Outline. Spectra of Fermi/LAT GRBs? Physical Origins of GeV emission? Summary

Outline Spectra of Fermi/LAT GRBs? Physical Origins of GeV emission? Summary

Prompt GRB Spectra : Clear Observations for >20 years 1991-2000 BATSE Power Law Breaks (Schaefer 1992) Band function (1993)

Prompt GRB Spectra : Clear Observation for >20 years 1991-2000 BATSE HETE2 Power Law Breaks (Schaefer 1992) Band function (1993) 2001-2005 INTEGRAL 2002-now Swift 2004-now HETE-2 040916 Integral 030227 Swift 101225A Power Law Cutoff-Pl Band Function

Prompt GRB Spectra : Clear Observation for >20 years 1991-2000 BATSE HETE2 Outliers? 911031 Thermal Bump Ryde 2006 Power Law Breaks (Schaefer 1992) Band function (1993) 2001-2005 INTEGRAL 2002-now Swift 2004-now HETE-2 040916 Integral 030227 Swift 101225A Power Law Cutoff-Pl Band Function

Prompt GRB Spectra : Clear Observation for > 20 years 1991-2000 BATSE HETE2 Power Law Breaks (Schaefer 1992) Band function (1993) 2001-2005 INTEGRAL 2002-now Swift 2004-now Fermi 2008- now HETE-2 040916 Integral 030227 Swift 101225A Power Law Cutoff-Pl Band Function

Fermi Era: Still a Featureless Band Function?

Fermi Era: Still a Featureless Band Function? Most likely NOT: photosphere thermal component second SSC bump pair-production cutoff

Asaf Pe'er, Peter Mészáros, and Martin J. Rees, 2006

What Does Fermi See? Zhang, Bin-Bin, et al. 2011 ApJ in press (Arixiv 1009.3338)

Highlights on Data Analysis : Time-Dependent Spectral Evolution in Finest Time Resolution for All The Fermi Bursts (This wok focus on LAT-only bursts)

Two Distinct Types Three Elemental Spectral Components

Two Distinct Types of GRBs 1 : Still Band Function

Two Distinct Types of GRBs 1 : Still Band Function 2 : Thermal Spectra

080916C: Band Function Does NOT narrow with reducing time bins

090902B Thermal : progressively narrowing with reducing time bins and extra PL

Three Elemental Spectral Components (observationally identified) I : Band-function component (extend up to GeV) See also in N. Omode s Talk II: Quasi-thermal Component III: extra non-thermal power law component extending to high energies. Observed spectra are combinations of the three.

component I only e.g 080916C, 14 out of 17

Components II+III 090902B, probably 090510

Components I+II X-Ray Excess 100724B, 38KeV, Sylvain Guiriec et al 2010 081221, 8keV, Shu-Jin Hou, 2010, in prep 911031, 56 kev, themal bump Ryde + 2006

Components II+III Possibly 090926A (Ackermann et al 2011, He Hao-Ning 2011 in prep)

Components I+II+III

Physical Origins of The Three Elemental Spectral Components

Band Function Component 14 out 17 bursts are fitted by Band Facts:

Band Function Component Facts: 14 out 17 bursts are fitted by Band Internal Non-thermal Emission 6-7 orders extension

Band Function Component Facts: 14 out 17 bursts are fitted by Band Internal Non-thermal Emission 6-7 orders extension Time-dependent Band Function in Finest time resolution

Band Function Component Facts imply : Poynting flux dominated outflow 14 out 17 bursts are fitted by Band Internal Non-thermal Emission 6-7 orders extension Time-dependent Band Function in Finest time resolution NO photosphere thermal component most energy carried in B filed, not in photons, photosphere emission is suppressed NO second SSC bump Compton parameter Y<<1 (because B energy density >> photon energy density, Y=U_ph/U_B<<1), so SSC is suppressed (L_ssc~YL_syn) Zhang & Pe'er 2009

Band Function Component How to make it? β: -2.2, high energy PL component :

Band Function Component How to make it? β: -2.2, high energy PL component : synchrotron

Band Function Component How to make it? β: -2.2, high energy PL component : synchrotron SSC (synchrotron self-compton) See also P. Beniamini D. Guetta E. Nakar, T. Piran, 2011, Arxiv : 1103.0745 and T. Piran s talk Compton up-scattering of a thermal photon source (Thompson et al 1994, Beloborodov 2010, Lazzati et al 2010, Kenji et al. 2010)

Band Function Component How to make it? β: -2.2, high energy PL component : - synchrotron - SSC (synchrotron self-compton) See also P. Beniamini D. Guetta E. Nakar, T. Piran, 2011, Arxiv : 1103.0745 - Compton up-scattering of a thermal photon source Issue: can not easily produce the >GeV photons (Thompson et al 1994, Beloborodov 2010, Lazzati et al 2010, Kenji et al. 2010. - See also B. Morsony s talk )

Band Function Component How to make it? α: -1, Low energy PL component hard to explain

Band Function Component How to make it? α: -1, Low energy PL component hard to explain - thermal emission gives : +1

Band Function Component How to make it? α: -1, Low energy PL component hard to explain - thermal emission gives : +1 --with sub-photosphere heating gives 0.4 Beloborodov,A 2010 ; Deng, W (poster ) Mizuta,A 2011 & poster. - syn gives -1.5 - not simple multi-color BB effect (cf Toma+ 2010) - high-latitude emission effect of PS emission : too late - Deng, Wei (poster), Asaf & Ryde, 2010 Asaf & Ryde, 2010

Beloborodov & Vurm s talks (picture taken from Beloborodov,A s talk)

(picture taken from Beloborodov,A s talk)

Quasi-Thermal Component Thermal Emission from Fireball Photosphere [Photosphere emission when relativistic outflow turn optically thin] ( Paczynski 1986; Goodman 1986; Rees & Meszaros, 2005; Pe'er et al. 2006; Thompson et al. 07; Pe'er & Ryde 2010; Beloborodov, 2010; Lazzati et al 2009,2011; Toma et al 2010, Ioka 2010, Ryde et al 2011 arixv: 1103.0708 )

Extra Simple Power law component Not Straightforward See also Medvedev s talk

Extra Simple Power law component Not Straightforward Accompany with BB, or Band

Extra Simple Power law component Not Straightforward Accompany with BB, or Band Extended to GeV, also existed in low energies

Extra Simple Power law component Not Straightforward Accompany with BB, or Band Extended to GeV, also existed in low energies NOT straightforward to expected since theoretically non-thermal GRB spectra should be curved (Pe'er et al 2006, Gupta & Zhang 2007, Asano & Terasawa 2009 ) For 090902B, it might be a combination of Syn emission (low energy dominated) and SSC and Comptonization of thermal photons (Pe'er 2010)

Extra Simple Power law component Not Straightforward Tracking of thermal and non-thermal suggests that they might be from the same origin Accompany with BB, or Band Extended to GeV, also existed in low energies NOT straightforward to expected since theoretically non-thermal GRB spectra should be curved (Pe'er et al 2006, Gupta & Zhang 2007, Asano & Terasawa 2009 ) For 090902B, it might be a combination of Syn emission (low energy dominated) and SSC and Comptonization of thermal photons (Pe'er 2010)

Origin of GeV Emission

PS Where dose the High Energy emission come from? Internal? - External?

A Dilemma (e.g, 080916C) GBM GBM LAT LAT In spectral domain, one single component (GeV component seems to be natural extentension of Lower energy parts)

A Dilemma (e.g, 080916C) GBM GBM LAT LAT In spectral domain, one single component (GeV component seems to be natural extentension of Lower energy parts) In light curve domain, two different components. (Different decay slopes)

A Dilemma (e.g, 080916C) GBM GBM LAT LAT In spectral domain, one single component (GeV component seems to be natural extentension of Lower energy parts) How to solve the dilemma? In light curve domain, two different components. (Different decay slopes)

A Dilemma (e.g, 080916C) Internal GBM GBM External LAT LAT In light curve domain, two different components. (Different decay slopes)

A Dilemma (e.g, 080916C) Internal GBM GBM External LAT LAT Have to interpret.

Solution 1: Superposition in Spectrum Internal GBM GBM External LAT LAT Have to interpret.

Solution 1: Superposition in Spectrum A Big Issue: Since observation shows Band function in every time slice. ---> have to assume superposition of external and internal components in each time slice to make a Band function umar, & Barniol Duran, 2010.

A Dilemma of Superpositions Superposition in Spectrum Plus, our simulation shows that the external shock model can NOT reproduce a so steep light curve during the prompt emission phase (Amanda Maxham, Bin-Bin Zhang And Bing Zhang, 2011 submitted,arxiv:2011.0144 )

Maxham, Zhang B-B and Zhang,B, 2011, MNRAS, in press Using the MeV prompt emission (GBM) data record the energy output from the central engine as a function of time

Maxham, Zhang B-B and Zhang,B, 2011, MNRAS, in press Using the MeV prompt emission (GBM) data record the energy output from the central engine as a function of time (assuming a constant radiative efficiency) track energy accumulation in the external shock calculate the high energy emission lightcurve in the LAT band Our Simulation shows that : The late time LAT light curves after T90 can be well fit by the model. The early external shock emission cannot account for the observed GeV flux level.

A Dilemma (e.g, 080916C) GBM GBM LAT LAT Have to interpret

Solution 2 Superposition in Light Curve GBM GBM LAT LAT Have to interpret

Solution 2 Superposition in Light Curve a dominant internal emission component + gradually enhancing external shock component Maxham, Zhang B-B and Zhang,B 2011, MNRAS, in press Prompt Emission Phase

Further Evidence for Internal Origin of Prompt GeV Rough Tracking Light Curve btw GBM and LAT

Further Evidence for Internal Origin of Prompt GeV Rough Tracking Light Curve btw GBM and LAT For 080916C, GeV peak Coincides the Second Peak of GBM light Curve (logscale) cf, talk by Ghisellini

If Prompt GeV emission is internal, how to interpret the onset delay? Possibility 1: Possibility 2: hange of particle acceleration echanism arly on, the particle acceleration process ay not be so efficient, so the electron nergy spectral index is steep

If Prompt GeV emission is internal, how to interpret the onset delay? Possibility 1: hange of particle acceleration echanism arly on, the particle acceleration process ay not be so efficient, so the electron nergy spectral index is steep Possibility 2: Change of Opacity particle acceleration process is the same, the pair-production opacity changes early on. supporting evidence: 1) GBM alone spectra give similar beta with later epochs 2) gradual increase of the maximum energy with time

Summary Prompt GeV Emission: Internal or External? Internal External

Summary Prompt GeV Emission: Internal or External? Band Extends to GeV Internal External Spectrum Slope: IS and ES spectra mimic a same BAND function in ALL the Time Bins (Kumar & Barniol Buran 2009) MORE contrived

Summary Prompt GeV Emission: Internal or External? Band Extends to GeV Onset delay of LAT emission Internal External Spectrum Slope: IS and ES spectra mimic a same BAND function in ALL the Time Bins (Kumar & Barniol Buran 2009) MORE contrived Change of particle acceleration Mechanism Change of Opacity

Summary Prompt GeV Emission: Internal or External? Band Extends to GeV Onset delay of LAT emission LC : Roughly tracking behaviors (MeV vs GeV) Internal Change of particle acceleration Mechanism Change of Opacity External Spectrum Slope: IS and ES spectra mimic a same BAND function in ALL the Time Bins (Kumar & Barniol Buran 2009) MORE contrived

Summary Prompt GeV Emission: Internal or External? Band Extends to GeV Onset delay of LAT emission LC : Roughly tracking behaviors (MeV vs GeV) Single Decay Slope of LAT LC from Prompt to AG Internal Change of particle acceleration Mechanism Change of Opacity superposition effect Light Curve Slope : require same decay slope at the transition epoch [contrived] gradual die-off of the central engine activity External Spectrum Slope: IS and ES spectra mimic a same BAND function in ALL the Time Bins (Kumar & Barniol Buran 2009) MORE contrived (but can not be that steep, during prompt phase, Maxham et al 2010)

Summary Prompt GeV Emission: Internal or External? Band Extends to GeV Onset delay of LAT emission LC : Roughly tracking behaviors (MeV vs GeV) Single Decay Slope of LAT LC from Prompt to AG For 080916C, GeV peak Coincides the Second Peak of GBM light Curve Internal Change of particle acceleration Mechanism Change of Opacity superposition effect Light Curve Slope : require same decay slope at the transition epoch [contrived] gradual die-off of the central engine activity External Spectrum Slope: IS and ES spectra mimic a same BAND function in ALL the Time Bins (Kumar & Barniol Buran 2009) MORE contrived (but can not be that steep, during prompt phase, Maxham et al 2010) ES Deceleration time= 2nd central engine activity time Fine Tuned Bulk Lorentz Factor --Highly contrived

A bright GRB co-triggered by Fermi LAT/GBM and Swift will help

A bright GRB co-triggered by Fermi LAT/GBM and Swift will help 090510? But too weak Pelassa et al 2010

PS Where dose the High Energy emission come from? Internal? - External? - Both ()

PS Where dose the High Energy emission come from? Internal? - Entire Band Function is from Sync Emission - External? - Sync emission from External forward shock - Both ()

Summary Three Elemental Spectral Components are observed in Fermi/LAT GRBs Prompt GeV emission is likely of internal origin. Long term GeV emission is likely of external origins. Further Co-triggered Bursts by Swift and Fermi would help to narrow down the physical origins of GeV emissions.

Q&A 1. Band or BB: due to Binning effect? 2. How to Change the particle acceleration Mechanism 3. What Changes the Opacity 4. What are the possibilities for GeV Onset Delay? 5. Constrains from Observed kev-gev Band Function 6. How to choose Models?

Band or BB: due to Binning effect? No Time bins, Δt/(1+z), in rest frame for 080916C and 090902B are comparable

How to Change the particle acceleration Mechanism UNKNOWN May be related to the initial B configurations

What Changes the Opacity Lorentz factor are smaller in early times?

What are the possibilities for GeV Onset Delay? Up-scanter cocoon (Kenji Toma,2010) Hardronic (pγ, pp, etc) process, Razzaque et al 2008,2010) External Forward Shock Change of particle acceleration Mechanism Change of Opacity

Constrains from Observed kev-gev Band Function NO photosphere thermal component most energy carried in B filed, not in photons, photosphere emission is suppressed NO second SSC bump Compton parameter Y<<1 (because B energy density >> photon energy density, Y=U_ph/U_B<<1), so SSC is suppressed (L_ssc~YL_syn) NO pair-production cutoff Poynting flux : larger emission radius reduce 2-photon annihilation opacity and increase the pair cutoff energy

How to choose spectral models? The following spectral functions are considered : PL, CPL, BB, Band (1) If a one-component model can adequately describe the data (giving reasonable reduced 2, say, between 0.75 and 1.5), two-component models are not considered; (2) for onecomponent models, if a function with less free parameters can describe the data adequately, it is favored over the models with more parameters. (3) In addition, the Akaike s Information Criterion2 (AIC, Akaike 1974) is calculated to evaluate each model by considering both the fitting goodness and the complexity of the model.