High Energy Emission from Gamma-Ray Bursts The 3rd International Workshop on the High Energy cosmic-Radiation Detection Facility High Energy Emission from Gamma-Ray Bursts Xuefeng Wu Purple Mountain Observatory 2016.1.18-21, Xi’an

Basic Features of GRBs -- detection rate: -- temporal features 1-2 events per day by CGRO/BATSE -- detection rate: -- temporal features -- spectral features Profiles Complicated Durations 5 ms ─ 10 s ─ 5×103 s Variability 0.1ms ─1ms , No repetition -- spatial features Photon Energy 1 eV ─ 1 MeV ─ 30 GeV Non-thermal: N(E) ∝ E-α High Energy Tail: no cutoff above 1 MeV Fluence: 10-10 ─ 10-8 J /m2 log Isotropic Lack of weak GRBs GRB 940217: 低能波段只持续了约180秒，其间观测到约10个高能光子，高至4GeV 但1.6 小时之后仍观测到高能光子（200MeV －20GeV） 图中的粗黑线表示地球遮挡无法观测的时间 GRB 941017: 观测到高达200MeV的幂律谱成份（Gonzalez et al. 2003, Nature, 424, 749） GRB 970417A: 观测到有0.1TeV的光子 （Atkins et al. 2000, ApJ, 533, L119

Gehrels, Piro & Leonard 2002, Scientific American

Gehrels, Piro & Leonard 2002, Scientific American

Why GeV-TeV emission so important? 10-4 10-5 10-6 10-7 10-8 10-9 10-10 Zhang, S. N., et al., arXiv:1407.4866 radiation mechanism for >GeV emission particle acceleration responsible for the >GeV emission GRB physics (Lorentz factor, jet composition, emission site)

High energy emission from GRBs: Pre-Fermi era Little is known about GRB emission above ~100 MeV EGRET detections from a few GRBs: GRB941017: HE extra component (up to 200 MeV), with different temporal evolution in. Inconsistent with the synchrotron model ! GRB940217‏: GeV photons detected up to 90 minutes after trigger Pre-Fermi Record holders: GRB990123: highest measured Eiso ~ 2x1054 ergs GRB971214: highest lower limit on bulk Lorentz factor: >410 AGILE observation of GRB080514B Need for more data… GRB940217 (Hurley et al. 94) GRB941017 (Gonzalez et al. 03) GRB080514B, AGILE (Giuliani et al. 08) -18 to 14 sec 14 to 47 sec 47 to 80 sec 80-113 sec 113-211 sec BATSE - LAD EGRET - TASC

The Fermi Observatory Spacecraft : Large Area Telescope (LAT) : Low-Earth Circular Orbit (altitude 550 km) with 28.5° inclination Large Area Telescope (LAT) : Energy range : 20 MeV to >300 GeV Large field of view : ~2.4 sr at 1 GeV Full sky coverage every 3 hours Localization, spectroscopy and GRB trigger capabilities (on board and ground) Gamma-ray Burst Monitor (GBM): Full unocculted sky coverage : >9.5 sr On board triggers 8 keV to 40 MeV 12 NaI (8 keV to 1 MeV) Localization (on board & ground) Spectroscopy 2 BGO (200 keV to 40 MeV) Spectroscopy

GRBs detected with Fermi (2rd GRB catalogue by GBM, 2008-2012) GBM: 250 GRB/yr LAT: 10 GRB/yr

I. Delay High-Energy Emission onset (>100 MeV) Case of the long GRB 080916C Abdo et al. 2009, Science 323, 1688 Case of the short GRB 090510 Abdo et al. 2009, Nature 462, 331 The first LAT peak coincides with the second GBM peak Delay in HE onset: ~4-5 s The first few GBM peaks are missing but later peaks coincide Delay in HE onset: ~0.1-0.2 s

LAT Fluence vs GBM Fluence Comparable LE and HE gamma-ray outputs for short GRBs Long GRBs seem to emit ~5-20 times less at HE than at LE w.r.t. short GRBs 10

II. Long Lived GeV Emission Case of GRB 090926A Case of GRB 090328 t1.690.0 5σ Detection up to 4800s 5σ Detection up to 1600s

II. Long Lived GeV Emission Case of GRB 090510 (De Pasquale et al 2010, ApJL 709, 146) t1.380.07 LAT emission until 200 s No spectral evolution (photon index -2.1 ± 0.1) Fit of the SED at T0 + 100s >4.5 σ break at HE‏ beta2 = 0.62 +0.08 -0.06 beta3 = 1.14 +0.10 -0.09 Eb2 in [10 – 130] MeV Forward shock model can reproduce the spectrum from the optical up to GeV energies! (non thermal synchrotron emission from the decelerating blast wave) Extensions needed to arrange the temporal properties

Synchrotron LAT afterglow scenario ? (Kumar & Barniol Duran 09, 10) Pros Can easily explain the simple decay Can explain the delayed onset as the onset of the HE afterglow The flux level matches the observations Cons Maximum synchrotron energy is ~50 MeV in the shock rest frame (Bohm acceleration approximation) Observer frame： 50MeV*Γ/(1+z), where >10 GeV photons strongly challenge the synchrotron scenario (e.g. Piran & Nakar 10; Wang et al. 13)

LAT afterglow by inverse Compoton scattering? Liu, Wang, Wu 2013, ApJL

Hadronic model? Hadronic model: require very high isotropic-equivalent proton energies >10^55 erg/s (Asano & Meszaros 09)

III. Spectral Evolution Case of GRB 080916C (Abdo et al. 2009, Science 323, 1688) Alpha -1.02+/-0.02 Beta -2.21 +/-0.03 Epeak 1170+/-142 keV Amp. 0.0354+/-0.001 photons/s-cm2-keV REDUCED CHISQ = 0.963, PROB = 0.698 Consistent with a single Band function from 10 keV to 10 GeV. Global soft-hard-soft evolution.

IV. Additional Spectral Component Case of the short GRB 090510 Ackermann et al. 2010, ApJ 716, 1178 Case of the long GRB 090902B Abdo et al. 2009, ApJL 706, 138 Spectral deviation from the standard Band function adequately fit with an additional power law in long and short GRBs. The Extra-PL is usually not present during the all burst duration. Usually, the extra-PL over-power the Band function at low (< ~tens of keV) and high (> ~tens of MeV) energy. Possible PL break in a time resolved spectrum of GRB 090902B.

Additional Spectral Component in GBM Data Only Case of GRB 090227B (Guiriec et al, ApJ) Band (Cstat: 699/607 dof) Cutoff PL + PL (Cstat: 689/606 dof) νFν (ph/keV/cm2/s) α β Epeak σ 10000 1000 100 10 6 -6 Energy (keV) PL Cutoff PL Cutoff PL+PL prefered over the standard Band function usually used The additional component dominates the standard “Band function” at both low and high Energy 18

High-Energy Spectral Cutoff in the Extra-PL Component of GRB 090926A The extra-PL overpower the standard Band function above ~10 MeV. Existence of a 6σ spectral cutoff at ~1.4 GeV in the extra-PL. Break shape not constraint.

Lower Limit on the Jet Lorentz-Factor : Γmin Compactness problem : GRBs have a high luminosity in the gamma-ray energy range (Liso ~ 1050-1053 erg/s). Small emission region (computed from the variability time-scale observed in the light curves). Non thermal gamma-ray emission should be blocked due to the γ- γ pair production opacity. A relativistic jet reduces the photon seed population above the pair production threshold (compatible with the observations) Estimation of the jet Lorentz-Factor lower-limit : - Hypothesis : uniform, isotropic and time-independent seed photon field. More realistic models (i.e. Granot, 2008) give significantly lower values (~3 times). Sylvain Guiriec – Workshop GRBs,- Toulouse, 2010

Lower Limit on the Jet Lorentz-Factor : Γmin Case of the short GRB 090510 Result Summary Lower limit of the bulk Lorentz factor >900. GRB 090926A : it is the first direct measurement of the bulk Lorentz factor if the high-energy power-law cutoff is due to the γ-γ pair production opacity => Γ ~ 200-700 (model dependent).

Contraints on the Lorentz Invariance Violation Case of GRB 090510 (Abdo et al. 2009, Nature 462, 331) Some QG models predict a violation of the Lorentz invariance (Vph(E) ≠ c) The LAT is mostly sensitive to linear variation (n=1) and maybe quadratic variations (n=2) Sylvain Guiriec – Workshop GRBs,- Toulouse, 2010

Contraints on the Lorentz Invariance Violation Case of GRB 090510 (Abdo et al. 2009, Nature 462, 331) Some QG models predict a violation of the Lorentz invariance (Vph(E) ≠ c) HERD is more sensitive than Fermi/LAT at >20 GeV band The LAT is mostly sensitive to linear variation (n=1) and maybe quadratic variations (n=2) Sylvain Guiriec – Workshop GRBs,- Toulouse, 2010

Contraints on the EBL models HERD is more sensitive than Fermi Most models are optically thin for the 33 GeV photon from GRB 090902B. “Baseline” and “fast evolution” models rejected at 3.6 σ.

Polarization of GRBs -high polarization degree pros and cons： Rutledge & Fox (2004) – con: a) RHESSI---low signal /noise， b) CB03 overestimated the scattered photon numbers by 10 times, c) CB03 did not considered systematic errors。 Wiggle et al. (2004) : P=41+57-44%， data quality is not enough good。 GRB 021206 P =80+/-20% Addition: Willis et al. (2005) , CGRO/BATSE detections， GRB 930131: P >35%, GRB 960924: P >50%, however, systematic errors are big。 Coburn & Boggs, 2003, Nature, 423, 415

Polarization of GRBs -high polarization degree GRB 041219A: INTEGRAL detection McGlynn et al., 2007, A&A, 466, 895

Polarization of GRBs -high polarization degree GRB 100826A: GAP (70-300 keV) onboard IKAROS P=27+/-11%（2.9σ）， evidence for evolution of P. A. (Polarization Angle) Yonetoku et al., 2011, arXiv: 1111.1779

Polarization of GRBs -high polarization degree GRB 110301A: GAP (70-300 keV) onboard IKAROS P=70+/-22% (3.7σ)， no evidence for P.A. evolution Yonetoku et al., 2012, arXiv: 1208.5287

Polarization of GRBs -high polarization degree GRB 110721A: GAP (70-300 keV) onboard IKAROS P=84+16/-28% (3.3σ)， no evidence for P. A evolution Yonetoku et al., 2012, arXiv: 1208.5287

LIV birefringence constraint with GRB prompt emission polarization 3 GRBs with poln detection with GAP (Toma et al., 2013, PRL) GRB 110721A: HERD is more sensitive than Fermi/LAT at >20 GeV band!

Why HERD so important for GRB emission beyond 10 GeV? 10-4 10-5 10-6 10-7 10-8 10-9 10-10 Zhang, S. N., et al., arXiv:1407.4866 radiation mechanism for >GeV emission particle acceleration responsible for the >GeV emission GRB physics (Lorentz factor, jet composition, emission site) with >10 GeV or even >100 GeV photons Constraining EBL much severely Constraining LIV (time of flight, birefringence) at least 10-100 better than previous

Thank You