Von Wietersheim-Kramsta, Maximilian;
(2024)
Forward-Simulations of Large-Scale Structure for Cosmological Inference.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
Von Wietersheim-Kramsta_10185306_thesis_revised.pdf Download (12MB) | Preview |
Abstract
Measuring the density of fluctuations in the large-scale structure of the Universe has become a powerful tool in constraining and testing cosmological models. Current and upcoming galaxy surveys are pushing the precision limits of current physical models, while measuring unprecedented amounts of data. This thesis presents statistical simulations of large-scale structure and inference techniques for estimating cosmological parameters and calibrating systematic effects from galaxy survey data. First, I present a novel method to calibrate magnification bias observed within galaxy clustering and weak gravitational lensing measurements regardless of the selection applied to the galaxy data. This method addresses the need for estimating this systematic within the Kilo-Degree Survey’s (KiDS) cosmological analysis. Secondly, I show a suite of statistical forward-simulations of large-scale structure which is designed to model galaxy survey observations, while including relevant physical and observational biases. These simulations create realistic catalogues of galaxy observations based on underlying matter density fields consistent with a given cosmological model. Next, I describe how these simulations are used to conduct the first Bayesian simulation-based inference (SBI) of cosmological parameters from weak lensing data from KiDS at a similar precision as standard analyses. This SBI analysis allows dropping the common assumption of a Gaussian likelihood, fully propagating uncertainty from data to parameter posteriors at a comparable computational cost as standard weak gravitational lensing analysis. Thus, this may facilitate the efficient extraction of information from surveys such as Euclid or the Vera Rubin Observatory. Lastly, I use an altered form of the forward-simulations to test signal and uncertainty modelling conducted for KiDS’s analysis. This shows the sufficiency of analytical modelling when considering systematics such as variations in observational depth. In summary, in this thesis, I present novel techniques to estimate systematic uncertainties in inferring cosmological parameters from galaxy surveys and I show how fast realistic forward-simulations may be used for SBI and model testing in current and future surveys.
| Type: | Thesis (Doctoral) |
|---|---|
| Qualification: | Ph.D |
| Title: | Forward-Simulations of Large-Scale Structure for Cosmological Inference |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| Additional information: | Copyright © The Author 2024. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
| Keywords: | cosmology, physics, astronomy, astrophysics, bayesian, machine learning |
| UCL classification: | UCL UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Maths and Physical Sciences > Dept of Physics and Astronomy |
| URI: | https://discovery.ucl.ac.uk/id/eprint/10185306 |
Archive Staff Only
 |
View Item |
