Theory of prompt and afterglow emission Robert Mochkovitch (IAP) Gamma-Ray Bursts in the Multi-Messenger Era (Paris, June 2014)

central engine relativistic jet photosphere internal dissipation: prompt emission RS FS afterglow emission jet break What we know for (almost) sure….

central engine relativistic jet photosphere internal dissipation: prompt emission RS FS afterglow emission jet break Unsettled issues: acceleration/energy content of the jet: thermal/magnetic? dissipation mechanism at work? respective contributions of the forward and reverse shocks to the afterglow surprises in the early afterglow What we know for (almost) sure….

The prompt emission Brief observational summary Temporal properties: hard X-rays diversity of light curves bimodal duration distribution variability down to the ms time scale long bursts: collapsars short bursts: merging of NS

Temporal properties: optical → diversity of behaviors GRB GRB GRB B GRB : optical not correlated to hard X-rays GRB : optical and hard X-rays correlated optical flux consistent with extrapolation of the hard X-ray spectrum at low energy GRB b : correlated (?) optical flux 100 times brighter than extrapolation of the hard X-ray spectrum at low energy (the naked eye burst) RAPTOR

Spectral properties GRB spectra are (too) simple : broken power-laws    +2  +2 EpEp grey area: bright BATSE bursts solid line: Fermi data (Lu et al, 2012) Phenomenological Band function:

Going beyond the Band function: → indications of the presence of an underlying thermal (photospheric) component Spectral properties Polarization positive detection in a few events: - GRB A:  ~ 4 – 40% (IBIS; Götz et al, 2009) - GRB A, A, A:  ~ 20 – 80% (GAP; Yonetoku et al, 2011,2012) Additional power-law in some cases (excess at low (keV) and high (>10 MeV) energy) (Guiriec et al, 2010) GRB B GRB A

Models Basic requirements z → D L + observed flux → E ,iso = – erg short time scale variability → compact source Relativistic outflow to avoid opacity problem:  → e + e -  min  ~ 100 – 1000 Acceleration of the flow Thermal :  = r acc ~  r 0 m: entrained mass (fireball model) (Hascoët et al, 2012) T 0 ~ a few MeV

Acceleration of the flow Magnetic : slower acceleration: initially → Acceleration may not be completed at the photosphere → Thermal emission can be much reduced if → Remaining magnetization at infinity ?.. (Tchekhovskoy et al, 2010) r 1/3

Dissipation processes below the photosphere: “photospheric models” np collisions shocks → energetic electrons (and positrons) reconnection IC on thermal photons at a few optical depths below the photosphere + synchrotron contribution/geometrical effect at low energy IC syn E E2N(E)E2N(E) (Vurm et al, 2011) Planck → Band

Dissipation processes above the photosphere: internal shocks variable Lorentz factor in the outflow  r  r 1>1>  rr dissipate 10 – 20 % of the flow KE Redistribution of the dissipated energy :  e x E diss : into a non thermal (power law) distribution of electrons  B x E diss : in magnetic energy → synchrotron emission (Daigne & Mochkovitch, 1998)

Above the photosphere: reconnection  ∞ > 1 difficult and uncertain physics → few predictions except ICMART (Internal-Collision-induced MAgnetic Reconnection and Turbulence) Potentially large efficiency : 30 – 50 % ? (Zhang & Zhang, 2014)

Evaluating the models Internal shocks : many predictions in good agreement with observations : hardness – duration, HIC, HFC, W(E) Potential problems efficiency requires cooling electrons in “fast cooling regime” → low energy slope  = -1.5 while obs ~ -1 (see however Derishev, 2007; Daigne et al, 2011; Uhm & Zhang, 2013) acceleration of electrons: much energy into a small (1%) fraction of the electrons magnetic acceleration required to avoid bright photospheric emission but then what about  ∞ and the existence/efficiency of shocks ? Pulse width Time lags Energy [keV] keV 260 keV-5 MeV >100 MeV >1 GeV GBM LAT Time [s] Photon flux [ph/cm 2 /s] (Bošnjak & Daigne 2013)

Reconnection: natural model if magnetic acceleration with large  ∞ ? uncertainties with the spectrum: general shape, low energy spectral slope  Photospheric models: less uncertain input physics requires an “adaptable dissipative process” should work for a full range of L, E p (X-ray flashes, X ray emission during quiescence in gamma) → Looking for tests of the various models…

Temporal tests steep decay at the end of the prompt phase high latitude emission tbtb tbtb  -3 IS, ICMART :  t ~ t b →  ~ -3 Photospheric models :  t << t b In photopheric models the initial decay must correspond to an effective behavior of the central engine RR (Hascoët et al, 2012)

Spectral tests additional thermal (photospheric) component in the spectra ? expected in internal shock, reconnection models… …but a priori not in photospheric models where the spectrum is the (modified) photospheric emission (Guiriec et al, 2013)

Optical emission Bursts where  /opt are correlated suggest similar emission radii : R em,  ~ R em,opt → risk of self-absorption in photospheric models High energy emission  → e + e - :  min depends on R GeV R GeV  R MeV possible with IS, ICMART but not in photospheric models GRB C (Abdo et al. 2009) GBM : keV-MeV LAT >100 MeV >1 GeV

Polarization Models with synchrotron emission (internal shocks and reconnection)  Large  possible if:  - ordered in emission region i.e.  B > 1/  ( lines anchored at the source)   - Jet viewed on the edge :  v ~  j (within 1/  ) (random ) Photospheric models Polarization averages to ~ 0 except if the jet is viewed on the edge (but synchrotron contribution can be present →  ≠ 0 )  (Toma, 2014)

Conclusions (prompt emission) Best and worst for each model: Photospheric emission B: reliable physics; good spectra W: early steep decay ≠ high latitude emission ; optical prompt self-absorbed Internal shocks B: large set of predictions agrees with observations W: acceleration of electrons ;  ∞ ∞  ; low energy spectral index Reconnection B: natural if  ∞ is large ; possibly large efficiency W: few predictions ; spectra ? territories: friendly hostile Terra incognita internal shockslargequite largesmall photosphericmedium reconnectionsmall large Model geography

The afterglow … results from the deceleration of the flow by the external medium (uniform or stellar wind) The pre-Swift era: afterglows looked pretty simple ! Forward shock dynamics described by the Blandford-McKee solution (at deceleration radius: swept up mass ) uniform medium stellar wind R -3/2-1/2 t -3/8-1/4 Shock dissipated energy injected into a non-thermal distribution of electrons:  e, p,   and amplifies the magnetic field:  B  Injection Lorentz factor of the electrons: m  B  m 2   Post-shock magnetic field: → c  B  c 2 

Afterglow spectra and light curves are made of consecutive power-law segments multi-wavelength fit of the afterglow → E K,  e,  B, p, n/A * Hz  c < m m < c day E K = erg ; n = 0.14 cm -3  e = ;  B = ; p = 2 (Sari, Piran, Narayan, 1998) (Panaitescu & Kumar, 2000) (Panaitescu & Kumar, 2002) spectra light curves

Concerns: robustness of the results → constant microphysics parameters ? → uniform external medium often found; at odds with expectation for a WR progenitor The Swift era: surprises in the early afterglow plateau phase, flares, steep slopes, optical/X-rays: (a)chromatic behaviors… t -3.2 New ingredients/paradigm needed !

New ingredients Plateaus and flares: a late activity of the central engine ? Extended plateaus require large amounts of energy to be injected into the forward shock E 0 = erg E inj = k  E 0 with k = 2, 10, 100 E fs = E 0 + E inj → efficiency crisis for the prompt phase but Ex: f mes = 0.1 ; k = 10, 100 → f true = 0.53, 0.92 ! Flares : late internal shocks ? But how to explain that ~ 100 s ~ 1000 s ~ s

New ingredients Plateaus: an initially inefficient afterglow ? “missing” energy” Let us assume: a wind external medium  e ∝ n - (n > n crit ) and constant for n < n crit ∝ R 2 → flat plateau for ~ 1 E fs (Hascoët et al, 2014) =1 ; A * =1 (Margutti et al, 2012)

New paradigm Making the early afterglow with a long-lived reverse shock Standard picture: the reverse shock is short-lived; it rapidly crosses the high  ejecta, heats the electrons → slow cooling electrons radiate in optical/IR: early optical flash ( GRB ; Sari & Piran, 1999 ) Alternative proposal: ejecta has a low  tail → the reverse shock is long-lived (Genet et al, 2007; Uhm & Beloborodov, 2007) Emission from the reverse shock sensitive to the distribution of energy in the ejecta → great flexibility in light curve shapes (Uhm et al, 2012) Plateaus injected power in the tail FS

Steep slopes But what about the steepest slopes? (internal plateaus) → magnetar activity ? (Lyons, O’Brien, Zhang et al, 2010) (luminosity/duration of the plateau ↔ energy reservoir of a magnetar) -9

Flares After completion of internal shocks the ejecta is highly structured Additional assumption: anisotropy of the radiation field in comoving frame FS (simulation by F. Daigne)

High energy afterglow emission Forward shock synchrotron emission… + inverse Compton (for the highest energy photons) ( s and ks in GRB ) Alternative: pair loading and heating at the blast wave (Vurm, Hascoët, Beloborodov, 2014) The pairs make: synchrotron emission → optical flash IC scaterring with prompt/early afterglow photons → GeV emission GRB

Conclusion How to make new progress? Expect a Rosetta stone burst: GRB A ? Enter a new era: SVOM (2020) an improved spectral coverage of the prompt emission GWACs: a real-time coverage in optical of ECLAIRs fov GFTs: dedicated follow-up telescopes The multi-messenger era neutrinos, cosmic rays, gravitational waves… → new clues on GRB physics ? (shock waves, magnetization) GWAC GFT