Profile Instabilities of the Millisecond Pulsar PSR J1022+1001

Abstract

We present evidence that the integrated proÐles of some millisecond pulsars exhibit severe changes that are inconsistent with the moding phenomenon as known from slowly rotating pulsars. We study these proÐle instabilities in particular for PSR J1022]1001 and show that they occur smoothly, exhibiting longer time constants than those associated with moding. In addition, the proÐle changes of this pulsar seem to be associated with a relatively narrowband variation of the pulse shape. Only parts of the integrated proÐle participate in this process, which suggests that the origin of this phenomenon is intrinsic to the pulsar magnetosphere and unrelated to the interstellar medium. A polarization study rules out proÐle changes due to geometrical e †ects produced by any sort of precession. However, changes are observed in the circularly polarized radiation component. In total we identify four recycled pulsars that also exhibit instabilities in the total power or polarization proÐles due to an unknown phenomenon (PSRs J1022]1001, J1730 [2304, B1821[24, and J2145[0750). The consequences for highprecision pulsar timing are discussed in view of the standard assumption that the integrated proÐles of millisecond pulsars are stable. As a result we present a new method to determine pulse times of arrival that involves an adjustment of relative component amplitudes of the template proÐle. Applying this method to PSR J1022]1001, we obtain an improved timing solution with a proper-motion measurement of [17^2 mas yr~1 in ecliptic longitude. Assuming a distance to the pulsar as inferred from the dispersion measure, this corresponds to a one-dimensional space velocity of 50 km s~1.

Single-pulse and profile-variability study of PSR J1022+1001

“…The polarisation properties are in agreement between the two sites, with a correlation coefficient of 0.999 (as defined in Liu et al 2012). Given the profile instability of this pulsar, especially in observations separated by several years (Purver 2010), the polarisation components qualitatively agree with the previous observations (Kramer et al 1999;Ramachandran & Kramer 2003;van Straten 2013).…”

“…Here, following the idea of previous analyses (e.g. Kramer et al 1999), we calculate the amplitude ratio of the leading and trailing components of total intensity profiles for 10-min integrations, and plot them against time. For each component, the amplitudes were defined as the total intensity within a fixed phase range (1.7% of a period), after subtracting the baseline.…”

“…The same analysis at MJD 55909 data led to ρw = 0.55 when using the reception model and ρw = 0.60 when applying single-axis calibration, also consistent with the result based on calibrators. Further investigation based on independent Effelsberg data for the Large European Array for Pulsars project (Kramer & Stappers 2010), shows that the circular receiver does not suffer from strong crosscoupling between the two feeds (Lee et al in prep. ), which validates the use of the single-axis model in this case.…”

“…Liu et al 2012;Shannon & Cordes 2012), and temporally correlated variations induced by other phenomena (e.g. Kramer et al 1999;Edwards & Stappers 2003a). The contribution of jitter to the pulsar timing noise could be mitigated by simply increasing the exposure time or by correcting the biased integrated profiles with suitable algorithms (Cordes & Shannon 2010;Os lowski et al 2013).…”

“…The amplitude ratio of the two components has been shown to change significantly as a function of frequency (Ramachandran & Kramer 2003). There is also evidence that the ratio is unstable across time and may even evolve on short timescales of tens of minutes (Kramer et al 1999;Purver 2010). Meanwhile, as the trailing component of the pulse profile is highly polarised, profile instability could also arise from improper polarisation calibration due to an imperfect receiver model .…”