Search

Galaxy rotation has been studied for over a century, using spectroscopic measurements to construct rotational models that describe the motions of their gas and stars. However, only a small subset of this work has recognized that galaxies are not entirely radially symmetrical, but are instead often disrupted in ways that simple velocity field models cannot capture. In this dissertation, I describe my work modelling nonaxisymmetric galaxy rotation and the astrophysical insights gained from these models.Gravitational lensing distorts a galaxy's velocity field in a manner distinct from the distortions of its photometric shape, an effect called kinematic weak lensing. This allows for the construction of a model that uses this difference to extract lensing information about the system. I detail the properties and strengths of such a model, finding that for mock observations of source galaxies at moderate redshifts, the signal-to-noise of the lensing measurements improves by up to a factor of six over previous works, enabling the possibility of future lensing studies independent of current weak lensing systematic constraints.I also describe the development of Nirvana, a Bayesian nonparametric velocity field fitting code designed to describe the bisymmetric motions present in barred galaxies. Using a sample of barred galaxies from the MaNGA survey, I construct the Nirvana-MaNGA sample, which is comprised of velocity field models of >1000 local barred galaxies, as well as a matched control sample of unbarred galaxies. Nirvana determines bar strength and location independent of imaging, providing an independent and direct test of dynamical models of higher-order noncircular motions in bars, agreeing with visual bar classifications on bar angle. I also find direct evidence of flattening in stellar population gradients along bar kinematic axes as compared to surrounding disk regions at the same radii, verifying results reliant on visual classifications and affirming the dynamic connection between the presence of bars and radial mixing of stellar populations.

We perform the first harmonic space joint analysis of Galaxy Clustering and Weak Lensing cosmological probes with the Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) public data. This project has two main goals: (i) set precise constraints to ΛCDM cosmological parameters analysing the HSC-SSP public data in this multi-probe study and (ii) develop measurement and analysis tools that will be required when the Rubin Observatory Legacy Survey of Space and Time (LSST) starts operating at the end of 2023. HSC is optimal as a precursor of the LSST Dark Energy Science Collaboration (DESC) as it is the deepest Stage III survey and uses a prototype version of the Rubin Science Pipeline. One of the highlights of our process is that we apply mode deprojection as a method to correct for the impact of observing conditions in the galaxy density field and in the spin-2 shear field. Another important ingredient in a multi-experiment analysis is the covariance matrix, necessary to account for the correlation between the different probes. We describe the implementation of the Gaussian and Super-Sampling terms. As our main result, we present preliminary constraints in the plane Ωm – S8 that are competitive with the current state of the art.

Gatti, M., et al. (DES Collaboration), We present and characterize the galaxy shape catalogue from the first 3 yr of Dark Energy Survey (DES) observations, over an effective area of 4143 deg2 of the southern sky. We describe our data analysis process and our self-calibrating shear measurement pipeline metacalibration, which builds and improves upon the pipeline used in the DES Year 1 analysis in several aspects. The DES Year 3 weak-lensing shape catalogue consists of 100 204 026 galaxies, measured in the riz bands, resulting in a weighted source number density of neff = 5.59 gal arcmin-2 and corresponding shape noise σe = 0.261. We perform a battery of internal null tests on the catalogue, including tests on systematics related to the point spread function (PSF) modelling, spurious catalogue B-mode signals, catalogue contamination, and galaxy properties., Based in part on observations at Cerro Tololo Inter-American Observatory at NSF’s NOIRLab (NOIRLab Prop. ID 2012B-0001; PI: J. Frieman), which is managed by the Association of Universities for Research in Astronomy (AURA) under a cooperative agreement with the National Science Foundation. The DES data management system is supported by the National Science Foundation under grants AST-1138766 and AST-1536171. The DES participants from Spanish institutions are partially supported by MINECO under grants AYA2015-71825, ESP2015-66861, FPA2015-68048, SEV-2016-0588, SEV-2016-0597, and MDM-2015-0509, some of which include ERDF funds from the European Union. IFAE is partially funded by the CERCA programme of the Generalitat de Catalunya. Research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Program (FP7/2007-2013) including ERC grant agreements 240672, 291329, and 306478. We acknowledge support from the Brazilian Instituto Nacional de Ciência e Tecnologia (INCT) e-Universe (CNPq grant 465376/2014-2).

The cross-correlation of gravitational wave strain with upcoming galaxy surveys probe theories of gravity in a new way. This method enables testing the theory of gravity by combining the effects from both gravitational lensing of gravitational waves and the propagation of gravitational waves in spacetime. We find that within 10 years, the combination of the Advanced-LIGO and VIRGO detector networks with planned galaxy surveys should detect weak gravitational lensing of gravitational waves in the low redshift Universe ($z, 16 pages, 7 figures. Minor changes in the text and added new references. Matches the version accepted for publication in MNRAS

In this work we present updated forecasts on parameterised modifications of gravity that can capture deviations of the behaviour of cosmological density perturbations beyond $\Lambda$CDM. For these forecasts we adopt the SKA Observatory (SKAO) as a benchmark for future cosmological surveys at radio frequencies, combining a continuum survey for weak lensing and angular galaxy clustering with an HI galaxy survey for spectroscopic galaxy clustering that can detect baryon acoustic oscillations and redshift space distortions. Moreover, we also add 21cm HI intensity mapping, which provides invaluable information at higher redshifts, and can complement tomographic resolution, thus allowing us to probe redshift-dependent deviations of modified gravity models. For some of these cases, we combine the probes with other optical surveys, such as the Dark Energy Spectroscopic Instrument (DESI) and the Vera C. Rubin Observatory (VRO). We show that such synergies are powerful tools to remove systematic effects and degeneracies in the non-linear and small-scale modelling of the observables. Overall, we find that the combination of all SKAO radio probes will have the ability to constrain the present value of the functions parameterising deviations from $\Lambda$CDM ($\mu$ and $\Sigma$) with a precision of $2.7\%$ and $1.8\%$ respectively, competitive with the constraints expected from optical surveys and with constraints we have on gravitational interactions in the standard model. Exploring the radio-optical synergies, we find that the combination of VRO with SKAO can yield extremely tight constraints on $\mu$ and $\Sigma$ ($0.9\%$ and $0.7\%$ respectively), which are further improved when the cross-correlation between intensity mapping and DESI galaxies is included., Comment: 24 pages, 10 figures

Weak gravitational lensing is one of the most promising probes to constrain dark energy parameters. It corresponds to the distortion of a source image, induced by the bending of space-time, thus of the light path, generated by the presence of mass along the line of sight. This effect is small and can only be detected measuring the correlation of the shapes of a group of background galaxies. The value of this correlation yields the value of the cosmic shear. Several surveys dedicated to its study are going to start soon, such as the Vera Rubin Legacy Surveys of Space and Time (LSST), and Euclid. These surveys have been built with particular requirements, especially about the treatment of systematic errors, in order to reach an extreme precision on the dark energy equation of state parameters. This thesis happened in the context of the Dark Energy Science Collaboration (DESC), inside the LSST experiment. It focuses on the issue of blending, the overlap of astronomical sources. The blending of galaxies introduces a dominant systematic uncertainty on the cosmic shear measurement. This systematic effect impacts the shape and the redshift measurements, both necessary for weak gravitational lensing analysis. We propose two avenues based on Bayesian deep learning methods. The first one is a deblending algorithm which uses a deep generative network called variational autoencoder. This neural network allows to learn a prior for the generation of isolated galaxy images. The latter is used in a second network to perform the deblending of the centred galaxy on images of simulated galaxies. We show that the pixel joint analysis of LSST and Euclid data decreases the median error on galaxy shape reconstruction from 8 to 47%. Blending of sources being closely linked to detection, we demonstrate that our method is robust to small decentring. Also we test our method on images of real galaxies artificially blended showing the interest of transfer learning from a neural network trained on simulated galaxy images. An iterative process is then designed in order to separate all the galaxies in an image going through detection, classification and deblending of sources. Finally this deblending algorithm is tested on images extracted from the DC2 simulation, generated within the DESC to prepare for the analysis of futures images taken by LSST. Our results show an improvement of 70 to 120% on the median error on galaxy shape reconstruction compared to the generic method used in the current LSST pipeline. Then, inspired by this first method, we propose a neural network allowing for the direct estimation of galaxy shape and redshift parameters from DC2 images, without going through the deblending. We show that this neural network allows for a precise measurement of these parameters, even when sources are blended, and we compare these results to the ones obtained with deblending. Then we present the first application of a Bayesian neural network to galaxy shapes estimation. This kind of network provides the estimation of the epistemic uncertainty which characterizes the uncertainty coming from the training sample. It can be viewed as a confidence level in the measurement performed by the network, a primordial information for the application of neural networks in science and in cosmology. It can be used to decrease the importance (or reject) of an incorrect measurement in the analysis.; Le lentillage gravitationnel faible est l'une des sondes cosmologiques les plus prometteuses pour contraindre les paramètres de l'énergie noire. Il correspond à la modification de l'image d'une source, induite par la courbure de l'espace-temps, et donc du passage des rayons lumineux, générée par la présence de masse le long de la ligne de visée. Cet effet est très faible et ne peut être détecté qu'en mesurant la corrélation des formes d'un ensemble de sources en arrière-plan. La valeur de cette corrélation donne la mesure du cisaillement gravitationnel. Plusieurs relevés dédiés à son étude vont bientôt démarrer, tels que le Vera Rubin Legacy Survey of Space and Time (LSST), et Euclid. Ces relevés ont été construits avec des exigences particulières, notamment sur le traitement des erreurs systématiques, afin d'atteindre une précision extrême sur les paramètres de l'équation d'état de l'énergie noire. Cette thèse s'est déroulée dans le contexte de la Dark Energy Science Collaboration (DESC), au sein de l'expérience LSST. Elle se focalise sur la problématique de la superposition de sources. Cet effet génère une incertitude systématique dominante dans la mesure du cisaillement gravitationnel. Cette systématique impacte à la fois les mesures de forme et celles du décalage vers le rouge des galaxies, deux mesures nécessaires à l'étude du lentillage gravitationnel faible. Nous proposons deux solutions basées sur les méthodes bayésiennes d'apprentissage profond. La première est un algorithme de séparation de galaxies qui utilise un réseau de neurones génératif appelé autoencodeur variationnel. Ce réseau de neurones permet l'apprentissage d'un prior pour la génération d'images de galaxie isolée. Celui-ci est utilisé dans un second réseau qui fait la séparation de la galaxie centrale sur des images de galaxies simulées. Nous montrons que l'analyse jointe des pixels des données LSST et Euclid permet une diminution de l'erreur médiane de reconstruction des formes des galaxies de 8 à 47%. La superposition de sources étant étroitement liée à l'étape de détection, nous démontrons la robustesse de notre méthode à de petits décentrages. Nous avons aussi testé notre méthode sur des images de vraies galaxies artificiellement superposées, montrant l'intérêt du transfert d'apprentissage à partir d'un réseau de neurones entraîné sur des images de galaxies simulées. Une processus itératif est ensuite mis en place afin de pouvoir séparer toutes les galaxies d'une image en passant par des étapes de détection, classification, et séparation des sources. Enfin, cet algorithme de séparation de galaxies est testé sur des images extraites de la simulation DC2, générée au sein de DESC dans le but de préparer à l'analyse des futures images prises par LSST. Nos résultats montrent une amélioration de 70 à 120% sur l'erreur médiane de reconstruction des formes comparée à la méthode générique utilisée dans la pipeline LSST actuellement. Ensuite, inspirés de cette première méthode, nous proposons un réseau de neurones permettant de faire directement l'estimation des paramètres de forme et de décalage vers le rouge d'une galaxie à partir des images DC2, sans appliquer de méthode de séparation de sources. Nous montrons que ce réseau permet des mesures précises de ces paramètres, même lorsque les sources sont superposées, et comparons ces résultats à ceux obtenus avec séparation de source. Par la suite, nous présentons une première application d'un réseau de neurones bayésien à l'estimation des formes des galaxies. Ce type de réseau permet l'estimation de l'incertitude épistémique qui représente l'incertitude liée à l'échantillon d'entraînement. Celle-ci peut être vue comme un niveau de confiance dans la mesure réalisée par le réseau, une information cruciale à l'utilisation des réseaux de neurones en science et en cosmologie. Elle permet de diminuer l'importance (voir de rejeter) d'une mesure qui a de fortes probabilités d'être erronée dans l'analyse.

This is an amazing time for Cosmology. In the last 60 years astronomical observations and achievements of particle physics laid the foundation for the Big Bang model, letting Cosmology evolve from the myth to science. The Big Bang idea was originally proposed at the end of the ?20s with the discovery of the Universe expansion by Edwin Hubble. In the ?60s this hypothesis got consolidated with the observations of the Cosmic Microwave Background Radiation (CMB) and the measurement of light-element abundance. In the last 25 years a quantitative model was developed, the so called Standard Model of Cosmology or ?CDM model. The ?CDM model relies on two pillars of the XX century physics: Einstein?s theory of General Relativity, which describes gravity by means of curved spacetime, and the Standard Model of Particle Physics, the relativistic quantum field theory which describes the electromagnetic, weak, and strong interactions. The ?CDM model, in its fundamental aspects, can be easily explained. About 13.8 billion years ago, when the density and temperature were incredibly high, the Universe started inflating. The first moments are not yet understood since we do not know the physical laws governing this high-energy regime, but after some picoseconds the Universe cooled down to a temperature about 10 15 K. From this point on, General Relativity, the Standard Model of Particle Physics, and Thermodynamics let us predict what happened: unstable particles disappeared and the Universe became a hot plasma made by protons, neutrons, electrons, photons, neutrinos, and hypothetical dark matter particles. Within the first three minutes, nuclear reactions led to the formation of light nuclei, then the Universe continued expanding and cooling for 300 thousands years, when it became cold enough so that electrons and protons became bound to form electrically neutral hydrogen atoms. After this process, matter became transparent to light, which started to free-stream across the Universe, becoming what now is known as the Cosmic Microwave Background. From this moment, gravity got the upper hand and dark matter played a crucial role in the formation of the large scale structure of the Universe. After some billion years, while the Universe expansion looked to be slowing down, a new character enters the scene, the Dark Energy, giving a new impetus to the Universe expansion. This is the history of the Universe as we know it, from its very beginning, through its evolution and until the present configuration of the cosmic web with galaxies, stars, and planets: in one of them we are wondering about how all this happened. The ?CDM model gives a consistent explanation of all these observations: it well explains the hydrogen and helium abundances and predicts with great accuracy the CMB temperature fluctuations. The agreement between data and predictions is astonishing and confirms that we are capturing at least small sparks of the truth. Although its success, we are aware that the ?CDM model is an incomplete theory, with several points left unsettled. In particular, visible matter gives a contribution of about 5% to the Universe energetic budget and we explain the remaining part with the two unknown entities already mentioned, Dark Matter and Dark Energy. The former behaves like an invisible slow (cold) matter which is not predicted by the Standard Model of Particles, hence the name Cold Dark Matter. The latter, usually indicated by the letter ?, is even more mysterious as represents a fluid with negative pressure permeating the Universe: we have no solid ideas about its origin. In order to shed light on the dark side of the Universe, several experiments are being led in the world and more are coming. These experiments try to produce Dark Matter particles in accelerators or to detect them in underground detectors. We have not a convincing explanation of the Universe first moments. Although the inflation era, a very early period of extremely rapid expansion, is by now quite commonly accepted by the scientific community, we still need to characterize its mechanism in detail. Furthermore, as the astronomic datasets available grew up in quality and details, the cosmological community started to make high precision measurements using different kind of probes. Although the qualitative picture of the ?CDM has been confirmed, there are still several tensions among different measurements of cosmological parameters. In particular, there is a tension on the exact value of the parameter which quantifies the current expansion rate of the Universe, the Hubble constant. In order to understand these problems, the cosmological community is working along many directions. New theoretical tools are being developed, in order to sharpen our comprehension of the Cosmos and make accurate predictions. New experiments are being carried out and designed. In the next decades, a wealth of new observations will be available via new CMB temperature and polarization probes, gravitational wave observatories, underground dark matter and neutrino detectors, and large scale galaxy surveys. These are among the most prominent tools that will be available to study our Universe with unprecedented accuracy, and possibly unveil the nature of its dark components and/or new physics. In order to extract as much information as possible, we need to develop also the correct data analysis tools, identifying the most interesting observables and understanding how to correctly model summary statistics. One of the forthcoming galaxy survey is Euclid, a mission of the European Space Agency (ESA). It will be one the widest redshift survey ever performed, covering an area of 15000 square degrees, corresponding to one third of the sky, up to a redshift z ? 2.5. Euclid will comprise two main instruments, one to measure light in the visible band and another one dedicated to both photometric and spectroscopic measurements in the near-infrared band. The main cosmological probes are weak gravitational lensing and galaxy clustering. The former, also called Cosmic Shear, is the gravitational deviation of photon paths due to the gravitational interaction with matter along the line of sight and is measured from galaxy shapes. The latter is produced by the galaxy trend not to be randomly distributed in space, but rather to cluster because of gravity, building up a complicated cosmic large scale structure. The joint analysis of these two cosmological probes will put unprecedented constraints on cosmological parameters and sharpen our knowledge of the Cosmos. However, weak lensing and galaxy clustering do not capture all the information encoded in a galaxy survey and there are a number of further additional probes that can be used to extract additional information. Cosmic voids are among the most promising novel probes of the large scale structure of the Universe; in particular, they are sensitive to the effects of Dark Energy and neutrinos. In this thesis, for the first time in literature, I performed a forecast on the Euclid measurement errors of cosmological parameters using the cross-correlation between weak lensing and cosmic voids, based on state-of-the-art theory and N-body simulations. This work confirms that using cosmic voids we can push further the sensitivity of the Euclid mission.

International audience; We introduce a new approach to estimate the point spread function (PSF) field of an optical telescope by building a semi-parametric model of its wavefront error. This method is particularly advantageous because it does not require calibration observa- tions to recover the wavefront error and it naturally takes into account the chromaticity of the optical system. The model is end-to-end differentiable and relies on a diffraction operator that allows us to compute monochromatic PSFs from the wavefront information.; Nous présentons une nouvelle approche pour estimer le champ de la fonction d'étalement du point d'un télescope optique en construisant un modèle semi-paramétrique de son erreur de front d'onde. Cette méthode est particulièrement avantageuse car elle ne nécessite pas d'observations de calibration pour récupérer l'erreur de front d'onde et elle prend naturellement en compte la chromaticité du système optique. Le modèle est différentiable de bout en bout et s'appuie sur un opérateur de diffraction qui nous permet de calculer les fonction d'étalement du point monochromatiques à partir des informations du front d'onde.

We study the morphology of convergence maps by perturbatively reconstructing their Minkowski Functionals (MFs). We present a systematics study using a set of three generalised skew-spectra as a function of source redshift and smoothing angular scale. Using an approach based on pseudo-$S_{\ell}$s (PSL) we show how these spectra will allow reconstruction of MFs in the presence of an arbitrary mask and inhomogeneous noise in an unbiased way. Our theoretical predictions are based on a recently introduced fitting function to the bispectrum. We compare our results against state-of-the art numerical simulations and find an excellent agreement. The reconstruction can be carried out in a controlled manner as a function of angular harmonics $\ell$ and source redshift $z_s$ which allows for a greater handle on any possible sources of non-Gaussianity. Our method has the advantage of estimating the topology of convergence maps directly using shear data. We also study weak lensing convergence maps inferred from Cosmic Microwave Background (CMB) observations; and we find that, though less significant at low redshift, the post-Born corrections play an important role in any modelling of the non-Gaussianity of convergence maps at higher redshift. We also study the cross-correlations of estimates from different tomographic bins., 16 pages, 12 figures

To the best of our (the authors) knowledge this is the first set of joint spherical reconstructions of all (excluding the now released HSC observation patch) public weak lensing observational data. Maps included are: - Mask - Number density map - e1/e2 shear maps - Spherical Kaiser-Squires convergence estimate (both raw and with 25 arcmin smoothing) - DarkMapper (Hierarchical Bayesian Sparse optimisation) convergence estimate, {"references":["Price et al 2020 (arXiv:2004.07855)"]}

Gravitational lensing has been identified as a powerful tool to address fundamental problems in astrophysics at different scales, ranging from exoplanet identification to dark energy and dark matter characterization in cosmology. Image simulations have played a fundamental role in the realization of the full potential of gravitational lensing by providing a means to address needs such as systematic error characterization, pipeline testing, calibration analyses, code validation, and model development. We present a general overview of the generation and applications of image simulations in strong and weak gravitational lensing, Comment: Accepted for publication in Symmetry, MDPI, special Issue on Gravitational Lensing

We introduce the skew-spectrum statistic for weak lensing convergence $\kappa$ maps and test it against state-of-the-art high-resolution all-sky numerical simulations. We perform the analysis as a function of source redshift and smoothing angular scale for individual tomographic bins. We also analyse the cross-correlation between different tomographic bins. We compare the numerical results to fitting-functions used to model the bispectrum of the underlying density field as a function of redshift and scale. We derive a closed form expression for the skew-spectrum for gravity-induced secondary non-Gaussianity. We also compute the skew-spectrum for the projected $\kappa$ inferred from Cosmic Microwave Background (CMB) studies. As opposed to the low redshift case we find the post-Born corrections to be important in the modelling of the skew-spectrum for such studies. We show how the presence of a mask and noise can be incorporated in the estimation of a skew-spectrum., Comment: 16 pages, 11 figures

The coherent distortions in the observed shapes of distant galaxies, a consequence of the spacetime curvature induced by the intervening large-scale structure of the Universe, is an abundant reservoir of cosmological information. Via this phenomenon of weak gravitational lensing, and a number of other independent cosmological probes, the parameters of the standard model, ΛCDM, have been inferred, now with uncertainties approaching the per cent level. In this era of precision cosmology, however, we face new challenges. Elements of tension have emerged between the measurements of the cosmological parameters from lowand high-redshift probes, seemingly implying either a failure to account for all relevant systematics, or perhaps even an incompleteness in the ΛCDM paradigm. In this thesis, I develop novel methodologies in weak lensing, to enhance the cosmological information extracted from current and future data sets. In this pursuit, I adopt a three-pronged approach, combining new advances in theoretical modelling, cutting-edge numerical simulations and recent developments in machine learning. Applying this trinity of techniques to three distinct bodies of research, described below, I construct new routes to improving the constraining power of this cosmological probe. A notable shortfall of the standard two-point statistics conventionally used in weak lensing, is their inability to capture all of the information contained in the non-linear cosmological fields of the real Universe. In answer to this problem, I develop the use of “clipping” transformations, which suppress the signal from the highest density regions observed. I present the first “clipped” cosmic shear measurement using data from the Kilo-Degree Survey (KiDS-450), and employ a suite of numerical simulations to calibrate and explore the cosmological dependence of this novel statistic. I show that these transformations improve constraints on S8 = σ8(Ωm/0.3)0.5, where Ωm is the mass energy density and σ8 is the amplitude of matter density fluctuations, when used in combination with conventional, “unclipped” two-point statistics, by 17% in the case of the KiDS-450 data. Clipping is but one member of the non-Gaussian statistics family, which have great potential for improving cosmological constraints, but are reliant both on numerical simulations, and a robust means to interpolate the statistics measured in the simulations to arbitrary cosmologies for comparison to the data. In this thesis, I develop a general framework to facilitate this, by designing and training a Gaussian process emulator, employing Bayesian supervised machine learning, on the state-of-the-art cosmo-SLICS suite, consisting of 26 different wCDM cosmologies. I demonstrate that this emulator achieves per cent level interpolation accuracy, in turn yielding unprecedented precision in the estimation of non-Gaussian statistics. I subsequently show how the cosmo-SLICS emulator might be employed within a likelihood analysis to constrain the cosmology of next-generation lensing data using these non-standard statistical probes. Taking clipped shear correlation functions as an example, I find that the low levels of noise present in the cosmo-SLICS emulator’s predictions facilitate improved constraints on cosmological parameters when the clipped and unclipped two-point probes are combined, not only for S8, but also for Ωm, and the Hubble and dark energy equation of state parameters, by 18%-26%. Finally, I combine the emulator approach with recent progress in theoretical modelling, to create a comprehensive framework for accurately predicting the non-linear matter power spectrum in arbitrary models of cosmology. This requires only a suite of vanilla ΛCDM N-body simulations with their initial conditions suitably tailored, such that the late-time non-linear power spectrum deviates from the standard model within a range permitted by observational constraints. These “pseudo” power spectra serve as the training set for the emulator, the predictions from which can be rescaled by reaction functions, analytically computed from the halo model, to obtain per cent level accurate non-linear predictions in a broadclass of beyond-ΛCDM cosmologies. In this proof-of-concept analysis, with a halofit training set substituting the simulation suite, I find that the emulator recovers the power spectra corresponding to f(R) gravity, massive neutrino cosmologies, combinations thereof, and even artificially generated departures from the ΛCDM prediction, with errors ≲ 1% deep in the highly non-linear regime. This work thus demonstrates a flexible and powerful method to not only test the validity of the standard model in the non-linear regime with next-generation cosmological data, but to also limit our reliance on costly numerical simulations in the future.

As light propagates through the Universe, its path is altered by the presence of massive objects. This causes a distortion of the images of distant galaxies. Measuring this effect, called weak gravitational lensing, allows us to probe the large scale structure of the Universe. This makes it a powerful source of cosmological insight, and can in particular be used to study the distribution of dark matter and the nature of Dark Energy. The European Space Agency’s upcoming Euclid mission is a spaceborne telescope with weak lensing as one of its primary science objectives.In practice, the weak lensing signal is recovered from the measurement of the shapes of galaxies. The images obtained by any optical instrument are altered by its Point Spread Function (PSF), caused by various effects: diffraction, imperfect optics, atmospheric turbulence (for ground-based telescopes)… Since the PSF also alters galaxy shapes, it is crucial to correct for it when performing weak lensing measurements. This, in turn, requires precise knowledge of the PSF itself.The PSF varies depending on the position of objects within the instrument’s focal plane. Unresolved stars in the field provide a measurement of the PSF at given positions, from which a PSF model can be built. In the case of Euclid, star images will suffer from undersampling. The PSF model will thus need to perform a super-resolution step. In addition, because of the very wide band of its visible instrument, variations of the PSF with the wavelength of incoming light will also need to be accounted for.The main contribution of this thesis is the building of novel PSF modelling approaches. These rely on sparsity and numerical optimal transport. The latter enables us to propose the first method capable of building a polychromatic PSF model, using no information other than undersampled star images, their position and spectra. We also study the propagation of errors in the PSF to the measurement of galaxy shapes., Le chemin parcouru par la lumière, lors de sa propagation dans l’Univers, est altéré par la présence d’objets massifs. Cela entraine une déformation des images de galaxies lointaines. La mesure de cet effet, dit de lentille gravitationnelle faible, nous permet de sonder la structure, aux grandes échelles, de notre Univers. En particulier, nous pouvons ainsi étudier la distribution de la matière noire et les propriétés de l’Energie Sombre, proposée comme origine de l’accélération de l’expansion de l’Univers. L’étude de l'effet de lentille gravitationnelle faible constitue l’un des objectifs scientifiques principaux d'Euclid, un télescope spatial de l’Agence Spatiale Européenne en cours de construction.En pratique, ce signal est obtenu en mesurant la forme des galaxies. Toute image produite par un instrument optique est altérée par sa fonction d’étalement du point (PSF). Celle-ci a diverses origines : diffraction, imperfections dans les composantes optiques de l’instrument, effets atmosphériques (pour les télescopes au sol)… Puisque la PSF affecte aussi les formes des galaxies, il est crucial de la prendre en compte lorsque l’on étudie l’effet de lentille gravitationnelle faible, ce qui nécessite de très bien connaître la PSF elle-même.Celle-ci varie en fonction de la position dans le plan focal. Une mesure de la PSF, à certaines positions, est donnée par l’observation d’étoiles non-résolues dans le champ, à partir desquelles on peut construire un modèle de PSF. Dans le cas d’Euclid, ces images d’étoiles seront sous-échantillonnée ; aussi le modèle de PSF devra-t-il contenir une étape de super-résolution. En raison de la très large bande d’intégration de l’imageur visible d’Euclid, il sera également nécessaire de capturer les variations en longueur d’onde de la PSF.La contribution principale de cette thèse consiste en le développement de méthodes novatrices d’estimation de la PSF, reposant sur plusieurs outils : la notion de représentation parcimonieuse, et le transport optimal numérique. Ce dernier nous permet de proposer la première méthode capable de fournir un modèle polychromatique de la PSF, construit uniquement à partir d’images sous-échantillonnées d’étoiles et leur spectre. Une étude de la propagation des erreurs de PSF sur la mesure de forme de galaxies est également proposée.

Square Kilometre Array Cosmology Science Working Group: et al., arXiv:1811.02743v1, We present a detailed overview of the cosmological surveys that we aim to carry out with Phase 1 of the Square Kilometre Array (SKA1) and the science that they will enable. We highlight three main surveys: a medium-deep continuum weak lensing and low-redshift spectroscopic HI galaxy survey over 5 000 deg2; a wide and deep continuum galaxy and HI intensity mapping (IM) survey over 20 000 deg2 from to 3; and a deep, high-redshift HI IM survey over 100 deg2 from to 6. Taken together, these surveys will achieve an array of important scientific goals: measuring the equation of state of dark energy out to with percent-level precision measurements of the cosmic expansion rate; constraining possible deviations from General Relativity on cosmological scales by measuring the growth rate of structure through multiple independent methods; mapping the structure of the Universe on the largest accessible scales, thus constraining fundamental properties such as isotropy, homogeneity, and non-Gaussianity; and measuring the HI density and bias out to z = 6. These surveys will also provide highly complementary clustering and weak lensing measurements that have independent systematic uncertainties to those of optical and near-infrared (NIR) surveys like Euclid, LSST, and WFIRST leading to a multitude of synergies that can improve constraints significantly beyond what optical or radio surveys can achieve on their own. This document, the 2018 Red Book, provides reference technical specifications, cosmological parameter forecasts, and an overview of relevant systematic effects for the three key surveys and will be regularly updated by the Cosmology Science Working Group in the run up to start of operations and the Key Science Programme of SKA1., RB, CD, and SH acknowledge support from an STFC Consolidated Grant (ST/P000649/1). SC is supported by the Italian Ministry of Education, University and Research (MIUR) through Rita Levi Montalcini project ‘PROMETHEUS—Probing and Relating Observables with Multi-wavelength Experiments To Help Enlightening the Universe’s Structure’, and by the ‘Departments of Excellence 2018-2022’ Grant awarded by MIUR (L. 232/2016). PGF acknowledges support from ERC Grant No: 693024, the Beecroft Trust and STFC. RM and MGS acknowledge support from the South African Square Kilometre Array Project and National Research Foundation (Grant Nos. 75415 and 84156). LW is supported by an ARC Discovery Early Career Researcher Award (DE170100356). YA acknowledges support from the Netherlands Organization for Scientific Research (NWO) and the Dutch Ministry of Education, Culture and Science (OCW), and also from the D-ITP consortium, a programme of the NWO that is funded by the OCW. YA is also supported by LabEx ENS-ICFP: ANR-10-LABX-0010/ANR-10-IDEX0001-02 PSL∗. HP’s research is supported by the Tomalla Foundation. AR has received funding from the People Programme (Marie Curie Actions) of the European Union H2020 Programme under REA grant agreement number 706896 (COSMOFLAGS). TS and DJS gratefully acknowledge support from the Deutsche Forschungsgemeinschaft (DFG) within the Research Training Group 1620 ‘Models of Gravity’.

Accurate shape measurements are essential to infer cosmological parameters from large area weak gravitational lensing studies. The compact diffraction-limited point-spread function (PSF) in space-based observations is greatly beneficial, but its chromaticity for a broad band observation can lead to new subtle effects that could hitherto be ignored: the PSF of a galaxy is no longer uniquely defined and spatial variations in the colours of galaxies result in biases in the inferred lensing signal. Taking Euclid as a reference, we show that this colourgradient bias (CG bias) can be quantified with high accuracy using available multi-colour Hubble Space Telescope (HST) data. In particular we study how noise in the HST observations might impact such measurements and find this to be negligible. We determine the CG bias using HST observations in the F606W and F814W filters and observe a correlation with the colour, in line with expectations, whereas the dependence with redshift is weak. The biases for individual galaxies are generally well below 1%, which may be reduced further using morphological information from the Euclid data. Our results demonstrate that CG bias should not be ignored, but it is possible to determine its amplitude with sufficient precision, so that it will not significantly bias the weak lensing measurements using Euclid data., 13 pages, 14 figures, MNRAS accepted, comments welcome

The unprecedented quality, the increased dataset, and the wide area of ongoing and near future weak lensing surveys allows to move beyond the standard two points statistics thus making worthwhile to investigate higher order probes. As an interesting step towards this direction, we expolore the use of higher order moments (HOM) of the convergence field as a way to increase the lensing Figure of Merit (FoM). To this end, we rely on simulated convergence to first show that HOM can be measured and calibrated so that it is indeed possible to predict them for a given cosmological model provided suitable nuisance parameters are introduced and then marginalized over. We then forecast the accuracy on cosmological parameters from the use of HOM alone and in combination with standard shear power spectra tomography. It turns out that HOM allow to break some common degeneracies thus significantly boosting the overall FoM. We also qualitatively discuss possible systematics and how they can be dealt with., Comment: 25 pages, 5 figures, This paper extends and replaces arXiv:1606.03892. It matches PRD accepted version after minor changes

Cosmic shear is one of the primary probes to test gravity with current and future surveys. There are two main techniques to analyse a cosmic shear survey; a tomographic method, where correlations between the lensing signal in different redshift bins are used to recover redshift information, and a 3D approach, where the full redshift information is carried through the entire analysis. Here we compare the two methods, by forecasting cosmological constraints for future surveys like Euclid. We extend the 3D formalism for the first time to theories beyond the standard model, belonging to the Horndeski class. This includes the majority of universally coupled extensions to $\Lambda$CDM with one scalar degree of freedom in addition to the metric, still in agreement with current observations. Given a fixed background, the evolution of linear perturbations in Horndeski gravity is described by a set of four functions of time only. We model their time evolution assuming proportionality to the dark energy density fraction and place Fisher matrix constraints on the proportionality coefficients. We find that a 3D analysis can constrain Horndeski theories better than a tomographic one, in particular with a decrease in the errors of the order of 20$\%$. This paper shows for the first time a quantitative comparison on an equal footing between Fisher matrix forecasts for both a fully 3D and a tomographic analysis of cosmic shear surveys. The increased sensitivity of the 3D formalism comes from its ability to retain information on the source redshifts along the entire analysis., Comment: 15 pages, 10 figures, matches published MNRAS version

Galaxy clusters are the most massive collapsed structures in the Universe and their properties offer a crucial insight into the formation of structure. High quality observational data is forthcoming with ongoing and upcoming surveys, but simulations are needed to provide robust theoretical predictions for comparison, as well mock data for testing observational techniques. Numerical simulations are now able to accurately model a range of astrophysical processes. This is highlighted in the BAHAMAS and MACSIS simulations, which have successfully reproduced the observed scaling relations of galaxy clusters. We use these simulations to quantify the impact baryons have on the mass distribution within galaxy clusters, as well as the bias in X-ray and weak lensing mass estimates. It is shown that baryons have only a minor affect on the spins, shape and density profiles of galaxy clusters and they have no significant impact on the bias in weak lensing mass estimates. When using spectroscopic temperatures and densities, the X-ray hydrostatic mass bias decreases as a function of mass, leading to a bias of ~40% for clusters with M_500 > 10^15 solar masses. In the penultimate chapter, we use the EAGLE and C-EAGLE simulations to construct more realistic mock cluster observations. The EAGLE simulations have been shown to successfully reproduce the properties of field galaxies and they are complemented by the C-EAGLE project, which extends this work to the cluster scale. We use these simulations to construct a cluster lightcone that accounts for the impact of uncorrelated large scale structure on cluster observables, including weak lensing mass estimates, the Sunyaev-Zel'dovich parameter and X-ray luminosity.

Galaxy clusters are essential cosmological and astrophysical tools, since they represent the largest and most massive gravitationally bound structures in the Universe. Through the study of their mass function, of their correlation function, and of the scaling relations between their mass and different observables, we can probe the predictions of cosmological models and structure formation scenarios. They are also interesting laboratories that allow us to study galaxy formation and evolution, and their interactions with the intra-cluster medium, in dense environments. For all of these goals, an accurate estimate of cluster masses is of fundamental importance. I studied the accuracy of the optical richness obtained by the RedGOLD cluster detection algorithm (Licitra et al. 2016) as a mass proxy, using weak lensing and X-ray mass measurements. I measured stacked weak lensing cluster masses for a sample of 1323 galaxy clusters in the CFHTLS W1 and in the NGVS at 0.2