Thesis title
Galaxy Clusters as Cosmological Probes : Halo Mass Function Systematics, Extreme Value Statistics, and Impact of Clustering Dark Energy
Composition of the jury
David ELBAZ, Université Paris Cité, examiner
Jochen WELLER, referee
Etienne POINTECOUTEAU, Université de Toulouse, referee
Sunayana BHARGAVA, Université de la Côte d’Azur, examiner
Jenny SORCE, Université de Lille, examiner
Linda BLOT, Université de Tokyo, invited member
Pier-Stefano CORASANITI, Observatoire de Paris, thesis supervisor
Summary
Galaxy clusters constitute a powerful probe for testing and constraining theoretical cosmological models. They result from the cosmological evolution of the highest matter-density peaks in the primordial density field.
Given the complexity of the physical and cosmological processes involved in their formation and evolution, galaxy clusters are subject to numerous systematic biases, both observational and theoretical. Within the framework of the Euclid space mission, we conducted an in-depth study of a theoretical bias related to the modelling of their dark matter distribution and its impact on cosmological parameter constraints. Using high-resolution numerical simulations, we showed that assuming a deterministic mass profile can introduce a bias in cosmological inference based on cluster number counts. Indeed, observed mass distributions can systematically deviate from a prescribed parametric form and, consequently, adopting a non-parametric and stochastic approach can help mitigate this type of bias.
We then focused on the cosmological properties of the most massive galaxy clusters by adapting a statistical framework emerging from extreme value theory to their study. We investigated the statistical relationships between their masses ranked in ascending order. This analysis revealed a connection between the halo mass function and the universal probability distribution of extreme values, as well as a new statistical approach based on the magnitude between two mass ranks. This framework enables us to test the consistency of a given cosmological model while accounting for and mitigating systematic errors in observed masses. Using this formalism, we were able to assess the accuracy of cluster mass measurements from the Planck satellite and their consistency with the LCDM model.
Finally, we contributed to the development and scientific exploitation of the first numerical simulation of inhomogeneous dark energy modelled as a fluid, using the newly developed Nefertiti code. This simulation enabled us to investigate the impact of dark energy non-linearities across different scales and on different probes. We showed that inhomogeneous dark energy models have a significant influence on the nonlinear regime of structure formation and leave a detectable signature on the scale of massive clusters, for example, in their density profiles and masses.
