1: \documentclass[12pt,preprint]{aastex}
2: %\documentclass[preprint2]{emulateapj}
3:
4: % v1.0 09/24/06 K.Mori 1st circulation to all the collaborators
5:
6: \newcommand{\vdag}{(v)^\dagger}
7: \newcommand{\myemail}{katsuda@ess.sci.osaka-u.ac.jp}
8:
9:
10:
11: %% You can insert a short comment on the title page using the command below.
12:
13: %\slugcomment{Not to appear in Nonlearned J., 45.}
14:
15:
16: \shorttitle{X-Ray--Emitting Ejecta Knots in Puppis~A}
17: \shortauthors{Katsuda et al.}
18:
19:
20: \begin{document}
21:
22:
23: \title{Discovery of Fast-Moving X-Ray--Emitting Ejecta Knots in the
24: Oxygen-Rich Supernova Remnant Puppis~A}
25:
26:
27: \author{S. Katsuda\altaffilmark{1}, K. Mori\altaffilmark{2},
28: H. Tsunemi\altaffilmark{1}, S. Park\altaffilmark{3},
29: U. Hwang\altaffilmark{4,5}, D. N. Burrows\altaffilmark{3},
30: J. P. Hughes\altaffilmark{6}, and P. O. Slane\altaffilmark{7}}
31:
32:
33: \altaffiltext{1}{Department of Earth and Space Science, Graduate School
34: of Science, Osaka University,\\ 1-1 Machikaneyama, Toyonaka, Osaka,
35: 60-0043, Japan; katsuda@ess.sci.osaka-u.ac.jp}
36:
37: \altaffiltext{2}{Department of Applied Physics, Faculty of Engineering,
38: University of Miyazaki, 1-1 Gakuen Kibana-dai Nishi, Miyazaki, 889-2192,
39: Japan}
40:
41: \altaffiltext{3}{Department of Astronomy and Astrophysics, Pennsylvania
42: State University, 525, Davey Laboratory, University Park, PA 16802
43: }
44: \altaffiltext{4}{NASA Goddard Space Flight Center, Code 662, Greenbelt
45: MD 20771
46: }
47: \altaffiltext{5}{Department of Physics and Astronomy, The Johns Hopkins
48: University, 3400 Charles Street, Baltimore, MD 21218
49: }
50: \altaffiltext{6}{Department of Physics and Astronomy, Rutgers
51: University, 136 Frelinghurysen Road, Piscataway, NJ 08854-8019
52: }
53: \altaffiltext{7}{Harvard-Smithsonian Center for Astrophysics, 60 Garden
54: Street, Cambridge, MA 02138, USA
55: }
56:
57:
58: \begin{abstract}
59:
60: We report on the discovery of fast-moving X-ray--emitting ejecta knots
61: in the Galactic Oxygen-rich supernova remnant Puppis~A from {\it
62: XMM-Newton} observations. We find an X-ray knotty feature positionally
63: coincident with an O-rich fast-moving optical filament with
64: blue-shifted line emission located in the northeast of Puppis~A. We
65: extract spectra from northern and southern regions of the
66: feature. Applying a one-component non-equilibrium ionization model
67: for the two spectra, we find high metal abundances relative to the
68: solar values in both spectra. This fact clearly shows that the
69: feature originates from metal-rich ejecta. In addition, we find that
70: line emission in the two regions is blue-shifted. The Doppler
71: velocities derived in the two regions are different with each other,
72: suggesting that the knotty feature consists of two knots that
73: are close to each other along the line of sight. Since fast-moving
74: O-rich optical knots/filaments are believed to be recoiled metal-rich
75: ejecta, expelled to the opposite direction against the high-velocity
76: central compact object, we propose that the ejecta knots disclosed
77: here are also part of the recoiled material.
78:
79:
80: \end{abstract}
81: \keywords{ISM: abundances --- ISM: individual (Puppis~A) --- supernova remnants
82: --- X-rays: ISM}
83:
84:
85: \section{Introduction}
86:
87: Puppis~A is categorized as an ``oxygen-rich'' supernova remnant (SNR)
88: based on optical spectroscopy (Winkler \& Kirshner 1985). The
89: group of O-rich remnants contains only two
90: other members in our Galaxy, Cassiopeia~A (Chevalier \& Kirshner 1979)
91: and G292.0+1.8 (Goss et al.\ 1979), and a few in the Magellanic Clouds
92: (Blair et al.\ 2000 and references there in). At optical wavelengths, these
93: SNRs show fast-moving metal-rich ejecta knots ($v >
94: 1000\,\mathrm{km\,sec^{-1}}$) which are typically enriched in O and
95: Ne. High-Z elements, like Ar, Ca, and Fe, are dominantly
96: generated in a Type-{\scshape I}a supernova (Nomoto et al.\
97: 1984), while low-Z elements, like O, Ne, and Mg, are found in
98: core-collapse SN debris (Thielemann et al.\ 1996). These facts
99: suggest that O-rich SNRs are core-collapse origin. The detection of
100: metal-rich ejecta from these SNRs thus provides us with an opportunity
101: to make a direct test of core-collapse SN nucleosynthesis models.
102:
103: O-rich fast-moving optical knots (hereafter, OFMKs) in Puppis~A, whose
104: proper motions are all consistent with
105: undecerelated expansion from a common center (Winkler \& Kirshner
106: 1985; Winkler et al.\ 1988), have been found only in the northeastern
107: quadrant, suggesting asymmetric mass ejection during SN explosion
108: which produced Puppis~A. On the other hand, proper motion of a
109: central compact object (CCO), RX J0822-4300, associated with the
110: Puppis~A SNR was recently measured to be directed toward the southwest
111: with a high velocity of $\sim$1600\,km\,sec$^{-1}$ (Winkler \& Petre
112: 2007). The CCO is considered to be kicked by asymmetric mass ejection
113: during the SN explosion according to momentum conservation, from which
114: we can study SN explosion mechanisms (e.g., Scheck et al.\ 2006).
115:
116: The so-called ``omega'' filament (Winkler \& Kirshner 1985) is one of
117: the best studied OFMKs in Puppis~A.
118: Winkler \& Kirshner (1985) observed the ``omega'' filament and found
119: that the spectrum was dominated by O lines, with rather weak Balmer
120: lines. The mass ratio of O to H was estimated to be 30, i.e., $\sim$2000
121: times the solar value. Furthermore, they found that the O lines were
122: blue-shifted by 1500$\pm100$\,km\,sec$^{-1}$. These facts led
123: them to consider that the ``omega'' filament was nearly pure O ejecta
124: from the core of the SN progenitor that had remained more or less
125: intact.
126:
127: In X-ray wavelengths, Puppis~A generally showed sub-solar metal
128: abundances, suggesting that the X-ray emission from Puppis~A was
129: dominated by the swept-up interstellar medium (Tamura 1995; Hwang et
130: al.\ 2008). The complex interstellar environment such as a gradient in
131: the ambient density (Petre et al.\ 1982) as well as the relatively old
132: age of $\sim$4000 years (Blair et al.\ 2003) might have made it
133: difficult to detect clear signs of metal-rich ejecta in its spectra.
134: However, recent {\it Suzaku} observations revealed metal-rich ejecta
135: features in the northeastern part of the remnant (Hwang et al.\ 2008).
136: Here, we also report of the discovery of the X-ray--emitting
137: fast-moving ejecta knots positionally coincident with the optical
138: ``omega'' filament.
139:
140: \section{Observations and Data Reduction}
141:
142: Parts of Puppis~A have been observed several times with {\it XMM-Newton}.
143: Among them, two observations (ObsIDs.\ 0113020101 and 0113020301)
144: cover the ``omega'' filament (Winkler \& Kirshner 1985). The fields of
145: view (FOV) of the two {\it XMM-Newton} observations are shown on a
146: {\it ROSAT} HRI image of the entire Puppis~A in Fig.~\ref{fig:hri}.
147: All the raw data were processed
148: with version 6.5.0 of the XMM Science Analysis Software. We use only
149: MOS data, since the pn data were obtained in PrimeSmallWindow mode so
150: that the pn FOV covered only a small region around the CCO that is not
151: focused in this paper. We select X-ray events corresponding to
152: patterns 0--12. We further
153: clean the data by rejecting high background (BG) intervals and
154: remove all the events in bad columns listed in Kirsch~(2006). The
155: summed exposures from the good time intervals (GTIs) after data cleaning
156: are 22\,ks for ObsID.\ 0113020101 and 10.8\,ks for ObsID.\ 0113020301,
157: respectively. After the filtering, the data were vignetting-corrected
158: using the sas task {\tt evigweight}.
159:
160: \section{Analysis and Results}
161:
162: Fig.~\ref{fig:opt xmm} {\it Left} shows the optical O {\scshape III}
163: image. We can see an $\Omega$-shaped filament, so-called ``omega''
164: filament in the central portion of the figure. Fig.~\ref{fig:opt xmm}
165: {\it Right} shows an X-ray three-color image of the
166: merged MOS1/2 data which covers the same area. Red, green, and blue
167: represent 0.4--0.7 (for O lines), 0.7--1.2 (for Ne lines), and
168: 1.2--5.0 (for hard band)\,keV, respectively. We find an
169: X-ray--emitting knotty feature in the central
170: portion of Fig.~\ref{fig:opt xmm} {\it Right}. The southern region
171: of the feature is positionally coincident with the optical ``omega''
172: filament as indicated by box regions in Fig.~\ref{fig:opt xmm}.
173:
174: We extract spectra from the north and south of the X-ray knotty feature
175: because it is divided into two regions in terms of color:
176: the NE region of the feature shows white color indicating a significant
177: mixture of spectrally hard emission, while the SW region of the feature
178: shows soft red emission. The two regions are shown as a white solid
179: circle (radius of 30$^{\prime\prime}$) and a white solid ellipse
180: (angular sizes of x and y-axis of 20$^{\prime\prime}$ and
181: 40$^{\prime\prime}$, respectively) in Fig.~\ref{fig:opt xmm} {\it
182: Right}. Note that the southern region in the feature includes the
183: ``omega'' filament. We simultaneously fit the MOS1 and MOS2 spectra,
184: allowing the normalizations between the two detectors to vary by
185: introducing the constant model in XSPEC (Arnaud 1996).
186: As BG, we subtract the X-ray emission in their surrounding regions
187: (i.e., areas enclosed by dashed lines around each region shown in
188: Fig.~\ref{fig:opt xmm} {\it Right}) after normalizing the intensities at the
189: ratios of source areas to BG areas. In this way, we attempt to
190: obtain the spectra of the knotty feature itself. We note that our
191: method is not unprecedented; the same method was successfully
192: performed in analyses of ejecta knots in Cas~A SNR (Hwang \& Laming
193: 2003; Laming \& Hwang 2003). The count rates per area of the source
194: and BG regions are summarized in Table~\ref{tab:rate}.
195:
196: We apply an absorbed single component non-equilibrium ionization
197: (NEI) model for the two spectra (the {\tt wabs} (Morrison \& McCammon
198: 1983) and {\tt vpshock} model (NEI version 2.0) (e.g., Borkowski et
199: al.\ 2001) in XSPEC v\,12.3.1). Initially, we allow the individual
200: element abundances to be freely fitted relative to H and find that
201: enhanced values are required\footnote{In this fitting, we obtain
202: only lower limits of metal abundances that are typically several
203: hundred times the solar values.}. This fact
204: clearly shows that the knotty feature consists of metal-rich
205: ejecta. The northern region is rich in O, Ne, Mg, Si, S, and Fe,
206: whereas the southern region is rich in O, Ne, and Mg. We note that
207: Hwang et al.\ (2008) independently detect metal-abundance enhancements
208: positionally coincident with the X-ray knotty feature disclosed here
209: from their {\it Suzaku} observations, although the relatively low
210: spatial resolving power of {\it Suzaku} prevented them from
211: identifying the metal-abundance enhancements with the ``omega'' filament.
212: We then set O/H at the optically determined value of 2000 (Winkler \&
213: Kirshner 1985) in order to obtain element abundances relative to O.
214: Free parameters are hydrogen column density,
215: $N_\mathrm{H}$; electron temperature, $kT_\mathrm{e}$; ionization
216: timescale, $n_{\rm e}t$; volume emission measure (VEM; VEM $=\int
217: n_\mathrm{e}n_\mathrm{H} dV$, where $n_\mathrm{e}$ and $n_\mathrm{H}$
218: are number densities of electrons and protons, respectively and $V$ is
219: the X-ray--emitting volume); abundances of Ne, Mg, Si, S, Fe, and
220: Ni. Above, $n_{\rm e}t$ is the electron density times the elapsed
221: time after shock heating and the {\tt vpshock} model assumes a range of
222: $n_{\rm e}t$ from zero up to a fitted maximum value. We set the
223: abundance of Ni equal to that of Fe. Abundances of the other elements
224: are fixed to the solar values \cite{Anders1989}.
225:
226: The two spectra, along with the best-fit models, are shown in
227: Fig.~\ref{fig:ej_spec} (left column). Note that we exclude data in the
228: energy range below 0.65\,keV where apparent differences between MOS1
229: and MOS2 data are seen. The difference might be due to the
230: contamination on the MOS1 chip (Pradas \& Kerp 2005). The best-fit
231: parameters and fit statistics are summarized in
232: Table~\ref{tab:param}. We find that the fit level for the northern
233: spectrum is far from acceptable while that for the southern one is
234: moderate. We notice that the residuals in the northern spectrum have
235: wavy structures around several lines such as Ne He$\alpha$, Mg
236: He$\alpha$, or Si He$\alpha$. These features suggest that the line
237: center energies are systematically shifted from those expected by the
238: best-fit model.
239:
240: We thus introduce a variable redshift in the same {\tt vpshock} model. The
241: two spectra with the revised best-fit model are also shown in
242: Fig.~\ref{fig:ej_spec} (right column). The revised model significantly
243: improves the fits for both of the two spectra (with a significance
244: level of greater than 99.9\% based on the $F$-test). In particular,
245: the wavy structures of residuals seen in the previous fit, where the
246: redshift is fixed to zero, for the northern spectrum are greatly reduced in
247: the revised fit, where the redshift is free. Table~\ref{tab:param}
248: summarizes the best-fit parameters for the two spectra obtained from
249: the revised model fitting. The derived redshift parameter shows
250: negative, i.e., blue-shifted values in the two regions. The value of
251: the blueshift for the northern region of the X-ray knotty feature is higher
252: than that for the southern region. We find that a relatively low
253: temperature and little evidence of Si and S abundances in south of the
254: feature, while relatively high temperatures and significant Si and S
255: abundances in the north of the feature. Those facts result in the
256: color variation seen Fig.~\ref{fig:opt xmm} {\it Right}. Ionization
257: timescales for the two regions are significantly lower than that
258: expected for the collisional ionization equilibrium.
259:
260: Doppler velocity measurements with X-ray CCD detectors have been
261: successfully demonstrated by several different groups (e.g.,
262: Willingale et al.\ (2002) using data from {\it XMM-Newton} MOS). However,
263: small redshifts of $\lesssim 10$\,eV require careful investigation of
264: systematic uncertainties and/or artifacts due to the detector
265: calibration (the uncertainty of absolute energy scale is $\lesssim$\,5\,eV;
266: Kirsch 2007). To estimate systematic errors of the derived Doppler
267: shifts, as well as to ensure
268: that the line shifts are not due to calibration uncertainties, we
269: further perform spectral analysis in the following three cases: (1)
270: individual fit of MOS1 and MOS2 spectra, (2) individual fit of spectra
271: from the two observations (whose rotation angles are different from
272: each other), (3) inhomogeneous intensities of BG spectra, in case that the
273: intensities of the BG spectra in the source regions are different from
274: those in the BG regions currently selected. We examine
275: 50\% higher or lower than the original intensity since there are
276: inhomogeneities of X-ray intensities of $\sim$50\% around the X-ray
277: knotty feature. We summarize all the derived values of the redshift in
278: Table~\ref{tab:uncertainty}. We find significant blueshifts in all
279: the cases for the two regions. The systematic
280: uncertainties are greater than the statistical ones and dominate the
281: significance of the redshift. The Doppler velocities, considering all
282: the systematic uncertainties, are
283: $-3400^{+1000}_{-800}$\,km\,sec$^{-1}$ for the north of the X-ray
284: knotty feature and $-1700^{+700}_{-800}$\,km\,sec$^{-1}$ for the south
285: of the feature. The dominant uncertainty comes from the systematic
286: uncertainty between the two detectors and/or that introduced by
287: possible variations of the BG intensities. We should note that the
288: values of metal abundances vary by factors of $\sim$3 in the three
289: cases of BG intensities examined. This systematic uncertainty is
290: considered to be conservative errors of the abundances obtained.
291:
292: Next, we investigate whether or not the calibration uncertainty of the
293: MOS energy scale is serious in our data. We measure line center
294: energies in the two regions and the local BG region for each region,
295: applying a phenomenological model, i.e., a thermal continuum and 14
296: Gaussian line profiles. The center energy, line width, normalization
297: of each Gaussian line, as well as the temperature and normalization of
298: the thermal continuum, were treated as free parameters. The derived
299: line center energies for Ne He$\alpha$, Mg He$\alpha$, and Si
300: He$\alpha$ are summarized in Table~\ref{tab:lineE}. We confirm
301: that the line center energies obtained in the local-BG-subtracted
302: spectra are indeed significantly higher than those obtained in the
303: surrounding BG regions. Furthermore, the blueshifts implied by simply
304: comparing the source and BG line centers are quite similar to that
305: found in the NEI best-fit models. These facts strongly
306: support the suggestion that the line shifts seen in the ejecta feature
307: are not due to calibration uncertainties of the MOS energy scale.
308:
309: Finally, we investigate whether the variations of line center energies
310: are caused by the plasma condition (i.e., ionization states, electron
311: temperature) rather than Doppler shifts. According to the NEI code in
312: Hughes et al.\ (2003), line center energies of the K-shell complex,
313: including all lines from all charge states excluding Ly$\alpha$ and
314: all higher energy lines are calculated to be 0.872--0.918\,keV for Ne,
315: 1.269--1.350\,keV for Mg, and 1.744--1.862\,keV for Si. The results in
316: Table~\ref{tab:lineE} show that the shifts in line centroids from the
317: northern region cannot be related to temperature/ionization effects,
318: while the effects cannot be ruled out for the south region. However,
319: the agreement with optical Doppler velocity makes it plausible that we
320: detect blueshift in the south region.
321:
322: In this context, we are convinced that the line shifts observed in
323: the two regions are due to neither instrumental origin nor plasma
324: conditions (i.e., $kT_{\mathrm e}$ and $n_{\rm e}t$) but reflect the
325: Doppler-shifts caused by their fast motion toward us. In conclusion,
326: we observe blue-shifted line emission from the north and south of the
327: X-ray knotty feature with Doppler velocities of
328: $-3400^{+1000}_{-800}$\,km\,sec$^{-1}$ and
329: $-1700^{+700}_{-800}$\,km\,sec$^{-1}$, respectively. The Doppler
330: velocity estimated in the south of the X-ray knotty feature, where the
331: optical ``omega'' filament is included, is consistent with that
332: measured for the optical ``omega'' filament
333: (1500$\pm$100\,km\,sec$^{-1}$; Winkler \& Kirshner 1985).
334:
335: \section{Discussion}
336:
337: We find an ejecta-dominated X-ray bright knotty feature on the
338: position of the optically O-rich filament discovered by Winkler \&
339: Kirshner (1985) and Winkler et al.\ (1988). Due to its fast motion
340: toward us, we detect blue-shifted lines from the feature.
341: The Doppler velocity in the northern region of the feature is
342: higher than that in the southern region. If they are moving with
343: different velocities, distances from the explosion center are also
344: different with each other. This means that we see two different ejecta
345: knots that are close to each other along the line of sight. Follow-up
346: work with data of better spatial resolution (i.e., {\it Chandra} data)
347: might reveal clear morphological separation between the north knot and
348: south knot.
349:
350: \subsection{Mass and Origin of the Fast-Moving Ejecta Knots}
351:
352: Under the assumption of metal abundance of O/H to be 2000 times the
353: solar value, the electrons are dominantly supplied by O or other
354: metals such as Si or S. We assume that the value of $n_\mathrm{e}$ is
355: 7 times that of the O ion density, $n_\mathrm{O}$, because O ions mainly
356: exist in the He-like or H-like ionization states at the electron
357: temperature and the ionization timescale obtained for
358: the knots (Table~\ref{tab:param}). Assuming that the depth of the
359: X-ray--emitting plasma is the same as the physical scale corresponding
360: to the apparent angular size of the knots (1$^\prime$) at a distance of
361: 2.2\,kpc (Reynoso et al.\ 1995), we estimate elemental densities and
362: masses in each region; these are summarized in Table~\ref{tab:dens_mass}.
363: We obtain the total masses in the northern and southern knots to be
364: $\sim$0.07\,M$_\odot$, $\sim$0.08\,M$_\odot$, respectively.
365:
366: The masses of individual OFMKs seen in Cas~A and Puppis~A are
367: estimated to be 1$\times 10^{-4}$ M$_{\odot}$ (Raymond 1984) and
368: 1$\times 10^{-2}$ M$_{\odot}$ (Winkler et al.\ 1988), respectively.
369: The masses of Fe-rich knots in Cas~A and O and Ne-rich knot in G292.0+1.8
370: are estimated to be 4$\times10^{-5}$--1$\times10^{-3}$\,M$_{\odot}$
371: (Hwang \& Laming 2003) and 1$\times 10^{-3}$ M$_{\odot}$ (Park et al.\
372: 2004), respectively. The masses of Vela shrapnels A (Si-rich knot)
373: and D (O, Ne, and Mg-rich knot) are estimated to be 5$\times10^{-3}$ and
374: 0.1\,M$_{\odot}$, respectively (Katsuda \& Tsunemi 2006; Katsuda \&
375: Tsunemi 2005). Therefore, the ejecta knots disclosed here are
376: categorized as the most massive ejecta knots.
377:
378: We investigate the origin of these ejecta material by comparing the
379: observed metal abundance ratios with those of theoretical models. We
380: employed models for various progenitor masses
381: with initial composition of solar values. We examine two sets of
382: models by Rauscher et al.\ (2002) and Tominaga et al.\ (2007).
383: For the southern knot, we compare the relative abundances
384: of Ne and Mg to O with those expected in the models, and find that
385: some models by both of the two groups can reproduce the derived
386: relative abundances in the explosive Ne/C-burning cores. On the other
387: hand, the composition of the northern knot turns out not to be
388: reproduced in any models at any mass radius. If we compare the mass
389: ratios of S and Fe to Si to those of the models, we find that all the
390: models examined can reproduce the mass ratios in incomplete explosive
391: Si-burning layers. The progenitor mass and specific burning process
392: are not well constrained by the current data analysis. This is
393: partly because the local BG subtraction affects the abundance
394: determination by a factor of $\sim$3, and partly because emission
395: from north and south knots are contaminating with each other in our
396: current data. Follow-up observations with better spatial resolution
397: are needed to reveal the origins of the two knots as well as the
398: progenitor mass.
399:
400: \subsection{Recoil to the High-Velocity CCO}
401:
402: Considering that OFMKs in Puppis~A are believed to be recoiled
403: materials to the high-velocity CCO (Hui \& Becker 2006; Winkler \&
404: Petre 2007), we suggest that the ejecta knots which we find
405: near the OFMKs are also part of the recoiled materials. Winkler \&
406: Petre (2007) measured the proper motion of the stellar remnant and
407: estimated the momentum to be
408: $\sim4\times10^{41}\mathrm{g\,cm\,s^{-1}}$ in the plane of the sky.
409: They also estimated approximate total
410: momentum for the 11 measured OFMKs toward
411: the opposite direction to the traveling direction of the CCO to be
412: $\sim1.3\times10^{41}\mathrm{g\,cm\,s^{-1}}$, comprising about
413: $\sim$30\% of the required momentum to balance that of the CCO.
414: We estimate that for the X-ray--emitting ejecta knots to be
415: $\sim0.5\times10^{41}\mathrm{g\,cm\,s^{-1}}$, comprising about 10\% of
416: the required momentum. The rest of the required momentum might be
417: explained by unidentified $\sim$20 OFMKs whose existence was
418: suggested in O {\scshape III} image of this remnant (Winkler \&
419: Petre 2007).
420:
421: There are mainly three mechanisms proposed to explain the origin
422: of high-velocity neutron stars such as the CCO in Puppis~A: binary
423: disruptions, natal or postnatal kicks from the SN explosions (see, Lai
424: et al.\ 2001 for a review). Recent theoretical studies have focused
425: on natal kicks imparted to neutron stars at birth rather than the
426: other two mechanisms. A natal kick could
427: be due to global hydrodynamical perturbations in the SN core (e.g.,
428: Goldreich et al.\ 1996), or it could be a result from asymmetric
429: neutrino emission in the presence of super strong magnetic fields
430: ($B \gtrapprox 10^{15}$ G) in the proto--neutron star (e.g., Blandford et
431: al.\ 1983). The two models lead to distinct predictions: the measured
432: neutron star velocity should be directed opposite to the momentum of
433: the gaseous SN ejecta caused by linear momentum conservation in the
434: hydrodynamically-driven mechanisms (e.g., Scheck et al.\ 2006) while
435: neutrino-driven mechanisms predict the motion of the ejecta
436: roughly in the same direction of the neutron star kick (Fryer \&
437: Kusenko 2006). Therefore, our data as well as the distributions of
438: OFMKs suggest that the hydrodynamically- (or ejecta-)driven mechanisms
439: are at work to produce the high-velocity CCO in Puppis~A.
440: However, we should care about the possibility that the lack of
441: OFMKs in the southwestern portion of the remnant might otherwise be
442: explained by complicated structures of the ambient medium (e.g.,
443: Petre et al.\ 1982). Also, the existence of ejecta knots disclosed
444: here is not inherently convincing evidence for a lack of undetected
445: ejecta in other directions, although diffuse X-ray emission is
446: generally reported to have low metal-abundances (Tamura 1995; Hwang
447: et al.\ 2008). To obtain further conclusive evidence that the
448: ejecta-driven mechanism is at work, we need detailed spatially
449: resolved spectral analysis for the entire remnant.
450:
451: \section{Summary}
452:
453: We have analyzed archival {\it XMM-Newton} data of the Puppis~A SNR.
454: We disclose an X-ray knotty feature on the position of one of OFMKs
455: discovered by Winkler \& Kirshner (1985) and Winkler et al.\ (1988).
456: We find that the X-ray knotty feature is metal-rich ejecta with
457: blue-shifted emission lines. The composition in the northern region
458: of the feature is different from that in the southern region: O, Ne,
459: Mg, Si, S, Fe-rich in the northern region, whereas O, Ne, and Mg-rich
460: in the southern region. Also, the Doppler velocity in the northern region,
461: $-3400^{+1000}_{-800}$\,km\,sec$^{-1}$, is different from that in the
462: southern region, $-1700^{+700}_{-800}$\,km\,sec$^{-1}$. These facts
463: lead us to consider that the feature consists of two different knots
464: that are close to each other along the line of sight. Current data
465: are not sufficient to constrain the origins of the two knots as well
466: as to precisely determine the Doppler velocities in the two knots.
467: Further observations with better spatial resolution and better
468: observational condition, especially considering that the knots are
469: located at the edge of FOV in the current data, are strongly required
470: to reveal possible morphological separation between the north knot and
471: south knot, and to accurately determine the compositions as well as
472: Doppler velocities in the two knots.
473:
474:
475: \acknowledgments
476:
477: This work is partly supported by a Grant-in-Aid for Scientific Research
478: by the Ministry of Education, Culture, Sports, Science and Technology
479: (16002004). This study is also carried out as part of the 21st Century
480: COE Program, \lq{\it Towards a new basic science: depth and
481: synthesis}\rq. The work of K.M.\ is partially supported by the
482: Grant-in-Aid for Young Scientists (B) of the MEXT (No.\ 18740108).
483: S.P.\ was supported in part by the NASA grant under the contract
484: NNG06GB86G. P.O.S.\ acknowledges support from NASA Contract
485: NAS8-39073. S.K.\ is supported by JSPS Research Fellowship for Young
486: Scientists.
487:
488: \clearpage
489:
490: \begin{thebibliography}{}
491: \bibitem[Anders \& Grevesse 1989]{Anders1989}
492: Anders, E., \& Grevesse, N. 1989, Acta, 53, 197
493: \bibitem[Arnaud (1996)]{Arnaud1996}
494: Arnaud, K. A. 1996, ASP Conf. Ser., 101, 17
495: \bibitem[Blair et al.\ (2000)]{Blair2000}
496: Blair, W. P., Morse, J. A., Raymond, J. C., Kirshner, R. P.,
497: Hughes, J. P., Dopita, M. A., Sutherland, R. S., Long, K. S.,
498: \& Winkler, P. F. 2000, ApJ, 537, 667
499: \bibitem[Blair et al.\ (2003)]{Blair2003}
500: Blair, W. P., Sankrit, R., Ghavamian, P., Raymond, J. C., and
501: Morse, J. A. 2003, AAS, 203, 3912B
502: \bibitem[Blandford(1983)]{Blandford1983}
503: Blandford, R. D., Applegate, J. H., \& Hernquist, J. 1983,
504: MNRAS, 204, 1025
505: %\bibitem[Borkowski(1994)]{Borkowski1994}
506: % Borkowski, K. J., Sarazin, C. L., \& Blondin, J. M. 1994, ApJ,
507: % 429, 710
508: \bibitem[Borkowski et al.\ (2001)]{Borkowski2001}
509: Borkowski, K. J., Lyerly W. J., \& Reynolds, S. P. 2001, ApJ,
510: 548, 820
511: %\bibitem[Brickhouse(1993)]{Brickhouse1993}
512: % Brickhouse, N. S., Raymond, J. C., \& Smith, B. W. 1993, BAAS, 25, 864
513: %\bibitem[Burrows(1996)]{Burrows1996}
514: % Burrows, A., \& Hayes, J. 1996, Phys, Rev. Lett., 76, 352
515: %\bibitem[Burrows et al.\ 2007]{Burrows2007}
516: % Burrows, A., Livne, E., Dessart, L., Ott, C. D., Murphy,
517: % J. 2007, ApJ, 655, 416
518: %\bibitem[Canizares(1981)]{Canizares1981}
519: % Canizares, C. R., \& Winkler, P. F. 1981, ApJ, 246, L33
520: \bibitem[Chevarier(1979)]{Chevarier1979}
521: Chevarier, R., \& Kirshner, R. 1979, ApJ, 233, 154
522: %\bibitem[Dopita(1984)]{Dopita1984}
523: % Dopita, M. A. \& Tuohy, I. R. 1984, ApJ, 282, 135
524: \bibitem[Fryer(2006)]{Fryer2006}
525: Fryer C. L. \& Kusenko, A. 2006, ApJ, 163, 335
526: %\bibitem[Fujita(2004)]{Fujita2005}
527: % Fujita, Y., Sarazin, C. L., Reiprich, T. H., Andernach, H.,
528: % Ehle, M., Murgia, M., Rudnich, L., \& Slee, O. B. 2004, ApJ,
529: % 616, 157
530: \bibitem[Goldreich(1996)]{Goldreich1996}
531: Goldreich, P., Lai, D., \& Sahrling, M., 1996, in Unresolved
532: Problems in Astrophysics, ed. J. N. Bahcall, \& J. P. Ostriker
533: (Princeston; Princeton Univ. Press), 269
534: \bibitem[Goss(1979)]{Goss1979}
535: Goss, W. M., Shaver, P. A., Zealey, W. J., Murdin, P., and
536: Clark, D. H. 1979, MNRAS, 188, 357
537: %\bibitem[Hamilton(1983)]{Hamilton1983}
538: % Hamilton, A. J. S., Sarazin, C. L., \& Chevalier, R. A. 1983,
539: % ApJS, 51, 115
540: %\bibitem[Hughes(2000)]{Hughes2000}
541: % Hughes, J. P., Rakowski, C. E., Burrows, D. N., Slane,
542: % P. O. 2000, ApJ, 528, L109
543: \bibitem[Hughes(2003)]{Hughes2003}
544: Hughes, J. P., Rakowski, C. E., Decourcelle, A. 2003, ApJ, 543, L61
545: %\bibitem[Hui(2006a)]{Hui2006a}
546: % Hui, C. Y., \& Becker, W. 2006, A\&A, 454, 543 (Hui \&
547: % Becker 2006a)
548: \bibitem[Hui(2006)]{Hui2006}
549: Hui, C. Y., \& Becker, W. 2006, A\&A, 457, L33
550: %\bibitem[Hwang(2000)]{Hwang2000}
551: % Hwang, U., Holt, S. S., and Petre R. 2000, ApJ 537 L119
552: \bibitem[Hwang(2003)]{Hwang2003}
553: Hwang, U., \& Laming, J. M. 2003, ApJ, 597, 362
554: %\bibitem[Hwang(2005)]{Hwang2005}
555: % Hwang, U., Flanagan, K. A., \& Petre, R. 2005, ApJ, 635, 355
556: \bibitem[Hwang et al.\ (2008)]{Hwang2008}
557: Hwang, U., Petre, R., \& Flanagan, K. A. 2008, ApJ, in press
558: \bibitem[Katsuda(2005)]{Katsuda2005}
559: Katsuda, S., \& Tsunemi, H. 2005, PASJ, 57, 620
560: \bibitem[Katsuda(2006)]{Katsuda2006}
561: Katsuda, S., \& Tsunemi, H. 2006, ApJ, 642, 917
562: %%\bibitem[Katsuda et al.\ (2007)]{Katsuda2007}
563: % Katsuda, S., Tsunemi, H., Miyata, E., Mori, K., Namiki, M.,
564: % and Nemes, N. 2007 PASJ in press
565: \bibitem[Kirsch(2006)]{Kirsch2006}
566: Kirsch, M. 2006, XMM-EPIC status of calibration and data
567: analysis, XMM-SOC-CAL-TN-0018, issue 2.5
568: \bibitem[Kirsch(2007)]{Kirsch2007}
569: Kirsch, M. 2007, XMM-EPIC status of calibration and data
570: analysis, XMM-SOC-CAL-TN-0018, issue 2.6
571: %\bibitem[Kirshner(1989)]{Kirshner1989}
572: % Kirshner, R. P., Morse, J. A., Winkler, P. F., \& Blair,
573: % W. P., 1989, ApJ, 342, 260
574: %\bibitem[Kotake(2005)]{Kotake2005}
575: % Kotake, K, Yamada, S., \& Sato K. 2005, ApJ, 618, 474
576: \bibitem[Lai(2001)]{Lai2001}
577: Lai, D., Chernoff, D. F., \& Cordes, J. M. 2001, ApJ, 549, 1111
578: \bibitem[Laming(2003)]{Laming2003}
579: Laming, J. M., \& Hwang, U., 2003, ApJ, 597, 347
580: %\bibitem[Leonard(2006)]{Leonard2006}
581: % Leonard, D. C, Filippenko, A. V., Ganeshalingam, M., Serduke,
582: % F. J. D., Li, W., Swift, B. J., Gal-Yam, A., Foley, R. J., Fox,
583: % D. B., Park, S., Hoffman, J. L., Wong., D. S. 2006, Nature, 440, 505
584: %\bibitem[Liedahl(1995)]{Liedahl1995}
585: % Liedahl, D. A., Osterheld, A. L., \& Goldstein, W. H. 1995,
586: % ApJ, 438, L115
587: %\bibitem[Lasker(1978)]{Lasker1978}
588: % Lasker, B. M. 1978, ApJ, 223, 109
589: \bibitem[Morrison(1983)]{Morrison1983}
590: Morrison, R., \& McCammon, D. 1983, ApJ, 270, 119
591: \bibitem[Nomoto(1984)]{Nomoto1984}
592: Nomoto, K., Thielemann, F.-K., \& Yokoi, K. 1984, ApJ, 286, 644
593: %\bibitem[Motohara(2006)]{Motohara2006}
594: % Motohara, K., Maeda, K., Gerardy, C. L., Nomoto, K., Tanaka,
595: % M., Tominaga, N., Ohkubo, T., Mazzali, P. A., Fesen, R. A.,
596: % Hoflich, P., and Wheeler, J. C. 2006, astro-ph/0610303
597: %\bibitem[Park(2002)]{Park2002}
598: % Park, S., Peter W. A. R., Hughes, J. P., Slane, P. O., Burrows,
599: % D. N., Garmire, G. P., and Nousek, J. A. 2002, ApJ, 564 L39
600: %\bibitem[Park et al.\ (2003)]{Park2003}
601: % Park, S., Burrows, D. N., Garmire, G. P., Nousek, J. A.,
602: % Hughes, J. P., and Williams, R. M. 2003, ApJ, 586, 210
603: \bibitem[Park(2004)]{Park2004}
604: Park, S., Hughes, J. P., Slane, P. O., Burrows, D. N., Roming,
605: P. W. A., Nousek, J. A., and Garmire, G. P. 2004, ApJ, 602, L33
606: \bibitem[Petre(1982)]{Petre1982}
607: Petre, R., Kriss, G., A., Winkler, P. F., \& Canizares,
608: C. R. 1982, ApJ, 258, 22
609: %\bibitem[Petre(1996)]{Petre1996}
610: % Petre, R., Becker, C. M., \& Winkler, P. F. 1996, ApJ, 465, L43
611: \bibitem[Pradas(2005)]{Pradas2005}
612: Pradas, J. \& Kerp, J. 2005, A\&A, 443, 721
613: %\bibitem[Rauscher et al.\ 2002]{Rauscher2002}
614: % Rauscher, T., Heger, A., Hoffman, R. D., Woosley, S. E. 2002,
615: % ApJ, 576, 323
616: %\bibitem[Raymond(1977)]{Raymond1977}
617: % Raymond, J. C., \& Smith, B. W., 1977, ApJS, 35, 419
618: %\bibitem[Raymond (1984)]{Reynoso1984}
619: % Raymond, J. C. 1984, ARA\&A, 22, 75
620: \bibitem[Reynoso(1995)]{Reynoso1995}
621: Reynoso, E. M., Dubner, G. M., Goss, W. M. \& Arnal,
622: E. M. 1995, AJ, 110, 318
623: %\bibitem[Read(2003)]{Read2003}
624: % Read, A., M., \& Ponman, T. J. 2003, A\&A, 409, 395
625: %\bibitem[Sakhibov(1983)]{Sakhibov1983}
626: % Sakhibov, F. K. \& Smirnov, M. A. 1983, Soviet Astronomy, 27, 395
627: %\bibitem[Sato(2005)]{Sato2005}
628: % Sato, K., Furusho, T., Yamasaki, Y., Ishida M., Matsushida,
629: % K., and Ohashi, T. 2005, PASJ, 57, 743
630: \bibitem[Scheck(2006)]{Scheck2006}
631: Scheck, L., Kifonidis, K., Janka, H.-Th., \& Muller, E. 2006,
632: A\&A, 457, 963
633: %\bibitem[Smith(2001)]{Smith2001}
634: % Smith, R. K., Brickhouse, N. S., Liedahl, D. A., \& Raymond,
635: % J. C. 2001, ApJ, 556, L91
636: %\bibitem[Scheck(2004)]{Scheck2004}
637: % Scheck, L., Plewa, T., Janka, H.-T., Kifonidis, K., \& Muller,
638: % E. 2004, Phys. Rev. Lett., 92, 011103
639: \bibitem[Tamura(1995)]{Tarumra1995}
640: Tamura, K. 1995 Ph.D. thesis, Osaka Univ.
641: \bibitem[Thielemann(1996)]{Thielemann1996}
642: Thielemann, F.-K., Nomoto, K., \& Hashimoto, M. 1996, ApJ, 460, 408
643: \bibitem[Tominaga et al.\ 2007]{tominaga07}
644: Tominaga, N., Umeda, H., Nomoto, K. in prep.
645: %\bibitem[Tsunemi et al.\ 2007]{Tsunemi2007}
646: % Tsunemi, H., Katsuda, S., Nemes, N., and Miller, E. D. 2007,
647: % ApJ in press
648: \bibitem[Willingale et al.\ 2002]{Willingale2002}
649: Willingale, R., Bleeker, J. A. M., van der Heyden, K. J.,
650: Kaastra, J. S., and Vink, J. 2002, A\&A, 381, 1039
651: \bibitem[Winkler(1985)]{Winkler1985}
652: Winkler, P. F. \& Kirshner, R. P. 1985, ApJ, 299, 981
653: \bibitem[Winkler(1988)]{Winkler1988}
654: Winkler, P. F., Tuttle, J. H., Kirshner, R. P., \& Irwin, M.,
655: J. 1988, in IAU Colloq. 101: Supernova Remnants and
656: the Interstellar Medium, ed. R. S. Roger \&
657: T. L. Landecker, 65-+
658: \bibitem[Winkler(2007)]{Winkler2007}
659: Winkler, P., F., \& Petre, R. 2007, ApJ, 670, 635
660: %\bibitem[Woosley \& Weaver 1995]{Woosley1995}
661: % Woosley, S. E. \& Weaver, T. A. 1995, ApJS, 101, 181
662: \bibitem[]{}
663:
664: \end{thebibliography}
665:
666: \clearpage
667:
668:
669: \begin{deluxetable}{lcccccc}
670: \tabletypesize{\scriptsize}
671:
672: \tablecaption{Count rates per area (counts\,sec$^{-1}$\,arcmin$^{-2}$)
673: in the source and BG regions.}
674: \tablewidth{0pt}
675: \tablehead{
676: \colhead{} &\colhead{North} &\colhead{North (BG)}&\colhead{South}
677: &\colhead{South (BG)}
678: }
679: \startdata
680: MOS1 &
681: 2.53$\pm$0.01 & 1.40$\pm$0.01 & 1.51$\pm$0.01& 0.99$\pm$0.01 \\
682: MOS2 &
683: 2.42$\pm$0.01 & 1.26$\pm$0.01& 1.37$\pm$0.01& 0.99$\pm$0.01\\
684: \enddata
685:
686: \tablecomments{Count rates are estimated in the energy range of
687: 0.65--4.0\,keV.}
688: \label{tab:rate}
689: \end{deluxetable}
690:
691:
692: %\begin{flushleft}
693: %\oddsidemargin -0.5cm
694: \begin{deluxetable}{lcccccccccccc}
695: \tabletypesize{\tiny}
696: \tablecaption{Spectral-fit parameters in the two regions shown in
697: Fig.~\ref{fig:opt xmm} {\it Right}}
698: \tablewidth{0pt}
699: \tablehead{
700: \colhead{Region}&
701: \colhead{$N_\mathrm{H}$}&
702: \colhead{$kT_\mathrm{e}$ (keV)}& \colhead{log($n_{\rm e}t$)}&
703: \colhead{Ne/O}&\colhead{Mg/O}& \colhead{Si/O}& \colhead{S/O} &
704: \colhead{Fe/O}& \colhead{VEM$^a$} &
705: \colhead{redshift ($\times10^{-3}$)}&\colhead{$\chi^2/d.o.f.$}
706: }
707: \startdata
708:
709: North&0.37$^{+0.03}_{-0.02}$& 1.36$^{+0.11}_{-0.08}$& 10.65$^{+0.05}_{-0.05}$
710: & 0.8$\pm0.1$ & 0.8$\pm0.1$ & 0.5$\pm0.1$
711: & 0.4$\pm0.3$ & 0.22$^{+0.08}_{-0.06}$ & $1.2\pm0.2$ & --- & 462/216 \\
712:
713: South&0.43$\pm0.02$& 0.36$\pm0.02$& 10.59$^{+0.06}_{-0.08}$
714: & 1.4$^{+0.3}_{-0.1}$ & 1.7$^{+0.5}_{-0.3}$ & $<$1.1
715: & $<$0.7 & $<$0.03 & 2.7$\pm0.3$ & --- & 290/191 \\
716:
717: \\
718:
719: North&0.40$^{+0.03}_{-0.06}$& 0.60$^{+0.08}_{-0.05}$& 11.20$^{+0.07}_{-0.06}$
720: & 1.04$^{+0.05}_{-0.06}$ & 0.95$^{+0.19}_{-0.05}$ & 0.7$\pm0.1$
721: & $<$0.8 & 0.16$\pm0.03$ & $2.2\pm0.1$ & $-11.4^{+0.4}_{-0.3}$
722: & 288/215 \\
723:
724: South&0.43$^{+0.01}_{-0.02}$& 0.35$\pm0.01$& 10.71$^{+0.02}_{-0.14}$
725: & 1.5$\pm0.1$ & 1.7$^{+0.2}_{-0.3}$ & $<$0.5
726: & $<$0.7 & $<$0.03 & 2.5$\pm0.3$ & $-5.5^{+0.1}_{-1.0}$ & 266/190 \\
727:
728: \enddata
729:
730: \tablecomments{The best-fit parameters for the two spectra in
731: Fig.~\ref{fig:opt xmm} {\it Right}. Results are from the {\tt vpshock} model in
732: which the redshift is fixed to zero (upper two rows) and allowed to vary
733: (lower two rows). The units of $N_{\rm H}$ and VEM are
734: $\times10^{22}$\,cm$^{-2}$ and
735: $\times10^{53}$\,cm$^{-3}$, respectively. VEM is calculated at a
736: distance of 2.2\,kpc (Reynoso et al.\ 1995). Quoted errors are at 90\%
737: confidence level.
738: The values of the abundance ratios are relative to those of the
739: solar ratios. Other elements are fixed to those of the solar values
740: \cite{Anders1989}. The errors are calculated after fixing the
741: $kT_\mathrm{e}$ to the best-fit value. Errors for VEM are
742: calculated after fixing the $kT_\mathrm{e}$ and Ne abundance to the
743: best-fit values.
744: }
745: \label{tab:param}
746: \end{deluxetable}
747: %\end{flushleft}
748:
749: \begin{table}
750: \begin{center}
751: %\tabletypesize{\scriptsize}
752: \caption{Redshift values ($\times10^{-3}$)}
753: \begin{tabular}{lcccccccc}
754: \tableline\tableline
755: Region &\multicolumn{2}{c}{Case-1$^a$}
756: &\multicolumn{2}{c}{Case-2$^b$} &
757: \multicolumn{3}{c}{Case-3$^c$}\\
758:
759: & MOS1&MOS2&Obs1&Obs2&0.5&1$^d$&1.5\\
760: \tableline
761: North & $-8.6^{+0.2}_{-0.7}$ & $-12.4^{+0.6}_{-0.4}$
762: & $-8.6\pm0.4$ & $-12.6^{+0.7}_{-0.5}$
763: & $-9.0^{+0.2}_{-0.1}$ & $-11.4^{+0.4}_{-0.3}$& $-13.9^{+0.1}_{-0.2}$\\
764:
765: South & $-5.6^{+1.6}_{-0.9}$ & $-5.6^{+2.2}_{-2.8}$
766: & $-5.1\pm0.2$ & $-6.7^{+3.5}_{-1.7}$
767: &$-5.7^{+0.1}_{-0.9}$ &$-5.5^{+0.1}_{-1.0}$& $-5.5^{+2.3}_{-1.2}$ \\
768:
769: \tableline
770: \end{tabular}
771: \tablecomments{Quoted errors are at 90\% confidence level. $^a$Case-1;
772: we individually analyze MOS1 and MOS2 data. In this case, we include
773: both of the two observations and subtract the local BG with original
774: intensity. $^b$Case-2; we separately analyze the data from the two
775: observations. Obs1 and Obs2 are ObsID.\ 0113020101 and ObsID.\
776: 0113020301, respectively. In this case, we use data from both MOS1 and MOS2
777: detectors and subtract the local BGs with original intensities.
778: $^c$Case-3; we artificially vary the intensities of the local BGs (0.5 and
779: 1.5 times the original intensity). In this case, we use all the
780: data sets (MOS1+MOS2+Obs1+Obs2). $^d$Same as listed in
781: Table~\ref{tab:param}. }
782:
783: \label{tab:uncertainty}
784: \end{center}
785: \end{table}
786:
787: \begin{deluxetable}{lcccccc}
788: \tabletypesize{\scriptsize}
789:
790: \tablecaption{Line center energies of spectra from regions shown in
791: Fig.~\ref{fig:opt xmm} {\it Right}.}
792: \tablewidth{0pt}
793: \tablehead{
794: \colhead{line} &\colhead{North} &\colhead{North (BG)}&\colhead{South}
795: &\colhead{South (BG)}
796: }
797: \startdata
798: Ne He$\alpha$ (eV) &
799: 930$^{+6}_{-5}$ & 917$\pm$2 &922$^{+5}_{-3}$ & 917$\pm$1 \\
800: Mg He$\alpha$ (eV) &
801: 1357$\pm$2 &1344$^{+1}_{-2}$ &1351$^{+3}_{-5}$ & 1346$^{+1}_{-3}$\\
802: Si He$\alpha$ (eV) &
803: 1878$^{+5}_{-8}$ & 1856$^{+4}_{-3}$ &ND$^a$ & 1857$^{+3}_{-2}$\\
804: \enddata
805:
806: \tablecomments{Quoted errors are at 90\% confidence level. $^a$We
807: could not determined the value due to the poor statistics.}
808: \label{tab:lineE}
809: \end{deluxetable}
810:
811: \begin{deluxetable}{lccccccc|cccccc}
812: \tabletypesize{\scriptsize}
813:
814: \tablecaption{Densities (cm$^{-3}$) and masses (M$_\odot$) in the two
815: regions in Fig.~\ref{fig:opt xmm} {\it Right}}
816: \tablewidth{0pt}
817: \tablehead{
818: \colhead{} &\colhead{$n_\mathrm{e}$}&\colhead{$n_\mathrm{O}$} &\colhead{$n_\mathrm{Ne}$}&\colhead{$n_\mathrm{Mg}$} &\colhead{$n_\mathrm{Si}$}&\colhead{$n_\mathrm{S}$} &\colhead{$n_\mathrm{Fe}$}&\colhead{M$_\mathrm{O}$} &\colhead{M$_\mathrm{Ne}$} &\colhead{M$_\mathrm{Mg}$} &\colhead{M$_\mathrm{Si}$} &\colhead{M$_\mathrm{S}$} &\colhead{M$_\mathrm{Fe}$}
819: }
820: \startdata
821: North&5&0.7&0.1&0.03&0.02&0.005&0.002&0.05&0.01&0.003&0.003&0.001&0.0015\\
822: South&5&0.7&0.15&0.05&0&0&0&0.06&0.015&0.007&0&0&0\\
823: \enddata
824: \tablecomments{Typical errors (which mainly come from the assumptions
825: of the plasma depth and the filling factor ) are about a factor of
826: 2.}
827: \label{tab:dens_mass}
828: \end{deluxetable}
829:
830: \begin{figure}
831: \includegraphics[angle=0,scale=0.8]{f1.eps}\hspace{1cm}
832: \caption{{\it XMM-Newton} FOV (white circles) overlaid on a {\it
833: ROSAT} HRI image of the entire Puppis~A SNR. The data have been
834: smoothed by Gaussian kernel of $\sigma = 15^{\prime\prime}$. The
835: intensity scale is square root. Optical O {\scshape III} and {\it
836: XMM-Newton} three-color images in the white box region are shown in
837: Fig.~\ref{fig:opt xmm}.
838: }
839: \label{fig:hri}
840: \end{figure}
841:
842: \begin{figure}
843: \includegraphics[angle=0,scale=0.8]{f2.eps}
844: \caption{{\it Left}: Linearly-scaled Optical O {\scshape III} image
845: for the box region
846: in Fig.~\ref{fig:hri}. The central $\Omega$-shaped filament
847: enclosed in a box region is a so-called ``omega'' filament (Winkler
848: \& Kirshner 1985). {\it Right}: {\it XMM-Newton} three-color image
849: in the same area. Red, green, and blue represent 0.4--0.7,
850: 0.7--1.2, and 1.2-5.0\,keV, respectively. The intensity scale is
851: square root. Spectral extraction regions are shown as solid (for
852: source spectra) and dashed (for BG spectra) white lines. The
853: location of the ``omega'' filament is indicated as a box
854: region that is the same box in Fig.~\ref{fig:opt xmm} {\it Left}.}
855: \label{fig:opt xmm}
856: \end{figure}
857:
858: \begin{figure}
859: \includegraphics[angle=0,scale=0.8]{f3.eps}
860: \caption{{\it Left column}: MOS1 (black) and MOS2 (red) spectra from the two
861: regions in Fig.~\ref{fig:opt xmm} with the best-fit model (redshift
862: = 0). The lower panels show the residuals. {\it Right column}:
863: Same as the left column but with the revised model (redshift is free).}
864: \label{fig:ej_spec}
865: \end{figure}
866:
867:
868: \end{document}
869: