Nonthermal Spectral Flattening in Early Impulsive Flares

1
Nonthermal Spectral Flattening in Early Impulsive
Flares
SH13A-0291

• Linhui Sui (CUA/NASA GSFC)Linhui.sui_at_gsfc.nasa.g
  ov,Gordon D. Holman Brian R. Dennis(NASA
  GSFC)

2
Motivation

• Search for evidence of a low-energy cutoff in
  solar flare nonthermal electron distributions.
• Estimate the total energy in nonthermal electrons
  with greater certainty.
• Study particle acceleration mechanisms based on
  the characteristics (such as the low-energy
  cutoff, total energy) of the accelerated
  electrons.

3
Normal Flares vs. Early Impulsive Flares
Normal Flares have significant plasma preheating
before the impulsive phase. Therefore, the
thermal component dominates the low-energy part
of the photon spectra. Even if there exists a
high low-energy cutoff to the nonthermal
electrons, flattening in the X-ray spectrum will
be difficult to detect. (Before RHESSI, the high
threshold of spectrometers and insufficient
spectral resolution also kept us from seeing the
flattening.)
Early Impulsive Flares have no apparent or very
limited preheating as seen in X-rays. The
thermal component is relatively weak at low
energies. The nonthermal power-law component can
be seen to down to low energies. These flares
offer the advantage that evidence for a
low-energy cutoff in the nonthermal electron
distribution can be identified down to lower
energies in the X-ray spectrum.
No preheating
Preheating
4
Early Impulsive Flares Observed by RHESSI in 2002

• 136 early impulsive flares were found in 2002.
• Criteria for early impulsive flares (1) the
  25-50 keV flux increases significantly in the
  flares(2) the time delay of flux increase at
  25-50 keV relative to 12-25 keV is less than 30
  s.
• 34 flares (1 GOES X-class flare, 9 M-class
  flares, 24 C-class flares) with a significant
  25-50 keV flux increase (peak to background flux
  ratio gt 3) were selected for this analysis.
• Multiple spectra (with 4s integration time) in
  each flare are fitted with a model consisting of
  an isothermal bremsstrahlung component, and a
  nonthermal thick-target bremsstrahlung component
  which is produced by a power-law electron
  distribution with a low-energy cutoff.

5
Flares Showing Spectral Flattening
Notes (1) The minimum cutoff energies are not
uniquely determined because of the lack of
spectral flattening in those time intervals.
(2) The radiated energy is the total radiated
energy from the soft X-ray emitting plasma
calculated using the GOES Temperature and
emission measure. (3) If the HXR spectral
flattening disappears after albedo correction,
such flattening is marked as likely albedo
flattening.
6
Case 1 June 02, 2002 (C9.4 flare)
ECutoff
flattens
flattens
no flattening
Cutoff 38 keV
Cutoff 30 keV
ENonthermal gt Eradiated 1029 ergs
2.9 X1028 ergs
7
Case 2 Nov 14, 2002 (C5.5 flare)
ECutoff
flattens
flattens
no flattening
Cutoff 38 keV
Cutoff 23 keV
ENonthermal gt Eradiated 3.8 X1029
ergs 9.3 X1028 ergs
8
Case 3 Feb 26, 2002 (C9.6 flare)
ECutoff
flattens
no flattening
flattens
Cutoff 41 keV
Cutoff 38 keV
The time profile of the cutoff energy seems to
have two peaks. The second peak preceded The
HXR peak by 12 s.
ENonthermal gt Eradiated 1.6 X1030
ergs 6.2 X1029 ergs
9
Spectral Fitting Results

• 10 of the 34 selected flares were found to show
  spectral flattening below 20 to 30 keV from the
  steeper power-law distribution at higher
  energies.
• The low-energy cutoff varies with time. In
  general, it correlates with the HXR (gt 25 keV)
  flux for most of the flares.
• Electron distributions with a sharp cutoff (i.e.,
  no electrons below Ecutoff) and a flat cutoff
  (i.e., constant flux below Ecutoff) fit the
  spectral data equally well.
• The estimated energy in nonthermal electrons is
  significantly larger than the radiated energy for
  all the flares.

10
Albedo

• Observed photon flux Iobs Isource Ialbedo
• Hard X-rays incident on the photosphere with
  energies in the range 10 - 100 keV are
  backscattered due to Compton collisions with
  electrons (Bai Ramaty 1978). The reflectivity
  has a maximum around 30-40 keV. So albedo will
  produce a broad hump in the photon spectra,
  similar to the effect of the low-energy cutoff
  (Zhang Huang, 2004).
• The RHESSI spectral analysis tool (i.e., OSPEX)
  offers an angle-dependent Greens function
  correction for albedo (Kontar et al. 2005).
• The obtained low-energy cutoff usually decreases
  by 1 to 3 keV after the isotropic albedo
  correction is applied.
• After the albedo correction, assuming isotropic
  emission, the spectral flattening in 2 flares
  goes away (marked as likely albedo flattening
  in the flare list table).
• The significance of the albedo contribution to
  the photon spectra needs to be investigated.

11
Isotropic Albedo Correction
Before Correction
After Correction
flattens
Spectral flattening remains in theJune 02, 2002
flare
flattens
Cutoff 36 keV
Cutoff 33 keV
Spectral flattening disappears in theSep 29,
2002 flare
flattens
no flattening
12
Conclusions

• RHESSI spectra of 34 early impulsive flares have
  been studied. 10 of them show spectral flattening
  below 20-30 keV.
• Because of the relatively weak thermal
  contribution, the existence of a low-energy
  cutoff (varying from 12 to 46 keV) is the likely
  explanation for the flattening.
• In general, the cutoff energy correlates with the
  HXR flux.
• Albedo from isotropically emitted photons cannot
  explain all the spectral flattening. Albedo from
  anisotropic emission could be a possibility.
  However, lack of flattening at high HXR fluxes
  does not support this possibility.

13
Case 3 Feb 25, 2002 (C4.1 flare)
flattens
flattens
no flattening
Cutoff 35 keV
Cutoff 46 keV
ENonthermal gt Eradiated 3.8 X1029
ergs 2.4 X1028 ergs