GRB emission models and multiwavelength properties Gabriele Ghisellini INAF-Osservatorio Astronomico di Brera - Italy with the help of: Z. Bosniak, D. Burlon, A. Celotti, C. Firmani, G. Ghirlanda, D. Lazzati, M. Nardini, L. Nava, F. Tavecchio
The standard model
The model: Internal/External Shocks Rees-Meszaros-Piran Shell still opaque Relativ. e- + B: synchrotron?? Relativ. e- + B: synchrotron
Why internal shocks? Counts/s Spikes have same duration Time [seconds] A process that repeats itself
Epeak Energy [MeV] Fishman & Meegan 1995 E F(E) Spectra Spectra
Kaneko+ 2006 [kev]
Nava PhD thesis 2009 Kaneko+ 2006 [kev]
Prompt radiation: Synchrotron?
Synchrotron -ray emission? The shell itself carries B-field B can also be produced and amplified by the internal shock The shock can accelerate e- to relativistic energies Synchrotron seems a good choice h syn ~ a few hundreds kev seems reasonable
Synchrotron -ray emission? Radiative cooling is (and must be) very rapid tcool ~ 10-7 3e ( /100) 2 MeV sec Extremely short - No way to make it longer tcool << tdynamical ~ 10-2 sec It must be short: if not, how can the flux vary?
Energy spectrum of a cooling electron Fast cooling + synchro: E( ) -1/2 N( ) -3/2 Photon index
Kaneko+ 2006
Line of death for non cooling e- Line of death for cooling e- Kaneko+ 2006 Nava PhD thesis 2009
Can it be rescued by: Reacceleration? No, in IS e- are accelerated only once. More generally, only few selected e- (and always the same) must be (re)accelerated Adiabatic losses? No, too small regions would be involved, large e- densities, too much IC Fast decaying B-field? No, synchro not efficient and too much IC Self absorption? No, lots of e- needed, too much IC Self Compton? No, tcool too small even in this case Small pitch angles? No, Very very small to avoid cooling, becomes inefficient External shocks to avoid cooling? No, same reason
Seeking alternatives Bulk Compton (Lazzati+ 2000; GG+ 2000) Quasi-thermal Comptonization (GG & Celotti 1999) Black-body from deep impacts (Thompson+2007; GG+ 2007; Lazzati+ 2009) Reconnection and continuous heating (Giannios 2008)
Seeds Wolf Rayet (about to explode) 100 Bulk Compton
Problems Time to refill the funnel of seeds Needs a lot of seeds, not clear if the funnel is sufficient Does not work for short (they do not have a SN)
Quasi thermal Componization Heating for r/c, not instantaneous acceleration sub-relat. T Cooling=Heating Synchro is self absorbed and produces seeds for quasi saturated Comptonization y needs to be >10
Quasi-saturated spectrum: 0+Wien
Problems Epeak ~ kt too high Needs time ( must be large) Wien peak not observed Seeds should be distributed in the center, e- more externally, to explain >-1
Deep impacts Thompson, Meszaros & Rees 2007 Lazzati, Morsony Begelman 2008 GG+ 2007 At R ~ Rstar the fireball dissipates part of its energy BB
Dissipation Wolf Rayet (about to explode) BB Transparency 1 j (i.e very small)
Problems fine tuned (and small) BB?? in Thompson+ 2008
There can be a Black Body but BATSE Time integrated spectrum BB Time resolved spectra Cu t of f pl aw l er w po 30 kev Ghirlanda+ 2007
There can be a Black Body but BATSE Time integrated spectrum BB Time resolved spectra pl aw l er w po Cu t of f The same occurs for ALL GRBs detected by BATSE and with WFC 30 kev Ghirlanda+ 2007
M B G I M R FE Ghirlanda+ 2009 Cutoff PL 30 kev
M B G I M R FE Ghirlanda+ 2009 BB+PL 30 kev
M B G I M R FE Ghirlanda+ 2009 Cutoff PL
BB+PL e h T e m a s l l a r o f s r u c oc e m i t s e c i l s Ghirlanda+ 2009 M B G I M R FE
Reconnection and continuous heating (Giannios 2008) 1 MeV Very promising, especially for magnetized fireballs, but.
Reconnecting regions should behave randomly
Reconnecting regions should behave randomly Instead there are trends
Ghirlanda+ 2009 Rate Epeak [kev] 1/2 L k Epeak = FERMI-GBM Luminosity [erg/s]
How to explain it? It is NOT DUE to selection effects!!!!!! It indicates something fundamental and very robust that we do not understand yet. Geometrical? Difficult Sequence of? Difficult Radiation process? Likely, but which one?
First conclusion We do not know yet what is the emission mechanism of the prompt
T A -L i rm e F GRB 090510 Short Very hard z=0.903 Detected by the LAT up to 31 GeV!! Well defined timing Delay: ~GeV arrive after ~MeV (fraction of seconds) Quantum Gravity? Violation of Lorentz invariance?
precursor 8-260 kev LAT all >100 MeV 0.6s 0.5s 31 GeV Time since trigger (precursor) >1 GeV Abdo et al 2009 0.26-5 MeV
precursor 8-260 kev 0.26-5 MeV LAT all Due to Lorentz invariance violation? >100 MeV 0.6s 0.5s 31 GeV Time since trigger (precursor) >1 GeV Abdo et al 2009 Delay between GBM and LAT
0.1 GeV Time resolved 2 30 GeV Different component 3 0.5-1s 4 1 Energy [kev] 3 Abdo et al 2009 F( ) [erg/cm2/s] Average
0.1 GeV Time resolved 2 30 GeV Different component 3 0.5-1s 3 If LAT and GBM radiation are cospatial: 4 >1000 to avoid photon-photon absorption 1 If >1000: deceleration of the fireball occurs early early afterglow! If >1000: large electron energies synchrotron afterglow! Energy [kev] Abdo et al 2009 F( ) [erg/cm2/s] Average
T A -L i rm e F t2 Signature Ghirlanda+ 2009 t-1.5
>1 GeV T-T* [s] Ghirlanda+ 2009 0.1-1 GeV
T-T* [s] Ghirlanda+ 2009
Strong limit to quantum gravity MQG > 4.7 MPlanck T-T* [s] Ghirlanda+ 2009 ~MeV and ~GeV emission are NOT cospatial. But the ~GeV emission is No measurable delay in arrival time of high energy photons: tdelay<0.2 s
Conclusions Paradigm : internal+external shocks, synchrotron for both: it does not work Problems: efficiency, spectrum, trends Fermi/LAT detection large Early high energy afterglow Violation of the Lorentz invariance? No (not yet)