In this study, the light curves and spectrum of the photospheric thermal radiation from ultrarelativistic gamma-ray burst (GRB) jets are calculated using 2D relativistic hydrodynamic simulations of jets from a collapsar. As the jet advances, the density around the head of the jet decreases, and its Lorentz factor reaches as high as 200 at the photosphere and 400 inside the photosphere. For an on-axis observer, the photosphere appears concave shaped due to the low density and high beaming factor of the jet. The luminosity varies because of the abrupt change in the position of the photosphere due to the internal structure of the jet. Comparing our results with GRB090902B, the flux level of the thermal-like component is similar to our model, although the peak energy looks a little bit higher (but still within a factor of 2). From the comparison, we estimate that the bulk Lorentz factor of GRB090902B is $\Gamma \sim 2.4 \times 10^2 (r/10^{12} \rm cm$) where $r$ is the radius of the photosphere. The spectrum for an on-axis observer is harder than that for an off-axis observer. There is a time lag of a few seconds for high energy bands in the light curve. This may be the reason for the delayed onset of GeV emission seen in GRB080916C. The spectrum below the peak energy is a power law and the index is $2.3 \sim 2.6$ which is softer than that of single temperature plank distribution but still harder than that of typical value of observed one.

We revisit the high-energy spectral cutoff originating from the electron-positron pair creation in the prompt phase of gamma-ray bursts (GRBs) with numerical and analytical calculations. We show that the conventional exponential and/or broken power law cutoff should be drastically modified to a shallower broken power-law in practical observations that integrate emissions from different internal shocks. Since the steepening is tiny for observations, this "smearing" effect can generally reduce the previous estimates of the Lorentz factor of the GRB outflows. We apply our formulation to GRB 080916C, recently detected by the Large Area Telescope detector on the Fermi satellite, and find that the minimum Lorentz factor can be ~600 (or even smaller values), which is below but consistent with the previous result of ~900. Observing the steepening energy (so-called "pair-break energy") is crucial to diagnose the Lorentz factor and/or the emission site in the future observations, especially current and future Cherenkov telescopes such as MAGIC, VERITAS, and CTA.

The spatial distribution of short Gamma-ray bursts (GRBs) in their host galaxies provide us an opportunity to investigate their origins. Based on the currently observed distribution of short GRBs relative to their host galaxies, we obtain the fraction of the component that traces the mergers of binary compact objects and the one that traces star formation rate (such as massive stars) in early- and late-type host galaxies. We find that the fraction of massive star component is 0.37\pm 0.13 from the analysis of projected offset distribution. This suggests that a good fraction of short GRBs still originate from merger events. From our analysis, we also conclude that the fraction of late-type hosts among the elliptical, starburst and spiral galaxy is 0.82\pm 0.05, which is consistent with the observed early- to late-type number ratio of host galaxies.

We study the survival of ultrahigh energy nuclei injected in clusters of galaxies, as well as their secondary neutrino and photon emissions, using a complete numerical propagation method and a realistic modeling of the magnetic, baryonic and photonic backgrounds. It is found that the survival of heavy nuclei highly depends on the injection position and on the profile of the magnetic field. Taking into account the limited lifetime of the central source could also lead in some cases to the detection of a cosmic ray afterglow, temporally decorrelated from neutrino and gamma ray emissions. We calculate that the diffusive neutrino flux around 1 PeV coming from clusters of galaxies may have a chance to be detected by current instruments. The observation of single sources in neutrinos and in gamma rays produced by ultrahigh energy cosmic rays will be more difficult. Signals coming from lower energy cosmic rays (E < 1 PeV), if they exist, might however be detected by Fermi, for reasonable sets of parameters.

The shock-acceleration theory predicts a power-law energy spectrum in the test particle approximation, and there are two ways to calculate a power-law index, Peacock's approximation and Vietri's formulation. In Peacock's approximation, it is assumed that particles cross a shock front many times and energy-gains for each step are fully uncorrelated. On the other hand, correlation of the distribution of an energy-gain factor for a particle is considered in Vietri's formulation. We examine how Peacock's approximation differs from Vietri's formulation. It is useful to know when we can use Peacock's approximation because Peacock's approximation is simple to derive the power-law index. In addition, we focus on how the variance of the energy-gain factor has an influence on the difference between Vietri's formulation and Peacock's approximation. The effect of the variance has not been examined well until now. For demonstration, we consider two cases for the scattering in the upstream: the large-angle scattering (model A) and the regular deflection by large-scale magnetic fields (model B). Especially there is no correlation among the distribution of an energy-gain factor for every step in model A. In this model, we see the power-law index derived from Peacock's approximation differs from the one derived from Vietri's formulation when we consider the mildly-relativistic shock, and the variance of the energy-gain factor affects this difference. We can use Peacock's approximation for a non-relativistic shock and a highly-relativistic shock because the effect of the variance is hidden. In model B, we see the difference of the power-law converging along the shock velocity.