IDV from Subimages

1
IDV from Subimages
Carl Gwinn UC Santa Barbara

• Outline
• IDV has some seemingly paradoxical properties
• Quasi-sinusoidal intensity variation
• Nearby scattering screen
• (tricky to explain via selection effects)
• Scintillations of PSR J0437-471 show
• Large decorrelation bandwidth ?small scattering
  angle
• Fine substructure
• Subimages can give rise to substructure
• Subimaging of IDV sources can explain some
  paradoxes
• Sinusoidal intensity variation
• Selects nearby scatterers
• SKA and LOFAR can observe subimages directly

2
Scintillation of 0405-385 Kedziora-Chudczer and
collaborators (1997) observe quasi-sinusoidal
intensity variations DI/I0.4, tDI1 hr,
VISM35 km/s

• Careful application of scattering theory yields
• Scattering near weak-strong transition
• qsource ?Dscreen (l Dscreen )1/2 VISM?tDI1010
  cm
• Dscreenlt50 pc (close).
• Strongly anisotropic scattering gt41
• --Rickett, Jauncey Kedziora-Chudczer 2002

Scintillation of 18193845 Dennett-Thorpe de
Bruyn (2002, 2003) see even more extreme
quasi-sinusoidal variations DI/I0.6, tDI1-10
hr, VISM30 km/s Dscreenlt10 pc (very
close) Anisotropy gt61
3

• Model for 4.8 GHz (Rickett et al. 2002)
• Source angular size
• qsource0.12 mas
• Fresnel angular scale
• (8 ln 2 l/D)1/20.2 mas
• Scattering angle
• qscatt0.14 mas
• Axial ratio lt0.25
• Screen distance
• D10 pc

• Distance of 10 pc is set by either or both of
• Scattering is near weak-strong transition, so
• -Fresnel angular scalesource angular size
• Timescale of scintillation and pattern speed
• -2-telescope measurements
• -annual modulation of timescale (18193845)

4
Selection Effects That Might Set the Screen
Distance are Tricky!

• We expect material to lie at typical ISM scales
  Dscatt500 pc.
• Scattering models (like Taylor Cordes 1993)
  concur, and suggest net scattering qscatt0.1 mas
  from the line of sight.

• But Scattering in the 10-pc screen must dominate
  the line of sight.
• If more distant screen(s) scatter more strongly,
  they will broaden the source to larger than the
  Fresnel angular scale at the 10-pc screen -- and
  so quench its scintillation.
• ?Note Interstellar scattering is optically
  thick every photon that passes through the
  screen is deflected.

5
Pulsar J0437-4715 lies 10 from PKS 0405-385.
At distance D150 pc, it lies somewhat beyond
material responsible for scattering PKS 0405-385.
(Shishov Smirnova 2004)

• We measure scintillation bandwidth DnISS15 MHz
  at observing frequency 330 MHz.
• Previous observations used too high a frequency
  or too narrow a bandwidth to detect scintillation
  this broad.
• Scintillation spectrum shows fine-scale structure.

Except Issur 2000, who found Dn consistent with
our wideband result, at f110 MHz.
From Hirano PhD thesis 2001
6
Substructure in Scintillation Spectrum of PSR
J0437-4715

• The fine structure visible in the scintillation
  spectrum is really there.
• Such substructure is not uncommon in pulsar
  dynamic spectra (Stinebring et al. 2001, Hill et
  al. 2003).
• We see characteristic bandwidth Dn0.5 MHz.
• All previous measurements for PSR J0437-4715
  found decorrelation bandwidth near this
  narrowband value.

Except for Issur 2000.
From Hirano PhD thesis 2001
7

• Stinebring et al. (2001) propose Subimages
  create substructure.

The subimages interfere with the primary
scattered image. In effect, subimage-primary
image pairs act as interferometers, to create a
fringe pattern in the observer plane. The fringes
are sinusoidal.
Pulsar
Scattering Screen
Observer
From Hill et al. 2003
8
Scintillation Bandwidth and Modulation Give
Image Parameters
Scintillation bandwidth and angular scale
Modulation
For PSR J0437-4715

• Diameter of primary scattering disk 0.4 mas
  (from DnISS)
• Distance to nearby subimages1.7 mas (from Dn for
  substructure)
• Intensity ratio Isub/Ipr0.16 (from substructure
  modulation of 70)

This scales to 0.0016 mas at f5 GHz
To create the quasi-sinusoidal variations seen
for 0405-385 from subimage(s) would require

• Angular distance to subimage0.2 mas
• Intensity ratio Isub/Ipr0.02 (from modulation of
  25)
• qscattlt0.1 mas along the line of sight

9
Subimaging is optically thin
Pulsar
Most photons arent bent by these large angles.

• Subimaging centers might lie all along the line
  of sight.
• To produce scintillation, a subimaging center
  must lie at transverse displacement y
• Beyond the first Fresnel zone ygt (l D)1/2
• Near enough that the source is unresolved (l
  /y)gtqsorc
• Subimaging centers outside this region dont
  matter.

Observer
Region for scintillation of subimages
Transverse scale y
Subimages
Scatterer distance D
10
How could SKA LOFAR find subimages in IDV?

• Low frequency capability (as well as high)
• Large fractional bandwidth
• esp at low frequency
• High angular resolution (as well as low)
• Good dynamic range

11
A fit to the correlation function for both epochs
yields
MHz
Estimated errors are from a Monte Carlo
simulation that fit models to simulated data,
evenly spaced in actual Dn. We fit only to model
spectra that had the highest intensity peak away
from the edge of the observed passband.
Results from Monte Carlo simulations yielding
Dn15 MHz.
Correlation function