1: %\documentclass[12pt,preprint]{emulateapj}
2: \documentclass[12pt,preprint]{aastex}
3: %\documentclass[12pt]{article}
4:
5: %\usepackage{emulateapj5}
6: \usepackage{graphicx}
7: \usepackage{fancyheadings}
8: \usepackage{color}
9: \usepackage{natbib}
10:
11:
12:
13:
14: \begin{document}
15:
16:
17: \title{Probing the Nature of the Vela X Cocoon}
18:
19: \author{Stephanie M. LaMassa\altaffilmark{1}, Patrick O. Slane\altaffilmark{2}, \& Okkie C. de Jager\altaffilmark{3,4}}
20:
21: \altaffiltext{1}{The Johns
22: Hopkins University}
23:
24: \altaffiltext{2}{Harvard-Smithsonian Center for
25: Astrophysics}
26:
27: \altaffiltext{3}{Unit for Space Physics, North-West University}
28:
29:
30: \altaffiltext{4}{South African
31: Department of Science \& Technology and National Research Foundation
32: Research Chair: Astrophysics \& Space Science}
33:
34: \begin{abstract}
35:
36: Vela~X is a pulsar wind nebula (PWN) associated with the active
37: pulsar B0833-45 and contained within the Vela supernova remnant
38: (SNR). A collimated X-ray filament (``cocoon'') extends south-southwest
39: from the pulsar to the center of Vela~X. VLA observations uncovered
40: radio emission coincident with the eastern edge of the cocoon and
41: H.E.S.S. has detected TeV $\gamma$-ray emission from this region as
42: well. Using XMM-\textit{Newton} archival data, covering the southern
43: portion of this feature, we analyze the X-ray properties of the
44: cocoon. The X-ray data are best fit by an absorbed nonequilibrium
45: plasma model with a powerlaw component. Our analysis of the thermal
46: emission shows enhanced abundances of O, Ne, and Mg within the
47: cocoon, indicating the presence of ejecta-rich material from the
48: propagation of the SNR reverse shock, consistent with Vela~X being
49: a disrupted PWN. We investigate the physical processes that excite
50: the electrons in the PWN to emit in the radio, X-ray and $\gamma$-ray
51: bands. The radio and non-thermal X-ray emission can be explained
52: by synchrotron emission. We model the $\gamma$-ray emission by Inverse
53: Compton scattering of electrons off of cosmic microwave background
54: (CMB) photons. We use a 3-component broken power law to model the
55: synchrotron emission, finding an intrinsic break in the
56: electron
57: spectrum at $\sim5 \times 10^{6}$ keV and a cooling break at $\sim$ 5.5
58: $\times 10^{10}$ keV. This cooling break along with a magnetic field
59: strength of 5 $\times 10^{-6}$ G indicate that the synchrotron break
60: occurs at $\sim$1 keV.
61:
62: \end{abstract}
63:
64: \keywords{ISM: individual (Vela X) --- pulsars: general --- pulsars:
65: individual(PSR B0833-45) --- supernova remnants}
66:
67: \section{Introduction}
68:
69: The Vela supernova remnant (SNR), at a distance of $\sim 290$ pc
70: \citep{Dod} is the closest SNR to contain an active pulsar, (PSR)
71: B0833-45. The Vela pulsar has a period (\textit{P}) of 89 ms and a
72: period derivative (\textit{\.{P}}) of 1.25 x 10 $^{-13}$s s$^{-1}$,
73: implying an age of 11,000 years and a spin-down luminosity
74: (\textit{\.{E}}) of 7$\times 10^{36}$ erg s$^{-1}$. The remnant
75: spans about 8$^\circ$ in diameter and contains regions of nonthermal
76: emission in the radio, X-ray and $\gamma$-ray bands, including Vela
77: X, a pulsar wind nebula (PWN) spanning a region of 2$^\circ$ $\times$
78: 3$^\circ$ south-southwest of the pulsar. The center of Vela~X is
79: offset by about 40' from the pulsar and emits in the radio, X-ray,
80: and $\gamma$-ray bands (Frail et al. 1997; Aharonian et al. 2006).
81:
82: \textit{ROSAT} observations of the Vela SNR (Figure~\ref{rosat})
83: reveal a collimated X-ray filament seen predominantly at higher
84: energies extending from the pulsar to the center of Vela~X, about
85: 45' in length. Markwardt and \"{O}gelman (hereafter M\"{O}) interpreted
86: the \textit{ROSAT} feature as an X-ray jet along which the pulsar
87: loses energy \citep{MO95}.
88: The ``jet head'', or interaction
89: between the jet and SNR shell, was later observed with \textit{ASCA}
90: by M\"{O}, and they found the spectrum of this region to be best
91: fit by a two component model, a lower temperature thermal plasma model
92: and a higher energy component, that could be fit by either a higher
93: temperature plasma model or power-law emission \citep{MO97}.
94:
95: An alternate explanation for the X-ray filament was proposed by
96: Gvaramadze (1999) who suggested that this feature represents a
97: projection effect of a Rayleigh-Taylor instability, which is caused
98: by the impact of the SN ejecta with the wind-driven shell from the
99: progenitor. Subsequent \textit{Chandra} observations revealed an
100: X-ray jet situated along the northwest-southeast pulsar spin axis,
101: coinciding with the pulsar proper motion and not overlapping the
102: M\"{O} ``jet'' \citep{Pavlov}.
103:
104: Though this collimated X-ray feature (hereafter ``cocoon'') is no
105: longer considered a ``jet,'' it warrants further investigation.
106: VLA observations reveal a bright radio filament coincident with the
107: eastern edge of this filament \citep{Frail}. H.E.S.S. observed TeV
108: $\gamma$-rays emanating from the \textit{ROSAT} cocoon; Aharonian
109: et al. 2006 use Inverse Compton scattering to fit this H.E.S.S.
110: spectral energy distribution (SED). Horns et al. 2006 attribute the
111: $\gamma$-ray emission from the cocoon to hadronic processes, which
112: hinges on a thermal plasma density of 0.6 cm$^{-3}$, consistent
113: with the previous findings of M\"{O} 1995 .
114:
115: Here we report on the analysis of XMM-\textit{Newton} archival data
116: from the interaction region of this filament with the SNR shell. In
117: \S2 we summarize the observations and data reduction techniques
118: applied to the associated spectra. In \S3 we describe our modeling
119: of the X-ray spectrum, where we investigate the thermal and nonthermal
120: emission observed from the cocoon. We investigate the broadband
121: spectrum from this region in \S4, using the results of our X-ray
122: analysis along with spectral information from the radio and very
123: high energy $\gamma$-ray bands. We consider these results in the
124: context of a model in which the reverse shock from the Vela SNR has
125: interacted with the PWN (Blondin, Chevalier, \& Frierson 2001). We
126: discuss our conclusions in \S5.
127:
128: \section{Observations and Data Reduction}
129:
130: The ``Vela jet head'' was observed by XMM-\textit{Newton} on May
131: 2, 2005 for 22 ks with the MOS1 and MOS2 detectors and for 20 ks
132: with the PN detector in Prime Full Window mode with the medium
133: filter (PI Gallant, ObsID 0094630101).
134: Data reduction was
135: performed with the XMM Science Analysis System (SAS) to filter the
136: data and remove flares. The good science time remaining after this
137: processing is as follows: 22 ks for MOS1, 21.6 ks for MOS2, and
138: 17.6 ks for PN.
139:
140:
141: The XMM field of view covers just the southernmost portion of the
142: cocoon detected by \textit{ROSAT}. The field of view for the PN
143: detector is shown in Figure~\ref{pn}, with the source and background
144: regions we analyzed overlaid. Spectra were extracted from roughly
145: the same source and background regions among all 3 detectors, using
146: the \textit{xmmselect} gui. The response files were generated with
147: the \textit{rmfgen} and \textit{arfgen} SAS tools.
148:
149: We extracted background spectra from a region within the
150: the SNR and the 2$^\circ$ x 3$^\circ$ radio PWN, but outside the cocoon (Figure 2).
151: For comparison, we extracted a second set of background spectra
152: from a region along the northwestern shell of the SNR using archival
153: data from observations performed on 9 November 2001 (PI Turner, ObsID
154: 0112690101). The latter region is centered at RA = 08:48:33 and Dec =
155: -42:23:45, with a radius of $1.8^\prime$.
156:
157: We applied the SAS tool evigweight to all spectra to correct for
158: vignetting. To account for the unvignetted internal background,
159: for both the source and background regions we subtracted similarly
160: processed spectra taken from available blank sky fields \footnote
161: {Blank sky files were taking in Prime Full Frame mode using the
162: medium filter (M1.M.FF.EVLIRP.FIT, M2.M.FF.EVLIRP.FIT, and
163: PN.M.FF.EVLIRA.FIT).}, each weighted by a factor determined by the
164: ratio of high-energy counts between the Vela and blank-sky spectra
165: (see Arnaud et al. 2002).
166:
167: The source region was also separated into upper and lower sections,
168: as seen in Figure~\ref{pn}. Spectra were extracted separately from
169: these areas to search for possible spectral changes with increasing
170: distance from the pulsar.
171:
172:
173: \section{Spectral Analysis}
174:
175: We modeled the spectra in XSPEC and found that a two-component model
176: was required to fit the entire source region, consistent with the
177: findings of M\"{O} in 1997. Using an adjacent background region,
178: the data for the full region were best fit by an absorbed (N$_H$
179: $\sim$ 1.6 $\times$ 10$^{20}$) nonequilibrium plasma model with a
180: powerlaw component to accommodate the hard emission, as shown in
181: Figure~\ref{spectrum}. A very high temperature thermal component proposed by M\"{O} (1997)
182: can reproduce the observed hard
183: emission in the XMM band, but given the observed emission from Vela~X at energies
184: in excess of 100 keV \citep{Mang}, it appears clear that this component
185: is nonthermal. The results of the
186: plasma model with a powerlaw component are presented in the first
187: column of Table 1. The modeling shows enhanced abundances of oxygen,
188: neon, and magnesium, higher than abundances seen in the SNR as a
189: whole, but consistent with abundances represented in ejecta fragments,
190: most notably Vela Region D \citep{Miceli}. The temperature of
191: $\sim$0.5 keV is also higher than that for most of the remnant,
192: which ranges between 0.12 and 0.18 keV \citep{Lu}, but is consistent
193: with temperatures of shocked ejecta fragments \citep{Asch}. The
194: hard component is best fit by a photon index of $\Gamma$ of 2.30
195: $\pm$ 0.04, consistent with synchrotron emission from energetic particles
196: injected by the pulsar.
197:
198: We compare these results using a background region from
199: within the northwestern rim of the SNR shell \citep{Miceli}.
200: By subtracting this background region, the source spectra retains
201: contribution from the PWN, but contributions from the background
202: and SNR are removed. The best-fit parameters from this fit
203: are shown in the second column of Table 1. The abundances are still
204: elevated after subtracting out SNR emission, suggesting that high
205: abundances are intrinsic to this portion of the
206: cocoon.
207:
208: Spectra were also extracted separately from the upper and lower
209: regions of the portion of the cocoon covered in this pointing, as
210: shown in Figure~\ref{pn}, to investigate whether the steepness of
211: the power law spectrum changes with distance from the pulsar. The
212: northern portion covers the region about 22.6' to 31.8'($\sim$2.0
213: pc) from the pulsar and the southern portion extends from about
214: 31.8' to 42.1' ($\sim$3.7 pc) from the pulsar. To facilitate fitting
215: the spectra, $N_H$ was set to our best fit value of 1.6$\times 10^{20}$ and
216: $\tau$ was set to 4.8$\times 10^{10}$. A variation in the spectral
217: index was found, with
218: $\Gamma_{upper}=2.13 \pm 0.03$ and $\Gamma_{lower}=1.90 \pm 0.02$. Setting
219: $N_H$ to the the value in the literature (2.26$\times 10^{20}$, M\"{O} 1997) shows
220: a similar trend, with $\Gamma_{upper}$ = 2.22$\pm$0.02 and $\Gamma_{lower}$ = 1.99$\pm$0.01.
221: The $\Gamma$ values above
222: indicate that the power-law spectrum does change with distance from
223: the pulsar, but in the opposite manner expected, flattening at
224: larger radii rather than steepening. Further
225: observations along the length of the cocoon are needed to fully
226: test whether this trend exists along the full filament or if this
227: is a local phenomenon related to turbulent disruption of the PWN
228: where crushing of the cocoon against the PWN could cause adiabatic
229: heating that hardens the spectrum.
230:
231:
232: \section{Modeling}
233:
234: We have modeled the radio and X-ray emission as synchrotron radiation
235: from energetic electrons within the cocoon,
236: and the $\gamma$-ray emission
237: as
238: Inverse Compton emission from the
239: scattering of the cosmic microwave background radiation off of the
240: same electron population. The domain
241: between the radio and X-ray energies is currently unexplored so we
242: have interpolated between these two bandpasses.
243: In order to reproduce the spectral index derived from the X-ray
244: described above, the measured spectral indices in the radio band (with
245: $\alpha_R$ ranging from $\sim 0$ to 0.38, where $S_\nu \propto \nu^{-\alpha_R}$),
246: and also the observed spectral shape of the TeV emission, we require an electron
247: spectrum with multiple slopes.
248: Treating the electron spectrum as a power law with a single break can adequately describe
249: the X-ray and $\gamma$-ray spectra, but either under-predicts the radio emission by a significant
250: factor, or requires a particle spectrum that leads to $\alpha_r \sim 0.7$, much steeper
251: than the measured values.
252: To fit the entire spectrum,
253: we thus consider a three-component power law:
254:
255: \[ \frac{dN}{dE} = \left\{ \begin{array}
256: {r@{\quad:\quad}l}
257: A_{1}E^{-\sigma_{1}} & E \le E_{1} \\
258: A_{2}E^{-\sigma_{2}} & E_{1} < E \le E_{2} \\
259: A_{3}E^{-\sigma_{3}} & E > E_{2} \\
260:
261: \end{array}\right. \]
262: A$_{1}$, A$_{2}$, and A$_{3}$ are the normalization factors and
263: $\sigma_{1}$, $\sigma_{2}$, and $\sigma_{3}$ are the indices of the
264: electron distribution within each energy range. The photon spectra
265: produced by the synchrotron and IC emission in ${\nu}F_{\nu}$ format are:
266:
267: \[{\nu}F_{\nu}(synch) = \frac{V_{E}}{4{\pi}D^{2}_{pc}}{\nu}P_{tot}({\nu},E),
268: \]
269:
270: \[{\nu}F_{\nu}(IC) = \frac{V_{E}}{4{\pi}D^{2}_{pc}}E^{2}\frac{dn_{\gamma}({\epsilon}_{\gamma})}{dt},\]
271: where $V_{E}\sim 10 ^{55}{\rm\ cm}^3$ is the electron emission volume
272: assuming a cylindrical volume with a radius of $10^\prime.44$ and
273: length of $19^\prime.5$; here we have assumed a distance $D_{pc} = 290$~pc
274: \citep{Mang}. $P_{tot}({\nu},E)$
275: and $dn_{\gamma}({\epsilon}_{\gamma})/dt$ are synchrotron and IC
276: emissivities \\ \citep{Lanz}.
277:
278: The VLA, H.E.S.S., and non-thermal components of the XMM spectra were
279: plotted along with the flux
280: model cited above. The VLA \citep{Frail} and H.E.S.S. data
281: points were scaled to coincide
282: with the XMM region size. The magnetic field strength, normalization, spectral index
283: values, and break energy values were varied until we
284: obtained a model that adequately described the data points, as shown
285: in Figure~\ref{model}; A$_{2}$
286: and A$_{3}$ were calculated by fixing A$_{1}$ and requiring the flux
287: to match at the endpoints of the broken power-law segments. The best-fit
288: parameters from this fit are shown in Table 2.
289:
290: A power law spectrum with multiple breaks has been used to describe
291: the emission from a number of PWNe, most notably the Crab Nebula.
292: The first break in the spectrum of the Vela X cocoon, at $5\times 10^{6}$
293: keV, most likely corresponds to an intrinsic break. Several factors can
294: contribute to this break; for the Crab nebula in particular, it
295: has been posited that this break results from two distinct electron
296: populations or from an elevated value of the
297: magnetization parameter \citep{Volpi}. It has been suggested by
298: de Jager (2007) that the radio-emitting and X-ray-emitting electrons
299: from the entire PWN form distinct populations. We note that our constraints
300: on both the maximum energy and the overall normalization of this low-energy
301: component are somewhat weak, and do not rule out this possibility for
302: the southern cocoon region.
303:
304: The second break in the spectrum at $5.5\times 10^{10}$ keV is most
305: likely a cooling break related to the synchrotron lifetime of the
306: electrons injected from the pulsar. For our analysis of this segment
307: of the Vela PWN, the magnetic field (5$\times10^{-6}$ G) and cutoff
308: energy (5.5$\times 10^{10}$ keV) corresponds to a critical frequency
309: of 2.4$\times 10^{17}$ Hz, or $\sim$1 keV, indicating that the
310: synchrotron break occurs in the X-ray band. This magnetic field
311: strength and synchrotron break are consistent with the values derived
312: by de Jager, Slane \& LaMassa (2008) which used \textit{INTEGRAL},
313: BeppoSAX, and H.E.S.S. observations of the full Vela X region to
314: constrain a model of a postshock injection spectrum. In this case,
315: the electron spectra are produced by solving the time dependent
316: transport equation with synchrotron losses.
317: Integration of the non-thermal electron spectrum from this region of
318: the cocoon gives an energy
319: of $\sim 7.7 \times10^{45}$~erg, which is about 0.3\% of the
320: energy input from the pulsar ($E \approx$ \textit{\.{E}} $\times \tau_c \sim 2.31 \times10^{48}$
321: erg, where $\tau_c$ is the characteristic age of the pulsar).
322:
323:
324: \section{Discussion and Conclusions}
325:
326: If the ambient medium into which a SNR expands is inhomogeneous,
327: simulations show that the SNR will expand asymmetrically, resulting
328: in a reverse shock that subsequently crushes the PWN asymmetrically
329: \citep{Blondin}. The Vela radio nebula extends southward of the
330: pulsar, possibly suggesting that the northern portion has been
331: disturbed by such an interaction with the SNR reverse shock.
332: In such a scenario, simulations show that ejecta-rich material can
333: be turbulently mixed into the PWN from the reverse shock interaction
334: \citep{Blondin}.
335: This is consistent with our observation of
336: enhanced abundances of O, Ne, and Mg in the cocoon,
337: similar to the observed abundances of ejecta fragments
338: in Vela (e.g. Vela Region D, Miceli et al. 2005),
339: providing support for the disrupted PWN interpretation.
340:
341:
342: Our modeling of the non-thermal emission from this segment of the cocoon
343: indicates that a leptonic model adequately describes the data, with
344: the radio and X-ray emission resulting from synchrotron radiation and
345: the $\gamma$-ray emission arising from inverse-Compton scattering.
346: We need a 3-component broken power law model to adequately fit the data, unless we
347: ignore the radio flux and spectral index values for Vela X.
348: The high energy break, at 5.5$\times 10^{10}$ keV, and derived magnetic
349: field of 5$\times 10^{-6}$ G, correspond to a break in the synchrotron
350: spectrum at $\sim$1
351: keV and are consistent with the findings of de
352: Jager, Slane \& LaMassa (2008) which obtained
353: obtain the same field strength and synchrotron break using
354: a different approach. The single break model to the X-ray and
355: $\gamma$-ray data, which could be
356: applicable if a separate electron population produces the radio
357: emission, results in a comparable magnetic field strength (4.6$\times 10^{-6}$
358: G) and cooling break energy (5.5$\times 10^{10}$ keV).
359:
360: Horns et al. (2006)
361: propose a hadronic model for the $\gamma$-ray emission from the Vela X
362: cocoon, wherein the emission is the result of the decay of neutral
363: pions produced in proton-proton collisions in the cocoon.
364: This model
365: requires a
366: number density
367: of $n > 0.6 {\rm\ cm}^{-3}$ for the target material.
368: However, the density of thermal emission from our observations is
369: only $\sim$0.09 cm$^{-3}$,
370: which is too low to accommodate the
371: hadronic scenario. Moreover, we have considered an electron/ion number
372: density ratio consistent with cosmic abundances; the presence of
373: ejecta will increase this ratio, further reducing the derived gas
374: density. Our observation only covers the southernmost portion of
375: the cocoon, so perhaps a higher density is present along the northern
376: parts. More X-ray observations of Vela X are needed to determine whether or not
377: the $\gamma$-ray emission from hadronic processes are feasible for other regions
378: of the PWN.
379:
380:
381:
382:
383:
384: \acknowledgments{PO Slane acknowledges support
385: from NASA contract NAS8-03060.}
386:
387:
388: \begin{thebibliography}{}
389: %\setlength\itemsep{0cm}
390:
391: \bibitem[Aharonian et al. 2006]{Ahar} Aharonian, F., et al.\
392: 2006, A\&A, 448L, 43A
393: \bibitem[Arnaud et al. 2001]{Arnaud} Arnaud, M. et al. \ 2001,
394: A\&A, 365L, 80
395:
396: \bibitem[Aschenbach et al. 1995]{Asch}
397: Aschenbach, B., Egger, R., \& Trumper J.\ 1995, Nature, 373, 587
398:
399: \bibitem[Blondin Chevalier \& Frierson 2001]{Blondin}
400: Blondin, J.~M., Chevlier, R.~A., \& Frierson, D.~M.\ 2001, ApJ,
401: 563, 806
402:
403: \bibitem[de Jager 2007]{Jag} de Jager, O.~C. \ 2007, ApJ, 658, 1177
404:
405: \bibitem[de Jager, Slane \& LaMassa 2008]{II} de Jager, O.~C., Slane,
406: P.~O., \& LaMassa, S.\ 2008, in press
407:
408: \bibitem[Dodson et al. 2003]{Dod} Dodson, R. et al. \ 2003, ApJ, 596, 1137
409:
410: \bibitem[Frail et al. 1997]{Frail} Frail, D.~A. et al.\ 1997, ApJ, 475, 224
411:
412: \bibitem[Gvaramadze 1999]{Gvar} Gvaramadze, Vasilii\ 1999, A\&A,
413: 352, 712
414:
415: \bibitem[Horns et al. 2006]{Horns} Horns, D. et al.\ 2006, A\&A,
416: 451, L51
417:
418: \bibitem[Lazendic et al. 2004]{Lanz} Lazendic, J.~S et al. \ 2004, ApJ, 602, 271
419:
420: \bibitem[Lu \& Aschenbach 2000]{Lu} Lu, F.~J., \&
421: Aschenbach, B.\ 2000, A\&A, 362, 1083
422: \bibitem[Mangano et al. 2005]{Mang} Mangano, V. et al. \ 2005, A\&A, 436, 917
423:
424: \bibitem[M\"{O} 1995]{MO95} Markwardt,
425: C.~B. \& \"{O}gelman, H.~B.\ 1995, Nature, 375, 40M
426:
427: \bibitem[M\"{O} 1997]{MO97} Markwardt,
428: C.~B. \& \"{O}gelman, H.~B.\ 1997, ApJ, 480, L13
429:
430: \bibitem[Miceli et al. 2005]{Miceli} Micel, M., Bocchino, F., Maggio,
431: A., \& Reale, F.\ 2005 A\&A, 442, 513
432:
433: \bibitem[Pavlov et al. 2001]{Pavlov} Pavlov, G.~G. et al.\ 2001, ApJ, 554, L189
434:
435: \bibitem[Volpi et al. 2008]{Volpi} Volpi, D., Del Zanna, L., Amato,
436: E., \& Bucciantini, N. \ arXiv:0804.1323v1
437:
438: \end{thebibliography}
439:
440: \clearpage
441:
442: \begin{table}
443: \caption{Best-fit Values}
444: \begin{center}
445: \begin{tabular}{lcc}
446: \hline
447: \noalign{\smallskip}
448: Parameter & Value using Background 1 & Value using Background 2 \\
449: \hline\noalign{\smallskip}
450: $N_H$ ($\times 10^{20}$) & 1.6$^{+0.3}_{-0.2}$ &
451: 1.6$^{+0.5}_{-0.5}$ \\
452: kT (keV) & 0.48$^{+0.05}_{-0.06}$ & 0.50$^{+0.07}_{-0.06}$ \\
453: $$[O]$^{1}$ & 1.7$^{+0.1}_{-0.1}$ & 1.6$^{+0.2}_{-0.1}$ \\
454: $$[Ne]$^{1}$ & 5.4$^{+0.4}_{-0.4}$ & 4.7$^{+0.4}_{-0.4}$\\
455: $$[Mg]$^{1}$ & 3.3$^{+0.6}_{-0.3}$ & 2.9$^{+0.8}_{-0.7}$\\
456: $$[Fe]$^{1}$ & 1.4$^{+1.0}_{-0.3}$ & \\
457: $\tau$ (10$^{10}$s/cm$^3$) & 4.8$^{+1.7}_{-0.8}$ & 4.0$^{+1.2}_{-0.9}$\\
458: $\Gamma$ & 2.3 $\pm0.04$ & 2.2$\pm0.1$ \\
459: Flux (total)$^{2}$ & 2.5$\pm0.4$ & 4.0$\pm0.4$ \\
460: Flux (unabsorbed)$^{2}$ & 2.7$\pm0.4$ & 4.2$\pm0.4$\\
461: Flux (power law)$^{2}$ & 1.5$\pm0.2$ & 2.5$\pm0.2$\\
462: \hline
463: \hline
464: \multicolumn{3}{l}{$^{1}$ Abundances relative to solar.}\\
465: \multicolumn{3}{l}{$^{2}$ Flux in units of 10$^{-11}$ erg
466: cm$^{-2}$s$^{-1}$ and averaged among the 3 detectors.}\\
467: \multicolumn{3}{l}{Background 1 is within the PWN.}\\
468: \multicolumn{3}{l}{Background 2 is outside the PWN.}\\
469: \end{tabular}
470: \end{center}
471: \end{table}
472:
473: \begin{table}
474: \caption{Synchrotron + Inverse Compton Scattering Model Values}
475: \begin{center}
476: \begin{tabular}{lc}
477: \hline
478: \noalign{\smallskip}
479: Parameter & Value \\
480: \hline
481: \noalign{\smallskip}
482: A$_{1}$ (particles/erg/cm$^3$ (s/g/cm)$^{-\sigma}$) & 2$\times10^{-29}$ \\
483: E$_{1}$ (keV) & 5$\times10^{6}$ \\
484: E$_{2}$ (keV) & 5.5$\times10^{10}$ \\
485: B (G) & 5$\times10^{-6}$ \\
486: $\sigma_{1}$ & 1.75 \\
487: $\sigma_{2}$ & 2.4 \\
488: $\sigma_{3}$ & 4.2 \\
489: \hline
490: \hline
491: \end{tabular}
492: \end{center}
493: \end{table}
494:
495: \clearpage
496:
497: \begin{figure}[f]
498: \plotone{f1.eps}
499: \caption{\label{rosat}Composite \textit{ROSAT} RASS image of the Vela SNR with
500: radio contours overlaid. The red band is the lower energy range,
501: from 0.1 - 0.4 keV and the green and blue bands are the higher energy range,
502: from 0.5 - 2.0 keV. All bands were smoothed using a Gaussian kernel
503: of 75 arcseconds. The white box indicates the XMM-\textit{Newton} region we analyzed.}
504: \end{figure}
505:
506: \begin{figure}[f]
507: \plotone{f2.eps}
508: \caption{\label{pn}XMM-\textit{Newton} image from the PN detector
509: of the southern portion of the Vela~X cocoon, filtered in energy
510: from 0.2 to 6.6 keV. The source and background regions are
511: indicated by the black boxes, where both the northern and southern
512: source regions indicate the regions used when investigating the
513: northern and southern portions of the cocoon separately. The
514: northern portion covers the region about 22.6' to 31.8' from the
515: pulsar and the southern portion extends from about 31.8' to 42.1'
516: from the pulsar. The image was
517: smoothed with a
518: gaussian kernel of 9.6''.}
519: \end{figure}
520:
521: \begin{figure}[f]
522: \plotone{f3.eps}
523: \caption{\label{spectrum}Epic PN spectrum from Vela X cocoon. The solid curve represents the best-fit model
524: summarized in Table 1. Individual model components are shown in red (thermal) and blue (nonthermal).}
525: \end{figure}
526:
527: \begin{figure}[f]
528: \plotone{f4.eps}
529: \caption{\label{model}Broadband modeling of cocoon emission. The upper panel shows a power law electron
530: spectrum with two spectral breaks. The lower panel shows a power law spectrum with a single break, using
531: a low-energy spectral index that either reproduces the observed radio flux (solid curve) or matches the
532: measured radio spectral index (dashed curve). The data points from the VLA and H.E.S.S. were scaled to match
533: the size of the region in the XMM observation.}
534: \end{figure}
535:
536:
537: \end{document}
538: