Studying the continuum spectrum of the parsec scale jets by multi ... · parsec scale jets by...
Transcript of Studying the continuum spectrum of the parsec scale jets by multi ... · parsec scale jets by...
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Studying the continuum spectrum of theparsec scale jets by multi-frequency VLBI
measurements
Tuomas SavolainenMax-Planck-Institut für Radioastronomie; Alexander von Humboldt fellow
in collaboration withK. Wiik1, E. Valtaoja1 & M. Tornikoski2
1Tuorla Observatory, University of Turku2Metsähovi Radio Observatory, Helsinki University of Technology
EVN Symposium 2008, Bologna
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Analysis of Multi-frequency VLBI DataIntroduction
Why is it important to study the 2-D spectrum of pc scale jets?In order...
to study the jet physics: measuring the synhrotronself-absorption turnover and the component size gives youa handle on the important physical parameters: B and N0.And there is no need to assume equipartition!to predict how much SSC X-rays one can expect from thesource and from which part of the jet these come from.to test shock models by measuring the spectral evolution ofthe moving components.to study the possible absorption by circumnuclear gas.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Analysis of Multi-frequency VLBI DataIntroduction
Why is it important to study the 2-D spectrum of pc scale jets?In order...
to study the jet physics: measuring the synhrotronself-absorption turnover and the component size gives youa handle on the important physical parameters: B and N0.And there is no need to assume equipartition!to predict how much SSC X-rays one can expect from thesource and from which part of the jet these come from.to test shock models by measuring the spectral evolution ofthe moving components.to study the possible absorption by circumnuclear gas.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Analysis of Multi-frequency VLBI DataIntroduction
Why is it important to study the 2-D spectrum of pc scale jets?In order...
to study the jet physics: measuring the synhrotronself-absorption turnover and the component size gives youa handle on the important physical parameters: B and N0.And there is no need to assume equipartition!to predict how much SSC X-rays one can expect from thesource and from which part of the jet these come from.to test shock models by measuring the spectral evolution ofthe moving components.to study the possible absorption by circumnuclear gas.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Analysis of Multi-frequency VLBI DataIntroduction
Why is it important to study the 2-D spectrum of pc scale jets?In order...
to study the jet physics: measuring the synhrotronself-absorption turnover and the component size gives youa handle on the important physical parameters: B and N0.And there is no need to assume equipartition!to predict how much SSC X-rays one can expect from thesource and from which part of the jet these come from.to test shock models by measuring the spectral evolution ofthe moving components.to study the possible absorption by circumnuclear gas.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Analysis of Multi-frequency VLBI DataIntroduction
Why is it important to study the 2-D spectrum of pc scale jets?In order...
to study the jet physics: measuring the synhrotronself-absorption turnover and the component size gives youa handle on the important physical parameters: B and N0.And there is no need to assume equipartition!to predict how much SSC X-rays one can expect from thesource and from which part of the jet these come from.to test shock models by measuring the spectral evolution ofthe moving components.to study the possible absorption by circumnuclear gas.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems:Target sources are typically variable⇒ need(quasi)simultaneous multi-λ measurements. OK with VLBAAccuracy of flux density calibration. OK at ν ≤ 22 GHz,need checks at higher freqs.Image alignment. After phase self-cal, there is no absoluteposition information in the data. Can be dealt with.Uneven (u, v) coverage. Our new method alleviates thisproblem.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems:Target sources are typically variable⇒ need(quasi)simultaneous multi-λ measurements. OK with VLBAAccuracy of flux density calibration. OK at ν ≤ 22 GHz,need checks at higher freqs.Image alignment. After phase self-cal, there is no absoluteposition information in the data. Can be dealt with.Uneven (u, v) coverage. Our new method alleviates thisproblem.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems:Target sources are typically variable⇒ need(quasi)simultaneous multi-λ measurements. OK with VLBAAccuracy of flux density calibration. OK at ν ≤ 22 GHz,need checks at higher freqs.Image alignment. After phase self-cal, there is no absoluteposition information in the data. Can be dealt with.Uneven (u, v) coverage. Our new method alleviates thisproblem.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems:Target sources are typically variable⇒ need(quasi)simultaneous multi-λ measurements. OK with VLBAAccuracy of flux density calibration. OK at ν ≤ 22 GHz,need checks at higher freqs.Image alignment. After phase self-cal, there is no absoluteposition information in the data. Can be dealt with.Uneven (u, v) coverage. Our new method alleviates thisproblem.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems:Target sources are typically variable⇒ need(quasi)simultaneous multi-λ measurements. OK with VLBAAccuracy of flux density calibration. OK at ν ≤ 22 GHz,need checks at higher freqs.Image alignment. After phase self-cal, there is no absoluteposition information in the data. Can be dealt with.Uneven (u, v) coverage. Our new method alleviates thisproblem.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI Data
Problems: Uneven (u, v) coverage
VLBI networks cannot be scaled⇒ (u, v) plane coveragediffers significantly at different observing frequenciesClassic solution: match the (u, v) coverages either bythrowing away long spacings at high frequencies ortapering the data to common resolution⇒ throw away a lotof data and lose angular resolutionBroad frequency coverage, which is needed to observe thespectral turnover, exacerbates the problem!
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI DataModel-fitting Based Method
Solution: brightness distribution template
Use a priori knowledge of the source structure from thehigh(est) frequency map⇒ fit the lower frequency datawith a brightness distribution templateAngular resolution now depends on the SNR of the visibilitydata⇒ if the data have high SNR, it is possible to derivesizes and flux densities of even those emission featuresthat have mutual separations less than the Rayleigh limitSignificantly relieves the problem with uneven (u, v)coverage
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI DataHow Is It Done in Practice?
1 Form a model of the source brightness distribution at thefrequency giving the best angular resolution and goodSNR (a small number of Gaussians)
2 Transfer the model to lower frequency, keep relativepositions of the components fixed (a priori knowledge)
3 Align the model with the data by using optically thin jetfeatures (fine-tune the alignment, if necessary)
4 Fit the model letting Sν and Θ to vary5 Add new components, if new emission regions have
emerged. Merge components that have mutual separation< 1/5 of the beam size.
Repeat 2− 5 for all the frequencies.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extraction of the Spectra from Multi-λ VLBI DataHow Is It Done in Practice?
1 Assume smooth variations in component size overfrequency and estimate the size-frequency relation (a prioriknowledge)
2 Fix the size and fit the model with Sν as the only freeparameter
⇒ Final spectra
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Multifrequency VLBA Observations of 3C 273
5-epoch multifrequency VLBAmonitoring carried outconcurrently with the INTEGRALcampaign in 2003 (Courvoisier etal 2003, A&A, 411, 343)Simultaneous observations at 5,8.4, 15, 22, 43, and 86 GHz,including polarisation at everyfrequency
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Examples of the Measured Spectra
10 1000.1
110
A2
110
10 1000.1
110
B4
10 1000.1
110 10
B3
10 1000.1
110
a3
10 1000.1
110
0.1
110
a2
VLBA5−86 GHz
C1
C2 3C273
10 100
0.1
0.1
110
Core
10 1000.1
110
a1
10 1000.1
110
1 mas
A3
15 GHz43 GHz
1 mas10010
10010
100100.1
1
B2
10 100
110
0.1
A1
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Decomposition of the Core Region
43 GHz
May 11 2003
C1C2
Sum
S1Core
Elongated, rather flat-spectrum core can be decomposed into aseries of self-absorbed synctrotron components
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Evolution
May 11 2003
Feb 28 2003
C2
Consistent evolution gives confidence on the method!In some components synchotron losses dominate? Inothers adiabatic?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Evolution
May 11 2003
Feb 28 2003
C1
Consistent evolution gives confidence on the method!In some components synchotron losses dominate? Inothers adiabatic?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Evolution
May 11 2003
Feb 28 2003
B4
Consistent evolution gives confidence on the method!In some components synchotron losses dominate? Inothers adiabatic?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Evolution
May 11 2003
Feb 28 2003
B2
Consistent evolution gives confidence on the method!In some components synchotron losses dominate? Inothers adiabatic?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Evolution
May 11 2003
Feb 28 2003
B3
Consistent evolution gives confidence on the method!In some components synchotron losses dominate? Inothers adiabatic?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Spectral Turnover
May 11 2003
Feb 28 2003
Core
May 11 2003
Feb 28 2003
Core
The synchrotron peak frequency decreases as function ofdistance from the core: νm ∝ r−0.7±0.1
Confirms the composite nature of the flat radio spectrum
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Physical Parameters of the Plasma
We have measured Sm, νm, α and Θ; Doppler factor δ isavailable from VLBI monitoring (Savolainen et al. 2006,A&A, 446, 71)Assuming standard synchrotron theory (e.g. Marscher1987, in Superluminal Radio Sources):B = 10−5b(α)Θ(νm)4ν5
mS−2m
δ1+z [G]
N0 = n(α)D−1GpcΘ(νm)−(7−4α)ν
−(5−4α)m S3−2α
m
×(1 + z)2(3−α)δ−2(2−α) [erg−2α cm−3]Ure ≈ f (α, ν2/ν1)D−1
GpcΘ(νm)−9ν−7m S4
m(1+z)7δ−5 [erg cm−3]
No need to assume equipartition!Highly non-linear equations:
Need accurate input parametersMonte Carlo methods needed in estimating uncertainties
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Magnetic Field Density
log(B) vs. log(r)
May 11 2003
Feb 28 2003b = −2
b = −1
In the core, B ∼ 1 GExcluding B2,B ∝ r−1
Significant Bgradient across thejet at 1.5 mas fromthe core: Northernside ∼ 1 mG whileSouthern side∼ 50 mG
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Magnetic Field Density
log(B) vs. log(r)
May 11 2003
Feb 28 2003B = 1 mG
B = 50 mGS−side
N−side
In the core, B ∼ 1 GExcluding B2,B ∝ r−1
Significant Bgradient across thejet at 1.5 mas fromthe core: Northernside ∼ 1 mG whileSouthern side∼ 50 mG
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Gradients in B and Γ across the Jet
43 GHz VLBA monitoring: there is avelocity gradient across the jetbetween 1-2 mas (Savolainen et al.2006, A&A, 446, 71):
Northern component B2 has Γ ≈ 7Southern components B3 and B4have Γ ≈ 17Jorstad et al. (2005, AJ, 130, 1418)report a similar velocity gradient fordifferent components and differenttime; according to them knots on thenorthern side have Γ ≈ 8 and knotson the southern side have Γ ≈ 14
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Gradients in B and Γ across the Jet
Coincident gradients in B and Γ across the jet: Fast southerncomponents with Γ ∼ 17 have magnetic field density of ∼ 50 mG whilethe slow northern component B2 has Γ ∼ 7 and B ∼ 1 mG;spine-sheath structure?
Velocity field (Γ)
B2
B4
B3
Magnetic field density
May 11 2003
Feb 28 2003B = 1 mG
B = 50 mGS−side
N−side
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Finally: A word of Caution
Limitations of the spectral extraction methodThe results are sensitive to the image alignment. Needs tobe accurate!There is a limit to how far one can extrapolate in spatialfrequencies: it depends on the SNR of the data and on theproperties of the model.It is implicitly assumed that there are no spectral indexgradients over the individual components i.e. theirbrightness centroids are co-spatial at different frequencies.
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Summary
Multifrequency VLBA observations can yield high-quality2-D spectra of the jets in parsec scalesMeasured spectral evolution in 3C 273 is very consistent.Gives confidence on the method.Magnetic field density in the jet of 3C 273 was mappedwithin first 2 mas; in the core B ∼ 1 GThere are coincident magnetic field density and bulkvelocity gradients across the jet at ∼ 1.5 mas from thecore: fast southern components have stronger magneticfield than the slow northern component; Do magnetic jetlaunching models predict this?
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Extra Slides
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Simultaneous Data at 6 Frequencies
B+CA1
a3
a1
a1
C
A3A2
A1
a3
B+CA1
a1
a2
a3
a2A2+A3+a4
a5
a2
A2+A3+a4+a5
Ba4a5
C
a1
a3
A1
B
a4
A3
C
B4
B1+B3
B2
a5
a5
a4
A3A2
A2
C B2
B4
B1+B3
a2
A1
Observed on February 28, 2003
Savolainen et al. 2-D spectra of the pc scale jets
Mapping the Jet Spectrum on Parsec ScaleThe Case of 3C 273
Summary
Energy Density of the Relativistic Electrons
log(Ure) vs. log(r)
−2 −1.5 −1 −0.5 0 0.5 1−8
−7
−6
−5
−4
−3
−2
−1
0
1
2
log10
(r) [mas]
log 10
(Ure
) [er
g cm
−3]
B3
B2
log(UB/Ure) vs. log(r)
−2 −1.5 −1 −0.5 0 0.5 1−8
−6
−4
−2
0
2
4
6
log10
(r) [mas]lo
g 10(U
B/U
re)
B2
B3
〈Ure〉 ∼ 10−4 erg cm−3
Out of equipartition: Core (B dom.), B2 (particle dom.)Limit for γmin: if the source is required to be rest-mass dominatedbeyond the core, γmin must be below 10
Savolainen et al. 2-D spectra of the pc scale jets