2 Identification and results

We took positions of pulsars from the updated catalog of Taylor et al. (1993). PSR names in J2000, and B1950 if applicable, are given in the Cols. (1) and (2) of the Table 1. Their positions are given in Cols. (3) and (4), generally with an accuracy better than 0.1'', but occasionally up to a few arcsec. These positions were determined by timing observations or interferimetric measurements at epoch for the position in Col. (5). For comparison, we list in Col. (6) the flux density at 1.4 GHz from pulsar catalog, which were normally obtained from the average of several pulsar observation sessions to overcome scintillation effects. We searched for radio sources in the NVSS catalog within 30'' angular distance around each of the 520 pulsar positions. Only 106 radio sources were found to match the positions and are probably pulsars. The positions of the NVSS sources are listed in Cols. (7) and (8). The angular offset from pulsar positions "$\Delta$'' in arcsec is given in Col. (9). The flux density and polarization parameters of the NVSS sources extracted from the NVSS catalog are listed in Cols. (10)-(13). A blank in these columns indicates no significant detection above the sensitivity limit of linear polarization of the NVSS ($\sim$0.5 mJy). We marked in Col. (14) if there was any further consideration during identification.

Note that the epochs for pulsar position in the pulsar catalog differ from that of NVSS observations. However, even if a pulsar has the largest proper motion, e.g. 400 mas per year, then after 20 years, the position offset would be 8''. So, our search in 30'' should not miss any known pulsar if it is detectable by the NVSS.

On the other hand, the NVSS was done over a long period, from $\sim$1993 to $\sim$1996. We will take an approximate epoch MJD 49718 ($\sim$1995.0) in following discussion. There should be only a very small position offset (<1'') caused by pulsar proper motions, if any, over the NVSS observation period, much smaller than the position uncertainties of the NVSS sources listed in Table 1. If the position of a pulsar was measured at an epoch later than MJD 47000, we will not consider its proper motion during the identification process for the same reason.

  

Table 1: NVSS sources around pulsar positions

 

Table 1: continued

The first step for identification is to check the position offset $\Delta$.At this stage, we ignored the proper motion. If $\Delta$ is smaller than twice of the total position uncertainty, i.e. , then we attribute the NVSS source as being a positive identification of a pulsar. This process yielded the first 90 positive detections. If any pulsar position was obtained at an epoch several years ago, the pulsar must have had only a very small proper motion so that the position offsets are not significant.

Now we consider the remaining 16 sources more carefully, which are marked with "?'' in Col. (14) of Table 1.

Nine Confusion cases: (a) PSR B0531+21 (Crab) and PSR B1951+32 are confused by their associated supernova remnants. We marked them in Notes, i.e. Col. (14), of Table 1 with "SNR''. (b) PSRs B1112+50, B1829-10 and B1831-00 are confused by their nearby strong sources which have much larger flux density (more than 10 times) than that from the pulsar catalog. One NVSS source was detected 28.1'' (formally $7.8\sigma$) away from PSR B1920+21, too large to be proper motion for this distant pulsar (distance $\sim$12.5 kpc). We consider these detections unlikely and mark with "no'' in Notes to stand for "no detection''. (c) PSRs B1744-24A and J2129+1210A, (maybe also B1745-20 as indicated by Kaplan et al. 1998), are confused by other continuum sources in the host globular clusters Terzan-5 and M 15, (and NGC 6440?), respectively. They are marked with "glbc''. (d) PSR B1718-35 is a marginal case, maybe confused by a source $19\hbox{$.\!\!^{\prime\prime}$}4$ away, with 4.7$\sigma$ for position offset and 27.7 mJy in flux (pulsar: 10.0 mJy).

Seven detection cases: (a) The position offset of the NVSS source to PSR B1831-04 is only 4.75'' (formally $2.3\sigma$, or $2.6\sigma$ rather than $14\sigma$ using the new position in Kaplan et al. 1998), much smaller than the beam size of the NVSS. Although Kaplan et al. (1998) suggested otherwise, we believe the pulsar is detected. The consistent flux densities of the pulsar and the NVSS source confirm the identification. We mark such a case as "yes" in the Notes. (b) PSRs B0823+26, B1133+16, B2016+28, (and B2154+40) have small position offsets caused by proper motions (see Sect. 3.3). (c) PSR B1820-31 is detected with a position offset of $12.5'' = 2.6\sigma$, as confirmed by consistent flux density, and more importantly, by the highly linear polarization of the NVSS source. (d) A marginal case is the strong pulsar PSR B2020+28. The NVSS survey detected a very weak source $2.2\sigma$ away, too weak to believe the identification (see more discussion below). However, highly linear polarization of the source suggests that it is the pulsar. We mark in the Notes "yes?" for this case.

In all, the NVSS detected 97 pulsars, including the 73 which appeared in Kaplan et al. (1998) and 24 new identifications marked with "*" in the Notes.