Speaker
Description
The radiation from astrophysical sources such as blazars originates mainly from various particle interactions in the jet or in other regions of the blazar. The spectral energy distribution of a blazar shows a two-bump structure. While the low-energy emission is explained by synchrotron radiation of highly energetic electrons, the high-energy part can contain contributions from hadrons too. Therefore, the radiation of a blazar can be described by leptonic, hadronic or lepto-hadronic models. Depending on the model there are various values describing properties of both the radiation source itself and the particles that produce the radiation. These parameters are used for simulating the radiation by solving the particle transport equation numerically, so that the resulting blazar spectra can be obtained. The modeled spectrum can then be compared with the observed data from the blazar. With large parameter scans, each parameter can be varied so that many spectra are obtained in the end. Several different parameter sets may describe the radiation well but lead to different physical properties of the source.
In this project, we are investigating leptonic models and focus on the parameter space overall. We compare the models that describe the data equally well. If there are large differences between the parameter sets, this could mean the properties of one and the same radiation source also differ greatly depending on the model. Furthermore, we are interested in how to optimize the modeling with these findings. Using the blazar PKS 0735+178 as an example, we visualize the parameter sets resulting from the modeling of the radiation. We apply machine learning methods for mapping the parameter space and visualization of multidimensional data. We present the maps of the parameter space and discuss the physical implications of the obtained results.
