Evidence has been accumulated on the existence of a thermal-like component during the prompt phase of GRBs. This component, often associated with the GRB jets photosphere, is usually subdominant compared to a much stronger non-thermal one. The prompt
emission of Fermi GRB 131014A provides a unique opportunity to study this thermal-like component. Indeed, the thermal emission in GRB 131014A is much more intense than in other GRBs and a pure thermal episode is observed during the initial 0.16 s. The thermal-like component cools monotonically during the first second while the non-thermal emission kicks off. The intensity of the non-thermal component progressively increases until being energetically dominant at late time. This is a perfect scenario to disentangle the thermal component from the non-thermal one. A low-energy spectral index of +0.6 better fit the thermal component than the typical index value +1 corresponding to a pure Planck function. The non-thermal component is adequately fitted with a Band function whose low and high energy power law indices are ~-0.7 and <~-3, respectively; this is also statistically equivalent to a cutoff power law with a ~-0.7 index. This is in agreement with our previous results. Finally, a strong correlation is observed between the time-resolved luminosity of the non-thermal component, L$_i^{nTh}$, and its corresponding rest frame spectral peak energy, E$_{peak,i}^{rest,nTh}$, with a slope similar to the one reported in our previous articles. Assuming this relation to be universal for all GRBs we estimate a redshift of ~1.55 for GRB 131014A that is a typical value for long GRBs. These observational results are consistent with the models in which the non-thermal emission is produced well above the GRB jet photosphere but they may also be compatible with other scenarios (e.g., dissipative photosphere) that are not discussed in this article.
We reanalyze the prompt emission of two of the brightest Fermi GRBs (080916C and 090926A) with a new model composed of 3 components: (i) a thermal-like component--approximated with a black body (BB)--interpreted as the jet photosphere emission of a m
agnetized relativistic outflow, (ii) a non-thermal component--approximated with a Band function--interpreted as synchrotron radiation in an optically thin region above the photosphere either from internal shocks or magnetic field dissipation, and (iii) an extra power law (PL) extending from low to high energies likely of inverse Compton origin, even though it remains challenging. Through fine-time spectroscopy down to the 100 ms time scale, we follow the smooth evolution of the various components. From this analysis the Band function is globally the most intense component, although the additional PL can overpower the others in sharp time structures. The Band function and the BB component are the most intense at early times and globally fade across the burst duration. The additional PL is the most intense component at late time and may be correlated with the extended high-energy emission observed thousands of seconds after the burst with the Fermi/Large Area Telescope (LAT). Unexpectedly, this analysis also shows that the additional PL may be present from the very beginning of the burst. We investigate the effect of the three components on the new time-resolved luminosity-hardness relation in both the observer and rest frames and show that a strong correlation exists between the flux of the non-thermal component and its E$_{peak}$ only when the three components are fitted simultaneously to the data (i.e., F$_i^{NT}$-E$_{peak,i}^{NT}$ relation). In addition, this result points toward a universal relation between those two quantities for all GRBs when transposed to the central engine rest frame (i.e., L$_i^{NT}$-E$_{peak,i}^{rest,NT}$ relation).
Observations of GRB 100724B with the Fermi Gamma-Ray Burst Monitor (GBM) find that the spectrum is dominated by the typical Band functional form, which is usually taken to represent a non-thermal emission component, but also includes a statistically
highly significant thermal spectral contribution. The simultaneous observation of the thermal and non-thermal components allows us to confidently identify the two emission components. The fact that these seem to vary independently favors the idea that the thermal component is of photospheric origin while the dominant non-thermal emission occurs at larger radii. Our results imply either a very high efficiency for the non-thermal process, or a very small size of the region at the base of the flow, both quite challenging for the standard fireball model. These problems are resolved if the jet is initially highly magnetized and has a substantial Poynting flux.
From July 2008 to October 2009, the Gamma-ray Burst Monitor (GBM) on board the Fermi Gamma-ray Space Telescope (FGST) has detected 320 Gamma-Ray Bursts (GRBs). About 20% of these events are classified as short based on their T90 duration below 2 s. W
e present here for the first time time-resolved spectroscopy at timescales as short as 2 ms for the three brightest short GRBs observed with GBM. The time-integrated spectra of the events deviate from the Band function, indicating the existence of an additional spectral component, which can be fit by a power-law with index ~-1.5. The time-integrated Epeak values exceed 2 MeV for two of the bursts, and are well above the values observed in the brightest long GRBs. Their Epeak values and their low-energy power-law indices ({alpha}) confirm that short GRBs are harder than long ones. We find that short GRBs are very similar to long ones, but with light curves contracted in time and with harder spectra stretched towards higher energies. In our time-resolved spectroscopy analysis, we find that the Epeak values range from a few tens of keV up to more than 6 MeV. In general, the hardness evolutions during the bursts follows their flux/intensity variations, similar to long bursts. However, we do not always see the Epeak leading the light-curve rises, and we confirm the zero/short average light-curve spectral lag below 1 MeV, already established for short GRBs. We also find that the time-resolved low-energy power-law indices of the Band function mostly violate the limits imposed by the synchrotron models for both slow and fast electron cooling and may require additional emission processes to explain the data. Finally, we interpreted these observations in the context of the current existing models and emission mechanisms for the prompt emission of GRBs.
