The diversity in observed planets and planetary systems has raised the question of whether they can be explained by a single model of planet formation or whether multiple models are required. The work presented in this thesis aims to examine the oligarchic growth scenario, to determine whether the core accretion model, where planets form bottom-up, can recreate the observed diversity. I begin by exploring how changing model parameters such as disc mass and metallicity influence the types of planetary systems that emerge. I show that rapid inward migration leads to very few planets with masses mp > 10M⊕ surviving, with surviving planetary systems typically containing numerous low-mass planets. I examine what conditions are required for giant planets to form and survive migration, finding that for a planet similar to Jupiter to form and survive, it must form at an orbital radius rp > 10 au. In the second project in this thesis, I update the physical models before examining whether a broader range of parameters can produce planetary systems similar to those observed. I find that compact systems of low-mass planets form in simulations if there is sufficient solid material in the disc or if planetesimals are small, thus having increased mobility. I also find that giant planets can form when the solid abundance and mobility of planetesimals are high, however they all undergo largescale migration into the magnetospheric cavity located close to the star. For the final project of this thesis, I examined the effects that disc radial structuring has on the formation of giant planets. I find that by including radial structures, numerous giant planets are able to form at large orbital radii and survive migration. The observed period valley between 10-100 days is also recreated, of which I attribute to disc dispersal late in the disc's lifetime.

<p>Planets are believed to form in primordial gas-dust discs surrounding newborn stars. An important breakthrough in our understanding of planetary formation was the discovery of extra-solar planets around sun-like stars, especially the frequent occurrence of giant planets on close orbits (hot Jupiters). The mechanisms involved in the formation of these objects remain uncertain, however the difficulties associated with their formation at their observed orbital radius has awoken an interest in theories for the migration of protoplanetary cores due to gravitational interaction with the disc. There are three fundamental regimes of planet migration. The type I and II migration regimes, driven by the differential Lindblad torques, result mostly in inward migration and concern low- and high-mass planets respectively. Type III migration, driven by the co-orbital gas flow, concerns an intermediate range of planetary masses and does not have a predefined direction.</p><p>In this thesis the orbital evolution of a high-mass, rapidly (type III) migrating planet is investigated using numerical hydrodynamical simulations. For these simulations we used the state-of-the-art hydrodynamics code FLASH. We focus on the physical aspects of type III migration. However, the problem of rapid migration of such massive planets is numerically challenging, and the disc model has to be chosen carefully, using numerical convergence as a discriminator between models (Paper I). We simulate both inward and outward directed migration (Papers II and III) and provide an extensive description of the co-orbital flow responsible for driving the migration, as well as its time evolution. The migration rate due to type III migration is found to be related to the mass of the planet's co-orbital region, making inward and outward directed migration self-decelerating and self-accelerating processes respectively (for a standard disc model). Rapid migration depends strongly on the flow structure in the planet's vicinity, which makes it sensitive to the amount of mass accumulated by the planet as it moves through the disc. This quantity in turn depends on the structure of the accretion region around the planet. The results of the numerical simulations show a good agreement with the analytical formulation of type III migration (Paper IV).</p>

Planets are believed to form in primordial gas-dust discs surrounding newborn stars. An important breakthrough in our understanding of planetary formation was the discovery of extra-solar planets around sun-like stars, especially the frequent occurrence of giant planets on close orbits (hot Jupiters). The mechanisms involved in the formation of these objects remain uncertain, however the difficulties associated with their formation at their observed orbital radius has awoken an interest in theories for the migration of protoplanetary cores due to gravitational interaction with the disc. There are three fundamental regimes of planet migration. The type I and II migration regimes, driven by the differential Lindblad torques, result mostly in inward migration and concern low- and high-mass planets respectively. Type III migration, driven by the co-orbital gas flow, concerns an intermediate range of planetary masses and does not have a predefined direction. In this thesis the orbital evolution of a high-mass, rapidly (type III) migrating planet is investigated using numerical hydrodynamical simulations. For these simulations we used the state-of-the-art hydrodynamics code FLASH. We focus on the physical aspects of type III migration. However, the problem of rapid migration of such massive planets is numerically challenging, and the disc model has to be chosen carefully, using numerical convergence as a discriminator between models (Paper I). We simulate both inward and outward directed migration (Papers II and III) and provide an extensive description of the co-orbital flow responsible for driving the migration, as well as its time evolution. The migration rate due to type III migration is found to be related to the mass of the planet's co-orbital region, making inward and outward directed migration self-decelerating and self-accelerating processes respectively (for a standard disc model). Rapid migration depends strongly on the flow structure in the planet's vicinity, which makes it sensitive to the amount of mass accumulated by the planet as it moves through the disc. This quantity in turn depends on the structure of the accretion region around the planet. The results of the numerical simulations show a good agreement with the analytical formulation of type III migration (Paper IV).

Made available in DSpace on 2016-03-15T19:35:53Z (GMT). No. of bitstreams: 1 DIRCEU YURI SIMPLICIO NETTO.pdf: 5745714 bytes, checksum: 3210bc6ef066c76d9c2b76320ed3b1d8 (MD5) Previous issue date: 2015-06-25 / Fundo Mackenzie de Pesquisa / Currently it is possible to estimate the rotation profile of a star that hosts a planet in an orbit such that it eclipses the star periodically. During one of these transits, the planet may occult a spot on the photosphere of the star, causing small variations in its light curve. By detecting the same spot in a later transit, it is possible to estimate the stellar rotation period. A total of more than 1200 planets, from almost 1900 detected, are known to eclipse their host star. Some stars with planets that were detected by the Kepler and CoRoT satellites have been analyzed and their rotation periods and differential rotation determined. Kepler-63 is a sun-like star, very young. It has a planet in a polar orbit, transiting the host star at many different latitudes. The results show that Kepler-63 has differential rotation of 0.133 (rd/d) and a relative differential rotation of 11.4%. Continuous monitoring of the starspots also show that the star probably has an activity cycle, with high spot concetration at the pole. / O perfil de rotação de uma estrela pode se estimado caso esta possua um planeta em órbita que a eclipse periodicamente. Durante um destes trânsitos, o planeta pode ocultar uma mancha na fotosfera da estrela, causando pequenas variações na curva de luz da mesma. Realizando o monitoramento das posições destas manchas em trânsitos posteriores, é possível estimar o período de rotação de uma estrela. Atualmente são confirmados um total de mais de 1900 planetas, onde um pouco mais de 1200 planetas eclipsam a sua estrela hospedeira. Algumas estrelas com planetas detectadas pelos satélites CoRoT e Kepler já foram analisadas por este método e seus perfis de rotação determinados. Kepler- 63 é uma estrela, do tipo solar, muito jovem. Possui um planeta que apresenta órbita polar, transitando a estrela em várias latitudes. Os resultados mostram que Kepler-63apresentarotaçãodiferencialde 0, 133 (rd/d) e um valor de rotação diferencial relativa de 11,4%. O monitoramento das manchas indica também que a estrela provavelmente possui um ciclo de atividade, com alta concentração de manchas no polo.

Dans le cadre de ma thèse, j'ai effectué l'analyse spectrale d'un échantillon massif de spectres stellaires dans le contexte du programme exoplanète de la mission CoRoT. J'ai tiré avantage des instruments Flames/GIRAFFE pour observer près de 2000 étoiles dans le but de comprendre les populations stellaires dans les champs CoRoT. Dans ce but, j'ai implémenté, calibré, testé, et appliqué une chaîne de traitement et de réduction fiable et efficace afin de réduire et d'analyser automatiquement (en utilisant l'algorithme de paramétrisation MATISSE, Gazzano et al. 2010) un large échantillon de spectres stellaires. J'ai déterminé la vitesse barycentrique radiale, une estimation de la vitesse de rotation projetée sur la ligne de visée, la température effective, la gravité de surface, de la métallicité global et l'enrichissement des éléments par rapport au fer pour 1 227 étoiles dans trois des champs CoRoT. Ainsi, j'ai construit un des premiers échantillons affranchis de biais de sélection pour toute étude concernant la relation planète métallicité dans les champs CoRoT et démontré que le nombre des étoiles naines a été généralement sous-estimé par la classification photométrique (Exo-Dat, Deleuil et al. 2009). J'ai appliqué la relation reliant le nombre de planètes détectées à la métallicité de l'étoile hôte (Udry & Santos 2007), parfaitement en accord avec le nombre actuel de détection planétaire dans les champs CoRoT correspondant (Gazzano et al. 2010). En utilisant les paramètres atmosphériques MATISSE, nous avons déterminé les distances et nous les avons combinées avec des informations cinématiques (les mouvements propres du catalogue PPMXL - Roeser et al. (2010), et l'astrométrie). Nous avons dérivé les composantes de cinématique Galactique : position et vitesse. Cela m'a permis d'étudier les populations stellaires dans les champs CoRoT /exoplanète et de quantifier le gradient de métallicité dans la Galaxie (Gazzano et al. En préparation). J'ai par ailleurs montré que les étoiles à planètes détectées dans les champs considérés pour ma thèse sont exclusivement des étoiles de disque mince. J'ai également participé au suivi des candidats planète CoRoT à l'aide de spectroscopie à haute résolution. J'ai effectué l'analyse spectrale, avec le logiciel VWA (Bruntt et al. 2010b,a), des étoiles hôtes pour la mission spatiale CoRoT. Ces études ont conduit à la détermination des paramètres fondamentaux de l'étoile, qui est une étape indispensable pour la caractérisation complète de la planète. / During my Ph.D., I performed the spectral analysis of a massive sample of stellar spectra in the context of the CoRoT /Exoplanet mission. We took advantage of the Flames/GIRAFFE multi-fibre instrument to observe almost 2 000 stars with the aim of understanding the stellar populations in the CoRoT fields. To these purposes, I implemented, calibrated, tested and applied an automatic pipeline to reduce and analyse automatically (using the parameterization algorithm MATISSE, Gazzano et al. 2010) a large sample of stellar spectra. I derived the barycentric radial velocity, an estimate of the rotational velocity projected on the line of sight, the effective temperature, the surface gravity, the overall metallicity and the -enhancement for 1227 stars in three of the CoRoT fields. Hence, I built one of the first unbiased samples for any study regarding planet metallicity relationship in the CoRoT fields and demonstrated that the amount of dwarf stars was generally underestimated by the photometric classification (in Exo-Dat, Deleuil et al. 2009). I applied the relationship linking the number of planets as a function of the metallicity of the host star (Udry & Santos 2007), totally in agreement with the current number of planetary detection in the corresponding CoRoT fields (Gazzano et al. 2010). Using MATISSE atmospheric parameters, we determined distances and combining them with kinematics information (proper motions from PPMXL catalogue - Roeser et al. (2010) and astrometry), we derived Galactic kinematics components : position, velocities and orbits. This allowed me to study the stellar populations in the CoRoT /Exoplanet fields and quantify the metallicity gradient in the Galaxy (Gazzano et al. in preparation). I also participated to the spectroscopic follow-up observations of CoRoT planetary candidates with high resolution spectroscopy, to the spectroscopic analysis, with the VWA software (Bruntt et al. 2010b,a), of planet hosting stars for the CoRoT space mission and to their characterisation and publication. Indeed, the determination of the fundamental parameters of the star is a mandatory step for the complete characterisation of the planet.

La formation, l’évolution et la structure des planètes géantes font parties des grandesproblématiques de l’astrophysique moderne. Les planète géantes ont un rôle majeur carelles possèdent la plupart de la masse des systèmes planétaires et donc influencent leursévolutions dynamiques. Mon travail de thèse s’inscrit dans une démarche observationnellequi est essentielle pour apporter des contraintes sur la diversité des systèmes exoplanétaires.Mes premiers résultats de thèse sont issus d’une campagne d’observations sur trois ansréalisées avec l’instrument NaCo au VLT. Mes observations de HD142527 excluent laprésence d’une planète géante dans le disque et favoriseraient plutôt un système multiplede faible masse pour expliquer les structures de ce disque de transition. J’ai égalementdétecté une planète géante autour de HD95086. Cette planète possède des propriétés atmosphériquesparticulières. Sa présence fait de HD95086 un rare exemple de systèmesimagés possédant un disque de débris et une planète géante. Enfin, j’ai réalisé une étudestatistique sur l’ensemble du relevé et montré que les planètes géantes sur des orbiteséloignées sont rares (10 − 20 %) et ne peuvent pas s’être formées majoritairement pareffondrement direct du gaz dans un disque instable.La dernière partie de mon travail de thèse a été consacrée à l’étude du mode d’imageriedifférentielle simultanée spectrale. J’ai identifié les biais particuliers liés à la réductionde ce type de données et ai montré que leurs analyses nécessitent l’utilisation de modèlesd’évolution. Cette étude permettra d’exploiter les données de l’instrument IRDISde SPHERE installé au VLT. / How giant exoplanet form, evolve and are made of is one of the biggest challenge of modernastronomy. They play an important role as they carry most of the planetary systemmasses. Therefore, they strongly impact their dynamics and the fate of these systems tohost life. My PhD thesis falls within an observational approach that is mandatory to bringconstraints on the diversity of planetary systems.I got my first results from a three-year survey, with NaCo at VLT. My observations ofHD142527 excluded the presence of massive giants planets to explain the structures of thistransitional disk and might favor a light multiple system. I also detected a giant planetaround HD95086 and I showed that it has particular atmospheric properties. Finally, Icarried out a statistical analysis of the whole survey to show that giant planets on wideorbitsare rare (10 − 20 %) and could not be formed via direct collapse of unstable disks.I dedicated the last part of my work to investigate the spectral simultaneous differentialimaging mode. I pointed out the biases associated to the reduction of these data andshowed that evolutionary models have to be used to analyze them. This study might helpto exploit the full potential of SPHERE/IRDIS data.