Hard X-ray Footpoint Motionsin connection to the

Reconnection Process

Chang Liu

Collaborators:Jeongwoo Lee, Rui Liu, Ju Jing,Na Deng, Haimin Wang3-Aug-10, RHESSI 10th workshop, Annapolis, MD

Types of HXR Footpoint Motion

Bogachev et al. (2005)

Both Yohkoh (e.g., Sakao et al. 1998; Bogachev et al. 2005) and RHESSI (e.g., Gan et al. 2008; Yang et al. 2009) observations reveal three fundamental types of HXR footpoint motion:

1. perpendicular away from the neutral line (NL) — standard 2D flare model2. anti-parallel motions along the NL also, the early converging motion (Ji et al. 2006) — relaxation of sheared fields3. motions in the same direction along the NL — motion of the acceleration region in the corona

Depending on specific magnetic field structure, two or more types of motions can occur in a flare.

This talk concentrates on the type 3, which may represent a non-standard behavior missing in the 2D model.

1. (13%)

2. (26%)

3. (35%)

Krucker et al. (2003) studied the 2002 Jul 23 X4.8 flare, where one footpoint moves along the NL and generally correlates with HXR flux. Other two do not.

Grigis & Benz (2005) studied 2002 Nov 9 M4.9 flare, where its two HXR footpoints move smoothly along the NL without correlation with HXR flux. They proposed a moving trigger along the NL similar to opening of a zipper.

Lee & Gary (2008) presented a generalized framework in which both the parallel and perpendicular motions represent the magnetic flux change in the corona.

Recently, the parallel motion was related to the phenomenon called asymmetric filament eruption (Grigis & Benz 2005; Tripathi et al. 2006; Liu et al. 2009)

Parallel Motion of HXR Footpoints along the NL

Courtesy R. Liu

Only one end of the filament erupts upward with the other end still anchored

Sequential magnetic reconnection occurring along the NL (important clue!)

2005 Jan 15 X2.6 Flare at NOAA AR 10720

This event features the followings:

Multiple (Four) stages of HXR footpoint evolution, especially parallel motion along the NL.

A compact M1.0 flare precedes the X2.6 flare, along with an asymmetric while failed filament eruption.

The X2.6 flare occurred half an hour later, along with a fast halo CME.

We discuss the implications of the HXR footpoint motions on the eruption using a non-linear force-free field (NLFFF) extrapolation from the active region.

M1.0X2.6

http://sprg.ssl.berkeley.edu/~tohban/browser

BBSO high-resolution (~0.6”/pixel, 1-2 mins cadence) Hα-0.8 Å images

NLFFF modeling using a BBSO DVMG pre-flare vector magnetograma. Polarization saturation alleviated by filling central umbra fields using MDI datab. Azimuthal ambiguity resolved using the “minimum energy” method (Metcalf 2006)c. Projection effects removed via transforming to heliographic coordinatesd. Lorentz forces & torques minimized by preprocessing (Wiegelmann et al. 2006)e. NLFFF model constructed within a box of 248”x248”x248” using the weighted optimization method (Wiegelmann 2004) adapted to BBSO data (Jing et al. 2009)

RHESSI:40-100 keV HXRs; integration time 20 s; CLEAN images (grids 3-9)

SOHO/EIT 195 Å images (5.26”/pixel, ~12 mins cadence)

2005 Jan 15 X2.6 Flare: Data Set and Reduction

NOAA/SWPC

Before the X2.6 event, an M1.0 flare was triggered by the asymmetric eruption of filament f from west to east.

The M1.0 is most probably a compact flare due to its motionless flare kernels.

Two converging flare kernels N1/S1 of the subsequent X2.6 flare originated from those of the M1.0 flare.

2005 Jan 15 X2.6 Flare in Hα blue wing

The asymmetric eruption of filament f was also seen in EUV images.

It’s most probably a failed eruption because (1) overarching fields seem not to be immediately opened; (2) M1 is confined.

Instead, overarching fields expand during 22:00 to 22:24 UT, leading to the subsequent fast halo CME and the X2.6 flare.

2005 Jan 15 X2.6 Flare in EUV 195 Å

We divide the HXR source evolution into four phases II-V:II: N1/S1 converge mainly in the direction parallel to the NL;III: N1/S1 move side-by-side parallel to the NL;IV: N1 move perpendicular away from the NL;V: N2/S2 show up in distinct, eastern section of the NL.

2005 Jan 15 X2.6 Flare: HXR Footpoint Motions (I)

Phase II: Hα kernels and HXR footpoints converge (~20-30 km/s ) mainly parallel to PIL, with decreasing angle of flare shear.

Phase III: Parallel motion of N1/S1 along the NL. For N1, the speed (up to ~100 km/s) has a peak-to-peak correlation with its HXR flux. The overall Pearson correlation coefficient between them is ~0.71. S1 behaves similarly.

Phase IV: Standard separation motion of N1 away from the NL.

2005 Jan 15 X2.6 Flare:HXR Footpoint Motions (II)

2005 Jan 15 X2.6 Flare: Eruption Mechanism

Magnetic breakout model and tether-cutting reconnection model are not considered because there are no obvious chromospheric brightenings before the flare, either in remote regions or within the initial flaring site.

Kink instability is excluded here because writhing motion of the filament is not apparent.

For this AR with continuous flux emergence, the torus instability (Kliem & Török 2006) is considered most appropriate.

Fan & Gibson (2007)

The likelihood of eruption is gauged by decay indexwith a critical range 1.5-2.0,

when the outward hoop force of the flux rope can no longer be balanced by the confining force.

•NLFFF model reveals distinct field structure of the four phases of HXRs. •We calculate the decay index n along the NL, and find (1) a low value in the filament region (failed eruption); (2) subsequent reconnection evolves towards regions of increasing n, i.e., weaker magnetic confinement.

2005 Jan 15 X2.6 Flare: Magnetic Field Structure(pre-flare NLFFF model + Flare timing using HXR footpoint motion)

1. We find in the 2005 Jan 15 X2.6 flare in NOAA AR 10720 that (1) The parallel motion of HXR footpoints shows a good correlation (r=0.71) with the HXR flux, suggesting that the parallel motion not only maps the propagation of the trigger but also represents the primary flare energy release, in complement to the standard 2D model.

(2) An asymmetric eruption progressively opens the overlying field from one end of the NL to the other, resulting in magnetic reconnection proceeding along the NL. A new intriguing property is that the eruption tends to progress toward the region of weaker magnetic confinement.

2. Studying HXR footpoint motion with an aid of modeling of (pre-)flaring magnetic field structure can help to ultimately understand the reconnection process associated with flares/CMEs.

Main Conclusions