The work presented in this thesis focuses on developing compression techniques to exploit fully the constraining power of high-order statistics when applied to the cosmological observable of interest. We present four different methods in the three-point (3pt) case. The mathematical theoretical framework is first de- veloped and then followed, for all the methods, by application on real data. In particular we use data from the CMASS sample of the Sloan Digital Sky Survey III BOSS Data Releases 11 and 12. Our compression results are compared to those obtained via standard analysis, for example Markov chain Monte Carlo (MCMC) sampling. First, we consider the three-point auto-correlation function as an integrated compressed version of the standard correlation one. We derive analytic expres- sions including corrections for the Primordial non-Gaussianity. We then test the model on data to constrain cosmological parameters. Secondly, we explore two methods of compressing the redshift-space galaxy power spectrum and bispectrum with respect to a chosen set of cosmological parameters. Both methods transform the original data-vector into a compressed one with dimension equal to the number of model parameters considered using the Multiple Optimised Parameter Estimation and Data compression algorithm (MOPED) algorithm. Analytic expressions for the covariance matrix are derived in order both to compress the data-vector and to test the compression perfor- mance by comparing with standard MCMC sampling on the full data-vector. Finally, we apply our compression methods to the galaxy power spectrum monopole, quadrupole and bispectrum monopole measurements from the BOSS DR12 CMASS sample. We derive an analytic expression for the covariance ma- trix of the new data-vector. We show that compression allows a much longer data-vector to be used, returning tighter constraints on the cosmological param- eters of interest.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:756284 |
| Date | January 2018 |
| Creators | Gualdi, Davide |
| Publisher | University College London (University of London) |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://discovery.ucl.ac.uk/10054949/ |
Page generated in 0.0179 seconds