Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.539376
Title: Detecting large-scale structure in the era of petabyte/gigaparsec astronomy
In this thesis, we present a study of the identification of large-scale structure in optical astronomical surveys. This encompasses the detection of large connected structures of alaxies in spectroscopic datasets and of galaxy clusters in deep photometric surveys. Beginning with a survey featuring full 3D galaxy data, in chapter 2 we present a method to identify filamentary structure after accounting for the line-of-sight velocity distortions characteristic of the virialised systems we search for. We compare data from a real galaxy survey to a series of realistic mocks. Despite broad similarities between the two, we find models do not reproduce the argest observed structures. To evaluate the exploration of a multi-band survey lacking spectroscopy, we simulate the effects of photometric redshift uncertainties on galaxy redshifts. Our findings provide limits on the accuracy of photometric redshift estimators required to recover the same diverse range of structures detected in the original spectroscopic survey. As an alternative means of exploiting the deep multi-band photometric data common to wide-area observational campaigns, in chapter 3 we present a red sequence-based algorithm to detect galaxy clusters with Voronoi diagrams. This algorithm makes no prior assumptions about cluster properties other than the similarity in colour of their members, and an enhanced projected surface density. Testing the algorithm with mock galaxy survey data reveals a detection performance equalling or exceeding that of alternative detection algorithms. Chapter 4 describes the application of this algorithm to a $270{\rm deg^2}$ survey with deep SDSS photometry. The scientific exploitation of $4,000\ {\rm z}\ \leq\ 0.6$ cluster detections from this survey is ongoing, but work presented here shows: agreement with the red sequence slope evolution derived from semi-analytic galaxy models, evidence stellar age is not responsible for responsible for the sequence slope, and a well-defined colour-colour track of potential use in photometric cluster redshift stimation. We detail improvements made to the cluster algorithm in chapter 5. Through a series of case studies we verify our approach successfully identifies galaxy clusters in a diverse range of surveys, from volumes spanning $2h^{-3}{\rm Gpc}^{3}$ to deep near-IR detections at ${\rm z}\sim 1$. based on our findings, we expect the Pan-STARRS $3\pi$ large-area survey to identify over $10^{5}$ clusters and groups. In chapter 6, we explore the characteristics of randomly-distributed noise in Voronoi diagrams. We verify the model traditionally used to describe the distribution of Voronoi cell areas in Poisson data fails to describe the frequency of high-density random cells. Because high-density cells resemble those expected from a population of galaxy cluster members, using a large dataset generated in this study we propose an alternative model that better estimates the frequency of their areas. This new model may in the future be used to improve Voronoi-based recovery of clustered data in a diverse range of applications, both astronomical and otherwise.