1. Slow heating, fast cooling in gamma-ray bursts Juri Poutanen University of Oulu, Finland +Boris Stern + Indrek Vurm

2. 1. Gamma-ray burst spectral properties 2. Synchrotron or Compton, cooling or not? 3. Slow heating/reacceleration model - Simple physics - Numerical codes - Synchrotron - Synchrotron self-Compton - Quasi-thermal Comptonization 4. Summary Plan

3. 1. Gamma-ray burst spectral properties 2. Synchrotron or Compton, cooling or not? 3. Slow heating/reacceleration model - Simple physics - Numerical codes - Synchrotron - Synchrotron self-Compton - Quasi-thermal Comptonization 4. Summary Plan

4. `Band´ (or `GRB´) function: dN/dE ~ E   exp( – E/E 0 ) E < (  –  ) E 0 dN/dE ~ E  E > (  –  ) E 0 EF E E Energy spectra

5. Preece et al. (2000), 156 GRBs, ~5000 spectra (different time bins)  E0E0  GRB spectral properties

6. Imamura & Epstein (1987): The steeply rising X-spectra can be regarded as one of the securely determined aspects of gamma-ray bursts, since it has been independently measured by several research groups, and there are no reported counterexamples. GRB hardness

7. GRB spectral evolution Ford et al. (1995)

8. Thermal components Ryde & Pe’er (2008)

9. Thermal components Ryde et al. (2006) Power-law indices after and before subtracting a blackbody

10. Gonzales et al. (2003)   = - 0.84  = - 1.1 113 to 211 s  = -0.96 80 to 113 47 to 80 s 14 to 47 s -18 to 14 s  = - 1.06  = - 0.79  = - 1.08 High-energy emission

11. Briggs et al. 1999; Galama et al. 1999 GRB 990123: prompt optical emission

12. GRB 080319B “the naked eye burst” Simultaneous optical and gamma-ray emission. Optical emission is 4 orders of magnitude brighter than continuation of the gamma-ray spectrum Racusin et al. 2008 t=3 sec 17 sec 32 sec

13. Spectral properties  Broad distribution of low energy spectral indices with peak at photon index a ≈ -1  and the hardest spectra are with a ≈   Hard-to-soft spectral evolution during pulses  Sometimes (always?) strong `thermal’ component.  High-energy emission at 100 MeV with a ≈ -1 in GRB 941017, softer spectra in Fermi GRBs. Delayed relative to MeV. Second component?  Prompt optical emission (GRB 990123, GRB 080319B) orders of magnitude larger than the continuation of gamma-ray emission. (Delayed relative to MeV emission, maybe produced later.)

14. 1. Gamma-ray burst spectral properties 2. Synchrotron or Compton, cooling or not? 3. Slow heating/reacceleration model - Simple physics - Numerical codes - Synchrotron - Synchrotron self-Compton - Quasi-thermal Comptonization 4. Summary Plan

15. Synchrotron shock model Electrons are assumed to obtain a fraction e e <1 of available energy instantaneously (acceleration time << cooling time), BUT Electron cooling time << light-crossing time

16. Spectra from cooling electrons n, frequency n F n F n ~ n 1/3 Instantaneous synchrotron spectrum of mono-energetic electrons a = –2/3 ”cooling” spectrum N n ~F n / n ~ n a a = –3/2 time (e.g. Ghisellini & Celotti 1999; Ghisellini, Celotti, Lazzati 2000)

17. a=–2/3 a =–3/2 cooling deathline synchrotron deathline Cooling vs. Synchrotron “deathline”

18. a=–2/3 a =–3/2 cooling deathline synchrotron deathline Cooling vs. Synchrotron “deathline” FORBIDDEN >90 % of GRB spectra are inconsistent with instantaneous particle injection, following by cooling via synchrotron or Compton scattering. Should we stop writing and accepting papers ignoring those simple facts?

19. How to get a GRB spectrum? Fast cooling models with “instantaneous” injection can be ruled out. How then can we get a prompt GRB? 1.Photospheric emission can give hard spectra. Maybe then the underlying continuum is softer? Consistent with cooling? 2.Electrons could be heated at a time-scale >> cooling time-scale (stochastic `slow ’ heating: 2nd order Fermi, magnetic dissipation, etc.). Emission : synchrotron self-Compton. 1+2. Photospheric emission + slow heating + SSC

20. 1. Gamma-ray burst spectral properties 2. Synchrotron or Compton, cooling or not? 3. Slow heating/reacceleration model - Numerical codes - Synchrotron - Synchrotron self-Compton - Quasi-thermal Comptonization 4. Summary Plan

21. Heating / acceleration Electrons / Pairs Radiation Cooling by synchrotron and inverse Compton Energy balance L / t T = R 2 c(U B + U rad ) g 2 Heating = cooling y =t T g 2 ~a few Slow heating / reacceleration of particles

22. (simple) energy balance for electrons

23. If Thomson optical depth t T ~10 -8 then electron Lorentz factor g ~10 4 and synchrotron is the main cooling mechanism If t T ~10 –2 -10 –4 then g ~ 10-100 and synchrotron self-Compton dominates (Stern & Poutanen 2004, Vurm & Poutanen 2009) If t T ~1 then g~ 1-2 and quasi-thermal Comptonization dominates (Ghisellini & Celotti 1999; Stern 1999) Dominant radiative processes

24. 1. Large Particle Monte- Carlo code (LPMC) Stern et al. (1995)  = 0.1 R Slab geometry PROCESSES INCLUDED: Cyclo-synchrotron emission Compton scattering Pair production, pair annihilation Particle thermalization by Coulomb scattering and synchrotron self-absorption HEATING: LPMC: Electrons/pairs obtain equal amount of energy per unit time. Kinetic: stochastic acceleration Time-dependent simulations Sphere R 2. Kinetic equations (for electrons, positrons, photons) Vurm & Poutanen (2009)

25. Kinetic code (Vurm & Poutanen) One-zone, isotropic particle distributions, tangled B- field Processes: oCompton scattering:  exact Klein-Nishina scattering cross-sections for all particles  diffusion limit at low energies oCyclo-synchrotron radiation: exact emissivity/absorption for photons and heating/cooling (thermalization) for pairs. opair-production, exact rates oCoulomb scattering Time-dependent, coupled kinetic equations for electrons, positrons and photons. Contain both integral and (up to 2nd order )differential terms Discretized on energy and time grids and solved iteratively as a set of coupled systems of linear algebraic equations Accurate energy conservation.

26. Kinetic code (Vurm & Poutanen) Main equations 1. Processes: 2. Form:

27. Synchrotron model t T = 2 x 10 -8 l = 0.3, B = 10 G Produces too soft gamma-ray spectra Optical depth is very small 0 < t < 0.35 0.35 < t < 0.6 0.6 < t < 1 EF E Low-energies High-energies Optical depth BATSE g ~10 4

28. SSC (Stern & Poutanen 2004) synchrotron a =0 pairs 1st Compton Comoving photon spectrum 2nd Compton t T = 6 x 10 -4 l = 3  B =0.3 (B = 10 3 G) R´= 10 13 cm  =130

29. SSC (Vurm & Poutanen 2009) optical / UV MeV GeV electrons synchrotron 1st Compton Comoving photon spectrum 2nd Compton positrons pairs

30. l= 300 50 keV (comoving) 5 MeV observed ! t T ~1 Quasi-thermal Comptonization (Stern 2003; Stern 1999; Ghisellini & Celotti 1999) optical might be too weak? Soft gamma-ray spectra, peaking at high energies Photospheric emission will change the spectrum g~ 1-2 t T ~1

31. Summary: Spectra and models  Synchrotron or Compton from cooling electrons produce too soft X- ray spectra. (How long are we going to ignore this fact?)  Cooling can be avoided if electrons are heated over the dynamical time. If it is so, synchrotron still produces too soft spectra, it also requires very small optical depths. SSC by nearly monoenergetic electrons (e.g. Maxwellian, see Stern & Poutanen 2004, Vurm & Poutanen 2009) can produce hard enough spectra.  Photospheric emission (Pe’er et al. 2006) might solve the problem of hard spectra. What is then the radiation mechanism producing underlying powerlaw: synchrotron or Compton? Electrons: cooling distribution or Maxwellian? It seems still that spectra are harder than the cooling ones.  Origin of prompt optical emission: synchrotron from larger radius, reverse shock?  Origin on GeV emission: IC of MeV photons? Just a tail of MeV photons? Prompt or already afterglow?

32. Questions  Energy dissipation. Electron heating / acceleration mechanism. 1.Internal shocks? 2.Magnetic reconnection? 3.Neutron decay?  Radiative processes 1.Synchrotron 2.Compton (tau>1) 3.SSC (tau<1) 4.Photosphere + SSC  How many spectral components? optical, MeV, GeV