GRB prompt emission: spectral energy distribution and light curve profile Francesco Massaro Harvard (SAO – CfA) Thanks to: Barcellona J. Grindlay, R. Preece, N. Omodei June 2011
Modeling GRBs Since the discovery of GRBs… Two major descriptions of their SEDs have been used: Band fucntion and Blackbody+powerlaw while the lightcurves have been described with the Norris profile We propose a new view for both the 1. Spectral Energy Distributions (SEDs) and 2. Light Curves (LCs)
FACTS…and NOT ARTIFACTS Observational evidences (random order): 1. SEDs curved and broadly peaked 2. Single pulse or Spiky lightcurves 3. Cosmological distances 4. Variability classes (long and short GRBs) 5. Presence of afterglows 6. Relativistic effects 7. Supernova connection 8. HIC during the LC decay phase 9. GeV emission (not all of them) 10. Presence of X-ray flares (not all of them)
Spectral Energy Distribution of the GRB prompt emission
AIMS We aim to find a model for the SED that: 1. Could describe at least 90% of GRBs 2. Must describe the TIME RESOLVED SPECTRA (not only the time integrated) 3. Must interpret the majority of the observed FACTS 4. Must work directly on the observed spectra (not on the “deconvolved” SED) 5. Must have a strong physical background 6. MUST NOT IMPROVE THE NUMBER OF PARAMETERS Ockham Razor (Frustra fit per plura quod fieri potest per paucior)
Since 1992…The Band Function SWIFT, Fermi GBM, Multifrequency ToO obs. new planned missions (we hope) … but still …
Since 1992…The Band Function Band et al. 1993 Tavani et al. 1996 The SED shape is well described by this phenomenological model
The thermal emission Physical background !!! (Ryde 2004) My view…it still has some problems 1. No signatures exponential cutoff 2. Power-law always necessary 3. Extrapolation of the power-law 4. No BB photon index at low energies 5. Connection low-high energies 6. It cannot describe all GRBs 7. No high values of curvature
The log parabolic spectral shape Massaro, Grindlay & Paggi 2010 Log-parabola Energy dependent photon index A new physical parameter: the spectral curvature b
The log parabolic spectral shape Log-parabolic means log-normal Parabola is the natural way to approximate functions around a minimum or a maximum --> e.g. Taylor series
The physics!!! Since Kardashev 1962 The general solution of the kinetic equation is well approximated by a log-parabolic function when: 1. Not only Systematic but also stochastic acceleration 2. Radiative cooling + adiabatic expansion ……etc. etc. Similar ideas: Ellison et al. 2001, Pelletier et al. 2003, Stawarz & Petrosian 2006.
The log-parabolic synchrotron spectra N( γ )=N 0 ( γ / γ 0 ) - s - r Log γ / γ 0 Curvatures b ~ r/5 F( ν )=F 0 ( ν / ν 0 ) - a - b Log ν / ν 0 b (BL Lacs): 0.05 – 0.5 b (GRBs): 0.2 – 1.2 (time resolved spectra)
GRB 090902B
GRB 090902B
GRB 090902B
GRB 090902B
GRB 090902B
Adiabatic expansion Hp. Self similar scenario: Possible interpretation of the hardness intensity correlation (HIC): so The HIC has a peak index of ~ 1.6 Massaro & Grindlay 2011
Using the log-parabola…. We can test this idea
Spectral curvature behavior Simulated spectral evolution Observed spectral behavior Adiabatic losses do not change the shape of the SED We do not see drastic variations of the curvature during GRB single pulses Massaro & Grindlay 2011
A new feature of GRBs No high values of spectral curvature (Blackbody expected b~10)
A note on the synchrotron scenario
Synchrotron: line of death or small pitch angle? Photon index A clear signature of Synchrotron emission Then LCs…. Llody & Petrosian 2002
Light curves of the GRB prompt emission
Since 1996 and 2005… The Norris profile Single long pulse GRBs Norris et al. 2005 It is always asymmetric The LC profile can be described by this phenomenological model
Since 1996 and 2005… The Norris profile An example of an artifact….. Rise and decay time ratios is ~ 1/2 1,000,000 of Montecarlo simulations with uniform distribution Massaro & Grindlay 2011 in prep.
Modified Beta Function Vetere et al. 2006 3C 273 Abdo et al. 2010 Massaro & Grindlay 2011 in prep.
FACTS Observational evidences: 1. SEDs curved and broadly peaked 2. Spiky lightcurves 3. Cosmological distances 4. Fast variability (long and short GRBs) 5. Presence of afterglows 6. Relativistic effects 7. Supernova connection 8. HIC during the LC decay phase 9. No large variations of the spectral curvature 10. GeV emission (not all of them) 11. Presence of X-ray flares (not all of them) 12. Modified Beta profile to describe LCs (symmetric and asymmetric profiles)
CONCLUSIONS 1. Log-parabola vs Band From the statistical point of view 4 or 5 vs 3 parameters (in agreement with Fermi LAT GRBs detections) From the physical point of view a priori physical background 2. Time resolved spectra are very well described in terms of log-parabolic model (up to now no exceptions) 3. No drastic variation of the spectral curvature during GRB single pulses (CAREFUL must be tested) (signatures of adiabatic expansion) 4. A new idea to describe the LCs: modifed Beta function More versatile than the exponential profile and without degeneracies or biases
AIMS for the SED model And we came out with a model 1. Could describe at least 90% of GRBs 2. Must describe the TIME RESOLVED SPECTRA (not only the time integrated) 3. Must interpret the majority of the observed FACTS 4. Must work directly on the observed spectra (not on the “deconvolved” SED) 5. Must have a strong physical background 6. MUST NOT IMPROVE THE NUMBER OF PARAMETERS Ockham Razor (Frustra fit per plura quod fieri potest per paucior)
And thanks for your attention
Backup slides
GRB conference in Rome 2004 R. Blandford concluding remarks
Curved spectra in jets Vela plerion (Mangano et al. 2005) Cygnus A (FR II) (Carilli et al. 1991) Crab Pulsar (Campana et al. 2008)
Curved spectra in jets Mrk 501 BL Lacs (Massaro et al. 2006, 2008) Mrk 501 Log-parabolic model Mrk 421
Curved spectra in jets Radio galaxies (Katz-Stone et al. 1993) GPS radio sources (Ostorero et al. 2009) High Frequency Peakers (Maselli et al. 2009)
The synchrotron line of death Preece et al. 1998
The synchrotron line of death Asymmetric log-parabola
X-ray flares in GRB afterglows SWIFT time resolved spectral analysis 1. No drastic variations of b 2. Inconsistent with thermal (i.e. Blackbody) emission 3. Same model adopted for GRB prompt emission
Recommend
More recommend
