1: \documentclass[12pt,preprint]{aastex}
2:
3: \slugcomment{}
4: \shorttitle{Pulsar Wind Nebula in DA 495}
5: \shortauthors{Arzoumanian et al.}
6:
7: \bibpunct[; ]{(}{)}{,}{a}{}{;}
8:
9: \newcommand{\nh}{N_{\rm H}}
10: \newcommand{\etal}{et al.\ }
11: \newcommand{\asca}{{\em ASCA}}
12: \newcommand{\cxo}{{\em Chandra}}
13: \newcommand{\kothes}{KLR+08}
14:
15:
16: \begin{document}
17:
18: \title{{\em Chandra\/} Confirmation of a Pulsar Wind Nebula in DA 495}
19:
20: \author{Z. Arzoumanian\altaffilmark{1,2}}
21:
22: \email{zaven@milkyway.gsfc.nasa.gov}
23:
24: \author{S. Safi-Harb\altaffilmark{3}}
25:
26: \author{T. L. Landecker\altaffilmark{4}}
27:
28: \author{R. Kothes\altaffilmark{4,5}}
29:
30: \and{}
31:
32: \author{F. Camilo\altaffilmark{6}}
33:
34: \altaffiltext{1}{CRESST and X-ray Astrophysics Laboratory, NASA-GSFC, %Code 662,
35: Greenbelt, MD 20771.}
36:
37: \altaffiltext{2}{Universities Space Research Association, Columbia, MD 21044.}
38:
39: \altaffiltext{3}{Canada Research Chair. Department of Physics and Astronomy, University of
40: Manitoba, Winnipeg, MB, R3T 2N2, Canada.}
41:
42: \altaffiltext{4}{National Research Council of Canada, Herzberg
43: Institute of Astrophysics, Dominion Radio Astrophysical Observatory,
44: Penticton, British Columbia, V2A 6J9, Canada.}
45:
46: \altaffiltext{5}{Department of Physics and Astronomy, University of
47: Calgary, Calgary, AB, Canada.}
48:
49: \altaffiltext{6}{Columbia Astrophysics Laboratory, Columbia
50: University, New York, NY 10027.}
51:
52:
53: \begin{abstract} As part of a multiwavelength study of the unusual
54: radio supernova remnant DA 495, we present observations made with
55: the {\em Chandra X-ray Observatory}. Imaging and spectroscopic
56: analysis confirms the previously detected X-ray source at the heart
57: of the annular radio nebula, establishing the radiative properties
58: of two key emission components: a soft unresolved source with a
59: blackbody temperature of 1 MK consistent with a neutron star,
60: surrounded by a nonthermal nebula 40$''$ in diameter exhibiting a
61: power-law spectrum with photon index $\Gamma = 1.6\pm0.3$, typical of a
62: pulsar wind nebula. The implied spin-down luminosity of the
63: neutron star, assuming a conversion efficiency to nebular flux
64: appropriate to Vela-like pulsars, is $\sim 10^{35}$ ergs~s$^{-1}$,
65: again typical of objects a few tens of kyr old. Morphologically, the
66: nebular flux is slightly enhanced along a direction, in projection
67: on the sky, independently demonstrated to be of significance in radio
68: polarization observations; we argue that this represents the
69: orientation of the pulsar spin axis. At smaller scales, a narrow
70: X-ray feature is seen extending out 5$''$ from the point source, a
71: distance consistent with the sizes of resolved wind termination
72: shocks around many Vela-like pulsars.
73: Finally, we argue based on synchrotron lifetimes in
74: the estimated nebular magnetic field that DA 495 represents
75: a rare pulsar wind nebula in which electromagnetic flux makes up a
76: significant part, together with particle flux, of the neutron star's
77: wind, and that this high magnetization factor may account for the
78: nebula's low luminosity.
79: \end{abstract}
80:
81: \keywords{ISM: individual (DA 495, G65.7+1.2) --- radiation
82: mechanisms: nonthermal --- stars: neutron --- supernova remnants ---
83: X-rays: ISM}
84:
85: \section{Introduction}
86:
87: With the {\em Chandra X-ray Observatory}'s first glimpse of the
88: Crab pulsar \citep{2000ApJ...536L..81W}, a new window was opened
89: onto the interactions of neutron stars with their surroundings.
90: Rotation-powered pulsars drive outflows of particles and
91: electromagnetic energy into the immediate environment; depending
92: primarily on the pulsar's age, this environment might simply be the
93: interstellar medium or, in the case of young, energetic pulsars, the
94: circumstellar debris of the supernova explosion that accompanied the
95: birth of
96: the neutron star. The astrophysical value of these interactions lies
97: in their uniqueness as a probe of the content of pulsar winds, which
98: originate in the incompletely understood magnetospheres of neutron
99: stars, and of the physics of particle acceleration and shock mixing
100: of relativistic and non-relativistic fluids. In contrast to the
101: optical and radio observations that provided early insights into
102: these phenomena, X-ray studies, in particular with {\em Chandra}'s
103: unequalled angular resolution, sample the properties of energetic and short-lived
104: particles nearest the accelerating engines that produce them.
105:
106: In the class of filled-center supernova remnants (SNRs), the bulk of the
107: radiation from a remnant can be traced back energetically to the
108: rapidly rotating neutron star at its core, with emission from the
109: shocked neutron star wind often defining a pulsar wind nebula (PWN).
110: Shell-type remnants, by contrast, expend the energy of the
111: supernova explosion's blast wave. Evidence for the transfer of a
112: neutron star's rotational kinetic energy to its surroundings is
113: provided by the axisymmetric structures---arcs, rings, tori, and
114: jets---seen in high-resolution X-ray images of the environs of a number of pulsars.
115: Together with amorphous nonthermal emission also found around
116: energetic pulsars, these structures are brought to light through synchrotron
117: radiation by the pulsar's ultrarelativistic particle wind where it
118: is confined and perturbed by the surrounding medium; see, e.g.,
119: \citet{2006ARA&A..44...17G} for a review of PWNs. The morphological
120: and spectral properties of PWNs carry valuable information about the
121: physical mechanisms of energy transfer, the local conditions
122: (particle densities, flow velocities, magnetic field strengths,
123: etc.), and the nature of the particle-accelerating central engine.
124:
125: Although its identification as a pulsar wind nebula is now virtually
126: certain, the DA 495 radio nebula (G65.7+1.2; \citealp{1983AJ.....88.1810L})
127: is highly unusual: its morphology is distinctly
128: annular, with a radial profile that rises rapidly from a central
129: emission deficit (the ``hole'') then falls gradually with increasing distance
130: from the center; see Figure~\ref{fig:radio} for a contour map, and
131: \citet{klr+07} for a detailed discussion of the radio properties.
132: We described in an earlier paper \citep[hereafter,
133: ASLK04]{2004ApJ...610L.101A} X-ray observations of DA 495 that
134: revealed the central engine apparently powering the radio nebula, the {\em ROSAT}
135: source 1WGA J1952.2+2925, and provided some evidence for extended
136: X-ray emission consistent with a PWN. To confirm the existence of a
137: high-energy nebula, and because lingering doubts remained about a
138: possible ``thick shell'' thermal origin for the radio annulus
139: \citep{1989JApA...10..161V}, we set out to perform with \cxo\ an
140: X-ray observation sensitive to
141: compact nonthermal structures as a means of
142: strengthening the identification of DA 495 as a PWN, through
143: the detection in particular of jets, toroidal structures, or other
144: markers of large-scale axial symmetry, such as those seen around the
145: Crab, Vela, and similar pulsars.
146:
147:
148: \section{Observations}
149:
150: The \cxo\ observation of DA 495 was carried out on 2002 December 9.
151: Anticipating the need for a large field of view---to search for
152: X-ray emission from the entire $\sim 20'$-diameter radio nebula---we
153: used the Advanced CCD Imaging Spectrometer (ACIS) instrument with
154: the putative pulsar, 1WGA J1952.2+2925 (see ASLK04), placed at the
155: aimpoint of the I3 CCD. The 4-chip ACIS-I array neatly captured most
156: of the radio extent of DA 495 within its $16.9'$ square field of
157: view. Two additional CCDs, the front-illuminated chips ACIS-S2 and
158: -S4 were also activated to provide a measure of the X-ray background
159: in fields adjacent to the radio nebula;
160: ultimately, only S2, which
161: lies closer to the telescope's optical axis, was used.
162:
163: X-ray photon events were acquired in timed-exposure (TE) readouts of
164: the ACIS CCDs, and telemetered to the ground in Very Faint mode. The
165: starting points for our analyses were event lists, aspect solutions,
166: and other standard data products resulting from the third
167: reprocessing (REPRO-III) of raw data at the \cxo\ X-ray Center. We
168: used the CIAO software package (version 3.3.0.1;
169: \citealp{2006SPIE.6270E..60F}) and associated calibration database
170: (CalDB version 3.2.3) for reduction of the ACIS data.
171:
172:
173: \section{Results}
174:
175: Following standard data preparation procedures\footnote{Standard
176: procedures are described in CIAO software analysis guides and
177: ``threads'' available at {\tt http://cxc.harvard.edu/ciao/guides/}.}
178: for filtering (e.g., rejection of events due to particle background
179: rather than cosmic X-rays), effective area and energy calibrations,
180: and mapping of detector to sky coordinates, final ACIS photon event
181: lists were used to cumulate images, spectra, and lightcurves for
182: further analysis. For an observation duration of 25,160 s, a total
183: filtered exposure of 24,843 s was acquired. We found that photon events
184: outside the 0.6--7 keV energy band were consistent with background
185: and provided little useful information; unless noted otherwise, all
186: of the following discussion refers to analysis and results in this
187: energy range.
188:
189: \subsection{Image Analysis}
190:
191: The map of X-ray counts on the sky for all CCDs active during the 25
192: ks \cxo\ observation, binned by a factor of 16, is shown in
193: Figure~\ref{fig:radio}, with radio contours superposed. The basic
194: result of earlier {\em ROSAT} and \asca\ observations (ASLK04), that an
195: X-ray source lies near the center of the DA 495 annulus, is readily
196: confirmed (source `A'). ASLK04 argued further that the \asca\ images
197: were consistent with a combination of soft emission from a compact source
198: and harder emission from a nebula barely resolved by the
199: available $\gtrsim 1'$ angular resolution.
200: The leap to {\em Chandra}'s arcsecond-level imaging upholds this
201: conclusion as well, demonstrating however (as discussed below) that
202: it was fortuitously reached: the extent of the diffuse X-ray
203: emission is too small to have been resolved by \asca. Instead, a
204: small number of unrelated nearby sources were blended by the \asca\
205: point-spread function, giving the appearance of extended nebulosity.
206:
207: To allow for morphological analysis with the best available angular
208: resolution, we formed ACIS images with pixel randomization (used by
209: default in standard CIAO data reduction to avoid spatial
210: digitization effects) disabled; we also applied the sub-pixel event
211: localization algorithm described by \citet{2001ASPC..251..576M}. In
212: practice, we found that these techniques produced no discernible
213: change in the morphology of the diffuse X-ray emission, but yielded
214: a somewhat improved ($\sim 10$\%) concentration of counts onto the
215: point-like source from its immediate vicinity. The coordinates of
216: the unresolved source on the sky are (J2000) RA 19$^{\rm h}$52$^{\rm
217: m}$17\fs04, Dec +29\arcdeg25\arcmin52\farcs5, good to the nominal
218: \cxo\ astrometric uncertainty of $0.6''$.
219:
220: In the full-resolution ACIS-I image, Figure\ \ref{fig:twoenergies},
221: the X-ray nebulosity is revealed to be $\sim 40''$ in diameter. The
222: extended emission surrounds the point-like source but is not
223: centered on it. Instead, there is a strong asymmetry, roughly along
224: an east-west line, in both flux and radial extent, with most of the
225: diffuse flux lying to the east of the unresolved source. To quantify
226: this asymmetry, we sum the counts detected within the eastern and
227: western halves of a $20''$-radius circle centered on the point
228: source, neglecting the latter's flux by excluding the central
229: $1''$-radius circle. Of the 474 counts that define the X-ray nebula
230: and are contained within the larger circle, 314 lie to the east and
231: 160 to the west of the north-south bisector. As shown in Figure
232: \ref{fig:twoenergies}, we choose an extraction region for spectral
233: analysis that is not centered on the point source, in order to
234: better match the asymmetric shape of the nebula. The
235: background-subtracted countrate within this extraction circle was
236: $18.4\pm1.5$ cts~ks$^{-1}$. For the unresolved source, 175 counts
237: were detected in a $1''$-radius region centered midway between the
238: two brightest pixels, for a background-subtracted countrate of
239: $7.0\pm 1.3$ cts ks$^{-1}$; the same region was used to accumulate a
240: spectrum. Figure \ref{fig:twoenergies} displays the \cxo\
241: images in two energy bands, with and without smoothing; the softness
242: of the point source relative to the extended emission is evident.
243: Radial surface brightness profiles of the observed emission, in
244: comparison to the \cxo\ point-spread function, are shown in
245: Figure~\ref{fig:psf}.
246:
247:
248: The low surface brightness of the X-ray nebulosity (0.36
249: cts~arcsec$^{-2}$) and the brevity of the observation conspire
250: to make difficult the identification of any well-defined structures
251: within the nebula. Smoothing of a raw counts image can increase the
252: significance of faint features if their extents are well-matched to
253: the smoothing scale, but the resulting images can also be misleading
254: if over-interpreted. Nevertheless, we experimented with Gaussian and adaptive
255: smoothing techniques to attempt to discern compact structures such
256: as those exemplified by the Crab nebula. Two candidate features
257: emerge. The first is an apparent arc of emission, resembling an
258: inverted hook, 10 pixels ($4.92''$) in extent and immediately
259: adjacent to the point source: it extends initially to the east and
260: then curves south (Fig.~\ref{fig:jet}). Whether this feature is truly curved is unclear,
261: but it is sufficiently bright in the unsmoothed image (61 counts in
262: a $15\times9$-pixel elliptical extraction region) relative to the local
263: background (an average of 21 counts in the same
264: elliptical region positioned at four nearby locations) that it
265: is statistically significant at the $4\sigma$ level. The apparent
266: curvature results from a deficit of counts immediately to the SE of
267: the point source, but statistical fluctuations could account for
268: this deficit.
269:
270: The second structural detail suggested by the data, near the
271: detection limit for low surface-brightness features, is slightly
272: enhanced flux along a line (labeled ``NE Enhancement'' in Fig.\ \ref{fig:jet})
273: extending from the point source to the northeast. In Figure
274: \ref{fig:jet}, we plot the counts contained within a number of
275: identical rectangular strips, centered on the point source and
276: rotated at $10^\circ$ intervals. The resulting measurements are
277: found to be roughly consistent with a constant flux, but although
278: the deviations from the mean are not highly significant, the trend
279: is suggestive of an enhancement along the line at $50^\circ$ north
280: through east. Consistent with the
281: overall east-west asymmetry in the nebular flux, the enhanced
282: emission that may exist to the northeast is clearly nonexistent in
283: the opposite direction, to the southwest of the point source;
284: repeating the counting analysis shown in Figure \ref{fig:jet} with
285: a rectangular region that only samples the semicircular region to
286: the east of the point source (out to the same radius) results in a
287: statistically more significant peak at $50^\circ$.
288:
289: While the evidence for this
290: directionally enhanced flux
291: in the \cxo\ data
292: alone is
293: meager, an independent measurement lends
294: support to make
295: it more compelling (cf.\ Fig.\ \ref{fig:jet}). In the work of
296: \citet[hereafter, KLR+08]{klr+07}, a multi-frequency radio
297: polarization study of DA 495 allows for the measurement of its
298: spatially-varying rotation measure. Combined with modeling of the
299: particle densities within the remnant, the orientation of the
300: intrinsic magnetic field can be derived. The result favors a dipolar
301: field oriented at $50^\circ$, identically to our NE enhancement, in
302: projection on the sky. Thus, two distinct results point to a
303: preferred direction and the possibility of axisymmetry in the
304: morphology of DA 495.
305:
306: We note with interest that the two diffuse features, the
307: hook and the NE enhancement, appear to lie in orthogonal directions, suggestive
308: of equatorial and polar structures seen in many PWNs. We discuss
309: this possibility further below.
310:
311: In our analysis of \asca\ data
312: (ASLK04) we argued---without prior knowledge of the radio nor \cxo\
313: results---that a marginally resolved hard source was extended
314: along a NE-SW direction. Figure\ \ref{fig:jet} overlays the
315: innermost \asca-GIS smoothed-intensity contour onto the \cxo\ image.
316: Although the coincidence of the ASCA contour with the NE enhancement
317: and radio-derived magnetic axis is striking, {\em Chandra}'s
318: superior angular resolution reveals a collection of pointlike, hard
319: X-ray sources that collectively account for the elongated \asca\
320: feature, which is thus an artifact of the telescope's optical
321: limitations rather than a true morphological feature.
322:
323: Aside from the unresolved and diffuse X-ray emission believed to be
324: associated with DA 495, 48 additional sources are seen in
325: the ACIS-I field of view.
326: Two of these lie within the confines of
327: the DA 495 radio nebula but their faintness and spectral properties
328: suggest that they are extragalactic background sources:
329: \begin{description}
330: \item CXOU J195217.1+292532 is unresolved, faint (ten counts in a
331: $3\times3$-pixel area), hard, and lies $21''$ south of source A---it
332: is visible in the hard-band images in Figure~\ref{fig:twoenergies}.
333: No counterpart at another wavelength has been cataloged.
334: \item CXOU J195217.4+292745 (Source C) consists of nine counts in a
335: $3\times3$-pixel area, all between 2 and 8 keV, and is coincident
336: with an unresolved feature in radio images. Future high-resolution
337: radio maps will better distinguish between the nebular radio flux of
338: DA 495 and the contribution of source C, a likely background AGN.
339: \end{description}
340: A third notable source is the brightest in the field, CXOU
341: J195253.1+292633 (Source B). Astrometric coincidence suggests that
342: it is the X-ray counterpart of the star 2MASS J19525309+2926337.
343:
344: \subsection{Spectral Analysis}
345:
346: Spectral analysis was carried out for three source regions: the
347: point source at the heart of DA 495, the surrounding diffuse
348: emission within a $20''$-radius circle, and a larger annulus
349: encompassing the brightest parts of the radio nebula to search for
350: thermal X-ray emission associated with a ``thick shell'' supernova
351: remnant. Suitable background regions, in some cases more than one,
352: were identified for each source region; serendipitous sources were
353: excised and background spectra subtracted from accumulated source
354: spectra prior to fitting with the {\tt Sherpa} software. Results for
355: all spectral fits are summarized in Table~\ref{tab:spec}.
356:
357: Following the results of \kothes, we adopt an assumed distance to DA
358: 495 of 1 kpc when estimating luminosities of the various emission
359: components.
360:
361: \subsubsection{Nebula and Point Source}
362:
363: Point-source and nebular spectra were formed using the {\tt
364: psextract} and {\tt specextract} scripts, respectively, with
365: response matrices for the regions of interest generated using the
366: {\tt mkacisrmf} tool. For the point source, a circular extraction
367: region with a radius of 2 pixels ($\sim 1''$), centered midway between the two
368: brightest pixels, was defined to minimize contamination from the
369: nebula, and extracted events were grouped in energy so
370: that a minimum of 15 photons fell into each spectral bin. For the
371: diffuse emission, the extraction circle (Fig.\
372: \ref{fig:twoenergies}) was $40''$ in diameter and centered $4''$
373: to the northeast of the point source; events were grouped to a
374: minimum of 20 counts per energy bin. To allow for consistency
375: checks, two regions were
376: used to estimate the contribution of background emission to the
377: source spectra (see Fig.\ \ref{fig:radio}): an annular arc to the
378: north of the nebula with inner and outer radii $2.1'$ and $2.5'$
379: respectively, centered on the point source and extending out to the
380: chip gaps to its east and west; and a circular region (radius
381: $1.2'$) at the northern corner of the I3 chip, where there is
382: essentially no radio emission from the remnant.
383:
384: Events contained within the point source region reflect an evidently
385: soft-spectrum origin: just 4 counts were detected in the energy band
386: above 2.5 keV, consistent with the surface brightness of the
387: surrounding diffuse emission, so that the point source is cleanly
388: detected only at low energies (e.g., Fig.~\ref{fig:psf}). A blackbody
389: spectral model adequately describes this emission, with best-fit
390: parameter values that do not
391: depend, to within estimation uncertainties, on which background
392: region is used.
393: In the diffuse emission, there were too few counts to
394: allow for quantitative analysis of any features or sub-regions other
395: than the integrated nebula, which exhibits a hard tail consistent
396: with an absorbed power-law spectral model.
397:
398: Statistically, the most well-constrained spectral fits were those
399: made for the point-source (blackbody) and diffuse (power-law)
400: emissions simultaneously, with a common interstellar absorption
401: parameter, the neutral hydrogen column density $\nh$. We verified
402: that fits to each component separately produced consistent results,
403: finding however that poor statistics prevented measurement of $\nh$
404: in either fit alone---even in the simultaneous analysis, the best-fit
405: value $\nh = (0.23\pm0.17)\times 10^{22}$~cm$^{-2}$ amounts to an
406: upper limit at the $2\sigma$ level. Observed and model spectra
407: derived from the simultaneous analysis are shown in Figure
408: \ref{fig:spec}, together with confidence contours projected onto
409: two-dimensional slices of the full five-dimensional parameter grid.
410: Least squares minimization produced the best fit at a reduced
411: $\chi^2 \simeq 0.5$ for 27 degrees of freedom.
412:
413: For the point source, power-law fits were formally acceptable
414: (reduced $\chi^2 \simeq 0.7$), but produced steep spectra (photon
415: indices $\Gamma \sim 3.5$) when $\nh$ was fixed at the nebular
416: value, and steeper still when $\nh$ was allowed to vary up to five
417: times the nebular value. We prefer instead thermal models: a
418: blackbody spectrum yields a best-fit temperature of
419: $0.23^{+0.06}_{-0.04}$ keV (uncertainties are $1\sigma$ throughout),
420: for a bolometric unabsorbed flux of $\sim 8\times
421: 10^{-14}$~ergs~cm$^{-2}$~s$^{-1}$. The isotropic luminosity of the
422: source is then $L_X \sim 10^{31}\,(d / 1\;{\rm kpc})^2$
423: ergs~s$^{-1}$ for distance $d$, implying for a simple blackbody an
424: emitting area with a characteristic radius of $\sim 0.3\,(d /
425: 1\;{\rm kpc})$ km. This size is consistent with the area of a
426: magnetic polar cap, heated by bombardment of particles accelerated
427: in the magnetosphere, on the surface of a neutron star with a rotation
428: period of several tens of milliseconds (i.e., similar to the Vela
429: pulsar). Propagation effects
430: in stellar atmospheres, however, are believed to significantly
431: distort thermal spectra, making temperature and size estimates
432: unreliable, with pure blackbody fits tending to overestimate
433: temperatures by up to a factor of two depending on the magnetic
434: field strength and composition of the atmosphere.
435:
436: To investigate the possible impact of an atmosphere, we used the
437: {\tt XSPEC} NSA model
438: \citep{1995lns..conf...71P,1996A&A...315..141Z} to fit the
439: point-source spectrum to presumed emission from the entire stellar
440: surface for nonmagnetic atmospheres as well as $B = 10^{12}$ and
441: $10^{13}$ G conditions. For canonical neutron star mass (1.4
442: $M_\sun$) and radius (10 km) assumptions, these fits also provide a
443: distance estimate, as the emitting area seen by a distant observer
444: is fixed. The fits demonstrated that reasonable values of effective
445: surface temperature and distance (shown in Table \ref{tab:spec};
446: their implications are discussed in Sec.~\ref{sec:thermdisc})
447: could be inferred from the available data, but we found that the NSA
448: model parameters, effective temperature and overall normalization,
449: were highly covariant with one another in every case, with the
450: $\chi^2$ surface describing an extended, shallow minimum, precluding a
451: well-bounded measurement of either quantity by itself. The implied
452: luminosity, which depends on both temperature and the normalization,
453: is more reliably constrained. These ill-defined fits are likely the
454: result of poor statistics and the necessarily coarse binning of our
455: observed spectrum; we note, however, that the number of fitting
456: parameters was the same whether we used the NSA models or a simple
457: blackbody, where the latter produced much more robust parameter
458: estimations. Also unlike the blackbody model, the NSA fits all
459: required a hydrogen absorbing column $\nh$ some three times higher
460: than is supported by our preferred power-law fit to the nebular
461: emission surrounding the point source, as described below. When fit
462: simultaneously, $\nh$ remains high while the nebular spectrum
463: exhibits unmodeled excess flux at energies $\lesssim 1$ keV. Attempts
464: to model this excess with additional thermal or nonthermal spectral
465: model components were inconclusive.
466: A deeper observation will be needed to obtain reliable results from
467: neutron star spectral models incorporating atmospheric effects.
468:
469:
470: For the nebula, the best-fit photon index is $\Gamma = 1.63\pm0.27$,
471: for 0.5--10 keV and 2--10 keV unabsorbed fluxes of $2.8\times
472: 10^{-13}$ and $1.8\times10^{-13}$~ergs~cm$^{-2}$~s$^{-1}$,
473: respectively. The implied 0.5--10 keV isotropic luminosity is $L_X
474: \simeq 3.3\times 10^{31}\,(d/1\;{\rm kpc})^2$ ergs~s$^{-1}$. These
475: results are largely in line with the \asca\ measurements reported by
476: ASLK04: the power-law photon index is virtually identical, but the
477: source luminosity derived from \asca\ data is a factor of 4 higher
478: than the \cxo\ value. This discrepancy
479: is attributable to
480: contamination from the point source and other, unrelated, sources
481: that could not be distinguished from the PWN
482: in the low-angular-resolution \asca-GIS observation.
483: The newly measured $\nh$ is also consistent with the
484: \asca-derived value; at roughly one-sixth of the full Galactic
485: column density, it supports prior evidence of a small distance to DA
486: 495.
487:
488: \subsubsection{Constraints on nebular thermal emission}
489:
490: The only significant diffuse X-ray emission in our \cxo\ image is the
491: nebula that lies within $\sim 40''$ of the point source, i.e.,
492: largely inside the radio ``hole.'' Fitting this emission to
493: thermal models yields statistically acceptable results---with, e.g.,
494: a bremsstrahlung temperature of 10 keV---but these temperatures are
495: significantly higher than those typically seen in shock-heated
496: supernova remnant interiors ($\lesssim 3$ keV).
497: A thermal origin of the X-ray
498: nebula is therefore disfavored. More importantly, in a ``thick
499: shell'' interpretation of the radio emission as being thermal in
500: nature, one would expect significant spatial correspondence between
501: thermal X-ray emission and the bright radio annulus. No X-ray
502: emission in an annular or shell-like structure is seen.
503:
504: To quantitatively constrain X-ray flux
505: coincident with the radio annulus, we defined an annular region with
506: inner and outer radii of $1.7'$ and $5.5'$ respectively, centered
507: on the radio hole, from which to extract source photons. Chip gaps
508: were excluded, and unresolved sources detected by the {\tt vpdetect}
509: tool were removed. To estimate the local background, we defined two
510: regions (as a check on the consistency of the results), one at the
511: northern corner of the I3 chip, and another taking up most of the S2
512: chip (cf.\ Fig.\ \ref{fig:radio}). The countrates in both of these
513: detector regions were degraded by $\simeq 15$\% due to mirror
514: vignetting, resulting in an overestimate of the source flux in a
515: background-subtracted analysis. For the purpose of setting a
516: conservative upper limit on the source flux, however, this is
517: acceptable. We find that the maximum allowable count rate in a
518: diffuse component within the radio-bright annulus is 0.048 cts
519: s$^{-1}$ ($3\sigma$) in the 0.5--5 keV band, for a surface
520: brightness $<15$ cts~arcmin$^{-2}$ in 25 ks; corrected for
521: vignetting, the upper limits on photon flux are $9.5\times
522: 10^{-5}$ cts~cm$^{-2}$~s$^{-1}$ and $1.4\times 10^{-6}$
523: cts~cm$^{-2}$~arcmin$^{-2}$.
524:
525:
526:
527:
528:
529:
530: \subsection{Timing Analysis}
531:
532: The photon time tagging resolution provided by the \cxo\ CCDs in
533: full imaging mode is 3.2 s, much too poor to allow for a periodicity
534: search for a typical young, rotation-powered pulsar. We did,
535: however, attempt to confirm the putative 10.9 s pulse period we
536: identified, with low statistical significance, in the \asca-GIS data
537: (ASLK04). In Fourier and Bayesian epoch-folding \citep{gl92b}
538: analyses, we detected no flux modulations at that, or any other,
539: pulse period $\gtrsim 10$ s, nor was there any evidence of flaring
540: or other variability in the point source and nebular fluxes.
541:
542: \section{Discussion}
543: \label{sec:disc}
544:
545:
546: \subsection{Thermal properties of the point source and nebula}
547: \label{sec:thermdisc}
548:
549: For the putative pulsar powering the DA 495 nebula, power-law
550: spectral fits are statistically acceptable but yield
551: uncomfortably large photon indices and significantly larger absorption
552: columns than are supported by the spectral fits to the nebula.
553: For nominal fits to the NSA atmosphere models, the bolometric
554: luminosity for an assumed age may be compared with neutron star
555: cooling curves from theory and the blackbody temperatures and
556: luminosities of other pulsars with similar ages
557: \citep{2002ApJ...571L.143T}. \kothes\ estimate the age of DA 495 at
558: 20 kyr. For this assumed age, the putative neutron star in DA 495 is
559: underluminous by a factor $\sim 5$ compared with expectations from
560: standard cooling models, $L_X \sim 10^{33}$ ergs~s$^{-1}$, assuming the
561: NSA fit results are reliable.
562: The Vela pulsar, among others, is
563: similarly cooler and dimmer than expected \citep{2001ApJ...554L.189P}.
564: For the blackbody fits, the implied area and temperature are not
565: unreasonable for emission from a heated polar cap, but if the
566: observed flux is dominated by polar caps,
567: any contribution from
568: cooling over the remaining surface would have to be significantly
569: (at least an order of magnitude) less luminous, further exacerbating
570: the discrepancy with cooling theory. It seems reasonable to assume, therefore,
571: that the observed thermal emission is dominated by whole-surface
572: cooling emission and that atmospheric effects are important.
573:
574: In the context of attempts to describe DA 495 as a ``thick shell''
575: SNR \citep{1989JApA...10..161V},
576: the absence of detectable diffuse thermal emission, either on large
577: scales or as in 3C58 where thermal and nonthermal emission are mixed
578: \citep{2007ApJ...654..267G}, suggests two possibilities. Either no
579: reverse shock has developed, and therefore the PWN is expanding into
580: cool surroudings (presumably ejecta), as in the Crab, or emissions
581: driven by the reverse shock have already dissipated.
582: That a reverse shock, in the former scenario, has not yet produced
583: significant heating of ejecta nor interacted with the PWN in a $\sim
584: 20$ kyr-old remnant would be somewhat surprising, but the absence of
585: a detectable limb-brightened shell surrounding DA 495 may be
586: significant---the existence of a forward shock is a prerequisite for
587: the formation of a reverse shock.
588: The second case, in which
589: the hot ejecta that signal the passing of a reverse shock have
590: already cooled and dissipated, is difficult to reconcile with a
591: remnant age of just a few tens of kyr. Interaction with
592: a reverse shock would also be expected to significantly compress and distort a PWN
593: \citep{2001A&A...380..309V,2001ApJ...563..806B}, resulting in
594: asymmetries and, sometimes, spatial offsets between the radio and X-ray
595: emissions such as are found in the Vela and SNR G327.1$-$1.1
596: PWNs \citep{2006ARA&A..44...17G}.
597: The E-W asymmetry, relative to the central source, in the
598: morphology of the DA 495 X-ray nebula (a hint of which is also
599: evident in the radio morphology) could conceivably be
600: attributed to distortion from passage of a reverse shock, but there
601: is no measurable offset between the radio and X-ray nebulae,
602: both appearing to be tied to the current pulsar position to within,
603: at worst,
604: the $\sim 45''$ angular resolution of the available radio images.
605: On balance, it seems unlikely that interaction with a reverse shock
606: has occurred, so that DA 495 is not an evolved SNR.
607: The X-ray results therefore are at odds with a thermal origin for the
608: radio emission; \kothes\ provide further
609: arguments against the thick shell interpretation.
610:
611: \subsection{Termination shock}
612:
613: Pulsar wind nebulae are synchrotron structures that depend on the
614: perturbation of free-flowing particles at a standing shock.
615: Of immediate interest in an observed nebula, then, is the
616: possibility of locating this wind-termination shock. Whether the wind
617: is confined by ram pressure induced by the pulsar's motion or simply
618: ambient pressure in a medium, the shock is associated with a
619: transition from an underluminous (free-flowing) zone to a brighter
620: region. The only feature in our \cxo\ image of DA 495 that
621: approximately matches this description is the ``hook'' near the
622: central point source: there is a hint of a deficit of X-ray flux
623: where the hook curves southward. If the deficit is significant,
624: the hook may plausibly be associated with
625: the standing shock where the pulsar wind is
626: equatorially confined.
627: The extent of the hook, $\sim 4''$,
628: corresponds to a physical size of $r_{\rm hook} \sim 0.02\,d_1$ pc,
629: where $d_1 \equiv d/(1\;{\rm kpc})$ for distance $d$.
630:
631: An alternative explanation for the hook is that it represents
632: post-shock emission from an equatorial outflow similar to the torus
633: around the Crab pulsar---for which the wind termination shock is more reliably
634: identified with an inner ring of emission---or an associated wisp,
635: as suggested for PSR B1509$-$58 by
636: \citet{2002ApJ...569..878G}\footnote{Such
637: wisps may arise from thermal instabilities due to
638: synchrotron cooling
639: \citep{1998MmSAI..69..883H} or compression of particle pairs by
640: magnetic reflections of heavy ions \citep{1994ApJ...435..230G}.}.
641: Indeed, the equatorial features seen in most X-ray PWNs are
642: thought to be extended toroidal structures a short distance downstream
643: of the termination shock, rather than ``inner ring'' markers of the
644: shocks themselves. In these cases, one-half of the observed radius
645: serves as a reasonable estimate of the true wind
646: termination shock radius $r_s$ (e.g., \citealp{2007ApJ...660.1413K}).
647: We adopt this interpretation for the DA 495 hook as well, so that
648: $r_s = r_{\rm hook}/2 \sim 0.01\,d_1$ pc, which is
649: consistent with (the low end of) the equatorial termination shock radii around several
650: Vela-like pulsars (see, e.g., Table 2 of \citealp{2007ApJ...660.1413K}).
651: The combination of a $4''$-radius torus surrounding a $2''$-radius ring
652: around an unresolved source would be difficult to
653: disentangle even with \cxo. We cannot rule out the possibility that
654: the termination shock is smaller still than we have assumed, that it
655: is unresolved and cannot be distinguished from the point source. The
656: Crab's inner ring is detected with surface brightness at least as
657: high as the downstream torus---in our dataset, the 4 counts above
658: 2.5 keV within the point-source region represent a surface
659: brightness 50\% higher than that in the hook for the same energy
660: band, and Poisson fluctuations could allow an even higher unresolved
661: nonthermal flux. Nevertheless, our estimate of $r_s$ is at
662: worst a useful upper limit: no known PWNs
663: have estimated termination-shock radii smaller than 0.01 pc, and
664: arguments derived below from $r_s$ estimates generally require an
665: upper limit in any case.
666:
667: Given, then, the tentatively identified shock radius, a
668: constraint on the spin-down luminosity of the pulsar is available
669: through pressure-balance,
670: \begin{equation}
671: \dot E = 4\pi c r_s^2 P_{\rm PWN}/f_\Omega,
672: \end{equation}
673: where $P_{\rm PWN}$ is the nebular pressure, and the factor
674: $f_\Omega$ accounts for a non-isotropic wind;
675: a value $f_\Omega
676: \simeq 3/2$ corresponding to an equatorially-dominant wind is
677: typically assumed (see, e.g., \citealp{2002AstL...28..373B}).
678: Assuming equipartition between magnetic and particle pressures,
679: $P_{\rm PWN}$ may be estimated from the
680: radio-derived average magnetic field, $B_{\rm neb} \simeq
681: 1B_1$ mG (\kothes, and see Sec.~\ref{sync.sec}), where
682: $B_1 \equiv B/(1\;{\rm mG})$,
683: such that $P_{\rm PWN} \simeq 8\,B_1^2\times 10^{-8}$ erg~cm$^{-3}$. The
684: resulting $\dot E$ is $2\,d_1^2\,B_1^2\times10^{37}$ ergs s$^{-1}$.
685: An independent estimate of $\dot E$ is available, in
686: principle, by assuming an efficiency of conversion
687: from spin-down to nebular luminosity, $\eta \equiv L_X/\dot E$,
688: between $10^{-5}$ and $10^{-3}$, consistent with the PWNs seen
689: around Vela-like pulsars (i.e., those with characteristic ages of
690: 10--30 kyr; \citealp{2007ApJ...660.1413K}). For our measured 0.5--8
691: keV nebular luminosity, $L_X = 2.9\times 10^{31}$ ergs~s$^{-1}$,
692: the implied $\dot E$ lies between $3\times 10^{34}$ and $3\times
693: 10^{36}$ ergs~s$^{-1}$, well within the range of known Vela-like
694: pulsars, but a factor of 10--1000 below the value implied by the
695: putative termination shock radius, which is also consistent with the
696: same set of pulsars. Reconciliation of these $\dot E$ estimates
697: would require one of the following to hold true:
698: (1) the termination
699: shock is actually unresolvable ($\lesssim 1''$) and $\log\dot E\lesssim
700: 36$; (2) $\dot E$ is indeed high ($\log\dot E\sim 37$) as required
701: by the hook, but the conversion efficiency $\eta$ to nebular flux is
702: only a few times $10^{-6}$; (3) the pulsar wind flows predominantly
703: into a solid angle much less than assumed by the factor $f_\Omega$, reducing the
704: pressure-derived $\dot E$ estimate by an order of magnitude or more;
705: (4) the nebular pressure estimate, which is based on
706: radio-wavelength measurements made at radii beyond the extent of the
707: observed X-ray nebula, is too high by an order of magnitude or more,
708: and $\log\dot E\sim 35$. Option (1) is possible, as argued above,
709: but would place DA 495 outside the company of Vela-like PWNs in its
710: termination shock radius,
711: while option (2) is difficult to entertain because a
712: $10^{37}$ ergs~s$^{-1}$ pulsar would seem likely to be detectable as
713: a radio or gamma-ray pulsar, or both, and would be significantly
714: younger than the estimated 20 kyr age of DA 495. Options (3) and
715: (4), or a combination of them, therefore appear most reasonable.
716: The assumption of equipartition in deriving $P_{\rm PWN}$, in
717: particular, may be suspect: in the PWN model of
718: \citet{1984ApJ...283..694K}, equipartition holds only at a single
719: post-shock distance that depends on the ratio of electromagnetic to
720: particle flux in the wind, with particle pressure dominating closer
721: in and magnetic pressure further out. Typical equipartition radii
722: are 5--$20 r_s$. Because the radio nebula is brightest at $\gtrsim
723: 30 r_s$, the pressure inferred from the radio-derived magnetic field
724: may significantly overestimate the confining pressure at the wind
725: termination shock. For pressures that correspond to equipartition
726: field strengths comparable to those of PWNs around other Vela-like
727: pulsars, $B_1 \simeq 0.1$, the two methods of estimating $\dot E$
728: are brought in line.
729:
730: \subsection{Morphology and connection to radio properties}
731:
732: Lending further credence to the identification of DA
733: 495 as a PWN is the fact that the combination of a soft point source
734: surrounded by a hard nebula is not uncommon among pulsars that are
735: tens of kyr old. In particular, the Vela pulsar, PSR B1706$-$44, and
736: G189.22+2.90 (IC443) all share this property
737: (\citealp{2001ApJ...554L.189P}, \citealp{2002ApJ...567L.125G}, and
738: \citealp{2006ApJ...648.1037G}, respectively), where Vela and
739: B1706$-$44 inflate wind nebulae confined by the ambient medium while
740: IC443's supersonic motion produces a ram-pressure confined nebula.
741: Intriguingly, sensitive radio imaging and polarimetry of the Vela
742: supernova remnant has shown \citep{2003MNRAS.343..116D} that the
743: pulsar and PWN lie within a radio emission depression straddled by a
744: pair of symmetric polarized lobes, both of which are also features
745: of DA 495.
746:
747: That the pulsar wind in DA 495 is confined by ambient rather than
748: ram pressure is supported by two observations: (1) the X-ray PWN
749: morphology does not resemble a bow shock or narrow ``trail'' of
750: emission (cf.\ PSR B1957+20; \citealp{2003Sci...299.1372S}), and (2)
751: despite the spatial offset of the point source from the center of
752: the radio annulus, a low velocity---at most a few tens of
753: km~s$^{-1}$---is suggested by the anchoring of
754: the polarized radio emission (which is due to long-lived particles
755: injected soon after the pulsar was born) at the current position of
756: the pulsar (\kothes). The radio hole, then, may plausibly be
757: attributed to an absence of synchrotron-emitting particles in a
758: region similar to the so-called optical ``bays'' seen in the Crab
759: nebula \citep{1992ApJ...399..599F}. Regrettably, the spacecraft roll
760: angle during the observation was such that the coordinates of the
761: radio hole largely fell upon one of the chip gaps in the ACIS-I
762: array (Fig.~\ref{fig:radio}), resulting in poor sensitivity to any
763: X-rays from the hole region.
764:
765: Virtually all known PWNs are axisymmetric in their morphologies, and
766: many---e.g., 3C58 \citep{2004ApJ...616..403S} and the
767: PWN in the G54.1+0.3 SNR \citep{2002ApJ...568L..49L}---
768: are notably elongated in X-rays along their apparent pulsar spin axis directions.
769: The only feature in the DA 495 X-ray PWN that hints at axisymmetry is the
770: putative NE enhancement; its existence depends on just a
771: dozen or so detected photons in our \cxo\ image, but the symmetry axis
772: that it suggests is supported by an independent radio measurement.
773: Whether the NE enhancement in DA 495 represents an elongation
774: of the PWN's faint outer reaches or is simply a brightened region is difficult to
775: establish conclusively from the available data. If its
776: orientation does correspond to the spin axis of the neutron star,
777: the feature may plausibly be explained by increased synchrotron lifetimes
778: in a weak axial field relative to the equatorial direction,
779: or perhaps a variation on the proposed process
780: \citep{2002MNRAS.329L..34L,2004A&A...421.1063D} that collimates
781: axial outflows into the narrow relativistic jets that appear to
782: emanate from the rotational poles of the Vela pulsar, B1706$-$44,
783: and many others.
784:
785:
786: Aside from orientation, the radio model of \kothes\ provides
787: further evidence in support of the NE enhancement in DA 495:
788: the radio polarization-derived inclination of the magnetic field
789: vector to the line of sight, $\sim 55^\circ$, shows the field
790: pointing away from us to the SW, so that the \cxo\ brightening to
791: the NE (and perhaps much of the overall E-W asymmetry in X-ray flux)
792: is consistent with Doppler boosting, especially given the
793: relativistic flow speeds implied by the synchrotron cooling time
794: (see below). If the ``hook'' to the east and
795: south represents a portion of the wind termination shock or an
796: equatorial torus, then the
797: DA 495 X-ray structure would be reminiscent of the PWN in G54.1+0.3
798: and other objects (e.g.,
799: \citealp{2004ApJ...601..479N}) in that both the equatorial and polar
800: (jets, in most cases) structures, ostensibly orthogonal to one
801: another, are brightened on the same side of the pulsar, a situation
802: difficult to reconcile entirely with Doppler boosting.
803:
804: Comparison of the radio and X-ray spectra is also instructive.
805: \kothes\ find that the broadband radio spectrum of DA 495 shows a
806: break near 1.3 GHz; the measured spectral slopes below and above the
807: break are $\alpha_{\rm Rlo} = -0.45 \pm 0.2$ and $\alpha_{\rm Rhi} =
808: -0.87 \pm 0.1$ respectively (defined such that flux $S$ at frequency $\nu$
809: is expressed as $S \propto \nu^\alpha$).
810: \kothes\ weigh the available evidence
811: bearing on the origin of the break, whether it is due to synchrotron
812: cooling or is intrinsic to the spectrum of radiating particles
813: injected by the pulsar. A key discriminant is the spectrum of the X-ray
814: PWN, which suggests through the following reasoning that the break
815: is in fact due to cooling and not a consequence of the injected spectrum
816: (see \kothes\ for additional arguments). The radiative lifetimes of
817: synchrotron-emitting particles are such that the bulk of the
818: observed radio emission must arise from particles that were injected
819: into the nebula soon after the pulsar was
820: born, when spin-down luminosity was at its peak.
821: In contrast, the X-ray emission is dominated by
822: particles that have only recently been injected (see below). Thus,
823: if the origin of the break lies in synchrotron cooling, if the
824: injected particle spectrum is unbroken between the radio and
825: X-ray bands, and finally if the injected spectrum has not
826: changed substantially since the pulsar was born (despite
827: declining in particle luminosity) then one expects that the
828: X-ray spectral slope $\alpha_X \equiv 1-\Gamma = -0.63\pm 0.27$
829: should lie somewhere between the two radio spectral slopes, and
830: perhaps closer to $\alpha_{\rm
831: Rlo}$. This is indeed observed, albeit with
832: large combined uncertainties. The general agreement in
833: spectral slopes suggests that the conditions above are not violated.
834: For example, the existence of an injected break at an intermediate
835: frequency between radio and X-ray can be discounted at the $2\sigma$
836: level if it is to resemble those with spectral slope differences
837: $\Delta\alpha > 0.5$ seen at tens of GHz in three filled-center SNRs
838: (G29.7$-$0.3, 3C58, and G16.7+0.1; see \kothes).
839: Similarly, attribution of the observed break at 1.3 GHz to the
840: injected spectrum is formally allowed but somewhat
841: contrived: it would require that $\alpha_X$, $\alpha_{\rm Rhi}$,
842: and $|\alpha_{\rm Rhi} - \alpha_{\rm Rlo}|$ {\em all\/} take on
843: values at or beyond their $1\sigma$ uncertainties so that the X-ray slope
844: is at least as steep as the high-frequency radio
845: slope, and that the magnitude of the 1.3 GHz break is greater than
846: 0.5, for consistency with other injected breaks in the radio band.
847:
848: \subsection{\label{sync.sec}Wind magnetization}
849:
850: Based on the conclusion that the 1.3 GHz spectral break is due
851: to radiation losses, \kothes\ compare the energetics of DA 495 with
852: those of other PWNs and infer a nebular field strength
853: of $B_{\rm neb} \sim 1.3$ mG, a value that exceeds the Crab Nebula's magnetic field
854: by a factor of four. In addition to the radio probes, such as rotation
855: measure, of $B_{\rm neb}$, an independent determination is possible in
856: principle through steepening of the X-ray emission spectrum with
857: increasing radial distance from the pulsar, but
858: because of the short \cxo\ exposure and low surface brightness of
859: DA 495 in X-rays, a significant measurement was not possible.
860: The vastly different sizes of the X-ray and radio nebulae suggest,
861: however, that synchrotron cooling is very efficient, consistent with
862: the high field derived by \kothes---the Crab Nebula, for
863: instance, has a radio-to-X-ray size ratio of roughly 3.3, whereas
864: for DA 495 the ratio is $>25$. Because there is substantial
865: uncertainty in the value of $B_{\rm neb}$, however, we assume a
866: value of 1 mG and carry through scalings with
867: $B_1 \equiv B/(1\;{\rm mG})$
868: when estimating derived quantities.
869:
870: The synchrotron lifetime of particles emitting at energy
871: $\epsilon$ keV in a magnetic field $B$ $\mu$G is $\tau_{\rm
872: synch} = 39 B^{-3/2} \epsilon^{-1/2}$ kyr.
873: For a 1 mG field
874: and emission at $\epsilon = 1.5$ keV, the radiating
875: lifetime is $\tau_{\rm synch} = 1\,B_1^{-3/2}$ yr.
876: Thus, the DA 495 PWN
877: imaged with \cxo\ is dominated by particles roughly one
878: year old. The velocity necessary to
879: traverse the $\gtrsim 20''$, or $0.1\,d_1$ pc,
880: radius of the nebula within the radiating lifetime is $0.33\,B_1^{3/2}c$.
881: \citet{2002ApJ...568L..49L} find a velocity a substantial
882: fraction of $c$ for the
883: G54.1+0.3 PWN in the flow immediately downstream of the
884: termination shock;
885: at larger post-shock distances, a velocity of $\sim 0.01 c$ has been estimated
886: for PSR B1046$-$58 \citep{2006ApJ...652..569G}.
887: In the case of DA 495,
888: the relativistic flow must extend to a great distance post-shock,
889: given that the hook feature provides a reliable upper
890: bound on the termination shock radius $r_s$ at roughly one-tenth the PWN
891: radius.
892:
893: Relativistic bulk velocity at a large downstream
894: distance has been inferred for a linear feature in the N157B PWN
895: \citep{2006ApJ...651..237C}, but is thought to arise for that
896: object from strong
897: pressure confinement due to a nearby dense cloud.
898: In the absence of such external influence, the
899: work of \citet{1984ApJ...283..694K} suggests that relativistic bulk
900: motion at large radii is only possible for values of the
901: ``magnetization parameter,'' the ratio of electromagnetic to
902: particle energy flux in the pulsar wind, of $\sigma \simeq 0.3$
903: which, we argue, applies to DA 495. By
904: contrast, for G54.1+0.3, $\sigma = 0.06$, while for the Crab it is
905: smaller still, $\sigma = 0.003$.
906: In fact, most estimates of $\sigma$
907: to date have found that the observed PWNs are driven by strongly
908: particle-dominated winds (e.g., \citealp{2007ApJ...662..988P}).
909: A notable exception is the Vela X-ray PWN, for which
910: \citet{2001ApJ...556..380H} argue that $\sigma$ is nearly unity,
911: based on the \cxo-resolved termination shock and an earlier estimate
912: of the nebular magnetic field. \citet{2003ApJ...593.1013S} reach a
913: similar conclusion from modeling of the broadband spectrum. Thus, DA
914: 495 provides, potentially, a second example of a pulsar wind in
915: which Poynting and particle flux appear
916: to be comparable.
917: This conclusion is based on an inferred
918: nebular magnetic field and the observed size of the X-ray PWN;
919: it is worth keeping in mind that spin-down luminosity
920: considerations suggest that the nebular pressure, and thus magnetic
921: field strength, in the X-ray nebula may be
922: smaller than implied by the radio $B$-field measurement made at
923: larger distances from the neutron star, underscoring the need for a direct measurement of
924: synchrotron lifetimes through future, high-sensitivity spectroscopy
925: of the X-ray nebula to constrain the burn-off rate via radial
926: steepening of the spectrum.
927:
928: Importantly, \citet{1984ApJ...283..694K}
929: show that, naturally, particle dominated winds are necessary for
930: efficient conversion of wind luminosity to synchrotron luminosity.
931: \citet{2004HEAD....8.0807G} has pointed out that pulsars
932: with spin-down luminosity less than $\simeq 4\times 10^{36}$
933: ergs~s$^{-1}$ host substantially dimmer PWNs than their more luminous
934: counterparts. It is tempting to speculate that the magnetization
935: fraction $\sigma$ varies either with spin-down luminosity or
936: with time such that older, less luminous pulsars produce dim
937: nebulae because Poynting flux contributes substantially to their winds. Current
938: understanding of magnetospheric physics for rotation-powered pulsars
939: argues for a strongly field-dominated wind at the light-cylinder,
940: $\sigma_{\rm LC}\sim 10^4$.
941: The much smaller measured values
942: $\sigma \lesssim 0.01$ in PWNs imply conversion to a
943: strongly particle-dominated wind by the time the flow reaches the
944: termination shock, at least for the youngest and most energetic
945: pulsars. Candidate mechanisms for effecting this conversion have
946: been proposed \citep{2002ApJ...566..336C,2007A&A...466..301C}, but
947: confrontations with observation have not yet been possible. If
948: future PWN studies confirm present suggestions that the winds of
949: Vela-like pulsars emerge from confinement with $\sigma$ of order
950: unity, the implied reduction in the efficacy of the conversion
951: mechanism for older, less luminous pulsars will be an important test
952: of these models.
953:
954:
955:
956:
957: \section{Conclusions}
958:
959: The soft spectrum of the unresolved source near the center of the DA
960: 495 radio nebula, its implied blackbody temperature and luminosity,
961: and its lack of variability on timescales of a few seconds through
962: the 25 ks duration of the observation are highly suggestive of
963: emission from the surface of an isolated neutron star. Taken
964: together with the properties of the surrounding extended, nonthermal
965: emission,
966: no doubt remains that DA 495 is a pulsar wind nebula,
967: driven by the rotation-powered magnetosphere of a neutron star.
968: Its unusual radio
969: characteristics can be explained as the result of aging in the absence of
970: any evident interaction with the supernova explosion's reverse
971: shock, such that DA 495 presents a valuable example of a pristine
972: PWN at an advanced age.
973:
974:
975: \acknowledgements
976: We thank the anonymous referee for a rigorous review that resulted
977: in a much-improved manuscript.
978: Support for this work was provided by the National Aeronautics and
979: Space Administration through \cxo\ Award Number GO3-4092A issued by
980: the \cxo\ X-ray Observatory Center, which is operated by the
981: Smithsonian Astrophysical Observatory for and on behalf of the
982: National Aeronautics and Space Administration under contract
983: NAS8-03060. SSH is supported by the Natural Sciences and Engineering
984: Research Council (NSERC) of Canada and the Canada Research Chair
985: program. The CGPS is a Canadian project with international
986: partners and is supported by the Natural Sciences and Engineering
987: Research Council (NSERC).
988:
989:
990: \bibliographystyle{apj}
991: \bibliography{apj-jour,myapj-jour,psrrefs,./ms}
992:
993: \clearpage
994:
995: \begin{deluxetable}{@{}cccccccc}
996:
997:
998:
999:
1000: \tablecaption{\label{tab:spec}Spectral fitting results for X-ray features in DA 495}
1001:
1002: \tablenum{1}
1003:
1004: \tablehead{\colhead{Emission} & \colhead{Model} & \colhead{$\nh$} & \colhead{$\log(T_{\rm eff})$} & \colhead{$\Gamma$} & \colhead{$\chi^2/\nu$} & \colhead{Dist.} & \colhead{$\log(L_X)$} \\
1005: \colhead{feature} & \colhead{} & \colhead{($10^{21}$ cm$^{-2}$)} & \colhead{(K)} & \colhead{} & \colhead{} & \colhead{(kpc)} & \colhead{(ergs s$^{-1}$)} }
1006:
1007: \startdata
1008: Pt. source & BB & $2.7\pm2.0$ & $6.40\pm0.10$ & \ldots & 4.4/7 & \ldots & 31.0 \\
1009: & H (0 G) & $6.4\pm0.7$ & $\sim5.8$ & \ldots & 5.2/7 & 0.8 & 31.7 \\
1010: & H ($10^{12}$ G) & $6.9\pm0.9$ & $\sim6.0$ & \ldots & 5.9/7 & 1.5 & 32.3 \\
1011: & H ($10^{13}$ G) & $6.8\pm0.9$ & $\sim6.0$ & \ldots & 5.1/7 & 1.4 & 32.3 \\
1012: \hline
1013: Nebula & PL & $1.9\pm1.7$ & \ldots & $1.58\pm0.32$ & 8.4/19 & \ldots & 31.4 \\
1014: \hline
1015: Pt. source & BB & $2.3\pm1.7$ & $6.42\pm0.08$ & \ldots & 12.9/27 & \ldots & 31.0 \\
1016: + nebula & + PL & & \ldots & $1.63\pm0.27$ & & \ldots & 31.3 \\
1017: \hline
1018: Pt. source & H ($10^{13}$ G) & $6.0\pm0.7$ & $\sim 6.0$ & \ldots & 18.7/27 & 1.8 & 32.4 \\
1019: + nebula & + PL & & \ldots & $2.10\pm0.24$ & & \ldots & 31.8 \\
1020: \enddata
1021:
1022:
1023: \tablecomments{Models BB and PL refer to blackbody and power-law spectra,
1024: respectively. H refers to the hydrogen-atmosphere NSA models %in {\tt XSPEC}
1025: with magnetic field strengths $B$ given in parentheses. Luminosities, which
1026: for thermal and nonthermal models are bolometric and in the 2--10 keV band,
1027: respectively, assume a distance of 1 kpc if the spectral model does not provide
1028: an independent estimate.}
1029:
1030:
1031: \end{deluxetable}
1032:
1033: \clearpage
1034:
1035: \begin{figure}
1036: \epsscale{0.95}
1037: \plotone{f1.eps}
1038: \figcaption{\label{fig:radio}
1039: \cxo-ACIS image of the DA 495 field in a 25 ks exposure:
1040: X-ray counts in the 0.6--7 keV band, binned into $8''$-pixels, are
1041: mapped on a logarithmic grayscale with a maximum of 218 counts.
1042: Superposed contours show radio (1.4 GHz Canadian Galactic Plane
1043: Survey [CGPS]; see ASLK04) brightness temperatures between 1 and 6 K
1044: in 1 K steps. Three X-ray sources are labeled. Source A is the
1045: proposed neutron star and wind nebula near the center of DA 495. Source B
1046: is a Guide Star Catalog star previously detected by {\em ROSAT}
1047: (1WGA J1952.8+2926). Source C, newly designated CXOU
1048: J195217.3+292744, is apparently associated with an unresolved radio
1049: enhancement to the north of source A. Event extraction regions used
1050: for spectral background estimates are depicted by dashed circles and
1051: arcs, while the solid rectangles show regions, affected by gaps
1052: between ACIS-I chips, that were excluded from further analysis. }
1053: \end{figure}
1054:
1055: \begin{figure}
1056: \plotone{f2.eps}
1057: \figcaption{\label{fig:twoenergies}
1058: \cxo\ images of the central X-ray source in DA 495.
1059: {\em Upper panels:\/} Raw counts in two energy bands. The grayscale is
1060: logarithmic, with a maximum of 50 counts for the central pixel of
1061: the unresolved source in the 0.6--2 keV image. Solid circles depict
1062: the extraction region for spectral analysis of the nebula; for
1063: clarity, the $1''$-radius extraction region centered on the point
1064: source is not shown. {\em Lower panels:\/} The same images
1065: smoothed by convolving with a Gaussian of width $\sigma = 3$ pixels
1066: ($1.5''$). The grayscale is linear, truncated at 0.6 counts
1067: pixel$^{-2}$.
1068: } \end{figure}
1069:
1070: \begin{figure}
1071: \plotone{f3.eps}
1072: \figcaption{\label{fig:psf}
1073: Radial surface brightness profiles of the background-subtracted
1074: X-ray emission from DA 495 in a series of annuli centered on the
1075: unresolved source; the annuli have widths in radius of $2.5''$
1076: within $25''$ (50 pixels) of the center, and $5''$ beyond. Photon
1077: energy bands for the dotted, dashed, and solid lines are shown in
1078: the legend. The thin solid line corresponds
1079: to the radial profile of the \cxo\ point-spread function simulated
1080: for a source at the same position as DA 495 on the ACIS-I detector
1081: and with the same spectrum. By construction, the integrated area
1082: under the two solid curves is the same.
1083: }
1084: \end{figure}
1085:
1086: \begin{figure}
1087: \plotone{f4.eps}
1088: \figcaption{\label{fig:jet}
1089: Adaptively smoothed 2--8 keV \cxo\ image in the context
1090: of an earlier \asca\ observation and the magnetic field
1091: orientation derived from radio polarization measurements.
1092: For smoothing, a circular convolution kernel of varying size,
1093: requiring a minimum of 5 counts within the kernel diameter, was
1094: applied. The single
1095: heavy contour depicts the 0.73 counts~pixel$^{-2}$ level in the smoothed
1096: \asca-GIS image (Seq.\ 57054000, 45'' Gaussian smoothing; see
1097: ASLK04). \cxo\ resolves a handful of hard X-ray sources that are
1098: responsible for the apparent extent of the low-resolution \asca\
1099: morphology. The dashed lines show the polar ({\em arrow}) and
1100: equatorial directions of the dipole field, on the plane of the sky,
1101: inferred from the radio rotation measure
1102: (\kothes), consistent with the ``NE Enhancement'' in the X-ray PWN
1103: and coincidentally in line with the \asca\ contour. {\em Inset,
1104: upper right:\/} Close-up of the DA 495 PWN in the 0.6--7 keV band,
1105: with adaptive smoothing as above.
1106: The greyscale is linear, truncated at 1
1107: count~pixel$^{-2}$. {\em Inset, lower left:\/} Summed counts within a
1108: rectangular extraction region 5 pixels wide and 150 pixels long,
1109: centered on the point source but excluding it, as a function of its
1110: orientation angle north through east.
1111: } \end{figure}
1112:
1113:
1114: \begin{figure}
1115: \epsscale{0.78}
1116: \plotone{f5.eps}
1117: \figcaption{\label{fig:spec}
1118: Photon flux spectra of the unresolved X-ray source ({\em top left
1119: panel}) and surrounding nebula ({\em bottom left}), fit
1120: simultaneously to blackbody and power-law models respectively, with
1121: a common neutral-hydrogen absorption ($\nh$) parameter. To the
1122: right, projections of $\chi^2$ in three planes of interest are
1123: plotted as contours representing 1, 2, and $3\sigma$ uncertainties
1124: in the parameter values, where $kT$ is the blackbody temperature of
1125: the point source and $\Gamma$ is the photon index of the nebula's
1126: power-law spectrum.
1127: }
1128: \end{figure}
1129:
1130: \end{document}
1131: