Magnetic-field production

by cosmic rays drifting

upstream of SNR shocks

Martin Pohl, ISU

with Tom Stroman, ISU, Jacek Niemiec, PAN

Supernova remnants

SNR can be resolved in TeV-band gamma rays!

TeV band (HESS) or IC keV band (ASCA) synchrotron

Supernova remnants

Young SNR are ideal laboratories

Important questions:

• Particle acceleration and magnetic turbulence

• What produces strong magnetic turbulence?

Supernova remnants

Relative drift

Magnetic turbulence

Magnetic field amplification

Observation:

Nonthermal X-raysin filaments

Requires strongmagnetic field

Magnetic turbulencerelated toparticle acceleration?

Magnetic field amplification

X-ray filaments involve strong magnetic field

Origin unknown

Fate unknown

Shock? Energetic particles?

should be turbulent

If persisting, MF must be very strong

Turbulent field should cascade away …

Not seen in radio polarimetry…

How strong and where is it?

Magnetic field amplification

X-ray filaments suggest B/B >> 1

Decay by cascading downstream! (MP et al. 2005)

Magnetic

filaments

arise!

B not determined

Magnetic field amplification

Estimate magnetic-field strength using spectra?

Depends on what electron spectrum you assume…..

Factor 3 variation

Voelk et al. 2008,

modified by MP

Magnetic field amplification

Clues from X-ray variability? (Uchiyama et al. 2007)

Energy losses

require a few

milliGauss!

BUT:

Damping gives

same timescale

Magnetic field amplification

Strong field in entire SNR?No!

RX J1713-3946:

X-ray variability a few milliGauss

(Uchiyama et al. 2007)

Produces too muchradio emission fromsecondaries

(Huang & Pohl 2008)

Magnetic field amplification

Radio polarization at rim of Tycho (Dickel 1991)

• Radial fields at 6cm• Polarization degree 20-30%

Doesn’t fit to turbulently amplified field!

Models require homogeneous radial field (Stroman & Pohl, in prep.)

Support for rapid damping?

Magnetic turbulence

Level and distribution of amplified MF unclear

What produces strong magnetic turbulence?

Upstream:

Relative motion

of cosmic rays

and cool plasma

Magnetic turbulence

Most important: Saturation process and level

• Electrons and ions don’t form single fluid

• Coupling via electromagnetic fields

• Changes in the distribution functions

• Small-scale physics dominates large-scale structure

Particle-in-Cell simulations

Magnetic turbulence

MHD simulations:

Brms >> B0

CR current assumed constant

Knots and voids in NL phase

MHD can’t do vacuum

Analytical theory (e.g. Tony Bell):

• Streaming cosmic rays produce purely growing MF

• Wave-vector parallel to streaming

Magnetic turbulence

Earlier PIC simulations: no Brms >> B0

3-D 2-D, larger system

Niemiec et al. 2008

Magnetic turbulence

• Magnetic-field growth seen

• Saturation near B ~ B0

• No parallel mode seen

but << g not maintained!

• CR back-reaction: drift disappearsB larger when CR back-reaction turned off!

Particle distributions

Establish common bulk motion

New simulations

2.5-D only!

Parameters:

Ni / NCR = 50 CR = 10

Vdrift = 0.3 c max / g,i = 0.3

See poster by Tom Stroman

New simulations

Parallel mode seen!

By

Ni

New simulations

Drifts speedsalign to 0.06 c

Overshoot indrift speed?

Im = 0.25 max

Peak MF ~ 12 B0

Decays to ~ 6 B0

Conclusions

New simulations with << g

• Parallel mode seen!

• Saturation still through changes in bulk speed

• Saturation level still at a few B0 … may be enough

• Substantial density fluctuations

Conclusions of Niemiec et al. (2008) still hold

Back-up slides

Particle distributions

Energytransferredtobackground plasma

Particle distributions

Isotropyroughlypreserved

Heating possibly artificial