Speaker
Description
The spectral energy distributions (SEDs) of blazars are dominated by synchrotron and inverse Compton radiation. The origin of blazar variability can be investigated from the time variation of the SEDs. However, it is difficult to estimate the optimal model parameters of SEDs and their uncertainties because some parameters are strongly correlated and the standard method may be trapped by local solutions. In this study, we applied a Markov chain Monte Carlo method (MCMC) to this problem. Our experiments using artificial data demonstrate that at least one prior probability is required to uniquely determine the solution. We used simultaneous observations of Mrk 421 with Fermi-LAT, Swift-XRT, and the 1.5-m optical telescope, Kanata from 2009 to 2011 reported in Itoh et al. (2015). We succeeded in estimating the optimal parameters and their uncertainties by using the prior probability of the time-scale and Doppler factor. As a result, we found that the break energy of electron energy distribution is proportional to X-ray flux and the normalization of electron distribution is inversely proportional. These correlations suggest that the X-ray variations were caused by the fluctuations of the break energy rather than the number of electrons.