1: %\documentclass[12pt,preprint]{aastex}

2: 

3: \documentclass{emulateapj}

4: %\documentclass[12pt,preprint]{aastex}

5: %\usepackage{./apjfonts}

6: %%

7: %% 0737Pol - final(?) version - 04jan31

8: 

9: %\usepackage[]{natbib}

10: %\documentstyle[12pt,aasms4]{article}

11: %\documentstyle[11pt,aaspp4]{article}

12: 

13: \usepackage{epsfig}

14: 

15: % Here's some slug-line data.  The receipt and acceptance dates will be 

16: % filled in by the editorial staff with the appropriate dates.  Rules will 

17: % appear on the title page of the manuscript until these are uncommented 

18: % out by the editorial staff.

19: \newcommand\etal{{\it et al}}

20: \def\etal{{\it et al}}

21: \def\deg{^{\circ}}

22: \def\P3hat{{\mathaccent 94 P}_3}

23: \def\parhang{\noindent\hangindent=0.4 true in \hangafter=1}

24: \def\eg{{\it e.g.,}}

25: \def\ie{{\it i.e.}}

26: \def\cf{{\it cf.}}

27: \def\A{{\bf A}~}

28: \def\B{{\bf B}~}

29: \def\An{{\bf A}}

30: \def\Bn{{\bf B}}

31: \def\pA{pulsar {\bf A}~}

32: \def\pB{pulsar {\bf B}~}

33: \def\pAn{pulsar {\bf A}}

34: \def\pBn{pulsar {\bf B}}

35: \def\snr{{\tt S/N~}}

36: \def\phorb{{\phi_{\rm orb}}}

37: \def\xa{{X_{\rm A}(t,\nu)}}

38: \def\xb{{X_{\rm B}(t,\nu)}}

39: \def\pa{{\tt PA~}}

40: \def\pas{{\tt PA}s}

41: %\received{7 March 1999}

42: %\accepted{23 September 1988}

43: %\journalid{337}{15 January 1989}

44: %\articleid{11}{14}

45: 

46: %\slugcomment{Not to appear in Nonlearned J., 45.}

47: 

48: % The right head is a modified title of up to roughly 44 characters.  Running heads

49: % are not printed.

50: 

51: \shorttitle{Morphological properties of J0737--3039 pulsars}

52: 

53: \shortauthors{R. Ramachandran et al.}

54: 

55: \begin{document}

56: 

57: \title{Fluctuation and morphological properties of the pulsars in J0737--3039 system}

58: 

59: \author{R. Ramachandran, D. C. Backer, P. Demorest} \affil{Department

60: of Astronomy, University of California, Berkeley, CA 94720-3411, USA}

61: 

62: \author{S. M. Ransom$^{1}$, V. M. Kaspi$^{1,2}$}

63: 

64: \affil{Department of Physics, McGill

65: University, Montreal, QC H3A 2T8, Canada} \altaffiltext{1}{Center for

66: Space Research, Massachusetts Institute of Technology, Cambridge, MA

67: 02139} \altaffiltext{2}{Canada Research Chair, Steacie Fellow, CIAR

68: Fellow}

69: 

70: \begin{abstract}

71: We describe the morphological and fluctuation properties of the

72: pulsars in the double neutron star system, PSR J0737--3039. Pulsar \B

73: is seen in almost all orbital phases, except in the range of $\sim

74: 6\deg$ to $65\deg$. This may be interpreted as an {\it eclipse} of

75: \pBn's signal by its own magnetopause region produced by interaction

76: with \pAn's relativistic wind. No modulation of the emission of \pB is

77: found at the period of \pAn. This places a constraint on the models

78: that propose that \pAn's beamed radiation is directly responsible for

79: \pBn's emission. Modulation index values indicate that the pulse to

80: pulse variations in the two objects are mostly intrinsic. Pulsar A

81: shows significant differential modulation index within its pulse

82: profile.

83: \end{abstract}

84: 

85: \keywords{pulsars: general -- pulsars: individual (J0737-3039) --

86: radiation mechanisms: non-thermal}

87: 

88: \section{Introduction}

89: \label{sec-intro}

90: Recent discovery of the double pulsar system, reported by Burgay et

91: al. (2003) and Lyne et al. (2004) is one of the most important

92: discoveries in the history of pulsar astronomy. This high inclination system,

93: which consists of a short period recycled pulsar and an ordinary slow

94: pulsar revolving with an orbital period 2.45 hrs,

95: gives us almost an ideal situation to probe many fundamental properties

96: of neutron star astrophysics.

97: 

98: The distance between the two pulsars is so small (2.8 light sec), that the

99: light cylinder volume (magnetosphere of a solitary pulsar) of PSR

100: J0737--3039B (hereafter \Bn) subtends about 20 degrees from PSR

101: J0737--3039A (hereafter \An).

102: 

103: It is unclear whether the measured value of $\dot{P}$ of \pB is

104: intrinsic to the pulsar, given the mechanical luminosity ($\dot{E}$)

105: of \pAn. It is conceivable that the relativistic wind from \pA

106: influences the characteristics of \pBn.  In spite of this, \pB

107: continues to emit radio radiation, clearly demonstrating the

108: robustness of the fundamental radio emission mechanism.  According to

109: Lyne et al. (2004), this system may well have indicated that the seed

110: acceleration responsible for this emission comes from deep inside the

111: magnetosphere, clearly ruling out outer magnetospheric emission

112: models.

113: 

114: The measured flux of \pB significantly changes as a

115: function of orbital phase\footnote{Orbital phase ($\phorb$) is given

116: as $(\omega+\eta)$, where $\omega$ and $\eta$ are the position angle

117: of periastron and the ``true'' anomaly.} (Lyne et al. 2004). 

118: 

119: Models have been proposed to suggest that the wind from \pA creates a

120: magnetopause region around \pB due to the interaction with \pBn's

121: magnetosphere (Arons et al. 2004). In this model, the synchrotron

122: absorption optical

123: depth of this

124: magnetopause region is the primary reason for the observed eclipse of

125: \pAn, and possibly for the light curve of \pBn.

126: 

127: In this work, we present some of the basic morphological and

128: fluctuation properties of these two pulsars. Section \ref{sec-flux}

129: describes the properties of average pulses of \pA and \pBn. The

130: fluctuation properties of the two pulsars are described in Section

131: \ref{sec-fluct}. Implications of our results are discussed in Section

132: \ref{sec-discuss}.

133: 

134: \section{Observations}

135: \label{sec-obsvn}

136: Observations were performed at the Green Bank Telescope (GBT) on 2003

137: December 11, 19, 23, \& 24, and 2004 January 1. With the BCPM

138: (Berkeley-Caltech Pulsar Machine; see e.g., Camilo et al. 2002), we

139: observed at 430, 800, 1400 \& 2200 MHz frequencies to record total

140: power (Stokes-I parameter) at a time resolution of 72$\mu$sec. The

141: bandwidths for the first two bands were 48 MHz (96 channels of 0.5 MHz

142: each), and the last two were 96 MHz (96 channels of 1 MHz each).

143: At 430 and 800 MHz, we also used the Spigot card with a

144: bandwidth of 50 MHz.

145: 

146: The data was corrected for interstellar dispersion as part of our

147: offline processing. We then added all channels together to produce a

148: single time series. We used the ephemeris given by Burgay et

149: al. (2003) for analysing the data from December 11th, 2003, and

150: produced an improved ephemeris for \pAn. This was then used for

151: generating an ephemeris for \pB for all our subsequent observations.

152: 

153: \section{Profile characteristics and light curves}

154: \label{sec-flux}

155: 

156: The two pulsars in this binary system show entirely different

157: characteristics in their average properties. Apart from an eclipse of

158: $\sim$30 sec duration, \pA shows no significant variation in its

159: light curve\footnote{Observed flux as a function of time}, whereas \pB

160: exhibits a spectacular orbital phase dependent variation in its light

161: curve. Pulsar \A shows significant average profile variation as a

162: function of radio frequency which is seen in many other MSPs. Pulsar

163: \B shows significant profile variation not only as a function of

164: frequency, but also as a function of orbital phase. We summarise all

165: these properties in this section.

166: 

167: \subsection{Average profiles of \pAn}

168: \label{sec-profa}

169: Pulsar \A does not show any significant flux variation as a function

170: of orbital phase outside of the eclipse region at its superior

171: conjunction, whose characteristics are described by Lyne et al. (2004)

172: and Kaspi et al. (2004). However, \pA shows considerable change in its

173: average profiles as a function of radio frequency. We observed this

174: pulsar at four different frequency bands, namely 427, 820, 1400 \&

175: 2200 MHz. Average profiles constructed from these data sets are given

176: in Figure \ref{fig:profA}.

177: 

178: As Burgay et al. (2003) \& Lyne et al. (2004) reported, we have a

179: two-featured profile. Although these features appear like the classic

180: main pulse and interpulse of an orthogonal rotator, the most favored

181: configuration is likely an almost aligned rotation and magnetic axes,

182: where our sight line cuts the emission cone in two different sides of

183: the cone (Demorest et al. 2004). At the lowest frequency, the first

184: feature is dominant, with less pronounced structures. However, at all

185: the other frequencies, the second feature is dominant. Also, the first

186: feature develops complex component structures. Although this is quite

187: strikingly different from the behavior of ordinary long period

188: pulsars, it is by no means uncommon among the shorter period pulsar

189: population.

190: 

191: \begin{figure}

192: \begin{center}

193: \epsfig{file=fig1.ps,height=10cm}

194: \caption{Average profiles of \pA at (top to bottom) 427, 820, 1400 \&

195: 2200 MHz. The profile at 427 MHz is from the SPIGOT card, and the rest

196: are from BCPM. The effective time resolution for each of the profiles

197: is indicated as a horizontal error bar. See text for details.}

198: \label{fig:profA}

199: \end{center}

200: \end{figure}

201: 

202: The amplitude reversal of components as a function of frequency has

203: been noted among several short period pulsars (millisecond pulsars).

204: For instance, PSRs J0751+1807 \& J2145--0750 are known to exhibit such

205: reversal (Kramer et al. 1999). PSR J1022--1001 exhibits even a more

206: complex evolution, where the component strengths reverse twice, while

207: going from 400 MHz to a few GHz (Ramachandran \& Kramer 2003). On the

208: other hand, there are also counter examples like J0621+1002 \&

209: J1744--1134, where the pulsar exhibits little change in the profile

210: structure.

211: 

212: We estimated the width and separation of the two features at all the

213: frequency bands, and they do not seem to exhibit significant change.

214: As it turns out, average profiles in all four bands could be fitted

215: with seven gaussian components (4 for first feature, and 3 for the

216: second). In order to quantify the profile span in longitude

217: completely, we measured the width between the ``base'' (corresponds to

218: approximately 1/50th of the peak value) of the raising and the

219: trailing edge. This turned out to be 66$\deg$ and 60$\deg$ for the two

220: features in all frequencies, and these values coincided within

221: 2-3$\deg$. Similarly, the separation of the outer edges of the two

222: features was $\sim$240$\deg$, and the separation of inner edges was

223: $\sim$114$\deg$.

224: 

225: Several investigators have studied the variation of pulse width and

226: component separation as a function of radio frequency. The classical

227: `conal-single' pulsars are known to exhibit a power law behavior,

228: where the width of the pulse depends on radio frequency as $\nu^{-b}$,

229: where $b$ is in the range of $\sim 0.1-0.3$ (Rankin 1983). Hankins \&

230: Fowler (1986) studied the frequency dependence of separation between

231: the main and the interpulse of eight pulsars with interpulses. They

232: find that although the main and the interpulse show considerable

233: frequency dependence in their shape, the separation between them is

234: frequency independent. Although this lack of frequency dependence

235: tends to favor magnetic and rotation axes being almost orthogonal to

236: each other (observed emission coming from both poles, with main pulse

237: from one pole and the interpulse from the other pole: {\it two-pole

238: model}), with their polarization observations, they point out that

239: they favor the geometry where the two axes are almost aligned ({\it

240: single-pole model}).

241: 

242: However, for recycled pulsars, the observed frequency dependence is

243: more complex. The separation between the two features of PSR B1259--63

244: changes from $\sim$160$\deg$ to 139$\deg$ while going from 660 MHz to

245: 1.5 GHz. Then, at 4.7 GHz, the separation is 107$\deg$, and it stays

246: constant above this frequency (Manchester \& Johnston 1995). As

247: Manchester \& Johnston state, the most favored model for this pulsar

248: is the single-pole model.

249: 

250: In the case of PSR J1022+1001, although the strength of the components

251: evolve significantly as a function of radio frequency, and the profile

252: itself shows significant instability over time scales of several

253: minutes, the location of components and the width of the profile

254: remains almost constant over a wide range of 328 MHz to about 4.8 GHz

255: (Ramachandran \& Kramer 2003). Kramer et al. (1999) present a very

256: good summary of recycled pulsar profiles and their frequency

257: evolution. In fact, their fig. 16 demonstrates clearly that the

258: observed $b$ values for recycled pulsars are very close to zero, with

259: a few exceptions.

260: 

261: Therefore, one can conclude from the behavior of average pulses of \pA

262: as a function of radio frequency that it is no different from the

263: general behavior of short period pulsar population. The lack of

264: frequency dependence of either the pulse width or the main and the

265: interpulse separation cannot be considered as a strong evidence for

266: the two-pole model.

267: 

268: \subsection{Light curves of \pBn}

269: \label{sec-lightcurve}

270: As Lyne et al. (2004) indicate, \pBn's flux varies systematically as a

271: function of orbital phase. In some phases, the pulsar is so bright

272: that single pulses are detectable. However, in most of the orbital

273: phase range, the emission is faint. Figure \ref{fig:lc} shows that the

274: pulsar is prominent in four orbital phase ranges (windows I to

275: IV). For clarity, we have given the relative flux in logarithmic units

276: and have repeated the light curve for two orbital periods. The

277: inferior conjunction of \pB (when \B is in front of \An) occurs when

278: $\phorb = 270\deg$. 

279: 

280: There are several features here that are noteworthy. Each sample in

281: Figure \ref{fig:lc} is produced by finding the area of the pulse after

282: averaging some number of pulses. Owing to \snr considerations, we have

283: taken unequal phase intervals for defining each sample in the

284: plot. The shortest averaging length that we have taken is when the

285: pulsar is very bright ($\phorb\sim180\deg$) where we have taken an

286: average of 20 pulses. The longest length is when the pulsar flux was

287: the weakest ($\phorb\sim 400\deg$), where we have considered 700

288: pulses to find an upper limit to the average pulse flux.

289: 

290: It is worth emphasizing one important aspect of this light curve. In

291: the orbital phase range of $\sim6\deg$ to 65$\deg$ (modulo 360$\deg$

292: in Figure \ref{fig:lc}), it appears that the pulsar signal exhibits

293: what may be considered as an ``eclipse''. We did not detect

294: significant flux in this range at all. Although we cannot define the

295: boundary of this ``eclipse'' region clearly due to signal to noise

296: ratio considerations, the boundary appears to be very sharp. The

297: physical mechanisms responsible for this behavior will be discussed in

298: detail elsewhere (Arons et al. 2004).

299: 

300: Pulsar profiles reach a stable state only after averaging some

301: thousand pulses (Helfand et al. 1975). Therefore, it is likely that

302: our flux estimation in the phase ranges where we have taken very short

303: averaging are somewhat in error. However, our emphasis here is on the

304: overall systematic relative variations, and not on the absolute flux

305: itself. As we will see in the next section, this behavior is not

306: restricted to only the light curve, but also to the detailed profile

307: structure.

308: 

309: There is some hint of chromaticity in the light curve during

310: windows III and IV, where we observed lower flux at 1400 MHz than at

311: 800 MHz. However, when the pulsar is bright, the light curve seems to

312: be achromatic. A possible reason for the observed chromaticity in

313: windows III and IV may be due to the line of sight passing close to

314: the edge of the emission cone (impact angle $\beta$ approximately

315: equal to emission cone radius $\rho$), where the effective emission

316: beam size is smaller at higher frequencies due to radius to

317: frequency mapping.

318: 

319: \begin{figure*}[t]

320: \begin{center}

321: \epsfig{file=fig2.ps,height=15cm,angle=-90}

322: \caption[]{Light curve of \pB at 820 MHz (filled circles) and 1400 MHz

323: (open circles). The four bright windows are indicated as I, II, III \&

324: IV. The dotted line has been drawn just for clarity. See text for

325: details.}

326: \label{fig:lc}

327: \end{center}

328: \end{figure*}

329: 

330: \subsection{Average profiles of \pBn}

331: In contrast to \pAn, \pB shows significant change in its pulse profile

332: structure as a function of orbital phase. For instance, the average

333: profiles seen in window I are very different from those in window

334: II. To demonstrate this effect, we have presented the average profiles

335: derived at three of our bands, 800, 1400 and 2200 MHz, in Figure

336: \ref{fig:profB}. The orbital phase range corresponding to the four

337: panels are given in the figure caption. The width of window I

338: corresponds to approximately 830 seconds ($\sim$35$^{\circ}$ of

339: orbital phase -- panels $Ia$ to $Ic$). During this interval, the

340: profile evolves significantly from having a slow rising edge (sharp

341: trailing edge) to sharp rising edge (slow trailing edge). This is

342: consistently observed in all the three frequencies. Although there are

343: some differences in the average profiles in a given orbital phase

344: range between one orbit to the next, these differences can be

345: attributed to insufficient averaging in producing the average pulse.

346: 

347: In the fourth panel which corresponds to window II the profile

348: displays two components, and these two components seem to be well

349: separated at higher frequencies.

350: 

351: \begin{figure*}

352: \begin{center}

353: \epsfig{file=fig3.ps,height=15cm,angle=-90}

354: \caption{Average profiles of \pBn. The four panels (Ia, Ib, Ic and II)

355: correspond to orbital phase ranges of 171$\deg$-183$\deg$,

356: 183$\deg$-195$\deg$, 195$\deg$-207$\deg$ \& 240$\deg$-273$\deg$,

357: respectively. The three profiles in each panels are at 820 (top), 1400

358: (middle) and 2200 MHz (bottom), respectively. Pulse longitude given

359: in the x-axis has arbitrary reference. See text for details.}

360: \label{fig:profB}

361: \end{center}

362: \end{figure*}

363: 

364: %Given this complex behavior, it is impossible to understand the

365: %geometric orientation of \pB with respect to the line of sight. 

366: %The

367: %measured rotation measures (RM) towards \pA and \pB are $-112.3 \pm

368: %1.5$ rad m$^{-2}$ and $-118\pm 12$ rad m$^{-2}$, respectively.

369: 

370: Among classical conal single pulsars, it is common to see profiles

371: evolving into double profiles at lower frequencies. This is generally

372: interpreted as due to radius-to-frequency mapping. Also, this model

373: would certainly predict increasing width of the profiles with

374: decreasing frequency. However, in the case of panel $(d)$ (window II),

375: the two components seem to be more pronounced with increasing

376: frequency, which is the opposite of what we expect from

377: radius-to-frequency mapping.

378: 

379: Having said this, it is also important to appreciate that in the case

380: of the classical pulsar population, pulsars are solitary, and there is

381: no influence on their magnetospheres from the surroundings. However,

382: as Lyne et al. (2004) state, mechanical luminosity of \pA is

383: significant enough to influence the magnetospheric properties of

384: \pBn. In fact, \pBn's proximity to \pA ensures that \pBn's

385: magnetosphere is altered by \pAn's relativistic plasma wind.

386: 

387: 

388: There are three effects that can alter the emission amplitude and

389: morphology of \pB (Demorest et al.  2004; Arons et al. 2004):

390: \begin{enumerate}

391: \item Orbital phase dependence of \pA wind pressure on \pBn's

392: magnetosphere owing to the orientation of \pAn's magnetic and rotation

393: axes.

394: \item Orbital phase dependence of the response of pulsar \pBn's

395: magnetosphere to the \pA wind pressure, owing to the orientation and

396: spin phase of \pB as we observe it.

397: \item Variable expansion of the \pB magnetosphere (at the spin phase

398: of \pB when we observe \pB) into the downstream cavity created by the

399: impinging \pA wind.

400: \end{enumerate}

401: In all cases the variable magnetosphere shape will likley translate to

402: variable polar cap size that, in turn, defines the active region of

403: beamed radio emission which we parameterize with impact angle

404: ($\beta$) and emission cone radius ($\rho$). While the effects in

405: cases 1 and 2 are likely to produce rather small variations owing to

406: the likely $P^{1/12}$ dependence of polar cap size on pressure, we

407: expect that case 3 could lead to substantial variations.

408: 

409: \section{Fluctuation parameters}

410: \label{sec-fluct}

411: It is well known that individual pulses from pulsars fluctuate in

412: their intensity, location and shape, although the average pulse

413: profiles of almost all pulsars show a great deal of stability. These

414: fluctuations exhibit a spectacular variety in their spectral

415: characteristics. The intensity variations at a given pulse longitude

416: are usually noiselike, although in many cases, there are

417: quasi-periodic features. These quasi-periodic variations are usually

418: associated with one of the most remarkable properties seen in pulsar

419: signals -- drifting subpulses. In this section, we will address

420: various aspects of single pulse fluctuations seen in signals from \pA

421: and \pBn.

422: 

423: \subsection{Modulation Index}

424: \label{sec-modind}

425: One of the ways to quantify the amount of fluctuations seen in a given

426: signal is by measuring modulation index, which is measured as

427: \[

428: m = \frac{\sqrt{\sigma_{\phi}^2 -

429: \sigma^2_{ \circ}}}{\mu_{\phi}}, 

430: \]

431: \noindent

432: where $\sigma_{\phi}$ and $\sigma_{\circ}$ are the variance seen in a

433: given pulse longitude ($\phi$) and the offpulse region, respectively;

434: $\mu_{\phi}$ is the mean flux measured at that longitude. With our 820

435: MHz data, for our analysis on \pAn, we considered one hundred thousand

436: pulses. In Figure \ref{fig:modind}, we give the measured modulation

437: index as a function of pulse longitude. The left panel corresponds to

438: \pAn, and the middle and the right panels correspond to windows I and

439: II of \pBn. The measured value of modulation index comes to almost

440: hundred percent across the pulse profile. That is, the

441: r.m.s. variation in the flux level is as high as the mean level

442: itself.

443: 

444: There are a few investigations reported in the literature that have

445: measured intrinsic modulation index of pulsars. For instance, Weisberg

446: et al. (1986) have measured modulation index values of many classical

447: pulsars. They conclude that core components, on the average, have

448: lower modulation index than the conal components. While the reason for

449: this is not completely clear, perhaps it is due to a combination of

450: intrinsic instability of conal component emissions of pulsars, and

451: pulse nulling. As Rankin (1986) points out, pulse nulling has {\it

452: never} been seen before in pulsars with only core emission. There is

453: also a study of two MSPs by Jenet and his collaborators (Jenet \& Gil

454: 2004; Jenet et al. 2001). They conclude that PSR B1937+21 has very low

455: -- if not negligible -- intrinsic modulation index, whereas PSR

456: J0437--4715 exhibits a high modulation index. In general, it is

457: unclear what the fluctuation properties of MSPs are when compared to

458: ordinary pulsars, as no comprehensive study exists to date.

459: 

460: Interstellar scintillation is one well known source of extrinsic

461: modulation that can contribute to the measured value of $m$. The

462: expected value of $m$ from the interstellar scintillation is given by

463: $m_{\rm ism} = 1/\sqrt{N_{\rm scint}}$, where $N_{\rm scint}$ is the

464: number of scintles in the band (Jenet et al. 2001). The observed

465: modulation index is then given by $\sqrt{m^2+m^2_{\rm ism}}$. The

466: value of $N_{\rm scint}$ is estimated as $N_{\rm scint}\, = \,(1 \, +

467: \, \eta B/\Delta\nu)$, where $B$ is the bandwidth of observation,

468: $\Delta\nu$ is the decorrelation bandwidth, and $\eta$ is the packing

469: fraction. The measured decorrelation bandwidth of \pA at 800 MHz is

470: 0.1$\pm$0.2 MHz, and at 1400 MHz is 1.8$\pm$0.6 MHz (Ransom et

471: al. 2004). Given the observed packing fraction of $\sim$0.4, the

472: estimated modulation index due to interstellar scattering is about 7\%

473: at 800 MHz. In Figure \ref{fig:modind}, the inner regions of both

474: features in the \pA profile seems to show significantly higher

475: modulation index than the outer regions. Such variations in modulation

476: index as a function of pulse longitude clearly suggests that there is

477: certainly some intrinsic variation, and not all of it comes from the

478: interstellar scattering. The high values of modulation index in the

479: inner regions of the two features also possibly indicate that they are

480: intermittent.

481: 

482: In windows I and II of \pBn, although modulation index does not vary

483: much as a function of pulse longitude, it seems to be much higher than

484: what is expected from interstellar scattering. Therefore, we conclude

485: that both the pulsars have significant intrinsic variability, whatever

486: the spectral characteristics may be.

487: 

488: \begin{figure*}

489: \begin{center}

490: \epsfig{file=fig4a.ps,height=7cm}

491: \epsfig{file=fig4b.ps,height=7cm}

492: \epsfig{file=fig4c.ps,height=7cm}

493: \caption{Modulation index ($m$) for \pA \& \pBn. ($a$) Upper panel

494: gives average profile produced with five hundred thousand pulses of

495: \pAn. Modulation parameter is plotted as a function of pulse longitude

496: in the lower panel. ($b$) \& ($c$) give the same information for \pB

497: in window I \& II, respectively. Error bars indicate $1-\sigma$

498: errors. The longitude resolution in all the plots have been chosen to

499: optimize signal to noise ratio. See text for details.}

500: \label{fig:modind}

501: \end{center}

502: \end{figure*}

503: 

504: \subsection{Fluctuation properties}

505: Having established in \S\ref{sec-modind} that \pA and \pB show

506: significant intrinsic pulse fluctuation, we will assess the spectral

507: nature of these fluctuations. For instance, these fluctuations could

508: be almost fully associated with possible drifting subpulses as seen in

509: many long period pulsars. One of the ways to quantify these

510: fluctuations is by computing the {\it longitude-resolved} (LR; Backer,

511: Rankin \& Campbell 1975) and {\it harmonic-resolved} (HR; Deshpande \&

512: Rankin 1999) spectra.  In general pulsars show wide ranging spectral

513: properties of pulse-to-pulse fluctuations, from very high coherence

514: (like in the case of drifting subpulses in PSR B0943+10 \& B0809+74

515: (Backer, Rankin \& Campbell 1975; Deshpande \& Rankin 2001 and

516: references therein), or broad quasi-coherent features as in the case

517: of PSR B2016+28. These (quasi-)coherent features not uncommon among

518: pulsar population. All `conal-single' pulsars are known to exhibit

519: drifting subpulse phenomenon (Rankin 1986). Interestingly, highly

520: coherent fluctuation features are seen in some components of

521: `multiple-component' pulsars like PSR B1237+25 (Backer, Rankin \&

522: Campbell 1975; Rankin \& Ramachandran 2003). These coherent features

523: are thought to represent some highly ordered kinematics in pulsar

524: magnetosphere.

525: 

526: In the case of \pB, there is also another reason for looking for

527: possible coherent intermodulations. Jenet \& Ransom (2004) have

528: recently proposed a scenario to explain the light curve of \pBn. In

529: their model, the near alignment of the magnetic and the rotation axes

530: of \A as suggested by Demorest et al. (2004) would mean that the

531: emission cone of \A is directed towards \B in two phase ranges in the

532: orbit. Then, if one assumes that \pAn's high energy emission is beamed

533: the same way as the radio emission, then it is conceivable that the

534: $\gamma$-ray emission from \pA influences \pBn. They attribute the

535: bright patches (windows I to IV) of \pBn's light curve to this direct

536: influence from \pAn.

537: 

538: If \pAn's beamed emission is directly responsible for \pBn's emission,

539: then it is important to look for possible intermodulation between the

540: two signals. That is, modulation of \pBn's signal with \pAn's

541: periodicity. Since the rotation periods of these two pulsars are

542: incommensurate with each other, if this modulation exists, it will

543: manifest itself as {\it drifting subpulses} in \pBn's signal. Of

544: course, this assumes that the environment of \pB does not influence

545: the beamed radiation from \pA to reduce the effect of

546: modulation. However, finding such a modulation will be a direct proof

547: for such an interaction between the two pulsars.

548: 

549: The computed L-R and H-R spectra at 820 MHz and 1400 MHz failed to

550: show any significant features in windows I \& II, where we had

551: adequate \snr. Owing to only a limited length of time in each

552: window, we had only 256 pulses ($\sim$12 min). The corresponding

553: spectral resolution in the LR spectrum is $(1/256P_1)$, and the

554: spectral width is $(1/2P_1)$, where $P_1$ is the rotation period.

555: our sensitivity to detect any fluctuation feature was limited to

556: one part in 540 in window I, and one part in 220 in window II.

557: 

558: We also looked for possible features in the LR and HR spectra of

559: \pAn. Although the possibility of having any influence of \pBn's

560: radiation on \pA is low, \pA may have intrinsic spectral nature of

561: its fluctuation properties that show (quasi-)coherent features in

562: the LR and HR spectra. However, our analysis failed to reveal

563: any such features within our sensitivity limit. Our sensitivity was

564: one part in $3\times 10^4$, where we considered 500 thousand pulses

565: for our analysis.

566: 

567: We conclude from these analyses that within our sensitivity limits, we

568: do not detect any systematic intermodulation of the signals from \pA

569: and \pBn, and the intrinsic contribution to modulation index comes

570: from fluctuations of broad spectral nature.

571: 

572: \section{Conclusions}

573: \label{sec-discuss}

574: The measured flux from \pB shows systematic variation as a function of

575: orbital phase. The pulsar is visible is all phase ranges except in

576: the range of $\phi_{\rm orb}\sim 6\deg$ to $65\deg$, where we have

577: obtained only an upper limit to the flux. This void region may be

578: interpreted as an {\it eclipse} of \pB by its own magnetopause region

579: which is produced by interaction with the relativistic wind from \pAn.

580: It is this magnetopause region that may also be responsible for the

581: eclipse of \pA around its superior conjunction (Arons et al. 2004).

582: Although it is difficult for us to produce reliable light curve around

583: this \pBn's apparent {\it eclipse} owing to signal-to-noise considerations,

584: it is clear that the drop in flux is very rapid (within several pulses!)

585: 

586: With our observations at 800 and 1400 MHz, we see some indication of

587: frequency dependence in \pBn's light curve. During windows III and

588: IV, the flux at 1400 MHz seems to be systematically lower than the

589: value at 800 MHz. This needs to be validated with further high

590: signal-to-noise observations.

591: 

592: Modulation index measurements indicate variation of single pulse flux

593: in both these objects. These variations are significantly more than

594: what is expected from the measured decorrelation bandwidth (due to

595: interstellar scattering) of this system. Therefore, these variations

596: are intrinsic to the two pulsars. The inner regions in \pA seems to

597: have significantly higher modulation index than the brighter outer

598: regions.

599: 

600: Pulsar \A exhibits a complex evolution of its average profile as a

601: function of radio frequency. However, these properties do not seem to

602: be different from the characteristics that short period pulsar

603: population seems to show.

604: 

605: Pulsar \Bn's profile shows dramatic variations as a function of

606: orbital phase, making it impossible to solve for its geometric

607: orientation without sensitive polarization measurements. The profile

608: evolution as a function of frequency seems to reveal mainly a double

609: component profile, where the components become more pronounced at

610: higher frequencies. This is quite the opposite of what one would

611: expect from simple dipole geometry, combined with radius to frequency

612: mapping.

613: 

614: Given the possible geometrical orientation of \pA derived on the basis

615: of polarimetric observations, it appears that there will be an orbital

616: phase dependent wind pressure on \B (Demorest et al. 2004). It is also

617: plausible that the diameter of \pBn's emission cone changes due to

618: the effect of orbital phase dependent pressure, there by changing the

619: fractional value of impact angle ($\beta$) with respect to the radius

620: of the emission cone ($\rho$). This can introduce complex profile

621: changes as a function of orbital phase, as we see for \pBn. It is very

622: important to understand the geometrical orientation of \pBn, for which

623: sensitive polarimetric observations are necessary.

624: 

625: Within our sensitivity limits, there is no significant intermodulation

626: between signals from the two objects. We do not detect any significant

627: feature in the L-R \& H-R spectra, indicating that the spectral

628: characteristics of the intrinsic pulse fluctuations is

629: ``noiselike''. This constrains the models proposing direct influence

630: on the emission mechanism of \pB by radiation from \pAn.

631: 

632: \acknowledgements 

633: 

634: We thank the GBT staff, and in particular Carl Bignell, Frank Ghigo,

635: Glen Langston and Karen O'Neil, for extensive help with the

636: observations and very useful discussions. V.M.K. is supported from

637: NSERC Discovery Grant 228738-03, NSERC Steacie Supplement 268264-03, a

638: Canada Foundation for Innovation New Opportunities Grant, FQRNT Team

639: and Centre Grants, and the Canadian Institute for Advanced

640: Research. V.M.K. is a Canada Research Chair and Steacie Fellow.  RR

641: would like to thank A. A. Deshpande for valuable discussions.

642: 

643: \begin{thebibliography}{}

644: \bibitem[]{} Arons, J.A., Spitkovsky, A., Backer, D. C., Kaspi,

645: V. M. 2004, ApJ, in preparation

646: \bibitem[]{} Backer, D. C., Rankin, J. M., Campbell, D. B. 1975, ApJ,

647: 197, 481

648: \bibitem[]{} Backer, D.C., Wong, T. \& Valanju, J. 2000, ApJ, 543, 740

649: \bibitem[]{} Burgay, M., D'Amico, N., Possenti, A. et al. 2003,

650:              Nature, 426, 531

651: \bibitem[]{} Camilo, F., Stairs, I. H., Lorimer, D. R. et al. 2002,

652: ApJ, 571, 41

653: \bibitem[]{} Demorest, P., Ramachandran, R., Backer, D. C., Ransom,

654: S. M., Kaspi, V. M., Arons, J., Spitkovsky, A. 2004, ApJ, Submitted

655: (astro-ph/0402025)

656: \bibitem[]{} Deshpande, A. A. \& Rankin, J. M. 1999, ApJ, 524, 1008

657: \bibitem[]{} Deshpande, A. A., Rankin, J. M. 2001, MNRAS, 322, 438

658: \bibitem[]{} Jenet, F. A., Gil, J., 2004, ApJ, 602, L89

659: \bibitem[]{} Jenet, F. A., Anderson, S. B., Prince, T. A. 2001, ApJ,

660: 546, 394

661: \bibitem[]{} Jenet, F. A., Ransom, S. M. 2004, Nature, in press

662: \bibitem[]{} Hankins, T. H., Fowler, L. A. 1986, ApJ, 304, 256

663: \bibitem[]{} Helfand, D. J., Manchester, R. N., Taylor, J. H. 1975,

664: ApJ, 198, 661

665: \bibitem[]{} Kaspi, V. M., Ransom, S. M., Backer, D. C., Ramachandran,

666: R., Demorest, P., Arons, J., Spitkovsky, A. 2004, ApJ, Submitted

667: (astro-ph/0401614)

668: \bibitem[]{} Kramer, M., Lange, C., Lorimer, D. R. et al. 1999, ApJ,

669: 526, 957

670: \bibitem[]{} Lyne, A. G., Burgay, M., Kramer, M., Possenti, A.,

671: Manchester, R. N., Camilo, F., McLaughlin, M. A., Lorimer, D. R.,

672: D'Amico, N., Joshi, B. C., Reynolds, J., Freire, P. C. C.  2004,

673: Science, 303, 1153

674: \bibitem[]{} Manchester, R. N. \& Johnston, S. 1995, ApJ, 441, L65

675: \bibitem[]{} Ramachandran, R. \& Kramer, M., 2003, A\&A, 407, 1085

676: \bibitem[]{} Rankin, J. M. 1983, ApJ, 274, 359

677: \bibitem[]{} Rankin, J. M. 1986, ApJ, 301, 901

678: \bibitem[]{} Rankin, J. M., Ramachandran, R. 2003, ApJ, 590, 411

679: \bibitem[]{} Ransom, S. M., Kaspi, V. M., Ramachandran, R., Demorest,

680: P., Backer, D. C., Pfahl, E. D., Ghigo, F. D., Kaplan, D. L. 2004,

681: ApJ, submitted (astro-ph/0404149)

682: \bibitem[]{} Weisberg, J. M., Armstrong, B. K., Backus, P. R., Cordes,

683: J. M., Boriakoff, V., Ferguson, D. C. 1986, AJ, 92, 621

684: \end{thebibliography}

685: 

686: \end{document}

687: