IDV scattering screens: clues from pulsar spectroscopy

1
IDV scattering screens clues from pulsar
spectroscopy

• Mark Walker
• (Sydney Uni)
• Don Melrose Dan Stinebring

2
Outline

• Why study IDV screens with pulsars?
• Key observed phenomena arcs and arclets
• Interpretation distant IDV screens
• Current directions prospects
• The screen population very local examples

3
Background

• Rapid scintillation of PKS0405, J1819 PKS1257
  is due to local phase screens
• Ordinary sources, but extraordinary sightlines
  (Tb 1012 K favours screens which are nearby)
• Pulsars ( Tb gtgt 1012 K) avoid this bias
• Pulsar spectra exhibit fully modulated
  interference fringes from multipath propagation

4
Normal Dynamic Spectrum

• PSR B113316 (Ooty data Gupta, Bhat Rao 1999)
• Mottled appearance due to random interference
  maxima

5
Anomalous spectrum

• Obvious fringe pattern (multiple imaging)
• Natural representation is the power-spectrum of
  the dynamic spectrum

6
Another anomalous spectrum

• PSR B192910 Arecibo data
  Hill et al 2003 ApJ
• Criss-cross patterns what do they tell us?

7
More anomalies
B083406 Arecibo data showing arclets whose
apexes follow a parabola Courtesy Dan
Stinebring
8
Recent recognition that
(Stinebring et al 2001 ApJL)

• Parabolic arcs (and arclets) are generic features
  of pulsar secondary spectra.
• Parabolic distribution of power is natural
  Doppler-shift prop to qx
  (Geometric) delay prop to qx2
  qy2
• Thin screens are responsible.

9
Expected power distribution
Isotropic Kolmogorov model
Anisotropic Kolmogorov model
(Purely geometric delays assumed.)
10
Snapshots
11
Implications

• Patchy illumination --gt power concentrations on
  scales lt 0.15 AU (083406)
• Power concentrations may be diffractive
  (localised scattering) or refractive (lens-like)
• Scattering angles comparable to what is expected
  from IDV screens
• Vscr not large compared with Vpsr

12
Summary

• Pulsar spectroscopy often reveals anomalous
  scattering from one or more thin screens
• Scattering angles as expected for IDV screens
  (l2.2 scaling)
• Strong anisotropy is common
• Velocities are not large
• Consistent with distant, IDV-type screens
• Starting to tap into a rich information source
  --- great variety of power distributions seen

13
Current pulsar directions

• Image with N speckles exhibits N2 interference
  terms in the secondary spectrum, so complete
  reconstruction of the image is possible.
• Veff variations through orbit (Earth/PSR)
  determine Vscr and Dscr
• Multi-station (VLBI) cross-power spectrum yields
  direct measures of image structure, orientation
  and angular scale

14
PKS0405-385 (Kedziora-Chudczer et al 2001 ApSS)
15
Connection between IDV and lensing events?

• Estimate tlens 1 / 2000, Rlens 2 AU, so
  if path-length 1 kpc, then
  n 1000 pc-3

16
Are there very local screens?

• J1819 is a line-of-sight with an unusually close
  screen t n s D1819 1 / 700 (MASSIV)
• Closest screen at Dmin / D1819 8 (R / D1819)
  2/3
• Estimate s p R2 ,
  with R gt 5 yrs at 30 km/s --gt 30 AU
• Hence n lt 2,000 (D1819 / 10 pc)-1 screens / pc3
• There are lt 107 (D1819 / 10 pc)2 closer
  screens

17
Can we find them?

• Closest screen at 0.05 pc if R 30 AU
• Would cause fast variability even below 1 GHz
• For screen at D D1819 / 100, expect
  tvar (0.25, 2) hr at (843, 327) MHz
• Need SUPER-MASSIV 100 x MASSIV, with 105
  compact sources
• SUPER-MASSIV SUMSS? WENSS?