Institute of Astronomy,

1
Observational Constraints on Flare
Acceleration Working Group 5 Arnold Benz
Institute of Astronomy, Radio Astronomy and
Plasma Physics Group
Eidgenössische Technische Hochschule Zürich Swiss
Federal Institute of Technology, Zürich
2
"Flares are unique in the astrophysical realm for
the great diversity of diagnostic data that are
available." Miller et al. 1997
3
MHD perspective 1. Cusp 2. Two interacting
loops 3. Statistical flare
But Particle acceleration is not MHD
Reconnection in Earth's magnetotail is
collisionless
4
Acceleration ???
Acceleration ???
Forbes
5
1. Constraints on Acceleration
Radio observations Interplanetary
particles X-rays Gamma rays
6
Radio Observations

• Diagnostics on high-frequency phenomena
• Diagnostics of non-thermal velocity distributions
• Diagnostics on high-frequency waves in plasma
• Radio emission at 100 GHz and beyond
• ? highly relativistic electrons in gt 1000 G
  magnetic fields

7
Interplanetary Particles

• Ion acceleration up to 100 MeV per nucleon
• Enrichments of Ne, Mg, Si, Fe relative to C, N, O
  over coronal ratios
• Enrichment of 3He relative to 4He

8
Hard X-Rays

• Flare energy release up to 1032 erg, largest part
  into kinetic particle energies.
• Number of accelerated electrons up to 1038 per
  flare (number problem!)
• Significant number of electrons at tens of MeV
• Acceleration in lt 0.1 s (SMM, BATSE)

9
Gamma Rays

• Ratio electrons/ions varies in different flares
• Ion accelerated up to 10 GeV per nucleon
• Footpoints at different locations from X-rays

10
2. Signatures of Accelerated Electrons
Incoherent emissions - bremsstrahlung (soft
and hard X-rays) - synchrotron,
gyro-synchrotron, cyclotron emission (dm and cm
radio) Coherent emissions (radio waves via
wave-wave coupling) - bump-on-tail
instability of electron beams, Langmuir waves -
loss-cone instability of trapped electrons,
upper-hybrid waves or electron cyclotron waves
(maser) Others?
11
A. Incoherent emissions
non-thermal
thermal
RHESSI much energy in non-thermal, thick
target electrons
?
Te
P. Saint-Hilaire
12
?Ekindt
Eth
Saint-Hilaire B., 2005
13
spectral index
flux
Grigis B.
14
P. Grigis
15
Battaglia et al. 2005
16
B. Coherent emissions - Electrons have
anisotropic velocity distribution (free energy) -
High-frequency plasma waves in resonance with
energetic electrons - Plasma waves couple to
propagating radio waves 1. Bump-on-tail
instability
f(v?)
Wave growth E12 Eth2 e-2?t ? ?p
vne ? (vb/?v)2 (nb/ne)?
Resonance vb ?/k
vb
v?
17
Coherent emissions by velocity space instabilities
v
T
loss cone
Result high-frequency waves that easily couple
into radio waves
v?
bump on tail
Expect good correlation between non-thermal
electrons and radio emissions
18
3. Proposed Mechanism for Acceleration
1. Shock acceleration 2. Parallel electric
field 3. Stochastic waves and fields
Reviews by Ramaty, 1980 Heyvaerts,
1981 Vlahos et al., 1986 Melrose,
1990 Benz et al., 1994 Miller et al.,
1997
19
Proposed Mechanisms
1. Shock acceleration - flare shock, CME
front (secondary phase) - reconnection jet
termination shock (rare, controversial)
1. Shock acceleration 2. Parallel electric
field 3. Stochastic waves
20
radio-quiet flare
6 12 keV
12 25 keV
25 50 keV
50 100 keV
GOES class M1.0
21
Frequent Absence of Radio Emissions

• 17 of flares gt C5.0 have no assoicated
  coherent radio emission (excl. type I 22) .
• Only 33 of the flares have classic type III
  events at meter waves.
• Detailed correlations with decimeteric spikes and
  pulsations exist, but rare.

Suggests bulk heating (energization) rather than
acceleration.
Flares are often well contained.
Suggests multiple acceleration sites per flare.
Benz et al. 2005
22
Proposed Mechanisms
1. Shock acceleration 2. Parallel electric
field 3. Stochastic waves
23
Proposed Mechanisms
2. Parallel electric field - reconnection
(?B/?t) (number and energy problems) -
current interruption ( I const.) (auroral
zone)
24
Proposed Mechanisms
1. Shock acceleration 2. DC electric field 3.
Stochastic waves
25
Proposed Mechanisms
3. Stochastic waves and fields - resonant waves
(s 1) (heavy ions) - transit time damping
(s0) (electrons) - stochastic parallel
electric fields (non-resonant)
26
Resonant Waves
Resonance ? sO k v 0 Wave
damping sO/v- ?/?p- k?/?p
f(p,p -) lt 0 Quasi-linear diffusion
?lt f gt ? ?
?t ?v
?v Diffusion coefficient D
W(k,t)k?k/v
(D lt f gt)
?p2 mnv
27
Transit Time Damping (s 0)
Second order Fermi acceleration - Particle
gains energy in head-on collisions, but loses
in overtaking collision - Overtaking collisions
are more frequent - In second order, a particle
gains more than it loses.
Stochastic Parallel Electric Fields
- Consider random, stationary electric fields. -
Particles (electrons and ions) move randomly and
interact. - Not second order Fermi
acceleration - Particle gains energy in qvE,
but loses in antiparallel interactions - Both
interactions are equally frequent - Particle
gains energy by stochastic diffusion in energy
space.
Different processes! ? different diffusion
constants? ? different observational
signatures?
28
Observational Constraints on Flare Particle
Acceleration

• Absence of radio emission in 17 of flares does
  not support violent acceleration processes, such
  as single shocks or single DC fields.
• Consistent with heating processes (bulk
  energization).
• RHESSI observations show that flares start with
  soft non-thermal spectrum. In the beginning it is
  difficult to distinguish from a thermal spectrum
  (? 8).

4. The spectrum of non-thermal electrons gets
harder with flux of non- thermal electrons
both in time during one flare, as well as with
peak flare flux (Battaglia et al.
2005). 5. The evidence supports stochastic bulk
energization to hot thermal distribution
and, if driven enough with power-law wings.

29
Open Questions

• What acceleration mechanism?
• How to prove stochastic acceleration?
• Magnetic (MHD) or parallel electric turbulence?
• Observable differences?
• Why power-law energy distribution?

30
(No Transcript)
31
(No Transcript)