Le SuperNova Legacy Survey (SNLS) est un programme de détection et de suivi photométrique de plusieurs centaines de supernovæ de type Ia (SNe Ia) dont l'objectif est de retracer l'histoire de l'expansion cosmologique afin d'en déduire une caractérisation de la nature de l'énergie noire, c'est-à-dire une mesure de son paramètre d'état wDE. La prise de données du SNLS est arrivée à son terme en juillet 2008 après 5 ans de programme. Le travail réalisé dans le cadre de cette thèse a consisté en l'analyse de ces 5 ans de données SNLS et la photométrie des 419 SNe Ia détectées et spectroscopiquement identifiées. Pour chaque supernova, les courbes de lumière dans les bandes gMrMiMzM sont produites, calibrées et ajustées par un modèle spectrophotométrique. Une nouvelle méthode de photométrie, sans rééchantillonnage des pixels des images, est également implémentée dans le cadre de cette thèse. En préservant les propriétés statistiques des pixels, elle permet de mieux contrôler les incertitudes sur la mesure des flux, et par là même, la précision sur la mesure des paramètres cosmologiques qui s'en déduit. Les performances des deux méthodes sont testées et comparées sur les étoiles de calibration et les supernovae. Si la photométrie sans rééchantillonnage permet une estimation plus exacte des incertitudes de mesure du flux, la précision et la stabilité des deux méthodes sont quant à elles similaires. Un lot de SNe Ia de statistique et de qualité inédites est maintenant disponible pour l'analyse de cosmologie. Avec le complément de supernovæ proches de programmes extérieurs, une contrainte à 5% sur la nature du paramètre d'état de l'énergie noire est donc pour la première fois envisageable.

Avec la découverte étonnante que l'univers se trouve à présent dans une phase d'expansion accélérée, il y a dix ans, la cosmologie entra dans ce que l'on peut nommer l'ère de la cosmologie à haute précision. Les contraintes actuelles indiquent un modèle cosmologique à géométrie plate, où la plus grande partie du contenu en masse-énergie de l'univers est contribuée par une composante inconnue, souvent appelée l'`énergie noire', qui contribue environ soixante-dix pour cent à la densité totale de l'univers. Dans ce modèle, la matière baryonique ordinaire et la radiation ne contribuent qu'environ cinq pour cent, et la matière noire contribue vingt-cinq pour cent. Les propriétés mesurées de l'énergie noire étant consistant avec celles d'une Constante Cosmologique, $Lambda$, ce modèle standard cosmologique est connu sous le nom du modèle `$Lambda$-Cold-Dark-Matter' (`$Lambda$CDM'). Malgré son succes, ce modèle souffre de plusieurs problèmes. L'existence d'une Constante Cosmologique soulève des problèmes fondamentaux concernant sa nature physique, et beaucoup d'auteurs traitent le `problème de coîncidence'. Des essais de la décrire comme la contribution du vide quantique faillissent quantitativement. En conséquent, un grand nombre de modèles alternatifs a été développé, qui tentent décrire la composante d'énergie noire: des loi modifiés de la gravitation, de dimensions supplémentaires, les modèles de Quintessence. Aussi, des effets astrophysiques qui miment une expansion accélérée ont été considérés. Dans ce manuscrit, on expose les bases théoriques et observationneles du modèle $Lambda$CDM et les divers approches théoriques à expliquer l'énergie noire. Un autre problème du modèle standard provient de la dépéndance des résultats de l'analyse des données sur des hypothèses qui sont présentes dans les analyses pour l'extraction des paramètres. Il s'agit des hypothèses sur la physique, mais aussi des dépéndances des paramétrages notamment des propriétés de l'énergie noire. Aujourd'hui, des analyses combinées de divers sondes cosmologiques sont effectuées afin d'extraire les paramètres du modèle, dont le nombre peut s'élever jusqu'à vingt, dépendant des suppositions de modèle. De différentes hypothèses (géométrie plate, équation d'état de l'énergie noire constante, hypothèses sur la physique du CMB comme la vitesse du son ou le spectre de puissance initial,...) sont appliquées, qui peuvent dangereusement biaiser les résultat de l'analyse. La présence d'une mauvaise hypothèse, où une application d'un paramétrage non-approprié de la physique, pourra entraîner qu'on mesure à haute précision quelque chose qui n'est pas là. Nous montrons que, dû à la haute précision des mesures cosmologiques modernes, des approches purement cinématiques à la cosmologie ne permettent plus d'extraire des résultats fiables sur l'expansion de l'univers. Dans l'analyse des données cosmologiques on doit par conséquent se servir de la relation (exacte) intégrale pour les distances cosmologiques. Nous discutons le problème de dégénéréscence analytique entre les paramètres cosmologiques qu'introduit l'utilisation de cette relation. Puis, les résultats principaux de ce travail sont présentés. Ils concernent notamment la validité du paramétrage de l'équation d'état de l'énergie noire de Chevallier, Polarski, et Linder, et les éffets d'une évolution en redshift des magnitudes apparentes des Supernovae du type Ia.

Supernovae are some of the most energetic phenomena in the Universe and they have throughout history fascinated people as they appeared as new stars in the sky. Supernova (SN) 1987A exploded in the nearby satellite galaxy, the Large Magellanic Cloud (LMC), at a distance of only 168,000 light years. The proximity of SN 1987A offers a unique opportunity to study the medium surrounding the supernova in great detail. Powered by the dynamical interaction of the ejecta with the inner circumstellar ring, SN 1987A is dramatically evolving at all wavelengths on time scales less than a year. This makes SN 1987A a great ``laboratory'' for studies of shock physics. Repeated observations of the ejecta-ring collision have been carried out using the UVES echelle spectrograph at VLT. This thesis covers seven epochs of high resolution spectra taken between October 1999 and November 2007. Three different emission line components are identified from the spectra. A narrow (~10 km/s) velocity component emerges from the unshocked ring. An intermediate (~250 km/s) component arises in the shocked ring, and a broad component extending to ~15,000 km/s comes from the reverse shock. Thanks to the high spectral resolution of UVES, it has been possible to separate the shocked from the unshocked ring emission. For the unshocked gas, ionization stages from neutral up to Ne V and Fe VII were found. The line fluxes of the low-ionization lines decline during the period of the observations. However, the fluxes of the [O III] and [Ne III] lines appear to increase and this is found to be consistent with the heating of the pre-shock gas by X-rays from the shock interactions. The line emission from the ejecta-ring collision increases rapidly as more gas is swept up by the shocks. This emission comes from ions with a range of ionization stages (e.g., Fe II-XIV). The low-ionization lines show an increase in their line widths which is consistent with that these lines originate from radiative shocks. The high-ionization line profiles (Fe X-XIV) initially show larger spectral widths, which indicates that at least a fraction of the emission comes from non-radiative shocks.

Type Ia supernovae can be used as distance indicators for probing the expansion history of the Universe. The method has proved to be an efficient tool in cosmology and played a decisive role in the discovery of a yet unknown energy form, dark energy, that drives the accelerated expansion of the Universe. The work in this thesis addresses the nature of dark energy, both by presenting existing data, and by predicting opportunities and difficulties related to possible future data. Optical and infrared measurements of type Ia supernovae for different epochs in the cosmic expansion history are presented along with a discussion of the systematic errors. The data have been obtained with several instruments, and an optimal method for measuring the lightcurve of a background contaminated source has been used. The procedure was also tested by applying it on simulated images. The future of supernova cosmology, and the target precision of cosmological parameters for the proposed SNAP satellite are discussed. In particular, the limits that can be set on various dark energy scenarios are investigated. The possibility of distinguishing between different inverse power-law quintessence models is also studied. The predictions are based on calculations made with the Supernova Observation Calculator, a software package, introduced in the thesis, for simulating the light propagation from distant objects. This tool has also been used for investigating how SNAP observations could be biased by gravitational lensing, and to what extent this would affect cosmology fitting. An alternative approach for estimating cosmological parameters, where lensing effects are taken into account, is also suggested. Finally, it is investigated to what extent strongly lensed core-collapse supernovae could be used as an alternative approach for determining cosmological parameters.

Cette thèse, effectuée à l'Institut de Physique Nucléaire de Lyon (IPNL), s'inscrit dans le cadre de l'expérience SNfactory, dédiée à l'observation précise d'un grand nombre de supernovae de type Ia. Ces explosions thermonucléaires de naines blanches, très lumineuses, sont aujourd'hui les meilleures chandelles standards connues et ont de ce fait un grand intérêt pour la cosmologie.<br />Les travaux exposés ici contribuent à l'étalonnage de SNIFS, le spectrographe à champ intégral utilisé par SNfactory pour l'observation des supernovae de type Ia. Une étude des CCD a été effectuée ainsi que la collection de spectres d'étoiles de référence pour l'étalonnage en flux de l'instrument. La contribution principale a consisté en la réalisation d'un code complet pour l'extraction du spectre d'une source ponctuelle à partir du cube pré-étalonné de données. Le procédé choisi est la photométrie de PSF (Point Spread Function), nécessaire ici du fait de la petite taille du champ observé. Le flux perdu peut ainsi être évalué, et l'on atteint une qualité de mesure du flux de l'ordre du pourcent, objectif de l'expérience.

The character of the first galaxies at redshifts z [greater-than or equal to] 10 strongly depends on the star formation which takes place during their assembly. Conducting cosmological hydrodynamics simulations, we study how the radiative output and chemical enrichment from the first stars impacts the properties of the first galaxies. We find that the radiative feedback from the first stars suppresses the star formation rate at redshifts z [greater-than or equal to] 15 by a factor of only a few. In turn, this suggests that a large fraction of the first galaxies may form from gas which has already been enriched with the first heavy elements ejected by primordial supernovae. In order to characterize the properties of primordial dwarf galaxies, we carry out radiation hydrodynamics simulations which allow to determine how the luminosities in hydrogen and helium emission lines depend on the initial mass function of the stars in the galaxy. As well, we show that the chemical abundance patterns observed in metal-poor Galactic halo stars contain the signature of the first supernovae, and we use this data to indirectly probe the properties of the first stars. / text

The discovery of superluminous supernovae in the past decade challenged our understanding of explosive stellar death. Subsequent extensive observations of superluminous supernova light curves and spectra has provided some insight for the nature of these events. We present observations of one of the most luminous self-interacting supernovae ever observed, the hydrogen-rich SN 2008am discovered by the Robotic Optical Transient Search Experiment Supernova Verification Project with the ROTSE-IIIb telescope located in the McDonald Observatory. We provide theoretical modeling of superluminous supernova light curves and fit the models to a number of observed events and similar transients in order to understand the mechanism that is responsible for the vast amounts of energy emitted by these explosions. The models we investigate include deposition of energy due to the radioactive decays of massive amounts of nickel-56, interaction of supernova ejecta with a dense circumstellar medium and magnetar spin-down. To probe the nature of superluminous supernovae progenitor stars we study the evolution of massive stars, including important effects such as rotation and magnetic fields, and perform multi-dimensional hydrodynamics simulations of the resulting explosions. The effects of rotational mixing are also studied in solar-type secondary stars in cataclysmic variable binary star systems in order to provide an explanation for some carbon-depleted examples of this class. We find that most superluminous supernovae can be explained by violent interaction of the SN ejecta with >1 Msun dense circumstellar shells ejected by the progenitor stars in the decades preceding the SN explosion. / text

This thesis presents the CfAIR2 data set, which includes over 4000 near-Infrared (NIR) \(JHK_s\)-band measurements of 104 Type Ia Supernovae (SN Ia) observed from 2005-2011 using PAIRITEL, the 1.3-m Peters Automated InfraRed Imaging TELescope at the Fred Lawrence Whipple Observatory (FLWO) on Mount Hopkins, Arizona. While the discovery of dark energy and most subsequent supernova cosmology has been performed using optical and Ultraviolet wavelength observations of SN Ia, a growing body of evidence suggests that NIR SN Ia observations will be crucial for future cosmological studies. Whereas SN Ia observed at optical wavelengths have been shown to be excellent standardizeable candles, using empirical correlations between luminosity, light curve shape, and color, the CfAIR2 data set strengthens the evidence that SN Ia at NIR wavelengths are essentially standard candles, even without correction for light-curve shape or for reddening. CfAIR2 was obtained as part of the CfA Supernova Program, an ongoing multi-wavelength follow-up effort at FLWO designed to observe high-quality, densely sampled light curves and spectra of hundreds of low-redshift SN Ia. CfAIR2 is the largest homogeneously observed and processed NIR data set of its kind to date, nearly tripling the number of individual \(JHK_s\) band observations and nearly doubling the set of SN Ia with published NIR light curves in the literature. Matched only by the recently published Carnegie Supernova Project sample, CfAIR2 complements the large and growing set of low-redshift optical and NIR SN Ia observations obtained by the CfA and other programs, making this data set a unique and particularly valuable local universe anchor for future supernova cosmology. / Astronomy

Astrophysical flows have an enormous dynamic range of relevant length scales. The physics occurring on the smallest scales often influences the physics of the largest scales, and vice versa. I present a detailed study of the multiscale and multidimensional behavior of three high-energy astrophysical flows: jet-driven supernovae, massive black hole accretion, and current-driven instabilities in gamma-ray burst external shocks. Both theory and observations of core-collapse supernovae indicate these events are not spherically-symmetric; however, the observations are often modeled assuming a spherically-symmetric explosion. I present an in-depth exploration of the effects of aspherical explosions on the observational characteristics of supernovae. This is accomplished in large part by high-resolution, multidimensional numerical simulations of jet-driven supernovae. The existence of supermassive black holes in the centers of most large galaxies is a well-established fact in observational astronomy. How such black holes came to be so massive, however, is not well established. In this work, I discuss the implications of radiative feedback and multidimensional behavior on black hole accretion. I show that the accretion rate is drastically reduced relative to the Eddington rate, making it unlikely that stellar mass black holes could grow to supermassive black holes in less than a Hubble time. Finally, I discuss a mechanism by which magnetic field strength could be enhanced behind a gamma-ray burst external shock. This mechanism relies on a current-driven instability that would cause reorganization of the pre-shock plasma into clumps. Once shocked, these clumps generate vorticity in the post-shock plasma and ultimately enhance the magnetic energy via a relativistic dynamo process. / text

Supernovae are cosmic explosions of cataclysmic proportion that signify the death of a star. While being interesting phenomena in their own right, their brightness also make them excellent probes of the early universe. Depending on the type of the progenitor star and the origin of the explosion different subjects can be investigated. In this dissertation the work I have done on the detection, characterisation and rate measurements of supernovae in the Stockholm VIMOS Supernova Search is presented. We have discovered 16 supernovae that exploded billions of years ago (or, equivalently, at high redshift, z). The observed brightness and colour evolution have been used to classify the supernovae into either thermonuclear (type Ia) or core collapse (type II) supernovae. The accuracy of the classification code is high, only about 5% of the supernovae are mistyped, similar to other codes of the same kind. By comparing the observed frequency of supernovae to simulations the underlying supernova rate at these high redshifts have been measured. The main result reported in this thesis is that the core collapse supernova rate at high redshift matches the rates estimated from looking at the star formation history of the universe, and agree well with previous studies. The rate of Ia supernovae at high redshift have been investigated by several projects, our results show a somewhat higher rate of Ia supernovae than expected. Proper estimates of the systematic errors of rate measurements are found to be very important. Furthermore, by using novel techniques for reducing and stacking images, we have obtained a galaxy sample containing approximately 50,000 galaxies. Photometric redshifts have been obtained for most of the galaxies, the resulting accuracy below z=1 is on the order of 10%. The galaxy sample has also been used to find high redshift sources, so called Lyman Break Galaxies, at z=3-5. / At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Submitted. Paper 3: Manuscript. Paper 4: Manuscript.