In this thesis we present a time-dependent multi-zone radiative transfer code and its applications to study the multiwavelength emission of blazars. The multiwavelength variability of blazars is widely believed to be a direct manifestation of the formation and propagation of relativistic jets, and hence the related physics of the black hole - accretion disk - jet system. However, the understanding of these variability demands highly sophisticated theoretical analysis and numerical simulations. Especially, the inclusion of the light travel time effects(LTTEs) in these calculations has long been realized important, but very difficult. The code we use couples Fokker-Planck and Monte Carlo methods, in a 2 dimensional (cylindrical) geometry. For the first time all the LTTEs are fully considered, along with a proper, full, self-consistent treatment of Compton cooling, which depends on the LTTEs. Using this code, we studied a set of physical processes that are relevant to the variability of blazars, including electron injection and escape, radiative cooling, and stochastic particle acceleration. Our comparison of the observational data and the simulation results revealed that a combination of all those processes is needed to reproduce the observed behaviors of the emission of blue blazars. The simulation favors that the high energy emission at quiet and flare stages comes from the same location. We have further modeled red blazars PKS 1510-089. External radiation, which comes from the broad line region (BLR) or infrared torus, is included in the model. The results confirm that external Compton model can adequately describe the emission from red blazars. The emission from BLR is favored as the source of Inverse Compton seed photons, compared to synchrotron and IR torus radiation.

<p dir="ltr"> This article-based dissertation provides an account of two distinct classes of expansive astrophysical outflows and techniques to interpret their observations using numerical modeling. The primary purpose of this dissertation is to provide an extensive description of the research projects I undertook during my tenure as a Graduate Research Assistant, under the guidance of my advisor Prof. Paul Duffell.</p><p dir="ltr">Chapter 1 provides a brief introduction to numerical hydrodynamics and techniques of modeling expanding flows numerically. I also introduce the aforementioned classes of astrophysical outflows, namely relativistic jets from Active Galactic Nuclei (AGN), and supernova remnants (SNRs). I provide a general overview of the theoretical picture, and the general strategy used in this work to model them.</p><p dir="ltr">Chapter 2 describes my investigation on the connection of kiloparsec scale AGN jet properties to their intrinsic parameters and surroundings, based on an article published in The Astrophysical Journal. Using a suite of over 40 relativistic hydrodynamic jet models, we find that the dynamics of relativistic jets can be described in terms of only two parameters, the jet to ambient medium energy density ratio, and the jet opening angle. The former is found to strongly control the Fanaroff-Riley (FR) morphological dichotomy, which was previously thought to be tied to the magnitude of the jet luminosity. We also suggest a purely hydrodynamical origin of bright spots observed in some AGN jets. Our models were tested against and found to be consistent with the observations of the jets in M87 and Cygnus A.</p><p dir="ltr">In chapter 3, I present my moving-mesh hydrodynamics code Sprout, also described in an article published in The Astrophysical Journal Supplements. Sprout solves the equations of ideal hydrodynamics on an expanding Cartesian mesh. The expanding mesh can follow fluid outflows for several orders of magnitude with very little numerical diffusion. This allows Sprout to capture expanding flows with very high dynamic range. Sprout is thus particularly suitable for studying expanding outflows such as supernova remnants and active galactic nuclei. Relative to other moving mesh codes, the simple mesh structure in Sprout is also convenient for implementing additional physics or algorithms. I discuss many code tests that were performed to test the accuracy and performance of the numerical scheme.</p><p dir="ltr">Chapter 4 details my study of hydrodynamic instabilities in supernova remnants (SNRs) as they expand against the circumstellar medium (CSM). This is based on an article published in The Astrophysical Journal. A suite of 3D hydrodynamical SNR models, generated using my hydro code \sprout, was used to study the impact of the stellar ejecta density profile and seed anisotropies in the ejecta and the CSM on formation of turbulent structures in the SNRs. We found that most of the turbulent power in these models resides at a typical angular mode or scale that is determined by the ejecta density structure. It was also found that clumps or anisotropies in either the ejecta or CSM do not imprint upon these turbulence structures unless they are massive and form large-scale coherent structures.</p><p dir="ltr">In chapter 5, I discuss the implementation of a technique to measure anisotropies in observed SNRs just using 2D high-resolution images. This technique is calibrated using 3D hydro SNR models and synthetic images derived from them. As seen in Chapter 4, we find a similar dominant angular scale of turbulent structures dictated by the ejecta density structure. Both the 3D models and the synthetic images yield the same value of this scale, which validates the image analysis technique used in this work. As an example of how this technique can be applied to observations, we analyze observations of a known supernova remnant (Tycho's SNR) and compare with our models. Our technique picks out the angular scale of Tycho's fleece-like structures and also agrees with the small-scale power seen in Tycho.</p><p dir="ltr">PhChapter 6 summarizes the results, conclusions, and future prospects of all the research work described so far. It is followed by a bibliography, my curriculum vita, and a list of publications.</p>

Les noyaux actifs de galaxie (NAG) sont les objets les plus énergétiques de l'univers. Cette incroyable puissance provient de l'énergie gravitationnel de matière en rotation autour d'un trou noir super-massif siégeant au centre des galaxies. Environ 10% des NAG sont pourvus de jets relativistes émanant de l'objet central (trou noir et matière environnante) et s'étalant sur des échelles de l'ordre de la galaxie hôte. Ces jets sont observés à toutes les longueurs d'ondes, de la radio aux rayons gamma les plus énergétiques. En dépit de nombreuses études et d'instruments de plus en plus précis depuis leur découverte dans les années 1950, les NAG sont encore très mal compris et la formation, la composition et l'accélération des jets sont des questions encore pleinement ouvertes. Le modèle le plus répandu visant à reproduire l'émission des NAG, le modèle "une zone" repose souvent sur des hypothèse ad-hoc et ne parvient pas à apporter une modélisation satisfaisante.Le paradigme du "two-flow" (deux fluides) développé à l'IPAG et basé sur une idée originale de Sol et al (1989) a pour but de fournir une vision unifiée et cohérente des jets de NAG. Cette théorie repose sur une l'hypothèse principale que les jets seraient en fait composés de deux fluides co-axiaux: une colonne centrale composée d'un plasma purement leptonique (électrons/positrons) se déplaçant à des vitesses relativistes et responsable pour la grande partie de l'émission non thermique observée entourée par une enveloppe composée d'un plasma baryonique (électrons/protons), régie pas la magnéto-hydrodynamique, se déplaçant à des vitesses sous-relativistes mais transportant la majorité de l'énergie. Cette hypothèse est basée sur des indices observationnels ainsi que sur des arguments théoriques et permet d'expliquer nombre des caractéristiques des NAG.Afin d'étudier plus en profondeur le paradigme du two-flow, un modèle numérique basé sur ses concepts et produisants des observables comparables aux observations est nécessaire.Durant ma thèse, j'ai participé au développement de ce modèle, m'intéressant notamment à la diffusion Compton inverse de photons provenant de l'extérieur du jet. Ce processus, primordial dans la modélisation des NAG, est aussi central dans le paradigme du two-flow car il est à l'origine de l'accélération de la colonne via l'effet fusée Compton. Pour cela, j'ai du développer des nouvelles approximations analytiques de la diffusion Compton d'une distribution thermique de photons.En m'intéressant à l'effet fusée Compton, j'ai pu montré que dans le champ de photon thermique d'un NAG, le facteur de Lorentz d'ensemble du plasma pouvait être sujet à des variations le long du jet en fonction de la distance à l'objet central. Ces variations peuvent avoir un effet important sur l'émission observée et peuvent induire de la variabilité spatiale et temporelle. J'ai également montré que les facteurs de Lorentz terminaux obtenus étaient compatibles avec les conditions physiques attendus dans les jets et avec les observations.Le modèle complet produit des DES directement comparables aux observations. Néanmoins, le modèle est par nature erratique et il est quasiment impossible de relier directement les paramètres du modèles avec les DES produites. Malheureusement, les procédures standards d'adaptation automatique aux données (e.g. basé sur les méthodes de gradient) ne sont pas adaptées au modèle à cause de son grand nombre de paramètres, de sa non-linéarité et du temps de calcul important. Afin de palier à ce problème, j'ai développé une procédure d'adaptation automatique basée sur les algorithmes génétiques. L'utilisation de cet outil a permis la reproduction de plusieurs DES par le modèle. J'ai également montré que le modèle était capable de reproduire les DES observées avec des facteurs de Lorentz d'ensemble relativement bas, ce qui pourrait potentiellement apporter une harmonisation entre les observations et les nécessités théoriques. / Active galactic nuclei (AGN) are the most energetic objects known in the universe. Their fantastic energy is due to efficient conversion of gravitational energy of mass accreted on super-massive black-holes at the center of galaxy into luminous energy. 10% of AGN are even more incredible as they display relativistic jets on galaxy scales. Those jets are observed at all energies, from far radio to highest gamma-rays. Despite intense study since their discovery in the 50's and more and more observations, favored by rapid progress in instrumentation, AGN are still widely misunderstood. The questions of formation, composition, and acceleration of jets are central but still a matter of debates. Models aiming at reproducing observed emission have been developed throughout the years. The most common one, the one-zone model, often relies on ad hoc hypothesis and does not provide a satisfactory answer.The two-flow paradigm developed at IPAG and based on an original idea from Sol et al (1989) aims at giving a more coherent and physical representation of AGN jets. The principal assumption is that jets are actually composed of two coaxial flows: an inner spine made of a pure pair plasma, moving at relativistic speed and responsible for the non-thermal observed emission surrounded by an external sheath, made of a baryonic MHD plasma, midly relativistic but carrying most of the power. The two-flow paradigm finds roots in observations as well as theoretical arguments and has been able to explain many AGN features.During my PhD, I studied this paradigm and contributed to the development of a numerical model based on its concepts. I have been particularly interested in the inverse Compton scattering of thermal photons, fundamental process in the modeling of AGN emission, as well as the Compton rocket effect, key to the acceleration of the spine in the two-flow paradigm.However, taking into account the inverse Compton emission, with the complete cross-section (including the Klein-Nishina regime) and the anisotropy can be very time consuming. To accomplish fast and efficient computation of the external Compton emission, I have had to formulate new analytical approximations of the scattering of a thermal distribution of photons.I have also studied the Compton rocket effect, responsible for the acceleration of the inner spine in the two-flow paradigm. I showed that the resulting bulk Lorentz factor of the flow in the complex photon field of an AGN is subject to variations along the jet as a function of the distance to the central engine. These variations can have drastic effects on the observed emission and could induce variability, both spatially and temporally.I also showed that the terminal bulk Lorentz factor obtained are compatible with physical conditions expected in jets and with observations.The complete model produce spectral energy distribution (SED) comparable to observed ones. However, the model is by nature erratic and it is difficult to make a direct link between the model parameters (input) and the SED (output). Unfortunately, standard data fitting procedures (e.g. based on gradient methods) are not adapted to the model due to its important number of parameters, its important computing time and its non-linearity. In order to circumvent this issue, I have developed a fitting tool based on genetic algorithms. The application of this algorithm allowed me to successfully fit several SED. In particular, I have also showed that the model, because based on a structured jet model, can reproduce observations with low bulk Lorentz factor, thus giving hope to match observations and theoretical requirements in this matter.