Precision Cosmology with Inaccurate Data: Assessing and Addressing Systematic Errors in Large-Scale Structure
Weaverdyck, Noah
2021
Abstract
Cosmology as a field has made rapid progress over the last 30 years, driven in large part by a massive increase in observational data. Large-scale structure (LSS) surveys have played a key role in this rise of `Precision Cosmology', having grown from mapping the location of thousands of galaxies to the hundreds of millions. As statistical errors plummet, however, progress in the field increasingly hinges on our ability to model and control systematic errors to an exquisite degree. As statistical errors plummet, however, progress in the field increasingly hinges on our ability to model and control systematic errors to an exquisite degree. First we present a detailed study of the Integrated Sachs-Wolfe (ISW) effect and how errors in the large scale photometric calibration of LSS surveys impact estimation of the ISW signal. The ISW effect is an imprint of dark matter in the Cosmic Microwave Background and contains important information about dark energy, including possible signatures for modifications to General Relativity. We quantify the necessary levels of calibration to produce accurate reconstructions of the ISW map and power spectrum for next-generation surveys. We provide a roadmap for ISW reconstruction, including the optimization of survey configuration and an improved estimator to render the analysis more robust to calibration errors. Next, we perform a detailed study of the leading methods for removing spatially-dependent systematic errors in galaxy surveys, such as those induced by interstellar dust, variable atmospheric conditions, and other effects that modulate the observed number of galaxies across the sky. We recast them into a common statistical framework, elucidating assumptions implicit within each method and characterize their performance on a suite of simulations. We propose extensions to current methods that are more robust, simpler to implement, and exhibit greater suppression of systematic errors. We further derive uncertainty estimates for the galaxy-level corrections, enabling the propagation of errors from the correction methods into the LSS galaxy catalogs and any subsequent analyses that use them. The final portion of this thesis focuses on small scale systematic errors in LSS analyses, such as arise from theoretical uncertainties in the non-linear growth of dark matter, baryonic effects, and other astrophysical phenomena. We characterize how errors in modeling such small scales impact our ability to accurately infer the primordial power spectrum of curvature fluctuations, which initially seeded structure in the early Universe. We show that if unaccounted for, current and predicted modeling uncertainties can strongly bias measurements of the ``runnings" of the spectral index, key parameters for testing single-field slow-roll models of inflation, thought to be responsible for the rapid, early expansion of the Universe. We compare methods designed to mitigate such small-scale systematic errors and demonstrate that, even with optimistic improvements in small scale modeling, only exotic models of inflation will be testable via constraints on the runnings from near-future LSS surveys. These three studies represent important steps for continued progress in the field and towards ensuring that analyses of large-scale structure are robust and accurate in the era of Precision Cosmology.Deep Blue DOI
Subjects
Cosmology Large-scale structure Statistical methods Inflation Systematic errors Data analysis
Types
Thesis
Metadata
Show full item recordCollections
Remediation of Harmful Language
The University of Michigan Library aims to describe library materials in a way that respects the people and communities who create, use, and are represented in our collections. Report harmful or offensive language in catalog records, finding aids, or elsewhere in our collections anonymously through our metadata feedback form. More information at Remediation of Harmful Language.
Accessibility
If you are unable to use this file in its current format, please select the Contact Us link and we can modify it to make it more accessible to you.