1. Observation of impulsive Solar Flares with the Fermi Large Area Telescope ! Nicola.Omodei@stanford.edu Fermi LAT and GBM Collaborations

2. Nicola Omodei – Stanford/KIPAC Gamma-ray emission from the Sun The sun is a steady, faint source of gamma-rays (produced by the interactions of CR with the solar atmosphere and with the solar radiation field); High-energy emission (up to GeV) has been observed in solar flares: In the past decades, only two long-lived (hours long) gamma-ray emissions were observed by EGRET (e.g. Kanbach+93, Ryan00) It was unclear where, when, how the high-energy (HE) particles responsible for gamma-ray emission are accelerated Complex flare build up (magnetic filed structures, loops and Active Region); Magnetic fields reconnect releasing energy which accelerate particles; Particles are trapped by magnetic field lines and interact with the solar atmosphere, producing gamma-rays; Some of the particles have access to the open filed line and escape into interplanetary space; They can also be accelerated by the CME shock and some of them can travel back to the sun and also produce gamma rays;

3. Nicola Omodei – Stanford/KIPAC Fermi as solar observatory Large Area Telescope (LAT) Observes 20% of the sky at any instant, views entire sky every 3 hrs 20 MeV - 300 GeV, includes unexplored region between 10 - 100 GeV Gamma-ray Burst Monitor (GBM) Two type of detectors (12 NaI and 2 BGO) Observes entire unocculted sky Detects transients from 8 keV - 40 MeV Fermi's Excellent Solar Capabilities: High-sensitivity at energy > 100 MeV Improved angular resolution at high energy Broader energy range, covering interesting features of the X-ray to gamma-ray spectrum Wide field of view, optimal for solar flares

4. Nicola Omodei – Stanford/KIPAC Impulsive & Long Duration flares June 12, 2010: Gamma-Ray temporally associated with impulsive hard X-ray emission. Particles accelerated up to ~ 300 MeV in few seconds; Hard X-ray pile up in ACD causes suppression of the standard LAT event rate (on-ground classification of gamma-rays) Signal recovered in LAT Low Energy Events (looser selection cut) Sustained gamma-ray emission not observed Ackermann et al. 2012, ApJ...745..144A March7/8 2011: Sustained emission associated to one impulsive episode in X-rays; Accompanied by modest SEP, but very fast (~2000 km/s) CME; Continuous interaction of particles with the sun for hours after the impulsive flare; PRELIMINARY SOL2010-06-12T00:57

5. Nicola Omodei – Stanford/KIPAC Let’s focus on impulsive events: SOL2010-06- 12T00:57 Data analysis of the join GBM and LAT data provides useful information about the underlying accelerated particle distributions: –Electron Bremsstrahlung dominates at < 1 MeV energies Not a simple powerlaw: hardening followed by a roll-off (at 2.4 MeV); not compatible with transport effects alone; –Protons/ions: gamma-ray spectral features as a proxy of the accelerated ion spectrum GBM LAT Ackermann et al. 2012, ApJ...745..144A 2.2 MeV neutron Nuclear lines 511 keV e + e - Pion decay 3.3 <1.2 2.4 MeV 3.2 4.3 ComponentEnergy of gamma-ray Energy of the ionsDerived accelerated ion spectrum Neutron Capture2.2 MeV10-50 MeV 3.2 (10-50 MeV) Nuclear lines5-20 MeV50-20 MeV 4.3 (50 -300 MeV) Pions>300 MeV>280 MeV 4.5 (>300 MeV) 4.5 ~2

6. Nicola Omodei – Stanford/KIPAC Combined GBM and LAT timing studies High energy gamma-rays delayed by few seconds: –~3 seconds onset and ~10 seconds delay in the peak time; –Similar “double peaked” structure in X- and Gamma- rays; This implies: –Protons/electrons began reaching >100 MeV energies after only few seconds after ~100 keV electrons; –Build-up process, lasting approximately 10 seconds; GBMLAT Ackermann et al. 2012, ApJ...745..144A

7. Nicola Omodei – Stanford/KIPAC High Significance Detections Date Type GOES ClassCME speed 2010-06-12 Impulsive M-class<500 km/s 2011-03-07 ~15h M-class~2000 km/s 2011-06-07 ~3h M-class~1200 km/s 2011-08-04 ~3h X-Class~1000 km/s 2011-08-09 Impulsive X-Class~1700 km/s 2011-09-06 Impulsive + 3h X-Class~1000 km/s 2011-09-07 ~6h X-Class~700 km/s 2011-09-24 Impulsive X-Class <500 km/s+1500 km/s (delayed) 2012-01-23 ~9h M-class~1500 km/s 2012-01-27 ~3h X-ClassMULTIPLE >1500 km/s 2012-03-05 ~6h X-Class~1700 km/s 2012-03-07 Impulsive + ~20 h X-ClassMULTIPLE >1700 km/s 2012-03-09 ~6h M-class~1000 km/s 2012-03-10 ~3h M-class~1700 km/s 2012-05-17 ~3h M-class~1700 km/s PRELIMINARY

8. Nicola Omodei – Stanford/KIPAC Impulsive Flare detection > 20 MeV events, with lose event selection (LLE); Flare events visible on- top of the varying background (celestial and local CR); Flare of both GOES X and M classes seen; 2011 Aug 09 08:01:01 2010 June 12 00:55:06 2011 Sept 06 22:17:17 2011 Sept 24 09:34:38 PRELIMINARY

9. Nicola Omodei – Stanford/KIPAC Long duration flares March 7/8, 2011 Solar Flare - M Class March 7, 2012 Solar Flare - X5 & X1 Class High energy gamma-rays (>100 MeV) observed up to 20 hours PRELIMINARY

10. Nicola Omodei – Stanford/KIPAC A very bright Solar Flare was detected on March 7, exceeding: 1000 times the flux of the steady Sun; 100 times the flux of Vela; 50 times the Crab flare; High energy emission (>100 MeV, up to 4 GeV) lasts for ~20 hours Softening of the spectrum with time The longest lasting gamma-ray emission: March 7, 2012 PRELIMINARY LAT 1 day all sky data >100 MeV

11. Nicola Omodei – Stanford/KIPAC March 7, 2012 Normally the LAT observe the sky with a different rocking angle every orbit (+50/-50) A modified rocking profile with fixed rocking angle can also be adopted for increasing exposure

12. Nicola Omodei – Stanford/KIPAC On shorter time scales: the first hour Orbit Sunrise X5.4 flare X1 flare PRELIMINARY Hardening of the proton spectrum: new injection of particles with the X1 flare Enhancement of the BGO spectrum (nuclear lines) 2.2 MeV neutron Nuclear lines 511 keV e + e - PRELIMINARY Pion decay Sun In the LAT FoV Softening of the proton spectrum Flux > 100 MeV Proton index

13. Nicola Omodei – Stanford/KIPAC Location computed during the first orbit ~2400 seconds of dataLocalization Events corrected for the “fish- eye-effect” 95% CL error circle with systematic error added in quadrature Interval analyzed: first orbit ~2400 seconds of data Location of the gamma-ray emission consistent with the location of the Active Region 11429 Continuous acceleration of particles for ~20 hours PRELIMINARY

14. Nicola Omodei – Stanford/KIPACLocalization Events corrected for the “fish- eye-effect” 95% CL error circle with systematic error added in quadrature Interval analyzed: first orbit ~2400 seconds of data Location of the gamma-ray emission consistent with the location of the Active Region 11429 Continuous acceleration of particles for ~20 hours Localized gamma-ray emission consistent with the Active Region at later time Locations computed during multiple orbits (one per orbit) PRELIMINARY

15. Nicola Omodei – Stanford/KIPACSummary Fermi LAT has detected >100 MeV gamma-rays from solar flares, including the most energetic gamma-rays and the longest-duration emission; – Long Lasting emission flare and Impulsive flare events detected; – Joint LAT-GBM observations unveil the properties of the accelerated particles, such as spectrum and time scales of the accelerated particles; – Thanks to the LAT’s improved angular resolution, we can now localize time-extended gamma-ray emission to the site of the X-ray flare for the first time; – As the solar cycle progresses toward the maximum of Cycle 24 (mid-2013), the number of extreme energetic flares will increase; 12

16. Backup Slides

17. Possible scenario 1 Trap and precipitation of HE particles produced during the impulsive phase via magnetic reconnection (e.g., Kanbach et al. 1993)  Pitch angles of trapped particles are altered by Coloumb collisions  Trapping time is limited by Coulomb collision time  The observed duration of ~12 hours require the coronal density of <10 11 cm -3 (under calculation)  Is it possible to explain the LAT rising profile? Yokoyama & Shibata 2001

18. Possible scenario 2 Precipitation of HE protons accelerated stochastically within the loop (2 nd order Fermi) (e.g., Ryan & Lee 1991)  MHD turbulence work as a scatter, and particles are stochastically accelerated  Assuming the loop size of 10^5 km and Alfven velocity of 1000 km/s, MHD turbulence is maintained only for ~100 s, unless there is an energizing process in flaring loops  How the MHD turbulence is maintained? ( Is it possible to explain the long-lived signature?) SOHO EUV image © NASA MHD turbulence

19.  Maximum energy of protons  Acceleration time of 1 GeV proton  The fast CME-driven shock can easily and immediately accelerate protons up to >300 MeV  Note that the CME-driven shock proceeds away from the Sun  Is it possible to explain the LAT rising profile? Possible scenario 3 Return of protons accelerated by CME-driven shock (1 st order Fermi) (Murphy et al. 1987, Cliver et al. 1993) Cliver+1993 20