Observations and NLFFF Modeling of AR 10953

1
Observations and NLFFF Modeling of AR 10953
Yingna Su1,2 Collaborators A. A. Van
Ballegooijen1, E. E. Deluca1, Leon Golub1 P.
Grigis1, B. Lites3, G. L. Huang2
1. Harvard-Smithsonian Center
for Astrophysics, USA
2. Purple Mountain Observatory, China
3. High Altitude Observatory,
USA AGU/SPD, Fort Lauderdale, 05/29/2008
2
Outline

• Background
• NLFFF Modeling of (pre-flare) AR 10953
• Flux Rope Insertion Method
• Data SOHO/MDI, Hinode/SOT, KSO/H-alpha,
  Hinode/XRT
• Observations of a C8.5 Flare in AR 10953
• Data Hinode/XRT, TRACE, RHESSI, MLSO/H-alpha
• Interpretations and Conclusions

3
Background

• Existing methods for constructing NLFFFs
  (non-linear force free fields)
• Most methods extrapolating photospheric vector
  fields to the corona
• (Schijver et al. 2006).
• The method we adopted flux rope insertion
  method (van Ballegooijen 2004
• Bobra etal. 2008) which requires line of sight
  magnetograms. This method was tested by
• Bobra et al. (2008), and the model was
  constrained by TRACE observations.
• What is the 3D pre-flare magnetic configuration?
  Where and how
• is the flare initiated?
• In this work, we construct NLFFF models for the
  pre-flare state
• of AR 10953, and the model is constrained by
  multiple non-potential
• X-ray loops observed by XRT.

4
Flux Rope Insertion Method
2007-May-02 1730 UT
2007-May-02 1131 UT
2007-May-02 1459 UT
MDISOT/SP
KSO/H-alpha
XRT MDISOT/SP
Magneto

• PF model

Insert Flux Rope
NLFFF Model
Friction
Van Ballegooijen 2004 Bobra et al. 2008
5
Model Constraint X-ray Loops

• Best fit model field lines for four non-potential
  X-ray loops
• Model Free Parameters Axial flux and Poloidal
  flux of the flux rope

6
Calculation Solution Set
Table 1 The Average Distance (in solar radii) of
the projected best-fit model field lines from the
observed X-ray loops for various models with
fixed Poloidal Flux (1e10 Mx/cm).

• Best Fit Model Axial Flux7e202e20 Mx ,
  Poloidal Flux 1e9 to 1e11 Mx/cm (larger range)
  Eruption for Axial Flux 15e20 Mx
• Loop 4 may be in a non-stable state.

7
Comparison with SOT/SP Vector Magnetogram
Blue Vector Observation Black Vector Model
Best Fit Model Axial Flux7e20 and 9e20 Mx
8
Vector Magnetogram Obs. Versus Mod.
worst fit good fit
worse fit
Blue Vector Observation Black Vector Model
Best Fit Model Axial Flux7e20 and 9e20 Mx
9
Observations of C8.5 Flare
MLSO/H-alpha
TRACE/171
XRT/Ti_poly

• Filament activation (2330 UT) associated with
  the flare was seen in H-alpha and EUV, not
  X-ray.
• Two-ribbon flare unsheared-sheared-unsheared.
• Not associated with filament eruption and CME.

10
Light Curves of C8.5 Flare
EUV flare starts about 20 minutes later than
the X-ray (XRT and RHESSI) flare, Why?
11
Pre-EUV flare X-ray brightenings
XRT XRT RHESSI XRT TRACE
RHESSI Spectral fitting suggests that the pre-EUV
flare X-ray sources are mainly caused by thermal
bremsstrahlung emission.
12
Where Does the Flare Start?
flare loop
XRT 2007-May-02 230924 UT
Z14
flare loop

Reconnection site?
Current layer
At flare onset, the flare loop is unsheared. This
suggests the flare starts near the outer edge of
the flux rope.
13
Conclusions (I)

• Comparison of modeled and observed X-ray loops
  for pre-flare (C8.5) state of AR 10953
• Axial Flux 7e202e20 Mx (good constraints from
  X-ray loops)
• Poloidal Flux 1e9-1e11 Mx/cm (wide range)
• Upper limit Axial Flux15e20 Mx
• Free Energy 0.85e32 erg, Potential Energy
  9.62e32 erg
• Similar result obtained by comparisons of
  observed and modeled photospheric vector
  magnetograms.
• The axial flux (7e202e20 Mx) in the flux rope
  is far away from the upper limit (15e20 Mx) for
  eruption, which is consistent with the fact that
  no successful filament eruption occurred in this
  active region.

14
Conclusions (II)

• The X-ray brightenings appear about 20 minutes
  earlier than the EUV flare associated with a
  filament activation, which may be caused by the
  localized coronal heating.
• Open question Why does the heating not
  propagate to the chromosphere where EUV flare
  occurs?
• This flare starts from unsheared brightenings,
  unlike most two-ribbon flares (Su et al. 2007).
• Interpretation the flare starts near the outer
  edge of the flux rope, Not at the inner side or
  at the bottom as suggested in the standard flare
  model (e.g., Moore et al. 2001).

15
Thank you for your attention !