Discovery and timing of three millisecond pulsars in radio and gamma-rays with the GMRT and Fermi-LAT
B. Bhattacharyya, J. Roy, T. J. Johnson, P. S. Ray, P. C. C. Freire, Y. Gupta, D. Bhattacharya, A. Kaninghat, B. W. Stappers, E. C. Ferrara, S. Sengupta, R. S. Rathour, M. Kerr, D. A. Smith, P. M. Saz Parkinson, S. M. Ransom, P. F. Michelson
DDraft version February 9, 2021
Typeset using L A TEX modern style in AASTeX63
Discovery and timing of three millisecond pulsars in radio and γ -rayswith the GMRT and Fermi -LAT
B. Bhattacharyya, J. Roy, T. J. Johnson, P. S. Ray, P. C. C. Freire, Y. Gupta, D. Bhattacharya, A. Kaninghat, B. W. Stappers, E. C. Ferrara, S. Sengupta,
4, 8
R. S. Rathour,
1, 9
M. Kerr, D. A. Smith, P. M. Saz Parkinson,
11, 12, 13
S. M. Ransom, and P. F. Michelson National Centre for Radio Astrophysics, Tata Institute of Fundamental Research, Pune 411 007,India George Mason University, resident at U.S. Naval Research Laboratory U.S. Naval Research Laboratory, Washington, DC 20375, USA Max-Planck-Institut f¨ur Radioastronomie, Bonn, D-53121, Germany Inter-University Centre for Astronomy and Astrophysics, Pune 411 007, India Jodrell Bank Centre for Astrophysics, School of Physics and Astronomy, The University ofManchester, Manchester M13 9PL, UK UMD and NASA/GSFC Indian Institute of Technology, Kharagpur, West Bengal 721302 Nicolaus Copernicus Astronomical Centre, Polish Academy of Sciences, Bartycka 18, PL-00-716Warsaw, Poland Centre d’ ´Etudes Nucl´eaires de Bordeaux Gradignan, IN2P3/CNRS, Universit´e Bordeaux, BP120,33175 Gradignan, France Santa Cruz Institute for Particle Physics, Department of Physics and Department of Astronomyand Astrophysics, University of California at Santa Cruz, Santa Cruz, CA 95064, USA Department of Physics, The University of Hong Kong, Pokfulam Road, Hong Kong, China Laboratory for Space Research, The University of Hong Kong, Hong Kong, China National Radio Astronomy Observatory, 1003 Lopezville Road, Socorro, NM 87801, USA W. W. Hansen Experimental Physics Laboratory, Kavli Institute for Particle Astrophysics andCosmology, Department of Physics and SLAC National Accelerator Laboratory, StanfordUniversity, Stanford, CA 94305, USA
ABSTRACTWe performed deep observations to search for radio pulsations in the directions of375 unassociated
Fermi
Large Area Telescope (LAT) γ − ray sources using the Gi-ant Metrewave Radio Telescope (GMRT) at 322 and 607 MHz. In this paper wereport the discovery of three millisecond pulsars (MSPs), PSR J0248+4230, PSRJ1207 − − ∼ − − ∼
62 days about a companion of minimum mass 0.32 M (cid:12) . We also presentmulti-frequency pulse profiles of these MSPs from the GMRT observations. PSRJ1536 − γ -ray pulsations a r X i v : . [ a s t r o - ph . H E ] F e b Bhattacharyya et al. from these three MSPs, confirming them as the sources powering the γ -ray emission.For PSR J1536 − − γ -ray timing using ∼ γ -ray pulse times of arrivals (TOAs) along with the radio TOAs. PSR J1536 − γ -ray emission out to above 25 GeV, confirming earlierassociations of this MSP with a ≥
10 GeV point source. The multi-wavelength pulseprofiles of all three MSPs offer challenges to models of radio and γ -ray emission inpulsar magnetospheres. INTRODUCTIONThe Large Area Telescope (LAT, Atwood et al. 2009), the primary instrumenton the
Fermi Gamma-ray Space Telescope , has been surveying the GeV γ -ray skysince its scientific activation on August 4, 2008. This has dramatically increased thenumber of known γ ray sources and with each catalog release there have been anincreasing number of sources unassociated with any known counterpart likely to bepowering the γ -ray emission. Particularly at high Galactic latitude, many of thesesources have proven to be hitherto unknown millisecond pulsars (MSPs) (Ray et al.2012). Searching for pulsations of unknown MSPs in the γ -ray band is extraordinarilycomputationally expensive, particularly in the case of binaries. While it has provenpossible in a few cases (e.g., Nieder et al. 2020, who used astrometric and orbital dataprovided by optical observations to greatly reduce the necessary number of trials),it is generally far more efficient to first search for radio pulsars in the direction ofthese sources. Targeted searches for radio pulsations at the position of unassociatedLAT point sources coordinated by the Fermi
Pulsar Search Consortium (PSC, Rayet al. 2012) have resulted in the discovery of 95 radio MSPs so far, including the onesreported here. Finding pulsars powering these sources is important for identifyingthe nature of the γ -ray sources and for the astrophysics made possible by timing thenewly-discovered pulsars. An identification also rules out more exotic possible sourcessuch as dark matter subhalos (Coronado-Blazquez et al. 2019).Using the LAT sources to guide searches is a powerful technique. It allows deepsearches through long observations as well as allowing multiple visits per source. Thisis valuable because a pulsar can be missed in a single observation due to scintillation,eclipses, or acceleration in a binary system. The Giant Metrewave Radio Telescope(GMRT ) − a multi-element aperture synthesis telescope consisting of 30 antennaseach of 45 m diameter, having maximum baseline length of 25 km (Swarup et al. 1997) − is particularly well suited to this task. The low frequency observing capabilities(300–600 MHz) of the GMRT are optimal for sensitive detection of MSPs having steepspectra and typically low values of the dispersion measures. Its design, featuring alarge array of small telescopes, provides multiple advantages: (1) wide field of viewwith incoherent beam (FWHM of 80 (cid:48) at 322 MHz, and 40 (cid:48) at 607 MHz; ideal for pulsar http://gmrt.ncra.tifr.res.in iscovery and timing of three MSPs ∼ (cid:48)(cid:48) ) using the imaging capability, even on search observations. The semi-major axis ofthe 95% confidence error ellipses of the Fermi -LAT sources are about ± (cid:48) (althoughthe exact value is a function of location and integration time). Hence the larger beamwidth of the GMRT at lower frequencies is of considerable help. This wide beamallows the GMRT to search in a single observation faint LAT sources that are notwell localized, something that cannot be done with large single dish telescopes. Theprospect of the GMRT in pulsar searches has been demonstrated by the discovery of30 pulsars in targeted and blind searches at an encouraging pulsar-per-square degreediscovery rate (e.g., Ray et al. 2012; Bhattacharyya et al. 2013, 2016, 2019).In this paper we present the GMRT discoveries, follow-up timing and subsequentdiscovery of γ -ray pulsations for three MSPs which are associated with Fermi -LATsources. Section 2 details the target selection and provides limiting flux densities forthe sources from which pulsations were not detected. A list of all the GMRT pointingsand corresponding detection limits are presented in the appendix. Section 3 of thispaper details the search and timing observations with the GMRT. Section 4 detailsthe discoveries. Section 5 presents the more accurate position estimates of thesethree MSPs by localizing them in the image plane with the GMRT interferometricarray. Results from follow up timing studies of the discovered pulsars are reported inSection 6. Section 7 presents the results from γ -ray analysis of these pulsars. Section8 presents the discussion and Section 9 a summary. A list of all the GMRT pointingsand corresponding detection limits are presented in the Appendix. SOURCE SELECTIONAs part of a broader effort coordinated by the
Fermi
PSC, we selected sources fromearly versions of the
Fermi -LAT catalogs (Abdo et al. 2010; Nolan et al. 2012 analysisthat were not associated with likely γ -ray emitting counterparts and were visible fromthe GMRT (see the Appendix for details).Using the GMRT, we have performed a targeted radio pulsar survey of 375 unas-sociated γ -ray point sources detected by the Fermi -LAT. The survey was conductedwith observations at 322 MHz and 607 MHz. We aimed to observe the relatively highlatitude pointings at 322 MHz as the dispersion broadening and scattering contri-butions are comparatively lower for the target sky. For this we selected the sourcesavailable at a given time span either at 322 or at 607 MHz. In general, we observedeach target once with the GMRT either at 322 MHz or at 607 MHz. However, incase of marginal detection of possible millisecond pulsations we conduct confirmationobservations. In this effort, we have discovered four MSPs associated with the
Fermi -LAT γ -ray sources. One MSP, the black widow PSR J1544+4937, is the first GalacticMSP discovered by the GMRT and has been published elsewhere (Bhattacharyya etal. 2013). In Section 4 of this paper we present discovery details of the remaining Bhattacharyya et al.
PSR J0248+4230 W e i g h t e d C o un t s ( . - G e V ) PSR J1207 5050
322 MHz607 MHzLAT 0.1 GeV0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Pulse Phase
PSR J1536 4948 I n t e n s i t y Figure 1.
Radio and γ -ray pulse profiles for PSRs J0248+4230 (top), J1207 − − ◦ of the respective pulsar position with the blackdashed horizontal line giving the estimated background level (calculated as in Abdo et al.2013). The solid red lines on the right half of each plot show the results of the pulse profilefits described in the text. The green dashed curves show the 322 MHz radio profiles, whenavailable, and the blue dash-dot curves show the 607 MHz profile (both for highest signal-to-noise detection and using the right y-axis giving relative intensity in arbitrary units).The data spanning pulse phases 0 to 1 is duplicated over pulse phases from 1 to 2, to moreeasily show features occurring near a pulse phase of 1. three MSPs. We have also independently detected MSP J1446 − − − Fermi -LAT γ -ray sources. The distances of these pulsars from the pointingcentres of the Fermi -LAT sources are 57 (cid:48) , 10 (cid:48) and 26 (cid:48) respectively (Refer to Table2 of Roy & Bhattacharyya 2013). These serendipitous discoveries will be reportedin a follow up paper (Bhattacharyya et al. in preparation ). The details of our radioobservations for all 375
Fermi -LAT sources and corresponding 10 σ detection limit foreach source are also presented in the Appendix. OBSERVATIONS AND PULSATION SEARCH ANALYSIS iscovery and timing of three MSPs Figure 2.
The on-off gating image of PSR J0248+4230 using the coherently dedispersedMSP gating correlator. The MSP is localised with ± (cid:48)(cid:48) accuracy at 13 σ detection signifi-cance. Table 1.
Parameters of pulsars discovered in
Fermi directed survey with the GMRTPulsar P ˙ P DM S † S ‡ spectral index α (ms) (s/s) (pc cm − ) (mJy) (mJy)PSR J0248+4230 2.600 1.68 × − − − × − < ∗ > − × − − † Flux density at 322 MHz without primary beam correction. ‡ Flux density at 607 MHz without primary beam correction. ∗ σ non-detection limit at 322 MHz for 30-minutes of GMRT observations using 17antennas in phased array α The numbers in the parenthesis are uncertainties in preceding digit. Error on spectralindex is calculated considering a typical 10% uncertainty in the flux measurement.
Bhattacharyya et al.
Figure 3.
Radio timing residuals for PSR J0248+4230 from the GMRT observations at322 MHz (black points) and 607 MHz (red points) with bandwidth of 32 MHz using theGMRT legacy system.
The search observations were performed between 2010 November and 2013 Septem-ber with the GMRT Software Back-end (GSB, Roy et al. (2010)) producing simultane-ous incoherent and coherent beam filter-bank outputs of 512 × . µ s. Details of the observational configuration are described in Bhat-tacharyya et al. (2013). Positional uncertainty associated with the Fermi -LAT sourcescan easily be covered by the wider incoherent beam of the GMRT In addition to thewider incoherent beam, data from the more sensitive coherent beam were simulta-neously recorded with much narrower beam width ( ∼ ± (cid:48) at 322 MHz and ∼ ± (cid:48)(cid:48) at 607 MHz), which is useful in cases where the pulsar happens to be close enoughto the pointing center. However, this is not likely considering the typical positionaluncertainty ( ∼ ± (cid:48) ) of the Fermi -LAT sources.Using parameters of 32 MHz bandwidth, 10% duty cycle, incoherent array gain of2.3 K/Jy, for a 30-minute observing time, we estimate the search sensitivity usingradiometer equation (Lorimer et al. 2004) for a 5 σ detection as (66K + T sky )/(335K) iscovery and timing of three MSPs Figure 4.
Radio timing residuals for PSR J1207 − − mJy at 322 MHz and (92K + T sky )/(335K) mJy at 607 MHz, where 66K and 92K arethe receiver temperatures at the respective frequencies. Thus considering | b | > ◦ and T sky ∼ σ detectionis 0.3–0.9 mJy. Whereas, considering | b | > ◦ and T sky ∼ σ detection at 607 MHz is 0.3–0.4 mJy. The sky temperatureis estimated from the all-sky 408 MHz image by Haslam et al. (1982). This skytemperature is then scaled to the observing frequency using an assumed spectralindex of − − upto 350 pc cm − , which is the limiting DM for pulsars at | b | > ◦ up to distance of 8kpc (according to NE2001, Cordes & Lazio (2001)). Since we are observing at lowfrequencies we are only sensitive to nearby MSPs; at higher DMs the survey sensi- Bhattacharyya et al.
Figure 5.
Timing residuals for PSR J1536 − tivity decreases because of dispersive smearing within the channels. A linear drift ofup to 200 Fourier-frequency bins for the highest summed harmonic was allowed. Thepowerline, 50 Hz, and its subsequent harmonics were excised.The newly discovered MSPs can be localised in the image plane with the GMRTinterferometric array with an accuracy of better than ± (cid:48)(cid:48) (half of the typical syn-thesized beam used in the image made at 322 MHz) using gated imaging of pulsars(Roy & Bhattacharyya (2013), see Figure 2) and the multipixel beam former (Royet al. 2012) which is detailed in Section 5. Once the MSPs are localised in the imageplane, we use the coherent array for follow up observations with a smaller field ofview but with enhanced sensitivity. Using the coherent array with the central core ofthe GMRT having 17 antennas (i.e. gain of ∼ σ detection. After discovery we started a regular timing campaign at322 and 607 MHz over ∼ iscovery and timing of three MSPs Table 2.
Timing parameters of PSR J0248+4230, J1207 − − Name J0248+4230 J1207 − − ∗ Right ascension (J2000) . . . . . . . . . . . . 02 h m s (7) 12 h m s (5) 15 h m s (10)Declination (J2000) . . . . . . . . . . . . . . . . +42 ◦ (cid:48) (cid:48)(cid:48) (4) − ◦ (cid:48) (cid:48)(cid:48) (10) − ◦ (cid:48) (cid:48)(cid:48) (10)Parameters from radio timing ∗ Right ascension (J2000) . . . . . . . . . . . . 02 h m . s h m . s h m . s ◦ (cid:48) . (cid:48)(cid:48) − ◦ (cid:48) . (cid:48)(cid:48) − ◦ (cid:48) . (cid:48)(cid:48) − ). . . . − − − ) . . − − f (Hz). . . . . . . . . . . . . . . . . . . 384.49193525267(4) 206.493931730035(8) 324.68438438109(6)Frequency derivative ˙ f (Hz s − ) . . . . − × − − × − − × − Period epoch (MJD) . . . . . . . . . . . . . . . 56588.0 56478.0 56530.6Dispersion measure DM (pc cm − ) 48.2634(1) 50.67 38.00125(4)DM 1 s t derivative DM1 . . . . . . . . . . . . . − − − n d derivative DM2 . . . . . . . . . . . . − − − DDHOrbital period P b (days). . . . . . . . . . . . − − x (lt-s) − − T ASC (MJD) − − h ( µ s) . . . . − − − − − × − × − × − Total time span (yr) . . . . . . . . . . . . . . . 4.6 5.6 11.5Energy loss rate ˙ E (erg/s) . . . . . . . . . . 3.8 × × × ˙ E with kinematic corrections (erg/s) − × × Characteristic age (yr) . . . . . . . . . . . . . 2.4 × × × Surface magnetic field (Gauss). . . . . . 2.1 × × × DM distance (kpc) ‡ . . . . . . . . . . . . . . . 1.8 1.5 1.8DM distance (kpc) ‡† . . . . . . . . . . . . . . 2.5 1.3 0.98Companion mass M (cid:12) . . . . . . . . . . . . . . − − − ∗ Errors correspond to 1 σ . ‡ using the Cordes & Lazio (2001) model of electron distribution ‡† using the Yao et al. (2017) model of electron distributionWe note that the calculated DM distance is model dependent.Timing uses DE421 solar system ephemeris.The numbers in the parenthesis are uncertainties in preceding digits.4. DISCOVERY OF THREE MSPSPSR J0248+4230 was discovered in a 30 − minute observing run with the GMRT at322 MHz targeted at the γ -ray source 4FGL J0248.6+4230 (Abdollahi et al. 2020).Here, and throughout the paper, we refer to the 4FGL names of the associated LATsources even though our initial observations were targeted at sources from earliersource lists and catalogs. It is a 2.60 ms MSP with DM of 48.25 pc cm − having fluxdensity of 7.5 mJy at 322 MHz and 0.96 mJy at 607 MHz. We estimate a spectralindex of − Bhattacharyya et al.
Table 3. γ − ray resultsName PSR J0248+4230 PSR J1207 − − N (10 − cm − s − GeV − ) 0.4 ± ± ± ± ± ± E C (GeV) . . . . . . . . . . . . . . . . . 1.5 ± ± ± F (10 − cm − s − ) . . . . . . . . 1.1 ± ± ± G (10 − erg cm − s − ) . . . 1.6 ± ± ± L γ NE2001 (10 erg s − ) . . . . 6.2 ± ± ± L γ YMW2017 (10 erg s − ) . . 12 ± ± ± η γ NE2001 (%) . . . . . . . . . . . . . . 1.6 ± ± ± η γ YMW2017 (%) . . . . . . . . . . . . 3.2 ± ± ± φ P . . . . . . . . . . . . . . . . . . . . . . . 0.151 ± ± ± w P . . . . . . . . . . . . . . . . . . . . . . . 0.055 ± ± ± φ P . . . . . . . . . . . . . . . . . . . . . . . 0.830 ± ± ± w P . . . . . . . . . . . . . . . . . . . . . . . 0.085 ± ± ± φ P . . . . . . . . . . . . . . . . . . . . . . . − − ± w P . . . . . . . . . . . . . . . . . . . . . . . − − ± φ P . . . . . . . . . . . . . . . . . . . . . . . − − ± w P . . . . . . . . . . . . . . . . . . . . . . . − − ± ± ± ± φ B . . . . . . . . . . . . . . . . . . . . . . . − − ± w B . . . . . . . . . . . . . . . . . . . . . . . − − ± φ B . . . . . . . . . . . . . . . . . . . . . . . − − ± w B . . . . . . . . . . . . . . . . . . . . . . . − − ± γ -ray luminosity ( L γ ) and efficiency ( η γ ) values (see Abdo et al. 2013) are reported inrows 7 through 10 of the spectral fit results, with subscripts indicating which distanceestimate was used. For η γ , we use the kinematically corrected ˙ E values, when available.The uncertainties in rows 7 through 10 use the uncertainties on G only. The ∆ valuelisted in row 9 of the pulse profile fitting results is the difference in phase between the firstand last γ -ray peak. All uncertainties are statistical only. of this MSP profile (Figure 1). The leading component has a significantly higheramplitude than the following component. Due to dispersion smearing, the profilecomponents are not well resolved at 322 MHz.PSR J1207 − − minute observing run with the GMRT at607 MHz, targeted at the γ -ray source 4FGL J1207.4 − − , having flux density of 0.5 mJy iscovery and timing of three MSPs σ non-detection limit at 322 MHz is 0.38 mJy for 30 − minute observing timeusing 17 antennas in phased array. Considering this limiting flux density we estimatespectral index of ∼ − − minute observing run with the GMRT at322 MHz, targeted at the γ -ray source 4FGL J1536.4 − − having flux density of 12 mJyat 322 MHz. We have also detected this MSP at 607 MHz having a flux density 4mJy. We estimate a spectral index of − > ◦ ) with 3 components, but due to dispersionsmearing ( ∼
20% of pulse period) the profile components are not well resolved at 322MHz. Table 1 summarizes the discovery parameters of these three MSPs. LOCALIZATION OF THE MSPSFollowing the discoveries with the GMRT incoherent array (half power beam width ∼ (cid:48) for 322 MHz, 40 (cid:48) for 607 MHz), using the techniques of multi-pixel beam-forming (Roy et al. 2012) and MSP gating correlator (Roy & Bhattacharyya 2013),we could significantly improve the large positional uncertainties allowing us to usethe sensitive coherent array (4 to 5 times more than incoherent array for the GMRT)for the follow-up timing observations which substantially reduces the use of arraytelescope time (16 to 25 times for the GMRT).Since PSR J0248+4230 is a relatively weak MSP, we were not expecting to get sig-nificant signal-to-noise in the continuum image plane. Because the radio pulse profilesMSPs are dispersion broadened, specially for this MSP with wide emission compo-nents, we used a coherently dedispersed gating correlator, with proper optimisation,when selecting average on and off visibility phase bins. This MSP is unambiguouslydetected in a 30 − minute observing run with 13 σ detection significance in an on − offgated image (Figure 2). The pulsar is found as the only point source in the imagewith the precise position being 2 h m s (7), +42 ◦ (cid:48) (cid:48)(cid:48) (4).Roy & Bhattacharyya (2013) reported a precise position for PSR J1207 − − Bhattacharyya et al. timing observations with the coherent array mode of the GMRT detailed in Section6. TIMING STUDYFollowing the precise astrometric localization of these three MSPs described in Sec-tion 5, we conducted dense observations using the GMRT coherent array to time theMSPs. This timing campaign allowed us to construct timing models that describewell the pulse times of arrival (TOAs). These timing models can be extended tonearby epochs, and allow us to start observing the MSPs more sparsely. We usedthe highest signal − to − noise ratio profiles as templates for extracting TOAs. ForPSR J0248+4230 and J1536 − − − − − , which uses the tempo timing software to explore all thecombinations of rotation numbers between unconnected observations that result intiming solutions with a low residual χ . The algorithm soon determined the correctset of rotation numbers between all observations, which yields the timing solution.The timing solutions from ∼ tempo2 , are presented in Table 2. The timingresiduals (the observed TOAs minus the prediction of the model for the TOAs) aredisplayed in Fig. 3, 4 and 5, for PSR J0248+4230, J1207 − − µ s respectively.For timing of PSR J1536 − γ -rays (see section 7.3). The residuals showed a small secular driftbetween the radio and γ -ray TOAs, caused by long-term change in DM, which is dueto the relative motion of the pulsar and the Earth changing the column density ofionized gas between both. This variation of the DM is confirmed by comparison ofthe radio TOAs at frequencies of 322 and 607 MHz, and can be modeled well withtwo DM derivatives, which are also listed in Table 2. https://github.com/pfreire163/Dracula http://tempo.sourceforge.net/ iscovery and timing of three MSPs − − (cid:12) , which, assuming a pulsar mass of 1.4 M (cid:12) andorbital inclinations of 90, 60 and 25 ◦ would yield companion masses of 0.27, 0.32and 0.74 M (cid:12) respectively. We have not seen any evidence of eclipsing from this widebinary, suggesting that the companion is a Helium white dwarf (WD). For the orbitalperiod of this system, the expectation of the Tauris & Savonije (1999) model is aHelium WD mass of ∼ . M (cid:12) ; which suggests that the orbital inclination is close tothe median of 60 ◦ .The DDH fit provides a weak (2 σ ) detection of the orthometric amplitude of theShapiro delay ( h ). In the absence of other post-Keplerian parameters, this is notenough for a determination of the mass of the pulsar or the mass of the companion,nor of the orbital inclination. We also note that the current timing precision (withthe observations using the GMRT Software Back-end having 32 MHz bandwidth) isnot sufficient to make reliable estimation of the Shapiro delay. Ongoing observationswith the upgraded GMRT wide band system (Gupta et al. 2017) will allow to increasethe timing span with more precise TOAs for better estimation of Shapiro delay.Pulsar distance estimates come from comparing observed DM with the Galacticelectron density n e ( (cid:126)x ) integrated along the line of sight (LoS) to the pulsar, usingthe models for n e ( (cid:126)x ) provided by Yao et al. (2017) and Cordes & Lazio (2001).Comparing the DM distances with those obtained by other methods suggests thatfor most pulsars, the uncertainty is roughly gaussian, with a standard deviation near30%. Unfortunately, the difference distribution has very broad tails: for some pulsarsthe disagreement is a factor of a few to several. The tools described in Theureau etal. (2011) and Hou et al. (2014) attempt to identify the aberrant cases by comparingthe models with HI, CO, and H α observations.We examined the LoS to our three pulsars to see if they might crossunmodeled electron over- or under-densities. Nothing is unusual for PSRsJ0248+4230 and J1536 − − ∼
20 pc cm − step could easily be off by a factorof two, which could shift the distance by ± a few hundred pc. Next is that Yao et al.(2017) highlight PSR J1227 − − − ∼
10 pc cm − step due to the edge of the Local Bubble. But the 600 pcdiscrepancy appears to not exist: Jennings et al. (2018) show that the Gaia parallaxmeasurements of PSR J1227–4853’s optical companion give a distance matching both4 Bhattacharyya et al.
DM distances. This would bolster confidence in PSR J1207 − H α maps of Finkbeiner (2003) show that PSR J1207 − H α glow, presumably created by the stars and thus well in thepulsar’s foreground. Unmodeled extra electrons would mean that the pulsar is closerthan predicted by the electron models. This does not seem to be the case, becausethe ratio of both observed H α intensity and calculated emission measure at the twopulsar positions is the same. We conclude that the DM distance to PSR J1207 − GAMMA-RAY DETECTIONTo confirm identification of the newly detected radio MSPs with the correspondingunassociated LAT sources, we need to detect significant pulsations at the spin periodin the γ -ray data. We therefore performed spectral and timing analyses of the LATdata, using the radio timing solutions, as described below.7.1. LAT data preparation
We analyzed LAT Pass 8 data (Atwood et al. 2013; Bruel et al. 2018) within 15 ◦ ofthe best-fit radio timing position of each MSP, separately, starting from the beginningof the mission, 2008 August 4, and ending 2020 April 27. We kept all events withreconstructed energies from 0.05 to 500 GeV, zenith angles less than 90 ◦ , and be-longing to the SOURCE event class. We filtered the data to create good time intervalswhen the spacecraft was in nominal science operations mode, the data were flaggedas good, and to avoid LAT-detected solar flares and gamma-ray bursts.7.2.
Gamma-ray spectral analysis and results
We created spatial and spectral models of the regions around each MSP using the
Fermi -LAT Fourth source catalog (4FGL, Abdollahi et al. 2020), including all sourceswithin 25 ◦ of the pulsar and the corresponding diffuse emission components. For allthree MSPs, the position of the associated 4FGL source was ≤ . (cid:48) from the timingposition, consistent within the 4FGL positional uncertainty. We chose to move the4FGL source associated with each MSP (4FGL J0248.6+4230, 4FGL J1207.4 − − γ -ray spectrum of eachsource associated with one of our MSPs was modeled using an exponentially cutoffpower-law shape of the form described in Eq. 1, observed to describe the spectra ofmost γ -ray pulsars well (Abdo et al. 2013). dNdE = N (cid:16) EE (cid:17) − Γ exp (cid:110) − (cid:16) EE C (cid:17) b (cid:111) (1)In Eq. 1, N is a normalization parameter with units of GeV − cm − s − and is cal-culated from the 4FGL information to be the value of the differential counts spectrum iscovery and timing of three MSPs E , Γ is the low-energy photon index, E C is the cutoff energy, and b is an exponential index controlling how quickly the spectrum cuts off. We chose tofix b to a value of 1, but did explore other values, as discussed later in this section.For each MSP, we performed a binned maximum likelihood fit, using the P8R3 SOURCE V2 instrument response functions , in which we allowed the spectralparameters to vary for all sources within 6 ◦ of the pulsar that were found to have anaverage significance of ≥ σ in the 4FGL catalog. The spectral parameters of theGalactic and isotropic diffuse emission components were also allowed to vary in thefits. For sources not meeting the previous criteria which were flagged as significantlyvariable in the 4FGL catalog, we allowed their spectral normalizations to be free inthe fits if they were within 8 ◦ of the corresponding pulsar position. The spectralanalysis was done over the energy range of 0.1 to 300 GeV but the exposure productswere calculated over the entire energy range of our data, with 10 bins per decade, toallow for the use of energy dispersion .After an initial fit, we examined the spatial residuals to determine if the spectralparameters of any sources we had fixed to the 4FGL values needed to be allowed tovary and if there was evidence for new sources not in the 4FGL catalog. In doing so,for the region around PSR J1536 − − . ◦ for known γ -ray pulsars and fixes the b parameter to a valueof 2/3 based on what is observed in the spectra of the brightest γ -ray pulsars. In ourfitting, we found that the Γ parameter for PSR J0248+4230 was unstable and oftenfit to ≈
0. When we instead set b to a value of 1, the Γ parameter was more well-behaved, but we found a strong dependence on the starting value of other parametersand chose to instead switch to the formulation given in Eq. 1. To be consistent, weused this spectral shape for all three MSPs.We performed likelihood analysis with the b parameter free, as well as fixed to thevalue of 2/3 used in the 4FGL catalog. For PSRs J0248+4230 and J1207 − b = 2 / b free, when compared tofits with b = 1. For PSR J1536 − b , but the final result was very dependent on the starting value of E C , possibly dueto issues with modeling the diffuse emission in this region, so we chose to use andreport only the b = 1 results.The resulting best-fit spectral parameters for each MSP are given in Table 3 aswell as the derived integral photon and energy flux values, F and G , respectively.Our best-fit parameters are not directly comparable to those reported in the 4FGL See https://fermi.gsfc.nasa.gov/ssc/data/analysis/documentation/Pass8 edisp usage.html. See the entry for the
PLSuperExpCutoff2 model at https://fermi.gsfc.nasa.gov/ssc/data/analysis/scitools/source models.html. Bhattacharyya et al. catalog, due to the differences in functional form, for the sources associated with theseMSPs. However, our fits with the same spectral model were and the results in Table3 yield compatible values of F and G .7.3. Gamma-ray pulsation detection and timing
Using the best-fit models of the regions, we selected events within 3 ◦ of each MSPwith energies from 0.1 to 300 GeV and calculated spectral weights representing theprobability that each event came from the pulsar of interest. Use of these weightswith the H test (de Jager et al. 1989; de Jager & B¨usching 2010) has been shown toenhance the sensitivity of searches for γ -ray pulsations in LAT data (Kerr 2011). Theresulting weighted H-test values resulted in significant detections of γ -ray pulsationsfrom all three pulsars, with values of 60.4 (6 . σ ), 188.8 (12 . σ ), and 3780.4 (60.4 σ )for PSRs J0248+4230, J1207 − − − ◦ radius selection to construct γ -ray TOAs which werecombined with the radio TOAs (as discussed in Section 6) to produce an improvedtiming model, yielding a weighted H-test result of 4644.5 (67.1 σ ), which indicates asignificant improvement; this removed the drift in phase seen in the early LAT data.The resulting γ -ray pulse profiles, as well as the radio profiles, are shown in Figure 1.7.4. Gamma-ray pulse profile characterization
Once we had the final timing solutions , we fit the γ -ray pulse profiles using themaximum likelihood method described in Abdo et al. (2013) but restricted our modelto Gaussian functional forms for the peaks. The resulting fits are shown on the righthand side of each panel in Figure 1 as the solid red line. The fit values are given inTable 3 where we have labeled the peaks in the order in which they appear in phase.For PSRs J0248+4230 and J1207 − − ∼
45% of the total emission.Although these results depend on accurate photon weights, the quality of the spectralfit is good, and the excess emission is present even at higher energies where the LATis more capable of distinguishing sources from backgrounds. This component thuslikely represents bona fide, nearly constant emission from the magnetosphere. Thismeasurement comes directly from the maximum likelihood fit to the photon weights,but is in good agreement with the background level drawn in Figure 1, which is anempirical estimator following the prescription in Section 5.1 of Abdo et al. (2013). These timing solutions will be made available at https://fermi.gsfc.nasa.gov/ssc/data/access/lat/ephems/. iscovery and timing of three MSPs E [3.0, 9.0) GeV
PSR J0248+4230 W e i g h t e d C o un t s E [1.0, 3.0) GeV
Pulse Phase
E [0.1, 1.0) GeV E n e r g y ( G e V ) Figure 6.
The γ -ray pulse profile of PSR J0248+4230 in multiple energy bands, as indicatedin the plot. The dashed horizontal lines in each panel indicate the estimated background(derived as in Abdo et al. 2013). In the top panel, we show only one, representative error bar,the rest are smaller than this by as much as 40%. The scatter plot in the top panel showsthe phases and energies, right y-axis, of individual events with marker sizes proportional tothe spectral weight values. Fitting the γ -ray pulse profile for PSR J1536 − ∼
7% of the pulsar emission,but given the numerous peaks spanning most of the pulse phase, it is possible thismight just reflect wings/tails of the peaks which are not entirely fit with Gaussians.7.5.
Pulse Profile Energy Evolution
Similar to what is observed at radio wavelengths, γ -ray pulse profiles can showinteresting evolution when the data are split into smaller energy bands. Figures 6, 7,and 8 show the γ -ray pulse profiles in different energy bands for PSRs J0248+4230,J1207 − − Bhattacharyya et al.
E [5.0, 12.5) GeV
PSR J1207-5050 W e i g h t e d C o un t s E [1.0, 5.0) GeV
Pulse Phase
E [0.1, 1.0) GeV E n e r g y ( G e V ) Figure 7.
The γ -ray pulse profile of PSR J1207 − pulsed photon energy, we searched through the events within 3 ◦ of the radio position,requiring the spectral weight to be ≥ − E C near 1 GeV in Table 3. PSR J1536 − γ -ray pulsations from the Vela and Crab pulsars above100 GeV and even out to TeV energies (e.g., Aliu et al. 2011; Ansoldi et al. 2016; iscovery and timing of three MSPs E [12.5, 60.0) GeV
PSR J1536-4948
E [5.0, 12.5) GeV W e i g h t e d C o un t s E [1.0, 5.0) GeV
Pulse Phase
E [0.1, 1.0) GeV E n e r g y ( G e V ) Figure 8.
The γ -ray pulse profile of PSR J1536 − Abdalla et al. 2018), there has been growing interest in determining which other γ -raypulsars might have detectable pulsations at energies out to 100 GeV or beyond (e.g.,Saz Parkinson et al. 2017). In order to assess the possibility for detection of PSRsJ0248+4230, J1207 − − Bhattacharyya et al. observed feature in the pulse profile. If we constrain our search to the pulse profilepeaks (phases φ ∈ [0 . , . ∪ [0 . , . − − γ -ray pulse profile, it does havea low weight. This could be due, in part, to the choice of spectral model. The nexthighest energy event occurs at phase 0.468, within the second tallest peak, and has anenergy of 46.1 GeV and a spectral weight of 0.757. Events in the highest energy bandhave spectral weights ranging from 0.107 to 0.996. Taking events corresponding tothe top panel of Figure 8, we calculate a weighted H test value of 220, correspondingto a detection of 13 σ .To further quantify evidence for pulsed emission above 10 GeV, we performed ananalysis similar to that reported in Ackermann et al. (2013) and Saz Parkinson etal. (2017). For each pulsar, we selected all events with energies from 1 to 10 GeVand with spectral weights ≥ ≥ . ◦ . ◦ − − × − forevents above 10 GeV, suggesting significant pulsed emission above this energy. Whenapplying the same analysis to events above 25 GeV for this pulsar, the likelihood testyields a p-value of 0.02, suggesting there is significant pulsed emission even above 25GeV. These results are in agreement with those of Saz Parkinson et al. (2017), whoclaimed evidence for pulsed γ -ray emission above 25 GeV for five MSPs, includingPSR J1536 − Fermi -LAT sources (3FHL), Ajello et al. (2017) anal-ysed 7 years of LAT data above 10 GeV looking for hard sources and associated thispulsar with the source 3FHL J1536.3 − iscovery and timing of three MSPs
21a long-term timing solution had not yet been constructed. Fitting data from 10 to300 GeV, modeling the spectrum PSR J1536 − ± × − cm − s − , energy flux of (6.5 ± × − erg cm − s − , and photon index of 3.5 ± − − DISCUSSIONAn important, derived quantity for understanding high-energy pulsar emission isthe γ -ray luminosity L γ = 4 πd f Ω G , where f Ω is a beaming factor typically assumedto be near 1 (Abdo et al. 2013). From this luminosity, we can calculate the efficiencywith which rotational energy is turned into γ -rays as η γ = L γ / ˙ E . Table 3 reportsvalues for L γ and η γ using both the distance estimate from Cordes & Lazio (2001)and Yao et al. (2017), and the values of ˙ E with kinematic corrections, when available.Known γ -ray MSPs span a large range in η γ (Abdo et al. 2013), from 1% to > γ -ray Pulsars(Abdo et al. 2013), only three MSPs have a higher L γ , using the NE2001 distance,than PSR J1536 − − − η γ < − − η γ > f Ω less than 1, that the distances usedare overestimated, or that the “standard” values of neutron star mass and radius usedmay not be representative (e.g., using a radius of 14 km instead of 10 km will increase˙ E by a factor of 2).The prevailing models of γ -ray emission from rotation-powered pulsars posit thatparticles are accelerated along magnetic field lines near, or beyond, the light cylinderradius ( c P/2 π , where co-rotation with the star requires moving at the speed of light c ). Particle acceleration may happen within the light cylinder in relatively narrowvacuum gaps above the last open field line (e.g., Cheng et al. 1986; Muslimov &Harding 2004) or over the full open volume above the polar cap (e.g., Harding etal. 2005). Alternatively, the emission may originate outside the light cylinder in astriped wind (e.g., Kirk et al. 2002) or in regions near an equatorial current sheet(e.g., Kalapotharakos et al. 2014). Features in the predicted pulse profiles depend onthe assumed structure of the magnetosphere used in each model. Thus, testing themodels is one way to better understand the complex magnetic fields of neutron stars.Fitting the observed γ -ray pulse profiles using one of these models is one methodof estimating the viewing geometry of the system, namely the inclination angle ofthe magnetic axis and the observer viewing angle, both relative to the spin axis (e.g.,Johnson et al. 2014; Chang & Zhang 2019). When combined with profiles at otherwavelengths or geometry constraints from different methods, these fitting methods2 Bhattacharyya et al. can be useful tests of the different emission models. The pulse profiles of PSRsJ0248+4232, J1207 − − γ -ray pulsars, the main radio peak is recorded at an earlier phase than thefirst γ -ray peak. A small fraction of MSP γ -ray pulse profiles, however, are observedto have their first peak occur before the radio. This is predicted by models in whichthe full open volume above the polar cap is available for particle acceleration (Hardinget al. 2005). The γ -ray profile for PSR J0248+4232 might fall into this category ifwe consider what we have called the second peak in Table 3, based on the order theyappear with our choice of phasing, to be the first peak. However, the models thatpredict that the γ -ray peak should precede the radio also predict much broader peaksand have difficulty matching the relatively sharp peaks we observe (Venter et al. 2009;Johnson et al. 2014).The γ -ray pulse profile of PSR J1207 − γ -ray pulse profile of PSR J1536 − γ -ray emission out to >
25 GeV in Figure 8. Harding etal. (2018) have modeled the spectrum and pulse profile of the Vela pulsar out to 100TeV. In their model, the GeV emission is curvature radiation and the highest energyphotons are produced via inverse Compton interactions between accelerated particlesand infrared-optical photons. Harding et al. (2018) use the maximum Lorentz factorof accelerated particles necessary to argue against models in which GeV γ -rays arethe result of synchrotron radiation. While PSR J1536 − SUMMARYWe report the GMRT discoveries of three MSPs, PSR J0248+4230, PSRJ1207 − − Fermi -LATsources, 4FGL J0248.6+4230, 4FGL J1207.4-5050 and 4FGL J1536.4-4948 respec-tively. Considering the discovery of four MSPs with γ -ray associations, (one MSP, iscovery and timing of three MSPs Fermi directed targeted survey with the GMRTis ∼ ∼ two orders of magnitude better accuracy than the discovery po-sition associated with the Fermi error boxes. These precise positions allowed us toconduct sensitive, follow-up timing observations in phased array mode at 322 and 607MHz while optimising telescope time usage. PSR J0248+4230 and PSR J1536 − − − − −
607 MHz. A detailed spectralstudy of this MSP with the upgraded GMRT wide band system is in progress.We have presented phase-connected timing models for each MSP from ∼ ∼ γ -ray TOAs for PSR J1536 − − − ∼
62 days and companion mass of ∼ (cid:12) foran inclination of 60 ◦ . We report a weak (2 σ ) detection of the orthometric amplitudeof the Shapiro delay ( h ), which is not enough to determine the mass of the pulsaror mass of the companion in absence of other post-Keplerian parameters. Ongoingcoherently dedispersed observations of these MSPs using the ugpraded GMRT willallow us to reduce the TOA uncertainities and will enable better constrains on thebinary parameters. This may lead to possible determination of Shapiro delay ( h ) forPSR J1536 − γ -ray pulsations from these threeMSPs, which confirms that the pulsars are the engines powering the previously unas-sociated γ -ray sources. For some of the relatively weak γ -ray sources associated radiopulsars are relatively bright, indicating that radio flux is uncorrelated with the γ -rayflux and even faint new LAT sources can harbor bright radio MSPs. Such detectionsprovide strong justification to continue radio observations as new unassociated LATsources are revealed in analysis of longer data sets.Ongoing radio polarimetric studies of these MSPs will be helpful to probe the possi-ble emission geometry enabling further constraints on possible models explaining theobserved radio and γ -ray emission. Profile modeling will also be aided by ongoinginvestigation of profile evolution of these MSPs for wider radio frequency range with4 Bhattacharyya et al. the upgraded GMRT. The ongoing timing observations with the upgraded GMRTwill reveal the prospect of using these MSPs in the pulsar timing array which will bereported in a future publication. To conclude, in this paper we present the discov-ery of three radio MSPs with the GMRT in
Fermi directed targeted searches. Thediscovery was followed by long term radio timing and subsequent discovery of γ -raypulsations. We also present a study of phase aligned radio, γ -ray profiles of theseMSPs. In addition, we provide a list of target pointings and the detection limits forthe Fermi -LAT point sources that were observed with the GMRT, which will help toplan future observations for these sources. iscovery and timing of three MSPs
Fermi directed searches with the GMRT during between 2010 Novem-ber and 2013 September as part of an effort coordinated by the
Fermi
PSC. Based onseveral criteria such as the γ -ray spectral index, the amount of variability seen, thesignificance of detection etc, the PSC has rank-ordered the unassociated γ -ray sourcesaccording to the probability of them being pulsars. Out of these we considered sourceswith | b | > ◦ to limit the effects of scatter broadening. In addition we choose a greaterfraction of the sources in the declination range − ◦ to − ◦ , which is outside the skycoverage of other active PSC searches (e.g. GBT, Effelsberg). Since the LAT pointsource catalogs evolved during the span of 2011 − − γ -ray sources and many promising high and mid Galactic latitudesources were still left to be searched for millisecond pulsations.Table A-1 present details of the GMRT observations for all 375 Fermi -LAT sourcesin this survey. This table includes the observing epoch, frequency and duration.Additionally, to guide planning of future follow up observations, we have included a10 σ detection limit for each source, calculated using the radiometer equation (Lorimeret al. 2004) with the GMRT ETC calculator . Table A-1 . Summary of the GMRT observations † : 10 σ detection threshold calculated with the GMRT ETC Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J2323-4919 23 h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .28 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 +66 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .9 +60 ◦ (cid:48) (cid:48)(cid:48) h m s .4 +09 ◦ (cid:48) (cid:48)(cid:48) h m s .38 +63 ◦ (cid:48) (cid:48)(cid:48) Bhattacharyya et al.
Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J0515.6-4404 05 h m s .5 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 +22 ◦ (cid:48) . (cid:48)(cid:48) h m s .35 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s .93 − ◦ (cid:48) (cid:48)(cid:48) h m s .12 − ◦ (cid:48) (cid:48)(cid:48) h m s +69 ◦ (cid:48) (cid:48)(cid:48) h m s .2 − ◦ (cid:48) (cid:48)(cid:48) h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s .28 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 +66 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .9 +60 ◦ (cid:48) (cid:48)(cid:48) h m s .38 +63 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 +22 ◦ (cid:48) (cid:48)(cid:48) h m s .5 − ◦ (cid:48) (cid:48)(cid:48) h m s .35 − ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s .12 − ◦ (cid:48) (cid:48)(cid:48) h m s .01 − ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .2 − ◦ (cid:48) (cid:48)(cid:48) h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s .28 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .31 +74 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s +69 ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s .93 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .01 − ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s .28 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) iscovery and timing of three MSPs Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J0515-4404 05 h m s .5 − ◦ (cid:48) (cid:48)(cid:48) h m s .9 +60 ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s .12 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s .29 − ◦ (cid:48) (cid:48)(cid:48) h m s .86 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .93 − ◦ (cid:48) (cid:48)(cid:48) h m s .71 +49 ◦ (cid:48) (cid:48)(cid:48) h m s .92 − ◦ (cid:48) (cid:48)(cid:48) h m s .7 − ◦ (cid:48) (cid:48)(cid:48) h m s .6 − ◦ (cid:48) (cid:48)(cid:48) h m s .6 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .9 − ◦ (cid:48) (cid:48)(cid:48) h m s .9 +60 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .92 − ◦ (cid:48) (cid:48)(cid:48) h m s .9 − ◦ (cid:48) (cid:48)(cid:48) h m s .2 +10 ◦ (cid:48) (cid:48)(cid:48) h m s .29 − ◦ (cid:48) (cid:48)(cid:48) h m s .86 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 − ◦ (cid:48) (cid:48)(cid:48) h m s .84 − ◦ (cid:48) (cid:48)(cid:48) h m s .01 − ◦ (cid:48) (cid:48)(cid:48) h m s .08 − ◦ (cid:48) (cid:48)(cid:48) h m s .3 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +32 ◦ (cid:48) (cid:48)(cid:48) h m s .9 +60 ◦ (cid:48) (cid:48)(cid:48) h m s .35 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .90 +04 ◦ (cid:48) (cid:48)(cid:48) h m s .10 − ◦ (cid:48) (cid:48)(cid:48) h m s .93 − ◦ (cid:48) (cid:48)(cid:48) h m s .3 +23 ◦ (cid:48) (cid:48)(cid:48) h m s .1 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) Bhattacharyya et al.
Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J1726-0724 17 h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .71 +49 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .40 +75 ◦ (cid:48) (cid:48)(cid:48) h m s +78 ◦ (cid:48) (cid:48)(cid:48) h m s +61 ◦ (cid:48) (cid:48)(cid:48) h m s .47 +67 ◦ (cid:48) (cid:48)(cid:48) h m s +40 ◦ (cid:48) (cid:48)(cid:48) h m s +33 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .6 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .30 − ◦ (cid:48) (cid:48)(cid:48) h m s .90 − ◦ (cid:48) (cid:48)(cid:48) h m s .84s − ◦ (cid:48) (cid:48)(cid:48) h m s .39 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .80 00 ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .82 − ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .98 − ◦ (cid:48) (cid:48)(cid:48) h m s .53 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .90 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .43 − ◦ (cid:48) (cid:48)(cid:48) h m s .41 − ◦ (cid:48) (cid:48)(cid:48) h m s .52 − ◦ (cid:48) (cid:48)(cid:48) h m s .71 +49 ◦ (cid:48) (cid:48)(cid:48) h m s .70 +32 ◦ (cid:48) (cid:48)(cid:48) h m s +35 ◦ (cid:48) (cid:48)(cid:48) h m s .4 +37 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +27 ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) iscovery and timing of three MSPs Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J1725-0509 17 h m s .30 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .29 +07 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .94 +16 ◦ (cid:48) (cid:48)(cid:48) h m s .29 +07 ◦ (cid:48) (cid:48)(cid:48) h m s +10 ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .52 − ◦ (cid:48) (cid:48)(cid:48) h m s .19 +58 ◦ (cid:48) (cid:48)(cid:48) h m s .79 +63 ◦ (cid:48) (cid:48)(cid:48) h m s .39 +67 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +35 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +02 ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +37 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +53 ◦ (cid:48) (cid:48)(cid:48) h m s .99 +00 ◦ (cid:48) (cid:48)(cid:48) h m s .99 +25 ◦ (cid:48) (cid:48)(cid:48) h m s +75 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +54 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +04 ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .80 +29 ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .39 − ◦ (cid:48) (cid:48)(cid:48) h m s .59 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) Bhattacharyya et al.
Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J1704-4618 17 h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .59 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .19 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 +29 ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +27 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +32 ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 +23 ◦ (cid:48) (cid:48)(cid:48) h m s .07 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 +06 ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 +00 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +43 ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .07 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +01 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +43 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +63 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +66 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +52 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .36 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .79 − ◦ (cid:48) (cid:48)(cid:48) h m s .75 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +43 ◦ (cid:48) (cid:48)(cid:48) iscovery and timing of three MSPs Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) J1306-4028 13 h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +37 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +52 ◦ (cid:48) (cid:48)(cid:48) h m s .19 +58 ◦ (cid:48) (cid:48)(cid:48) h m s .80 +39 ◦ (cid:48) (cid:48)(cid:48) h m s +75 ◦ (cid:48) (cid:48)(cid:48) h m s .99 +25 ◦ (cid:48) (cid:48)(cid:48) h m s .19 +68 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +56 ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .19 +68 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +41 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +63 ◦ (cid:48) (cid:48)(cid:48) h m s +69 ◦ (cid:48) (cid:48)(cid:48) h m s +68 ◦ (cid:48) (cid:48)(cid:48) h m s +61 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +63 ◦ (cid:48) (cid:48)(cid:48) h m s .94 +35 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +42 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +85 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +32 ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s +20 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +37 ◦ (cid:48) (cid:48)(cid:48) h m s .80 +17 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +54 ◦ (cid:48) (cid:48)(cid:48) h m s +14 ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .76 +57 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +31 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +46 ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 +6 ◦ (cid:48) (cid:48)(cid:48) Bhattacharyya et al.
Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) P73Y2387 15 h m s .80 +49 ◦ (cid:48) (cid:48)(cid:48) h m s .80 +76 ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .22 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +28 ◦ (cid:48) (cid:48)(cid:48) h m s +66 ◦ (cid:48) (cid:48)(cid:48) h m s .00 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 ++11 ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 +13 ◦ (cid:48) (cid:48)(cid:48) h m s .80 +6 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +38 ◦ (cid:48) (cid:48)(cid:48) h m s +66 ◦ (cid:48) (cid:48)(cid:48) h m s +69 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +3 ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s − ◦ (cid:48) (cid:48)(cid:48) h m s +17 ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) h m s .20 +70 ◦ (cid:48) (cid:48)(cid:48) h m s .80 +36 ◦ (cid:48) (cid:48)(cid:48) h m s .20 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 +42 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +2 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +42d29 (cid:48) (cid:48)(cid:48) h m s ++63 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +38 ◦ (cid:48) (cid:48)(cid:48) h m s .60 +6 ◦ (cid:48) (cid:48)(cid:48) h m s +24 ◦ (cid:48) (cid:48)(cid:48) h m s .40 +37 ◦ (cid:48) (cid:48)(cid:48) h m s .80 − ◦ (cid:48) (cid:48)(cid:48) h m s .60 − ◦ (cid:48) (cid:48)(cid:48) iscovery and timing of three MSPs Source Right ascension Declination MJD Frequency Duration S min † (MHz) (min) (mJy) P73Y0411 2 h m s .40 +42 ◦ (cid:48) (cid:48)(cid:48) h m s +51 ◦ (cid:48) (cid:48)(cid:48) h m s +42 ◦ (cid:48) (cid:48)(cid:48) h m s .20 +49 ◦ (cid:48) (cid:48)(cid:48) h m s .35 − ◦ (cid:48) (cid:48)(cid:48) h m s .40 − ◦ (cid:48) (cid:48)(cid:48) h m s .16 − ◦ (cid:48) (cid:48)(cid:48) h m s .11 − ◦ (cid:48) (cid:48)(cid:48) h m s .15 − ◦ (cid:48) (cid:48)(cid:48) h m s .73 +40 ◦ (cid:48) (cid:48)(cid:48) h m s .74 − ◦ (cid:48) (cid:48)(cid:48) h m s .49 − ◦ (cid:48) (cid:48)(cid:48) h m s .8 +20 ◦ (cid:48) (cid:48)(cid:48) h m s .44 − ◦ (cid:48) (cid:48)(cid:48) h m s .80 +03 ◦ (cid:48) (cid:48)(cid:48) h m s .74 +06 ◦ (cid:48) (cid:48)(cid:48) h m s .99 − ◦ (cid:48) (cid:48)(cid:48) h m s .52 − ◦ (cid:48) (cid:48)(cid:48) h m s .28 − ◦ (cid:48) (cid:48)(cid:48) h m s .72 − ◦ (cid:48) (cid:48)(cid:48) ACKNOWLEDGMENTSWe acknowledge support of the Department of Atomic Energy, Government of In-dia, under project no.12-R&D-TFR-5.02-0700. The GMRT is run by the NationalCentre for Radio Astrophysics of the Tata Institute of Fundamental Research, India.We acknowledge the support of GMRT telescope operators for observations. We alsoacknowledge the generous support of the HPC systems of IUCAA and NCRA. The
Fermi -LAT Collaboration acknowledges generous ongoing support from a number ofagencies and institutes that have supported both the development and the operationof the LAT as well as scientific data analysis. These include the National Aeronauticsand Space Administration and the Department of Energy in the United States, theCommissariat `a l’Energie Atomique and the Centre National de la Recherche Scien-tifique / Institut National de Physique Nucl´eaire et de Physique des Particules inFrance, the Agenzia Spaziale Italiana and the Istituto Nazionale di Fisica Nucleare inItaly, the Ministry of Education, Culture, Sports, Science and Technology (MEXT),High Energy Accelerator Research Organization (KEK) and Japan Aerospace Explo-ration Agency (JAXA) in Japan, and the K. A. Wallenberg Foundation, the SwedishResearch Council and the Swedish National Space Board in Sweden. This work per-formed in part under DOE Contract DE-AC02-76SF00515.4
Bhattacharyya et al.
Additional support for science analysis during the operations phase is gratefullyacknowledged from the Istituto Nazionale di Astrofisica in Italy and the Centre Na-tional d’ ´Etudes Spatiales in France. The National Radio Astronomy Observatory isa facility of the National Science Foundation operated under cooperative agreementby Associated Universities, Inc. PCCF gratefully acknowledges continuing supportfrom the Max Planck Society. SMR is a CIFAR Fellow and is supported by the NSFPhysics Frontiers Center award 1430284. We thank Ismael Cognard, Philippe Brueland Gu¨olaugur J´ohannesson for their comments. BB acknowledges the commentsfrom Dale Frail and Dave Thompson.
Fermi -LAT work at NRL is supported by NASA.
Facilities:
GMRT, Fermi LAT
Software:
GMRT Software Backend(Roy et al. 2010), presto(Ransomet al. 2002), Tempo(Nice et al. 2015), Tempo2(Edwards et al. 2006),Dracula(Freire & Ridolfi 2018)
REFERENCES
Abdalla, H., Aharonian, F., Ait Benkhaliet al. 2018, A&A, 620, A66.Abdo, A. A., Ackermann, M., Ajello, M.et al. 2010, ApJS, 187, 460.Abdo, A. A., Ackermann, M., Ajello, M.et al. 2009, Science, 325, 840.Abdo, A. A., Ackermann, M., Ajello, M.et al. 2009, Science, 325, 848.Abdo, A. A., Ajello, M., Allafort, A., etal. 2013, ApJS, 208, 17.Abdollahi, S., Acero, F., Ackermann, M.,et al. 2020, ApJS, 247, 33.Acero, F., Ackermann, M., Ajello, M., etal. 2016, ApJs, 223, 26.Ackermann, M., Ajello, M., Allafort, A.,et al. 2013, ApJS, 209, 34.Ajello, M., Atwood, W. B., Baldini, L., etal. 2017, ApJs, 232, 18.Aliu, E., Arlen, T., Aune T. et al. 2011,Science, 334, 69.Alpar, M., A., Cheng, A., F., Ruderman,M., A., Shaham, J., 1982, Nature, 300,728.Ansoldi, S., Antonelli, L. A., Antoranz,P., et al. 2016, A&A, 585, A133.Archibald A., M., Stairs, I., H., Ransom,S., M., et al. 2009, Nature, 324, 1411. Atwood, W., B., Abdo, A., A.,Ackermann, M., et al. 2009, ApJ, 697,1071.Atwood, W. B., Baldini, L., Bregeon, J.,et al. 2013, ApJ, 774, 76.doi:10.1088/0004-637X/774/1/76Bhattacharya D., van den Heuvel E. P. J.,1991, Physics Reports, 203, 1.Bhattacharyya, B., Roy, J., Ray, P. S., etal., 2013, ApJ Letters, 773, 12.Bhattacharyya, B., Cooper, S., Malenta,M., et al., 2016, ApJ, 817, 130.Bhattacharyya, B., Roy, J., Stappers, B.W., et al., 2019, ApJ, 881, 1.Bruel, P., Burnett, T. H., Digel, S. W., etal. 2018, arXiv:1810.11394Chang, S., & Zhang, L. 2019, MNRAS,483, 1796.Cheng, K. S., Ho, C., & Ruderman, M.,1986, ApJ, 300, 500.Cordes, J. M., Lazio, T. J. W., 2001(astro-ph/0207156).Coronado-Blazquez, J.,Sanchez-CondeVenter, M. A.,Dominguez A., et al. 2019, JCAP, 07,20.Dai, S., Hobbs, G., Manchester, R. N.,2015, MNRAS, 449, 3223. iscovery and timing of three MSPs Damour, T., & Deruelle, N. 1986, Ann.Inst. Henri Poincar´e Phys. Th´eor, 44,263.de Martino, D., Casares, J., Mason, E. etal. 2014, MNRAS 444, 3004.de Jager, O. C., Raubenheimer, B. C., &Swanepoel, J. W. H., 1989, A&A, 221,180.de Jager, O. C. & B¨usching, I., 2010,A&A, 517, 9.Edwards, R. T., Hobbs, G. B.,Manchester, R. N. 2006, MNRAS, 372,1572.Faucher-Giguere C. A., Kaspi V. M.,2006, ApJ, 643, 332.Finkbeiner, D. 2003, ApJ, 146, 407.Frail, D. A., Mooley K., P., JagannathanP. et al. 2016, MNRAS, 461, 1062.Freire, P. C. C., & Ridolfi, A. 2018,MNRAS, 476, 4794.Freire, P. C. C., & Wex, N. 2010,MNRAS, 409, 199.Gupta Y., Ajithkumar B., Kale H. S. etal., 2017, Current Science, 113, 707.Hou, X., Smith, D. A., Guillemot, L. etal. 2014, A&A, 570, 44.Harding, A. K., Usov, V. V., & Muslimov,A. G., 2005, ApJ, 622, 531.Harding, A. K., Kalapotharakos, C.,Barnard, M., et al. 2018, ApJL, 869,L18.Haslam, C. G. T.,Salter, C. J., Stoffel, H.,Wilson, W. E. 1982, A&A, 47, 1.Jennings, R. J., Kaplan, D. L.,Chatterjee, S. et al. 2018, ApJ, 864, 26.Johnson, T. J., Venter, C., Harding,A. K., et al. 2014, ApJS, 213, 6.Kalapotharakos, C., Harding, A. K., &Kazanas, D., 2014, ApJ, 793, 97.Keith, M. J., Johnston, S., Bailes, M., etal. 2011, MNRAS, 419, 1752.Kerr, M. 2011, ApJ, 732, 38.Kirk, J. G., Skjæraasen, O., & Gallant,Y. A. 2002, A&A, 388, L29.Kramer, M., Xilouris, K. M., Lorimer, D.R., et al. 1998, ApJ, 501, 270. Lange, C., Camilo, C., Wex, N. et al.,2001, MNRAS, 326, 274.Lorimer, D. R., & Kramer, M., 2004,Handbook of Pulsar Astronomy, Vol. 4.Cambridge, UK.Nice, D., Demorest, P., Stairs, I. et al.2015, Astrophysics Source CodeLibrary.Ransom, S., M., Eikenberry, S., S., &Middleditch, J. 2002, AJ, 124, 1788.Ray, P. S., Kerr, M., Parent, D., etal. 2011, ApJS, 194, 17.Ray, P. S., Abdo, A. A., Parent, D. et al.,2012, ,arXiv:1205.3089.Roy J., Gupta Y., Ue-Li Pen et al. 2010,Experimental Astronomy, 28, 55.Roy J., Bhattacharyya B., Gupta Y.,2012, MNRASL, 427, 90.Roy J. & Bhattacharyya B., 2013, ApJL,765, 45.Roy J., Ray P. S., Bhattacharyya B. etal., 2015, ApJL, 800, 12.Saz Parkinson, P., Belfiore, A., Fidalgo,D., et al. 2017, Proceedings of the 7thInternational Fermi Symposium, 8.Swarup, G., Ananthakrishnan, S.,Subrahmanya, C., R. 1997, in