Dick Manchester (ATNF)

The 20-cm multibeam receiver at Parkes was originally conceived with the aim of detecting galaxies through their 21-cm HI emission and has been very successful in that endeavour. However, it did not take pulsar astronomers long to realise the potential of the instrument for pulsar surveys. The outstanding success of pulsar surveys using this receiver has made Parkes the premier telescope world-wide for pulsar discoveries. Counting earlier surveys, Parkes has discovered about two-thirds of the 1750 or so known pulsars or, put another way, Parkes has discovered twice as many pulsars as the rest of the world's telescopes put together! More than three-quarters of these pulsars, or more than half the total number of pulsars known, have been discovered in the past eight or nine years using the multibeam receiver.

What are the principal reasons for this outstanding success? There are three. Firstly, the location of Parkes is ideal for Galactic surveys, with the Galactic Centre passing almost overhead. The spatial density of pulsars increases greatly toward the Galactic Centre and so searches of the inner parts of our Galaxy are generally more productive. Secondly, thanks to the foresight and skill of our engineers and scientists, the receivers and back-end systems installed at Parkes are at the forefront in development of innovative technologies. In particular, the multibeam receiver was years ahead of the competition (in fact, we had to build one for them!). With its 13 beams and extremely good sensitivity, this receiver is a highly effective instrument for pulsar (and HI) surveys, increasing the effectiveness of such surveys by more than an order of magnitude. Finally, the international team responsible for the major multibeam pulsar surveys is highly experienced in developing the equipment and signal-processing techniques needed for pulsar surveys.

The Parkes 20-cm Multibeam Pulsar System

The 20-cm multibeam system has 13 beams arranged in a double hexagon about a central beam (Figure 1). The beams are spaced on the sky by approximately two beamwidths, but by combining four interleaved pointings, essentially complete sky coverage can be obtained. Each of the 26 receiver channels (two orthogonal linear polarisations per beam) has a bandwidth of nearly 300 MHz centred at about 1375 MHz. As well as having very wide bandwidths, the first-stage amplifiers (designed and constructed at Jodrell Bank Observatory, University of Manchester) have exceptionally good noise figures, with system temperatures of about 21 K. The 26 signals are amplified and down-converted to feed a massive filterbank/digitiser system which was designed at Jodrell Bank Observatory and constructed there and at the Astronomical Observatory, Bologna. Each filterbank has 96 channels, each 3 MHz wide. Corresponding channels from the two polarisations are summed to give 1248 data streams which are one-bit digitised with a sample interval typically of 125 or 250 microseconds and written to tape for subsequent analysis.

Several different pulsar surveys have been undertaken with the multibeam system and all used basically the same analysis procedure. Data from each beam are "dedispersed", that is, data from different frequency channels are delayed to compensate for the effects of interstellar dispersion and then summed, for a range of dispersion measures. Each dedispersed data stream is then Fourier-transformed to give the modulation spectrum of the signal. Peaks in this spectrum may result from various forms of interference or may represent a pulsar. Real pulsar signals can generally be distinguished from interference by their dependence on frequency, time and dispersion. After rejecting those thought to be due to interference, details of strong signals are saved as pulsar candidates. These are re-observed at the telescope to show whether or not they are real pulsars. Confirmed pulsars are then observed at intervals of a few weeks over a year or 18 months to determine the precise pulsar period and other parameters, including binary parameters if the pulsar is a member of a binary system.

Multibeam Pulsar Surveys

The most extensive survey is the Parkes Multibeam Pulsar Survey (PMPS), a search of a 150^o by 10^o strip along the southern Galactic Plane. A large international team with members from the UK (Jodrell Bank Observatory), Italy (Bologna Astronomical Observatory/Cagliari Astronomical Observatory), USA (Columbia University, Massachusetts Institute of Technology, Haverford College), Canada (University of British Columbia, McGill University) and the ATNF was responsible for this survey which commenced in mid-1997 and took six years to complete. A total of 2670 pointings, each of 35 minutes duration, was needed to cover the survey region and more than 3 terabytes of data were recorded. The data were processed using computer clusters at the collaborating institutions. Recently, all of the data have been reprocessed using the COBRA cluster at Jodrell Bank Observatory to give improved rejection of interference, improved sensitivity to long-period pulsars and pulsars in short-period binary orbits and to search for isolated dispersed pulses. Overall, the survey has been extremely successful, finding more than 750 pulsars, by itself nearly doubling the number of known pulsars.

A more limited survey at higher Galactic latitudes was carried out by a team from Swinburne University with an extension done in collaboration with Caltech. This survey was optimised for discovery of millisecond pulsars (MSPs) and found 16 such pulsars among a total of 95 discoveries. The PMPS collaboration also carried out a high-latitude survey with parameters similar to that of the Swinburne survey. This survey found 17 pulsars, including the now-famous "double pulsar", PSR J0737-3039A/B, voted by Science magazine to be one of the top ten scientific breakthroughs of 2004. The high efficiency of the multibeam system was also exploited in a survey of the Magellanic Clouds which discovered 14 very weak pulsars, 12 of which are believed to be associated with the Clouds. With the eight previously known pulsars, two of which were discovered at X-ray wavelengths, this brings the total number of Magellanic Cloud pulsars known to 20. So far, these are the only known pulsars outside of our Galaxy.

Figure 2 shows the distribution in Galactic coordinates of all known radio pulsars (excepting those in globular clusters) with pulsars discovered in the principal multibeam surveys marked. Although this distribution is strongly affected by observational selection, there is a clear concentration of pulsars along the Galactic equator and in the inner Galactic quadrants. Because of their high density on the sky, most of the PMPS pulsars are unresolved in this plot.

Pulsars and Magnetars

Pulsars are renowned as highly precise celestial clocks. But their periods are not constant - all pulsars slow down as a result of loss of energy and angular momentum to "magnetic-dipole" radiation (electromagnetic radiation at the pulsar frequency) and charged-particle winds. The rate of period increase is tiny, typically a fraction of a microsecond a year, and it varies considerably from pulsar to pulsar. The plot of rate of period increase, that is, the first period time-derivative, versus pulsar period (the "P - Pdot diagram") is a basic tool of pulsar astrophysics. Figure 3 shows this plot for all known pulsars except those in globular clusters. (The observed period derivative for globular cluster pulsars is often affected by acceleration of the pulsar in the cluster gravitational potential and so it doesn't represent the intrinsic spindown of the pulsar.) This figure clearly shows the huge number of pulsars discovered in the PMPS!

Pulsars in the top-right of the diagram marked with an open star are the AXPs, Anomalous X-ray Pulsars, which are evidently rotating neutron stars just like pulsars, but with enormous magnetic fields, greater than 10¹⁵ G in some cases, and relatively long periods, between 5 and 12 seconds. They are only detectable at hard X-ray and γ-ray wavelengths. Despite their rapid spindown, their integrated X-ray and γ-ray luminosity exceeds the power available from the loss of rotational kinetic energy and it is believed that they are powered by decay of the super-strong magnetic fields. Because of this, they are often known as "magnetars".

As well as greatly increasing the available sample for studies of the evolution and Galactic distribution of pulsars and interstellar medium studies, the multibeam surveys have uncovered new populations of pulsars and many interesting individual objects. A whole new population of young, highly magnetised and relatively long-period radio pulsars has been found by the PMPS. These pulsars have magnetic fields of 10¹³ G or more and periods as long as 7.7 s, which puts some of them at least in the same part of Figure 3 as the AXPs. Yet these are radio pulsars with very different properties to the AXPs, a difference which is not currently understood. The large Pdot of these pulsars means that they evolve very quickly and, despite their relatively small numbers, they account for a large fraction of the total pulsar birthrate.

RRATs

Another completely new and very interesting population of pulsars, the so-called Rotating Radio Transients, or RRATs for short, was revealed by the single-pulse analysis of the PMPS data. Like the AXPs, these objects have very different emission properties to normal pulsars and are believed to be rotating neutron stars with relatively long periods. RRATs are distinguished by the fact that just single pulses of emission are seen at intervals which range from several minutes to several hours. Eleven of these sources were found in the reanalysis of the PMPS data. Figure 4 shows an individual dispersed pulse from two of the sources. A careful analysis of the intervals between the pulses from a given source showed that there was a common factor, much shorter than the intervals between the pulses and typically a few seconds. This common factor is identified as the rotation period of the under-lying neutron star. Observed periods range from 0.44 to 6.9 s.

This identification of RRATs as rotating neutron stars was confirmed by long-term timing of three of the most frequently pulsing objects, which showed the slow period increase typical of pulsars. In fact, for one of the objects at least, the implied magnetic field is greater than 10¹³ G, putting the RRATs among the long-period high-magnetic-field pulsars discovered by the PMPS. The emission from RRATs must be beamed as for normal pulsars, otherwise there would be no periodicity. However unlike other pulsars, the emission process is extremely intermittent, being active for only a small fraction of the neutron-star rotation period - in no case have two consecutive pulses ever been seen. It is not known why the emission process is so different. RRATs have a similar Galactic distribution to normal pulsars. They could be relatively young neutron stars born with different properties to normal pulsars or maybe they represent a terminal stage in the life of a normal pulsar. At least in the former case, their discovery means that the population of neutron stars in the Galaxy is much greater than previously realised. Their extremely intermittent nature means that many have been missed by pulsar surveys and so the underlying population must be large, maybe more than that of normal pulsars. If true, this would have significant implications for our understanding of mechanisms for formation of neutron stars. Finding more of these enigmatic objects will help to solve these mysteries but unfortunately it will not be easy.

Binary Pulsars and Tests of Gravitational Theories

In total, the Parkes multibeam surveys have discovered 40 binary pulsars, nearly half of the 87 known in the Galactic disk. Many interesting systems have been discovered. For example, PSR J1909-3744 is a 2.95-ms pulsar in a 1.53-day orbit around a white dwarf companion discovered in the Swinburne mid-latitude survey. The main interest of this pulsar comes from its very narrow (half-power width 42 microseconds) and relatively strong pulse which leads to very precise timing of the pulsar and hence detection of a number of interesting effects. Daily averages for this pulsar yield an rms timing residual of only 74 ns, the most accurate so far obtained. As well as determining a very precise position and proper motion for the system, timing measurements have given an accurate measurement of the annual parallax (0.88 ± 0.03 mas) and hence the distance to the system (1.14 ± 0.04 kpc). They have also allowed measurement of the Shapiro delay, that is, the delay due to the ray path passing through curved space-time close to the companion star (Figure 5). The shape and amplitude of this curve give both the inclination angle i of the orbit and the mass of the companion white dwarf. Fortuitously, the orbit of this system is viewed almost edge-on (inclination angle i = 86^o.8) and hence the Shapiro delay is large, giving an accurate value for the companion mass, 0.2038 ± 0.0022 solar masses. This, combined with the mass function, a relation between the masses and orbital parameters derived from Kepler's third law, gives a value for the pulsar mass, 1.438 ± 0.024 solar masses, very similar to neutron-star masses derived from observations of double-neutron-star (DNS) systems. This value is interesting as it shows that a pulsar can become highly recycled with accretion of a relatively modest amount of mass.

As Figure 3 shows, the PMPS has been especially successful in filling in what was something of a gap between MSPs and normal pulsars. Binary pulsars with these intermediate periods are very interesting as they are mainly DNS systems, formed from binary systems containing massive stars. These stars evolve quickly and there is insufficient time for the recycling process to spin up the first-born neutron star to short millisecond periods before the second supernova explosion and formation of the second neutron star. Two of the eight DNS systems currently known were found in the PMPS and another, the double pulsar, was discovered in the High-Latitude multibeam survey. These systems are important for a number of reasons. Because of the large system masses and compact orbits (most have orbital periods of less than one day), relativistic perturbations to the orbits are readily detectable, allowing a variety of tests of the gravitational theories used to interpret these effects. Furthermore, because of the highly compact nature of neutron stars (GM/Rc² ~ 0.1), gravitational theories can be tested under strong-field conditions that are inaccessible elsewhere. Another significant property of DNS systems is that their orbits decay relatively rapidly due to loss of orbital energy to gravitational radiation. This ultimately leads to a violent coalescence of the two neutron stars to form a black hole. These coalescence events are the principal astrophysical target of ground-based gravitational-wave detectors such as LIGO and VIRGO.

The Double Pulsar

Without doubt, the most exciting of these discoveries is the double pulsar, PSR J0737-3039A/B. This unique system consists of a 22-ms recycled pulsar (A) in a mildly eccentric orbit with a much slower but younger 2.7-s pulsar (B). The orbital period is just 2.4 hours, making it the most highly relativistic binary system known. For example, the rate of precession of periastron is 16.9 deg/yr, nearly four times larger than that of PSR B1913+16, the Hulse-Taylor binary pulsar. The detection of the second neutron star as a pulsar provided a spectacular confirmation of the ideas behind the recycling mechanism for formation of MSPs. It also makes the system a "double-line" binary, giving a direct measurement of the mass ratio of the two stars and providing an important constraint on gravitational theories. This system has by far the shortest coalescence time of all the DNS systems, about 85 Myr, leading to a greatly increased estimated rate for the detection of these events by gravitational-wave detectors such as LIGO. As if all this was not enough, fascinating interactions between the winds and magnetospheres of the two pulsars have been observed, opening up a new window on magnetospheric physics and the pulse emission process.

A total of five independent relativistic effects have now been observed in PSR J0737-3039A/B, more than for any other DNS system. Taken together with the mass ratio, these provide unprecedented tests of relativistic theories of gravity. Figure 6 shows the "mass - mass" diagram for PSR 0737-3039A/B with relativistic effects interpreted within the framework of Einstein's general theory of relativity (GR). All constraints are consistent with the very small blue region (visible on the inset), providing an outstanding confirmation that GR is an accurate theory of gravity under strong-field conditions. It also provides highly precise measurements of the masses of the two neutron stars. The precise agreement of the Shapiro delay inclination-angle constraint (s = sin i) with the masses determined by the intersection of the mass-ratio and periastron-precession constraints provides the most stringent test of GR under strong-field conditions currently available, verifying the theory to better than 0.1%. Furthermore, this is a test of a non-radiative prediction of GR and hence is qualitatively different to that from the Hulse-Taylor binary system.

All of these results have been obtained in less than three years of timing of this remarkable system. Future observations will lead to improvements in these constraints and, even more importantly, new and independent tests of other predictions of theories of relativistic gravitation. The measurement of orbit decay rate improves rapidly with data span and for PSR J0737-3039A/B it will exceed the precision of the measurement for the Hulse-Taylor binary system in 3 - 5 years. In contrast to the Hulse-Taylor system, the precision of the GR test in PSR J0737-3039A/B will not be significantly limited by the uncertainty in the relative acceleration of the binary system and the Sun in the Galactic gravitational field, at least for the next few decades. One exciting prospect is measurement of the effects of spin-orbit coupling on the observed precession of periastron. In principle, this can be used to constrain the moment of inertia of the neutron stars which would put significant limits on possible equations of state for neutron-star matter.

Conclusions

The Parkes multibeam pulsar surveys have had an extraordinary impact, not only on pulsar astronomy and astrophysics, but also on physics and astronomy in general. In just a few years they have doubled the number of known pulsars and revealed some fascinating individual objects, most notably the first-known double pulsar system, PSR J0737-3039A/B. Studies of these pulsars will continue for years to come and will undoubtedly reveal many new and important results over this diverse range of topics. This success is a tribute to the skill and dedication of many people including the engineers and scientists who designed, constructed and now maintain the system and the astronomers in the various collaborations involved in the surveys.

Acknowledgements

I thank my colleagues, especially those in the Parkes Multibeam Pulsar Survey and Parkes High Latitude Survey teams, without whose efforts this article could not have been written. Pulsar data shown in the figures were obtained from the ATNF Pulsar Catalogue (www.atnf.csiro.au/research/pulsar/psrcat).

Bibliography

The Multibeam Receiver:

Staveley-Smith, L., Wilson, W. E., Bird, T. S., Disney, M. J., Ekers, R. D., Freeman, K. C., Haynes, R. F., Sinclair, M. W., Vaile, R. A., Webster, R. L. & Wright, A. E., 1996. "The Parkes 21 cm Multibeam Receiver". PASA, 13, 43-248.

The Parkes Multibeam Pulsar Survey:

Manchester, R. N., Lyne, A. G., Camilo, F., Bell, J. F., Kaspi, V. M., D'Amico, N., McKay, N. P. F., Crawford, F., Stairs, I. H., Possenti, A., Morris, D. J. & Sheppard, D. C., 2001. "The Parkes multi-beam pulsar survey - I. Observing and data analysis systems, discovery and timing of 100 pulsars". MNRAS, 328, 17-35.

Morris, D. J., Hobbs, G., Lyne, A. G., Stairs, I. H., Camilo, F., Manchester, R. N., Possenti, A., Bell, J. F., Kaspi, V. M., Amico, N. D., McKay, N. P. F., Crawford, F. & Kramer, M., 2002. "The Parkes Multibeam Pulsar Survey - II. Discovery and timing of 120 pulsars". MNRAS, 335, 275-290.

Kramer, M., Bell, J. F., Manchester, R. N., Lyne, A. G., Camilo, F., Stairs, I. H., D'Amico, N., Kaspi, V. M., Hobbs, G., Morris, D. J., Crawford, F., Possenti, A., Joshi, B. C., McLaughlin, M. A., Lorimer, D. R. & Faulkner, A. J., 2003. "The Parkes Multibeam Pulsar Survey - III. Young pulsars and the discovery and timing of 200 pulsars". MNRAS, 342, 1299-1324.

Hobbs, G., Faulkner, A., Stairs, I. H., Camilo, F., Manchester, R. N., Lyne, A. G., Kramer, M., D'Amico, N., Kaspi, V. M., Possenti, A., McLaughlin, M. A., Lorimer, D. R., Burgay, M., Joshi, B. C. & Crawford, F., 2004. "The Parkes multibeam pulsar survey - IV. Discovery of 180 pulsars and parameters for 281 previously known pulsars". MNRAS, 352, 1439-1472.

Faulkner, A. J., Stairs, I. H., Kramer, M., Lyne, A. G., Hobbs, G., Possenti, A., Lorimer, D. R., Manchester, R. N., McLaughlin, M. A., D'Amico, N., Camilo, F. & Burgay, M., 2004. "The Parkes Multibeam Pulsar Survey - V. Finding binary and millisecond pulsars". MNRAS, 355, 147-158.

Lorimer, D. R., Faulkner, A., Lyne, A. G., Manchester, R. N., Kramer, M., McLaughlin, M. A., Hobbs, G., Possenti, A., Stairs, I. H., Camilo, F., Burgay, M., D'Amico, N., Corongui, A. & Crawford, F., 2006. "The Parkes multibeam pulsar survey: VI. discovery and timing of 142 pulsars and a Galactic population analysis". MNRAS, submitted.

Other Parkes Multibeam Surveys:

Edwards, R. T. & Bailes, M., 2001. "Recycled pulsars discovered at high radio frequency". ApJ, 553, 801-808.

Burgay, M., Joshi, B. C., D'Amico, N., Possenti, A., Lyne, A. G., Manchester, R. N., McLaughlin, M. A., Kramer, M., Camilo, F. & Freire, P. C. C., 2006. "The Parkes High-Latitude pulsar survey". MNRAS, 368, 283-292.

Jacoby, B. A., 2004. "Recycled pulsars". PhD thesis, Calfornia Institute of Technology.

Manchester, R. N., Fan, G., Lyne, A. G., Kaspi, V. M. & Crawford, F., 2006. "Discovery of 14 radio pulsars in a survey of the Magellanic clouds". ApJ, submitted.

Magnetars:

Thompson, C. & Duncan, R. C., 1996. "The soft gamma repeaters as very strongly magnetized neutron stars. II. Quiescent neutrino, X-ray, and Alfven wave emission". ApJ, 473, 322-342.

Kouveliotou, C., Dieters, S., Strohmayer, T., van Paradijs, J., Fishman, G. J., Meegan, C. A., Hurley, K., Kommers, J., Smith, I., Frail, D. & Murakami, T., 1998. "An X-ray pulsar with a superstrong magnetic field in the soft gamma-ray repeater SGR 1806-20". Nature, 393, 235-237.

McLaughlin, M. A., Stairs, I. H., Kaspi, V. M., Lorimer, D. R., Kramer, M., Lyne, A. G., Manchester, R. N., Camilo, F., Hobbs, G., Possenti, A., D'Amico, N. & Faulkner, A. J., 2003. "PSR J1847--0130: A radio pulsar with magnetar spin characteristics". ApJ, 591, L135-L138.

Woods, P. M., Kaspi, V. M., Thompson, C., Gavriil, F. P., Marshall, H. L., Chakrabarty, D., Flanagan, K., Heyl, J. & Hernquist, L., 2004. "Changes in the X-Ray Emission from the Magnetar Candidate 1E 2259+586 during its 2002 Outburst". ApJ, 605, 378-399.

RRATs:

McLaughlin, M. A., Lyne, A. G., Lorimer, D. R., Kramer, M., Faulkner, A. J., Manchester, R. N., Cordes, J. M., Camilo, F., Possenti, A., Stairs, I. H., Hobbs, G., D'Amico, N., Burgay, M. & O'Brien, J. T., 2006. "Transient radio bursts from rotating neutron stars". Nature, 439, 817-820.

Binary Pulsars and General Relativity:

Hulse, R. A. & Taylor, J. H., 1975. "Discovery of a pulsar in a binary system". ApJ, 195, L51-L53.

Stairs, I. H., Thorsett, S. E. & Arzoumanian, Z., 2004. "Measurement of Gravitational Spin-Orbit Coupling in a Binary-Pulsar System". Phys. Rev. Lett., 93, 141101-1 - 141101-4.

Weisberg, J. M. & Taylor, J. H., 2005. "The Relativistic Binary Pulsar B1913+16". In: "Binary Radio Pulsars", eds Rasio, F. & Stairs, I. H., Astronomical Society of the Pacific, San Francisco, 25-31.

Jacoby, B. A., Hotan, A., Bailes, M., Ord, S. & Kuklarni, S. R., 2005. "The mass of a millisecond pulsar". ApJ, 629, L113-L116.

The Double Pulsar:

Burgay, M., D'Amico, N., Possenti, A., A., Manchester, R. N., Lyne, A. G., Joshi, B. C., McLaughlin, M. A., Kramer, M., Sarkissian, J. M., Camilo, F., Kalogera, V., Kim, C.& Lorimer, D. R., 2003. "An increased estimate of the merger rate of double neutron stars from observations of a highly relativistic system". Nature, 426, 531-533.

Lyne, A. G., Burgay, M., Kramer, M., Possenti, A., Manchester, R. N., Camilo, F., McLaughlin, M. A., Lorimer, D. R., D'Amico, N., Joshi, B. C., Reynolds, J. & Freire, P. C. C., 2004. "A double-pulsar system: A rare laboratory for relativistic gravity and plasma physics". Science, 303, 1153-1157.

McLaughlin, M. A., Lyne, A. G., Lorimer, D. R., Possenti, A., Manchester, R. N., Camilo, F., Stairs, I. H., Kramer, M., Burgay, M., D'Amico, N., Freire, P. C. C., Joshi, B. C. & Bhat, N. D. R., 2004. "The Double Pulsar System J0737-3039: Modulation of A by B at Eclipse". ApJ, 616, L131-L134.

Kramer, M., Stairs, I. H., Manchester, R. N., McLaughlin, M. A., Lyne, A. G., Ferdman, R. D., Burgay, M., Lorimer, D. R., Possenti, A., D'Amico, N., Sarkissian, J. M., Reynolds, J. E., Joshi, B. C., Freire, P. C. C. & Camilo, F., 2006. "General relativity tested by the double pulsar". Nature, submitted.

Figure 1. The 20-cm multibeam receiver being hoisted into the Parkes focus cabin on 21 January, 1997, its first installation. With a diameter of 1.3 m and weight of 650 kg, the receiver is the heaviest and one of the largest ever installed at Parkes.

Figure 2. Galactic distribution of all known radio pulsars excepting those in globular clusters. The Milky Way runs along the horizontal axis with the Galactic Centre at the centre of the plot. Pulsars discovered in the principal multibeam surveys and the boundaries of those surveys are marked.

Figure 3. Plot of spin-down rate or period derivative (Pdot) versus pulsar period for all known pulsars (with measured Pdot) except those in globular clusters. Pulsars discovered in the principal multibeam surveys and AXPs are indicated and binary pulsars are marked with a circle around the symbol. Lines of constant characteristic age, t_c = P/(2Pdot), and surface magnetic-dipole field strength, ~ (P Pdot)^1/2, are shown along with the spin-up line which represents the minimum period that a pulsar can attain through "recycling" or accretion of mass from a companion star.

Figure 4. Single dispersed pulses from two RRATs, PSR J1443-60 and PSR J1819-1458. The lower part of each plot shows the dispersion of the pulse resulting from the frequency-dependent group delay in the interstellar medium and the upper plot shows the dedispersed pulse profile. (Image credit: M. McLaughlin)

Figure 5. Pulse timing residuals as a function of orbital phase for PSR J1909-3744 showing the Shapiro delay as the pulsar passes behind the companion. The upper part of the figure shows the observed residuals after fitting for all pulsar and binary parameters except Shapiro delay. The middle plot shows the total Shapiro delay term which peaks near orbital phase 0.25 when the pulsar is behind the companion. A portion of this Shapiro delay function is absorbed by other binary terms in the upper plot. The lower part shows the final residuals. (Jacoby et al. 2005)

Figure 6. Constraints on the masses of the two neutron stars in the double-pulsar system PSR J0737-3039A/B. The orange shaded areas are excluded by the condition that the sine of the orbital inclination angle cannot exceed 1.0 and the red line is from the measurement of the mass ratio R of the two stars. Other constraints are from relativistic effects interpreted within general relativity. Measurement of the precession of periastron gives the purple dashed lines, the blue dot-dashed lines are from variations in the relativistic time dilation and second-order Doppler effect as pulsar A moves around its eccentric orbit, the green lines are from measurements of the Shapiro delay as the signal from A passes over B and the black dash-double-dot lines are from measurement of orbit decay due to emission of gravitational waves from the system. The inset shows an expanded view of the region around the intersection of the various constraints. (Kramer et al. 2006)