1. L. Amati, E. Maiorano, E. Palazzi, R. Landi, F. Frontera, N. Masetti, L. Nicastro, M. Orlandini INAF-IASF Bologna (Italy) Unveiling GRB hard X-ray afterglow emission with Simbol-X

2. The GRB phenomenon  prompt emission

3.  afterglow emission

4. BeppoSAX era (1996-2002) prompt afterglow  prompt –> afterglow Swift era (2005-)

5. LONG (2 s – 2000s)  0.1 10 54 erg  possibly collimated emission  evidence of dense metal rich circum- burst environment  located in star forming region of high SFR host galaxies  GRB-SN (hypernovae) connection  origin: death of peculiar high mass stars (collapsar scenario)  still a few afterglow detections and z estimates (z < 1)  energy budget up to ~5x10 51  host galaxies: no clear distinction with those of long  origin: merging of compact objects (NS-NS, NS-BH, …) SHORT (0.001-2s)

6. … but prompt and afterglow emission mechanisms still to be settled !  ms time variability + huge energy + detection of GeV photons -> plasma occurring ultra-relativistic (  > 100) expansion (fireball)  non thermal spectra -> shocks synchrotron emission (SSM)  fireball internal shocks -> prompt emission  fireball external shock with ISM -> afterglow emission

7.  deviations form synchrotron predictions observed in prompt emission spectra of a fraction of GRBs  physics of prompt emission still not settled, various scenarios: SSM internal shocks, IC-dominated internal shocks, external shocks, photospheric emission dominated models, kinetic energy dominated fireball, poynting flux dominated fireball) Frontera et al. (2000)

8. Afterglow emission: less complex … Sari et al. (1998, 1999) The multi-wavelength afterglow emission is modeled as due to synchrotron. F(t, )  t -α -β with α and β depending on p, where N(E)  E -p is the electron energy distribution.

9. …but standard model not always works ! SED of GRB 970508: fit with standard synchrotron shock model in slow cooling regime is OK SED of GRB 000926: excess of X-ray emission with respect to synchrotron prediction: IC component ? Galama et al. (1997)Harison et al. (2001)

10. The puzzling case of GRB990123 - I Only one case of afterglow emission clear detection at energies > 15 keV: the bright GRB 990123 by BeppoSAX/PDS The 15-60 keV flux is inconsistent with the lower energy spectrum and synchrotron emission models predictions Maiorano et al. (2005)

11. The puzzling case of GRB990123 - II the fit with a synchrotron + IC component is more satisfactory, but still problems with the “closure relationships” between spectral and decay indices alternative explanations include peculiar circum-burst properties and/or peculiar shock physics this shows the relevance of sensitive measurements of GRB hard X-ray afterglow emission Corsi et al. (2005)

12. The ambiguous case of GRB 990806 Hard X-ray afterglow emission might have been detected also for the BeppoSAX GRB990806 but the presence of another X-ray source in the PDS field of view, as LECS and MECS images show (Fig. 5), make the detection quite uncertain relevance of hard X-ray imaging From BeppoSAX ASDC archive

13. Afterglow X-ray emission with Simbol-X - I Unprecedented sensitivity 15-60 keV: less than 1  Crab (several hundreds times better than BeppoSAX/PDS) Imaging capability comparable to lower energy X-ray telescopes at 11hr from the GRB 1/3 of GRB afterglows show a flux > 100 μcrab Critical issue: time needed to be on-target for a TOO observation (12 hours ? 1-2 days ?) Pareschi & Ferrando (2006)

14. 10% brightest BeppoSAX (and Swift) afterglows show a 2-10 keV flux > ~230 μCrab (~5x10 -12 erg/cm2/s) at 11hr from the burst by assuming the average photon index of 2.2 (~Crab-like) and the average temporal decay index (1.3), the expected 15-60 keV flux at 48 hr is about ~35 μCrab, and the average flux from 48hr to 76 hr (corresponding to a 100 ks long observation period) is ~25 μCrab even a 100 ks Simbol-X observation starting at 2 days after the GRB will allow a sensitive spectral measurement in 15-60 keV Afterglow X-ray emission with Simbol-X - II De Pasquale et al. (2006)

15. Simulation of Simbol-X image and spectrum of a bright afterglow (N>F ~10%) with no IC component observed with a 100 ks TOO starting 2 days after the GRB (assumed decay index 1.3, the average flux is 25  Crab) clear detection (about 18  ) in the image and well determined spectral shape in 10-60 keV this simulation is equivalent to a 100 ks observation of an afterglow of medium intensity observed after 12 hr from the GRB Afterglow X-ray emission with Simbol-X - III

16. Afterglow X-ray emission with Simbol-X - IV Simulation of Simbol-X spectrum of GRB 990123 (including IC component) with a 100 ks TOO starting 2 days after the GRB (decay index of 1.3 assumed, the average flux is 25  Crab) the IC excess in 20-60 keV is clearly visible in the residuals and has a 6.5  significance this simulation is equivalent to a 100 ks observation of an afterglow of medium intensity, showing an IC component and observed after 12 hr from the GRB Maiorano et al. (2005) BeppoSAX at 6-10 hr Simbol-X at 48-72 hr

17. Who will provide GRB detection and localization ? These simulations are very conservative: it will be likely possible to perform TOO observations of brigth afterglows at 12/24 hr from GRB onset (fluxes of 210-80  Crab for the brightest 10% assuming decay index 1.3) GRB detection and localization to a few arcmin is necessary: who will provide it in the >2012 time frame ?  Swift (operating since December 2004, detection and arcsec localization), Chandra, XMM, Suzaku (afterglow localization to a few arcsec, flux, decay index), AGILE (GRB detection and a few arcmin localization): who knows ?  GLAST (GRB detection and possibly few arcmin localization): likely  SVOM/ECLAIRS (GRB detection and few arcmin localization): likely  EDGE (GRB detection and localization, afterglow few arcsec localization, flux, decay index), Lobster (GRB detection and a few arcmin localization),…: maybe  Optical telescopes (afterglow localization, brightness, decay index): always Information on afterglow brightness and decay slope is also important to decide to perform a TOO; if X-ray information is lacking, these can be inferred from prompt emission intensity and optical afterglow intensity and decay slope

18. Conclusions  despite the enormous observational progress occurred in the last 10 years, the GRB phenomenon is still far to be fully understood  one of the main open issues is the understanding of physical mechanisms at the basis of prompt and afterglow emission  the case of GRB 990123 shows that measurements of the nearly unexplored GRB hard (> 15 keV) X-ray afterglow emission can provide very stringent test to emission models  thanks to its unprecedented sensitivity in the 15-60 keV energy band, Simbol-X can provide a significant step forward in this field  simulations based on observed distribution of X-ray afterglow fluxes and spectral and decay indices show that even a 100 ksTOO observation starting 2 days after the GRB can provide sensitive spectral measurements and allow to discriminate different emission components for a significant fraction of events  it is likely that significantly lower TOO stat times (12/24 hr) will be possible for a few event/year  the needed GRB detection and few arcmin localizations will be provided by space missions likely flying in the >2012 time frame and optical telescopes

19. ● ms time variability + huge energy + detection of GeV photons -> plasma occurring ultra-relativistic (Γ > 100) expansion (fireball) ● non thermal spectra -> shocks synchrotron emission ● fireball internal shocks -> prompt emission ● fireball external shock with ISM -> afterglow emission The “fireball” model

20. LONG ● energy budget up to >10 54 erg ● long duration GRBs ● metal rich (Fe, Ni, Co) wind circum-burst environment ● GRBs occur in star forming regions ● GRBs are associated with SNe ● naturally explained collimated emission ● energy budget up to 10 49 – 10 50 erg ● short duration GRBs (< 2 s) ● clean homogeneous circum-burst environment ● GRBs in the outer regions of the host galaxy SHORT

21. The “fireball” model Ultrarelativstically expanding source releases a huge amount of energy (~10 51 erg) in a rather small volume (R~10-1000 km), in form of a fireball expanding with relativistic velocity. The central engine produces several shells with different Lorentz factors that can overtake each other and collide, causing a relativistic blast wave (Blandford & McKee 1976). In the internal-external model, when shells with different velocity collide each other (internal shocks) produce the gamma-ray burst,while the afterglow occurs when the fireball hits the surrounding material (external shoks) (Sari 1997).

22. Prompt and afterglow light curves WFC (top and central panel) and GRBM (bottom panel) light curves. Hard-to-soft evolution is present. Atmospheric absorption in the GRB tail (~80 sec) affected soft X-ray data. (Corsi et al. 2005) Multiwavelength light curves from the prompt event to the afterglow. t 0 corresponds to the time of GRB onset.

23. GRBM+WFC average spectrum is well fit with the Band function α = -0.89 ± 0.08 β = -2.45 ± 0.97 E b = 703 ± 32 keV Simultaneous multiwavelength spectra derived at three times during the burst (ROTSE V-band,X-ray and γ- ray) The prompt event (Corsi et al. 2005) t = 32 s t = 58 s t = 7 s (Amati et al. 2002)

24. The broadband spectrum of the afterglow. The dashed line is the best-fit power-law describing optical and NIR data β opt = 0.60 ± 0.04 The solid line is the power-law which best fits the X-ray data β x = 0.94 ± 0.07 The spectral turnover between optical and X-ray bands is identified with the presence of the synchrotron cooling frequency at ν c = 0.47 keV = 1.14 x 10 17 Hz

25. Closure Relations β x = 0.94 β x =p/2 We expect: α X - α opt = (3p-2)/4 – 3(p-1)/4 = 1/4 β X - β opt = p/2 – (p-1)/2 = 1/2 We observed: α X - α opt = 0.36 ± 0.05 β X - β opt = 0.34 ± 0.08 p ~ 2 β opt = 0.60 ν c = 1.1 x 10 17 Hz β opt = (p-1)/2 = 0.5 ν o < ν c < ν x but Both consistent with the value we found Fully consistent with the value we found Assuming isotropically adiabatic expansion within homogeneous medium β opt / α opt = 2/3 = 0.54 ± 0.04 (β X – 1/3) /α X = 2/3 = 0.30 ± 0.05 Both the measured ratios are statistically inconsistent with the expectations α x = 1.46 α opt = 1.10

26. Conclusions: GRB990123 One of the most energetic events and first case of prompt optical emission WFC detection of the afterglow already 20 min after the GRB Smooth GRB-afterglow light curve connection First X-ray afterglow detected up to 60 keV in the PDS During the BeppoSAX observation the X-ray afterglow decays faster than the optical one On 24.65 January 1999 UT the broadband afterglow of GRB990123 is consistent with a synchrotron spectrum with ν c located at the lower energy border of the X-ray range covered by BeppoSAX A self-consistent interpretations of the afterglow with pure synchrotron emission is not viable. Also the presence of IC component in the X-ray band (Corsi et al. 2005) does not overcome the inconsistency. A more complex model is required to solve the “puzzle”.

27. ● ms time variability + huge energy + detection of GeV photons -> plasma occurring ultra-relativistic (Γ > 100) expansion (fireball) ● non thermal spectra -> shocks synchrotron emission ● fireball internal shocks -> prompt emission ● fireball external shock with ISM -> afterglow emission The “fireball” model

28. Headlines of this talk GRB afterglow emission Hard X-ray emission

29. The puzzling case of GRB990123 - I LECS,MECS,PDS first 20 ks spectrum: an absorbed power-law with photon index Γ = 1.94 ± 0.07 best fits the data. N H (Gal) = 1.98 x 10 20 cm -2 SFD : radio, IR and optical flux densities together with the 2-10 keV flux observed on 24.65 January 1999 UT WFC MECS WFC PDS 15-28 keV 2-10 keV The dashed line is the best- fit decay obtained from the X-ray afterglow data. The extrapolation smoothly reconnects with the late time WFC data points and upper limits suggesting that the X- ray afterglow had already started ~ 20 min after the prompt event.