1. Loop-top altitude decrease in an X-class flare A.M. Veronig 1, M. Karlický 2,B. Vršnak 3, M. Temmer 1, J. Magdalenić 3, B.R. Dennis 4, W. Otruba 5, W. Pötzi 5 1 Institute of Physics/IGAM, University of Graz, A-8010 Graz, Austria 2 Ondřejov Observatory, Czech Academy of Sciences, Czech Republic 3 Hvar Observatory, Faculty of Geodesy, HR-10000 Zagreb, Croatia 4 NASA Goddard Space Flight Center, MD 20771, U.S.A. 5 Kanzelhöhe Solar Observatory, University of Graz, A-9521 Treffen, Austria

2. Introduction  Recent RHESSI X-ray observations provided evidence for a loop-top altitude decrease during the early phase of a flare Sui & Holman 2003, Sui et al. 2004, Krucker et al. 2003, Liu et al. 2004 Afterwards the behavior changed to the generally observed upward growth of the flare loop system.  Here: Analysis of the flare loop system of the X3.9 flare on 2003 November 3 Aim: Extract further observational details (LT kinematics; LT plasma evolution) during the time of altitude decrease & Modelling in the frame of a collapsing magnetic trap Data: RHESSI (Reuven Ramaty High Energy Solar Spectroscopic Imager) GOES-12/SXI (Soft X-ray Imager) SoHO/EIT (Extreme-ultraviolet Imaging Telescope) Kanzelhöhe H 

3. GOES 3-day plot: 2003 November 2–5 X4 flare from NOAA 10488

4. Magnetic evolution of NOAA 10488 21 Oct – 4 Nov 2003 MDI Magnetograms + Flare locations Courtesy of Peter T. Gallagher

5. RHESSI soft and hard X-ray lightcurves

6. RHESSI image sequence Impulsive Phase: 09:47 UT – 10:01 UT Images: 12–15 keV Contours: 70–100 keV  2 footpoints at high energies + Loop-top source at low energies

7. KanzelhöheH  image sequence Kanzelhöhe H  image sequence Preflare, main & decay phase: 9 – 13 UT

8. H  loops in flare decay phase H  images from Public Observatory Rimavska Sobota (Slovakia) 12:46 UT 14:55 UT

9. GOES-12 SXI image time series GOES-12 SXI image time series Full day 2003 Nov 3: 2 X-class flares from same AR 10488

10. Evolution of flare and post-flare loop system Loop Height vs. Time

11. Centroids of RHESSI FPs and LT source on MDI continuum image Evolution of RHESSI footpoints and loop-top

12. Kinematics of RHESSI sources Impulsive phase:  Kinematics of LT & FPs is consistent  LT: higher energies at higher heights Time of LT altitude decrease:  Spectral change a) increase of T (thermal em) and/or b) spectral hardening (non-th)

13. Kinematics of RHESSI & SXI loop-top source

14. Loop-top altitude decrease: Kinematics Results from linear fits: Energy Initial Altitude Final Altitude Downward velocity (keV) (Mm) (Mm) (km/s) RHESSI 25 - 30 13.8 7.3 45 RHESSI 20 - 25 12.1 7.7 30 RHESSI 15 - 20 11.7 7.5 29 RHESSI 10 - 15 10.1 8.2 14 SXI 8.6 7.0 12  Distinct relation with X-ray energy (consistent with results of Sui et al. )

15. RHESSI spatially integrated spectra

16. Summary of observational results  At the very beginning the LT altitude decreased. The effect is stronger for higher X-ray energies (cf. Sui & Holman 2003, Sui et al. 2004). Decrease up to 50% of the initial height, mean „downward“ velocities up to 45 km/s.  Impulsive phase: LT source moved upward and FPs separated. At higher energies the LT source is located at higher altitudes.  Consistent with the standard reconnection model in which the energy release occurs higher and higher in the corona.  Simultaneously the LT spectrum changes. RHESSI spectra indicate thermal emission of a „superhot“ (Lin et al. 1981) plasma (35  45 MK) before the acceleration of fast particles!  X-ray and H  observations are indicative of very high densities in LT. Hot LT plasma at time of LT altitude decrease: n  10 10 cm  3 Hot LT plasma peak density: n  3·10 11 cm  3 H  post-flare LT plasma density: n  10 12 cm  3 (H  loop in emission against the solar disk: Heinzel & Karlický 1987, Švestka et al. 1987)

17. Discussion LT altitude decrease: Intrinsic process of magnetic reconnection?  Relaxation of newly reconnected field lines („field line shrinkage“) to form closed loops (Švestka et al. 1987, Lin et al. 1995, Forbes & Acton 1996, Lin 2004)  Plasma processes in a collapsing magnetic trap configuration (Somov & Kosugi, 1997, Karlický et al.)  Push down of the lower bound of the current sheet during the change from slow X-point to fast Petschek reconnection (Sui et al. 2004)  Plasma processes in a collapsing magnetic trap configuration (Somov & Kosugi, 1997, Karlický et al.)

18. Collapsing magnetic trap: Model Model based on Karlický and Kosugi (2004) Betatron mechanism: acceleration & heating (Brown & Hoyng 1975, Emslie 1981, Karlický & Kosugi 2004) Farady‘s law

19. Collapsing magnetic trap (1-D): Results X-ray Intensity for thermal bremsstrahlung as function of height at 3 times

20. Collapsing magnetic trap (1-D): Results Time evolution of emission centroid (for thermal bremsstrahlung) Height (t) Velocity (t)

21. Comparison of model results & observations Collapsing magnetic trap model can account for:  Altitude decrease of emission centroid (for thermal and nonthermal X-rays)  Structuring of X-rays with energies: Emission source of higher energy X-rays are located above lower energies – works only for the thermal case! For 2003 Nov 3 flare this is in agreement with RHESSI spectra  Higher downward „velocities“ for higher X-ray energies