1. Aristeidis Noutsos The Galactic Magnetic Field from Pulsar RMs and the Low-Frequency Arrays Aristeidis Noutsos Jodrell Bank Centre for Astrophysics, Manchester, UK

2. Other tracers of the Galactic magnetic field Transverse component (B ⊥ ) Synchrotron radiation from CRs ‣ Assumes energy equipartition of B and CRs. Polarimetry of dust or starlight ‣ Requires bright sources. Works up to ~ 2 kpc. Parallel component (B || ) Zeeman splitting of spectral lines ‣ Hard to measure. Difficult to translate the results into the large-scale component of the B-field.

3. Faraday Rotation of Pulsar Emission Pulsars are amongst the most polarised radio sources. Some, e.g. Vela, are ¼100% linearly polarised! Linear Circular Total Vela The plane of linearly polarised emission rotates (ΔPA) as highly-polarised pulsar emission propagates through the magnetised ISM. The amount of rotation across the observation band is expressed by the RM: PA Telescop e ISM Pulsa r B

4. d1d1 d2d2 LOS Potentially, field reversals can be revealed along the LOS B The Interstellar Magnetic Field from Pulsar Rotation Measures One can obtain the radial function of using pairs of nearly aligned pulsars, e.g. at d 1 and d 2 The average strength and direction of the B-field along the LOS to the pulsar is often estimated as B || B d

5. Advantages: Pulsars … The Interstellar Magnetic Field from Pulsar Rotation Measures … are highly polarised radio sources … are scattered throughout the entire Galactic volume … lie at approximately known distances (derived from DM + n e models) l=180 o 60 o 240 o Galactic Hammer–Aitoff projection North-pole projection of Galactic plane All known PSRs NE2001 n e model B synchrotron

6. where’s my pulsar?! (Actual pulsar position.) Galactic Hammer–Aitoff projection Problems, Problems … The Interstellar Magnetic Field from Pulsar Rotation Measures We have RMs for only ~ 1/3 of the known pulsars: Some are simply too weak or their pulses are too scattered to measure their polarisation properties Not all pulsars have measurable degree of polarisation Pulsar-distance estimates on the GP based on NE2001 can be up to ~ 20% in error. For high-b pulsars this error can be much higher (~ 50% in some cases!) (see Gaensler et al. 2008; PASA submitted)

7. … and more problems! The Galactic magnetic field can be seen as a regular, large-scale field mixed with a turbulent, small-scale component. The Interstellar Magnetic Field from Pulsar Rotation Measures ∝ RM∕DM only if n e (l) and B(l) are independent! For the turbulent field δ n e and δ B are correlated under/over estimation of (see Beck et al., 2003). The existence of small-scale magnetised regions affect the mapping of the ISM field from RMs. + ~ kpc ~ 10 – 100 pc HII region ~ 4 µG ~ 5 µG

8. New Pulsar Rotation Measures and the Galactic Magnetic Field

9. l=180 o 60 o 240 o We performed independent measurements of 150 pulsar RMs at 20cm with the 64m Parkes telescope (Noutsos et al. 2008): positivenegative l=180 o 60 o 240 o ‣ 46 new RMs ‣ 12 new RMs in Q1: a quadrant that benefited from the new sample ‣ ~ 20 RMs were revised from their previously published values

10. We plotted a map of the projected values of on the GP using all LOS with available pulsar data and looked for field reversals in the best sampled directions. B || towards observer B || away from observer Large-Scale Magnetic Field Reversals CCW CW CCW distance l = 305–310 o A field reversal is seen between Carina and Crux in Q4. CCW distance l = 305–310 o Between 6–8 kpc, in the Crux arm, the field appears to reverse from CCW to CW. CW CCW CW distance l = 280–285 o The field reverses from CW to CCW in the Carina arm region, where anomalous RM values have been reported (“Carina anomaly”; see e.g Han et al. 2006).

11. Local field Quadrant 1 Local field reverses from CW to CCW at ~ 1 kpc(confirms Lyne & Smith 1989). Quadrant 4 A reversal within r ☉ ~ 2 kpc towards ℓ = 285–290º is consistent with Frick et al. (2001) within the distance uncertainties. Local Neighbourhood Field Reversals Q1Q4

12. We selected 4 large-scale magnetic-field models to compare to the data 3 bisymmetric spirals + 1 dipolar–toroidal model face-on edge-on TT HMR PS Dipol.– Toroidal PS Dipol.– Toroidal B spiral + B halo B spiral + B halo + B toroidal + B dipolar B dipole + B toroidal TT HMR PS D.–T. Testing the Large-Scale Field Models

13. Comparison with the data revealed that the large-scale component alone cannot explain the B fluctuations. data PS TT HMR Dipol.– Toroid. Also, HII regions have a significant impact on the RMs and cause an ‘anomalous’ variation of on top of a smooth large-scale component (see e.g. Mitra et al 2003). Testing the Large-Scale Field Models

14. LOFAR MWA LWA RM Measurements with Low Frequency Arrays

15. Pulsar All-Sky Surveys ‣ Large effective area: ~ 10 5 m 2 (i.e. full-size original LOFAR design) ‣ Wide field-of-view: multi-beaming capabilities provide wide instantaneous sky coverage ‣ Optimal sensitivity at frequencies where most pulsar spectra peak: ~ 100–200 MHz (high-frequency band) Huge potential for discovering new pulsars But the distance on the Galactic plane to which pulsars will be discovered by low-frequency arrays will be limited by pulse scattering and sky background (high at low frequencies) Low frequency arrays will discover mostly low-DM pulsars: Nearby pulsars on the Galactic plane and high-latitude pulsars

16. All-Sky Surveys X (kpc) Y (kpc) –10–50510 0 5 Sun Known pulsar-RM sky van Leeuwen & Stappers (2008) LOFAR 60-day Survey Simulation X (kpc) Y (kpc) –10–50510 0 5 Sun 1000+ pulsars detected ~ 600 RMs measured GC Q1 Q2 Q1 Q2 New RMs in New Directions will help map the large-scale field A denser sample of RMs will increase our knowledge of the small-scale field ~ 100 pc

17. Polarimetry Advantages ‣ Large bandwidth (32 MHz) at low frequencies (20 ~ 300 MHz) RM = 0 ? Low-frequency arrays will have high sensitivity to small RMs, especially at their low-frequency bands: e.g. ~ 20–80 MHz. Hence, the small RMs (~ 1 rad m –2 ) of nearby, high-latitude pulsars will be accurately determined. RM = – 0.3 rad m –2

18. Multi-channel spectro-polarimetry: RM synthesis* (Brentjens & de Bruyn 2005) Decomposing RM Space ‣ High frequency resolution (up to 30,000 channels) 0 RM AB Pulsar Regions of Faraday rotation and polarised emission RM RMTF ~ 0.1 rad m –2 Bandwidth Low BandHigh Band Selected frequency ranges By appropriately selecting the frequency coverage, individual RMs can be resolved down to ~ 0.1 rad m –2 level f Side-lobes due to incomplete frequency coverage * RM Synthesis: decomposition of RM into discrete components (Fourier spectrum)

19. Ionospheric Faraday Rotation can contribute as much as ~ 5 rad m –2 (e.g. Junor et al. 2000). By using e.g. bright, polarised pulsars as calibrators, low-frequency arrays can improve the ionospheric electron-density models and help correct for the systematic effect caused by Faraday rotation through the ionospheric plasma: Ionospheric Calibration MWA Science Goals (website) Ionospheric Noon Sol. Max Ionospheric Noon Sol. Min Ionospheric Noon Sol. Max Ionospheric Noon Sol. Min Ionospheric Noon Sol. Max Ionospheric Noon Sol. Min This can reduce the systematic errors due to the ionosphere to as low as σ iono ~ 0.01 rad m –2, thus improving pulsar-RM measurements.

20. The High-latitude Sky By measuring the RMs of high-latitude pulsars, low-frequency arrays can shed light on the high-latitude Galactic B-field: LOFAR 60-day Pulsar Survey Simulation van Leeuwen & Stappers (2008) GC z = 1.8 kpc n e scale-height B field @ y = 8.5 kpc z < 0.5 kpc 0.5 < z < 1.0 kpc 1.0 < z < 1.5 kpc z > 1.5 kpc Latitudinal distribution of known pulsar RMs Studies of high-latitude pulsars will also help improve the electron-density models for high latitudes.

21. Summary ‣ Current efforts to map the Galactic Magnetic field using pulsar RMs show promise and have unveiled a number of features in the field’s structure (field direction, reversals, etc.) ‣ However, the sample of measured pulsar RMs is sparsely distributed across the sky, which makes the study of the large-scale Galactic magnetic field in certain directions difficult. ‣ Low frequency arrays (LOFAR, MWA, LWA, etc.) have the potential to discover many more nearby and high-latitude pulsars, many of which will provide RM measurements in follow-up polarisation observations. The sensitivity of low-frequency arrays to the small RMs expected from this objects will provide accurate measurements. ‣ The high frequency resolutions and high bandwidths of the low-frequency arrays open a new window into measuring the RM contributions of discrete sources of Faraday rotation along the LOS, as well as removing the systematic effects of the ionosphere. ‣ Current efforts to map the Galactic Magnetic field using pulsar RMs show promise and have unveiled a number of features in the field’s structure (field direction, reversals, etc.) ‣ However, the sample of measured pulsar RMs is sparsely distributed across the sky, which makes the study of the large-scale Galactic magnetic field in certain directions difficult. ‣ Low frequency arrays (LOFAR, MWA, LWA, etc.) have the potential to discover many more nearby and high-latitude pulsars, many of which will provide RM measurements in follow-up polarisation observations. The sensitivity of low-frequency arrays to the small RMs expected from this objects will provide accurate measurements.