precsion pulsar timing and the interstellar mediumpulsar spectroscopy p. demorest nrao postdoc...
TRANSCRIPT
Precsion Pulsar Timing and theInterstellar Medium
Paul Demorest, NRAO Charlottesville
April 30, 2009
Outline
P. Demorest NRAO Postdoc Symposium 2009
Pulsar Timing and Gravitational Waves
Systematic Timing Effects / ISM
Advances in Pulsar Spectroscopy
Recent Observational Results
Collaborators:
• GW: NANOGrav collaboration (http://www.nanograv.org)
• ISM: Mark Walker, Willem van Straten
Gravitational Waves
P. Demorest NRAO Postdoc Symposium 2009
GW Detectors
P. Demorest NRAO Postdoc Symposium 2009
• LIGO = Laser Interferometer Gravitational Wave Observatory
• νGW ∼ kHz• Main targets are galactic sources (WD, NS, BH binaries)• Currently operating, major upgrade planned.
• LISA = Laser Interferometer Space Antenna
• νGW ∼ mHz• Both galactic and extragalactic (MBH) sources.• Still in planning stages.
• PTA = Pulsar Timing Array(s)
• νGW ∼ nHz• Extragalactic/cosmological sources (MBH, cosmic strings)• Several ongoing international efforts: NANOGrav (N.
America), PPTA (Parkes), EPTA (Europe).
GW Spectrum
P. Demorest NRAO Postdoc Symposium 2009
Detector complementarity:
���
���
���
���
���
���
���
��� �� �� �� �� � � �
��� ���� ���� �������������
��������������������
� ���
��! �����
"#!$
"#%&
$'(���'�"#%&
)*$�
)!+� ��,,-�.��/�0�1"�22���
3$3&%�(������� 3$3&%�(����,)*$�����
Pulsars as GW detectors
P. Demorest NRAO Postdoc Symposium 2009
Pulsars as GW detectors
P. Demorest NRAO Postdoc Symposium 2009
Radio pulses travelling through GW to Earth acquire additionaldelays.
Pulsar Timing Array
P. Demorest NRAO Postdoc Symposium 2009
Pulsar Timing Array
P. Demorest NRAO Postdoc Symposium 2009
GW-induced timing fluctuations will be correlated between differentpulsars (Hellings & Downs, 1982):
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0 20 40 60 80 100 120 140 160 180
Cor
rela
tion
coef
ficie
nt
Angular separation
This method makes GW detection possible!
Timing Results
P. Demorest NRAO Postdoc Symposium 2009
Best NANOGrav MSP timing results have RMS ∼ 100 ns:
-2.5-2
-1.5-1
-0.5 0
0.5 1
1.5
2004.5 2005 2005.5 2006 2006.5 2007 2007.5 2008 2008.5
Res
idua
l (us
)
Year
PSR J1713+0747
AreciboGBT
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2004.5 2005 2005.5 2006 2006.5 2007 2007.5 2008 2008.5
Res
idua
l (us
)
Year
PSR J1909-3744, GBT
8201400
Order-of-magnitude GW sensitivity is ∼ δt/T .
Timing Array Limits
P. Demorest NRAO Postdoc Symposium 2009
Correlation analysis of first ∼4 years of NANOGrav data:hc(1 y−1) < 7 × 10−15
-100
-50
0
50
100
0 20 40 60 80 100 120 140 160 180
Cor
rela
ted
pow
er (
norm
aliz
ed)
Angular separation
Improvement with time
P. Demorest NRAO Postdoc Symposium 2009
GW Spectrum
P. Demorest NRAO Postdoc Symposium 2009
Improvement with time:
���
���
���
���
���
���
���
��� �� �� �� �� � � �
��� ���� ���� �������������
��������������������
� ���
��! �����
"#!$
"#%&
$'(���'�"#%&
)*$�
)!+� ��,,-�.��/�0�1"�22���
3$3&%�(������� 3$3&%�(����,)*$�����
Timing Array Summary
P. Demorest NRAO Postdoc Symposium 2009
• Current NANOGrav PTA limits are comparable to the 20-yearsingle pulsar limit.
• Another ∼3–5 years reaches into plausible detection territory.
• Jenet et al. (2005) estimate that we need 20–40 pulsars at100 ns for 5 years for a “guaranteed” robust (> 4σ) detection.
• More good-timing pulsars are needed.
• Searching for new pulsars.• Increasing BW, G/T to improve currently known pulsars.• Reduce systematic effects, improve analysis algorithms.
(See also Demorest et al. (2009) Astro2010 WP)
Systematic Timing Effects
P. Demorest NRAO Postdoc Symposium 2009
We compute arrival times (TOAs) by matching data:
-0.05
0
0.05
0.1
0.15
0.2
0.25
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
to a scaled, shifted “template”:
-5
0
5
10
15
20
25
30
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Any corruption in the measured profile shape can affect timing.100 ns ∼ 10−5 turns.
Systematic timing effects
P. Demorest NRAO Postdoc Symposium 2009
• Local:
• Polarimetry / calibration procedures• Radio-frequency interference• Instrumental effects
• ISM:
• DM variation with time (and maybe frequency)• Scattering/scintillation• Refraction
• Intrinsic:
• Pulse-to-pulse “jitter”• “Timing noise”
ISM
P. Demorest NRAO Postdoc Symposium 2009
ISM effects
P. Demorest NRAO Postdoc Symposium 2009
• Come from radio wave propagation through interstellar e−
plasma; strongly λ-dependent, based on plasma dispersionrelation.
• Effects include DM(t), scattering/scintillation.• One solution: Observe at higher RF, but psrs get weaker.
-0.005 0
0.005 0.01
0.015 0.02
0.025 0.03
0.035 0.04
0.045
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
-0.005 0
0.005 0.01
0.015 0.02
0.025 0.03
0.035 0.04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
DM Variation
P. Demorest NRAO Postdoc Symposium 2009
Total electron column density varies due to motions of Earth, PSR,and ISM:
(PSR B1937+21; Ramachandran et al. 2006)Can be measured/removed with multi-frequency timingmeasurements.
Scattering/Scintillation
P. Demorest NRAO Postdoc Symposium 2009
Electron density variation transverse to the line-of-sight causesconstructive/destructive interference:
(Walker et al. 2008)
Typical scales ν ∼ kHz–MHz,T ∼ sec–hours. Determinedby LOS/ISM velocity and ISMlength scales.
Can be thought of as a (time-varying) “filter” with transfer functionH(ν) or impulse response h(t). This affects profile shapes!
Pulsar Spectroscopy
P. Demorest NRAO Postdoc Symposium 2009
How to compute spectra from measured pulsar voltage signal x(t):
Standard pulsar spectroscopy:
• Signal is divided into many radio frequency channels.• In each channel, on-pulse and off-pulse flux are detected,
integrated for some time and subtracted (“gating”).• Results in dynamic spectrum S(t, ν).
Limitations of this approach:
• Gate width and frequency resolution are coupled (via usualuncertainty principle).
• Discards all pulse shape information.• Can only directly determine |H(ν)|2.
Pulsar Spectroscopy
P. Demorest NRAO Postdoc Symposium 2009
In cyclic spectroscopy, we compute correlations as a function ofpulse phase:
• Signal is delayed by many different lags τ .• Each is cross-multiplied with original signal and averaged
modulo the pulse period for some integration time (“folding”).• Results in periodic correlation C(t, τ, φ) → S(t, ν, φ).
Advantages of this approach:
• Number of lags (hence freq resolution) not constrained by pulsewidth or period.
• Makes full use of pulse shape information.• Directly recovers H(ν) with phase!• see reviews by Gardner (1991, 1992), Antoni (2007)
Cyclic Spectra
P. Demorest NRAO Postdoc Symposium 2009
PSR B1937+21 at 430 MHz, Arecibo/ASP:
Cyclic Spectra
P. Demorest NRAO Postdoc Symposium 2009
Same thing, zoomed in:
428.75
428.8
428.85
428.9
428.95
429
0.5 0.55 0.6 0.65 0.7 0.75 0.8
0.8 0.85 0.9 0.95 1 1.05 1.1 1.15 1.2
Fre
quen
cy (
MH
z)
Pulse Phase (turns)
Time (ms)
ISM De-scattering
P. Demorest NRAO Postdoc Symposium 2009
“Snapshot” ISM impulse response h(t) determined from data:
ISM De-scattering
P. Demorest NRAO Postdoc Symposium 2009
Comparison of uncorrected and intrinsic pulse profiles:
Dynamic Spectra
P. Demorest NRAO Postdoc Symposium 2009
B1937+21, ν=430MHz, BW=4 MHz, ∆ν=0.64 kHz, T∼1 hour
Secondary Spectra
P. Demorest NRAO Postdoc Symposium 2009
2-D FT of dynamic spectrum, shows clear “arc” structure.
Secondary Spectra
P. Demorest NRAO Postdoc Symposium 2009
Contrast with methods based on traditional dynamic spectra:
Future Directions
P. Demorest NRAO Postdoc Symposium 2009
• Get scattering-corrected TOAs, improve timing
• Investigate polarimetry aspects
• Explore more strongly scattered sources
• Work towards physical ISM images
• . . .