PARTICLE ACCELERATION
AND RADIATION IN
PULSAR WIND NEBULAE
Elena Amato
INAF-Osservatorio Astrofisico di Arcetri
Collaborators: Jonathan Arons, Niccolo’ Bucciantini,
Luca Del Zanna, Barbara Olmi, Delia Volpi
PULSAR WIND NEBULAE AT A
GLANCE
PLERIONS:
SUPERNOVA REMNANTS WITH A
CENTER FILLED MORPHOLOGY
FLAT RADIO SPECTRUM (R<0.5)
VERY BROAD NON-THERMAL EMISSION
SPECTRUM (FROM RADIO TO MULTI-TEV
-RAYS)
Kes 75 (Chandra)
(Gavriil et al., 2008)
WHY PWNe ARE INTERESTING
PULSAR PHYSICS: THEY ENCLOSE MOST OF THE
PULSAR SPIN-DOWN ENERGY (
,
,
·
·
-2
)
-10
L £10 E
Lradio £10
ER
g
R
·
LPWN ³ 0.1E R
CLOSE AND BRIGHT: BEST-SUITED LABORATORIES
FOR
THE PHYSICS OF RELATIVISTIC ASTROPHYSICAL
PLASMAS
PARTICLE ACCELERATION AT THE HIGHEST SPEED
SHOCKS IN NATURE (104<<107)
COSMIC RAYS: ONLY SOURCES SHOWING DIRECT
EVIDENCE FOR PEV PARTICLES
ANTIMATTER STORAGE ROOMS: AS MANY POSITRONS
AS ELECTRONS (PAMELA AND AMS02 EXCESS?)
THE CRAB NEBULA
(WHERE WE HAVE LEARNT MOST OF WHAT WE KNOW)
SYNCHROTRON AND ICS RADIATION BY
RELATIVISTIC PARTICLES
IN INTENSE (>few x 100 BISM) ORDERED
(HIGH POLARIZATION, IN RADIO, OPTICAL
AND EVEN -RAYS, Dean et al 08)
MAGNETIC FIELD
SOURCE OF B FIELD
AND PARTICLES:
NS SUGGESTED
BEFORE
PSR DISCOVERY
(Pacini 67)
Atoyan & Aharonian 96
Gaensler & Slane 2006
MAIN OPEN QUESTIONS
(personal view)
WE KNOW THAT:
THESE ARE THE MOST EFFICIENT ACCELERATORS
OBSERVED IN NATURE AND ACCELERATION TAKES PLACE
IN THE MOST HOSTILE ENVIRONMENT
WE DO NOT KNOW:
WHAT THE ACCELERATION MECHANISM(S) IS (ARE)
IN PRINCIPLE BOTH DEPEND ON
WHERE PARTICLE
ACCELERATION EXACTLY OCCURS
POSSIBILITIES DEPEND ON:
COMPOSITION (IONS? MULT.?)
MAGNETIZATION (=B2/4nmc2)
HOW TO GET CONSTRAINTS?
DETAILED DYNAMICAL AND RADIATION MODELING
PROPOSED ACCELERATION MECHANISMS
FERMI MECHANISM:
• RIGHT SLOPE FOR OPTICAL/X-RAY PARTICLES
• VIABILITY DEPENDS ON PLASMA MAGNETIZATION
(Spitkovsky 08, Sironi & Spitokvsky 11,12)
RESONANT CYCLOTRON ABSORPTION IN ION DOPED
OUTFLOW:
• MAGNETIZATION IS NOT VERY IMPORTANT
• REQUIRES IONS DOMINANCE
• PARTICLE SPECTRUM AND EFFICIENCY DEPEND ON
FRACTION OF ENERGY CARRIED BY IONS (Hoshino et al 92,
Amato & Arons 06, Stockem et al 12)
DRIVEN MAGNETIC RECONNECTION:
• CORRECT SLOPE FOR RADIO PARTICLES
• REQUIRES EXTREMELY LARGE MULTIPLICITY
• DIFFICULT TO MAKE SELF-CONSISTENT
(Lyubarsky 03, Lyubarsky & Liverts 08, Sironi & Spitkovsky 11)
CONSTRAINING THE WIND
MAGNETIZATION AT THE
SHOCK
(Weisskopf et al 00)
(Pavlov et al 01)
Fsin2()
Bsin()
Lyubarsky 02
Bogovalov & Khangoulian 02
THE PULSAR WIND
ANALYTIC SPLIT MONOPOLE SOLUTIONS (Michel 73; Bogovalov 99)
CONFIRMED BY NUMERICAL STUDIES IN THE
FORCE FREE (Contopoulos et al 99, Gruzinov 04, Spitkovsky 06)
AND RMHD REGIME (Bogovalov 01, Komissarov 06, Bucciantini et al 06)
Kirk et al 02
(Spitkovsky 06)
STREAMLINES ASYMPTOTICALLY RADIAL
BEYOND RLC
MAGNETIC FIELD COMPONENTS:
Br1/r2
Bsin()/r
MOST ENERGY FLOWS AT LOW LATITUDES:
Fsin2()
WIND IS HIGHLY MAGNETIZED (>>1)
CURRENT SHEET IN EQUATORIAL PLANE
OSCILLATING AROUND EQUATOR IN
OBLIQUE CASE
ANGULAR EXTENT DEPENDS ON
OBLIQUITY:
MAGNETIC DISSIPATION?
(Weisskopf et al 00)
JETS AND TORI
Fsin2()
(Pavlov et al 01)
Bsin()G()
Lyubarsky 02
Bogovalov & Khangoulian 02
AXISYMMETRIC RMHD
SIMULATIONS OF PWNE
Olmi et al 2014
Komissarov & Lyubarsky 03, 04
Del Zanna et al 04, 06
Bogovalov et al 05
Del Zanna et al 04
THE PHYSICS BEHIND THEM
Velocity
Del Zanna et al 04
Magnetization
=0.03
FOR SUFFICIENTLY HIGH , EQUIPARTITION IS REACHED IN
EQUATORIAL REGION
EQUATORIAL FLOW IS DIVERTED TOWARDS HIGHER LATITUDES
A FAST CHANNEL MAY THEN FORM ALONG THE AXIS
DEPENDENCE ON  OF FLOW
VELOCITY
(Del Zanna et al 04)
=0.003
=0.03
LOWER LIMIT
ON :
BEST FIT :
10 X LARGER THAN
WITHIN 1D
MHD MODELING
(Kennel & Coroniti 84)
BUT STILL <<1
>0.01 REQUIRED
FOR
JET FORMATION
=0.01
ISSUES WITH LOW 
=0.025
LOW  LEADS TO LOWER B FIELD THAN
REQUIRED TO FIT INTEGRATED
SPECTRUM
(Volpi et al 08, Olmi et al 14 vs
Atoyan & Aharonian 96, de Jager & Harding 92)
(Olmi et al 14)
=1.5
•STEEP PARTICLE
INJECTION SPECTRA:
Mori et al 04: X2.2;
Volpi et al 08: X2.7;
Olmi+ 14: X2.9
•LARGE NUMBER OF
PARTICLES REVEALED
BY
ICS COMPONENT:
Lsynne B2 , LICS ne Uph
MAYBE  IS NOT THAT
SMALL AFTER ALL
HOOP STRESSES
DECREASED BY KINK
INSTABILITIES
(Begelman 98, Porth et al 13)
INCREASING …
=1.5
LOW  REQUIRED
TO REPRODUCE
HIGH ENERGY
MORPHOLOGY
=0.025
RECENT PROGRESS
DIMENSIONALITY OF SIMULATIONS:
•IN 3D SAME  LOOKS VERY DIFFERENT!
•KINKS MAY MESS UP AND DISSIPATE FIELD IN THE PWN
(Begelman 98)
•CONSEQUENCES ON SPECTRUM AND POLARIZATION???
(Porth et al 13)
BOTTOM LINE:
 CAN ONLY
BE LARGER
THAN
INFERRED
FROM 2D
SO WHAT?
FERMI PROCESS AT RELATIVISTIC PAIR
SHOCKS
RESULTS FROM PI
SIMULATIONS BY
Spitkovsky 08,
Sironi & Spitkovsky 09, 11
EFFICIENT ACCELERATION AT UNMAGNETIZED
SHOCKS
DIFFUSIVE SHOCK ACCELERATION
IN WEIBEL MEDIATED (UNMAGNETIZED) e+-e- SHOCKS FERMI
ACCELERATION EFFECTIVE (Spitkovsky 08)
POWER LAW INDEX OK FOR THE OPTICAL/X-RAY SPECTRUM OF
CRAB (Kirk et al 00) BUT e.g. VELA SHOWS FLATTER SPECTRUM
(Kargaltsev & Pavlov 09)
ONLY PARTICLES
FROM REGIONS
WITH 10-3
REST OF
PARTICLES
SMALL FRACTION OF
THE FLOW SATISFIES
LOW MAGNETIZATION
(<10-3) CONDITION
MOST OF THE FLOW
HAS 0.1
RESONANT CYCLOTRON
ABSORPTION IN ION DOPED PLASMA
Configuration at the leading edge
~ cold ring in momentum space
Magnetic reflection mediates
the transition
Drifting e+-e--p
plasma
B increases
Coherent gyration leads to
collective emission of cyclotron waves
Pairs thermalize to
kT~mec2 over
10-100 (1/ce)
Plasma starts
gyrating
Ions take their time:
mi/me times longer
PARTICLE SPECTRA AND
ACCELERATION EFFICIENCY
IF
•IONS CARRY MOST OF THE ENERGY: <mi/me
•WIND SUFFICIENTLY COLD: u/u<me/mi
ACCELERATION EFFICIENCY:
~few% for Ui/Utot~60%
~30% for Ui/Utot~80%
SPECTRAL SLOPE:
>3 for Ui/Utot~60%
<2 for Ui/Utot~80%
MAXIMUM ENERGY:
~20% mic2 for Ui/Utot~60%
~80% mic2 for Ui/Utot~80%
RESULTS BY Amato & Arons 06 RECENTLY CONFIRMED BY Stockem et al 12
Constraints on Crab wind parameters
Lp = LB = Le ± = (1/3)L0
(Amato et al 03)
Synchrotron from
secondaries
-rays from
0 decay
OLD CALCULATION!
TO BE REDONE FOR ICECUBE!
VARIABILITY WITHIN RCA
RCA MECHANISM HAS AN INTRINSIC MAXIMUM ENERGY:
1
1
³
m ± g ± m ig i
w c ± ³ w ci Þ
BASED ON 1D PIC OF RCA
(Amato & Arons 06)
f~.1-.9 ( and Ei/Etot)
Caveat: in 3D?
FERMI STEADY
CUT-OFF
emax
Þ
g
loss
max
mi
G
Gwind » 6 ´10 9 wind 6
me
3 ´10
RCA
g max
=
= 1.2 ´10
10
f
»2.5 ´10 9
B-4
B-3
2
B æ g max ö
» 100 f1/ 2 MeV » 100 -3 ç
MeV
9÷
10 è 2.5 ´10 ø
Tvar AND max IN THE
æ e ph ö æ f ö-1æ B ö-3 / 2
» 3ç
÷ days BALLPARK FOR FIRST TWO,
÷ ç ÷ ç
BUT THE BIG ONE…
è100MeV ø è 0.5 ø è mG ø
1/ 2
Tvar
f1/ 2
SPECIAL RELATIVITY COULD HELP:
obs
rf
emax
» G emax
obs
rf
Tvar
» G-1 Tvar
DRIVEN MAGNETIC RECONNECTION
BROAD PARTICLE
SPECTRA
WITH =-1.5 IF
AND
BUT
REQUIRES K>107
RECONNECTION BEFORE
THE SHOCK?
(Kirk & Skjaeraasen 03)
(Sironi & Spitkovsky 11)
IF RADIO PARTICLES FROM
HIGH LATITUDES LOWER K
REQUIRED BUT STRIPES?
WIND COMPOSITION
·
·
E R = k N GJ m± c Gwind
2
æ
xmi ö
ç1+
÷(1+ s )
è k m± ø
·
k=
N
·
N GJ
B2
s=
2
4 p meff neff c 2 Gwind
• IF <mi/me IONS MAY CARRY MOST OF THE WIND
ENERGY
• CYCLOTRON ABSORPTION COULD PROVIDE THE
ACCELERATION
NOTE: IN THE CASE OF CRAB THESE WOULD BE PeV IONS
THE PULSAR
MULTIPLICITY
N(g)
A BROKEN
POWER-LAW
few 105
~1.5
~104
k~106
FROM RADIO EMISSION
(e.g. Bucciantini et al 11)
A GJ DENSITY OF IONS, IF
PRESENT, WOULD DOMINATE
THE WIND ENERGY FLOW
g
~2.2 ~3x106 k~104
FROM OPTICAL/X-RAY EMISSION
(e.g. Kennel & Coroniti 84)
NO WAY FOR IONS TO BE
ENERGETICALLY IMPORTANT
EVEN IF THEY WERE THERE
RADIO ELECTRONS MIGHT BE FOSSILE(Atoyan 99)
OR ACCELERATED ELSEWHERE
RADIO EMISSION MORPHOLOGY
SHOCK ACCELERATION
+
ADVECTION
UNIFORM
INJECTION
VERY SIMILAR MAP
Olmi et al 14
RADIO WISPS
MAGNETIC FIELD
DIFFERENCE BETWEEN IMAGES AT
DIFFERENT TIMES
Olmi et al 14
• RADIO EMISSION TRACES MAGNETIC FIELD
• RADIO WISPS EVEN FOR UNIFORM PARTICLE
DISTRIBUTION
ORIGIN OF RADIO PARTICLES
=1.5
4
EMAX0.1-0.5 TeV
1<EMIN/mc2<103 10-6<n(cm-3)<10-
PSR ORIGIN FOSSILE (Atoyan 99) WITH REACCELERATION
•NO CONSTRAINTS ON CURRENT K
•IF INJECTED AT LOW ENERGY NO VIOLATION OF ENERGY CONSTRAINTS
EVAPORATION FROM THERMAL FILAMENTS (Bucciantini et al 11, Komissarov 13)
ONLY e- IN THIS CASE: NO CONTRIBUTION TO e+ EXCESS
(PAMELA: Adriani et al 09; AMS02: Aguilar et al 13; Blasi & Amato 11)
BUT BOW SHOCK PWNe (Yusef-Zadeh & Gaensler 05, Ng et al 12)….
SPECTRAL CONTINUITY
WISPS AT MULTIWAVELENGTHS
ISOTROPIC
ACCELERATION
OBSERVATIONS
(Schweizer et al 13)
Olmi et al in prep
POLAR RADIO
EQUATORIAL X
SUMMARY AND CONCLUSIONS
• PWNe ARE THE MOST EFFICIENT ACCELERATORS SEEN IN NATURE
BUT PARTICLE ACCELERATION IN A VERY HOSTILE SETTING
• UNDERSTANDING THE ACCELERATION PHYSICS REQUIRES
PINNING DOWN THE PLASMA PROPERTIES
• DETAILED DYNAMICS AND RADIATION MODELING ALLOWS TO SET
LOWER LIMIT ON WIND MAGNETIZATION
X-RAY EMITTING PAIRS MUST BE ACCELERATED IN HIGH 
REGIONS: NO FERMI PROCESS AND MAGNETIC RECONNECTION
SEEMS INCONSISTENT
•RADIO MORPHOLOGY DOES NOT ALLOW TO ASSESS THE
ACCELERATION SITE OF RADIO PARTICLES
•ACCELERATION OF RADIO PARTICLES MAYBE UNRELATED TO WIND
AND SHOCK
CURRENT VALUE OF  MUCH LOWER
ION DOMINATED WIND STILL POSSIBLE
• MULTIWAVELENGTH VARIABILITY OF INNER NEBULA SUGGESTS
DIFFERENT ACCELERATION SITES IN DIFFERENT ENERGY RANGES
OR