© Daisuke YONETOKU (Kanazawa Univ.) Daisuke YONETOKU (Kanazawa Univ.) Toshio MURAKAMI (Kanazawa Univ.) Shuichi GUNJI （ Yamagata Univ. ） Tatehiro MIHARA (RIKEN), Kenji TOMA (Osaka Univ.) & GAP team SN & GRB Conference Kyoto, Japan(11-15, Nov., 2013) Study of prompt emission mechanism by gamma-ray polarization with IKAROS-GAP

GRB polarimetry How to release the huge amount of energy of erg in gamma-ray band in short time duration. How to release the huge amount of energy of erg in gamma-ray band in short time duration. Direction (spatial distribution), Time Variability, Spectrum. The another information of E-M Wave “Polarization”. Inter-Stellar Medium Internal Shock (prompt) External Shock (afterglow) □ X-ray ☑ Optical ☑ Radio (Non-detect) □ X-ray ☑ Optical ☑ Radio (Non-detect) Gamma □ Gamma □ X-ray ☑ Optical Gamma □ Gamma □ X-ray ☑ Optical Γ > 100 Central engine Γ < 10

GRB Steele et al. (2009) バースト発生後 秒後から P = 10.1±1.3% Covino et al. (2004) Q (%) U (%) ■ Optical afterglows: P = 1 ～ 5% ■ Optical Flash P = 10 % ■ Optical afterglows: P = 1 ～ 5% ■ Optical Flash P = 10 % Polarization in Opt. afterglows ~ 160sec after the burst GRB990510: Covino et al. (1999) Optical Flash P=1.7 +/- 0.2%

KANATA telescope (Hiroshima Univ.) Early afterglow T = 149 ~ 706 sec  = 10.4±2.5% Early afterglow T = 149 ~ 706 sec  = 10.4±2.5% Uehara et al. (2012) GRB091208B

X-ray Polarization Vector Bragg Reflection (OSO-8 : Crab Nebula, Weisskopf et al. (1978) Rotate Bragg Crystal (~ a few keV :) detector Compton Scattering (e.g. GAP, PoGO Lite, PHENEX) (~ 100 keV) drift plane Photo-electron Amplifier Signal readout ASIC GEM Photo-Electric Absorption (e.g. GEMS, PolariS) Polarization Vector (~ 10 keV)

Coburn & Boggs (2004) GRB021206A (RHESSI)  = 80±20 % P < 100 % Rutledge & Fox (2004) P = 41 (+57, -44) % Wigger et al. (2004) Rutledge & Fox (2004) RHESSI Ｘ ■ Complex Geometry ■ No onboard Coincidence for Compton events ■ Complex Geometry ■ No onboard Coincidence for Compton events RHESSI Coburn & Boggs (2003) Rutledge & Fox (2004) Wigger et al. (2004) Difficulty

INTEGRAL-SPI McGlynn et al. (2007) Kalemci et al. (2007)  = 63±30%, 98±33% (INTEGRAL-IBIS) Gotz et al. (2009)  < 4% GRB041219A (INTEGRAL) INTEGRAL (SPI) ? The situation is complicated. We need further observations with a reliable polarimeter. The situation is complicated. We need further observations with a reliable polarimeter. SPI

Count Rate of Compton Scat. 200 Scattering Angle (deg) GA mma-ray burst P olarimeter ■ Angular distribution of Compton Scat. ■ Geometrical symmetry r 0 : classical electron radius E 0 : energy of incident photon E : energy of scattered photon GAP (sensor unit) 20 cm, 3700g Polarimetry in 70 – 300 keV ■ KEK Experiment □ Geant 4 simulation ■ KEK Experiment □ Geant 4 simulation Yonetoku et al. (2006, 2011)

IKAROS Launched May 21, 2010 I nterplanetary K ite-craft A ccelerated by R adiation O f the S un

Systematic Uncertainty Sys. Error does not depend on the incident angle. 1.7% Average Systematic uncertainty (%) Incident angle (degree) Systematic Uncertainty for Off-Axis Incident Photon 1m 57 Co (122keV) 241 Am (59.5keV) Geant4 The systematics caused by imperfect tunings of parameters in the ground and in-orbit calibrations for the off-axis radiation. Comparing the experimental and simulated modulation curves, we estimated the systematic uncertainty is ~ 1.7% of the total coincidence gamma-rays. Comparing the experimental and simulated modulation curves, we estimated the systematic uncertainty is ~ 1.7% of the total coincidence gamma-rays. Proto-FM

No.GRBFluence (erg/cm 2 ) incident angle Other Obs. No.GRBFluence (erg/cm 2 ) incident angle Other Obs A a 8.8× K,F,W,M A -K,W A19K,I,W,M A a 3.7× K,F,W B145K,F A b 4.8× K,W,I,Sw A34K,F A -K A63K,F A a 2.3× K,F,W,Sw A-K ? -? A34K,F ? -? A b 3.0× K,F,W,M K A a 2.0× K,F A c 3.1× K,W,Sw A41K,F A b 6.1× K,F,Sw A26K,F A d 2× K,F,Sw A a 1.3× K,F,I,Sw A b 2.3× K,W,Sw A62K,F B 25F,K A b 3.0× K,Sw A a 3.5× K,F,I,M A63F A a 5.4× a : keV b : keV c : keV d : keV Data Samples Konus, Fermi, Swift, WAM, Integral, Mess.

Modulation Curve Coincidence Count (counts) Scattering Angle (degree) α = β = Ep = 606 keV Interval photons (polari. data) Interval-2 2,733 photons (polari. data) Modulation Curve Coincidence Count (counts) Scattering Angle (degree) α = β = Ep = 606 keV Interval photons (polari. data) Interval photons (polari. data) Δχ 2 マップ Confidence Contour Δχ 2 Polarization Angle (degree) Polarization Degree (%) Polarization Degree (including sys. uncertainty) Polarization Angle Interval-1 P = 25±15% (95.4% C.L.)PA = 159±18 deg Interval-2 P = 31±21% (89.0% C.L.) PA = 75±20 deg Combined Fit P = 27±11% (99.4% C.L.) 3.5σ Confidence Level 3.5σ Confidence Level GRB100826A Interval-1 Interval-2 Count Rate (counts/sec) Time since GRB trigger (sec) Very bright events with F = 3.0x10 –4 erg/cm2 Yonetoku et al. (2011) T90 ~ 100 sec χ2 = 21.8 for 19 d.o.f

GRB110301A & GRB110721A We detected the polarization from two bright GRBs with high significance. The polarization angles did not change during the prompt GRBs. GRB110301A GRB110721A 1,820 photons (polari. data) 1,092 photons (polari. data) T 90 = 7 sec T 90 = 11 sec GRB110301A GRB110721A P = 84(+16, -28) % PA = 160±11 deg 3.3σ P = 70±22% PA = 73±11 deg 3.7σ Polarization Angle (degree) Polarization Degree (%) χ2 = 14.0 for 10 d.o.f χ2 = 7.3 for 10 d.o.f

Lu et al. (2012) Fermi-GBM Spectral Evolution We simulated the model functions (detector responses) with the time resolved spectral parameters, , , Ep, flux, and combined them into one model.

Slide by Felix Ryde and Magnus Axelsson GAP’s energy range

■ Significant Polarization was detected from bright 3 GRBs. ■ GRB100826A : Polarization angle changed (3.5  confidence level.) ■ GRB110721A & GRB110301A : Polarization angle was stable. ■ Significant Polarization was detected from bright 3 GRBs. ■ GRB100826A : Polarization angle changed (3.5  confidence level.) ■ GRB110721A & GRB110301A : Polarization angle was stable. We need the emission model to explain both cases of change and no-change of polarization angle. Results of Polarization Analyses Polarization Degree (%) GRB Duration T90 (sec) Incident Angle (deg) 90% upper-limit 100

Jet opening angle : θj Relativistic beaming effect : 1/ Γ Helical Magnetic Fields to drive the relativistic jet Distribution of polarization Globally Random Magnetic Fields, But Locally Coherent The change of polarization angle can be explained with patchy structures of smaller than 1/ Γ. Inner Structures may exist in the Jet. Lazzati et al. (2003–2009) Toma et al. (2009)

Narrow: θ j < 1/ G ～ 0.01 rad Broad: θ j > 1/ G ～ 0.01 rad Observing entire jet surface. Pol. angle can change. Observing entire jet surface. Pol. angle can change. Observing the part of jet surface. Pol. angle can not strongly change. Observing the part of jet surface. Pol. angle can not strongly change. Jet Opening Angle Change/No-Change of Polarization Angle Jet Opening Angle Change/No-Change of Polarization Angle 1/ G We can explain both cases of change/no-change of pol. angle with the relation between θ j and 1/ G, and also the patches.

Photon Field highly polarized Electron Rest Frame Observer Frame Non-polarized highly polarized Compton Drag Lazzati et al. (2005) GRB Jet jj 1/  Dimmer, Highly Polarized Dimmer, Highly Polarized Brighter, Non-Polarized Brighter, Non-Polarized & Photospheric Emission The highly polarized gamma-rays: only when the observer is slightly outside of the jet. The prompt emission is dim. It is difficult to explain the existence of bright & highly polarized GRBs in the frame work of Compton drag and Photospheric emission mode.

GAP 70 – 300 keV GAP 70 – 300 keV Epeak – Polarization Degree Both low- and high-energy part of spectra are polarized. Polarization Degree (%) Epeak (keV) A Rapid Change of Polarization Angle A A A A A Average

GRB INTEGRAL (IBIS) Gotz et al. (2013) 250 – 800 keV 250 – 350 keV 350 – 800 keV 250 – 800 keV 250 – 350 keV 350 – 800 keV T 90 = 12 sec (almost single peak) Fluence = 2x10 -5 erg/cm 2 68% 90% 95% 99% Non-polarization was rejected with 4σ level

Gamma-Ray (Prompt – Forward shock) 27±11 % (GRB100826A) 70±22 % (GRB110301A) 84(+16, -28)% (GRB110721A) Gamma-Ray (Prompt – Forward shock) 27±11 % (GRB100826A) 70±22 % (GRB110301A) 84(+16, -28)% (GRB110721A) Optical Flash (Prompt – Reverse shock) 10.1±1.3 (GRB090102) Steele et al Optical Flash (Prompt – Reverse shock) 10.1±1.3 (GRB090102) Steele et al Early Optical Afterglow (Forward shock) < 8 % (GRB060418) Mundel et al. 10.4±2.5% (GRB091208B) Uehara et al Early Optical Afterglow (Forward shock) < 8 % (GRB060418) Mundel et al. 10.4±2.5% (GRB091208B) Uehara et al After Jet Break ?? % After Jet Break ?? % Time Brightness Evolution of polarization Late Optical Afterglow (Forward shock) 1 – 3 % (summarized in Covino et al. 2004) Late Optical Afterglow (Forward shock) 1 – 3 % (summarized in Covino et al. 2004) Prompt Afterglow

Summary ■ We detected the  -ray polarization from 3 bright GRBs, and set U.L. for 4 GRBs with GRB polarimeter “GAP”. ■ The emission mechanism of prompt GRBs are probably the synchrotron radiation in the coherent magnetic fields. (Our results favor theICMART model (Zhang’s group). We cannot exclude the photospheric and comptonized emission model) ■ Since the polarization angle rapidly changed, the multiple emission regions and/or the patchy structures with the scale of < 1/ Γ may exist in the relativistic jet. Next gamma-ray polarimeter aboard “TSUBAME” small satellite by the Tokyo-Tech team will be launched next year.

© Daisuke YONETOKU (Kanazawa Univ.) Inner structures & Magnetic Fields Image of Gamma-ray bursts