Lucy Oswald
Pulsar radio emission and its polarization are observed to evolve with frequency. This frequency dependence is key to the emission mechanism and the structure of the radio beam. With the new ultra-wideband receiver (UWL) on the Parkes radio telescope we are able, for the first time, to observe how pulsar profiles evolve over a broad continuous bandwidth of 700–4000 MHz. We describe here a technique for processing broad-band polarimetric observations to establish a meaningful alignment and visualize the data across the band. We apply this to observations of PSRs J1056–6258 and J1359–6038, chosen due to previously unresolved questions about the frequency evolution of their emission. Application of our technique reveals that it is possible to align the polarization position angle (PA) across a broad frequency range when constrained to applying only corrections for dispersion and Faraday rotation to do so. However, this does not correspond to aligned intensity profiles for these two sources. We find that it is possible to convert these misalignments into emission height range estimates that are consistent with published and simulated values, suggesting that they can be attributed to relativistic effects in the magnetosphere. In this poster we present the described work in the context of what is known and unknown about pulsar emission and describe what new information this work can uncover.
Plain text poster summary
Lucy Oswald, University of Oxford
Contact: lucy.oswald@physics.ox.ac.uk
Related publication: Oswald L., Karastergiou A., Johnston S., 2020, MNRAS, 496, 1418
Slide 1
Abstract: We create visualizations of broadband polarimetric data of two pulsars observed with the new Parkes ultra-wideband receiver. We use a Markov chain Monte Carlo to align the angle of linear polarization (PA) across frequency and so correct for the effects of the Interstellar Medium. For both pulsars, aligning the angle of linear polarization (PA) does not align the total intensity. One explanation, aberration/retardation (A/R), implies low frequency emission is produced higher above the pulsar surface. Our height predictions agree with a theory of how height range may evolve with spin-down energy loss rate.
[Image description: Figure 1. Waterfall plots (x axis is rotational phase; y axis is frequency; colour/shade is value) of PA across phase and frequency for PSR J1056–6258, before and after alignment.]
A pulsar is a spinning neutron star emitting a beam of radio waves. We detect the radio beam as a series of pulses as the pulsar rotates and the beam sweeps through space. Summing them together produces an integrated profile that is stable over time.
[Image description: cartoon pulsar represented by a green circle with two yellow triangles representing its beams of radio emission. Its axis and direction of rotation are indicated with arrows.]
Slide 2: Context
The angle of linear polarization, called the position angle or PA, has a shape that results from the pulsar geometry. This means that the shape of the PA is predominantly independent of frequency.
[Image description: Figure 2. PSR J1359–6038, 1410 MHz profile showing linear and circular polarization, and PA.]
The frequency evolution of pulse profiles is key to understanding the 3D structure of the beam. For example, aberration/retardation (A/R) effects cause the observed intensity and PA to misalign. Emission generated higher above the pulsar surface is misaligned more.
However, the Interstellar Medium (ISM) also introduces frequency evolution:
• Dispersion delays pulse arrival.
• Faraday rotation shifts the PA.
The proportionality constants of the frequency relationships of dispersion and Faraday rotation are called the dispersion measure (DM) and rotation measure (RM) respectively.
[Image description: Figure 3. Waterfall plot showing dispersed profile of PSR J1056–6258.]
Slide 3: Method
To study the properties of broadband observations, we must first identify the correct DM and RM. The constancy of the PA shape means we can align it across the observing band. We use a Markov chain Monte Carlo to sample the log-likelihood for PA alignment over the parameter space of ΔDM and ΔRM.
[Image description: Figure 4. Left: aligned pulsar profiles for PSR J1369–6038 (PA above, total intensity below), frequencies overlaid from lowest (blue) to highest (grey). Right: alignment probability distribution for ΔRM and ΔDM.]
Slide 4: Results
[Image description: Figures 5-8. Waterfall plots (set up as described for figures 1 and 3) showing PA and total intensity for PSRs J1359–6038 and J1056–6258. In both cases the PA is aligned, and total intensity is misaligned.]
Our best fit values align the PA profiles well. However, the total intensity profiles are not aligned, and the misalignment is larger at lower frequencies. Assuming A/R it implies that the 700 MHz emission may be produced 120 ± 120km higher up than that at 4000 MHz for PSR J1359–6038, and 430 ± 350km higher for J1056–6258. One theory suggests that narrow emission height range is linked to low spin-down energy loss rate (Ė) and vice versa. This fits with our calculations.