1: \documentclass[12pt,preprint]{aastex}
2:
3: %% manuscript produces a one-column, double-spaced document:
4: %% \documentclass[manuscript]{aastex}
5: %% preprint2 produces a double-column, single-spaced document:
6: %% \documentclass[preprint2]{aastex}
7:
8: \usepackage{graphics}
9:
10: \newcommand{\msun}{\mbox{$M_{\sun}$}}
11: \newcommand{\rsun}{\mbox{$R_{\sun}$}}
12: \newcommand{\pdot}{\mbox{$\dot{P}$}}
13: \newcommand{\pbdot}{\mbox{$\dot{P}_{\rm b}$}}
14: \newcommand{\usec}{\mbox{$\mu$s}}
15:
16: \shorttitle{Shapiro delay in the PSR~J1640+2224 system}
17: \shortauthors{O.~L\"ohmer et al.}
18:
19: \begin{document}
20:
21: \title{Shapiro delay in the PSR~J1640+2224 binary system}
22:
23: \author{Oliver L\"ohmer}
24: \affil{Max-Planck-Institut f\"ur Radioastronomie, Auf dem H\"ugel 69,
25: D-53121 Bonn, Germany}
26:
27: \author{Wojciech Lewandowski}
28: \affil{Toru\'n Centre for Astronomy, Nicolaus Copernicus University,
29: Gagarina 11, 87-100 Toru\'n, Poland}
30:
31: \author{Alex Wolszczan}
32: \affil{Department of Astronomy and Astrophysics, Pennsylvania State
33: University, University Park, PA 16802, USA}
34: \affil{Toru\'n Centre for Astronomy, Nicolaus Copernicus University,
35: Gagarina 11, 87-100 Toru\'n, Poland}
36: \affil{Max-Planck-Institut f\"ur Radioastronomie, Auf dem H\"ugel 69,
37: D-53121 Bonn, Germany}
38:
39: \author{Richard Wielebinski}
40: \affil{Max-Planck-Institut f\"ur Radioastronomie, Auf dem H\"ugel 69,
41: D-53121 Bonn, Germany}
42:
43:
44: \begin{abstract}
45: We present the results of precision timing observations of the binary
46: millisecond pulsar \object{PSR J1640+2224}. Combining the pulse
47: arrival time measurements made with the Effelsberg 100-m radio
48: telescope and the Arecibo 305-m radio telescope, we have
49: extended the existing timing model of the pulsar to search for a
50: presence of the effect of a general-relativistic Shapiro delay
51: in the data. At the currently attainable precision level, the
52: observed amplitude of the effect constrains the companion mass to
53: $m_2=0.15^{+0.08}_{-0.05}\, M_\sun$, which is consistent with the
54: estimates obtained from optical observations of the white dwarf
55: companion and with the mass range predicted by theories of binary
56: evolution. The measured shape of the Shapiro delay curve restricts the
57: range of possible orbital inclinations of the PSR J1640+2224 system to
58: $78^{\circ}\le i\le 88^{\circ}$. The pulsar offers excellent prospects
59: to significantly tighten these constraints in the near future.
60: \end{abstract}
61:
62:
63: \keywords{astrometry -- stars: neutron -- binaries: general --
64: pulsars: individual (PSR J1640+2224)}
65:
66:
67: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
68: \section{INTRODUCTION
69: \label{intro}}
70:
71: Precision timing measurements of binary millisecond pulsars (Phinney
72: \& Kulkarni 1994\nocite{pk94}) with sufficiently high (near edge-on)
73: orbital inclinations make it possible to detect the effect of a
74: general-relativistic time delay of the pulsar signal in the
75: gravitational field of the companion star. For a pulsar in a circular
76: orbit this ``Shapiro delay'' (Shapiro 1964\nocite{sha64}) is given by
77: %
78: \begin{equation}\label{shap}
79: \Delta t = -2\, m_2\; T_{\sun}\, \ln[1-\sin i\, \sin(\Phi-\Phi_0)]\ ,
80: \end{equation}
81: %
82: where $\Phi$ is the orbital phase in radians, $\Phi_0$ is the phase of
83: the ascending node, and $T_{\sun}=(G\,M_{\sun}/c^3)$. In practice,
84: Shapiro delay is conveniently expressed in terms of two observables,
85: the ``range'' $r = m_2\,T_{\sun}$ and the ``shape'' $s = \sin i$
86: (Ryba \& Taylor 1991\nocite{rt91a}), the post-Keplerian orbital
87: parameters which allow a determination of the companion mass, $m_2$,
88: and the orbital inclination, $i$.
89:
90: Because high inclination orbits are relatively rare, Shapiro delay has
91: been detected in only four pulsar-white dwarf (WD) binaries,
92: \object{PSR J1713+0747} (Camilo, Foster \& Wolszczan
93: 1994\nocite{cfw94}), \object{PSR B1855+09} (Ryba \& Taylor 1991),
94: \object{PSR J0437$-$4715} (van Straten et al. 2001\nocite{vbb+01}),
95: \object{PSR J1909$-$3744} (Jacoby et al.\ 2003\nocite{jbv+03}), and
96: possibly in \object{PSR J0751+1807} (Nice, Splaver \& Stairs
97: 2003\nocite{nss03}). Such detections offer very useful means to
98: measure masses of the companion stars and to calibrate the pulsar
99: spin-down models against the cooling models of white dwarfs. This is
100: accomplished by comparing the spin-down age of a pulsar obtained from
101: timing observations with the cooling age of a white dwarf estimated
102: from its mass, an optical measurement of its temperature and from an
103: appropriate cooling model (Kulkarni 1986\nocite{kul86}). In
104: particular, such comparisons are important in assessing the
105: temperature modifying effect of hydrogen left over after the white
106: dwarf formation (van Kerkwijk et al.\ 2000\nocite{vbkk00}).
107:
108: Shapiro delay has also been detected in two double neutron star
109: (NS--NS) systems, \object{PSR B1534+12} (e.g.\ Stairs et al.\
110: 2002\nocite{stt+02}) and \object{PSR J0737$-$3039A} (Burgay et
111: al. 2003\nocite{bdp+03}; Lyne et al.\ 2004\nocite{lbk+04}). In these
112: cases, the measured parameters $r$ and $s$, together with the other
113: two strong gravity related post-Keplerian parameters, the periastron
114: advance $\dot\omega$, and the time dilation and gravitational redshift
115: $\gamma$, provide a ``clean'' test of general relativity and other
116: theories of gravity, in the sense that it does not mix the
117: relativistic strong-field and the radiative effects (Damour \& Taylor
118: 1992\nocite{dt92}).
119:
120: Neutron stars in pulsar--WD binaries are thought to undergo extended
121: periods of transfer of mass and angular momentum from their companions
122: (Phinney \& Kulkarni 1994). As a result, they are spun up to
123: millisecond periods and may end up having masses significantly larger
124: than the canonical value of $1.35\, M_{\sun}$ derived for stars in the
125: NS--NS systems (Thorsett \& Chakrabarty 1999\nocite{tc99}). The
126: existing mass measurements for PSR J0437$-$4715 ($m_1=1.58\pm
127: 0.18\,M_{\sun}$, van Straten et al. 2001) and PSR B1855+09
128: ($m_2=1.57^{+0.12}_{-0.11}\,M_{\sun}$) (Nice, Splaver \& Stairs 2003,
129: 2004\nocite{nss04}) provide support for this idea. On the other hand,
130: the recent mass measurement of $1.3\pm 0.2\,M_{\sun}$ derived from
131: precision timing observations of PSR J1713+0747 (Splaver et al. 2004)
132: is in accord with the canonical value. Undoubtedly, more data for
133: similar systems are needed to improve the existing statistics.
134:
135: In this paper, we report new results of timing measurements of the PSR
136: J1640+2224 binary system. A tentative detection of the Shapiro delay
137: in the pulse arrival times from the pulsar allows us to set
138: preliminary constraints on the orbital inclination of the system and
139: the mass of the pulsar companion. In Sect.~\ref{obs} we describe the
140: timing observations made at Effelsberg and Arecibo, present the timing
141: analysis, and summarize the resulting best-fit timing model for PSR
142: J1640+2224. In Sect.~\ref{disc} we discuss the new findings and their
143: implications. In particular, we use simulated timing observations to
144: demonstrate the expected potential of a future sub-microsecond timing
145: of PSR J1640+2224 to verify the validity of the current best-fit model
146: and to improve the estimates of masses of the pulsar and its white
147: dwarf companion.
148:
149:
150: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
151: \section{OBSERVATIONS AND TIMING ANALYSIS \label{obs}}
152:
153: We have conducted systematic, high-precision timing observations of
154: PSR~J1640+2224 with the 100-m Effelsberg radiotelescope of the
155: Max-Planck-Institut f\"ur Radioastronomie in Bonn, Germany, and the
156: 305-m Arecibo radiotelescope of the National Astronomy and Ionosphere
157: Center in Puerto Rico, over a 7-year period from 1996 until 2003.
158:
159: At Effelsberg, PSR J1640+2224 was observed approximately once a month
160: using a 1300$-$1700~MHz tunable HEMT receiver at a centre frequency of
161: 1410~MHz. In order to monitor changes of the dispersion measure (DM)
162: we occasionally collected data at 860~MHz. As a backend, we used the
163: Effelsberg--Berkeley Pulsar Processor (EBPP), which corrects for the
164: dispersion smearing of the signal employing a coherent de-dispersion
165: technique (Hankins \& Rickett 1975\nocite{hr75}). In the total power
166: mode, the EBPP provided 32 channels for both senses of circular
167: polarization with a maximum total bandwith of 112~MHz, depending on
168: the DM and the observing frequency (Backer et al.\
169: 1997\nocite{bdz+97}). For PSR~J1640+2224, total bandwidths of 54~MHz
170: and 27~MHz were available at 1410~MHz and 860~MHz, respectively. The
171: output signals of each channel were fed into the de-disperser boards
172: for coherent on-line de-dispersion and were synchronously folded at
173: the pulse period over a 7~min integration time.
174:
175: At Arecibo, the timing observations of PSR J1640+2224 were made with
176: the dual-circular polarization receiving systems at 430 MHz, 1130~MHz
177: and 1410~MHz and the Penn State Pulsar Machine (PSPM). The PSPM pulsar
178: backend is a computer controlled processor with a $2\times 128\times
179: 60$~kHz filterbank designed to conduct fast sampled pulsar searches
180: and precision timing measurements. Technical details of the backend
181: are given in Cadwell (1997\nocite{cad97a}). For our timing
182: observations, the two signals of opposite circular polarizations were
183: added together, smoothed with a 32 $\mu$s time constant, 4-bit
184: quantized, folded synchronously with the topocentric pulse period, and
185: stored for further processing. The pulse integration times were 3
186: min. at 430 MHz and 5 min. at both 1130 MHz and 1410 MHz.
187:
188: Both Effelsberg and Arecibo data were time stamped using the
189: observatory hydrogen maser clocks and later synchronized to UTC(NIST)
190: using the signals from the Global Positioning System (GPS). In order
191: to calculate the pulse time-of-arrival (TOA), high signal-to-noise
192: template profiles of the pulse were constructed for each backend and
193: observing frequency and least-squares fitted to the observed profiles
194: in frequency domain (Taylor 1991\nocite{tay91}). A theory-independent
195: timing model for binary pulsars, devised by Damour \& Deruelle
196: (1986\nocite{dd86}), was least squares fitted to the combined TOAs,
197: weighted by their individual uncertainties, using the software package
198: {\sc tempo}\footnote{http://pulsar.princeton.edu/tempo} and the DE200
199: planetary ephemeris (Standish 1990\nocite{sta90}).
200:
201: In the fitting procedure, the TOA segments obtained with the EBPP and
202: the PSPM were fitted for an unknown offset between the two data sets
203: resulting from different templates and TOA reference points in the
204: profiles. Using the full TOA set at all frequencies we determined the
205: DM of the pulsar. In the subsequent analysis, we fixed the best-fit
206: value for the DM and used only the 1410~MHz TOAs from Effelsberg and
207: both 1130 and 1410~MHz TOAs from Arecibo, as these high frequency data
208: were not significantly affected by the observed DM variations. A
209: determination of an initial best-fit model for the PSR J1640+2224
210: timing data involved a set of 12 parameters including the astrometric,
211: and the rotational parameters of the pulsar and the orbital parameters
212: of the binary system. In order to achieve a uniform reduced $\chi^2=1$
213: for each data segment, we increased the TOA uncertainties by a
214: constant amount, approximately equal to the post-fit rms noise, by
215: adding it in quadrature to the actual TOA values.
216:
217: \placefigure{\ref{fig:chi2}}
218:
219: In order to examine the timing data for a possible presence of
220: the Shapiro delay we employed a grid search procedure used by Ryba \&
221: Taylor (1991\nocite{rt91a}). We searched the $m_2 - \cos i$ plane for
222: a global $\chi^2$ minimum, by fixing the Shapiro parameters $r$ and
223: $s$ at nodes of an appropriately defined 2D grid and repeatedly
224: fitting for all other parameters for each set of $(r,s)$ values. As
225: displayed in Fig.~\ref{fig:chi2}, the grid search produces a
226: well-defined global $\Delta\chi^2$ minimum, equivalent to the best-fit
227: $\cos i=0.11^{+0.09}_{-0.07}$ ($i\sim 84^{+4}_{-6}$ degrees) and
228: $m_2=0.15^{+0.08}_{-0.05}\,M_{\sun}$ (1$\sigma$
229: uncertainties). Clearly, the inclusion of a Shapiro delay in the
230: timing model for PSR J1640+2224 leads to astrophysically plausible
231: estimates of both the companion mass and the inclination of the pulsar
232: orbit. In Fig.~\ref{fig:shapiro}, the timing residuals for the
233: combined Arecibo and Effelsberg observations are plotted as a function
234: of orbital phase. Because the observed Shapiro delay is weak and
235: both $r$ and $s$ are strongly covariant with other model parameters,
236: the effect is not detectable in residuals from the best fit involving
237: the Keplerian orbit alone, as seen in Fig.~\ref{fig:shapiro}a. On the
238: other hand, in Fig.~\ref{fig:shapiro}b, showing a Shapiro delay
239: signature extracted from the grid search with all other timing effects
240: removed (see also Ryba \& Taylor (1991) and Camilo, Foster \&
241: Wolszczan (1994)), the amplitude of the effect in the PSR J1640+2224
242: TOA residuals significantly exceeds the TOA uncertainties, as expected
243: from the result of the $\chi^2$ search displayed in
244: Fig.~\ref{fig:chi2}.
245:
246: In principle, the observed signature could be induced by DM variations
247: over the pulsar orbit. If the pulsar's WD companion had an extended
248: envelope created by the pulsar wind and the high-energy photon flux,
249: as observed in eclipsing binary systems (e.g. Nice, Arzoumanian \&
250: Thorsett 2000\nocite{nat00}), the electron column density would
251: fluctuate periodically as a function of orbital phase. For highly
252: inclined orbits this would obviously cause periodic,
253: frequency-dependent TOA variations that could mimic the effect of
254: Shapiro delay. We have ruled out this possibility by verifying that
255: the effect has the same amplitude in the TOA measurements made at four
256: different frequencies.
257:
258: In the case of binary pulsars with nearly circular, low-inclination
259: orbits, Shapiro delay becomes covariant with Roemer delay and cannot
260: be measured (Lange et al.\ 2001\nocite{lcw+01}). PSR J1640+2224 has
261: the most eccentric orbit among the pulsar--WD binaries (see
262: Table~\ref{tab:params} and Edwards and Bailes 2001\nocite{eb01b}) and
263: its inclination angle of $i=84\pm6\degr$ derived from our analysis
264: appears to be high enough to allow the inclusion of Shapiro
265: delay in the timing model (Fig.~\ref{fig:shapiro}b). In any case,
266: further observations of the pulsar with higher timing precision are
267: clearly necessary to fully assess a statistical significance of our
268: detection.
269:
270: \placefigure{\ref{fig:shapiro}}
271:
272: \placefigure{\ref{fig:resid-epoch}}
273:
274: The parameters of the best-fit timing model for PSR J1640+2224 are
275: listed in Table~\ref{tab:params} along with the ones for which only
276: upper limits could be determined. In this case, the upper limits were
277: obtained by allowing the parameters to vary, one at a time, in the
278: global fit. Also included in the table are the most important
279: parameters derived from the final model. Finally, the behavior of the
280: post-fit timing residuals as a function of time, spanning a 7-year
281: period, is shown in Fig.~\ref{fig:resid-epoch}. Evidently, the
282: best-fit model including the Shapiro delay leaves no additional
283: systematic effects above the current post-fit rms residual of 2.0
284: $\mu$s.
285:
286: \placetable{\ref{tab:params}}
287:
288:
289: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
290: \section{DISCUSSION \label{disc}}
291:
292: A new timing model for the binary millisecond pulsar PSR J1640+2224
293: discussed in this paper is entirely consistent with the previous
294: models published by Wolszczan et al.\ (2000)\nocite{wdk+00} and
295: Potapov et al.\ (2003)\nocite{pio+03}. In addition, owing to a higher
296: timing precision, the new model provides further reduction of the
297: parameter estimation errors and, above all, it includes
298: astrophysically sensible estimates of the Shapiro delay parameters.
299:
300: The binary companion to PSR J1640+2224 is a white dwarf with the
301: estimated cooling age and mass of 7$\pm$2 Gyr and 0.25$\pm 0.10\,
302: M_{\sun}$, respectively, as determined from the Hubble Space
303: Telescope observations (Lundgren et al.\ 1996\nocite{lfc96}). A range
304: of masses predicted by the binary period--companion mass ($P_b-m_2$)
305: relationship based on the theory of low- and intermediate-mass binary
306: evolution is $0.35\le m_2\le 0.39\,M_{\sun}$ (Tauris \& Savonije
307: 1999\nocite{ts99a}), and the minimum companion mass from the mass
308: function, $f(m_1,m_2)=(m_2 \sin i)^3(m_1+m_2)^{-2}=
309: (2\pi/P_b)^2x^3/T_{\sun}=0.0059\,M_{\sun}$ is $m_2=0.25$\,\msun\ for
310: a $m_1=1.4\,M_{\sun}$ neutron star. At the presently attainable level
311: of accuracy, the best-fit companion mass of
312: $m_2=0.15^{+0.08}_{-0.05}\,M_{\sun}$ derived from our data is
313: consistent with the above estimates, as illustrated in
314: Fig.~\ref{fig:chi2}.
315:
316: \placefigure{\ref{fig:err}}
317:
318:
319: Among the pulsar--WD binaries with a detectable Shapiro delay, only
320: PSR J0437$-$4715, PSR J1713+0747, and PSR B1855+09 have the values of
321: $m_2$ and $\sin i$ determined with the accuracy that is high enough to
322: make them usable in setting a tight constraint on the pulsar mass (van
323: Straten et al.\ 2001; Nice, Splaver \& Stairs 2003, 2004; Splaver et
324: al. 2004) and in investigating the details of evolution of the
325: pulsar's WD companion (van Kerkwijk et al. 2000). We have examined a
326: future potential of the PSR J1640+2224 timing to become comparably
327: useful by generating artificial TOAs according to the model of
328: Table~\ref{tab:params} and analyzing the data over progressively
329: longer periods of time for several reasonable values of the timing
330: precision. Encouragingly, as shown in Fig.~\ref{fig:err}, in only four
331: years of monthly timing measurements with a 0.5 $\mu$s precision, the
332: estimation errors of $m_2$ and $\cos i$ approach the respective levels
333: of $0.01\,M_{\sun}$ and 0.001. Since the Arecibo timing measurements
334: using the PSPM and an 8 MHz receiver bandwidth are characterized by a
335: $\sim$1 $\mu$s long-term residual, it is quite conceivable that the
336: required $\le$0.5 $\mu$s precision can be achieved with a new
337: generation of broadband, 100 MHz bandwidth backends already available
338: at the telescope. Further observations at this level of precision will
339: quite conceivably allow a verification of the timing model presented
340: in this paper. PSR J1640+2224 is also likely to become a valuable
341: member of a set of the most accurate pulsar clocks that can be timed
342: either individually, or as an array to detect a low-frequency
343: background of gravitational radiation (e.g. Thorsett \& Dewey
344: 1996\nocite{td96}, Jaffe \& Backer 2003\nocite{jb03}).
345:
346:
347: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
348: \acknowledgments
349:
350: We are very grateful to all staff at the Effelsberg and Arecibo
351: observatories for their help with the observations. We thank O.\
352: Doroshenko, A.\ Jessner and M.\ Kramer for their assistance in the
353: Effelsberg timing project and helpful discussions about the PSR
354: J1640+2224 binary system. A.W.'s research was supported by the
355: Alexander von Humboldt Foundation and the National Science Foundation
356: under Grant No. PHY99-07949. Arecibo Observatory is part of the
357: National Astronomy and Ionosphere Center, which is operated by Cornell
358: University under contract with the National Science Foundation.
359:
360:
361: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
362: \bibliographystyle{apj}
363: %\bibliography{journals,modrefs,psrrefs,loehmer,crossrefs}
364:
365: \begin{thebibliography}{}
366:
367: \bibitem[Backer {\rm et~al.} (1997)]{bdz+97}
368: Backer~D.~C., Dexter~M.~R., Zepka~A., D.~N., Wertheimer~D.~J., Ray~P.~S.,
369: Foster~R.~S., 1997, PASP, 109, 61
370:
371: \bibitem[{Burgay} {\rm et~al.} (2003)]{bdp+03}
372: {Burgay}~M. {\rm et~al.}, 2003, \nat, 426, 531
373:
374: \bibitem[{Cadwell} (1997)]{cad97a}
375: {Cadwell}~B., 1997, PhD Thesis, Pennsylvania State University
376:
377: \bibitem[Camilo, Foster \& Wolszczan (1994)]{cfw94}
378: Camilo~F., Foster~R.~S., Wolszczan~A., 1994, ApJ, 437, L39
379:
380: \bibitem[Cordes \& Lazio (2002)]{cl02a}
381: Cordes~J.~M., Lazio~T.~J.~W., 2002, ApJ, submitted (astro-ph/0207156)
382:
383: \bibitem[Damour \& Deruelle (1986)]{dd86}
384: Damour~T., Deruelle~N., 1986, Ann. Inst. H. Poincar\'e (Physique Th\'eorique),
385: 44, 263
386:
387: \bibitem[Damour \& Taylor (1992)]{dt92}
388: Damour~T., Taylor~J.~H., 1992, Phys. Rev. D, 45, 1840
389:
390: \bibitem[Edwards \& Bailes (2001)]{eb01b}
391: Edwards~R.~T., Bailes~M., 2001, ApJ, 553, 801
392:
393: \bibitem[{Hankins} \& {Rickett} (1975)]{hr75}
394: {Hankins}~T.~H., {Rickett}~B.~J., 1975, in Methods in Computational Physics
395: Volume 14 --- Radio Astronomy.
396: \newblock Academic Press, New York, p.~55
397:
398: \bibitem[{Jacoby} {\rm et~al.} (2003)]{jbv+03}
399: {Jacoby}~B.~A., {Bailes}~M., {van Kerkwijk}~M.~H., {Ord}~S., {Hotan}~A.,
400: {Kulkarni}~S.~R., {Anderson}~S.~B., 2003, \apjl, 599, L99
401:
402: \bibitem[{Jaffe} \& {Backer} (2003)]{jb03}
403: {Jaffe}~A.~H., {Backer}~D.~C., 2003, \apj, 583, 616
404:
405: \bibitem[Kulkarni (1986)]{kul86}
406: Kulkarni~S.~R., 1986, ApJ, 306, L85
407:
408: \bibitem[Lange {\rm et~al.} (2001)]{lcw+01}
409: Lange~C., Camilo~F., Wex~N., Kramer~M., Backer~D., Lyne~A., Doroshenko~O.,
410: 2001, MNRAS, 326, 274
411:
412: \bibitem[Lundgren, Foster \& Camilo (1996)]{lfc96}
413: Lundgren~S.~C., Foster~R.~S., Camilo~F., 1996, in Johnston~S., Walker~M.~A.,
414: Bailes~M., eds, Pulsars: Problems and Progress, {IAU} Colloquium 160.
415: \newblock Astronomical Society of the Pacific, San Francisco, p.~497
416:
417: \bibitem[{Lyne} {\rm et~al.} (2004)]{lbk+04}
418: {Lyne}~A.~G. {\rm et~al.}, 2004, Science, 303, 1153
419:
420: \bibitem[Nice, Arzoumanian \& Thorsett~{2000}]{nat00}
421: Nice~D.~J., Arzoumanian~Z., Thorsett~S.~E., 2000, in Kramer~M., Wex~N.,
422: Wielebinski~R., eds, Pulsar Astronomy - 2000 and Beyond, {IAU} Colloquium
423: 177.
424: \newblock Astronomical Society of the Pacific, San Francisco, p.~67
425:
426: \bibitem[{Nice}, {Splaver} \& {Stairs} (2003)]{nss03}
427: {Nice}~D.~J., {Splaver}~E.~M., {Stairs}~I.~H., 2003, in {Bailes}~M.,
428: {Nice}~D.~J., {Thorsett}~S.~E., eds, ASP Conf. Ser. 302: Radio Pulsars.
429: \newblock p.~75
430:
431: \bibitem[{Nice}, {Splaver} \& {Stairs} (2004)]{nss04}
432: {Nice}~D.~J., {Splaver}~E.~M., {Stairs}~I.~H., 2004, in {Camilo}~F.,
433: {Gaensler}~B.~M., eds, Young Neutron Stars and Their Environments, {IAU}
434: Symposium 218.
435: \newblock Astronomical Society of the Pacific, San Francisco, p.~49
436:
437: \bibitem[Phinney \& Kulkarni (1994)]{pk94}
438: Phinney~E.~S., Kulkarni~S.~R., 1994, Ann. Rev. Astr. Ap., 32, 591
439:
440: \bibitem[{Potapov} {\rm et~al.} (2003)]{pio+03}
441: {Potapov}~V.~A., {Ilyasov}~Y.~P., {Oreshko}~V.~V., {Rodin}~A.~E., 2003,
442: Astronomy Letters, 29, 241
443:
444: \bibitem[Ryba \& Taylor (1991)]{rt91a}
445: Ryba~M.~F., Taylor~J.~H., 1991, ApJ, 371, 739
446:
447: \bibitem[Shapiro (1964)]{sha64}
448: Shapiro~I.~I., 1964, Phys. Rev. Lett., 13, 789
449:
450: \bibitem[Splaver et al. (2004)]{spl+04}
451: Splaver~F.~M., Nice,~D.~J., Stairs,~I.~H., Lommen,~A.~N., Backer,~D.~C.,
452: 2004, ApJ, accepted (astro-ph/0410488)
453:
454: \bibitem[{Stairs} {\rm et~al.} (2002)]{stt+02}
455: {Stairs}~I.~H., {Thorsett}~S.~E., {Taylor}~J.~H., {Wolszczan}~A., 2002, \apj,
456: 581, 501
457:
458: \bibitem[Standish (1990)]{sta90}
459: Standish~E.~M., 1990, A\&A, 233, 252
460:
461: \bibitem[Tauris \& Savonije (1999)]{ts99a}
462: Tauris~T.~M., Savonije~G.~J., 1999, A\&A, 350, 928
463:
464: \bibitem[Taylor (1991)]{tay91}
465: Taylor~J.~H., 1991, Proc. I. E. E. E., 79, 1054
466:
467: \bibitem[Thorsett \& Chakrabarty (1999)]{tc99}
468: Thorsett~S.~E., Chakrabarty~D., 1999, ApJ, 512, 288
469:
470: \bibitem[Thorsett \& Dewey (1996)]{td96}
471: Thorsett~S.~E., Dewey~R.~J., 1996, Phys. Rev. D, 53, 3468
472:
473: \bibitem[van Kerkwijk {\rm et~al.} (2000)]{vbkk00}
474: van Kerkwijk~M.~H., Bell~J.~F., Kaspi~V.~M., Kulkarni~S.~R., 2000, ApJ, 530,
475: L37
476:
477: \bibitem[van Straten {\rm et~al.} (2001)]{vbb+01}
478: van Straten~W., Bailes~M., Britton~M., Kulkarni~S.~R., Anderson~S.~B.,
479: Manchester~R.~N., Sarkissian~J., 2001, Nature, 412, 158
480:
481: \bibitem[Wolszczan {\rm et~al.} (2000)]{wdk+00}
482: Wolszczan~A. {\rm et~al.}, 2000, ApJ, 528, 907
483:
484: \end{thebibliography}
485:
486:
487: \clearpage
488:
489: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
490: \begin{figure}
491: \includegraphics[angle=-90,width=12cm]{f1.ps}
492: \caption{ \label{fig:chi2} Constraints on $m_2$ and $\cos i$ in the
493: PSR J1640+2224 system from a $\chi^2$ search for the best-fit Shapiro
494: delay parameters. The global $\chi^2$ minimum is indicated by a cross.
495: The contours of $\Delta\chi^2 = 1.0$ (solid), $\Delta\chi^2 = 4.0$
496: (dashed-dotted) and $\Delta\chi^2 = 9.0$ (dashed) have extrema
497: respectively corresponding to $1\sigma$, $2\sigma$, and $3\sigma$
498: errors on the individual parameters $\cos i$ and $m_2$. Lines of
499: constant $m_1$ are indicated. Horizontal lines denote the $m_2$--range
500: bounded by a lower limit obtained from optical observations ($m_2\ge
501: 0.15$\msun) and an upper limit allowed by the $P_b-m_2$ relationship
502: ($m_2\le 0.39$\msun).}
503: \end{figure}
504: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
505:
506:
507: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
508: \begin{figure}
509: \includegraphics[angle=-90,width=14cm]{f2.ps}
510: \caption{ \label{fig:shapiro} Timing residuals for PSR J1640+2224
511: observed with the Arecibo telescope at 1130 MHz and 1410 MHz (filled
512: circles) and with the Effelsberg telescope at 1410 MHz (open circles),
513: as a function of orbital phase. (a) Post-fit residuals for the
514: best-fit model involving only the five Keplerian parameters.
515: (b) The effect of Shapiro delay on the
516: TOA residuals calculated with $\sin i$ and $m_2$ set to zero and all
517: other parameters fixed at their best-fit values. The solid curve
518: represents the delay predicted by general relativity for the best-fit
519: Shapiro parameters.
520: (c) Post-fit residuals for the best-fit model
521: including the $\sin i$ and $m_2$ parameters (see also
522: Table~\ref{tab:params}).
523: }
524: \end{figure}
525: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
526:
527:
528: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
529: \begin{figure}
530: \includegraphics[angle=-90,width=12cm]{f3.eps}
531: \caption{ \label{fig:resid-epoch} Best-fit timing residuals for PSR
532: J1640+2224 as a function of observing epoch. See
533: Fig.~\ref{fig:shapiro} for further explanation.}
534: \end{figure}
535: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
536:
537:
538: %%%%%%%%%%%%%%%%%%%%%%%%%%%% FIGURE 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
539: \begin{figure}
540: \includegraphics[angle=-90,width=14cm]{f4.ps}
541: \caption{ \label{fig:err} Estimation errors of the companion mass and
542: orbital inclination from the simulated TOA measurements of PSR
543: J1640+2224 with a 0.5 $\mu$s precision. The initial error values are
544: equal to those currently observed.}
545: \end{figure}
546: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
547:
548:
549: \clearpage
550:
551:
552: %%%%%%%%%%%%%%%%%%%%%%%%%%%% Table 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
553: \begin{table*}
554: \begin{center}
555: \caption{Timing model for PSR~J1640+2224\label{tab:params}
556: }
557: \vspace{0.1cm}
558: \begin{tabular}{ll}
559: \hline\hline
560: \multicolumn{2}{c}{\it Measured parameters\,$^a$}\\
561: \hline
562: Right ascension, $\alpha$ (J2000) & $16\,\fh 40\,\fm 16\,\fs
563: 742307(10)$ \\
564: Declination, $\delta$ (J2000) & $22\degr 24\arcmin 08\,\farcs 9413(3)$\\
565: Proper motion, $\mu_{\alpha}$ (mas\,yr$^{-1}$) & $1.66(12)$ \\
566: Proper motion, $\mu_{\delta}$ (mas\,yr$^{-1}$) & $-11.3(2)$ \\
567: Pulse frequency, $\nu$ (s$^{-1}$) & $316.123984313238(2)$\\
568: Pulse frequency derivative, $\dot\nu$ (10$^{-16}$ s$^{-2}$) & $-2.8257(9)$\\
569: Pulse period, $P$ (ms) & $3.16331581791380(2)$\\
570: Period derivative, $\dot P$ (10$^{-20}$ s\,s$^{-1}$) & $0.28276(9)$\\
571: Epoch (MJD) & $51700.0$ \\
572: Dispersion measure, DM (pc\,cm$^{-3}$) & $18.4260(8)$ \\
573: Orbital period, $P_{\rm b}$ (d) & $175.46066194(7)$\\
574: Projected semi-major axis, $x$ (lt-s) & $55.3297198(4)$ \\
575: Eccentricity, $e$ & $0.000797262(14)$ \\
576: Epoch of periastron\,$^b$, $T_0$ (MJD) & $51626.1785(5)$\\
577: Longitude of periastron\,$^b$, $\omega$ (deg)& $50.7308(10)$ \\
578: Shape of Shapiro delay\,$^c$, $s$ & $0.99^{+0.01}_{-0.01}$\\
579: Range of Shapiro delay\,$^c$, $r$ ($\mu$s)& $0.74^{+0.39}_{-0.25}$\\
580: \hline
581: \multicolumn{2}{c}{\it Measured upper limits\,$^d$}\\
582: \hline
583: Parallax, $\pi$ (mas) & $<3.7$\\
584: Pulse frequency second derivative, $|\ddot\nu|$ (s$^{-3}$)& $<4\times 10^{-27}$\\
585: DM derivative, $|d({\rm DM})/dt|$ (pc\,cm$^{-3}$\,yr$^{-1}$) & $<1.3\times 10^{-3}$\\
586: Orbital period derivative, $|\dot P_{\rm b}|$ (s\,s$^{-1}$)& $<3\times 10^{-10}$ \\
587: Derivative of projected semi-major axis, $|\dot x|$ (lt-s s$^{-1}$) &
588: $<1.7\times 10^{-14}$ \\
589: Periastron rate of change, $|\dot\omega|$ (deg yr$^{-1}$) & $<1.1\times 10^{-3}$ \\
590: \hline
591: \multicolumn{2}{c}{\it Derived parameters}\\
592: \hline
593: Galactic longitude, $l$ & $41\fdg 051$\\
594: Galactic latitude, $b$ & $38\fdg 271$\\
595: DM distance$^e$ (kpc)& $1.16$ \\
596: Composite proper motion, $\mu$ (mas\,yr$^{-1}$) & 11.4(2)\\
597: Companion mass, $m_2$ (\msun) & $0.15^{+0.08}_{-0.05}$ \\
598: Orbital inclination angle, $i$ (deg) & $84^{+4}_{-6}$ \\
599: Mass function, $f_{\rm m}$ (\msun) & $0.0059074304(2)$\\
600: Number of TOAs & 314\\
601: Timing RMS (\usec) & 2.0 \\
602: \noalign{\smallskip}\hline
603: \end{tabular}
604: \end{center}
605: {\small $^a$ Figures in parentheses are $2\sigma$ uncertainties in the
606: last digits quoted (twice the formal {\sc tempo} errors).}\newline
607: {\small $^b$ $\omega$ and $T_0$ are highly covariant. Observers
608: should use $\omega=50.730834835740$ and $T_0=51626.178534099$}.\newline
609: {\small $^c$ Uncertainties are $1\sigma$ errors derived from the
610: $\chi^2$ analysis (see Sect.~2).}\newline
611: {\small $^d$ Upper limits represent 95\% C.L.}\newline
612: {\small $^e$ from the Cordes \& Lazio (2002\nocite{cl02a}) model of the
613: Galactic electron density distribution, with typical uncertainties of
614: 10\%}.\newline
615: \end{table*}
616: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
617:
618: \end{document}
619: