Exploring cosmology and structure formation via high-z galaxies

• Main Author: Nikoloudakis, Nikolaos
• Published: Durham University 2013
• Subjects: 530
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.586084

This thesis exploits the large-scale structure of the Universe via observations over a wide redshift range, with the aim of constraining the current cosmological models and galaxy formation physics. We present the eXtreme Multiplex Spectrograph (XMS), a proposed spectrograph that can map simultaneously 4000 Emission Line Galaxies (ELGs) and Luminous Red Galaxies (LRGs) in the range 0.4 ≤ z ≤ 0.8, with a success rate of 88%. Figures of merit clearly indicate that XMS is better or even competitive compared to future surveys for measurements of the gravitational growth rate, Baryon Acoustic Oscillations (BAO) and dark-matter halo mass function. Next, by selecting a unique photometric sample of 130, 000 LRGs in the Sloan Digital Sky Survey (SDSS) Stripe 82, with an estimated average redshift z ∼ 1, we perform a clustering analysis and compare the clustering evolution of the high-z Stripe 82 LRGs to lower-z LRGs. An immediate feature of the Stripe 82 LRGs clustering is a power excess at large scales. This behaviour is not expected within the ΛCDM model, making the conclusion of a slow clustering evolution as observed for the lower-z LRGs, non-trivial. Only Non-Gaussian models are able to describe the large scale clustering of the Stripe 82 LRGs. From follow up spectroscopic observations of a subsample of the Stripe 82 LRGs, we confirm that the average red- shift of our sample is z ∼ 0.9, while the slow clustering evolution of the LRGs is now slightly more favoured. However, Non-Gaussianity is still detected at a 2σ level. Finally, from the largest and deepest near-infrared field to present, the UltraVISTA survey, we select ≈ 4000 Distant Red Galaxies (DRGs) and study the largest galaxy separations ever probed with these massive galaxies. In agreement with previous results, UltraVISTA DRGs are strongly clustered objects. Furthermore, they show stronger clustering within their brighter K-limited samples, that could possibly imply luminosity segregation. Their connection to the local descendants cannot yet be established.