L’objectif de ma thèse porte sur les interactions entre les noyaux actifs de galaxies et le milieu interstellaire des galaxies. En particulier, je mets l’accent sur les deux mécanismes possibles responsables de la production des vents par les trous noirs : les jets et les vents produits par le rayonnement de ces trous noirs. Les simulations hydrodynamiques de haute résolution des galaxies comprenant la rétroac- tion d’un jet ont montré que l’activité des noyaux actifs peut conduire à une pression exces- sive sur les régions denses de formation stellaire dans les galaxies, et donc à augmenter la formation d’étoiles, conduisant à un effet positif de rétroaction. Je montre que ces noyaux actifs induits par pression régulée et formation d’étoiles peuvent aussi être une explica- tion possible des taux de formation stellaire élevés observés dans l’Univers à haut décalage spectral. De plus, j’ai également étudié en détails comment le rayonnement émis à partir d’un disque d’accrétion autour du trou noir agit efficacement avec le milieu interstellaire et entraîne un fort vent galactique, en simulant la propagation des photons à partir des équations hydrodynamiques du rayonnement. Les simulations montrent que la grande luminosité d’un quasar est en effet capable de conduire des vents à grande échelle et à grande vitesse. Le rayonnement infrarouge est nécessaire pour transérer efficacement le gaz par multi-diffusion sur la poussière dans les nuages denses. Le nombre typique de multi-diffusion diminue rapidement quand le nuage central de gaz central se dilate et se rompt, ce qui permet au rayonnement de s’échapper à travers les canaux à faible densité. / Supermassive black holes (SMBHs) are known to reside in the centres of most large galaxies. The masses of these SMBHs are known to correlate with large-scale properties of the host galaxy suggesting that the growth of the BHs and large-scale structures are tightly linked. A natural explanation for the observed correlation is to invoke a self-regulated mechanism involving feedback from Active Galactic Nuclei (AGN). The focus of this thesis is on the interactions between AGN outflows and the ISM and how the feedback impacts the host galaxy. In particular, it focuses on the two possible mechanism of outflows, namely, outflows related to AGN jets and outflows produced by AGN radiation. High resolution, galaxy scale hydrodynamical simulations of jet-driven feedback have shown that AGN activity can over-pressurise dense star-formation regions of galaxies and thus enhance star formation, leading to a positive feedback effect. I propose, that such AGN-induced pressure-regulated star formation may also be a possible explanation of the high star formation rates recently found in the high-redshift Universe. In order to study in more detail the effects of over-pressurisation of the galaxy, I have performed a large set of isolated disc simulations with varying gas-richness in the galaxy. I found that even moderate levels of over-pressurisation of the galaxy boosts the global star formation rate by an order of magnitude. Additionally, stable discs turn unstable which leads to significant fragmentation of the gas content of the galaxy, similar to what is observed in high-redshift galaxies. The observed increase in the star formation rate of the galaxy is in line with theoretical predictions. I have also studied in detail how radiation emitted from a thin accretion disc surrounding the BH effectively couples to the surrounding ISM and drives a large scale wind. Quasar activity is typically triggered by extreme episodes of gas accretion onto the SMBH, in particular in high-redshift galaxies. The photons emitted by a quasar eventually couple to the gas and drive large scale winds. In most hydrodynamical simulations, quasar feedback is approximated as a local thermal energy deposit within a few resolution elements, where the efficiency of the coupling between radiation of the gas is represented by a single parameter tuned to match global observations. In reality, this parameter conceals various physical processes that are not yet fully un- derstood as they rely on a number of assumptions about, for instance, the absorption of photons, mean free paths, optical depths, and shielding. To study the coupling between the photons and the gas I simulated the photon propagation using radiation-hydrodynamical equations (RHD), which describe the emission, absorption and propagation of photons with the gas and dust. Such an approach is critical for a better understanding of the coupling between the radiation and gas and how hydrodynamical sub-grid models can be improved in light of these results...

One of the key goals of Milky Way science is measuring the distribution of dark matter in the Galaxy. Through the study of Galactic dynamics, inferences can be made about the structure of the Galaxy, and hence the dark matter distribution. To this end, we present a study of methods useful for modelling and understanding dynamical systems in the Galaxy. A natural choice of coordinate system when studying dynamical systems is the canonical system of angle-action coordinates. We present methods for estimating the angle-actions in both axisymmetric and triaxial potentials. These fall into two categories: non-convergent and convergent. The non-convergent methods are fast approaches, mostly based on approximations to Stäckel potentials. We investigate the accuracy of these methods for realistic Galactic potentials. The slower convergent methods operate by constructing generating functions to take us from simple analytically-tractable potentials to our target potential. Tidal streams should prove useful for constraining the large-scale dark matter distribution in the Galaxy. Armed with our new angle-action tools, we investigate the properties of known streams in a realistic Galactic potential. We present a simple algorithm for constraining the Galactic potential using a tidal stream, which exploits the expected structure of a stream in the angle-frequency space of the true potential. We expand this approach into a fully probabilistic scheme that allows for handling of large errors, missing data and outliers. We close by discussing another tool useful for modelling the dynamics of the Galaxy: extended distribution functions for the Galactic disc. We present a simple extension of an action-based distribution function from Binney (2010) that includes metallicity information, and compare the model predictions with current data. These models are essential for incorporating the selection effects of any survey, and reveal the important chemo-dynamic correlations that expose the history and evolution of the Galaxy.

Hot Jupiters receive strong stellar irradiation, producing equilibrium temperatures of 1000-2500 K. Incoming irradiation directly heats just their thin outer layer, down to pressures of similar to 0.1 bars. In standard irradiated evolution models of hot Jupiters, predicted transit radii are too small. Previous studies have shown that deeper heating-at a small fraction of the heating rate from irradiation-can explain observed radii. Here we present a suite of evolution models for HD 209458b, where we systematically vary both the depth and intensity of internal heating, without specifying the uncertain heating mechanism(s). Our models start with a hot, high-entropy planet whose radius decreases as the convective interior cools. The applied heating suppresses this cooling. We find that very shallow heating-at pressures of 1-10 bars-does not significantly suppress cooling, unless the total heating rate is greater than or similar to 10% of the incident stellar power. Deeper heating, at 100 bars, requires heating at only 1% of the stellar irradiation to explain the observed transit radius of 1.4R(Jup) after 5 Gyr of cooling. In general, more intense and deeper heating results in larger hot-Jupiter radii. Surprisingly, we find that heat deposited at 10(4) bars-which is exterior to approximate to 99% of the planet's mass-suppresses planetary cooling as effectively as heating at the center. In summary, we find that relatively shallow heating is required to explain the radii of most hot Jupiters, provided that this heat is applied early and persists throughout their evolution.

In this work, the efficiency of the LUMBA UVES pipeline for processing of spectroscopic observations is tested by use on 23 high-resolution spectra of the open star cluster M67. An abundance trend discovered by Gavel et al. (2019) concerning iron abundances of giant and sub-giant stars of that cluster is examined. An initial run for a set of ”Gaia FGK benchmark stars”, as described in Blanco-Cuaresma et al. (2014) and Heiter et al. (2015) helps inspect the structure, method and output of the pipeline. Through Python language programming, processes are greatly automatized and the pipeline is run for a total of 460 weak and strong iron lines of our 23-star sample. The line fitting and efficiency of the pipeline is appreciated by statistically analyzing the results and looking into individual discrepant ones. The abundance trend is reproduced while using FeI lines, unlike runs using FeII lines. Trends in abundance over line strength plots also hint at bias through the Gaia-Eso Survey (GES) microturbulence relation. Using internal Data Release 6 (iDR6) and LUMBA-derived starting parameters, log(g) - Teff plots of our sample stars agree with a previously established 4.3 Gyr cluster age. An alternate run is performed for those stars, using LUMBA-derived starting parameters. The choice of starting parameters does impact abundance derivation, but is not the primary source of persistent systematic discrepancies.

Scientists' ability to generate and store simulation results is outpacing their ability to analyze them via ad hoc programs. We observe that these programs exhibit an algebraic structure that can be used to facilitate reasoning and improve performance. In this dissertation, we present a formal data model that exposes this algebraic structure, then implement the model, evaluate it, and use it to express, optimize, and reason about data transformations in a variety of scientific domains. Simulation results are defined over a logical grid structure that allows a continuous domain to be represented discretely in the computer. Existing approaches for manipulating these gridded datasets are incomplete. The performance of SQL queries that manipulate large numeric datasets is not competitive with that of specialized tools, and the up-front effort required to deploy a relational database makes them unpopular for dynamic scientific applications. Tools for processing multidimensional arrays can only capture regular, rectilinear grids. Visualization libraries accommodate arbitrary grids, but no algebra has been developed to simplify their use and afford optimization. Further, these libraries are data dependent—physical changes to data characteristics break user programs. We adopt the grid as a first-class citizen, separating topology from geometry and separating structure from data. Our model is agnostic with respect to dimension, uniformly capturing, for example, particle trajectories (1-D), sea-surface temperatures (2-D), and blood flow in the heart (3-D). Equipped with data, a grid becomes a gridfield. We provide operators for constructing, transforming, and aggregating gridfields that admit algebraic laws useful for optimization. We implement the model by analyzing several candidate data structures and incorporating their best features. We then show how to deploy gridfields in practice by injecting the model as middleware between heterogeneous, ad hoc file formats and a popular visualization library. In this dissertation, we define, develop, implement, evaluate and deploy a model of gridded datasets that accommodates a variety of complex grid structures and a variety of complex data products. We evaluate the applicability and performance of the model using datasets from oceanography, seismology, and medicine and conclude that our model-driven approach offers significant advantages over the status quo.

Thesis (MEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: In this thesis we set out to design an algorithm for solving the all-pairs fixed-radius nearest neighbours search problem for a massively parallel heterogeneous system. The all-pairs search problem is stated as follows: Given a set of N points in d-dimensional space, find all pairs of points within a horizon distance of one another. This search is required by any nonlocal or meshless numerical modelling method to construct the neighbour list of each mesh point in the problem domain. Therefore, this work is applicable to a wide variety of fields, ranging from molecular dynamics to pattern recognition and geographical information systems. Here we focus on nonlocal solid mechanics methods. The basic method of solving the all-pairs search is to calculate, for each mesh point, the distance to each other mesh point and compare with the horizon value to determine if the points are neighbours. This can be a very computationally intensive procedure, especially if the neighbourhood needs to be updated at every time step to account for changes in material configuration. The problem also becomes more complex if the analysis is done in parallel. Furthermore, GPU computing has become very popular in the last decade. Most of the fastest supercomputers in the world today employ GPU processors as accelerators to CPU processors. It is also believed that the next-generation exascale supercomputers will be heterogeneous. Therefore the focus is on how to develop a neighbour searching algorithm that will take advantage of next-generation hardware. In this thesis we propose a CPU - multi GPU algorithm, which is an extension of the fixed-grid method, for the fixed-radius nearest neighbours search on massively parallel systems. / AFRIKAANSE OPSOMMING: In hierdie tesis het ons die ontwerp van ’n algoritme vir die oplossing van die alle-pare vaste-radius naaste bure soektog probleem vir groot skaal parallele heterogene stelsels aangepak. Die alle-pare soektog probleem is as volg gestel: Gegewe ’n stel van N punte in d-dimensionele ruimte, vind al die pare van punte wat binne ’n horison afstand van mekaar af is. Die soektog word deur enige nie-lokale of roosterlose numeriese metode benodig om die bure-lys van alle rooster-punte in die probleem te kry. Daarom is hierdie werk van toepassing op ’n wye verskeidenheid van velde, wat wissel van molekulêre dinamika tot patroon herkenning en geografiese inligtingstelsels. Hier is ons fokus op nie-lokale soliede meganika metodes. Die basiese metode vir die oplossing van die alle-pare soektog is om vir elke rooster-punt, die afstand na elke ander rooster-punt te bereken en te vergelyk met die horison lente, om dus so te bepaal of die punte bure is. Dit kan ’n baie berekenings intensiewe proses wees, veral as die probleem by elke stap opgedateer moet word om die veranderinge in die materiaal konfigurasie daar te stel. Die probleem word ook baie meer kompleks as die analise in parallel gedoen word. Verder het GVE’s (Grafiese verwerkings eenhede) baie gewild geword in die afgelope dekade. Die meeste van die vinnigste superrekenaars in die wêreld vandag gebruik GVE’s as versnellers te same met SVE’s (Sentrale verwerkings eenhede). Dit is ook van mening dat die volgende generasie exa-skaal superrekenaars GVE’s sal implementeer. Daarom is die fokus op hoe om ’n bure-lys soektog algoritme te ontwikkel wat gebruik sal maak van die volgende generasie hardeware. In hierdie tesis stel ons ’n SVE - veelvoudige GVE algoritme voor, wat ’n verlenging van die vaste-rooster metode is, vir die vaste-radius naaste bure soektog op groot skaal parallele stelsels.

J'ai effectué une étude des éléments C, N, et O dans les atmosphères d'étoiles chimiquement particulières de la séquence principale, où ils sont sous-abondants et répartis de manière inhomogène. J'ai recensé les processus physiques qu'il faut inclure dans les calculs d'accélération radiative et montré leur importance relative. Des améliorations majeures par rapport aux approximations usuelles ont été obtenues grâce à l'utilisation systématique des données atomiques du projet OPACITY. Elles permettent de calculer précisément la dépendance en fréquence des opacités, et d'améliorer l'évaluation des largeurs de raies. Les contributions des raies et de la photoionisation sont calculées pour chaque ion et l'accélération totale sur un élément donné est obtenue grâce à un modèle prenant en compte les réactions rapides entre ions. Les accélérations radiatives calculées pour le carbone, l'azote, et l'oxygène, les poussant vers le haut, apparaissent inférieures à la gravité dans tous les modèles d'enveloppe considérés (étoiles de type A à F), pour une large gamme de paramètres, expliquant leurs déficits marquées à la surface de la plupart des étoiles chimiquement particulières. Des tables donnant d'une part l'opacité de ces éléments, d'autre part leur accélération radiative sur une grille contenant de nombreuses conditions de plasma permettent d'effectuer des calculs d'évolution stellaire prenant en compte tous les aspects de la diffusion des éléments C, N, et O, les plus abondants après H et He. <br /> <br />Je me suis ensuite intéressé aux étoiles de type lambda Bootis, un petit sous-groupe singulier d'étoiles chimiquement particulières, dont les anomalies d'abondance ne sont pas expliquées par le modèle de la diffusion radiative. Il s'agirait plutôt d'étoiles jeunes, encore entourées des restes du disque à partir duquel elles se sont formées, et dont elles accrèteraient un gaz appauvri en éléments lourds, ceux-ci s'étant condensés en grains. Afin de vérifier cette hypothèse, nous avons recherché la signature de matière circumstellaire dans le spectre de ces étoiles. Peu d'étoiles de notre échantillon montrent un tel indice et nos résultats suggèrent une anti-corrélation entre la présence de gaz ou de poussières, pouvant caractériser deux états différents dans l'évolution du disque protostellaire. Au cours de cette étude, nous avons découvert par hasard le premier cas de pulsations non radiales dans une étoile de type lambda Bootis, puis montré qu'elles sont communes dans ce groupe. L'identification des modes de pulsation permet de remonter à la structure interne de ces étoiles et à leur état d'évolution, permettant ainsi de tester le modèle d'accrétion. <br /> <br />Mon étude des environnements circumstellaires des étoiles lambda Bootis m'a conduit à m'intéresser aux disques protoplanétaires. Jusqu'à récemment, nous n'avions observé qu'un seul système solaire (le nôtre) dans lequel nous pouvions tester notre compréhension du processus de formation de planètes. Maintenant, plus d'une centaine de planètes ont été découvertes autour d'autres étoiles et les contraintes sur les modèles théoriques sont devenues très serrées. Nous savons que, dans la nébuleuse solaire, les particules de poussière de la taille du micron se sont agglomérées pour former des planètes, objets 10^13 à 10^14 fois plus grands. Bien qu'il y ait beaucoup de travail réalisé sur les dernières étapes de cette formation, et sur la migration de planètes déjà formées, peu de travail a été fait pour développer des modèles hydrodynamiques décrivant l'interaction du gaz et de la poussière dans les disques proto-planétaires. Nous développons un code hydrodynamique SPH permettant de modéliser cette interaction, principalement par la force de friction, entre deux phases: du gaz et des grains de poussière d'une taille donnée. Nous obtenons ainsi la répartition spatiale des grains dans le disque en fonction de leur taille. Ce travail correspond à la thèse de Laure Barrière-Fouchet, qui se termine en 2005. Nous projetons ensuite d'ajouter les mécanismes de coagulation, croissance, et évaporation des grains de poussière en modélisant plusieurs phases pour différentes tailles de grains et la variation du nombre de particules dans chaque phase qui en résulte. Ceci permettra de caractériser les zones du disque les plus favorables à la formation de planétésimaux. Ensuite, il s'agira d'explorer plus profondément les mécanismes de formation de planètes. En effet, si l'on arrive assez bien à faire croître les grains microscopiques jusqu'à une taille de l'ordre du centimètre, les collisions entre ces gros grains les refragmentent et empêchent de dépasser cette taille. Plusieurs solutions sont à envisager pour permettre de passer ce cap: diminution des vitesses de collisions dans les régions plus denses, rôle de la turbulence, etc... <br /> <br />Un peu à part de mes travaux précédents, avec mes collègues de l'ESO, j'ai observé et pris le premier spectre de la contrepartie optique du sursaut gamma GRB980425, qui s'est avéré être une supernova très particulière: SN1998bw. Son spectre en évolution rapide ne permettait pas de classer cette supernova, la première à être associée à un sursaut gamma, dans les types connus. Notre équipe a suivi régulièrement l'évolution de sa courbe de lumière et de son spectre, la somme de données recueillie ayant conduit à un modèle d'hypernova.

Extrasolar planets on eccentric short-period orbits provide a laboratory in which to study radiative and tidal interactions between a planet and its host star under extreme forcing conditions. Studying such systems probes how the planet's atmosphere redistributes the time-varying heat flux from its host and how the host star responds to transient tidal distortion. Here, we report the insights into the planet-star interactions in HAT-P-2's eccentric planetary system gained from the analysis of similar to 350 hr of 4.5 mu m observations with the Spitzer Space Telescope. The observations show no sign of orbit-to-orbit variability nor of orbital evolution of the eccentric planetary companion, HAT-P-2b. The extensive coverage allows us to better differentiate instrumental systematics from the transient heating of HAT-P-2b's 4.5 mu m photosphere and yields the detection of stellar pulsations with an amplitude of approximately 40 ppm. These pulsation modes correspond to exact harmonics of the planet's orbital frequency, indicative of a tidal origin. Transient tidal effects can excite pulsation modes in the envelope of a star, but, to date, such pulsations had only been detected in highly eccentric stellar binaries. Current stellar models are unable to reproduce HAT-P-2's pulsations, suggesting that our understanding of the interactions at play in this system is incomplete.

Cosmological N-body simulations indicate that the dark matter haloes of galaxies should be generally triaxial. Yet, the presence of a baryonic disc is believed to modify the shape of the haloes. The goal of this thesis is to study how bar formation is affected by halo triaxiality and how, in turn, the presence of the bar influences the shape of the halo. We performed a series of collisionless and hydrodynamical numerical simulations, using elliptical discs as initial conditions. Triaxial halos tend to become more spherical and we show that part of the circularisation of the halo is due to disc growth, but part must be attributed to the formation of a bar. We find that the presence of gas in the disc is a more efficient factor than halo triaxiality in inhibiting the formation of a strong bar. / As simulações cosmológicas de N-corpos indicam que os halos de matéria escura das galáxias devem ser em geral triaxiais. Contudo, acredita-se que a presença de um disco bariônico seja capaz de alterar a forma do halo. O objetivo desta tese é o de estudar como a formação de barras é afetada pela triaxialidade do halo e como, por sua vez, a presença da barra influencia a forma do halo. Nós realizamos uma série de simulações numéricas acolisionais e hidrodinâmicas, utilizando discos elípticos como condições iniciais. Os halos triaxiais tendem a se tornar mais esféricos e nós mostramos que parte da circularização do halo é devida ao crescimento do disco, mas parte precisa ser atribuída à formação da barra. Notamos que a presença de gás no disco é um fator mais eficiente do que a triaxialidade do halo em inibir a formação de uma barra forte.

Au delà du système solaire et de ses planètes, nous avons maintenant un catalogue de quasiment 1000 exoplanètes qui illustrent la grande diversité des planètes et des systèmes qu'il est possible de former. Cette diversité est un défi que les modèles de formation planétaire tentent de relever. La migration de type 1 est un des mécanismes pour y parvenir. En fonction des propriétés du disque protoplanétaire, les planètes peuvent s'approcher ou s'éloigner de leur étoile. La grande variété des modèles de disques protoplanétaires permet d'obtenir une grande variété de systèmes planétaires, en accord avec la grande diversité que nous observons déjà pour l'échantillon limité qui nous est accessible. Grâce à des simulations numériques, j'ai pu montrer qu'au sein d'un même disque, il est possible de former des super-Terres ou des noyaux de planètes géantes selon l'histoire de migration d'une population d'embryons. / In addition to the Solar System and its planets, we now have a database of nearly 1000 planets that emphasize the huge diversity of planets and systems that can be formed. This diversity is a challenge for planetary formation models. Type I migration is one of the mechanisms possible to explain this diversity. Depending on disc properties, planets can migrate inward or outward with respect to their host star. The huge parameter space of protoplanetary disc models can form a huge diversity of planetary systems, in agreement with the diversity observed in the nonetheless small sample accessible to us. Thanks to numerical simulations, I showed that within the same disc, it is possible to form super-Earths or giant planet cores, depending on the migration history of an initial population of embryos.