| Title: |
Découvrir l'univers local ; Unveil the Local Universe |
| Authors: |
Valade, Aurélien |
| Contributors: |
Institut de Physique des 2 Infinis de Lyon (IP2I Lyon); Université Claude Bernard Lyon 1 (UCBL); Université de Lyon-Université de Lyon-Institut National de Physique Nucléaire et de Physique des Particules du CNRS (IN2P3)-Centre National de la Recherche Scientifique (CNRS); Université Claude Bernard - Lyon I; Universität Potsdam; Anne Ealet; Matthias Steinmetz; Noam Libeskind |
| Source: |
https://theses.hal.science/tel-04717778 ; Cosmology and Extra-Galactic Astrophysics [astro-ph.CO]. Université Claude Bernard - Lyon I; Universität Potsdam, 2023. English. ⟨NNT : 2023LYO10089⟩. |
| Publisher Information: |
CCSD |
| Publication Year: |
2023 |
| Collection: |
HAL Lyon 1 (University Claude Bernard Lyon 1) |
| Subject Terms: |
Large scale structure; Dark Matter; Cosmology; Structure à grande échelle; Matière noire; Cosmologie; [PHYS.ASTR.CO]Physics [physics]/Astrophysics [astro-ph]/Cosmology and Extra-Galactic Astrophysics [astro-ph.CO] |
| Description: |
Galaxies in the Universe form a gigantic, complex edifice, called the Large Scale Structure (LSS). Still, the vast majority of the matter is thought to be dark, i.e.not directly observable by our telescopes and detectable solely through its gravitational interaction with its surrounding. The relationship between the distribution of galaxies and the total matter field is still not fully described and thus, the LSS cannot be reduced to the galaxies that inhabit it. Unveiling the matter distribution and the associated velocity field in the Local Universe is an ex- tremely difficult task. One approach consists in combining, for each galaxy, a measurement of redshift and estimation of distance to obtain its velocity with respect to its local environment. With the only source of motion on these scales being gravitation, and using the fact that the velocity field is tightly linked to the matter distribution, the two can be reconstructed together. Yet, estimations of distances, and thus of velocities, are difficult to make: data are sparse, tainted with errors and plagued with observational biases. Only the radial component of the velocity can be measured and the error size grows with the distance. Powerful mathematical methods need thus be employed. Our method follows the Bayesian inference approach developed in the last decade to over-come the short-falling of the Wiener Filter methodology, whose simplistic modeling of the data requires a somewhat ad-hoc treatment of the data beforehand. The first step in Bayesian inference is the description of the conditional probability of a set of parameters of a given model given a set of observations. The second step is the creation of a series of realizations of this probability law with a Monte Carlo method, on which summary statistics can be computed. However, this process is computationally very costly, and the previously developed methods were unable to face the growing size of the current and future problems. This work answers this issue thanks to two major innovations. ... |
| Document Type: |
doctoral or postdoctoral thesis |
| Language: |
English |
| Relation: |
NNT: 2023LYO10089 |
| Availability: |
https://theses.hal.science/tel-04717778; https://theses.hal.science/tel-04717778v1/document; https://theses.hal.science/tel-04717778v1/file/TH2023VALADEAURELIEN.pdf |
| Rights: |
https://about.hal.science/hal-authorisation-v1/ ; info:eu-repo/semantics/OpenAccess |
| Accession Number: |
edsbas.328B8940 |
| Database: |
BASE |