1: \documentclass{aastex}
2: %\documentclass[12pt,preprint]{aastex}
3: \usepackage{natbib}
4: %\usepackage{graphicx}
5: \usepackage{emulateapj5}
6:
7:
8: \slugcomment{Submitted to Astrophysical Journal Letters}
9:
10: \shorttitle{PSR J2021+3651}
11: \shortauthors{Roberts et al.}
12:
13: \begin{document}
14: \title{PSR J2021+3651: A YOUNG RADIO PULSAR COINCIDENT WITH AN UNIDENTIFIED EGRET $\gamma$-RAY SOURCE}
15:
16: \author{Mallory S. E. Roberts\altaffilmark{1},
17: Jason W. T. Hessels,
18: Scott M. Ransom\altaffilmark{1},
19: Victoria M. Kaspi\altaffilmark{1}}
20: \affil{Physics Department, McGill University, Rutherford Physics Building,
21: 3600 University Street, Montreal, QC, H3A 2T8, Canada}
22: \email{roberts@physics.mcgill.ca}
23:
24: \author{Paulo C. C. Freire}
25: \affil{NAIC, Arecibo Observatory, HC03 Box 53995, PR 00612, USA}
26: \author{Fronefield Crawford}
27: \affil{Physics Department, Haverford College, Haverford, PA 19041, USA}
28: \and
29: \author{Duncan R. Lorimer}
30: \affil{University of Manchester, Jodrell Bank Observatory, Macclesfield, Cheshire SK11 9DL, UK}
31: \altaffiltext{1}{Center for Space Research, Massachusetts Institute of Technology, Cambridge, MA 02139, USA}
32:
33: \begin{abstract}
34: We report on a deep search for radio pulsations
35: toward five unidentified {\it ASCA} X-ray sources coincident
36: with {\it EGRET} $\gamma$-ray sources.
37: This search has led to the discovery of a young and energetic pulsar
38: using data obtained with the new Wideband Arecibo Pulsar Processor.
39: PSR~J2021+3651 is likely associated
40: with the X-ray source AX J2021.1+3651, which in turn is likely associated with
41: the {\it COS~B} high energy
42: $\gamma$-ray source 2CG~075+00, also known as GeV~J2020+3658 or 3EG~J2021+3716. PSR~J2021+3651
43: has a rotation
44: period $P\cong 104$~ms and $\dot P \cong 9.6\times10^{-14}$, implying
45: a characteristic age $\tau_c\sim 17$ kyr and a spin-down luminosity
46: $\dot E\sim3.4\times 10^{36}$ ergs~s$^{-1}$. The dispersion measure DM$\simeq
47: 371$~pc cm$^{-3}$
48: is by far the highest of any observed pulsar in the Galactic longitude range
49: $55^{\circ} < l < 80^{\circ}$. This DM suggests a distance $d\ga 10$~kpc,
50: and a high $\gamma$-ray efficiency of $\sim$ 15\%, but the true distance may
51: be closer if there is a significant contribution to the DM from
52: excess gas in the Cygnus region.
53: The implied luminosity of the
54: associated X-ray source suggests the X-ray emission is dominated by a
55: pulsar wind nebula unresolved by {\it ASCA}.
56: \end{abstract}
57:
58: \keywords{pulsars: general --- pulsars: individual (PSR J2021+3651) --- stars: neutron --- X-rays: individual (AX J2021.1+3651) --- gamma-rays: individual (GeV 2020+3658)}
59:
60: \section{Introduction}
61:
62: The majority of high energy $\gamma$-ray sources observed by {\it EGRET}
63: and other telescopes have long escaped identification with lower energy
64: counterparts \citep{hbb+99}. Young pulsars remain the only Galactic source class (other than the Sun) unambiguously
65: shown to emit radiation in the 100~MeV -- 10~GeV range \citep{tho01}. It is likely that many of
66: the unidentified $\gamma$-ray sources at low Galactic latitudes are young
67: pulsars as well. Many of these sources have characteristics similar to
68: those of the known $\gamma$-ray pulsars, but have no known pulsars within their error boxes.
69: This fact, along with modelling of the multi-wavelength pulse profiles
70: and the still singular example of Geminga \citep{hh92,hel94}, has led to the suggestion that
71: a large fraction of the radio beams from $\gamma$-ray sources will miss
72: the Earth and appear radio quiet \citep{rom96a}.
73:
74: Recently, a number of young pulsars coincident with known $\gamma$-ray
75: sources have been discovered \citep{dkm+01_mal,cbm+01}. These new discoveries are largely a
76: result of greater sensitivity to pulsars with high dispersion measures (DM)
77: obtainable
78: with newer pulsar backends such as the Parkes multibeam system \citep{mlc+01}.
79: The recent detection of a young radio pulsar
80: in the supernova remnant 3C58 with a 1400~MHz flux density of only
81: $\sim 50$~$\mu$Jy \citep{csl+02} suggests many more faint radio pulsars
82: await discovery in deep, targeted searches.
83:
84: A major stumbling block in the identification of the {\it EGRET} sources is their
85: large positional uncertainty, which can be greater than $1^{\circ}$
86: across. We approach this problem by targeting
87: potential hard X-ray counterparts, whose size and positional
88: uncertainty are much smaller than the typical single dish radio beam.
89: Using as our guide the {\it ASCA} catalog of potential X-ray counterparts of
90: GeV sources (based on the \citet{lm97} catalog of
91: sources with significant flux above 1 GeV) by \citet{rrk01}
92: (hereafter, RRK), we have
93: searched five X-ray sources for radio pulsations using the 305-m Arecibo telescope
94: and the 64-m Parkes telescope (see Table~\ref{tab:observations}).
95: Previous searches of these targets were limited. In particular, two of the
96: three sources
97: observed at Parkes (AX J1418.7$-$6058 and AX J1809.8$-$2332)
98: were not previously the subject of any directed search and were
99: only observed as a matter of course during the Parkes Multibeam Galactic Plane Survey \citep{mlc+01}.
100: A survey of {\it EGRET} sources by \citet{ns97} looked at two of the sources
101: searched here (AX J1826.1$-$1300 and AX J2021.1+3651) with a limiting flux density for
102: slow pulsars of 0.5 -- 1.0~mJy at frequencies of 370 and 1390~MHz, but found no new pulsars.
103: Our search has led to the discovery of one young and energetic pulsar, PSR J2021+3658.
104: We argue that it is a likely counterpart to AX J2021.1+3651 and GeV J2020+3658 / 2CG 075+00.
105:
106: \section{Observations and Analysis}
107:
108: \subsection{Observations}
109:
110: On 2002 January 30 and 31, we observed the only two unidentified
111: sources in the RRK catalog visible from the Arecibo radio
112: telescope, AX J1907.4+0549 and AX J2021.1+3651,
113: using the Wideband Arecibo Pulsar Processor (WAPP).
114: The WAPP is a fast-dump digital
115: correlator with adjustable bandwidth (50 or 100~MHz) and variable numbers
116: of lags and sample times (for details see Dowd, Sisk, \& Hagen
117: 2000\nocite{dsh00}).
118: Our observations were made at 1.4~GHz with 100~MHz of bandwidth and summed
119: polarizations. The observational parameters are summarized in
120: Table~\ref{tab:observations}. The 16-bit samples were written to a disk array
121: and then tranfered to magnetic tape for later analysis.
122:
123: On 2001 February 11$-$15, the three extended hard X-ray sources listed by RRK
124: as potential pulsar wind nebulae, AX J1418.7$-$6058 (the Rabbit),
125: AX J1809.8$-$2333, and
126: AX J1826.1$-$1300, were searched for radio pulsations with the Multibeam
127: receiver on the Parkes radio telescope. Each source was observed once at a
128: central observing frequency of 1390~MHz with a
129: 512 channel filterbank spectrometer covering 256~MHz of bandwith, and once at a central observing frequency of 1373~MHz with 96 channels and 288~MHz of
130: bandwith (see Table~\ref{tab:observations}).
131: During each observation, signals from each channel were square-law
132: detected and added in polarization pairs before undergoing high-pass filtering.
133: The signals were one-bit digitized every 0.25~ms and recorded onto magnetic tape for
134: later analysis.
135:
136: \subsection{Analysis}
137:
138: Analysis of Arecibo observations was done using the {\tt PRESTO} software suite \citep{ran_thesis}
139: by first
140: removing obvious narrow band and/or short duration interference in both
141: the time and frequency domains. We
142: then dedispersed the data at 500 trial DMs between 10 and 510~pc cm$^{-3}$
143: for AX J2021.1+3651 and 540 trial DMs between 0 and 2695~pc cm$^{-3}$ for AX J1907.4+0549.
144: Employing harmonic summing, the FFTs of each time series were searched, and
145: interesting candidates were folded over a fine grid
146: in DM, period, and period derivative space to optimize the signal-to-noise.
147:
148: The Parkes observations were analyzed using standard pulse search software
149: and a similar procedure by searching
150: the 96 channel data at 279 trial DMs ranging from 0 to 1477~pc
151: cm$^{-3}$ and the 512 channel data at 501 trial DMs ranging from 0 to 670~pc
152: cm$^{-3}$.
153: We tested the system by observing a known bright pulsar
154: (PSR B1124$-$60) for 300 s, which was clearly
155: detected in the processing. Re-analysis of the data using {\tt PRESTO}
156: has not revealed any new candidates.
157:
158: \section{RESULTS}
159:
160: \subsection{PSR J2021+3651}
161:
162: A new highly dispersed 104-ms pulsar
163: was detected in the Arecibo observations made of AX J2021.1+3651;
164: it is clearly visible in both of the original search observations, and represents a $\sim$ 40 $\sigma$
165: detection in the longest observation. The pulse
166: profile is shown in Figure~\ref{fig:pulse}.
167:
168: A subsequent series of 7 observations performed between MJD 52405$-$52416
169: allowed us to determine a phase connected solution for some of the
170: pulsar parameters.
171: Integrated pulse profiles from these observations were convolved with a template
172: profile to extract 12 topocentric times of arrival (TOA). Using
173: {\tt TEMPO}\footnote{See http://pulsar.princeton.edu/tempo} and adopting the
174: ROSAT position for the pulsar (\S 4.1), the topocentric TOAs
175: were converted to TOAs at the solar system barycenter at infinite frequency
176: and fit simultaneously for pulsar period, period derivative,
177: and DM, with a residual rms of 91~$\mu$s. The measured and derived parameters
178: for this pulsar are listed in Table~\ref{tab:pulsar}.
179:
180: \subsection{Non-detections}
181:
182: No convincing pulsar candidates were detected in any of the search
183: observations conducted at Parkes. We estimate upper limits of
184: $S \cong 0.08$ mJy at 1.4~GHz for pulse
185: periods $P$ $\ga$ 10 ms for most of the 512 channel observations. A comparable
186: sensitivity was obtained for long periods in the 96 channel
187: observations. For DMs larger than about 100~pc cm$^{-3}$, the sensitivity
188: to fast pulsars (P $\la$ 50 ms) is significantly degraded in the 96
189: channel system. These sensitivity limits were estimated using a
190: sensitivity modeling technique described in detail elsewhere (e.g.,
191: Manchester et al. 2001\nocite{mlc+01}).
192:
193: Likewise, extensive searching of AX J1907.4+0549 yielded no convincing pulsar candidates.
194: We estimate an upper limit of $S\simeq 0.02$~mJy
195: at 1.4~GHz for a long period pulsar assuming a 10\% duty cycle.
196:
197: \section{Discussion}
198:
199: \subsection{PSR J2021+3651}
200:
201: Our search targeted AX J2021.1+3651, which was identified as a potential high-energy counterpart to GeV J2020+3658 by RRK.
202: The X-ray source is near the
203: {\it ASCA} field edge and so the positional uncertainty from {\it ASCA} is $\ga
204: 1^{\prime}$ \citep{guf+00}.
205: A subsequent search of the {\it ROSAT} All-Sky Survey Faint Source
206: catalog \citep{vab+00} revealed the source 1RXS J202104.5+365127 with
207: a smaller positional error of $24^{\prime \prime}$.
208: Given the rarity of such young, energetic pulsars and the small
209: size of the Arecibo beam (3$^\prime$ at FWHM),
210: an association with the X-ray source is highly probable.
211:
212: The DM of PSR J2021+3651 is by far the highest known in the
213: Galactic longitude
214: range $55^{\circ} < l < 80^{\circ}$ which is mainly an inter-spiral arm
215: direction. The \citet{tc93} model gives a distance
216: of $\sim 19$ kpc, well beyond the last spiral arm used in the
217: model. A revised model is currently in preparation by Cordes and Lazio which
218: includes an outer spiral arm at $d\sim 10$~kpc. Placing the pulsar at the far
219: edge of
220: this outer arm still does not account for all the
221: observed dispersion; and it is possible that there are further contributions
222: from clouds in the Cygnus region,
223: where there is known to be excess gas at $d\sim 1.5$~kpc (J. Cordes,
224: private communication). However, there are no obvious HII regions
225: within the Arecibo beam seen in either Very Large Array (VLA) 20-cm radio or
226: Midcourse Space Experiment (MSX) 8.3~$\mu$m images (available from
227: the NASA/IPAC Infrared Science Archive).
228:
229: The high DM is somewhat surprising given the X-ray absorption quoted by
230: RRK, n$_{\rm H}$=(5.0$\pm0.25)\times 10^{21}$~cm$^{-2}$, where the errors
231: represent the 90\% confidence region.
232: The total Galactic HI column density in this direction as estimated from the
233: FTOOL {\it nh}, which uses the HI map of \citet{dl90}, is $1.2\times 10^{22}$~cm$^{-2}$. This should be
234: a good approximation if the source is truly at the far edge of the
235: outer spiral arm. Noting that the {\it ASCA} image shows faint, softer emission
236: in the region (Figure~\ref{fig:xray}), and given the likely possibility of either
237: associated thermal X-ray flux from a supernova remnant or a nearby massive star, we fit the {\it ASCA} spectrum of RRK,
238: adding a thermal component to the absorbed power-law model.
239: Accounting for $\sim4$\% of the photon flux with a MEKAL thermal plasma model
240: of temperature $kT\sim 0.1$~keV in XSPEC \citep{arn96} statistically
241: improves the fit ({\it F}-test chance probability of 2.5\%). The
242: best-fit absorption for this three component model is
243: n$_{\rm H}$=7.6$\times 10^{21}$~cm$^{-2}$ with a 90\% confidence region
244: of (4.1 -- 12.3)$\times 10^{21}$~cm$^{-2}$, consistent
245: with the total Galactic column density. The best-fit photon index
246: is $\Gamma=1.86$, still consistent with the 1.47 -- 2.01 range in RRK
247: derived from the simple absorbed power-law model. Hence the X-ray
248: absorption does not force us to adopt a smaller distance than is suggested
249: by the DM.
250:
251: For a distance $d_{10}=d/10$~kpc, the inferred
252: isotropic X-ray luminosity $L_X=4.8\times 10^{34} d_{10}^2$ (2 -- 10~keV).
253: The X-ray efficiency $\eta_X=L_X/\dot E$ is 0.01$d^2_{10}$.
254: Compared to
255: the total pulsar plus nebula X-ray luminosity of other spin-powered pulsars
256: this is somewhat high, but within the observed scatter
257: \citep{pccm02,che00}.
258:
259: The pulsar's positional coincidence with the error box of the hard spectrum,
260: low variability {\it EGRET}
261: $\gamma$-ray source GeV J2020+3658 coupled with the high inferred spin-down
262: luminosity strongly suggests this pulsar emits pulsed $\gamma$-rays.
263: Unfortunately, confirming this by folding archival
264: {\it EGRET} data is problematic due to
265: the likelihood of significant past timing noise and glitches,
266: which make the back-extrapolation
267: of the rotational ephemeris uncertain. RRK noted that the chance
268: probability of an X-ray source as bright as AX J2021.1+3651 in the
269: {\it EGRET} error box was $\sim10$\%, but the nearby Wolf-Rayet star WR141
270: was equally bright in X-rays and also a potential $\gamma$-ray emitter.
271: However, young pulsars remain the only firmly established class of
272: Galactic {\it EGRET} sources. The known
273: $\gamma$-ray pulsars cluster at the top of pulsar lists rank-ordered
274: by spin-down flux $\dot E/d^2$, with $\gamma$-ray efficiencies
275: $\eta_\gamma=L_{\gamma}/\dot E$ mostly between 0.001 and 0.03
276: (assuming 1 sr beaming) with a tendency
277: to increase with pulsar age \citep{tbb+99}. The exception
278: is PSR B1055$-$52, with an apparent $\gamma$-ray efficiency $\eta_\gamma \sim 0.2$
279: given its nominal DM distance of 1.5 kpc. The inferred $\gamma$-ray efficiency
280: for PSR J2021+3651 is
281: $\eta_\gamma =0.15 d_{10}^2$ in the 100~MeV to 10~GeV range.
282: If the pulsar is located within the Perseus arm at
283: a distance of 5~kpc, then the inferred X-ray and $\gamma$-ray luminosities
284: would be fairly typical of the other pulsars with Vela-like spin-down
285: luminosities. While there is currently no observational evidence for
286: a distance this close, increased DM from an
287: intervening source in this relatively crowded direction would not
288: be surprising. We note that the DM derived distance for another young pulsar
289: recently discovered within an $EGRET$ error box, PSR J2229+6114, also
290: leads to an anomalously high inferred $\gamma$-ray efficiency \citep{hcg+01}.
291:
292: \subsection{Upper Limits Toward the Other Sources}
293:
294: Determining the fraction of radio-quiet versus radio-loud pulsars
295: is important for our understanding of $\gamma$-ray pulsar emission mechanisms.
296: The two leading classes of emission models, the outer-gap \citep{rom96a} and
297: polar-cap \citep{dh96} models, make very different estimates of the
298: fraction of $\gamma$-ray pulsars that should be seen at radio energies.
299: Out of the 25 brightest sources above 1 GeV not associated with blazars,
300: $\sim 10$ are now known to either be energetic radio pulsars or contain
301: such pulsars within their error boxes.
302: Searching the brightest unidentified X-ray
303: sources in five GeV error boxes, we
304: detected radio pulsations at the $\sim 0.1$~mJy level (similar to the limiting sensitivity of the
305: Parkes observations) from one of these
306: with Arecibo. This is well below the average flux level expected
307: for typical radio luminosities of young pulsars \citep{bj99}
308: and distances to star forming
309: regions statistically associated with $\gamma$-ray sources
310: \citep{yr97}.
311: Two of the sources observed with Parkes, AX J1418.7$-$6058 (the Rabbit) and AX J1809.8$-$2333,
312: have radio and X-ray properties that clearly identify them
313: as pulsar wind nebulae \citep{rrjg99,brrk02}, and the third, AX J1826.1$-$1300,
314: is an extended hard X-ray source that has few other source class
315: options. Therefore, all three remain viable candidates for $\gamma$-ray
316: loud, radio-quiet pulsars. Out of this same sample of 25 bright GeV
317: sources, the total number of reasonable
318: candidate neutron stars within the $\gamma$-ray error boxes
319: which have now been searched deeply
320: for radio-pulsations without success is $\sim 7$.
321: A current ``best guess" fraction of radio-loud $\gamma$-ray pulsars of
322: $\sim 1/2$ falls in between the predictions of the two main competing
323: models.
324:
325: \acknowledgments
326:
327: We thank Jim Cordes for useful discussions. We acknowledge support from
328: NSERC, CFI, an NSF CAREER Award, and a Sloan Fellowship.
329: M.S.E.R. is a Quebec Merit fellow. S.M.R. is a Tomlinson fellow.
330: J.W.T.H. is an NSERC PGS~A fellow, V.M.K. is a Canada Research Chair.
331: The Arecibo Observatory is part of the National Astronomy and Ionosphere
332: Center, which is operated by Cornell University under a cooperative
333: agreement with the National Science Foundation. The Parkes radio
334: telescope is part of the Australia Telescope, which is funded by the
335: Commonwealth of Australia for operation as a National Facility
336: managed by CSIRO.
337:
338:
339: \bibliographystyle{apj}
340: \bibliography{journals1,modrefs,psrrefs,crossrefs,J2020}
341:
342:
343: \begin{deluxetable}{c c c c c c c c c c c}
344: \rotate
345: \tablewidth{0pt}
346: \tabletypesize{\scriptsize}
347: \tablecaption{Observational Parameters \label{tab:observations}}
348: \tablehead{\colhead{Source} & \colhead{RA (J2000)} & \colhead{DEC (J2000)} & \colhead{Epoch} & \colhead{Telescope} &
349: \colhead{$\nu_c$\tablenotemark{a}} & \colhead{$\Delta\nu$\tablenotemark{b}} & \colhead{N$_{\rm ch}$\tablenotemark{c}} &
350: \colhead{t$_{\rm sam}$\tablenotemark{d}} & \colhead{T$_{\rm int}$\tablenotemark{e}} & \colhead{$S_{\rm min}$\tablenotemark{f}} \\
351: \colhead{Name} & \colhead{hh:mm:ss.s} & \colhead{dd:mm:ss} &\colhead{(MJD)} & & \colhead{(MHz)} & \colhead{(MHz)} & &
352: \colhead{($\mu$s)} & \colhead{(s)} & \colhead{(mJy)}}
353:
354: \startdata
355: AX J1418.7$-$6058 / GeV J1417$-$6100 & 14:18:41.5 & $-$60:58:11 & 51951.63 & Parkes & 1390 & 256 & 512 & 250 & 16900 & 0.08\\
356: & & & 51953.67 & Parkes & 1373 & 288 & 96 & 250 & 16900 & 0.08\\
357:
358: AX J1809.8$-$2332 / GeV J1809$-$2327 & 18:09:50.2 & $-$23:32:23 & 51951.83 & Parkes & 1390 & 256 & 512 & 250 & 16900 & 0.08\\
359: & & & 51954.77 & Parkes & 1373 & 288 & 96 & 250 & 16900 & 0.08\\
360:
361: AX J1826.1$-$1300 / GeV J1825$-$1310 & 18:26:04.9 & $-$12:59:48 & 51952.83 & Parkes & 1390 & 256 & 512 & 250 & 13234 & 0.09 \\
362: & & & 51955.83 & Parkes & 1373 & 288 & 96 & 250 & 14832 & 0.09 \\
363:
364: AX J1907.4+0549 / GeV J1907+0557 & 19:07:21.3 & +05:49:14 & 52305.58 & Arecibo & 1425 & 100 & 512 & 200 & 6480 & 0.02\\
365:
366: AX J2021.1+3651 / GeV J2020+3658 & 20:21:07.8 & +36:51:19 & 52304.67 & Arecibo & 1425 & 100 & 512 & 200 & 1627 & 0.04 \\
367: & & & 52305.66 & Arecibo & 1425 & 100 & 256 & 200 & 3000 & 0.03 \\
368:
369:
370: \enddata
371: \tablenotetext{a}{Central observing frequency.}
372: \tablenotetext{b}{Total bandwidth of the observation.}
373: \tablenotetext{c}{Number of frequency channels.}
374: \tablenotetext{d}{Sampling time.}
375: \tablenotetext{e}{Length of the observation.}
376: \tablenotetext{f}{Flux density sensitivity limit.}
377:
378:
379: \end{deluxetable}
380:
381:
382: \begin{deluxetable}{l c}
383: \tabletypesize{\scriptsize}
384: \tablewidth{0pt}
385: \tablecaption{Measured and Derived Parameters for PSR~J2021+3651 \label{tab:pulsar}}
386: \tablehead{\colhead{Parameter} & \colhead{Value}}
387:
388: \startdata
389:
390: Right ascension $\alpha$ (J2000) & 20$^{\rm{h}}$ 21$^{\rm{m}}$
391: 04$^{\rm{s}}$.5\tablenotemark{a}
392: \\
393: Declination $\delta$ (J2000) & +36$^\circ$ 51$'$ 27$''$.0\tablenotemark{a}\\
394: Galactic latitude $l$ & 75.$^{\circ}$23 \\
395: Galactic longitude $b$ & +0.$^{\circ}$11 \\
396: Pulse period $P$ (s) & 0.10372222480(17) \\
397: Period derivative $\dot P$ & $ 9.563(48) \times 10^{-14}$ \\
398: Pulse frequency $\nu$ (s$^{-1}$) & 9.641135271(16) \\
399: Frequency derivative $\dot{\nu}$ (s$^{-2}$) & $-8.889(45) \times 10^{-12}$ \\
400: Epoch (MJD) & 52407.389 \\
401: Dispersion Measure DM (pc cm$^{-3}$) & 371(3) \\
402: Pulse width at 50\% of peak w$_{50}$ (ms) & 9.9 \\
403: Pulse width at 10\% of peak w$_{10}$ (ms) & 18 \\
404: Flux density at 1425 MHz (mJy) & $\sim 0.1$ \\
405: Spin-down luminosity $\dot{E}$\tablenotemark{b} (erg s$^{-1}$) & $3.4 \times 10^{36}$ \\
406: Surface dipole magnetic field $B$\tablenotemark{c} (G) & $3.2 \times 10^{12}$ \\
407: Characteristic Age $\tau_{c} \equiv \frac{1}{2} P / \dot P$ (kyr) & 17 \\
408:
409: \enddata
410:
411: \tablecomments{ Figures in parentheses represent uncertainty in the least-significant digits quoted, equal to 3 times errors given by {\tt TEMPO}.}
412:
413: \tablenotetext{a}{ Coordinates of the $ROSAT$ X-ray source
414: 1RXS J202104.5+365127 with estimated positional error of $24^{''}$.}
415:
416: \tablenotetext{b}{ $\dot E = 4 {\pi}^2 I \dot P / P^3$ with $I = 10^{45}$ g cm$^2$. }
417:
418: \tablenotetext{c}{ Assuming standard magnetic dipole spindown:
419: $B = 3.2 \times 10^{19} (P \dot P)^{1/2}$ Gauss \citep{mt77}.}
420:
421: %\tablenotetext{e}{Calculated using the \citet{tc93} Galactic electron density model}
422:
423: \end{deluxetable}
424:
425: \clearpage
426:
427: \begin{figure}
428: \plotone{f1}
429: \figcaption{1.4~GHz pulse profile for PSR J2021+3651 from the MJD 52305 observation. The error bar represents the 1~$\sigma$ uncertainty. \label{fig:pulse}}
430: \end{figure}
431:
432: \begin{figure}
433: \plotone{f2}
434: \figcaption{ASCA GIS 2-10 keV image of the $\gamma$-ray source region.
435: The contours are the 68\%, 95\%, and 99\% confidence
436: regions of the $\gamma$-ray source position, derived from the $>$ 1~GeV photons
437: \citep{rrk01}. WR 141 is a Wolf-Rayet star also in the field of view. The circle
438: centered on AX J2021.1+3651 indicates the size of the $3'$ (FWHM) Arecibo beam.
439: \label{fig:xray}}
440: \end{figure}
441:
442: \end{document}
443:
444:
445: EXTRA BITS
446: and a source extent of
447: $32^{\prime \prime}$, which would be unresolved by {\it ASCA}.
448: We adopt the {\rm ROSAT} source position for the pulsar, and assume the X-ray
449: emission comes partially from an extended pulsar wind nebula, although
450: from only
451: $\sim 22$ {\rm ROSAT} counts, the true source extent and structure cannot
452: currently be determined.
453:
454:
455:
456: & 20:21:07.8 & +36:51:19 & 52343.58 & Arecibo & 1410 & 100 & 512 & 200 & 600 & $\dots$ \\
457: & 20:21:07.8 & +36:51:19 & 52344.55 & Arecibo & 1410 & 100 & 512 & 200 & 1800 & $\dots$ \\
458: & 20:21:04.5 & +36:51:27 & 52379.48 & Arecibo\tablenotemark{b} & 1425 & 100 & 512 & 256 & 1000 & $\dots$\\
459: & 20:21:04.5 & +36:51:27 & 52402. & Arecibo & 1475 & 100 & 512 & 180 & 1200 & $\dots$\\
460: & 20:21:04.5 & +36:51:27 & 52405. & Arecibo & 1475 & 100 & 512 & 180 & 1000 & $\dots$\\
461: & 20:21:04.5 & +36:51:27 & 52407. & Arecibo & 1400 & 100 & 512 & 180 & 2280 & $\dots$\\
462: & 20:21:04.5 & +36:51:27 & 52408. & Arecibo & 1400 & 100 & 512 & 200 & 2100 & $\dots$\\
463: & 20:21:04.5 & +36:51:27 & 52409. & Arecibo & 1425 & 100 & 512 & 200 & 2400 & $\dots$\\
464: & 20:21:04.5 & +36:51:27 & 52410. & Arecibo & 1425 & 100 & 512 & 200 & 2000 & $\dots$\\
465: