LOFAR's view on B0943+10

Anya Bilous

Transcript

1. LOFAR's view on B0943+10 Anya Bilous, Radboud Universiteit Nijmegen & LOFAR Pulsar Working Group
2. A pulsar: • Rapidly rotating ball of neutrons • Strong magnetic field (dipolar) • Spin and magnetic axes misaligned • Magnetosphere: particles move along field lines and rotate rigidly with the star • Light cylinder v_rot = c • Open field lines – do not close within light cylinder • Radio pulses: lighthouse effect Hewish et al, 1968
3. Average profile – stable in time, unique for each pulsar... But not always! Mode changing: some pulsars have 2 stable average profiles
4. Two faces of PSR B0943+10: Q-mode (quiet) B-mode (bright) • Different average profiles • Different single pulses • Abrupt switch (~1 spin period) Most of pulsar rotation energy loss goes through pulsar wind, magnetic dipole radiation and high-energy photons. Only tiny fraction of rotation energy goes to radio. So what?
5. Hermsen et al 2013 Expected: B-mode: bright radio → bright X-rays Q-mode: faint radio → faint X-rays Simultaneous observations of B0943+10 in X-rays and radio
6. Hermsen et al 2013 Expected: B-mode: bright radio → bright X-rays Q-mode: faint radio → faint X-rays Observed: No thermal X-rays at all Bright thermal X-rays Simultaneous observations of B0943+10 in X-rays and radio
7. So what is going on? Which parts of the magnetosphere are active in radio in B and Q modes? http://sci.esa.int/xmm-newton/51320-the-two-states-of-pulsar-psr-b0943-10-as-observed/
8. Classical approach: Radius to frequency mapping • Photons are emitted tangentially to field lines • Emission at given frequency comes from a single height Thanks to high-quality LOFAR data! (60x bandwidth, 100x observing time comparing to archive low- freq data from the other telescopes) Components merge together – line of sight misses the inner edge of the emission cone c B-mode Q-mode Αα = 12 deg B β = -5 deg (Deshpande & Rankin 2001)
9. Remark #1: plotting Plotting only the plane defined by spin and magnetic axes
10. We know angular radius of cone, but not the emission height (same radius for multiple lines starting at different magnetic latitude ϑ 0 ) r sin-2ϑ = R NS sin-2ϑ 0 ϑ 0 <0.013 rad (last open field line, ϑ p ) ϑ 0 >0.004 rad (absence of light travel time delay, aberration, magnetic line sweepback) Remark #2: which field line? On the next slides location of emission region is given for last open field line starting at ϑ p . For rest of lines coordinates scale as 1 < (ϑ p /ϑ 0 )2 < 10
17. c No information about the inner edge B-mode Q-mode 1 < ϑ p /ϑ 0 < 3.5 Radius of light cylinder : 52000 km Radius of the star: 10 km The difference in average profiles in B and Q modes at 25-80 MHz can be explained by shifting emitting region by few km along the field line.
18. • We used simple conventional assumptions to put constraints on the location of radio emitting regions in B and Q mode. • Radio emission comes from narrow range of heights. This heights are not very different between the modes.
19. Two faces of PSR B0943+10: Q-mode (quiet) B-mode (bright)
20. • Detected in B-mode only • Change in spin- down rate? Too big and cyclic • Changing magnetic field lines? Bending? Shift of pole? B-mode profile lag – about 2 ms/hr
21. Summary • The difference in average profiles in B and Q modes at 25-80 MHz can be explained by shifting emitting region by few km along the field line. If something dramatic happens to radio, it must be at much lower/higher frequencies or very close to the magnetic pole. • The midpoint between the components in B-mode drifts by about 2 ms/hr. The midpoint in Q-mode does not change within these limits. Why?
22. • Not a timing issue • Detected in B-mode only • Change in spin-down rate? Too big and cyclic • Changing magnetic field lines? Bending? Shift of pole? Shift of B-mode midpoint: about 2 ms / hr
24. LOFAR observations: LBA: f_central = 53.8 MHz BW = 78.1 MHz HBA: f_central = 155.2 MHz BW = 95.3 MHz Time resolution = 0.6 ms Stokes I LBA 09 March 2013 2hr Q 2hr B Q → B transition LBA 27 Feb 2013 10min Q 3.3 hr B Q → B transition HBA 30 Nov 2012 1.6hr B 2.9 hr Q B → Q transition LBA 28 Nov 2012 1 hr B
25. But we were lucky: the midpoint between 2 components shifts as ν-2 ~ f - 2 (φ c1 +φ c2 )/2 ~ kDMν-2 We can measure DM! Components in B-mode shift symmetrically around the midpoint (f1 = f2)
26. Spin phase Pulse # II) Drifting subpulses 1/P3 from FFT along constant spin phase line P1 == spin period In Q-mode – nothing In B-mode – strong peak at ~0.46 cycles/P1 Observed 1/P3 is an alias of its real value of ~0.53 cycles/P1 (Deshpande & Rankin, 2000) Serylak et al 2009
27. With increasing frequency, driftbands move closer both in P3 and P1 phases until they line up in a single driftband around 200 MHz Is the symmetry in P3 drift connected to the symmetry in profile evolution?... (recall B0809+74)
28. Mode switching happens within 1 pulse period: From chaotic pulses and broad profile To the lines of drifting subpulses and narrow profile Early (~ first ten of P1) drifting subpulses come from the phase range of the Q-mode average profile Another observed Q → B transit looks similar