High-energy emission from Gamma-Ray Bursts Frédéric Daigne Institut d Astrophysique de Paris, Université Pierre et Marie Curie HEPRO III High Energy Phenomena in Relativistic Outflows Barcelona, June 27 July 1, 2011
Gamma-Ray Bursts Duration : ms 1000 s (2 groups) Highly variable lightcurve Non-thermal spectrum (peak ~ kev MeV) Distance : z max,obs = 8.2! E γ,iso ~ 10 51 10 54 erg! Afterglow : minutes weeks Flux detection : X, optical, radio Fast decay : F ν t -α ν -β
The physics of GRBs Log( R ) [meters] Relativistic ejection Acceleration : Γ > 100 Photosphere Internal dissipation (shocks, reconnection?) Prompt γ-rays Reverse shock Contact discontinuity Lateral expansion Non-relativistic regime External shock Afterglow
(1) Detection of GRBs at high energy (GeV) Fermi-LAT : ~ 10 GRBs / year (to compare to GBM : ~ 250 GRBs / year) 4 brightest bursts : GRB z E γ,iso Group Refs 080916C 4.35 8.8 10 54 erg long Abdo et al. (2009a) 090510 0.9 1.1 10 53 erg short? Ackermann et al. (2010) 090902B 1.8 3.6 10 54 erg long Abdo et al. (2009b) 090926A 2.1 2.2 10 54 erg Long Ackermann et al. (2011) Low detection rate by the LAT : - no bright component in the GeV range - need for a cutoff at ~ 100 MeV? (see e.g. Le & Dermer 2009 ; Granot et al. 2010 ; Guetta et al. 2011 ; Beniamini et al. 2011)
(1) Detection of GRBs at high energy (GeV) An example : GRB 080916C GRB 080916C (Abdo et al. 2009) GBM : kev-mev LAT >100 MeV >1 GeV
Constraints on the Lorentz factor Compactness problem : short variability timescale + huge luminosities For a static source: γ-rays should not be able to escape due to photon photon annihilation: γγ e + e - (above m e c 2 = 511 kev for head-on collisions) c t var > R Alternative : the emitting source moves at a relativistic speed Size of the emitting region is larger lower photon densities Photons paths are almost parallel photon interaction less efficient (Rees 1966) GRBs : pre-fermi estimates (MeV observations) Γ min ~ 100-300 (see e.g. Baring & Harding 1997; Lithwick & Sari 2001) 2Γ 2 c t var > R
Constraints on the Lorentz factor Fermi-LAT detections in the GeV range : Stricter Lorentz factor constraints GRB 080916C : Γ min 887 (Abdo et al. 2009) GRB 090510 : Γ min 1200 (Ackerman et al. 2010) Such values of the Lorentz factor : - are challenging for most models of the central engine ; - have strong consequences on the GRB scenario (photospheric radius, deceleration radius, ). However, these estimates are based on simplified single zone models. GRB 080916C (Abdo et al. 2009)
Constraints on the Lorentz factor Detailed calculation : space/time/direction-dependent radiation field the estimate of Γ min is reduced by a factor ~ 2-3 (see Granot et al. 2008; Hascoët, Daigne, Mochkovitch & Vennin to be submitted) 1 MeV Bin b Model of bins a+b in GRB 080916C : Γ min ~ 360 (Hascoët et al. to be submitted) instead of ~900 (Abdo et al. 2009).
Constraints on the Lorentz factor Detailed calculation : space/time/direction-dependant radiation field the estimate of Γ min is reduced by a factor ~ 2-3 (see Granot et al. 2008; Hascoët, Daigne, Mochkovitch & Vennin to be submitted) If the GeV and the MeV emission are not produced at the same place : the constraint is even further reduced. (see Zhao et al. 2011; Zou et al. 2011)
Constraints on the Lorentz factor Detailed calculation : space/time/direction-dependant radiation field the estimate of Γ min is reduced by a factor ~ 2-3 (see Granot et al. 2008; Hascoët, Daigne, Mochkovitch & Vennin to be submitted) If the GeV and the MeV emission are not produced at the same place : the constraint is even further reduced. (see Zhao et al. 2011; Zou et al. 2011) A new approximate formula, more general, more accurate : (Hascoët, Daigne, Mochkovitch & Vennin to be submitted) correction factor (detailed modeling) single zone formula (Abdo et al. 2009) additional correction factor, if different MeV/GeV emitting regions = 1, if R GeV = R MeV < 1, if R GeV > R MeV
(2) Dominant spectral component The main component already known in the kev-mev range is dominant : GRB 080916C (Abdo et al. 2009)
(2) Dominant spectral component The main component already known in the kev-mev range is dominant : GBM LAT GRB 080916C (Abdo et al. 2009)
(3) A weak and soft thermal component? In at least one case, there is possibly the detection of a weak thermal component : GRB 100724B (Guiriec et al. 2011)
The physical origin of the prompt emission Fast variability : the prompt emission has an internal origin. Three possible reservoirs for internal dissipation : Thermal energy : radiated at the photosphere Flux (X- rays)? Pros : -no large theoretical uncertainty -high efficiency t Cons : -prompt spectrum is non-thermal : additional mechanisms are needed -origin of the steep decay in the X-ray afterglow? -there is a hint for a weak soft component in GRBs which is indeed thermal (Guiriec et al. 2011) (Paczynski 86; Goodman 86; Shemi & Piran 90; Meszaros & Rees 00; Meszaros et al. 02; Daigne & Mochkovitch 02; Zhang & Meszaros 02; Rees & Meszaros 05; Pe er et al. 06, 07, 08, 10; Ioka et al. 07; Beloborodov 10; Toma et al. 10; Vurm et al. 2011; ) TH Standard fireball : hot and bright photosphere NT Cold photosphere : magnetized outflow? (see GRB 100724B) NT TH
The physical origin of the prompt emission Fast variability : the prompt emission has an internal origin. Three possible reservoirs for internal dissipation : Thermal energy : radiated at the photosphere Flux (X- rays) Kinetic energy : -dissipation in shocks - radiation from shock accelerated electrons ~R/2Γ 2 c t Pros : -can reproduce well the temporal and spectral properties -origin of the early steep decay (X-ray afterglow) : high-latitude emission (Kumar & Panaitescu 2000) -no large theoretical uncertainties on the dynamics -the spectrum may have several components Cons : -low efficiency (Daigne & Mochkovitch 98 ; see however Beloborodov 00; Kobayashi & Sari 01) -large theoretical uncertainties for shock acceleration (Rees & Meszaros 94 ; Paczynski & Xu 94; Kobayashi et al. 97 ; Daigne & Mochkovitch 98, 00, 03 ; Meszaros & Rees 00; Pe er et al. 06; Bosnjak, Daigne & Dubus 09 ; )
The physical origin of the prompt emission Fast variability : the prompt emission has an internal origin. Three possible reservoirs for internal dissipation : Thermal energy : radiated at the photosphere Flux (X- rays) Kinetic energy : -dissipation in shocks - radiation from shock accelerated electrons ~R/2Γ 2 c t Magnetic energy : -dissipation by magnetic reconnection -particle acceleration radiation Only toy models are available : -efficiency may be high (Thomson 94 ; Spruit et al. 01 ; Drenkhahn & Daigne 02 ; Giannios 06 ; Giannios & Spruit 07 ; Giannios 08 ; ) -spectrum may have several spectral components (Giannios 2008) -lightcurves may show two typical timescales (Zhang & Yan 2011) -lightcurves may show too symmetric pulses (Lazar et al. 2009) More realistic and physically motivated simulations are needed
The physical origin of the prompt emission Fast variability : the prompt emission has an internal origin. Three possible reservoirs for internal dissipation : Thermal energy : radiated at the photosphere Flux (X- rays) Kinetic energy : -dissipation in shocks - radiation from shock accelerated electrons ~R/2Γ 2 c t Magnetic energy : -dissipation by magnetic reconnection -particle acceleration radiation Combinations are possible : -photospheric emission + internal shocks -photospheric emission + magnetic dissipation -magnetic dissipation + internal shocks : unlikely (shocks cannot propagate if the outflow is highly magnetized)
The physical origin of the prompt emission Dominant radiative process : synchrotron vs SSC? (non photospheric models) SSC :? GBM LAT? IC2 - Where is the strong IC2 component? or the strong syn component? Syn IC1 IC sca'. (Thomson) - Energy crisis Synchrotron : GBM LAT Fermi-LAT detection rate and observations clearly favor the synchrotron process. Syn IC sca'. (Klein- Nishina) IC (see e.g. Bošnjak, Daigne & Dubus 09; Piran, Sari & Zou 09) Synchrotron + IC scatterings in KN regime : low-energy slope α ~ -3/2-1 (Derishev et al. 2001; Bosnjak et al. 2009 ; Nakar et al. 2009 ; Daigne et al. 2011)
(4) Delayed onset of the GeV component GRB 080916C (Abdo et al. 2009) In several cases, there is a delayed onset of the GeV component :
(4) Delayed onset of the GeV component GRB 080916C (Abdo et al. 2009) In several cases, there is a delayed onset of the GeV component :
(5) Additional components at high energy In some cases, an additional component is needed GRB 000902B (Abdo et al. 2009)
(5) Additional components at high energy In some cases, an additional component is needed GRB 000902B (Abdo et al. 2009)
GeV delayed onset & additional components Intrinsic spectral evolution? (e.g. emergence of an IC component) (Bosnjak, Daigne & Dubus 2009)
GeV delayed onset & additional components Intrinsic spectral evolution? (e.g. emergence of an IC component) Similar conclusions are obtained by Asano & Meszaros (Asano s talk @ Fermi Symposium) They show that the late synchroton emission may explain the X-ray excess.
Delayed onset of the GeV emission Intrinsic spectral evolution? (e.g. emergence of an IC component) External inverse Compton (Toma et al. 2009; Toma et al. 2011) -seed photons = jet cocoon, photosphere, -relativistic electrons = internal shocks (Asano & Meszaros 2011 : Asano s talk @ Fermi Symposium) This model shows naturally several components in the spectrum.
Delayed onset of the GeV emission Intrinsic spectral evolution? (e.g. emergence of an IC component) External inverse Compton Emergence of a hadronic component? -delay = proton acceleration -GeV emission = hadronic cascade (Bötcher & Dermer 1998 ; Gupta & Zhang 2007 ; Asano & Inoue 2007 ; Asano, Inoue & Meszaros 2009 ; etc ) Efficiency is low and these models require a huge energy injected in protons.
Delayed onset of the GeV emission Intrinsic spectral evolution? (e.g. emergence of an IC component) External inverse Compton Emergence of a hadronic component? Delay : opacity effect? (Hascoët, Daigne, Mochkovitch & Vennin to be submitted) 1 MeV GBM t var 0.5 s t var 0.5 s t delay 5 s 3 GeV LAT t delay 5 s GRB080916C : model GRB080916C : Fermi obs. (Abdo et al. 2009)
(6) Cutoff at high energy Except in one case, there is no clear signature for a cutoff at high energy GRB 090926A (Ackermann et al. 2011) In GRB 090926A, the spectral shape of the cutoff cannot be well characterized: - γγ opacity (shape depends on the time interval)? - intrinsic curvature of the HE component?
(7) Long lasting emission in the LAT A long lasting emission is detected in many cases in the LAT after the end of the GBM emission : high-energy afterglow? GRB 080916C (Abdo et al. 2009)
Long lasting emission High-energy afterglow? Problem : no simultaneous observations of the early X-ray afterglow by Swift Where is the X-ray early steep decay? (in the high latitude emission scenario : prompt/afterglow transition) The long lasting emission associated with standard early afterglow observation may help in distinguishing between afterglow models)
Long lasting emission High-energy afterglow? An intringuing possibility : may the whole GeV emission be due to the external shock? (Kumar & Barniol Duran 2009, 2010 ; Gao et al. 2009 ; Corsi, Guetta & Piro 2010 ; de Pasquale 2010; Ghisellini et al. 2010 ; Ghirlanda et al. 2010 ; ) Possible issues : -afterglow must be in radiative regime pair enrichment? (see e.g. Beloborodov 2002) -exceeds maximum energy of synchrotron? (Piran & Nakar 2010) -needs Γ > 1000 (independent from γγ constraint) -needs very low density and magnetization in the external medium (Kumar & Barniol Duran 2009) Ghisellini et al. 2010 A possible test? Variability in GeV lightcurve
Summary Fermi observations : (1) low detection rate by the LAT (4 bright bursts in the GeV range) (2) MeV non-thermal component is dominant (3) weak & soft thermal component? (4) GeV delayed onset (5) additional HE component is some cases (6) HE cutoff in at least one case (7) long lasting emission in the LAT Models for the prompt emission : - photosphere + internal shocks or - (photosph.?) + magnetic dissipation? - dominant photosphere : spectral shape? Early X-ray afterglow? - dominant internal shocks : efficiency? Shock acceleration? synchrotron + IC in Klein Nishina regime is favored - dominant magnetic dissipation : more modelling is needed Models for the high-energy emission : - hadronic models have a low efficiency - leptonic models seem to be able to reproduce most observed features by a combination of intrinsic spectral evolution + opacity effects - an intringuing possibility : external origin of the whole GeV emission?