* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download ppt
Grand Unified Theory wikipedia , lookup
Monte Carlo methods for electron transport wikipedia , lookup
Bremsstrahlung wikipedia , lookup
ALICE experiment wikipedia , lookup
Atomic nucleus wikipedia , lookup
Nuclear structure wikipedia , lookup
Standard Model wikipedia , lookup
Future Circular Collider wikipedia , lookup
Relativistic quantum mechanics wikipedia , lookup
Eigenstate thermalization hypothesis wikipedia , lookup
Introduction to quantum mechanics wikipedia , lookup
ATLAS experiment wikipedia , lookup
Photoelectric effect wikipedia , lookup
Photon polarization wikipedia , lookup
Elementary particle wikipedia , lookup
Compact Muon Solenoid wikipedia , lookup
Theoretical and experimental justification for the Schrödinger equation wikipedia , lookup
Bulk Comptonization GRB Model and its Relation to the Fermi GRB Spectra Demosthenes Kazanas NASA/GSFC Apostolos Mastichiadis Un. Of Athens References: DK, M. Georganopoulos, A. Mastichiadis 2002 A. Mastichiadis, DK 2006 DK, A. Mastichiadis, M. Georaganopoulos 2007 A. Mastichiadis, DK 2009 There are (at least) two outstanding issues with the prompt GRB emission (Piran 2004): A. Dissipation of the RBW free energy. Energy stored in relativistic p’s or B-field. Sweeping of ambient protons stores significant amount of energy in p’s anyway. Necessary to store energy in non-radiant form, but hard to extract when needed. B. The presence of Epeak ~0.1 – 1.0 MeV. If prompt emission is synchrotron by relativistic electrons of Lorentz factor (LF) same as shock Ep ~G4, much too strong to account for the observations. We have proposed a model that can resolve both these issues simultaneously. The model relies: On a radiative instability of a relativistic proton plasma with B-fields due to the internally produced sychrotron radiation (Kirk & Mastichiadis 1992 ). 1. On the amplification of the instability by relativistic motion and scattering of the internally produced radiation by upstream located matter, a ‘mirror’ (Kazanas & Mastichiadis 1999). 2. D pg e+e- eB Bg In the blast frame the width of the shock D ~ R/G is comparable to its observed lateral width thereby considering all processes taking place in a spherical volume of radius D The instability involves 2 thresholds: (All particles behind the shock are assumed to have Lorentz factors equal to the shock one G, i.e. no accelerated particle populations!) 1. A kinematic threshold for the reaction p g e- e+ The photon at the proton rest frame must have energy greater than 2 me c2 . The synchrotron photon has an energy ES ~ bG2 at the plasma frame and ~ bG3 on the proton frame. So the condition reads b G3 > 2 me c2 or G > (2/b)1/3 2. A dynamic threshold : At least one of the synchrotron photons must be able to produce an e- e+ - pair before escaping the plasma volume. Because each electron produces Ng ~ G/bG2 ~ 1/bG photons, we obtain the following condition for the plasma column density (note similarity with atomic bombs!) nspgR > bG or nspgR G2 > bG3 ~ 2 Similarity of GRB/Nuclear Piles-Bombs The similarity of GRB to a “Nuclear Pile” is more than incidental: 1. They both contain lots of free energy stored in: Nuclear Binding Energy (nuclear pile) Relativistic Protons or Magnetic Field (GRB) 2. The energy can be released explosively once certain condition on the fuel column density (and not mass) is fulfilled (Note: no particle acceleration required!!). The instability requirements are greatly reduced if the radiation produced at the shock is mirrored by upstream located matter (because the energy of each photon increases by G2; DK, A. Mastichiadis 1999) to the following conditions: b G5 > 2 n spg R G4 > bG5 ~ 2 For n ~ 1, R ~ 1017R17, G ~ 211 (n R17)1/4, T ~ 40 sec THE SPECTRA Min(mec2 G1+1, bG6+1) ~ 100 GeV photons BC, SSC bG4+1 ~ 1 MeV photons (GBM photons are due to BC) BC RS bG2+1 ~ 10 eV O-UV photons Syn. RBW R/G2 Mirror We have modeled this process numerically. We assume the presence of scattering medium at R ~1016 cm and of finite radial extent. We follow the evolution of the proton, electron and photon distribution by solving the corresponding kinetic equations. We obtain the spectra as a function of time for the prompt GRB emission. The time scales are given in units of the comoving blob crossing time Dco/c ~ R / G2c ~ 2 R16/G2.6 sec. Prompt GRB Spectra (Mastichiadis & Kazanas 2006) BC: b G5 S : b G3 SSC : meG2 We have also modeled the propagation of a relativistic blast wave through the wind of a WR star (that presumably collapses to produce the relativistic outflow that produces the GRB). In this case we also follow the development of the blast wave LF and the radiative feedback on it. In this scenario the “scattering screen” necessary for the model to work is provided by the presupernova star wind, the length scales smaller, R0~1013 cm and the GRB is a “short GRB”. Evolution of the LF and the luminosity Evolution of Spectra with Time (eB~1) Evolution of Luminosity with Time (eB~0.01) Ep decreases with decreasing Luminosity. The BC component Decreases faster Than the SSC leavin The LAT flux after The GBM one is Gone. Aftreglow, GRB, XRF, Unification Inclusion of non-thermal particle populations and repeating the same arguments as above one obtains the evolution of Epeak with G or with time. A simple calculation gives (DK, AM, MG 2007) Ep ~ 4 10-2 [G(t)/50]2 ~ t-3/4 (talk/poster by R. Margutti on Monday) Conclusions The “nuclear pile” GRB model provides an over all satisfactory description of several GRB features, including the dissipation process, Ep and the Fermi observations. Provides an operational definition of the GRB prompt phase. No particle acceleration necessary to account for most of prompt observations (but it is not forbidden!). GBM photons due to bulk Comptonization ( => possibility of high polarization ~100%). It can produce “short GRB” even in situations that do not involve neutron star mergers. Exploration of the parameter space and attempt to systematize GRB phenomenology within this model is currently at work. Distribution of LE indices a The spectra of doubly scattered component (Mastichiadis & DK (2005)) S=1, a=-1 S=1/2, a=-3/2 S=2, a=0 120 G 12 G 1.2 G 0.12 G Fig. 6.| Plot of t he phot on luminosity evolut ion as t his is measured in t he comoving frame in t he EE case where np = ne = 104 cm¡ 3 while B varies (bot t om to t op) from 0.12 G t o 120 G by increment s of a fact or of 10. The rest of t he paramet ers are as st at ed in t he t ext . Time is measured in blob crossing t imes and t he value t = 0 has been set at t he inst ant when t he Eiso of the three different spectral components as a function of B for G=400 and np=105 cm-3. x 103 denotes the relative g-ray – O-UV normalization of GRB 990123, 041219a. 1 MeV 100 GeV X 103 O-UV Epeak as a function of the magnetic field B Variations If the “mirror” is in relative motion to the RBW then the kinematic threshold is modified to b G3 G2rel ~ 2; Grel is the relative LF between the RBW and the “mirror”. The value of Epeak is again ~ 1 MeV, however the synchrotron and IC peaks are higher and lower by G2rel than G2 . In the presence of accelerated particles the threshold condition is satisfied even for G< (2/b)1/5. This may explain the time evolution of GRB941017 (Gonzalez et al. 04) GRB flux is likely to be highly polarized (GRB 031206, Coburn & Boggs 03). This model applicable to internal shock model (photons from downstream shell instead of “mirror”). { 23 { Fig. 6.| Plot of t he phot on luminosity evolut ion as t his is measured in t he comoving frame in t he EE case where np = ne = 104 cm¡ 3 while B varies (bot t om to t op) from 0.12 G t o 120 G by increment s of a fact or of 10. The rest of t he paramet ers are as st at ed in t he t ext . Time is measured in blob crossing t imes and t he value t = 0 has been set at t he inst ant when t he RBW ent ers t he re° ect ion zone. Then .... to 4 > > ¾ p° R n » b¡ or ¾p° R n ¡ » 2 the source. Hence, the ¡ + e p° ! e e react ion 2 re N ' ° c =b° c = 1=b° c tron photons produced Shock – Mirror Geometry formation region, generally not much different than the reshold condit ion µ ¶ 1=5 2 5 > > ¡ b » 2 or ¡ » : b