[Fransson et al.\ 2005]{Fransson2005}
647: Fransson, C. et al.\ 2005, ApJ, 622, 991
648: %\bibitem[Ghavamian et al.\ 2001]{Ghavamian2001}
649: % Ghavamian, P., Raymond, John., Smith, R. C., \& Hartigan,
650: % P. 2001, ApJ, 547, L995
651: %\bibitem[Hamilton et al.\ 1983]{Hamilton1983}
652: % Hamilton, A. J. S., Chevalier, R. A., \& Sarazin, C. L. 1983,
653: % ApJS, 51, 115
654: %\bibitem[Hayashida et al.\ 2007]{Hayashida2007}
655: % Hayashida et al.\ 2007, this issue
656: %\bibitem[Hester et al.\ 1994]{Hester1994}
657: % Hester, J. J., Raymond, J. C., \& Blair, W. P. 1994, ApJ, 420, 721
658: \bibitem[Ishisaki et al.\ 2007]{Ishisaki2007}
659: Ishisaki, Y., et al.\ 2007, PASJ, 59, S113
660: \bibitem[Inoue et al.\ 1980]{Inoue1980}
661: Inoue, H., Koyama, K., Matsuoka, M., Ohashi, T., Tanaka, Y.,
662: Tsunemi, H. 1980 ApJ, 238, 886
663: \bibitem[Kahn et al.\ 1980]{Kahn1980}
664: Kahn, S. M., Charles, P. A., Bowyer, S., \& Blissett, R. J. 1980,
665: ApJ, 242, L19
666: \bibitem[Katsuda \& Tsunemi 2007]{Katsuda2007}
667: Katsuda, S., \& Tsunemi, H. 2007, Adv. Space Res. in press
668: \bibitem[Koyama et al.\ 2007]{Koyama2007}
669: Koyama, K., et al. 2007, PASJ, 59S, 221
670: %\bibitem[Ku et al.\ 1984]{Ku1984}
671: % Ku, W. H.-M., Kahn, S. M., Pisarki, R., \& Long, K. S. 1984,
672: % ApJ, 278, 615
673: \bibitem[Levenson et al.\ 1997]{Levenson1997}
674: Levenson, N. A. et al. 1997, ApJ, 484, 304
675: %\bibitem[Levenson et al.\ 1998]{Levenson1998}
676: % Levenson, N. A., Graham, J. R., Keller, L. D., \& Richter,
677: % M. J. 2002, ApJS, 118, 541
678: %\bibitem[Levenson et al.\ 1999]{Levenson1999}
679: % Levenson, N. A., Graham, J. R., and Snowden, S. L. 1999, ApJ, 526, 874
680: \bibitem[Leahy 2004]{Leahy2004}
681: Leahy, D. A. 2004, MNRAS, 351, 385
682: %\bibitem[Liedahl et al.\ 1995]{Liedahl1995}
683: % Liedahl, D. A., Osterheld, A. L., \& Goldstein, W. H. 1995,
684: % ApJ, 438, L115
685: %\bibitem[McCray \& Snow 1979]{McCray1979}
686: % McCray, R. \& Snow, T. P., Jr. 1979, ARA\&A, 17, 213
687: \bibitem[Meyer et al.\ 1998]{Meyer1998}
688: Meyer, D. M., Jura, M., \& Cardelli, J. A. 1998, ApJ, 493, 222
689: \bibitem[Miyata et al.\ 1994]{Miyata1994}
690: Miyata, E., Tsunemi, H., Pisarski, R., \& Kissel, S. E. 1994,
691: PASJ, 46, L101
692: %\bibitem[Miyata et al.\ 1998]{Miyata1998}
693: % Miyata, E., Tsunemi, H., Kohmura, T., Suzuki, S., and Kumagai,
694: % S. 1998, PASJ, 50, 257
695: \bibitem[Miyata et al.\ 1999]{Miyata1999}
696: Miyata, E., \& Tsunemi, H. 1999, ApJ, 525, 305
697: \bibitem[Miyata et al.\ 2007]{Miyata2007}
698: Miyata, E., Katsuda, S., Tsunemi, H., Hughes, J. P., Kokubun,
699: M., \& Porter, F. S. 2007, PASJ, 59S, 163 (Paper {\scshape I})
700: \bibitem[Mitsuda et al.\ 2007]{Mitsuda2007}
701: Mitsuda, K., et al. 2007, PASJ, 59, S1
702: \bibitem[Morrison \& McCammon 1983]{Morrison1983}
703: Morrison, R., \& McCammon, D. 1983, ApJ, 270, 119
704: %\bibitem[Raymond et al.\ 2003]{Raymond2003}
705: % Raymond, J. C., Ghavamian, P., Sankrit, R., Blair, W. P., and
706: % Curiel, S. 2003, ApJ, 584, 770
707: \bibitem[Rauscher et al.\ 2002]{Rauscher2002}
708: Rauscher, T., Heger, A., Hoffman, R. D., \& Woosley, S. E. 2002,
709: ApJ, 576, 323
710: %\bibitem[Rudolph et al.\ 2006]{Rudolph2006}
711: % Rudolph, A. L., Fich, M., Bell, G. R., Norsen, T., \& Simpson,
712: % J. P. 2006, ApJS, 162, 346
713: %\bibitem[Serlemitsos et al.\ 2007]{Serlemitsos2007}
714: % Serlemitsos, P. J., et al. 2007, PASJ, 59, S9
715: %\bibitem[Tsunemi et al.\ 1999]{Tsunemi1999}
716: % Tsunemi, H., Miyata, E., \& Aschenbach, B. 1999, PASJ, 51, 711
717: \bibitem[Tsunemi et al.\ 2007]{Tsunemi2007}
718: Tsunemi, H., Katsuda, S., Nemes, N., \& Miller E. D., ApJ in press
719: %\bibitem[]{}
720: %
721: %\bibitem[]{}
722: %
723: \end{thebibliography}
724:
725: \end{document}