Gamma-ray Burst Prompt Emissions: Fermi GBM Time-resolved - - PowerPoint PPT Presentation

4th FAN Workshop, University of Hong Kong, Hong Kong - 08.07.2013

Hoi-Fung Yu1

1Max-Planck-Institut für extraterrestrische Physik, Germany 2Universities Space Research Association, USA 3University of Huntsville in Alabama, USA

*Fermi GBM Principal Investigator

Gamma-ray Burst Prompt Emissions: Fermi GBM Time-resolved Spectroscopy

Collaborators: Jochen Greiner1*, William S. Paciesas2*, Robert D. Preece3, Charles A. Meegan3, Michael S. Briggs3, David Gruber1, Andreas von Kienlin1

Fermi GBM

• The Gamma-ray Burst Monitor

(GBM) is the second payload

• nboard Fermi, which is a joint

project of MPE and UAH

• GBM consists of 12 thallium

activated sodium iodide scintillation detectors (NaIs) and 2 bismuth germanate scintillation detectors (BGOs)

• NaIs cover lower energy

spectrum (8 keV - 1 MeV), BGOs cover higher energy spectrum (200 keV - 40 MeV)

NaI detector (Meegan et al. 2009) BGO detector (Meegan et al. 2009)

These detectors and the data read-out systems were actually built at MPE!

Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

Locations and orientations of the GBM detectors (Meegan et al. 2009)

• The axes of the NaI detectors are oriented

such that the positions of GRBs can be derived from the measured relative counting rates, a technique previously employed by CGRO BATSE (Meegan et al. 2009)

• Joint analysis of spectra and time histories of

Gamma-ray Bursts (GRBs) observed by both the GBM and LAT is possible for bright and hard GRBs

• Provide near-realtime burst locations
• nboard to permit repointing of the

spacecraft to obtain LAT observations of delayed emission from bursts, and to disseminate burst locations rapidly to the GRB community

• Provide excellent quality of spectral (128

energy channels CSPEC and TTE data) and temporal (up to 2 us TTE data) resolved data!

Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• Gamma-ray burst was first discovered in

1967 by Vela 3 & 4 satellites. In 1973 Klebesadel et al. estimated the rough sky positions of 16 bursts and ruled out solar system origin, thus confirming the cosmological nature of GRBs

• After ~ 40 years of research, we still don’t

have a consensus about the underlying physical processes producing GRBs

• There are two empirical types of GRBs,

namely the long and short GRBs. Usually the GRB research community adopts the definition that for short GRBs T90 < 2 s and for long GRBs T90 > 2 s (Kouveliotou et al. 1993)

• T90 is defined to be the time interval in

the observer’s frame such that the 5% - 95% counts of the prompt gamma-ray emission are observed within a certain energy range (usually 50-300 keV)

Kouveliotou et al. (1993) showing the bimodality of GRBs detected in the BATSE first year catalog Paciesas et al. (2012) showing the sky positions of 491 GRBs detected by Fermi GBM in the first 2 years of mission

GRBs

Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• Using dynamical arguments, the

prevailing physical explanation for GRBs is that the long GRBs are related to core-collapse massive stars, while the short GRBs are related to compact binary mergers (see, e.g., Kouveliotou, Woosley, & Wijers 2002; Woosley & Bloom 2006; Nakar 2007 for reviews)

• GRBs can release in gamma-ray

energies as high as Eiso ~ 1053 erg s-1 in the rest frame (Gruber et al. 2011)

• However, the emission process of

high energy photons is still poorly understood today

Gruber et al. (2011) showing the distribution of rest frame isotropic energies of 32 Fermi GBM bursts Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• The popular model among the GRB community is the so-called “fireball

model” (Cavallo & Rees 1978; Goodman 1986; Paczyński 1986) which assumes a large amount of energy is released in a small area resulting in high energy emissions in the gamma-ray energy band

• The prevailing explanation is that relativistic expanding shells of baryons colliding with

each other and with the interstellar medium (indeed not necessarily ISM, can also be the remaining materials from the progenitor), creating shock waves and converting kinetic and thermal energy into gamma radiation (see, e.g., Cohen et al. 1997)

• Observations suggest that the gamma-ray prompt emission is due to “internal shocks”

between expanding shells (for reviews see e.g. Piran 1999; Mészáros 2001) while the X-ray/optical/infrared/radio afterglow emission is due to “external shocks” between the shells and the ISM

Schematic drawing of the internal-external shock model of GRBs (Mészáros 2001) Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

GRB Spectral Properties

• The most frequently used model is the Band’s GRB function, first introduced by Band et al.

(1993):

• If beta < -2 then Epeak represents the peak energy in the vFv spectrum. The Band function has

been successfully fitting most of the GRB spectra (both time-averaged and time-resolved, e.g. Preece et al. 1998; Kaneko et al. 2006; Goldstein et al. 2012; Gruber et al. 2013, in prep.). However, it bears no physical origin and is completely empirical

• The simplest and most intuitive physical emission mechanism of the fireball model is the

synchrotron emission by electrons in the shocked materials, which provides theoretical predictions for the power-law indices and the break frequencies of the spectrum (Rees & Mészáros 1992, 1994; Mészáros & Rees 1993; Katz 1994; Tavani 1996)

• With certain physical assumptions, it predicts the value of alpha to be within -2/3 and -3/2

(the so-called “synchrotron line-of-death” and the “second line-of-death” respectively, see Preece et al. 2002). The difference in alpha and beta can also be used to constrain some physical parameters (Preece et al. 2002). There are also works done using physical models to fitting the data (e.g. the synchrotron cooling + blackbody model used in Burgess et al. 2011, 2013, in prep.)

Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• Two approaches to distinguish among various physical models:
• to fit multi-wavelength data simultaneously in order to obtain widest spectral

coverage (e.g. Kopac et al. 2013; Elliott, Yu et al. 2013, in prep.)

• to fit the data using time-resolved instead of time-averaged spectroscopy in
• rder to separate contributions from different emission episodes and to identify

spectral evolutions, e.g. Burgess et al. 2011; 2013, in prep.; Yu et al. 2013, in prep.)

• However, the progress of researches was slow until recently years, due to the

difficulty in obtaining multi-wavelength and high temporal resolution data for large GRB samples

Joint gamma-ray, X-ray, optical and infrared spectral fit of the recent GRB 121217A (Elliott, Yu et al. 2013, in prep.). This burst is unfortunately not bright enough to provide enough counts for a good time-resolved spectral analysis. preliminary Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• In this study we will give the most detailed time-resolved spectral analysis of Fermi GBM GRBs, thanks to the

high temporal (64 ms for TTE data) and spectral resolution (128 energy channels for both the NaIs and BGOs) of the Fermi GBM

• We selected bursts with 10 - 1000 keV fluence > 4 x 10-5 erg cm-2 and 50 - 300 keV peak flux > 55 ph s-1

cm-2, which consists the brightest Fermi GRBs for high quality time-resolved spectroscopy

• In the cases of weak BGO detections (i.e. relatively soft bursts), we include at least one of the BGOs to give

an upper limit in the > 1 MeV counts, which can constraint the high-energy index beta

• 54 out of the 954 GRBs in the 2nd Fermi GBM GRB catalog (von Kienlin et al. 2013, in prep.) match the above

selection criteria

Histogram plots of the energy fluence and peak photon flux of 954 GRBs detected in the first 4 years of Fermi mission (Yu et al. 2013, in prep.) preliminary preliminary > 4 x 10-5 erg cm-2 > 55 ph s-1 cm-2

Current Work: Fermi Time-resolved Spectral Catalog

Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

GRB 130606B

• We use the bright and hard burst

130606B to illustrate the works that could be showing in the time-resolved spectral catalog (Yu et al. 2013)

• We fitted individual time bins with the

following 3 spectral models and choose the best one with the lowest Casher C- statistics (C-stat, Cash 1979):

• Power-law
• Comptonized (high-energy cutoff)
• Band function
• Epeak traces the spectral hardness of the
• burst. There are basically two empirical

types of Epeak evolution: (1) power-law decay and (2) pulse-tracking behavior (see e.g. Preece et al. 2000; Kaneko et al. 2006)

• In this burst, the values alpha is consistent

with the -2/3 and -3/2 lines-of-death throughout the whole prompt emission

preliminary preliminary Spectral evolution of GRB 130606B with the 1 s binning CSPEC lightcurve overlaid Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• The temporal evolution of beta

can be used to trace the bulk lorentz factors of colliding shells (Lithwich & Sari 2001)

• In this burst the redshift is not

available, so we computed the minimum initial lorentz factor as a function of redshift, as described in Gruber et al. 2010, according to the formulae given in Lithwich & Sari 2001:

preliminary preliminary Estimation of the minimum (initial) lorentz factors of different emission episodes of GRB 130606B as a function of redshift Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013

• Spectral lags between high- and low-energy photons are observed in GRBs
• Using the cross-correlation function (CCF, Band 1997; Norris, Marani, & Bonnell 1997) for the

denoised lightcurves of different spectral ranges we can also compute the spectral lags between 25 - 50 keV and 100 - 300 keV (and higher energy ranges if the burst is hard, see e.g. Foley et al. 2008; Gruber et al. 2010)

preliminary preliminary preliminary preliminary Hoi-Fung Yu: Fermi GBM Time-resolved Spectroscopy - 08.07.2013