0810.0008/em.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass[apj]{emulateapj}
3: %\voffset 0.25in
4: 
5: \received{2008 August 13}
6: \accepted{2008 September 30}
7: %\journalid{}{}
8: %\articleid{}{}
9: 
10: %\usepackage{epsfig}
11: 
12: \def\ts{{\thinspace}}
13: \def\simgt{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
14: \def\simlt{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
15: \def\sp{\hspace{1.5pt}}
16: \def\jump{\hskip 0.2in}
17: 
18: \def\s{\phantom}
19: \def\ss{\phantom\phantom}
20: \def\sss{\phantom\phantom\phantom}
21: \def\ssss{\phantom\phantom\phantom\phantom}
22: \def\sp{\hspace{1.5pt}}
23: \def\etal{{et~al.}}
24: \def\amin{\ifmmode^{\prime}\else$^{\prime}$\fi}
25: \def\asec{\ifmmode^{\prime\prime}\else$^{\prime\prime}$\fi}
26: \def\ts{{\thinspace}}
27: \def\simgt{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
28: \def\simlt{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
29: \def\jump{\hskip 0.2in}
30: 
31: \newcommand{\loe}{\stackrel{<}{\sim}}
32: \newcommand{\goe}{\stackrel{>}{\sim}}
33: \newcommand\asca{{\it ASCA\/}}
34: \newcommand\chandra{{\it Chandra}}
35: \newcommand\glast{{\it GLAST\/}}
36: \newcommand\agile{{\it AGILE\/}}
37: \newcommand\cgro{{\it CGRO\/}}
38: 
39: \def\sp{\hskip 1.5pt}
40: \newcommand\pwna{G75.2+0.1}
41: \newcommand\pwnb{G74.9+1.2 = CTB 87}
42: \newcommand\mgro{MGRO~J2019+37}
43: \newcommand\asrc{CXOU~J20210546+3651048}
44: \newcommand\egra{3EG~J2021+3716}
45: \newcommand\egrb{3EG~J2016+3657}
46: \newcommand\gev{GeV~J2020+3658}
47: \newcommand\blazar{B2013+370}
48: \newcommand\cxo{CXOU~J183803.1$-$3651}
49: \newcommand\psr{PSR~J2021+3651}
50: \newcommand\agl{AGL~J2020.5+3653}
51: 
52: \newcommand\phndn{\phm{0.0}}
53: 
54: \slugcomment{The Astrophysical Journal Letters}
55: 
56: \shorttitle{Discovery of Gamma-Ray Pulsations from PSR J2021+3651}
57: \shortauthors{Halpern et al.}
58: 
59: \begin{document}
60: 
61: \title{Discovery of High-Energy Gamma-Ray Pulsations from PSR J2021+3651 with AGILE} 
62: 
63: \author{J. P. Halpern\altaffilmark{1}, F.~Camilo\altaffilmark{1},
64: A.~Giuliani\altaffilmark{2},
65: E.~V.~Gotthelf\altaffilmark{1}, M.~A. McLaughlin\altaffilmark{3},
66: R.~Mukherjee\altaffilmark{4},
67: A.~Pellizzoni\altaffilmark{2}, S.~M. Ransom\altaffilmark{5},
68: M.~S.~E. Roberts\altaffilmark{6},
69: and M.~Tavani\altaffilmark{7,8}}
70: 
71: \altaffiltext{1}{Columbia Astrophysics Laboratory, Columbia University,
72: New York, NY 10027.}
73: 
74: \altaffiltext{2}{INAF/IASF-Milano, I-20133 Milano, Italy.}
75: 
76: \altaffiltext{3}{Department of Physics, West Virginia University,
77: Morgantown, WV 26501.}
78: 
79: \altaffiltext{4}{Department of Physics and Astronomy, Barnard College,
80: Columbia University, New York, NY 10027.}
81: 
82: \altaffiltext{5}{National Radio Astronomy Observatory,
83: Charlottesville, VA 22903.}
84: 
85: \altaffiltext{6}{Eureka Scientific, Inc., Oakland, CA 94602.}
86: 
87: \altaffiltext{7}{INAF/IASF-Roma, I-00133 Roma, Italy.}
88: 
89: \altaffiltext{8}{Dipartimento di Fisica, Universit\`a ``Tor Vergata'', 
90: I-00133 Roma, Italy.}
91: 
92: \begin{abstract}
93: 
94: Discovered after the end of the {\it Compton Gamma-Ray Observatory}
95: mission, the radio pulsar \psr\ was long considered a likely
96: counterpart of the high-energy $\gamma$-ray source 2CG~075+00 = \egra\ = \gev,
97: but it could not be confirmed due to the lack of a contemporaneous
98: radio pulsar ephemeris to fold the sparse, archival $\gamma$-ray photons.
99: Here, we report the discovery of $\gamma$-ray pulsations from \psr\
100: in the 100--1500 MeV range
101: using data from the \agile\ satellite gathered over 8 months, folded
102: on a densely sampled, contemporaneous radio ephemeris obtained for this
103: purpose at the Green Bank Telescope.  The $\gamma$-ray pulse consists
104: of two sharp peaks separated by $0.47\pm 0.01$ cycles.  The single radio
105: pulse leads the first $\gamma$-ray peak by $0.165\pm 0.010$ cycles.
106: These properties are similar to those of other $\gamma$-ray pulsars,
107: and the phase relationship of the peaks can be interpreted in the
108: context of the outer-gap accelerator model for $\gamma$-ray emission.
109: Pulse-phase resolved images show that there is only one dominant source,
110: \agl\ = \psr, in the region previously containing confused sources
111: \egra\ and \egrb.
112: 
113: 
114: \end{abstract}
115: \keywords{gamma-rays: observations --- pulsars: individual (\psr)}
116: 
117: \section{Introduction}
118: 
119: Rotation-powered pulsars are the dominant class of high-energy
120: ($> 100$ MeV) $\gamma$-ray sources in the Galaxy.  The
121: Energetic Gamma-Ray Experiment Telescope (EGRET) on the
122: {\it Compton Gamma-Ray Observatory} (\cgro), 
123: which operated between 1991 and 2000, detected six pulsars,
124: and three others with less confidence.
125: (See \citealt{tho04} for a review of these EGRET results.)
126: %For young and middle-aged pulsars the maximum efficiency of
127: %converting rotational energy into radiation occurs in the
128: %100~MeV -- 1~GeV range, which makes this energy band of prime
129: %importance for pulsar physics.
130: It is widely assumed that a large
131: number of unidentified $\gamma$-ray sources in the Galactic
132: plane are also pulsars.  Some of these are positionally coincident
133: with radio pulsars, but do not have enough photons to recover the
134: pulse period, while others may be radio quiet.  An ephemeris
135: from radio or X-ray observations is generally
136: needed to fold the sparse $\gamma$-ray source photons over a
137: span of months to years in order to build up a statistically
138: significant pulsed
139: light curve.  In several cases, a radio pulsar candidate
140: was only discovered after the end of the \cgro\ mission, so that
141: the needed contemporaneous ephemeris does not exist.  The typical
142: ``timing noise'' of young pulsars causes a phase ephemeris to
143: fail when extrapolated over months to years.
144: 
145: Some of the still-unidentified Galactic $\gamma$-ray sources 
146: date from the 1975 {\it COS B} mission \citep{swa81}.
147: The target of this work is the {\it COS B} source 2CG 075+00
148: located in Cygnus.
149: EGRET possibly resolved 2CG 075+00
150: into two sources, \egra\ and the weaker \egrb\ \citep{har99}.
151: Multiwavelength study of these regions by \citet{muk00}
152: and \citet{hal01} uncovered possible counterparts for \egrb,
153: but no obvious candidate for \egra.  A smaller error region
154: was derived by \citet{lam97} by restricting analysis
155: to energies above 1~GeV, yielding a source \gev\
156: that overlapped with \egra.  An \asca\ observation
157: of \gev\ found two compact X-ray sources \citep{rob01},
158: one of which was discovered to contain a new, energetic
159: radio pulsar, \psr\ \citep{rob02}.  Its spin period $P = 0.103$~s,
160: characteristic age $\tau_c \equiv P/2\dot P$ = 17~kyr, and spin-down
161: luminosity $\dot E = 3.4 \times 10^{36}$ ergs s$^{-1}$ are in the
162: range of other young pulsars that were detected by EGRET, but its
163: initial distance estimate of $\sim 12$~kpc implied an unusually
164: large $\gamma$-ray efficiency (see \S 3.3) if \psr\ is
165: responsible for \egra.  It was not possible to extrapolate
166: the radio ephemeris backward to search
167: for the pulsar in EGRET $\gamma$-ray photons.
168: By searching in a range around the expected $P$ and $\dot P$,
169: \citet{mc04} possibly detected pulsations, but only in
170: two out of eight EGRET viewing periods.
171: Thus, the proposed identification of \psr\ with \egra\ and
172: \gev\ could not be confirmed, until now.
173: 
174: The \agile\ ({\it Astro-rivelatore Gamma a Immagini LEggero})
175: mission of the Italian Space Agency (ASI) was launched on 2007 April 23
176: into a low-Earth orbit \citep{tav08}.  \agile\ will allow a number of
177: new $\gamma$-ray pulsars to be found by folding on contemporaneous
178: pulsar ephemerides that are being constructed by our group
179: and others using several radio telescopes.  For the
180: analysis of \psr, we have been using the NRAO Green Bank Telescope
181: (GBT).  In this Letter, we apply the ephemeris of \psr\
182: to data obtained by \agile\ during the first year
183: of the mission.
184: 
185: \section{Observations and Analysis}
186: 
187: The \agile\ GRID (Gamma-Ray Imaging Detector) comprises 
188: a Silicon-Tungsten tracker \citep{bar01,pre03}
189: and a CsI(Tl) mini-calorimeter \citep{lab06}.  These provide effective
190: area of $\sim 500$~cm$^2$ in the 30 MeV--50 GeV range over
191: a $\sim 2.5$ sr field of view.  We used
192: GRID observations of the Cygnus region during five
193: periods between 2007 November 2 and 2008 June 30
194: that sum to 103 days of observation when
195: \psr\ was within $40^{\circ}$ of the pointing direction.
196: GRID photons are time-tagged with a precision of $\sim 1\ \mu$s,
197: and the arrival times were transformed to the solar system barycenter
198: in Barycentric Dynamical Time using GPS positioning of the
199: spacecraft and the \chandra\ measured position of the pulsar,
200: (J2000.0) R.A. = $20^{\rm h}21^{\rm m}05.46^{\rm s}$,
201: decl. = $+36^{\circ}51^{\prime}04.8^{\prime\prime}$, uncertainty
202: $\sim 0.5^{\prime\prime}$ \citep{hes04}.
203: The accuracy of this procedure is $\sim 10\ \mu$s, and the absolute
204: times were verified to a precision of $\sim 400\ \mu$s using observations
205: of the Crab pulsar and others \citep{pel08}.
206: %The latter paper describes
207: %additional details of \agile\ pulsar analysis on the Crab, Vela,
208: %Geminga, and PSR B1706$-$44.
209: 
210: A phase-connected rotational ephemeris for \psr\ contemporaneous with all
211: of the \agile\ observing periods was derived from 36 observations obtained
212: at the GBT between 2007 September 28 and 2008 June 30.  The pulsar was
213: observed with the GBT Pulsar Spigot \citep{kap05},
214: yielding total power samples every
215: $81.92\ \mu$s in each of 768 frequency channels spanning a bandwidth of
216: 600\,MHz centered on 1950\ MHz.  Each observation usually lasted for 5
217: minutes, from which we derived a time of arrival with typical uncertainty
218: of 0.2\,ms.  We used the
219: TEMPO\footnote{See http://www.atnf.csiro.au/research/pulsar/tempo.}
220: timing software to obtain the
221: ephemeris using standard methods.  Like many young pulsars, \psr\ exhibits
222: significant rotational instability.  We parameterized this by fitting for
223: the rotational frequency and its first six derivatives.  This timing
224: solution had small unmodeled residual features ($\chi^2_\nu = 2.5$),
225: with a post-fit rms of 0.33\ ms and maximum of 0.9\ ms.  This precision
226: is comparable to the absolute accuracy of the \agile\ timing, and ensures
227: that the effective time resolution of the folded $\gamma$-ray light curve
228: is $\simlt 0.01\,P$.  The observed radio dispersion measure ($\mbox{DM}
229: = 369.3\pm0.3$\ pc\ cm$^{-3}$, determined with the addition of a few
230: observations centered at 1550\ MHz) was used to correct the times
231: of arrival of the radio pulse to infinite frequency (with resulting
232: $1\sigma$ uncertainty of 0.6\ ms) for absolute phase comparison with
233: the $\gamma$-ray pulse.
234: 
235: In choosing the sky region around the pulsar position for extraction
236: of $\gamma$-ray photons, consideration of the point-spread
237: function (PSF) of the instrument, which increases with decreasing
238: energy, must be weighed against the high diffuse $\gamma$-ray
239: background in this region, which also increases with decreasing
240: energy, as well as possible contamination from nearby
241: point sources.
242: We find that the
243: pulsed power is maximized if we use an extraction region that is
244: comparable to the $68\%$ containment radius, which is
245: $3.5^{\circ}$ at 100~MeV and $1.2^{\circ}$ at 400~MeV \citep{tav08}.
246: We used $r = 1.2^{\circ}$ for $E \ge 400$ MeV,
247: and $r = 1.2^{\circ}(400/E)^{1/2}$ for $E < 400$ MeV.
248: %The latter scaling is somewhat arbitrary, as the energy resolution
249: %is only $\Delta E/E \sim 1$ at these energies.  In addition, for
250: %the light curve we
251: In addition. 
252: we use only the ``G'' class events, which are identified with good
253: confidence as photons, and exclude the ``L'' class which suffer
254: much higher particle background contamination.  This selection
255: limits the effective area to $\sim 250$~cm$^2$ at 100~MeV.
256: 
257: \begin{figure}[t]
258: \centerline{
259: \includegraphics[height=1.15\linewidth,clip=true]{f1.eps}
260: }
261: \caption{Folded $\gamma$-ray light curves ({\it histograms})
262: of \psr\ in two energy bands, and their sum.
263: Two cycles are plotted, with 40 bins per cycle.
264: The solid line in the bottom panel is the mean radio pulse at
265: 1950 MHz from the GBT. Phase lags are P2--P1=$0.47\pm 0.01$
266: cycles and P1--R=$0.165\pm 0.010$ cycles.}
267: \label{fig:pulse}
268: \end{figure}
269: 
270: \section{Results and Interpretation}
271: 
272: \subsection{Pulsed Light Curve}
273: 
274: Figure~\ref{fig:pulse} shows the resulting folded light curve
275: of \psr\ in
276: two energy bands, $100-300$~MeV and $300-1500$~MeV.
277: The pulse shape consists of two sharp peaks,
278: labeled P1 and P2, with 
279: separation of $0.47 \pm 0.01$ cycles \cite[consistent with the separation
280: in the tentative detections reported in EGRET data by][]{mc04}.
281: This structure
282: is similar to that seen with much better signal-to-noise ratio in
283: other pulsars.  For comparison, peak separations are 0.38,
284: 0.43, and 0.49 for the Crab, Vela, and Geminga, respectively.
285: As in the other $\gamma$-ray pulsars, the
286: radio pulse of \psr\ leads the $\gamma$-ray P1 by a 
287: significant phase, in this case $0.165 \pm 0.010$ cycles.
288: In Vela, the radio leads by 0.13 cycles.
289: The Crab pulsar has multiple broad radio peaks that
290: continue to high frequency, and coincide with
291: higher-energy emission from infrared through X-ray.  However,
292: if we identify the narrow low-frequency component
293: \citep[LFC:][]{mof96,mof99} as the usual polar-cap core emission analogous
294: to other pulsars, the $\gamma$-ray--radio lag is 0.11 cycles.  Alternatively,
295: the Crab's even lower frequency ``precursor'' pulse, which leads
296: the $\gamma$-ray P1 by 0.056 cycles, may be the core component.
297: Even in the case of PSR~B1706$-$44, where \agile\ resolved
298: two peaks separated by 0.24 cycles from the dominant ``bridge''
299: emission, the radio pulse leads P1 by 0.21 cycles \citep{pel08}.
300: 
301: The outer-gap model of $\gamma$-ray pulse shapes of
302: \citet{rom95} predicts an inverse dependence of the
303: $\gamma$-ray peak separation on the $\gamma$-ray--radio
304: lag, with the principal controlling parameter being viewing
305: angle $\zeta$ from the rotation axis.
306: Large $\zeta$ results in large peak separation,
307: and small $\gamma$-ray--radio lag.
308: \psr\ is roughly in accord with this prediction
309: (Fig.~\ref{fig:delta}).  In this picture,
310: pulsars that have peak separations close to 0.5, such as
311: Vela, Geminga, and \psr, have $\zeta$ close to $90^{\circ}$.
312: Such an angle for \psr\ is also supported by the
313: modeling of its X-ray pulsar wind nebula (PWN) \pwna\
314: \citep{hes04,van08}, which can be understood as a nearly
315: edge-on torus,
316: indicating a pulsar rotation axis nearly perpendicular
317: ($86^{\circ}$) to the line of sight.  In contrast, PSR B1706$-$44
318: has narrow $\gamma$-ray peak separation of 0.24 cycles,
319: and $\zeta = 53^{\circ}$
320: from its X-ray torus model.  \citet{ng08} summarize
321: the X-ray measurements and their implications for $\gamma$-ray
322: pulse profiles.
323: 
324: \begin{figure}[t]
325: \centerline{
326: \includegraphics[height=0.85\linewidth,clip=true]{f2.eps}
327: }
328: \caption{Reproduction of Figure~3 of \citet{rom95} comparing
329: $\gamma$-ray peak
330: separation $\Delta$ with $\gamma$-ray--radio lag $\delta$
331: for \cgro\  pulsars ({\it filled circles}).  The $\delta$
332: of the Crab pulsar (0531) is uncertain due to the
333: ambiguous identification of its core radio pulse.
334: The location of \psr\ is marked with an asterisk,
335: as is the closely situated PSR~B1951+32 \citep{ram05}.
336: Model predictions ({\it solid lines}) for different magnetic axis
337: inclination angles $\alpha$ are similar, while viewing angle $\zeta$
338: increases along the curves from lower right to upper left.
339: }
340: \label{fig:delta}
341: \end{figure}
342: 
343: \begin{figure}[t]
344: \centerline{
345: \includegraphics[height=0.9\linewidth,clip=true]{f3.eps}
346: }
347: \caption{Images of the \agile\ field centered on \psr\ ({\it cross})
348: in the 100--1500~MeV range.  Pulse-phase resolved images
349: were created as described in \S 3.2, and \psr\ was
350: isolated via difference imaging.  North is up and east is
351: to the left. The diameter of the field displayed is
352: $20^{\circ}$, and the pixel size is $0.125^{\circ}$.
353: Gaussian smoothing of $\sigma = 1.5$ pixels has been applied.
354: Each image is linearly scaled to its highest point (white), while
355: the inset stretches the scale to show the level of fluctuations
356: in the off-pulse image in the vicinity of the pulsar.
357: }
358: \label{fig:image}
359: \end{figure}
360: 
361: \begin{figure}[b]
362: \centerline{
363: \includegraphics[height=0.82\linewidth,angle=270.,clip=true]{f4.eps}
364: }
365: \caption{Photon spectrum of \agl\ with power-law fit
366: of $\Gamma=1.86$ to the first three points.
367: }
368: \label{fig:spec}
369: \end{figure}
370: 
371: \subsection{Position, Flux, and Spectrum}
372: 
373: In order to separate the contribution of \psr\ from other possible
374: $\gamma$-ray sources within its PSF, we made
375: images corresponding to on-pulse and off-pulse
376: intervals of its light curve (Fig.~\ref{fig:image}).
377: The on-pulse image includes six phase bins on each of P1 and
378: P2, while the off-pulse
379: image comprises the remaining 28 bins.
380: A difference image was created by
381: subtracting the off-pulse image from the on-pulse
382: image, scaled by the number of phase bins.
383: As expected, differencing
384: removes all diffuse background and sources except for
385: \psr.  We measure the
386: position of the resulting source \agl\ as 
387: (J2000.0) R.A. = $20^{\rm h}20^{\rm m}30^{\rm s}$,
388: decl. = $+36^{\circ}53.7^{\prime}$,
389: which is only $0.12^{\circ}$ from the X-ray position,
390: well within the current \agile\ systematic uncertainty
391: of $0.25^{\circ}$ \citep{tav08}.
392: We performed a standard spectral analysis on the total source
393: photons including exposure correction and modelling of
394: the diffuse Galactic background.  The energy range
395: 100$-$1000~MeV is fitted with a power law of photon index
396: $\Gamma = 1.86\pm 0.18$ and a flux of $(46\pm6)\times 10^{-8}$
397: photons cm$^{-2}$ s$^{-1}$, while a turndown is seen above 1.5~GeV
398: (Fig.~\ref{fig:spec}).
399: 
400: Using the counting statistics of the off-pulse image,
401: we estimate the level of fluctuation that would correspond
402: to a 3$\sigma$ point source at the location of \psr,
403: noting that there is no source present at this level.
404: Scaling from the pulsar flux, this corresponds to
405: upper limits of $1.0 \times 10^{-7}$ photons cm$^{-2}$ s$^{-1}$
406: for the off-pulse flux from \psr, and
407: $1.7 \times 10^{-7}$ photons cm$^{-2}$ s$^{-1}$ for a steady
408: point source within $1.2^{\circ}$ of \psr, 
409: such as the PWN \pwnb\ (of unknown spin parameters) or the blazar \blazar,
410: both located in the error circle of \egrb\ \citep{muk00,hal01}.
411: Since blazars are episodic $\gamma$-ray emitters, it cannot be ruled
412: out that \blazar\ was more active during the EGRET era.
413: However, we are not confident that \egrb\ was a source distinct from 
414: \egra.  The former is not included in the revised EGRET catalog
415: of \citet{cas08}, which instead lists only one source,
416: EGR J2019+3722.  The positions of \egra\ and EGR J2019+3722
417: are both inconsistent with \psr\ at their 99\% confidence
418: contours, while \agile\ sees one dominant source that is clearly
419: attributed to \psr\ by phase-resolved imaging.
420: 
421: \subsection{Gamma-Ray Efficiency}
422: 
423: Previous authors have noted that the $\gamma$-ray luminosity
424: of \egra\ is a large fraction $\eta_{\gamma}$ of the spin-down
425: luminosity of \psr\ if placed at the nominal DM
426: distance of 12~kpc using the \citet{cor02} electron-density
427: model of the Galaxy.
428: \cite{rob02} and \cite{hes04} calculated, assuming $1/4\pi$ beaming,
429: that $\eta_{\gamma} \equiv Fd^2/\dot E \approx 0.18\,d_{12}^2$,
430: where $F$ is the flux in the 100~MeV--10~GeV range.  Such
431: high efficiency deviates from the trend of other $\gamma$-ray
432: pulsars, which follow an approximate
433: $\eta_{\gamma} \propto \dot E^{-1/2}$ relation \citep{tho04}.
434: Pulsars of similar $\dot E$ to \psr\ have
435: $\eta_{\gamma} \sim 10^{-3}-10^{-2}$.
436: This contrast casts some suspicion on the accuracy of the
437: DM distance, but in addition, consideration of the X-ray
438: properties of \psr, including the thermal spectrum of
439: the neutron star and the size of its PWN torus,
440: led \citet{van08} to favor a distance as small as $3-4$ kpc.
441: 
442: Our results do not bear directly on the distance question,
443: while the inferred $\gamma$-ray efficiency of \psr\ remains
444: exceptional.
445: The \agile\ flux above 100~MeV is
446: $\approx 2.4 \times 10^{-10}$ ergs cm$^{-2}$ s$^{-1}$
447: yielding $\eta_{\gamma} \approx 0.10\,d_{12}^2$,
448: somewhat smaller than previous estimates 
449: because of the cutoff above 1.5~GeV.
450: In the models of \citet{rom95},
451: the beaming of \psr\ may be more severe than
452: in other $\gamma$-ray pulsars if its nearly
453: edge-on X-ray torus implies a large
454: magnetic inclination angle $\alpha$.
455: 
456: The TeV source \mgro\ \citep{abd07a,abd07b} 
457: of diameter $1.1^{\circ}\pm 0.5^{\circ}$
458: centered at $(\ell,b)=(75.0^{\circ},0.2^{\circ})$
459: may be an extension of \pwna, similar to many other
460: TeV PWNe that are spatially offset from their
461: smaller X-ray counterparts.  The flux of \mgro,
462: $\approx 5.6 \times 10^{-12}$ ergs cm$^{-2}$ s$^{-1}$,
463: represents $\approx 0.03\,d_{12}^2\,\dot E$ of \psr,
464: similar to other TeV PWNe \citep{gal08}.
465: 
466: \section{Conclusions}
467: 
468: After the demise of \cgro, several candidate
469: pulsar identifications for EGRET sources could not be confirmed
470: until it became possible again to obtain contemporaneous $\gamma$-ray
471: and radio timing.  Using the extensive exposure of \agile\
472: on the Cygnus region,
473: we detect $\gamma$-rays in the 100--1500~MeV
474: range from the first post-EGRET pulsar, \psr=\agl.
475: Its pulse shape is similar to other
476: $\gamma$-ray pulsars, with two sharp peaks separated by
477: slightly less than 0.5 cycles. This
478: morphology, as well as the lag between the
479: radio and first $\gamma$-ray peak, are in accord with simulations of
480: outer-gap accelerators viewed at large angles from the pulsar rotation axis,
481: which may help to explain its high $\gamma$-ray efficiency.  \psr\ is
482: the only $\gamma$-ray point source detected in the region of 2CG~075+00.
483: Timing noise of young pulsars such as \psr\ and other candidate
484: $\gamma$-ray pulsars demands frequent radio observations to
485: be able to discover and resolve the finest structure
486: in $\gamma$-ray light curves that will be accumulated
487: over months to years.  Independent discovery of pulsed
488: $\gamma$-rays from \psr\ by the {\it Fermi Gamma-ray Space Telescope}
489: will also be reported shortly.
490: 
491: \acknowledgements
492: 
493: We thank ASI for making these data available under the
494: Guest Observer Program of the \agile\ mission.
495: The GBT is operated by the National Radio Astronomy Observatory,
496: a facility of the National Science Foundation operated under
497: cooperative agreement by Associated Universities, Inc.
498: The referee David Thompson suggested
499: several improvements to this Letter.
500: 
501: \begin{thebibliography}{}
502: 
503: \bibitem[Abdo et al.(2007a)]{abd07a}
504: Abdo, A. A., et al. 2007a, \apj, 658, L33
505: 
506: \bibitem[Abdo et al.(2007b)]{abd07b}
507: ---------. 2007b, \apj, 664, L91
508: 
509: \bibitem[Barbiellini et al.(2001)]{bar01}
510: Barbiellini, G., et al. 2001, in AIP Conf. Ser. 587, Gamma 2001:
511: Gamma-Ray Astrophysics 2001, ed. S. Ritz, N. Gehrels,
512: \& C. R. Shrader (Melville, NY: AIP), 754
513: 
514: \bibitem[Casandjian \& Grenier(2008)]{cas08}
515: Casandjian, J.-M., \& Grenier, I. A. 2008, \aap, in press (arXiv:0806.0113)
516: 
517: \bibitem[Cordes \& Lazio(2002)]{cor02}
518: Cordes, J. M. \& Lazio, T. J. W. 2002, preprint (astro-ph/0207156)
519: 
520: \bibitem[Gallant \etal(2008)]{gal08}
521: Gallant, Y. A., et al. 2008, AIP Conference Ser. 983,
522: 40 Years of Pulsars: Millisecond Pulsars, Magnetars, and More,
523: ed. C. Bassa, Z. Wang, A. Cumming, \& V. M. Kaspi
524: (Melville, NY: AIP), 195
525: 
526: \bibitem[Halpern et al.(2001)]{hal01}
527: Halpern, J. P., Eracleous, M., Mukherjee, R., \& Gotthelf, E. V. 2001,
528: \apj, 551, 1016 (Erratum: 2002, \apj, 572, 693)
529: 
530: \bibitem[Hartman et al.(1999)]{har99}
531: Hartman, R. C., et al. 1999, \apjs, 123, 79
532: 
533: \bibitem[Hessels et al.(2004)]{hes04}
534: Hessels, J. W. T., Roberts, M. S. E., Ransom, S. M., Kaspi, V. M.,
535: Romani, R. W., Ng, C.-Y., Freire, P. C., \& Gaensler, B. M. 2004,
536: \apj, 612, 389
537: 
538: \bibitem[Kaplan et al.(2005)]{kap05}
539: Kaplan, D., Escoffier, R. P., Lacasse, R. J., O'Neill, K.,
540: Ford, J. M., Ransom, S. M., Anderson, S. B., Cordes, J. M.,
541: Lazio, T. J. W., \& Kulkarni, S. R. 2005, \pasp, 117, 643
542: 
543: \bibitem[Labanti et al.(2006)]{lab06}
544: Labanti, C. M., et al. 2006, in SPIE Conf. Ser. 6266, Space Telescopes
545: and Instrumentation II: Ultraviolet to Gamma-Ray, ed. M. J. L. Turner
546: \& G. Hasinger, 62663Q
547: 
548: \bibitem[Lamb \& Macomb(1997)]{lam97}
549: Lamb, R. C., \& Macomb, D. J. 1997, \apj, 488, 872
550: 
551: \bibitem[McLaughlin \& Cordes(2004)]{mc04}
552: McLaughlin, M. A., \& Cordes, J. M. 2004, in 
553: X-Ray and Gamma-Ray Astrophysics of Galactic Sources, ed. M. Tavani,
554: A. Pellizzoni, \& S. Vercellone (Rome: IASF), 67
555: 
556: \bibitem[Moffett \& Hankins(1996)]{mof96}
557: Moffett, D. A., \& Hankins, T. H. 1996, \apj, 468, 779
558: 
559: \bibitem[Moffett \& Hankins(1999)]{mof99}
560: ---------. 1999, \apj, 522, 1046
561: 
562: \bibitem[Mukherjee et al.(2000)]{muk00}
563: Mukherjee, R., Gotthelf, E. V., Halpern, J. P., \& Tavani, M.
564: 2000, \apj, 542, 740
565: 
566: \bibitem[Ng \& Romani(2008)]{ng08}
567: Ng, C.-Y., \& Romani, R. W. 2008, \apj, 673, 411
568: 
569: \bibitem[Pellizzoni et al.(2008)]{pel08}
570: Pellizzoni, A., et al. 2008, \apj, in press
571: 
572: \bibitem[Prest et al.(2003)]{pre03}
573: Prest, M., Barbiellini, G., Bordignon, G., Fedel, G., Liello, F.,
574: Longo, F., Pontoni, C., \& Vallazza, E. 2003, Nucl. Inst. and Meth.
575: A, 501, 280
576: 
577: \bibitem[Ramanamurthy et al.(1995)]{ram05}
578: Ramanamurthy, P. V., et al. 1995, \apj, 447, L109
579: 
580: \bibitem[Roberts et al.(2002)]{rob02}
581: Roberts, M. S. E., Hessels, J. W. T., Ransom, S. M., Kaspi, V. M.,
582: Freire, P. C., Crawford, F., \& Lorimer, D. R. 2002, \apj, 577, L19
583: 
584: \bibitem[Roberts et al.(2001)]{rob01}
585: Roberts, M. S. E., Romani, R. W., \& Kawai, N. 2001,
586: \apjs, 133, 451
587: 
588: \bibitem[Romani \& Yadigaroglu(1995)]{rom95}
589: Romani, R. W., \& Yadigaroglu, I.-A. 1995, \apj, 438, 314
590: 
591: \bibitem[Swanenburg et al.(1981)]{swa81}
592: Swanenburg, B. N., et al. 1981, \apj, 243, L69
593: 
594: \bibitem[Tavani et al.(2008)]{tav08}
595: Tavani, M., et al. 2008, \aap, submitted (arXiv:0807.4254)
596: 
597: \bibitem[Thompson(2004)]{tho04}
598: Thompson, D. J. 2004, in Cosmic Gamma Ray Sources, ed. K. S. Cheng
599: \& G. E. Romero (Dordrecht: Kluwer), 149
600: 
601: \bibitem[van Etten et al.(2008)]{van08}
602: van Etten, A., Romani, R. W., \& Ng, C.-Y. 2008, \apj, 680, 1417
603: 
604: \end{thebibliography}
605: 
606: \end{document}