>Magister Scientiae - MSc / Future radio surveys as the Square Kilometer Array (SKA) and its precursor, the "Meer" Karoo Array Telescope (MeerKAT), will map the Neutral Hydrogen (HI) in large areas of the sky using the intensity mapping (IM). HI IM is currently one of the most promising ways of accessing the Large-Scale Structure of the Universe. The distribution of matter in the Universe not only encodes its composition but also how it evolves and its initial conditions. An effect on the matter distribution that will be detected by the SKA on the post re-ionization Universe are the Baryonic Acoustic Oscillations (BAO). While it has been shown that in single dish mode the SKA can measure the BAO peak in the radial 21cm power spectrum at low redshifts, this possibility has not yet been studied in detail for the MeerKAT. In this thesis we construct a set of full sky simulations to test how well MeerKAT will be able to extract the BAO wiggles along the line of sight. These simulations are done for the frequencies corresponding to MeerKAT L-band. The maps combine the cosmological HI signal, systematic noise, cosmological foregrounds and the instrumental telescope beam. A model-independent estimator is used to extract the BAO wiggles by subtracting a smooth polynomial component from the 21cm radial power spectrum. We test with simulations if this estimator is biased and the signal to noise of the extraction. We conclude that we are able to remove contaminants and recover the cosmological HI signal while not risking the recovery of the BAO signal. We investigate the effects of varying the sky area and the observational hours on the signal to noise ratio for the BAO wiggles. We found that for a HI IM experiment using MeerKAT, the optimal sky area to detect the BAO along the line of sight is 50% of the sky. With a signal-to-noise ratio of 3.37. This can be achieved with 2000 hours of exposure time

The method of baryon acoustic oscillation (BAO) is among the best probes of the dark energy equation of state, and worldwide efforts are being invested in order to perform measurements that are accurate at the percent level. In current data analyses, however, estimates of the error about the BAO are based on the assumption that the density field can be treated as Gaussian, an assumption that becomes less accurate as smaller scales are included in the measurement. It was recently shown from large samples of N-body simulations that the error bars about the BAO obtained this way are in fact up to 15-20 per cent too small. This important bias has shaken the confidence in the way error bars are calculated, and is motivating developments of analyses pipelines that include non-Gaussian features in the matter density fields. In this thesis, we propose general strategies to incorporate non-Gaussian effects in the context of a survey. After describing the high performance N-body code that we used, we present novel properties of the non-Gaussian uncertainty about the matter power spectrum, and explain how these combine with a general survey selection function. Assuming that the non-Gaussian features that are observed in the simulations correspond to those of Nature, this approach is the first unbiased measurement of the error bar about the power spectrum, which simultaneously removes the undesired bias on the BAO error. We then relax this assumption about the similitude of the non-Gaussian natures in simulations and data, and develop tools that aim at measuring the non-Gaussian error bars exclusively from the data. It is possible to improve the constraining power of non-Gaussian analyses with `Gaussianizations' techniques, which map the observed fields into something more Gaussian. We show that two of such techniques maximally recover degrees of freedom that were lost in the gravitational collapse. Finally, from a large sample of high resolution N-body realizations, we construct a series of weak gravitational lensing distortion maps and provide high resolution halo catalogues that are used by the CFTHLenS community to calibrate their estimators and study many secondary effects with unprecedented accuracy.

The method of baryon acoustic oscillation (BAO) is among the best probes of the dark energy equation of state, and worldwide efforts are being invested in order to perform measurements that are accurate at the percent level. In current data analyses, however, estimates of the error about the BAO are based on the assumption that the density field can be treated as Gaussian, an assumption that becomes less accurate as smaller scales are included in the measurement. It was recently shown from large samples of N-body simulations that the error bars about the BAO obtained this way are in fact up to 15-20 per cent too small. This important bias has shaken the confidence in the way error bars are calculated, and is motivating developments of analyses pipelines that include non-Gaussian features in the matter density fields. In this thesis, we propose general strategies to incorporate non-Gaussian effects in the context of a survey. After describing the high performance N-body code that we used, we present novel properties of the non-Gaussian uncertainty about the matter power spectrum, and explain how these combine with a general survey selection function. Assuming that the non-Gaussian features that are observed in the simulations correspond to those of Nature, this approach is the first unbiased measurement of the error bar about the power spectrum, which simultaneously removes the undesired bias on the BAO error. We then relax this assumption about the similitude of the non-Gaussian natures in simulations and data, and develop tools that aim at measuring the non-Gaussian error bars exclusively from the data. It is possible to improve the constraining power of non-Gaussian analyses with `Gaussianizations' techniques, which map the observed fields into something more Gaussian. We show that two of such techniques maximally recover degrees of freedom that were lost in the gravitational collapse. Finally, from a large sample of high resolution N-body realizations, we construct a series of weak gravitational lensing distortion maps and provide high resolution halo catalogues that are used by the CFTHLenS community to calibrate their estimators and study many secondary effects with unprecedented accuracy.

J'ai démontrer la faisabilité de la sélection de la cible pour les galaxies en ligne des émissions lumineuses. Je comprends maintenant les principaux mécanismes physiques de conduite de l'efficacité d'une sélection, en particulier le rapport à la photométrie de parent. Une question reste perplexe, je ne pouvais pas encore estimer quantitativement l'impact de la poussière sur l'efficacité de la sélection. J'espère que d'aborder cette question avec l'ensemble des données décrites dans le chapitre 4.En dehors de la ligne de sélection de la cible de la galaxie d'émission, j'ai étudié, au premier ordre, les deux principales erreurs systématiques sur la détermination de l'échelle BAO nous attendent en raison de l'utilisation galaxies en ligne des émissions comme traceurs de la question. J'ai d'abord montré le caractère incomplet de la distribution redshift, en raison de la mesure du décalage spectral avec [Oii], est lié à la résolution instrumentale. Je trouve qu'il ya deux régimes intéressants. Pour une observation des plus brillants [OII] émetteurs, une résolution modérée est suffisante, alors que pour une enquête plus faible, la plus haute de la résolution le meilleur. Deuxièmement, j'ai estimé le biais de la galaxie linéaire des sélections discuté avant et je trouve qu'ils sont très biaisés. D'une part, ce sont d'excellentes nouvelles pour les observateurs, comme le temps nécessaire pour observer à un signal donné au bruit dans le spectre de puissance diminue avec le carré de la partialité. D'autre part, elle constitue un nouveau défi pour les algorithmes de reconstruction et la fabrication de catalogues simulacres. / In this PhD, I demonstrate the feasibility of the target selection for bright emission line galaxies. Also I now understand the main physical mechanisms driving the efficiency of a selection, in particular the relation to the parent photometry. A puzzling issue remains, I could not yet estimate quantitatively the impact of the dust on the selection efficiency. I hope to address this question with the data set described in chapter 4.Apart from the emission line galaxy target selection, I investigated, at first order, the two main systematic errors on the determination of the BAO scale we expect due to using emission line galaxies as tracers of the matter. First I showed the incompleteness in the redshift distribution, due to the measurement of the redshift with [Oii], is related to the instrumental resolution. I find there are two interesting regimes. For an observation of the brightest [Oii]emitters, a moderate resolution is sufficient, whereas for a fainter survey, the highest the resolution the best. Secondly, I estimated the linear galaxy bias of the selections discussed before and I find they are highly biased. On one hand, this is great news for the observers, as the time required to observed at a given signal to noise in the power spectrum decreases with the square of the bias. On the other hand, it constitutes a new challenge for reconstruction algorithms and the making of mock catalogs. The work in progress described in the last chapter shows I am starting to try and handle these questions in a robust manner.

Il y a presque dix ans que l'accélération de l'expansion de l'Univers a été mise en évidence grâce aux observations des supernovae de type Ia et du fonds diffus cosmologique. Cette découverte a changé notre compréhension du contenu énergétique de l'Univers puisque pour expliquer une telle accélération, une composante supplémentaire de matière (effective ou non) est nécessaire et contribue à hauteur de 70%. Cette dernière est appelé «énergie noire». Elle affecte aussi bien les mesures de distance, que la croissance des sur-densités de matières primordiales qui donnent naissance aux structures. Les principales sondes sensibles à ces deux dernières quantités sont les supernovae de type Ia, les amas de galaxies, les lentilles gravitationnelles, et les oscillations acoustiques des baryons (BAO). Afin de contraindre précisément les modèles théoriques (Constante Cosmologique, modification de la théorie de la relativité générale par exemple) qui tentent de déterminer la nature de l'énergie noire, l'observation de chacune de ces quatre sondes est indispensable. Le niveau de précision sur la mesure des paramètres des modèles d'énergie noire requis est tel qu'une nouvelle génération d'instruments va voir le jour dans les années à venir avec notamment le télescope LSST (Large Synoptic Survey Telescope). Le télescope LSST dont le miroir primaire fait 8.4 mètres de diamètre, produira un sondage couvrant la moitié du ciel observable dans 6 bandes photométriques pendant 10 ans. Sa caméra sera la plus grosse caméra jamais construite dans le monde avec un plan focal de 3.2 milliards de pixels. Cette thèse présente à la fois un aspect expérimental et phénoménologique. Le travail présenté porte tout d'abord sur l'élaboration du banc d'étalonnage de la caméra de LSST, et des premières mesures optiques validant le schéma de principe du banc. Nous présenterons ensuite la simulation des BAO dédiée à LSST tentant de prédire à quelle précision les paramètres d'énergie noire pourront être contraint. L'accent est mis sur la production d'un catalogue photométrique de galaxies simulé ainsi que sur une méthode de calcul des redshifts photométriques. La validation de la méthode grâce à des données spectro-photométriques du CFHTLS est également présentée. / More than ten years ago, the accelerated expansion of the Universe was discovered, by type Ia supernovae, and then confirmed by other probes. This discovery has changed our understanding of the energetic content of the Universe. Indeed, in order to explain such an acceleration, a new component has to be introduced and it must contribute to 70% of the total energy density. This component, the so called Dark Energy, affects both cosmological distances and the growth of structures from which galaxies originates. The main cosmological probes of dark energy are the type Ia supernovae, the galaxy cluster count, the weak gravitational lensing and the baryon acoustic oscillations (BAO). In order to precisely constrain theoretical models, such as the cosmological constant, a modify gravity or a new scalar field, joint observations of all four probes are very efficient. The required accuracy on cosmological measurements is so high that a new generation of instruments is growing, among which the Large Synoptic Survey Telescope (LSST). The telescope, with a primary mirror of 8.4 m diameter, will cover half of the optical sky in six photometric bandpasses. Its camera will be the world biggest camera ever constructed with a focal plane array composed of 3.2 Gpixels. This thesis treats both the experimental and phenomenological aspects. Firstly, the work presented here consists in the development of the LSST camera calibration optical bench. We have designed a system allowing an efficient commissioning of the camera before its installation on the telescope, and a precise calibration of the focal plane. Preliminary measurements validating the design of the bench will be presented. Secondly, a detailed Baryon Acoustic Oscillations simulation dedicated to LSST will be introduced. Its main goal is to predict the level of precision on the dark energy equation of state parameter reconstruction that will be reached with LSST. We will stress on the production of a mock photometric galaxy catalog and on the photometric redshifts computation. A validation of the method on real spectro-photometric from CFHTLS will also be shown.

This thesis presents work around 3 themes: dark energy, gravitational waves and Bayesian inference. Both dark energy and gravitational wave physics are not yet well constrained. They present interesting challenges for Bayesian inference, which attempts to quantify our knowledge of the universe given our astrophysical data. A dark energy equation of state reconstruction analysis finds that the data favours the vacuum dark energy equation of state $w {=} -1$ model. Deviations from vacuum dark energy are shown to favour the super-negative ‘phantom’ dark energy regime of $w {< } -1$, but at low statistical significance. The constraining power of various datasets is quantified, finding that data constraints peak around redshift $z = 0.2$ due to baryonic acoustic oscillation and supernovae data constraints, whilst cosmic microwave background radiation and Lyman-$\alpha$ forest constraints are less significant. Specific models with a conformal time symmetry in the Friedmann equation and with an additional dark energy component are tested and shown to be competitive to the vacuum dark energy model by Bayesian model selection analysis: that they are not ruled out is believed to be largely due to poor data quality for deciding between existing models. Recent detections of gravitational waves by the LIGO collaboration enable the first gravitational wave tests of general relativity. An existing test in the literature is used and sped up significantly by a novel method developed in this thesis. The test computes posterior odds ratios, and the new method is shown to compute these accurately and efficiently. Compared to computing evidences, the method presented provides an approximate 100 times reduction in the number of likelihood calculations required to compute evidences at a given accuracy. Further testing may identify a significant advance in Bayesian model selection using nested sampling, as the method is completely general and straightforward to implement. We note that efficiency gains are not guaranteed and may be problem specific: further research is needed.