1: %
2: % Ver 1.0 30 Aug, 2000 (started July 12, 1999)
3: %
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%55
5: %\documentstyle[12pt,aasms]{article}
6: %\documentstyle[11pt,aaspp]{article}
7: \documentstyle[emulateapj]{article}
8: %
9: \lefthead{Hughes}
10: \righthead{X-Ray Expansion of Tycho's SNR}
11: %
12: \submitted{Received 2000 September 6; accepted 2000 October 6}
13: \accepted{To Appear in the Astrophysical Journal Letters}
14: %\received{}
15: %\submitted{}
16: %
17: \def\einstein{{\it Einstein}}
18: \def\EO{{\it Einstein Observatory}}
19: \def\rosat{{\it ROSAT}}
20: \def\asca{{\it ASCA}}
21: \def\astroe{{\it Astro-E}}
22: \def\xmm{{\it XMM}}
23: \def\chandra{{\it Chandra}}
24: \def\mydegree{^\circ\mskip-5mu}
25: \def\myarcmin{^\prime\mskip-5mu}
26: \def\myarcsec{\mskip1mu^{\prime\prime}\mskip-7mu.\mskip2mu}
27: \def\lsim{\hbox{\raise.35ex\rlap{$<$}\lower.6ex\hbox{$\sim$}\ }}
28: \def\gsim{\hbox{\raise.35ex\rlap{$>$}\lower.6ex\hbox{$\sim$}\ }}
29:
30: \begin{document}
31:
32: \title{The Expansion of the X-ray Remnant of Tycho's Supernova
33: (SN1572)}
34:
35: \author{John P. Hughes}
36: \affil{Department of Physics and Astronomy, Rutgers
37: University, 136 Frelinghuysen Road, Piscataway, NJ 08854-8019
38: jph@physics.rutgers.edu}
39:
40: \begin{abstract}
41:
42: Two \rosat\ high resolution images separated by nearly five years have
43: been used to determine the expansion of the X-ray remnant of Tycho's
44: supernova (SN1572). The current expansion rate averaged over the
45: entire remnant is $0.124\pm0.011 \,\% \,\rm yr^{-1}$, which, when
46: combined with the known age of the remnant, determines the mean
47: expansion parameter $m$, defined as $R\propto t^m$, to be
48: $0.54\pm0.05$. There are significant radial and azimuthal variations
49: of the X-ray expansion rate. The radial expansion in particular shows
50: highly significant evidence for the more rapid expansion of the
51: forward blast wave as compared to the reverse-shocked ejecta, an
52: effect that has not been seen previously. The expansion parameter
53: varies from $m=0.71\pm0.06$ at the outermost edge of Tycho's supernova
54: remnant (SNR) to a value of $m=0.34\pm0.10$ on the inside edge of the
55: bright rim of emission. These values are consistent with the rates
56: expected for a remnant with constant density ejecta evolving into a
57: uniform interstellar medium during the ejecta-dominated phase of
58: evolution. Based on the size, age, and X-ray expansion rates, I
59: obtain values for the explosion energy and ambient density of
60: $E\approx 4-5\times 10^{50}\,\rm ergs$ and $n_0 \approx 0.35-0.45\,
61: \rm cm^{-3}$. As is also the case for Cas A and Kepler's SNR, the
62: X-ray expansion rate of Tycho's SNR appears to be significantly higher
63: than the radio expansion rate. In the case of Tycho's SNR, however,
64: the difference between radio and X-ray expansion rates is clearly
65: associated with the motion of the forward shock.
66:
67: \end{abstract}
68:
69: \keywords{
70: ISM: individual (Tycho's Supernova) --
71: shock waves --
72: supernova remnants --
73: X-rays: ISM
74: }
75:
76: \section{Introduction}
77:
78: Measurement of the current rate of expansion provides essential
79: information on the dynamical state of supernova remnants (SNRs). This
80: is particularly true for the so-called historical remnants, those for
81: which the date of the supernova (SN) explosion is known, since a
82: comparison of the average expansion rate with the current expansion
83: rate yields a measure of the deceleration. Unfortunately, the rate of
84: expansion, although rapid compared to most astrophysical objects in
85: the cosmos, is still rather long on human timescales, necessitating
86: measurements over the course of years, if not decades, in order to
87: attain accurate results. \rosat\ was the first X-ray observatory that
88: had sufficient angular resolution ($\sim$4$^{\prime\prime}$ half-power
89: radius for the High Resolution Imager) and operated for a long enough
90: time that significant measurements of the expansion rate of young SNRs
91: are possible.
92:
93: \par
94:
95: The time-averaged expansion rate of Tycho's SNR based on the outermost
96: extent of the remnant ($\sim$8$^\prime$ in diameter) in either the
97: radio or X-ray band and its well-known age is $\sim$0\farcs56
98: yr$^{-1}$. The observed proper motion of the optical filaments, which
99: are believed to trace the location of the blast wave, indicate much
100: lower current expansion rates ranging from 0\farcs18 yr$^{-1}$ to
101: 0\farcs28 yr$^{-1}$ (Kamper \& van den Bergh 1978). Evidently these
102: filaments are locations where significant deceleration of the SN blast
103: wave has occurred. From the width of the broad H$\alpha$ emission
104: Smith et al.~(1991) derive shock velocities in the range 1500--2800 km
105: s$^{-1}$ that, when combined with the proper motion measurements,
106: imply a distance of 1.5--3.1 kpc. This range is in good agreement
107: with other distance estimates to the remnant (Green 1984; however, see
108: Schwarz et al.~1995) and in the following I adopt a value of 2.3 kpc
109: for the distance. Because the optical filaments cover only a limited
110: portion of Tycho's SNR, the optical data are unable to provide a
111: comprehensive picture of the expansion of the remnant.
112:
113: % Distance estimates:
114: % 1.5-3.1 kpc Smith etal 1991
115: % 2.2 (-0.5,+1.5) Albinson etal 1986
116: % 3.0 Hamilton etal
117: % >2 kpc Black and Raymond 1984
118: % 2.3 kpc Green 1984 MNRAS 209 449
119: % ~2.5 kpc Raymond 1984
120:
121: \par
122:
123: The radio remnant of Tycho's SN has also been observed to be expanding
124: at the current epoch (Strom, Goss, \& Shaver 1982; Tan \& Gull 1985;
125: Reynoso et al.~1997, hereafter R97). Although in each successive
126: analysis the radio data have improved, the basic result of these
127: studies have remained in general agreement. The current radio
128: expansion rate, averaged over the outer rim, is 0.113 \% yr$^{-1}$ or,
129: expressed equivalently, the expansion parameter, defined as $R\propto
130: t^m$, is $m=0.471\pm0.028$. This result is between the free expansion
131: rate, $m\sim 1$, and the expansion rate expected for a remnant in the
132: Sedov phase of evolution, $m=0.4$. There is significant azimuthal
133: variation of the radio expansion rate, while interior features appear
134: to show the same expansion rate as the rim (R97).
135:
136: % SGS 0.47 +/- 0.07
137: % T&G 0.462 +/- 0.024 0.113 %/yr
138: % Ret 0.471 +/- 0.028 0.1126%/yr
139:
140: Vancura, Gorenstein, \& Hughes (1995), using data from two satellite
141: observatories, quoted a current X-ray expansion rate for the SNR based
142: on a 11.5 yr time baseline that was consistent with these other
143: values. Here I present more accurate results on the X-ray expansion of
144: Tycho's SNR using high resolution images accumulated by the \rosat\
145: satellite in two epochs separated by 4.55 yr. A preliminary report on
146: this work, using a different analysis approach, was given by Hughes
147: (1997).
148:
149: \section{Observations}
150:
151: The reduction of the X-ray data closely follows that done in an
152: earlier study of Kepler's SNR (Hughes 1999, hereafter H99); interested
153: readers are referred there for more details. A log of the high
154: resolution \einstein\footnotemark\footnotetext{
155: %
156: The \einstein\ observation is included here for completeness and in
157: fact results for only the two \rosat\ data sets are presented below.
158: Preliminary studies found that the \einstein/\rosat\ expansion results
159: were inconsistent with the \rosat/\rosat\ ones, yielding expansion
160: values a factor of 1.6--1.9 times higher. The \einstein\ and \rosat\
161: comparison is subject to more uncertainty due to differences in the
162: instrumental point-spread functions and spectral bandpasses. The
163: latter effect, combined with spatial variations in the X-ray spectrum
164: of Tycho's SNR (Vancura et al.~1995), is the likely cause of the
165: discrepancy.
166: %
167: }
168: and \rosat\ imaging observations of Tycho's SNR
169: is given in Table 1. The columns list the observatory, start date,
170: the Modified Julian Day (MJD) corresponding to the average date of the
171: observation, and the effective duration (live-time corrected). Figure
172: 1 shows the image from observation R2.
173:
174: \begin{small}
175: \begin{minipage}[t]{83mm}
176: \begin{center}
177: {\noindent{TABLE 1}}\\
178: {\noindent{\sc Observations of Tycho's SNR}}\\[6pt]
179: \begin{tabular}{@{}lccc@{}}
180: \hline\hline\\[-4pt]
181: Observatory
182: & Start Date
183: & Ave.~MJD
184: & Duration (s)\\[4pt]
185: \hline\\[-4pt]
186: {\it Einstein} & 1979 Feb 8 & 43,913.3 & $\phantom{0}50409.2$\\[2pt]
187: {\it ROSAT} (R1) & 1990 Jul 28 & 48,100.7 & $\phantom{0}22163.7$\\[2pt]
188: {\it ROSAT} (R2) & 1995 Feb 5 & 49,763.7 & $104332.1$\\[4pt]
189: \hline
190: \end{tabular}
191: \end{center}
192: \end{minipage}
193: \end{small}
194:
195: % Figure 1:
196: \begin{center}
197: \begin{minipage}[t]{0.47\textwidth}
198: \vspace{-0.5in}
199: \epsfxsize=1.1\textwidth \epsfbox{./tycho_heat.ps}
200: \figcaption{
201: %
202: {\it ROSAT\/} HRI image of Tycho's SNR. The data were smoothed by a
203: Gaussian function with $\sigma = 2^{\prime\prime}$ and are displayed
204: with a square-root intensity scaling. The plus sign marks the
205: geometric center of the remnant. The effective resolution of the map,
206: including both the instrumental point-spread function and the average
207: width of the smoothing kernel, is shown at the lower left.
208: %
209: }
210: \end{minipage}
211: \end{center}
212:
213: The \rosat\ high resolution imager (RHRI) data were processed in some
214: detail. The data were filtered in pulse height to reduce background.
215: Pulse height channels 1 to 11 were used for the first epoch \rosat\
216: observation (R1) and channels 1 to 12 were used for the second epoch
217: observation (R2). This reduced the background level by 5\%-7\% while
218: the source rates were nearly unaffected ($\sim$1\% change). Aspect
219: drift throughout an observation was corrected by aligning separate
220: images made from the data corresponding to each orbit. The images
221: from all the sub-intervals (typically 1500 s long) were registered to
222: the nearest $0\myarcsec5$ pixel, shifted, and added. For R1 the
223: initial registration of the individual maps from the standard analysis
224: was fairly good: all of the individual maps were already aligned
225: relative to each other to within $\sim$2$^{\prime\prime}$ or better.
226: In the case of R2 there was clear evidence for a drift in aspect
227: throughout the observation. The mean registration error was
228: $\sim$3$^{\prime\prime}$, although some of the individual images were
229: misregistered by up to 10$^{\prime\prime}$. For both epochs, the
230: shift-and-add alignment technique produced images with a noticeably
231: improved point response function.
232:
233: The grain scattering halo from Tycho's SNR (Mauche \& Gorenstein 1986;
234: Predehl \& Schmitt 1995) extends over the entire field of view of the
235: HRI, which makes background estimation difficult. The background
236: level was estimated by fitting a spatial power-law component (for the
237: scattering halo) plus a constant background level to the surface
238: brightness profile over the 5$^\prime$ to 15$^\prime$ radial
239: range. The fitted power-law components were consistent between the two
240: RHRI pointings (index of $-$2.3), although the background levels
241: differed by some 15\%, $4.2\times 10^{-3}$ cts s$^{-1}$ arcmin$^{-2}$
242: for R1 and $3.7\times 10^{-3}$ cts s$^{-1}$ arcmin$^{-2}$ for R2.
243: This difference is within the variation observed from field to field
244: for the RHRI (David et al.~1998).
245:
246: Exposure maps were generated for the RHRI observations as before (see
247: H99). Over the portion of the field containing the image of the
248: remnant the ratio of exposure between the first and second epochs
249: varied between 0.963 and 1.045. These corrections are small in
250: comparison to the flux differences I measure and are uncorrelated with
251: image structure in the remnant. Nevertheless, these maps are included
252: in the model fits described below.
253:
254: The exposure- and deadtime-corrected, background-subtracted RHRI count
255: rates of Tycho's SNR within a radius of 8$^\prime$ are $9.092\pm
256: 0.020\,\rm s^{-1}$ and $9.209\pm 0.009\,\rm s^{-1}$ from the first and
257: second epoch images, respectively. This difference in count rates
258: ($\sim$1\%) is consistent with the range of RHRI rates seen in
259: calibration observations of the SNR N132D (see H99).
260:
261: Finally I comment on possible changes in the plate scale using
262: observations of the Andromeda galaxy, which was observed by the RHRI
263: in July 1990, 1994, and 1995, and January 1996. I extracted these
264: data and used the positions of 10 isolated, moderately bright X-ray
265: point sources to constrain the relative rotation angle and plate scale
266: change between pairs of observations. I find no evidence for a change
267: in the RHRI plate scale and set a limit of $\sim$0.008\% yr$^{-1}$ on
268: any changes for timescales of 4 yrs or more.
269:
270: \vspace{-0.25in}
271:
272: \section{Expansion Results}
273:
274: The expansion rate was determined using fitting software that takes one
275: image (``model'') and compares it to another (``data'') as described
276: in H99. The model image was scaled in intensity, shifted in position,
277: and expanded (or contracted) in spatial scale to match the data using,
278: as the figure-of-merit function, the maximum likelihood statistic for
279: Poisson-distributed data. The fitted spatial scale factor yields the
280: global mean expansion rate, which is assumed to be uniform across the
281: entire remnant.
282:
283: Over the 4.55 yrs between observations R1 and R2, Tycho's SNR expanded
284: annually by an amount $0.124\pm0.011\,\% \,\rm yr^{-1}$. The error
285: bar is statistical at $1\,\sigma$ and includes uncertainty from Poisson
286: noise in both observations as well as plate scale changes assuming the
287: limit given above. This result is highly significant both in terms of
288: the final error bar and the reduction in the value of the likelihood
289: statistic for fits with and without any expansion.
290:
291: In order to determine the expansion rate as a function of radius and
292: azimuthal angle, one needs to know the position of the expansion
293: center. The current X-ray data do not allow for the determination of
294: this quantity (see H99), so I have opted to just define the geometric
295: center of the remnant at $0^{\rm h}25^{\rm m}19^{\rm s}\,\,
296: 64^{\circ}08^{\prime}10^{\prime\prime}$ (J2000) as the nominal center
297: of expansion. I then investigated how the results depended on this
298: choice by varying the center position by 40$^{\prime\prime}$ in each
299: of the four cardinal directions. I found that the different choices
300: of center could be nearly perfectly compensated for by appropriate
301: choices of the relative alignment of the two images. (Since there were
302: no serendipitous point sources in the field, it was not possible to do
303: an independent registration of the images.) This result is not
304: surprising since non-optimal image alignment or choice of expansion
305: center will each introduce a sinusoidal term in the expansion rate as
306: a function of azimuth. In effect what I have done is to remove any
307: such sinusoidal term from the results, regardless of origin. My
308: approach is different from that of R97 who used a
309: fixed expansion center (defined by the center of the nearly circular
310: western limb) to measure the radio expansion of Tycho's SNR. And
311: indeed their fractional expansion results do contain a significant
312: sinusoidal term. Because of these complications a detailed comparison
313: between the X-ray and published radio azimuthal variation is not
314: particularly enlightening and will not be pursued in this work.
315: However, since a sinusoid averages to zero over a full cycle, this
316: difference does not affect the comparison of the radio and X-ray
317: global mean expansion rates. As concerns the X-ray azimuthal
318: expansion, I note that the weighted average rate, $\sim$0.13\%
319: yr$^{-1}$, is consistent with the global mean X-ray rate and that
320: there are statistically significant azimuthal variations on angular
321: scales of 10$^\circ$ to 90$^\circ$ (similar to those in Hughes 1997).
322:
323: \begin{small}
324: \begin{minipage}[t]{83mm}
325: \begin{center}
326: {\noindent{TABLE 2}}\\
327: {\noindent{\sc X-ray Expansion Rates with Radius for}}\\
328: {\noindent{\sc Tycho's SNR}}\\[6pt]
329: \begin{tabular}{@{}ccc@{}}
330: \hline\hline\\[-4pt]
331: Radial range & Exp Rate & Sys Err \\
332: (arcmin) & (\% yr$^{-1}$) & (\% yr$^{-1}$) \\[4pt]
333: \hline\\[-4pt]
334: 0.0 -- 1.5 & $0.212\pm 0.088$ & ($-$0.090, $+$0.036) \\[2pt]
335: 1.5 -- 2.0 & $0.178\pm 0.057$ & ($-$0.019, $+$0.008) \\[2pt]
336: 2.0 -- 2.4 & $0.124\pm 0.051$ & ($-$0.017, $+$0.016) \\[2pt]
337: 2.4 -- 2.8 & $0.133\pm 0.033$ & ($-$0.008, $+$0.006) \\[2pt]
338: 2.8 -- 3.2 & $0.080\pm 0.022$ & ($-$0.000, $+$0.006) \\[2pt]
339: 3.2 -- 3.6 & $0.107\pm 0.015$ & ($-$0.003, $+$0.005) \\[2pt]
340: 3.6 -- 4.0 & $0.117\pm 0.012$ & ($-$0.004, $+$0.006) \\[2pt]
341: 4.0 -- 4.4 & $0.167\pm 0.012$ & ($-$0.010, $+$0.008) \\[2pt]
342: 4.4 -- 5.2 & $0.176\pm 0.043$ & ($-$0.004, $+$0.003) \\[4pt]
343: \hline
344: \end{tabular}
345: \end{center}
346: \end{minipage}
347: \end{small}
348:
349: In the remainder of this section I focus on the radial variation of
350: the X-ray expansion rate. Here the fits were carried out separately
351: for several different annular regions about the geometric center of
352: the remnant. Most of the radial bins were 24$^{\prime\prime}$ wide
353: although the innermost and outermost annuli were somewhat thicker.
354: There were only two fit parameters: the fractional expansion rate and
355: the change in intensity. Numerical values for the expansion rate are
356: given in Table 2 and are plotted in Figure 2 along with the change in
357: the X-ray flux and for reference the surface brightness profile. In
358: the bottom two panels the error bars show the statistical
359: uncertainties, while the small boxes surrounding the data points show
360: the range of values that come from fits using the four different
361: expansion centers. Only for the data point closest to the remnant's
362: center is this a significant error. Between the two epochs the X-ray
363: flux appears to have changed only very slightly over the image of the
364: remnant, i.e., by less than 2\%. The weighted average expansion rate
365: is $\sim$0.13\% yr$^{-1}$, although I can reject the hypothesis that
366: the expansion rate is constant with radius at more than the 99.9\%
367: confidence level ($\chi^2 = 17.65$ for 3 d.o.f.) based on the four
368: data points near the rim (radii between 2\farcm8 and 4\farcm4). I
369: find that the expansion parameter increases from $m = 0.34\pm0.10$
370: just inside the bright rim of Tycho's SNR to a value $m = 0.71\pm0.06$
371: at the outermost edge of the SNR. The rapid motion of the outermost
372: edge is the principal cause of the non-uniform radial expansion rate.
373: In fact, over the interior portion of the SNR, covering radii from
374: $2\farcm8$ to $4\farcm0$, the X-ray data are consistent with an
375: expansion parameter of $m=0.45 \pm0.04$. The large expansion rate for
376: the outer rim, plus evidence for a slower rate further in, was also
377: found by Hughes (1997) using an entirely different analysis technique.
378:
379: % Figure 2:
380: \begin{center}
381: \begin{minipage}[t]{0.47\textwidth}
382: %\epsfxsize=0.98\textwidth \epsfbox{./plots/sm_radexp2.ps}
383: \epsfxsize=0.98\textwidth \epsfbox{./sm_radexp2.ps}
384: \figcaption{Radial X-ray surface brightness profile of
385: Tycho's supernova remnant from the \rosat\ HRI (top panel) and the
386: change in X-ray flux (middle panel) and expansion rate as a
387: function of radius (bottom panel) from a comparison of the two \rosat\
388: HRI observations. The error bars show the statistical uncertainty,
389: while the boxes that surround each data point give an estimate of the
390: systematic uncertainty. The dashed line in the bottom panel is the
391: global mean X-ray expansion rate.
392: }
393: \end{minipage}
394: \end{center}
395:
396: \par
397:
398: The published radio expansion parameter of Tycho's SNR,
399: $m=0.471\pm0.028$ (R97), corresponds to the expansion of the rim.
400: This value disagrees, by more than $3\, \sigma$, with the expansion
401: rate of the X-ray rim just derived. Tycho's SNR thus joins the two
402: other youngest remnants of Galactic SN (Cas A and Kepler's SNR) in
403: showing considerably higher expansion rates in the X-ray compared to
404: the radio (see discussion and references in H99). For Tycho's SNR it
405: is clear that the main difference between the X-ray and radio results
406: occurs at the remnant's outermost edge where the forward shock is
407: plowing into the ambient interstellar medium (ISM). The slower motion
408: of interior features (i.e., the reverse-shocked ejecta) is consistent
409: across the radio ($m\approx 0.44$; R97) and X-ray ($m\approx 0.45$)
410: bands.
411:
412: \section{Discussion and Results}
413:
414: Numerous authors (Chevalier 1982; Dwarkadas \& Chevalier 1998;
415: Truelove \& McKee 1999) have modeled the expansion rates of young
416: supernova remnants. Dwarkadas \& Chevalier (1998) in particular
417: examined how different assumed density profiles for the SN ejecta
418: affect the evolution of the resulting remnant, assumed to be
419: interacting with a uniform density ISM. They considered three
420: principal cases: a power law profile ($\rho \propto r^{-7}$), an
421: exponential profile, and a constant density profile. The expansion
422: parameter uniquely defines the age and radius of the remnant,
423: conventionally expressed in scale-free variables. The constants of
424: proportionality between the scaled values and the true physical radius
425: and age depend on the three independent dimensional parameters:
426: explosion energy $E$, ejecta mass $M_{\rm ej}$, and ambient density
427: $\rho_0$. Thus, given the known size $R=2.8 (D/2.3\,\rm kpc)\,\rm pc$
428: and age $t=425\,\rm yr$ of Tycho's SNR it is possible to determine two
429: of the three dimensional parameters. Here I assume that the ejecta
430: mass is $1.4\,M_\odot$ and solve for the other two quantities.
431:
432: The power-law profile predicts a maximum expansion parameter of
433: $m=0.57$ which is too low to be consistent with the X-ray expansion
434: rate. The other two model ejecta profiles can accommodate the high
435: rate observed for the forward shock; however, the inferred values of
436: $E$ and $\rho_0$ are quite different in the two scenarios. Compared to
437: the uniform density case, the expansion parameter for the exponential
438: profile model falls more rapidly with time, so that for a given value
439: of the expansion parameter the scaled age and radius are smaller for
440: the exponential profile. The inferred dimensional parameters for the
441: exponential profile are $E_{51} = E/10^{51}\, {\rm erg} \approx (0.1 -
442: 0.2)\, (M_{\rm ej}/1.4\, M_\odot)\,(D/2.3\,{\rm kpc})^2$ and $n_0 =
443: \rho_0/\mu_{\rm H} \approx (0.004 - 0.08)\, (M_{\rm ej}/1.4\,
444: M_\odot)\,(D/2.3\,{\rm kpc})^{-3}\,\rm cm^{-3}$ ($\mu_{\rm H}$ is the
445: mean mass per hydrogen atom), rather low values for Tycho's SNR. On
446: the other hand, more appropriate values are obtained using the uniform
447: density ejecta model: $E_{51} \approx (0.4 - 0.5)\, (M_{\rm ej}/1.4\,
448: M_\odot)\,(D/2.3\,{\rm kpc})^2$ and $n_0 \approx (0.3 - 0.6)\, (M_{\rm
449: ej}/1.4\, M_\odot)\,(D/2.3\,{\rm kpc})^{-3}\,\rm cm^{-3}$. These are
450: fully consistent with the values that Hamilton, Sarazin, \& Szymkowiak
451: (1986) found in their study of Tycho's X-ray spectrum in which they
452: also assumed a uniform density ejecta profile. Thus it appears that
453: both the expansion of the forward shock and the X-ray emission
454: properties of Tycho's SNR can be well explained with this simple
455: model. What about the slower motion of the reverse-shocked ejecta?
456:
457: Truelove \& McKee (1999) have parameterized the evolution of both the
458: forward and reverse shocks in young SNRs for a number of cases,
459: including a uniform density ejecta model. For the range of scaled ages
460: that describe the expansion of the forward shock, the reverse shock
461: expansion parameter is $m_{\rm RS} = 0.51-0.63$. This rate is in
462: reasonable agreement with the measured expansion parameter of the
463: interior portions of Tycho's SNR ($m = 0.45\pm0.04$). (Although
464: equating the motion of the ejecta to the reverse shock is not strictly
465: correct, it serves as a reasonable first approximation.) An
466: acceptable joint fit ($\chi^2 = 2.7$ for 1 d.o.f.) to the measured
467: X-ray expansion parameters is obtained for best-fit values of
468: $m\approx0.64$ (forward) and $m\approx0.49$ (reverse). The inferred
469: values of $E$ and $n_0$ in this case are similar to those quoted
470: above.
471:
472: The forward shock expansion rate implies a shock velocity of
473: $4600\pm400\, (D/2.3\,{\rm kpc})\,\rm km\,s^{-1}$, which in turn
474: implies a mean post-shock temperature of $kT_S = {3\over 16}\, \mu m_p
475: v_S^2 = 25\pm4 (D/2.3\,{\rm kpc})^2\,\rm keV$ for a mean mass per
476: particle of $\mu =0.61$, which assumes a fully-ionized plasma with
477: 10\% helium. This temperature is quite a bit higher than the estimate,
478: by Hwang, Hughes, \& Petre (1998), of the post-shock electron
479: temperature of the blast wave in Tycho's SNR, $kT_e \approx 4\,\rm
480: keV$. This difference most likely arises from either the
481: non-equilibration of electron and ion temperatures at the shock front
482: or the partition of a significant fraction of the shock energy in
483: Tycho's SNR into relativistic particles as was recently found to be
484: the case in SNR E0102.2$-$7219 (Hughes, Rakowski, \& Decourchelle
485: 2000). Discriminating between these possibilities will be the focus
486: of a future article.
487:
488: \acknowledgments
489:
490: This research has made use of data obtained through the High Energy
491: Astrophysics Science Archive Research Center Online Service, provided
492: by the NASA/Goddard Space Flight Center. Partial support was provided
493: by NASA grant NAG5-6420.
494:
495: \begin{references}
496: \reference{}
497: Chevalier, R.~A.~1982, \apj, 258, 790
498: \reference{}
499: David, L.~P., et al.~1998, The ROSAT High Resolution Imager (HRI)
500: Calibration Report (http://hea-www.harvard.edu/rosat/rsdc-www/hricalrep.html)
501: \reference{}
502: Dwarkadas, V.~V., \& Chevalier, R.~A.~1998, ApJ, 497, 807
503: \reference{}
504: Green, D.~A.~1984, \mnras, 209, 449
505: \reference{}
506: Hamilton, A.~J.~S., Sarazin, C.~L., \& Szymkowiak, A.~E.~1986, \apj,
507: 300, 713
508: \reference{}
509: Hughes, J.~P.~1997, in Proc.~International Conf.~on X-Ray Astronomy --
510: \asca\ 3rd Anniv.-- X-Ray Imaging and Spectroscopy of Cosmic Hot
511: Plasmas, eds.~F.~Makino and K.~Mitsuda (Tokyo: Universal Academy), 359
512: \reference{}
513: Hughes, J.~P.~1999, \apj, 527, 298 (H99)
514: \reference{}
515: Hughes, J.~P., Rakowski, C.~E., \& Decourchelle, A.~2000, \apj, in press
516: (astro-ph/0007032)
517: \reference{}
518: Hwang, U., Hughes, J.~P., \& Petre, R.~1998, \apj, 497, 833
519: \reference{}
520: Kamper, K.~W., \& van den Bergh, S., 1978, \apj, 224, 851
521: \reference{}
522: Mauche, C.~W., \& Gorenstein, P.~1986, \apj, 302, 371
523: \reference{}
524: Predehl, P., \& Schmitt, J.~H.~M.~M.~1995, A\&A, 293, 889
525: \reference{}
526: Reynoso, E.~M., Moffett, D.~A., Goss, W.~M., Dubner, G.~M., Dickel,
527: J.~R., Reynolds, S.~P., \& Giacani, E.~B.~1997, \apj, 491, 816 (R97)
528: \reference{}
529: Schwarz, U.~J., Goss, W.~M., Kalberla, P.~M., \& Benaglia, P.~1995,
530: A\&A, 299, 193
531: \reference{}
532: Smith, R.~C., Kirshner, R.~P., Blair, W.~P., \& Winkler, P.~F.~1991,
533: \apj, 375, 652
534: \reference{}
535: Strom, R.~G., Goss, W.~M., \& Shaver, P.~A.~1982, \mnras, 200, 473
536: \reference{}
537: Tan, S.~M., \& Gull, S.~F.~1985, \mnras, 216, 949
538: \reference{}
539: Truelove, J.~K., \& McKee, C.~F.~1999, \apjs, 120, 299 (erratum:
540: \apjs, 128, 403)
541: \reference{}
542: Vancura, O., Gorenstein, P., \& Hughes, J.~P.~1995, \apj, 441, 680
543:
544: \end{references}
545:
546: \end{document}
547: