1: %%
2: %% Beginning of file 'sample.tex'
3: %%
4: %% Modified 2004 January 9
5: %%
6: %% This is a sample manuscript marked up using the
7: %% AASTeX v5.x LaTeX 2e macros.
8:
9: %% The first piece of markup in an AASTeX v5.x document
10: %% is the \documentclass command. LaTeX will ignore
11: %% any data that comes before this command.
12:
13: %% The command below calls the preprint style
14: %% which will produce a one-column, single-spaced document.
15: %% Examples of commands for other substyles follow. Use
16: %% whichever is most appropriate for your purposes.
17: %%
18: \documentclass[12pt,preprint]{aastex}
19: %\documentclass{emulateapj}
20:
21: \usepackage{graphicx}
22:
23: \newcommand{\chandra}{{\it Chandra}}
24: \newcommand{\rosat}{{\it ROSAT}}
25: \newcommand{\einstein}{{\it Einstein}}
26: \newcommand{\xmm}{{\it XMM-Newton}}
27: \newcommand{\HST}{{\it Hubble Space Telescope}}
28: \newcommand{\hst}{{\it HST}}
29: \newcommand{\hii}{\ion{H}{2}}
30: \newcommand{\hi}{\ion{H}{1}}
31: \newcommand{\ha}{H$\alpha$}
32: \newcommand{\nii}{\ion{N}{2}}
33: \newcommand{\half}{\small 1/2}
34: \newcommand{\sii}{\ion{S}{2}}
35: \newcommand{\oiii}{\ion{O}{3}}
36: \newcommand{\oii}{\ion{O}{2}}
37: \newcommand{\oi}{\ion{O}{1}}
38:
39: %% manuscript produces a one-column, double-spaced document:
40:
41: %% \documentclass[manuscript]{aastex}
42:
43: %% preprint2 produces a double-column, single-spaced document:
44:
45: %% \documentclass[preprint2]{aastex}
46:
47: %% Sometimes a paper's abstract is too long to fit on the
48: %% title page in preprint2 mode. When that is the case,
49: %% use the longabstract style option.
50:
51: %% \documentclass[preprint2,longabstract]{aastex}
52:
53: %% If you want to create your own macros, you can do so
54: %% using \newcommand. Your macros should appear before
55: %% the \begin{document} command.
56: %%
57: %% If you are submitting to a journal that translates manuscripts
58: %% into SGML, you need to follow certain guidelines when preparing
59: %% your macros. See the AASTeX v5.x Author Guide
60: %% for information.
61:
62: %%\newcommand{\vdag}{(v)^\dagger}
63: %%\newcommand{\myemail}{skywalker@galaxy.far.far.away}
64:
65: %% You can insert a short comment on the title page using the command below.
66:
67: %%\slugcomment{Not to appear in Nonlearned J., 45.}
68:
69: %% If you wish, you may supply running head information, although
70: %% this information may be modified by the editorial offices.
71: %% The left head contains a list of authors,
72: %% usually a maximum of three (otherwise use et al.). The right
73: %% head is a modified title of up to roughly 44 characters.
74: %% Running heads will not print in the manuscript style.
75:
76: \shorttitle{Multiwavelength Analysis of the Physical Structure of N49}
77: \shortauthors{Bilikova et al.}
78:
79: %% This is the end of the preamble. Indicate the beginning of the
80: %% paper itself with \begin{document}.
81:
82: \begin{document}
83:
84: %% LaTeX will automatically break titles if they run longer than
85: %% one line. However, you may use \\ to force a line break if
86: %% you desire.
87:
88: \title{Supernova Remnants in the Large Magellanic Clouds IX: \\
89: Multiwavelength Analysis of the Physical Structure of N49}
90:
91: %% Use \author, \affil, and the \and command to format
92: %% author and affiliation information.
93: %% Note that \email has replaced the old \authoremail command
94: %% from AASTeX v4.0. You can use \email to mark an email address
95: %% anywhere in the paper, not just in the front matter.
96: %% As in the title, use \\ to force line breaks.
97:
98: \author{J. Bilikova, R. N. M. Williams, Y.-H. Chu, R. A. Gruendl and
99: B. F. Lundgren}
100: \affil{Astronomy Department, University of Illinois at Urbana-Champaign \\
101: 1002 W. Green St. Urbana, IL 61801}
102:
103: %% Notice that each of these authors has alternate affiliations, which
104: %% are identified by the \altaffilmark after each name. Specify alternate
105: %% affiliation information with \altaffiltext, with one command per each
106: %% affiliation.
107:
108: %% Mark off your abstract in the ``abstract'' environment. In the manuscript
109: %% style, abstract will output a Received/Accepted line after the
110: %% title and affiliation information. No date will appear since the author
111: %% does not have this information. The dates will be filled in by the
112: %% editorial office after submission.
113:
114: \begin{abstract}
115:
116: We present a multiwavelength analysis of the supernova remnant N49 in the Large
117: Magellanic Cloud. Using high-resolution {\it Hubble Space Telescope} WFPC2
118: images of H$\alpha$, [\sii] and [\oiii] emission, we study the morphology of
119: the remnant and calculate the rms electron densities in different regions.
120: We detect an offset of [\oiii] and H$\alpha$ emission peaks of about 0\farcs5,
121: and discuss possible scenarios that could give rise to such high values.
122: The kinematics of the remnant is analyzed by matching individual kinematic
123: features in the echelle spectra obtained at CTIO with the morphological
124: features revealed in the WFPC2 images. We detect narrow H$\alpha$ emission
125: component and identify it as the diffuse pre-shock recombination radiation,
126: and discrete broad emission features that correspond to the shocked gas in
127: filaments. The overall expansion of the remnant is about 250 km s$^{-1}$.
128: The dense clouds are shocked up to line-of-sight velocities of 250 km s$^{-1}$
129: and the less dense gas up to 300 km s$^{-1}$. A few cloudlets have even
130: higher radial velocities, reaching up to 350 km s$^{-1}$. We confirm
131: the presence of the cavity in the remnant, and identify the center of
132: explosion.
133: Using archival {\it Chandra} and {\it XMM-Newton} data, we observe the same
134: trends in surface brightness distribution for the optical and X-ray images.
135: We carry out the spectral analysis of three regions that represent the most
136: significant optical features.
137:
138:
139:
140:
141: \end{abstract}
142:
143: \keywords{ISM: individual(\objectname{N49}) -- supernova remnants --
144: X-rays: individual(\objectname{SGR 0526-66})}
145:
146: %% From the front matter, we move on to the body of the paper.
147: %% In the first two sections, notice the use of the natbib \citep
148: %% and \citet commands to identify citations. The citations are
149: %% tied to the reference list via symbolic KEYs. The KEY corresponds
150: %% to the KEY in the \bibitem in the reference list below. We have
151: %% chosen the first three characters of the first author's name plus
152: %% the last two numeral of the year of publication as our KEY for
153: %% each reference.
154:
155: \section{Introduction}
156: %\input{intro.tex}
157:
158: N49 is the most optically bright supernova remnant (SNR) in the Large Magellanic Cloud (LMC). In contrast to a classic shell-type SNR, N49 exhibits largely asymmetric morphology at optical wavelengths. The increased brightness of this SNR in the southeast and its unique filamentary structure at optical wavelengths has long set N49 apart from other well-understood SNRs. Inhomogeneities in the optical morphology and kinematics of N49 have been observed in the past \citep{Shull83,CK88,V92}. However, the fine structure resolved from these kinematic observations was difficult to reconcile with the optical images taken from the ground. Recent mapping of molecular clouds in the region suggests that this SNR is expanding into a denser region toward the southeast, which has caused the asymmetric brightness enhancements in the shell \citep{B97}.
159:
160: The presence of such a filamentary structure suggests that dense material surrounded the progenitor star at the time of its supernova. In the case of a massive progenitor star, its strong stellar wind would have cleared most material from the environment during its lifetime. Considering the evidence for a relatively dense environment surrounding N49, a B-type star, which would lack a strong stellar wind, is the most likely candidate for the N49 progenitor \citep{Shull83}.
161:
162: This SNR also appears notably bright at the X-ray \citep{RW99, Park03} and radio \citep{D95} wavelengths, in which the complete spherical shell of the remnant is clearly visible. While the overall structure appears symmetric at X-ray wavelengths, the brightness peaks in the southeastern region as in the optical. In the {\it Chandra} observations, most of the optically bright filamentary structures have only diffuse X-ray counterparts. These observations also revealed a hard X-ray knot extending beyond the main shell in the southwest of N49, which has been suggested to be a fragment of the explosion accelerated beyond the shock boundary.
163:
164: An additional point of interest is the existence of a known soft gamma-ray repeater, SGR 0526-66, within the boundary of N49. Recent {\it Chandra} images have resolved a point source, suggested to be the X-ray counterpart to this SGR \citep{K03}. No optical or radio counterparts of this SGR have been detected \citep{D95,Setal98,Fetal98}, and the SGR does not seem to have strong influence on the overall optical morphology of the remnant.
165:
166:
167: In order to understand the kinematics of the remnant and the distribution of X-ray emission in terms of the detailed physical structure of the N49 SNR, we have obtained high-resolution {\HST} ({\it HST}) images in the H$\alpha$, [\oiii], and [\sii] emission lines, which reveal N49's bright filamentary structures with high precision. We have supplemented existing high-dispersion spectroscopy \citep{CK88} with additional deep spectra for N49 to investigate the full extent of velocity distribution. In this paper, we analyze the above observations in conjunction with archival {\it Chandra} and {\it XMM-Newton} observations for a comprehensive multi-wavelength study of the physical structure of N49. \S 2 describes the included observations, \S 3 analyzes the {\it HST} emission line images and the implications of the surface brightness and the line ratios, \S 4 analyzes the nebular kinematics and correlates the kinematic and morphological features, and \S 5 compares the X-ray and optical morphologies and discusses the relative distribution of hot and cool gas, and a summary is given in \S 6.
168:
169: \section{Observations}
170: %\input{observ.tex}
171:
172: %% In a manner similar to \objectname authors can provide links to dataset
173: %% hosted at participating data centers via the \dataset{} command. The
174: %% second curly bracket argument is printed in the text while the first
175: %% parentheses argument serves as the valid data set identifier. Large
176: %% lists of data set are best provided in a table (see Table 3 for an example).
177: %% Valid data set identifiers should be obtained from the data center that
178: %% is currently hosting the data.
179:
180:
181: \subsection{Optical Emission-Line Images}
182:
183: We imaged N49 using the {\hst} Wide Field Planetary
184: Camera 2 (WFPC2) with three filters: F656N
185: (H$\alpha$), F673N ([\ion{S}{2}] $\lambda\lambda$6717, 6731) and F502N
186: ([\ion{O}{3}] $\lambda$5007). The dates and exposure times of these observations are presented in Table 1. The data were reduced using the \textsc{stsdas} package within \textsc{iraf}\footnote{Image Reduction
187: Analysis Facility, maintained by NOAO}.
188: The continuum was not subtracted, as the line emission was so dominant
189: that the presence of continuum would not affect our conclusions.
190: Cosmic-ray events were removed by
191: combining multiple exposures, and the resulting files were bias-subtracted.
192: Each image was divided by its exposure time to produce a count-rate map.
193: These maps in turn were multiplied by a conversion factor, given by the
194: \textsc{photflam} parameter in the image header, to produce flux-density
195: maps. We used the \textsc{synphot} task to determine widths for each filter,
196: and multiplied the flux-density maps by the corresponding filter widths to produce flux
197: maps. The resulting files were then mosaicked together for the final images.
198:
199: \subsection{Optical Spectroscopy}
200:
201:
202: We obtained high-dispersion spectra of N49 with the echelle
203: spectrograph on the Blanco 4m telescope at Cerro Tololo Inter-American
204: Observatory (CTIO) in three observing runs.
205: A journal of our observations is given in Table 2.Ê We used the
206: 79 line mm$^{-1}$ echelle grating in the single-order, long-slit
207: observing configuration, with a flat mirror replacing the cross
208: disperser and a post-slit \ha\ filter ($\lambda_c$ = 6563 \AA,
209: $\Delta \lambda$ = 75 \AA) inserted to isolate a
210: single order.
211:
212: Due to the vignetting limitations imposed by the optics, the useful
213: spatial coverage is $\sim$3$'$.
214: The spectral coverage includes both the \ha\ line and the neighboring
215: [\nii]~$\lambda\lambda$6548, 6583 lines. We used a slit width of 250
216: $\mu$m (1\farcs64) for all observations.
217: The spectral dispersion and wavelength were calibrated by a Th-Ar lamp
218: exposure taken in the beginning of each night, and fine-tuned using the
219: geocoronal \ha\ and telluric OH lines present in the echellograms.
220:
221: The 1986 observing run used the Air Schmidt camera and a
222: GEC 385$\times586$ CCD. The 22 $\mu$m pixel size corresponds
223: to about 0.21 \AA\ and 0\farcs635 along the dispersion and spatial axes,
224: respectively.
225: The instrumental profile had a FWHM of $21\pm0.5$ km s$^{-1}$,
226: as measured from the Th-Ar calibration and sky lines.
227: The echelle spectra obtained in this observing run have been reported by
228: \citet{CK88}.
229:
230: The 1995 observing run used the red long-focus camera and a Tek
231: 2048$\times$2048 CCD.
232: The 24 $\mu$m pixel size corresponds to roughly 0.082 \AA\ and
233: 0\farcs26 along the dispersion and spatial axes, respectively.
234: The instrumental profile had a FWHM of $16.1\pm0.8$ km s$^{-1}$.
235:
236: The observations made in 2000 were recorded using the red long-focus
237: camera and the SITe2K\#6 2048$\times$2048 CCD.
238: The 24 $\mu$m pixel size corresponds to roughly 0.082 \AA\ and
239: 0\farcs26 along the diespersion and spatial axes, respectively.
240: The instrumental profile had a FWHM of $13.5\pm0.5$ km s$^{-1}$.
241:
242:
243:
244: \subsection{X-ray Images and Spectroscopy}
245:
246: X-ray observations of N49 were made with both the \xmm\ and \chandra\
247: observatories. We retrieved the pipeline-processed \xmm\ data from the
248: Science Operations Centre (SOC). Observations were made simultaneously
249: with multiple \xmm\ instruments; in this paper we will concentrate on
250: the European Photon Imaging Camera (EPIC) MOS and pn detectors. The
251: data were taken in 2001 April (Observation ID 0113000201; 26.9 ks;
252: PI: Kaastra) over multiple time intervals, using the Thick filter for two
253: EPIC-MOS1 exposures and the Medium filter for the remaining EPIC-MOS
254: and EPIC-pn exposures.
255: Initial reduction and analysis were carried out using the
256: Science Analysis Software package provided by the SOC, with
257: subsequent spectral analysis in \textsc{xspec}.
258:
259: The data were filtered to remove high background times or poor event
260: grades and merged to produce a single dataset for each instrument with
261: effective exposure times of 14.3 ks for EPIC-MOS1,
262: 19.5 ks for EPIC-MOS2 and 0.8 ks for EPIC-pn. Images were
263: then extracted from the filtered merged event files.
264: Spectra were extracted separately from each exposure, to avoid introducing
265: distortions due to the different filters used.
266: Background regions immediately surrounding the SNR but free of point
267: sources were used to produce background spectra, which were then scaled
268: and subtracted from the source spectra. Spectra from different
269: exposures were jointly fitted.
270:
271: \chandra\ observations used the Advanced CCD Imaging
272: Spectrometer (ACIS), primarily the S3 back-illuminated chip. Datasets
273: were sequence number 500107, observation 1041 (34.3 ks; PI: Garmire) and sequence
274: number 500134, observation 2515 (7.0 ks; PI: Kulkarni). Data were reprocessed
275: following procedures recommended by the Chandra X-ray Center:
276: we removed afterglow corrections, generated new bad pixel files, and
277: applied corrections for charge-transfer inefficiency and
278: time-dependent gain for an instrument temperature of 120 K. The two
279: datasets were filtered for high-background times and poor event
280: grades, resulting in total ``good time" intervals of 33.9 and 6.9 ks.
281:
282: Spectral results were extracted from the new event files. Background regions
283: were taken from source-free areas surrounding the SNR, and the spectra
284: of these background regions were scaled and subtracted from the source
285: spectra. Individual spectra for regions of interest, selected by comparison
286: of optical and X-ray images, and the corresponding primary and auxiliary
287: response files were extracted with the \textsc{ciao} acisspec script and
288: analyzed in \textsc{xspec}.
289: Spectra were rebinned by spectral energy to achieve a signal-to-noise ratio
290: of 4 in each bin.
291:
292:
293:
294: \section{Analysis of the HST Emission Line Images}
295: %\input{newopti2.tex}
296: \subsection{H$\alpha$ Surface Brightnesses and Electron Densities}
297:
298: The {\it HST} H$\alpha$ image (Fig. \ref{fig:has2o3}a) displays a complex
299: network of filaments with the brightest emission located toward the southeast.
300: In Figure \ref{fig:has2o3}b, we have marked the morphological features of the
301: remnant, which will henceforward be referred to as features i to vi.
302: The most prominent feature is the SE Ridge (feature i), a ridge of bright
303: filaments in the southeast quadrant, tracing a line from east to south
304: of the SNR.
305: From the SE Ridge, straight bright filaments extend toward the northwest over
306: lengths of 15--30$''$ (ii).
307: A network of bright filaments is present in the southwestern part of the
308: remnant (iii), where the
309: filaments are shorter than those in the SE Ridge.
310: Beyond this network toward the outer edge of the remnant, there exist a
311: number
312: of fainter filaments (iv).
313: Faint filaments are also seen in the northern and northeastern parts of the
314: SNR (v). No bright filaments exist in the western part of the remnant.
315:
316: The bright filaments are enveloped by a diffuse H$\alpha$
317: emission component.
318: This diffuse emission extends beyond the
319: radio/X-ray \citep{RW99, Park03, D95} boundary along the circumference from
320: the north through the east to the south.
321: On the west side of the remnant, only faint patches of H$\alpha$
322: emission exist (feature vi).
323: While the outermost emission patches follow the X-ray/radio
324: boundary, they
325: do not appear to be
326: embedded in a diffuse emission component, which is either absent
327: or below the detection limit.
328:
329: %Since the diffuse H$\alpha$ emission extends beyond the shock boundary,
330: %it is produced by recombination of hydrogen
331: %ionized by the UV precursor in the pre-shock region.
332: Due to its presence beyond the shock boundary, we identify the diffuse
333: H$\alpha$ emission as the recombination radiation of hydrogen ionized
334: by the UV precursor in the pre-shock region.
335:
336: The H$\alpha$ surface brightness ($SB$) of a 10$^4$ K ionized
337: gas can be expressed as
338: \begin{equation}
339: SB = 1.9 \times 10^{-18} n_{\rm e}^2~ L_{\rm pc} ~~
340: {\rm ergs ~s^{-1} ~cm^{-2} ~arcsec^{-2}},
341: \end{equation}
342: where $n_{\rm e}$ is the electron density in cm$^{-3}$ and
343: $L_{\rm pc}$ is the emitting path length in pc.
344: Using the flux-calibrated {\it HST} H$\alpha$ image\footnote{The WFPC2
345: H$\alpha$ filter (F656N) also transmits the [\nii] $\lambda$6548 line,
346: which is $\sim$8\% as bright as the H$\alpha$
347: line in N49 \citep{Vancura92M}. This [\nii] contamination has been corrected
348: for.}
349: and reasonable estimates
350: of the emitting path lengths, we may determine the gas densities.
351: For a filament, we adopt its length as upper limit and its width as lower
352: limit for the emitting path length.
353: For a diffuse emission region, we adopt its average diameter as the emitting
354: path length.
355:
356: In the brightest filaments, such as the SE Ridge,
357: the long
358: filaments extending to the northwest, and the bright
359: short filaments in the southwest,
360: $SB$ is $\sim (1 \pm 0.4) \times 10^{-13}$
361: ergs s$^{-1}$cm$^{-2}$ arcsec$^{-2}$, with the peak emission
362: reaching
363: $2 \times 10^{-13}$ ergs s$^{-1}$cm$^{-2}$arcsec$^{-2}$.
364: The rms densities in these bright filaments are between the limits of
365: 265 $\pm$ 20 and 515 $\pm$ 20 cm$^{-3}$, adopting the length and width of the
366: filaments as the emitting path lengths, respectively.
367:
368: The filaments in the central and northeastern regions, as well as the
369: brighter northern filaments, are slightly fainter than the SE Ridge,
370: with $SB$ of
371: $(3-8) \times 10^{-14}$ ergs s$^{-1}$cm$^{-2}$arcsec$^{-2}$.
372: The rms electron densities in these filaments are between
373: the limits of 115 $\pm$ 20 and 310 $\pm$ 20 cm$^{-3}$.
374: The $SB$ of the fainter filaments in the remnant, such as the northernmost
375: filaments, is $(0.9 - 2) \times 10^{-14}$
376: ergs s$^{-1}$cm$^{-2}$arcsec$^{-2}$.
377: The rms densities here are between the limits of 70 $\pm$ 20 and 225 $\pm$
378: 25 cm$^{-3}$.
379:
380: The faint patches in the western quadrant have $SB$
381: of $(1-4) \times 10^{-15}$ ergs s$^{-1}$cm$^{-2}$arcsec$^{-2}$.
382: Adopting their average diameters as the emitting path lengths, the rms
383: densities
384: in the patches are 90 $\pm$ 10 cm$^{-3}$.
385: The diffuse H$\alpha$ emission that envelops the bright filaments has a
386: similar $SB$, but a larger emitting path length.
387: Approximating the emitting path length by the radius of the remnant,
388: we get rms densities of about 10 $\pm$ 5 cm$^{-3}$.
389:
390:
391: Our derived rms densities are significantly lower than the densities of
392: 1000 -- 1800 cm$^{-3}$ calculated using
393: the [\sii] doublet ratios reported by \citet{Vancura92M}.
394: This difference can be explained by the different weightings
395: placed on density by these two methods.
396: The flux of an emission line is proportional to $\int n_{\rm e}^2~dL$,
397: thus an over-dense region along a line of sight will dominate
398: the emission and dictate the electron density determined from the
399: [\sii] doublet ratio.
400: On the other hand, the rms density is derived from the emissivity
401: averaged over the entire emitting path length, and therefore is
402: always lower than the peak value determined from the [\sii] doublet.
403: In general, the density determined from the [\sii] doublet is
404: representative of the densest regions, while the rms density better
405: represents the density over the entire emission path length.
406: The rms densities are particularly useful for region whose density
407: is $<$ 100 cm$^{-3}$, the low-density limit of the [\sii] doublet, such as
408: the faint emission regions.
409:
410: %below the low-density
411: %limit of the [\sii] doublet, 100 cm$^{-3}$,
412: %such as the faint emission regions.
413: %This discrepancy can be partly attributed to the difference in the regions
414: %selected for the density calculations.
415: %While Vancura et al. sampled the regions
416: %in which the [\sii]/ H$\alpha$ ratio is high, we have measured the surface
417: %brightness in the regions bright in H$\alpha$ that turn out to have relatively
418: %low [\sii]/ H$\alpha$ ratios (see $\sec$ next section).
419: %The regions with lower densities derived from H$\alpha$ $SB$ would require
420: %higher temperatures to maintain pressure equilibrium. At these temperatures,
421: %the
422: %H$\alpha$ recombination coefficient would however be lower, and therefore the
423: %observed $SB$
424: %would lead to higher emission measure and thus higher electron densities.
425: %Assuming a T of 20000 K in the H$\alpha$ bright regions, we can increase the
426: %measured densities by a factor of $\sqrt{2}$.This would somewhat attenuate the
427: %discrepancy between our calculation and Vancura et al.
428: %%(our highest densities would be 750)
429: %The observed surface brightnesses are
430: %consistent with the presence of a dense molecular cloud SE of the remnant
431: %\citep{B97}.
432:
433: \subsection{[\sii] Emission and [\sii]/H$\alpha$ Ratios }
434:
435:
436: Overall, the morphology of N49 in the [\sii] line (Fig. \ref{fig:has2o3}c)
437: resembles that in the H$\alpha$ line.
438: For example, the bright SE Ridge, the straight bright filaments extending to
439: the northwest, the fainter filaments in the north and west, and the faint
440: emission patches in the west are all observed in the [\sii] image.
441: The most notable morphological difference between the [\sii] and H$\alpha$
442: images is the lack of a diffuse [\sii] emission component enveloping
443: the bright filaments.
444:
445: As the morphologies of N49 in the H$\alpha$ and [\sii] lines are similar,
446: the ratio map (Fig. \ref{fig:has2o3}d) is fairly uniform, with a mean value
447: of 0.75 $\pm$ 0.05.
448: However, there are variations in this ratio, giving an overall range of
449: 0.4 -- 1.4. These variations result from different shock conditions.
450: The highest ratios are found in the bright filaments extending northward from
451: the SE Ridge, where the ratios reach values of 1 -- 1.4.
452: The lowest ratios can be seen in the outermost limb of the remnant, where
453: the ratios are below 0.4, which may be due to the lack of a diffuse [\sii]
454: component.
455: In addition, we find low ratios in the SE Ridge, where the H$\alpha$
456: filaments are the brightest.
457:
458:
459: %--------------------------------------------------------
460: \subsection{[\oiii] Emission and [\oiii]/H$\alpha$ Ratios}
461:
462: %While in general we see cospatial H$\alpha$ and
463: %[\sii] emission, the correlation between these two and the [\oiii]
464: %emission is weaker.
465: %Even though
466: The [\oiii] emission (Fig. \ref{fig:has2o3}e) is filamentary and peaks in the
467: southeast just like the H$\alpha$ and [\sii], but the filamentary
468: structures differ in details.
469: The [\oiii] emission associated with bright H$\alpha$ and
470: [\sii] filaments is usually concentrated in faint thin filaments
471: and shows an offset toward the leading edge of the shock fronts.
472: The most extreme difference is exhibited in the bright H$\alpha$
473: filament extending from the northern tip of the SE Ridge toward
474: northwest, where the [\oiii] emission is absent.
475:
476: These and other morphological differences are best demonstrated
477: by the [\oiii]/H$\alpha$ ratio map (Fig. \ref{fig:has2o3}f).
478: The ratios generally increase from the inner to the outer part of the remnant,
479: spanning a wide range from nearly 0 to about 2.5.
480: The ratios in the inner portions of the remnant range from 0 to about 0.5,
481: with a few regions reaching ratios of 0.7.
482: The ratio map in the inner portions of the remnant shows a filamentary
483: structure that is not as sharp as the direct images.
484: The brighter and fainter ``filaments'' on the ratio map have values of
485: 0.4 -- 0.7, and 0.05 -- 0.15, respectively.
486: Generally, the filaments bright in H$\alpha$ have very low [\oiii]/H$\alpha$
487: ratios. This can be seen in the SE Ridge, where the brightest
488: H$\alpha$ filaments are faint on the ratio map.
489: Ratios above 0.5 are found primarily in the outer portions of the remnant.
490: The most prominent feature of the ratio map is the increase in the ratios
491: toward the outermost edge of the remnant, where
492: they reach values between 0.8 and 1.5. A few small patches in the north,
493: northeast and southwest reach the values of up to 2.5 on their outer edge.
494:
495: The enhanced [\oiii]/H$\alpha$ ratios at the edges of the SNR
496: appear to be at least as much due to a diminishment in the H$\alpha$ emission
497: as to an increase in [\oiii] emission.
498: This morphology demonstrates an offset of [\oiii] emission toward the outer
499: edge of the remnant, seen also in the composite image of H$\alpha$, [\sii]
500: and [\oiii] emission (Fig. \ref{fig:color}b).
501: We evaluate the offset by comparing the peaks of
502: brightness profiles across the filaments in [\oiii] and H$\alpha$.
503: Sample SB profiles demonstrating this offset are shown in
504: Fig. \ref{fig:sbprofile}.
505: We find the offsets to range from 0 up to $\sim$ 0\farcs5, or 0.125 pc
506: (for a distance of 50 kpc; Feast 1999).
507:
508:
509: % in Halpha, we have less than 10% contamination by NII
510: % 4% unc in density. Ha SB should be reduced by 8% b/c N2 contributes
511: % 8% to the
512: % N6548 shifted to the band, contributes about 8% to SB. This has been
513: % corrected
514: On small scales, we would expect an offset between the [\oiii] and
515: H$\alpha$ filaments because the emissivity and ionic concentration
516: of a given line depend on the temperature and thus the distance
517: from the shock front.
518: In a plain, steady-shock model, the hydrogen recombination emission peaks
519: roughly where [\oiii]
520: and [\oii] emission diminishes and [\oi] emission rises \citep{Cox72}.
521: The time scale needed for a post-shock region to evolve from
522: [\oiii]-bright to [\oiii]-faint is $\sim$2000/$n_0$ yr, where $n_0$
523: is the pre-shock density in units of H-atom cm$^{-3}$ \citep{Raymond80}.
524: Therefore, we expect the post-shock [\oiii] emission peak and the
525: H$\alpha$ emission peak to exhibit a displacement of
526: $v \times (2000 {\rm yr}/n_0)$, where $v$ is the velocity
527: of the post-shock gas with respect to the shock front.
528: In case of an adiabatic (Sedov) shock, the post-shock gas expands downstream
529: with $\onequarter$ of the shock velocity; thus the shocked gas in the clouds
530: is observed to move with $v_c = \threequarters v_s$, where $v_s$ is the shock
531: speed.
532:
533: For a dense cloud with $n_0$ = 1000 cm$^{-3}$ (the lower limit for the filament
534: density calculated by \citet{Vancura92M}),
535: with an observed expansion velocity $v_c$ = 250 km~s$^{-1}$, the displacement
536: between [\oiii] and H$\alpha$
537: will be $1.7\times10^{-4}$ pc, or 0\farcs0007, which cannot be resolved
538: by the {\it HST}.
539: For the shock with the same velocity driven into the diffuse regions of N49,
540: where the density is about $n_0$ of 10 cm$^{-3}$,
541: the offset between [\oiii] and H$\alpha$ can be raised to 0.017 pc,
542: or 0\farcs07. This value still cannot be resolved by the {\it HST}.
543:
544: There are several other scenarios that would produce an offset between [\oiii]
545: and H$\alpha$ emission, each involving a departure from a steady-flow
546: shock model.
547: %In the first scenario, the offsets arise from the thermal instabilities
548: %in the region behind a radiatively cooling shock.
549: %The flow behind radiatively cooling shocks is unstable \citep{McCray75}, and
550: %\cite{Innes92} points out that the shocks with intermediate velocities are
551: %indeed subject to these thermal instabilities.
552: %This would result in differences
553: %of emission in regions with different temperatures \citep{Raymond80}.
554: %\cite{Innes92} further suggests that the instabilities can be dampened or
555: %suppressed by magnetic pressure, and that the observed offsets are due to the
556: % presence of secondary shocks.
557: Using observations from the MagIC camera at the Clay Telescope,
558: \cite{Rakowski07} compared the locations of the [OIII] and H$\alpha$ peaks,
559: finding separations of up to 0\farcs62. Our HST observations, which do not
560: suffer
561: from seeing distortions, show peak separations of up to 0\farcs5.
562: \cite{Rakowski07} attributed these offsets to thermal instabilities dampened by
563: magnetic field,
564: following the models of \cite{Innes92}, who associates the offsets with the
565: presence of secondary shocks.
566: These models have predicted maximum offsets of 3 $\times$ 10$^{17}$ cm,
567: %20\farcs $\times$ ($D$/1 kpc)$^{-1}$
568: which for the distance of the LMC yields roughly 0\farcs4 \citep{Rakowski07}.
569: However, the physical conditions for the models ($v_s$=175
570: km s$^{-1}$, $n_0$ = 1 cm$^{-3}$, $B_0$ = 3 $\mu$G) are different from the ones
571: in the LMC.
572: As \citet{Rakowski07} point out, the abundances of the LMC will increase the
573: cooling time by a factor of 3, which would tend to increase the offset by
574: this factor. Our derived pre-shock density is 10 cm$^{-3}$, which is 10 times
575: higher than the value used for the models, so the net result of
576: these 2 effects is a decrease in the offset by a factor of 10/3, giving an
577: offset of roughly 0\farcs15.
578: %sth about magntic field?
579: \citet{Rakowski07} argued that the fast shocks seen in N49 made this the most
580: probable scenario. However, without a substantial amount of magnetic support,
581: the derived values are still smaller than the maximum observed offsets.
582:
583: %The largest observed offsets between [\oiii] and H$\alpha$ peaks
584: %can be explained by shocks propagating into an inhomogeneous medium.
585: An alternate scenario was presented by Raymond et al. (1980, 1983) for such
586: offsets in the Cygnus Loop. In this scenario, the offsets arise due to
587: incomplete recombination zones behind a shock front driven into an
588: inhomogeneous environment.
589: For a SNR in a cloudy medium, the slow shocks driven into dense
590: clouds produce the [\oiii] and H$\alpha$ emission, while the fast
591: shocks propagating into the less-dense intercloud medium produce
592: X-ray emission. The H$\alpha$ and [\oiii] lines are both emitted
593: from shocked dense clouds, but their ``turn-on`` times are different
594: by $\sim$2000/$n_0$ yr. As the fast shocks advance through the
595: intercloud medium and encounter clouds, [\oiii] emission will be
596: produced almost immediately in the freshly shocked clouds, and
597: the H$\alpha$ emission is produced by clouds that were shocked
598: $\sim$2000/$n_0$ yr earlier. The displacement between the [\oiii]
599: and H$\alpha$ peaks is thus $\sim (v_{s,i} - v_{s,c}) \times 2000
600: {\rm yr}/n_0$, where $v_{s,i}$ is the velocity of a fast shock
601: through the intercloud medium and $v_{s,c}$ is the velocity
602: of a slow shock through a dense cloud \citep{Fesen82}.
603:
604: The velocity of the fast shock in the intercloud medium can be estimated
605: from the temperature of the post-shock, X-ray emitting gas, using
606: \begin{equation}
607: kT = \frac{3}{16} \mu m_{\rm H} v_{s,i}^{2} ,
608: \label{for:kT}
609: \end{equation}
610: where $m_{\rm H}$ is the mass of the hydrogen atom and $\mu$ is the mean
611: molecular weight.
612: For a fully ionized gas with He:H = 1:10, $\mu$ is 0.6.
613: Adopting the higher temperature component $kT$ = 1.0 keV \citep{Park03}, we
614: get a $v_{s,i}$ = 920 km s$^{-1}$.
615: The highest velocities seen in the dense shocked clouds, obtained from the
616: echelle spectra (see \S 4.), are about 250 km s$^{-1}$. The post-shock gas in
617: the clouds was accelerated to 0.75$v_{s,c}$.
618: %As the shocked cloud has already been cooled sufficiently, material must have
619: %been gradually accelerated to near 0.75$v_{s,c}$.
620: The pre-shock density is 10 cm$^{-3}$. Therefore, the expected
621: displacement between the [\oiii] and H$\alpha$ peaks is 0.12 pc, or 0\farcs5.
622: This offset is in excellent agreement with the observation.
623:
624:
625:
626:
627: \section{Analysis of Filament Velocity Structure with Echelle Spectra}
628: %\input{echelle2.tex}
629: To determine the exact slit positions of the echelle observations, we match
630: the $SB$ profiles along the slit to those extracted from the {\it HST}
631: H$\alpha$ image.
632: The high angular resolution of the {\it HST} images allows us for
633: the first time to see one-to-one correspondence between the kinematical and
634: morphological features of N49 (Fig. \ref{fig:echelle}).
635:
636: \subsection{Discrete Features}
637:
638: The echellograms in Figure \ref{fig:echelle} show
639: %the following kinematic features:
640: discrete broad emission features as well as a narrow component extending continuously along the slit.
641: %\begin{enumerate}
642: %narrow faint
643: %\item
644: The narrow component, detected by \citet{CK88} and \citet{Vancura92M} shows
645: $SB$ enhancements within the boundary of the SNR, indicating physical association; we therefore adopt its heliocentric velocity ($v_{\rm hel}$) of 315 km s$^{-1}$ as the SNR's systemic velocity.
646: Due to our extended slit coverage, we are able to pin down the location of this enhanced narrow component in the SNR. The $SB$ variation of this narrow component is similar to that of the diffuse H$\alpha$ emission, brighter in the east and fainter in the west.
647: We therefore associate this narrow component with the faint diffuse H$\alpha$ component identified morphologically in \S 3.1.
648: The narrow velocity width and the pre-shock location of this narrow compnent
649: are consistent with its origin as recombination radiation of hydrogen ionized
650: by the UV precursor of the SNR shocks.
651:
652: % MEASURE THE DISPERSION OF THE NARROW COMPONENT
653:
654: %\item
655: We also observe broad emission features which indicate the bulk expansion of about 250 km s$^{-1}$.
656: Some of the broad emission features have a ``head-tail'' structure; the bright emission at a head has a smaller radial
657: velocity offset ($\Delta v$) from the systemic velocity and the fainter emission in the tail stretches to larger $\Delta v$. Among the discrete broad emission features, the brighter ones have smaller $\Delta v$,
658: ranging from 0 to 160 km s$^{-1}$ at the ``head'' and up to about
659: 250 km s$^{-1}$ at the ``tail''.
660: %most of which are $\sim$ 0 kms$^{-1}$, some up to about 160 kms$^{-1}$.
661: They can be well matched to individual bright filaments.
662:
663: The structure of these emission features can be explained by the interaction between the shock and the encountered dense gas.
664: The shock gets dampened more by the dense gas in the filaments than in the lower-density cloudlets.
665: Therefore, the post-shock gas in the inner part of the dense bright regions was propelled to velocities significantly lower than the original shock velocity, thus producing smaller velocity spreads.
666: The presence of gas at systemic velocity suggests that the gas in these regions retarded the shock sufficiently to prevent its propagation throughout the whole filament, leaving the innermost gas unshocked.
667: %The dense bright filamnets were shocked to lower velocities than the cloudlets with lower densities, and thus they exhibit lower velocity offsets.
668: %The presence of gas at the systemic velocity suggests that some of the gas in the dense filaments has not been shocked yet.
669: Because it emits the H$\alpha$ recombination radiation, it must have been ionized by the UV precursor.
670: The fainter ``tail'' emission comes from the outer layers of these dense filaments that encounetered the undampened shock and have been accelerated to higher velocities.
671:
672: %The bright knots of emission in the echelle spectra can be matched to individual filaments. Their velocity spreads are generally lower(up to 160km/s), and most of the emission knots have a lower end of velocity spread at the systemic velocity of the remnant. These dense filamets were shocked to lower velocities than the diffuse component.
673: %component with offsets approaching 200 kms$^{-1}$ from the systemic velocity.
674: %The bright emission knots in 2000-1 all have a brightest component at the systemic velocity of the remnant and are shocked to lower speeds (200km/s) than the fainter emission blobs. The presence of gas at the systemic velocity suggests that it has not been shocked yet; however, as it emits Ha, it mush have been ionized by the UV precursor. The gas near the systemic velocity was not yet shocked by the remnant and thus is not accelerated. The emission from it probably caused by the UV precursor that ionizes the medium ahead of the shock front.
675:
676: %fast & faint
677: %\item
678: The less dense gas, which can be identified on echelle spectra as fainter emission component, can be accelerated by the schock to very high velocities. The lower-density gas is unable to significantly slow down the shock, and the post-shock gas is consequently accelerated to higher speeds. We observe velocity spreads of up to 300 km s$^{-1}$. Some of these cloudlets have no systemic velocity components, which implies that all of the gas is shocked.
679:
680: A few cloudlets have velocity spreads that exceed 300 km s$^{-1}$.
681: One example is a faint emission knot seen at the edge of the remnant in observation E2000-1, which has been marked in Figure \ref{fig:echelle} as feature a.
682: The clump has a radial velocity spread of 350 km s$^{-1}$.
683: Due to its location at the edge of the remnant, it likely has a significant tangential velocity component and thus the actual velocity of the blob is even higher. Assuming the two components to be comparable, the actual velocity of some of the gas in this blob could be up to $\sim$ 500 km s$^{-1}$.
684:
685: %\end{enumerate}
686:
687: \subsection{Overall Kinematic Structure}
688:
689: %ribbons
690: The observed distinct broad emission features have implications for the physical structure of the SNR.
691: The echelle spectra do not suggest the presence of an expanding shell, which would demonstrate itself as a simple bow-shaped line image.
692: %decrease in radial velocity from the center to the edge of the remnant, and divergence of the velocity component into two parts which correspond to the expansion of the shell in the near and the far side of the SNR.%We do not observe the decrease in the radial velocity from the center to the edge of the remnant, neither does the velocity component diverge into two, corresponding to the expansion of the shell in the near and the far side of the SNR.
693: The absence of this expanding shell signature in the echelle spectra implies that the observed filaments are not projection effects of an expanding sheet of gas, but actual long ribbon-like structures which indicate clumpiness in the surrounding gas prior to the explosion.
694: %we assume that the observed filaments are actual long ribbons and not projection effects of an expanding sheet of gas. The remnant thus probably had a ribbon-like structure prior to the explosion, possibly due to the presence of a magnetic field.
695:
696: %Balmer-dominated?
697: We have taken echelle spectra in the southern part of the remnant (E2000-2, E2000-3) to see whether the arc-like filament in the south is Balmer line dominated. We can however observe [\nii] $\lambda$6584 line in the spectra, and thus conclude that the arc is not Balmer dominated.
698:
699: %cavity
700: In observations E1986-2 and E2000-1, we observe very few broad component features at systemic velocity around the center of explosion determined from X-ray and radio data. In the blueshifted part of the echelle spectra, we see no broad features at all in E1986-2, and only few faint ones in E2000-1. In the redshifted part, we observe a number of broad features in both spectra, but only one of these in E2000-1 has the lower end of its velocity spread at the systemic velocity. The rest of the broad features are seen at higher velocities. Such spectra suggest the presence of a central cavity; its position matches the X-ray and radio observations. The asymmetry of the emission in this central region implies uneven density distribution of matter prior to the explosion, with higher density at the far side of the remnant.
701:
702: %image
703: The comparison of the echelle spectra with optical morphology of the SNR allows us to see three-dimensional geometry of the remnant. Using the echelle spectra, we were able to discern whether individual filaments are on the approaching or the receding side of the remnant. Assigning blue and red colors to the blue- and redshifted filaments, respectively, we have constructed a 3D image of N49 (Fig. \ref{fig:color}a).
704: %color to the blueshifter filaments in the near side of the SNR, and red to the receding filaments on the far side of the remnant. Using red and blue filters in the GIMP program, we have constructed a 3D image of N49, seen in Fig BLAH.
705: For the most part, we see the near and the far side of the remnant superimposed, but we observe a number of prominent red and blue filaments. The most prominent redshifted filaments are the straight long filaments extending form the SE Ridge northwards. The majority of the outer filaments is blueshifted, for example a filament extending westward from the SE Ridge.
706:
707: \section{Comparison of X-ray and Optical Morphology}
708: %\input{xray.tex}
709: To better trace the relationship of the warm ionized gas in the
710: optically-emitting filaments with the hot shocked gas, we compared
711: the {\hst} images with data from {\it XMM-Newton} and {\it Chandra}.
712: All of the X-ray images show a bright, point-like X-ray source to the
713: north corresponding to the location of the soft gamma-ray repeater
714: SGR 0525-66, as noted by \cite{Rotschild94}. This bright source appears
715: between optically bright features, in a relatively void optical
716: region, and it is not associated with any optical source.
717:
718: The X-ray image of N49 from {\it XMM-Newton} is presented in Figure
719: \ref{fig:xmm}a. The corresponding X-ray contours are overlaid on
720: the H$\alpha$ observation from {\hst} in Figure \ref{fig:xmm}b.
721: The 2$\sigma$ contour
722: provides a nicely defined boundary to the expanding shell. In general,
723: the comparison of {\it XMM-Newton} and optical images reveals a strong
724: relationship between X-ray brightness and the regions of highest optical
725: flux. As with the optical brightness, the X-ray flux declines toward the
726: west, although the western side of the SNR is still detectable. To the
727: southeast, the steeper X-ray contours indicate a substantial X-ray
728: brightening corresponding to the bright SE ridge in optical; this
729: emission appears to be resolved without any obvious point source. Westward from
730: this X-ray peak, the central cavity described in the velocity structure
731: analysis (\S4) appears somewhat brighter in X-rays than do regions
732: to the west and north of the SNR.
733:
734: A recent short {\it Chandra} exposure enables an X-ray/optical comparison
735: at higher resolution. The X-ray image of N49 from {\it Chandra} is
736: presented in Figure \ref{fig:chandra}a. Again, the corresponding X-ray
737: contours are overlaid on the H$\alpha$ observation from {\hst} in
738: Figure \ref{fig:chandra}b for a
739: multiwavelength comparison. As seen for the {\it XMM-Newton} observation,
740: the overall X-ray surface brightness distribution is similar to that in
741: H$\alpha$, in agreement with earlier {\it Chandra} ACIS observations by
742: \citet{Park03}. In particular, the ACIS observations show X-ray features
743: which roughly correspond to the brightest groups of optical filaments,
744: such as the SE ridge and the filaments that project from this region to
745: the northwest and southwest. Similarly, the ``patches" of optical
746: emission on the western face of the remnant are accompanied by a minor
747: X-ray enhancement in that region. The X-ray emission also outlines the
748: optical cavity region, with the emission decreasing toward the interior
749: of the region. Brighter X-ray emission traces the boundary with a radius
750: of curvature similar to that expected from both the optical and echelle
751: data. On the eastern and southern sides of the remnant, the faint
752: X-ray extent matches very well the SNR boundary seen in diffuse
753: H$\alpha$ emission.
754:
755: In addition to this morphological study, we examined the X-ray spectra
756: using {\it XMM-Newton} data for three regions corresponding to significant
757: optical features. These regions include one covering the SE Ridge
758: of optical emission, one along the optically faint region in the northwest,
759: and a third region corresponding to the cavity described in \S4. For each
760: region, spectra were extracted for each of the three \xmm\ instruments
761: (MOS1, MOS2, and pn) and fit jointly with spectral models for a thermal
762: plasma modified by a model of photoelectric absorption.
763:
764: The primary plasma model used is a plane-parallel shock, non-equilibrium
765: ionization model ({\it vpshock} in XSPEC). This model presumes a simple
766: shock structure with a constant postshock electron temperature $kT$.
767: Ionization timescales are modeled as a broad distribution described by
768: the ionization age $\tau=n_e t_s$, where $n_e$ is the post-shock electron
769: number density and $t_s$ is the time since the plasma was shocked. This
770: model provides a reasonable, if simplified, approximation to shocked
771: plasma conditions for relatively high-temperature shocks in limited
772: spatial regions of an SNR \citep{BLR01}.
773:
774: When describing broad spatial regions of an SNR, as we do here, one should
775: keep in mind that the spectra arise from a range of temperatures and densities
776: within that region. A single-temperature approximation is therefore of only
777: limited utility in correctly describing the ``characteristic" physical
778: conditions. In many cases, it is necessary to use at least two linked models
779: at different temperatures in order to better fit a spectrum which is in
780: reality produced by a superposition of shocks in different conditions.
781: However, particularly in smaller regions, too few counts are often present
782: to constrain a large set of parameters, as for instance those created by
783: adding extra temperature components. The model fits are therefore a compromise:
784: we use as few parameters as possible to characterize a given dataset, adding
785: complexity as the number of counts increases sufficiently to require - and
786: support - additional spectral components.
787:
788: The best-fit spectral models for the SNR and for the selected spatial regions
789: are given in Table~4. Listed in this table are the plasma model components
790: used and the fitted values for absorption column density ($N_H$), post-shock
791: temperature ($kT$), abundances of significant elements as a fraction of solar
792: values ($O/O_{\sun}$, etc.); and ionization age ($\tau$). Also listed are
793: normalization values, proportional to the emission measure, for each of the
794: \xmm\ instruments ($A_{MOS1}$, etc). Finaly, for each model fit the statistical
795: goodness-of-fit values are given: reduced $\chi^2$ ($\chi^2_{\rm red}$) and
796: degrees of freedom (dof). Considering the associated errors and the lower
797: spatial resolution of {\it XMM-Newton}, the results agree well with fits to
798: {\it Chandra} ACIS data made by \citet{Park03}.
799:
800: The X-ray spectra reflect the morphological differences between the
801: eastern and western sides of the SNR. To the west, a single-component
802: plasma model can adequately describe the spectrum. Likewise, the cavity
803: region requires only a single plasma component for a reasonable fit. To
804: the southeast, on the other hand, the spectrum requires two plasma
805: components for a reasonable fit ($\chi^2$/dof=1122/743). Notably, a
806: combination of plasma and power-law components provides a considerably
807: poorer fit ($\chi^2$/dof=1489/744), allowing us to rule out, at the 90\%
808: confidence level, the possibility that the high-energy component of
809: the spectrum is due to a nonthermal contribution.
810:
811: \citet{Park03} interpret this two-component fit to emission from
812: the southeastern SNR as follows: the low-temperature component
813: arising from the forward shock moving slowly into dense clouds, and
814: the high-temperature component originating from the reheating of
815: postshocked material by shocks reflected from dense clumps. Based
816: on our velocity analysis, however, we suggest a slight modification
817: of this picture. Rather than having two distinct temperature regimes,
818: this region of the SNR, we suggest, is the location of a distribution
819: of shock velocities. This picture is supported by the broad range of
820: measured optical expansion velocities, as well as by the modelling of
821: optical line emission by \citet{V92}. The two-temperature fit to
822: the X-ray emission, therefore, can be regarded as an approximation
823: to emission at a wide range of temperatures, generated by forward
824: and reflected shocks at various speeds. These variations in local
825: shock speed presumably arise from the encounter of the overall blast
826: wave with material at different densities, as would be expected where
827: the SNR is encountering the molecular cloud \citep{B97}.
828:
829:
830:
831:
832:
833:
834: \section{Summary}
835: %\input{summary.tex}
836:
837: We have conducted a multiwavelength study of SNR N49 using {\it HST} images
838: in H$\alpha$, [\oiii] and [\sii] emission lines, archival {\it Chandra}
839: and {\it XMM-Newton} images, and compared the morphological features to
840: SNR kinematics, using the high-dispersion echelle spectra obtained at CTIO.
841:
842: All emission line images reveal a highly filamentary morphology, with most
843: prominent features being the SE Ridge, long filaments extending NW and a
844: network of filaments in the SE. The optical emission from the western part
845: of the remnant is significantly lower.
846: We also detect a diffuse H$\alpha$ emission component protruding beyond the
847: shock boundary in all regions except for the west, where this emission is
848: absent or undetected. We identify this diffuse component as the recombination
849: radiation from the gas ionized by the UV precursor.
850: Using H$\alpha$ surface brightness, we have estimated the rms electron
851: densities in different regions of the remnant.
852:
853: Good correlation between [\sii] and H$\alpha$ morphology is demonstrated
854: in the uniformity of the ratio map, with the mean value of 0.75 $\pm$ 0.05,
855: and overall range of 0.4-1.4.
856: [\oiii] images and a wide range of values on the [\oiii]/H$\alpha$ ratio map
857: manifest morphological differences between the [\oiii] and H$\alpha$ emission.
858: The most prominent feature, an offset between [\oiii] and H$\alpha$ bright
859: filaments, is demonstrated on the ratio map as an enhancement of
860: [\oiii]/H$\alpha$ ratio toward the outer edge of the image, where ratios
861: reach between 0.8 and 1.5, with a few patches reaching up to 2.5. We observed
862: the highest offset of 0\farcs5.
863: We propose that such an offset would arise in one of the two scenarios:
864: (a) The offset is due to the thermal instabilities cushioned by magnetic fields
865: and the associated secondary shocks. The fast shocks observed in N49 make this
866: a likely scenario; however, a substantial amount of magnetic support is needed
867: to account for the values of maximum observed offsets.
868: (b) The offsets arise due to shocks propagating into a clumpy medium. The
869: fast shocks just encountering the cloud produce the [\oiii] emission, while
870: H$\alpha$ comes from previously shocked clouds that have cooled sufficiently.
871: The offset produced by this model matches the observations.
872: % (fast and slow shocks encountering clouds), but a model with secondary
873: % shocks could also produce an offset such as this.
874:
875:
876:
877: We have also carried out a detailed study of the remnant kinematics using
878: high-dispersion echelle spectra at 7 different slit positions throughout the
879: SNR. The spectra reveal a clumpy velocity structure, with discrete broad
880: emission features matching well the positions of filaments across the
881: slit, with radial velocity spreads reaching up to 160 km s$^{-1}$. We also
882: detect a number of fainter clumps with very high radial velocity spreads that
883: reach 350 km s$^{-1}$, and estimate the actual velocity of these clumps to
884: be up to 500 km s$^{-1}$.
885: In addition, we detect a narrow component extending throughout the slit,
886: and we associate it with the diffuse H$\alpha$ component identified
887: morphologically.
888:
889: We do not observe a bow-shaped signature of an expanding shell in the spectra.
890: This indicates a true filamentary characteristic of the remnant and implies a
891: clumpy structure of the gas surrounding the star prior to the explosion.
892:
893: We also confirm the presence of the cavity in the remnant which matches the
894: X-ray and radio data, and observe the asymmetric expansion of the gas in this
895: region, which we associate with irregular density distribution of the
896: surrounding matter prior to the explosion.
897: Matching discrete emission features to actual filaments allows us to identify
898: the blue- and redshifted filaments. Through a detailed comparison between the
899: echelle spectra and the optical images, we have constructed a rough 3D map of
900: the remnant.
901: %From the shapes of the broad emission features we concluded that the observed
902: %filaments are not just projection effects of expanding spheres, and thus the
903: %morphology of the ISM surrounding the star was likely filamentary prior to
904: %the explosion. We have also constructed a rough 3D map of the remnant.
905:
906: We have compared the archival {\it Chandra} and {\it XMM-Newton} images to the
907: H$\alpha$ morphology. The X-ray morphology of N49 correlates well with the
908: general trends seen in optical wavelengths. Even though we do not observe
909: filamentary structure in X-rays, the brightest emission in X-rays comes from
910: the SE, and the emission diminishes towards the west. The X-ray boundary
911: matches the extent of optical filaments. Unlike in the optical, we observe the
912: whole spherical envelope of the SNR in X-rays.
913:
914: In addition to the images, we have carried out spectral analysis of the
915: remnant. While the cavity region as well as the western side of the remnant
916: can each be fit by a single plasma component, the eastern and SE regions
917: require two component plasma. We rule out non-thermal radiation with 90$\%$
918: confidence.
919: We suggest the 2 plasma component to be an approximation to a range of
920: temperatures generated by forward and reflected shocks at various speeds.
921:
922:
923:
924: %\section{Conclusions}
925:
926: %% In this section, we use the \subsection command to set off
927: %% a subsection. \footnote is used to insert a footnote to the text.
928:
929: %% Observe the use of the LaTeX \label
930: %% command after the \subsection to give a symbolic KEY to the
931: %% subsection for cross-referencing in a \ref command.
932: %% You can use LaTeX's \ref and \label commands to keep track of
933: %% cross-references to sections, equations, tables, and figures.
934: %% That way, if you change the order of any elements, LaTeX will
935: %% automatically renumber them.
936:
937: %% This section also includes several of the displayed math environments
938: %% mentioned in the Author Guide.
939:
940: \acknowledgements
941:
942: The authors gratefully acknowledge the support of the Dean of the College
943: of Liberal Arts and Sciences of the University of Illinois for YHC.
944: RMW and YHC acknowledge support from STScI grant STI GO-8110.
945: This material is based upon work supported by the National
946: Aeronautics and Space Administration under grant NNG05GC97G
947: issued through the Long-Term Space Astrophysics Program. We would like
948: to thank the referee for useful suggestions to improve the paper.
949:
950: %% To help institutions obtain information on the effectiveness of their
951: %% telescopes, the AAS Journals has created a group of keywords for telescope
952: %% facilities. A common set of keywords will make these types of searches
953: %% significantly easier and more accurate. In addition, they will also be
954: %% useful in linking papers together which utilize the same telescopes
955: %% within the framework of the National Virtual Observatory.
956: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
957: %% for information on obtaining the facility keywords.
958:
959: %% After the acknowledgments section, use the following syntax and the
960: %% \facility{} macro to list the keywords of facilities used in the research
961: %% for the paper. Each keyword will be checked against the master list during
962: %% copy editing. Individual instruments can be provided in parentheses,
963: %% after the keyword, but they will not be verified.
964:
965: Facilities: \facility{HST (WFPC2)}, \facility{CXO (ACIS)},
966: \facility{XMM (EPIC)}, \facility{Blanco}.
967:
968: \begin{thebibliography}{}
969:
970: \bibitem[Banas et al.(1997)]{B97} Banas, K.~R., Hughes,
971: J.~P., Bronfman, L., \& Nyman, L.-A.\ 1997, \apj, 480, 607
972: %\bibitem[Blair et al.(2003)]{Blair03}Blair, et al. 2003
973: \bibitem[Blair et al.(2005)]{Blair05} Blair, W.~P., Sankrit,
974: R., \& Raymond, J.~C.\ 2005, \aj, 129, 2268
975: \bibitem[Borkowski et al.(2001)]{BLR01} Borkowski, K.~J.,
976: Lyerly, W.~J., \& Reynolds, S.~P.\ 2001, \apj, 548, 820
977: \bibitem[Chu \& Kennicutt(1988)]{CK88} Chu, Y.-H., \&
978: Kennicutt, R.~C., Jr.\ 1988, \aj, 95, 1111
979: \bibitem[Cox(1972)]{Cox72} Cox, D.~P.\ 1972, \apj, 178, 169
980: \bibitem[Dickel et al.(1995)]{D95} Dickel, J.~R., et al.\
981: 1995, \apj, 448, 623
982: \bibitem[Draine \& McKee(1993)]{Draine93} Draine, B.~T., \&
983: McKee, C.~F.\ 1993, \araa, 31, 373
984: %\bibitem[Fender et al.(1998)]{Fetal98} Fender, R.~P.,
985: %Southwell, K.~A., \& Tzioumis, A.~K.\ 1998, VizieR Online Data Catalog,
986: %729, 80692
987: \bibitem[Feast(1999)]{Feast99} Feast, M.\ 1999, New Views of
988: the Magellanic Clouds, 190, 542
989: \bibitem[Fender et al.(1998)]{Fetal98} Fender, R.~P.,
990: Southwell, K., \& Tzioumis, A.~K.\ 1998, \mnras, 298, 692
991: \bibitem[Fesen et al.(1982)]{Fesen82} Fesen, R.~A., Blair,
992: W.~P., \& Kirshner, R.~P.\ 1982, \apj, 262, 171
993: \bibitem[Ghavamian et al.(2000)]{Ghavamian00} Ghavamian, P.,
994: Hartigan, P., \& Raymond, J.~C.\ 2000, \apjs, 127, 323
995: \bibitem[Innes(1992)]{Innes92} Innes, D.~E.\ 1992, \aap, 256,
996: 660
997: \bibitem[Kulkarni et al.(2003)]{K03}Kulkarni, S.~R.,
998: Kaplan, D.~L., Marshall, H.~L., Frail, D.~A., Murakami, T., \& Yonetoku,
999: D.\ 2003, \apj, 585, 948
1000: \bibitem[McCray et al.(1975)]{McCray75} McCray, R., Kafatos, M.,
1001: \& Stein, R.~F.\ 1975, \apj, 196, 565
1002: \bibitem[McKee \& Ostriker(1977)]{McKee77} McKee, C.~F., \&
1003: Ostriker, J.~P.\ 1977, \apj, 218, 148
1004: %\bibitem[Page et al.(2004)]{Page04} Page, K.~L., Reeves,
1005: %J.~N., O'Brien, P.~T., Turner, M.~J.~L., \& Worrall, D.~M.\ 2004, \mnras,
1006: %353, 133
1007: \bibitem[Park et al.(2003)]{Park03} Park, S., Burrows, D.~N.,
1008: Garmire, G.~P., Nousek, J.~A., Hughes, J.~P., \& Williams, R.~M.\ 2003,
1009: \apj, 586, 210
1010: %\bibitem{2004AdSpR..33..409P} Park, S. et al. 2004, AdSpR, 33, 409
1011: \bibitem[Rakowski et al.(2007)]{Rakowski07} Rakowski, C.~E.,
1012: Raymond, J.~C., \& Szentgyorgyi, A.~H.\ 2007, \apj, 655, 885
1013: \bibitem[Raymond et al.(1980)]{Raymond80} Raymond, J.~C.,
1014: Hartmann, L., Black, J.~H., Dupree, A.~K., \& Wolff, R.~S.\ 1980, \apj,
1015: 238, 881
1016: %\bibitem[Richards et al.(2004)]{Rich04} Richards, G.~T., et
1017: %al.\ 2004, \apjs, 155, 257
1018: \bibitem[Rotschild, Kulkarni \& Lingenfelter (1994)]{Rotschild94} Rotschild,
1019: R.~E., Kulkarni, S.~R., \& Lingenfelter, R.~E.\ 1994, Nature, 368, 432
1020: \bibitem[Shull \& McKee(1979)]{Shull79} Shull, J.~M., \& McKee,
1021: C.~F.\ 1979, \apj, 227, 131
1022: \bibitem[Shull(1983)]{Shull83} Shull, P., Jr.\ 1983, \apj, 275, 611
1023: \bibitem[Shull et al.(1985)]{Shull85} Shull, P., Jr., Dyson,
1024: J.~E., Kahn, F.~D., \& West, K.~A.\ 1985, \mnras, 212, 799
1025: \bibitem[Smith et al.(1998)]{Setal98} Smith, I.~A., et al.\
1026: 1998, Advances in Space Research, 22, 1133
1027: \bibitem[STScI(2001)]{GSC}STScI, 2001; The Guide Star Catalog,
1028: Version 2.2 (GSC2.2)
1029: \bibitem[Vancura et al.(1992a)]{V92} Vancura, O., et al.\
1030: 1992, \apj, 401, 220
1031: \bibitem[Vancura et al.(1992b)]{Vancura92M} Vancura, O., Blair,
1032: W.~P., Long, K.~S., \& Raymond, J.~C.\ 1992, \apj, 394, 158
1033: %\bibitem[Wilkes et al.(1994)]{Wilkes94} Wilkes, B.~J.,
1034: %Tananbaum, H., Worrall, D.~M., Avni, Y., Oey, M.~S., \& Flanagan, J.\ 1994,
1035: %\apjs, 92, 53
1036:
1037: \bibitem[Williams et al.(1999)]{RW99} Williams, R.~M., Chu, Y.-H.,
1038: Dickel, J.~R., Petre, R., Smith, R.~C., \& Tavarez, M.\ 1999, \apjs,
1039: 123, 467
1040:
1041: \end{thebibliography}
1042:
1043: %(don't leave a space between the name and the paranthetical year in the
1044: %citation)
1045:
1046: %Then, wherever you'd cite Chu \& Kennicutt, you'd put in:
1047:
1048: %\citep{CK88} for citing something in parentheses (Chu \& Kennicutt 1988)
1049: %\citet{CK88} for citing something in the text. E.g., according to Chu
1050: % & Kennicutt (1988) blah blah....
1051:
1052: %% Appendix material should be preceded with a single \appendix command.
1053: %% There should be a \section command for each appendix. Mark appendix
1054: %% subsections with the same markup you use in the main body of the paper.
1055:
1056: %% Each Appendix (indicated with \section) will be lettered A, B, C, etc.
1057: %% The equation counter will reset when it encounters the \appendix
1058: %% command and will number appendix equations (A1), (A2), etc.
1059: \clearpage
1060: \begin{deluxetable}{lccccc}
1061: \tablecaption{Optical {\it HST} WFPC2 Observations}
1062: \tablehead{
1063: \colhead{Observation} &
1064: \colhead{Date} &
1065: \colhead{Line} &
1066: \colhead{Filter}&
1067: \colhead{Exposure} &
1068: }
1069: \startdata
1070: % 2000 Dec 6 & \ha\ + \ion{N}{2} & xxx s & E$-$W through ?? \\
1071:
1072: 1 & 2000 Jul 13 & \ha\ $\lambda$6563 & F656N & 2 $\times$ 500 s \\
1073: 2 & 2000 Jul 14 & [\sii] $\lambda\lambda$6716,6731 & F673N & 2 $\times$ 600 s \\
1074: 3 & 2000 Jul 14 & [\oiii] $\lambda$5007 & F502N & 720+637+800 s \\
1075:
1076:
1077:
1078: \enddata
1079: \label{ObsHST}
1080: \end{deluxetable}
1081:
1082: \begin{deluxetable}{lccccc}
1083: \tablecaption{CTIO 4m Echelle Observations}
1084: \tablehead{
1085: \colhead{Observation} &
1086: \colhead{Date} &
1087: \colhead{Exposure} &
1088: \colhead{Location} &
1089: \colhead{Position Angle}
1090: }
1091: \startdata
1092:
1093: %ra/dec, exposure, PA
1094:
1095: E1986-1 & 1986 Nov 20 & 300 s & N49 pos 1 & 90 \\
1096: E1986-2 & 1986 Nov 20 & 300 s & N49 pos 2 & 90 \\
1097: E1986-3 & 1986 Nov 20 & 300 s & N49 pos 3 & 90 \\
1098: E1986-4 & 1986 Nov 20 & 300 s & N49 pos 4 & 90 \\
1099: E1986-5 & 1986 Nov 20 & 300 s & N49 pos 5 & 90 \\
1100:
1101: E1995-1 & 1995 Jan 19 & 2$\times$600 s & N49 IRS pos & 42.5 \\
1102:
1103: E2000-1 & 2000 Dec 10 & 600 s & N49 & 90 \\
1104: E2000-2 & 2000 Dec 10 & 300 s & N49 south & 90 \\
1105: E2000-3 & 2000 Dec 10 & 300 s & N49 south, 3$''$N & 90 \\
1106:
1107:
1108: \enddata
1109: \label{Observations}
1110: \end{deluxetable}
1111:
1112: % normalized fluxes
1113: \begin{deluxetable}{lcccc}
1114: \tablecaption{X-ray fluxes for the regions normalized by area}
1115: \tablehead{
1116: \colhead{ } &
1117: \colhead{Area} &
1118: \colhead{Avg. Flux} &
1119: \colhead{Norm. Flux} & \\
1120: \colhead{} &
1121: \colhead{arcmin$^2$} &
1122: \colhead{ergs cm$^2$ s$^{-1}$} &
1123: \colhead{ergs cm$^2$ s$^{-1}$ arcmin$^2$} &
1124: }
1125: \startdata
1126:
1127:
1128:
1129: SE Limb & .269 & 2.23 $\times 10^{-11}$ & 8.3 $\times 10^{-11}$ \\
1130: NW Limb & .300 & 9.47 $\times 10^{-12}$ & 3.2 $\times 10^{-11}$ \\
1131: Cavity & .097 & 9.54 $\times 10^{-12}$ & 9.8 $\times 10^{-11}$ \\
1132:
1133:
1134: \enddata
1135: \label{NormFluxes}
1136: \end{deluxetable}
1137:
1138: % spectral models
1139: \begin{deluxetable}{l|ccccccccc}
1140: \tabletypesize{\scriptsize}
1141: \tablecaption{Best-fit spectral models to XMM data for N49 and sub-regions}
1142: \tablehead{
1143: \colhead{Parameter} &
1144: \multicolumn{2}{c}{Whole SNR} &
1145: \multicolumn{2}{c}{SE Limb} &
1146: \colhead{NW Limb} &
1147: \colhead{Cavity}
1148: }
1149: \startdata
1150: component & \multicolumn{2}{c}{vpshock + vpshock} & \multicolumn{2}{c}{vpshock
1151: + vpshock} & vpshock & vpshock \\[6pt]
1152: $N_{\rm H}$ (cm$^{-2}$) & \multicolumn{2}{c}{2.9$^{+0.1}_{-0.1} \times10^{21}$}
1153: & \multicolumn{2}{c}{3.2$^{+0.3}_{-0.3} \times10^{21}$}
1154: & 2.3$^{+0.1}_{-0.1} \times10^{21}$ & 2.2$^{+0.1}_{-0.1} \times10^{21}$
1155: \\[6pt]
1156: $kT$ (keV) & 0.41$^{+0.01}_{-0.01}$ & 0.98$^{+0.01}_{-0.01}$
1157: & 0.40$^{+0.03}_{-0.03}$ & 1.02$^{+0.01}_{-0.03}$
1158: & 0.66$^{+0.01}_{-0.02}$ & 0.54$^{+0.01}_{-0.01}$ \\[6pt]
1159: O/O$_\sun$ & \multicolumn{2}{c}{0.24$^{+0.01}_{-0.01}$}
1160: & \multicolumn{2}{c}{0.21$^{+0.01}_{-0.01}$}
1161: & 0.65$^{+0.05}_{-0.05}$ & 0.38$^{+0.03}_{-0.03}$ \\[6pt]
1162: Ne/Ne$_\sun$ & \multicolumn{2}{c}{0.31$^{+0.01}_{-0.01}$}
1163: & \multicolumn{2}{c}{0.29$^{+0.01}_{-0.02}$}
1164: & 0.62$^{+0.06}_{-0.05}$ & 0.38$^{+0.04}_{-0.02}$ \\[6pt]
1165: Mg/Mg$_\sun$ & \multicolumn{2}{c}{0.31$^{+0.01}_{-0.02}$}
1166: & \multicolumn{2}{c}{0.25$^{+0.02}_{-0.03}$}
1167: & 0.30$^{+0.06}_{-0.01}$ & 0.23$^{+0.03}_{-0.03}$ \\[6pt]
1168: Si/Si$_\sun$ & \multicolumn{2}{c}{0.57$^{+0.02}_{-0.02}$}
1169: & \multicolumn{2}{c}{0.65$^{+0.05}_{-0.04}$}
1170: & 0.41$^{+0.05}_{-0.06}$ & 0.48$^{+0.08}_{-0.04}$ \\[6pt]
1171: S/S$_\sun$ & \multicolumn{2}{c}{ 0.67$^{+0.05}_{-0.05}$}
1172: & \multicolumn{2}{c}{ 0.9$^{+0.1}_{-0.1}$}
1173: & 0.7$^{+0.2}_{-0.2}$ & 1.1$^{+0.5}_{-0.1}$ \\[6pt]
1174: Fe/Fe$_\sun$ & \multicolumn{2}{c}{0.34$^{+0.01}_{-0.01}$}
1175: & \multicolumn{2}{c}{0.34$^{+0.01}_{-0.02}$}
1176: & 0.24$^{+0.01}_{-0.02}$ & 0.16$^{+0.01}_{-0.02}$ \\[6pt]
1177: $\tau$ (cm$^{-3}$ s) & 3.1$^{+0.1}_{-0.3} \times10^{11}$ & $>1.0 \times10^{13}$
1178: & 2.8$^{+0.4}_{-0.2} \times10^{11}$ & $>1.2 \times10^{13}$
1179: & 9$^{+1}_{-1} \times10^{11}$ & 1.2$^{+0.2}_{-0.1} \times10^{12}$ \\[6pt]
1180: $A_{\rm MOS1}$ (cm$^{-5}$) & 4.8$^{+0.1}_{-0.1} \times10^{-2}$
1181: & 1.6$^{+0.1}_{-0.1} \times10^{-2}$
1182: & 2.8$^{+0.5}_{-0.4} \times10^{-2}$ & 6.2$^{+0.3}_{-0.2} \times10^{-3}$
1183: & 5.9$^{+0.1}_{-0.2} \times10^{-3}$ & 9.3$^{+0.4}_{-0.3} \times10^{-3}$
1184: \\[6pt]
1185: $A_{\rm MOS2}$ (cm$^{-5}$) & 5.6$^{+0.1}_{-0.1} \times10^{-2}$
1186: & 1.6$^{+0.1}_{-0.1} \times10^{-2}$
1187: & 3.0$^{+0.5}_{-0.4} \times10^{-2}$ & 6.0$^{+0.3}_{-0.3} \times10^{-3}$
1188: & 6.3$^{+0.2}_{-0.2} \times10^{-3}$ & 1.0$^{+0.1}_{-0.1} \times10^{-2}$
1189: \\[6pt]
1190: $A_{\rm pn}$ (cm$^{-5}$) & 5.8$^{+0.1}_{-0.1} \times10^{-2}$
1191: & 1.6$^{+0.1}_{-0.1} \times10^{-2}$
1192: & 3.1$^{+0.4}_{-0.3} \times10^{-2}$ & 6.8$^{+0.2}_{-0.3} \times10^{-3}$
1193: & 6.3$^{+0.3}_{-0.1} \times10^{-3}$ & 1.0$^{+0.1}_{-0.1} \times10^{-2}$
1194: \\[6pt]
1195: $\chi^2_{\rm red}$ & \multicolumn{2}{c}{2.13} & \multicolumn{2}{c}{1.54} & 1.32
1196: & 1.56 \\
1197: dof & \multicolumn{2}{c}{875} & \multicolumn{2}{c}{625} & 485 & 435 \\
1198: \enddata
1199: \tablecomments{Spectra cover the range between 0.3-8.0 keV. A is the
1200: normalization constant, defined as
1201: $A = 10^{-14}\int n_{\rm e}n_{\rm H}dV/(4 \pi D^2)$, where $n_{\rm e}$ and
1202: $n_{\rm H}$ are electron and proton densities, respectively, and $D$ is the
1203: distance, all in cgs units. Elements not listed are fixed to the LMC mean
1204: abundance. Errors give the
1205: 90\% confidence range for each fit parameter. Errors in $N_{\rm H}$, $kT$,
1206: $\tau$ and $A$ are estimated with all abundances except Fe fixed at their
1207: best-fit values; abundance errors are estimated with $N_{\rm H}$, $kT$ and
1208: $\tau$ fixed.}
1209: \label{tab:snrxspec}
1210: \end{deluxetable}
1211:
1212:
1213:
1214:
1215: \clearpage
1216:
1217: %\end{document}
1218:
1219: \begin{figure}
1220: \begin{center}
1221: %\includegraphics[width=0.5\textwidth]{Ha_axes.ps}
1222: %\includegraphics[width=0.43\textwidth]{Ha_rgns.ps}
1223: %\includegraphics[width=0.5\textwidth]{s2_axes.ps}
1224: %\includegraphics[width=0.43\textwidth]{s2ha.ps}
1225: %\includegraphics[width=0.5\textwidth]{o3_axes.ps}
1226: %\includegraphics[width=0.43\textwidth]{o3ha.ps}
1227: \includegraphics[width=1.0\textwidth]{f1.jpg}
1228: \caption{Mosaicked and flux-calibrated {\it HST} WFPC2 images and ratio maps
1229: of N49 displayed in linear greyscale.The panels are (a) H$\alpha$ image,
1230: (b) morphological features marked on the H$\alpha$ image,
1231: (c) [\sii] emission line image, (d) [\sii]/H$\alpha$
1232: ratio map, (e) [\oiii] emission line image, (f) [\oiii]/H$\alpha$ ratio map.}
1233: \label{fig:has2o3}
1234: \end{center}
1235: \end{figure}
1236:
1237: \begin{figure}
1238: \begin{center}
1239: \includegraphics[width=0.9\textwidth]{f2.jpg}
1240: \caption{Surface brightness profiles of H$\alpha$ (thin line) and [\oiii]
1241: (thick line) emission across selected filaments.
1242: The coordinates for the positions marked as zero are
1243: 05$^{\rm h}$26$^{\rm m}$01$\fs$34, $-$66$^{\circ}$05$'$29$\farcs$1
1244: (J2000), at PA = 30$^{\circ}$ for the top panel, and
1245: 05$^{\rm h}$26$^{\rm m}$04$\fs$4, $-$66$^{\circ}$05$'$18$\farcs$7
1246: (J2000) at PA = 135$^{\circ}$ for the bottom panel.}
1247: \label{fig:sbprofile}
1248: \end{center}
1249: \end{figure}
1250:
1251:
1252: \begin{figure}
1253: \begin{center}
1254: \includegraphics[width=0.85\textwidth]{f3.jpg}
1255: \caption{Echellograms of N49 for E2000-1 and E1995-1 data, with corresponding
1256: positions on the H$\alpha$ image marked by the solid horizontal and tilted
1257: lines, respectively.The feature a in the E2000-1 echelle spectrum and the
1258: optical image is a high-velocity clump discussed in \S 4.1. The dashed lines
1259: mark the positions of E1986-1, -2,-3, -4 and -5 (bottom to top) spectra taken
1260: by \cite{CK88}.}
1261: \label{fig:echelle}
1262: \end{center}
1263: \end{figure}
1264:
1265: \begin{figure}
1266: \begin{center}
1267: \includegraphics[width=0.49\textwidth]{f4a.jpg}
1268: \includegraphics[width=0.49\textwidth]{f4b.jpg}
1269: \caption{(a) A 3D cartoon of N49 with blue and red filaments corresponding to the near and the far side of the remnant, respectively. (b) An overlay {\it HST} image of H$\alpha$ (red), [\sii] (yellow) and [\oiii] (blue) emission.}
1270: \label{fig:color}
1271: \end{center}
1272: \end{figure}
1273:
1274:
1275: \begin{figure}
1276: \begin{center}
1277: \includegraphics[width=0.49\textwidth]{f5a.jpg}
1278: \includegraphics[width=0.49\textwidth]{f5b.jpg}
1279: \caption{(a) Smoothed contours of {\it XMM-Newton} X-ray data; (b) The same
1280: {\it XMM-Newton} contours overlaid on the optical H$\alpha$ image from
1281: {\it HST}. This image was binned in pixels of 1$\farcs$6 $\times$ 1$\farcs$6.
1282: Contours plotted on the image range from 3$\sigma$ -- 30$\sigma$
1283: above the background in 5 evenly spaced linearly increasing contours.}
1284: \label{fig:xmm}
1285: \end{center}
1286: \end{figure}
1287:
1288:
1289: \begin{figure}
1290: \begin{center}
1291: \includegraphics[width=0.49\textwidth]{f6a.jpg}
1292: \includegraphics[width=0.49\textwidth]{f6b.jpg}
1293: \caption{(a) Smoothed contours of {\it Chandra} ACIS X-ray data; (b) The same
1294: contours overlaid on the optical H$\alpha$ image from {\it HST}.
1295: This image was processed with a 2 pixel gaussian smoothing. Contours plotted
1296: on the image range from 2$\sigma$ -- 100$\sigma$ above the background in 5
1297: evenly spaced linearly increasing contours.}
1298: \label{fig:chandra}
1299: \end{center}
1300: \end{figure}
1301:
1302:
1303: \end{document}
1304:
1305:
1306:
1307: