Pulsar broadening measurements at low frequencies with LOFAR Kimon Zagkouris University of Oxford In collaboration with the LOFAR Pulsar Working Group Ierapetra June 2014 Image Credit: David A. Aguilar (CfA) / NASA / ESA
Dec 22, 2015
Pulsar broadening measurements at low frequencies with LOFAR
Kimon ZagkourisUniversity of Oxford
In collaboration with the LOFAR Pulsar Working Group
Ierapetra June 2014
Image Credit: David A. Aguilar (CfA) / NASA / ESA
Pulsar Scattering• Interstellar medium (ISM) is not
uniform or isotropic• The ISM causes radio waves to:
• Disperse• Scatter
• Scattering → “exponential like” tail.• Thin/thick screen or a uniformly distribution medium.• τd να
B1831-03 Löhmer et al. (2004)
Lorimer and Kramer (2005)
Pulsar Scattering
• Observed profile → I ⨂ ISMs ⨂ DMs ⨂ RsTraditional Measuring τd:
• Higher frequency profile → no scattering.• Convolve it with the ISM and instrument functions.• Fit on the observed profile.• Repeat for a range of τd. Best fit → τd .
Drawbacks:• Requires high frequency not scattered profile.• Profile evolution → Wrong τd.
Intrinsic profile
Observed profile
ISM ⨂ DMs ⨂ Rs
CLEAN based method
Developed by Bhat et al (2003)..
1. Find maximum of the profile.2. Multiply maximum with a gain factor (e.g. 5%).3. Convolve this with ISMs ⨂ DMs ⨂ Rs.4. subtract it from the profile.5. Repeat until the residual profile is noise like
• Repeated for a range of τd values. • Best τd → best noise like residual.
B2111+46 @ 122 MHz (LOFAR HBA)
Scattered “Cleaned”
Best value for τd:• Γ → skewness of the residual.• Fr → positivity of the residual.• Pvalue → Kolmogorov-Smirnov test value.• Nf → total number of iterations.
Minimising Γ + Fr → best τd value.
CLEAN method:• Finds the value of τd for a given ISM screen
model.• Finds the best screen model.• No high frequency profile needed.• Can return the intrinsic profile.
CLEAN based method
LOFAR Observations
• Bhat et al. (2004) measurements below DM=100 are from Δνd. 2π τd Δνd ≈ 1.
• LOFAR:• LBA 30 – 80 MHz• HBA 120 – 240 MHz• 80 MHz bandwidth.• Continuous band coverage.
• If τd ≈ P0 → pulsar might not be detected.
• LOFAR ideal to explore 10 – 200 DM region with direct measurements of τd.
• Measure α within LOFAR band.• Check for deviations in low frequency regime.
• LOFAR observed ~100 pulsars.
• 30 – 40 candidates for scattering measurement.
• Only 22 had scattering tails and enough SNR.
• 4 – 16 independent measurements within the band.
• 3 pulsars had their τd measured for the first time at these
frequencies.
• All 22 sources had only sporadically measurements at low
frequencies.
• Measured α within a continuous frequency band for the
first time!
This work
LOFAR Observations - Results
τ d
Frequency dependency of τd
Theory suggests:• Kolmogorov α ≈ -4.4• Gaussian α ≈ -4
Measurements:• Bhat (2004) α ≈ -3.86 ± 0.16• This work α ≈ -2.64 ± 1.28
Difference could be because:• We used a thin screen model in all cases.• Different scattering procedure at low frequencies.• Screen truncation Cordes and Lazio (2001) →
“flatter” spectrum.• Strong profile evolution.• Multiple screens.• Low SNR -> bigger error for τd and α. Direct measurements of τd (Lewandowski
2013). Purple filled circles are points of this work.
Τ d (m
sec)
Frequency (MHz)
B2217+47
Τ d (m
sec)
Frequency (MHz)
This work
B0611+22
Possible break!
Some pulsars (e.g. B0611+22) indicate possible break in the powerlaw. Mid-high frequency observations needed.
Three pulsar showed a steeper spectrum than expected α ≈ -5.7. Possibly a thin screen model is not the best choice for these cases.
Intriguing results
Conclusions and Future work
The story so far:• LOFAR is great to study scattering!• τd and α for 22 pulsars at low
frequencies.• Not all pulsars can be used for
scattering measurements.• Indications for a different scattering
behavior at low frequencies.• CLEAN based method → deconvolved
profile → useful for pulsars used in timing experiments.
The road ahead:• LOFAR Cycle 1 and 2 observations -> more
scattering measurements.• Time variability of scattering.• Telescopes such as (GBT, Arecibo, GMRT, LWA,
MWA) can fill in the frequency gaps to probe for breaks in the power law.
• Southern looking telescopes will help increase coverage and the analysis’ statistics.
• Cyclic spectroscopy (Demorest 2011) can measure the scattering timescales much more accurate and is the next thing to try!
Thank you!