Logo Logo
Help
Contact
Switch language to German
Cosmology with clusters of galaxies in the eROSITA era
Cosmology with clusters of galaxies in the eROSITA era
The nature of dark matter and dark energy is one of the most intriguing scientific questions of this century. Tiny density perturbations in the early Universe evolved under the action of gravity, growing into the structures we see nowadays, such as galaxies and clusters. Such growth throughout cosmic time depends on the nature and abundance of these mysterious dark components. Clusters of galaxies provide a great tool for studying the cosmological evolution of the Universe. Galaxy clusters are the most massive virialized objects in the Universe, they reside in the nodes of the cosmic web, and are a direct tracer of the large scale structure of the Universe. Their abundance as a function of mass and redshift (the halo mass function) provides a stringent test for different cosmological models and tight constraints on the total amount of matter and the amplitude of the density perturbations in the Universe, but it also depends on the nature of dark energy. The eROSITA X-ray telescope will provide the largest X-ray-selected sample of galaxy clusters ever, with the potential of studying dark matter and dark energy with unprecedented precision. A detailed understanding of the sources detected by eROSITA and the uncertainties in the data is fundamental to reach this goal. The first aim of this thesis is to produce and study a digital twin of the first eROSITA all-sky survey (eRASS1), to characterize the source catalog and test the science pipelines. We use state of the art models to predict the X-ray emission from clusters of galaxies and AGN, and model the X-ray background using real data as a starting point. With this setup, we run an X-ray software simulator to produce mock detected photons, accounting for technical details such as the instrument response and the observation strategy. We run the eROSITA Standard Analysis Software System (eSASS), which produces a source catalog with very similar properties to the one obtained from real data. We match the input and output catalogs with an accurate algorithm based on the origin of each photon. We detect all the brightest clusters and AGN in the simulation. The fraction of detected sources primarily depends on flux and exposure time. Secondary effects, such as the source size and the central emissivity, are relevant for clusters. We provide a detailed study of the trade-off between completeness and purity. We find that progressive thresholds of detection likelihood get rid of the background fluctuations in the source catalog, while cuts in extension likelihood are necessary to remove bright point sources contaminating the cluster sample. We discuss different selections according to a given science goal. For example, the inclusion of the majority of sources in the sample is best, if one wants to find interesting objects to study the astrophysics of faint clusters and groups. Instead, a more secure cluster sample with low contamination is required for a cosmological experiment. Once the cluster sample is selected, accurate halo mass function models are key to reaching precise constraints on cosmological models. The second aim of this thesis is to calibrate a model of the mass function that also includes variables related to the dynamical state of dark matter halos. We use the dark matter-only MultiDark suite of simulations and the high-mass objects hosting clusters of galaxies therein. We measure the mean relations of concentration, offset parameter (Xoff ), and spin as a function of dark matter halo mass and redshift. We investigate the distributions around the mean relations. We confirm the recent discovery of the concentration upturn at high masses and provide a model that predicts the concentration for different values of mass and redshift with one single equation. We find that the concentration of low-mass halos shows a faster redshift evolution compared to high-mass halos, especially in the high-concentration regime. We find that the offset parameter is systematically smaller at low redshift, in agreement with the relaxation of structures at recent times. The individual models are combined into a comprehensive framework, which predicts the mass function as a function of spin and offset. Our model recovers the fiducial mass function with great accuracy at redshift 0 and accounts for redshift evolution up to z ∼ 1.5. The generalized mass function framework allows marginalizing over selection effects related to the dynamical state of dark matter halos in a cluster count experiment. However, a link between observations and theoretical models is lacking. The third goal of this thesis is to study the dynamical state of clusters detected by eROSITA using the offset between the X-ray and the optical centers. We aim to connect the offset measured in eROSITA observations to predictions by hydrodynamical simulations and N-body models, assessing the astrophysical effects affecting the displacements. We measure the offset for clusters observed in the eROSITA Final Equatorial-Depth Survey (eFEDS) and eRASS1. We focus on a subsample of 87 massive eFEDS clusters at low redshift. We compare the displacements in this sample to those predicted by the TNG and the Magneticum simulations. We link the observations to the offset parameter Xoff measured for dark matter halos in N-body simulations, using the hydrodynamical simulations as a bridge. We find that on average the eFEDS clusters show a smaller offset compared to eRASS1 because the latter contains a larger fraction of massive and disturbed structures. The offset measured in the eFEDS subsample is in agreement with the predictions from TNG and Magneticum, and the distribution of the offset parameter from dark matter only simulations. However, the tails of the distributions are different. Baryonic effects cause a decrement (increment) in the low (high) offset regime compared to the Xoff distribution from dark matter-only simulations. Finally, we find a correlation between the offset predicted by hydro simulations and Xoff measured in their parent dark matter-only run and calibrate a relation between them, which allows us to recover the full Xoff distribution with excellent precision. The work developed in this thesis is essential to characterize the real eRASS1 sample, hat will soon be available to the public. The development of the innovative mass function framework and its link to data with the offset between different definitions of the cluster center will allow the minimization of uncertainties in cluster count experiments due to selection effects related to the cluster dynamical state.
eROSITA, galaxy clusters, cosmology, astrophysics
Seppi, Riccardo
2023
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Seppi, Riccardo (2023): Cosmology with clusters of galaxies in the eROSITA era. Dissertation, LMU München: Faculty of Physics
[thumbnail of Seppi_Riccardo.pdf] Licence: Creative Commons: Attribution 4.0 (CC-BY)
PDF
Seppi_Riccardo.pdf

36MB

Abstract

The nature of dark matter and dark energy is one of the most intriguing scientific questions of this century. Tiny density perturbations in the early Universe evolved under the action of gravity, growing into the structures we see nowadays, such as galaxies and clusters. Such growth throughout cosmic time depends on the nature and abundance of these mysterious dark components. Clusters of galaxies provide a great tool for studying the cosmological evolution of the Universe. Galaxy clusters are the most massive virialized objects in the Universe, they reside in the nodes of the cosmic web, and are a direct tracer of the large scale structure of the Universe. Their abundance as a function of mass and redshift (the halo mass function) provides a stringent test for different cosmological models and tight constraints on the total amount of matter and the amplitude of the density perturbations in the Universe, but it also depends on the nature of dark energy. The eROSITA X-ray telescope will provide the largest X-ray-selected sample of galaxy clusters ever, with the potential of studying dark matter and dark energy with unprecedented precision. A detailed understanding of the sources detected by eROSITA and the uncertainties in the data is fundamental to reach this goal. The first aim of this thesis is to produce and study a digital twin of the first eROSITA all-sky survey (eRASS1), to characterize the source catalog and test the science pipelines. We use state of the art models to predict the X-ray emission from clusters of galaxies and AGN, and model the X-ray background using real data as a starting point. With this setup, we run an X-ray software simulator to produce mock detected photons, accounting for technical details such as the instrument response and the observation strategy. We run the eROSITA Standard Analysis Software System (eSASS), which produces a source catalog with very similar properties to the one obtained from real data. We match the input and output catalogs with an accurate algorithm based on the origin of each photon. We detect all the brightest clusters and AGN in the simulation. The fraction of detected sources primarily depends on flux and exposure time. Secondary effects, such as the source size and the central emissivity, are relevant for clusters. We provide a detailed study of the trade-off between completeness and purity. We find that progressive thresholds of detection likelihood get rid of the background fluctuations in the source catalog, while cuts in extension likelihood are necessary to remove bright point sources contaminating the cluster sample. We discuss different selections according to a given science goal. For example, the inclusion of the majority of sources in the sample is best, if one wants to find interesting objects to study the astrophysics of faint clusters and groups. Instead, a more secure cluster sample with low contamination is required for a cosmological experiment. Once the cluster sample is selected, accurate halo mass function models are key to reaching precise constraints on cosmological models. The second aim of this thesis is to calibrate a model of the mass function that also includes variables related to the dynamical state of dark matter halos. We use the dark matter-only MultiDark suite of simulations and the high-mass objects hosting clusters of galaxies therein. We measure the mean relations of concentration, offset parameter (Xoff ), and spin as a function of dark matter halo mass and redshift. We investigate the distributions around the mean relations. We confirm the recent discovery of the concentration upturn at high masses and provide a model that predicts the concentration for different values of mass and redshift with one single equation. We find that the concentration of low-mass halos shows a faster redshift evolution compared to high-mass halos, especially in the high-concentration regime. We find that the offset parameter is systematically smaller at low redshift, in agreement with the relaxation of structures at recent times. The individual models are combined into a comprehensive framework, which predicts the mass function as a function of spin and offset. Our model recovers the fiducial mass function with great accuracy at redshift 0 and accounts for redshift evolution up to z ∼ 1.5. The generalized mass function framework allows marginalizing over selection effects related to the dynamical state of dark matter halos in a cluster count experiment. However, a link between observations and theoretical models is lacking. The third goal of this thesis is to study the dynamical state of clusters detected by eROSITA using the offset between the X-ray and the optical centers. We aim to connect the offset measured in eROSITA observations to predictions by hydrodynamical simulations and N-body models, assessing the astrophysical effects affecting the displacements. We measure the offset for clusters observed in the eROSITA Final Equatorial-Depth Survey (eFEDS) and eRASS1. We focus on a subsample of 87 massive eFEDS clusters at low redshift. We compare the displacements in this sample to those predicted by the TNG and the Magneticum simulations. We link the observations to the offset parameter Xoff measured for dark matter halos in N-body simulations, using the hydrodynamical simulations as a bridge. We find that on average the eFEDS clusters show a smaller offset compared to eRASS1 because the latter contains a larger fraction of massive and disturbed structures. The offset measured in the eFEDS subsample is in agreement with the predictions from TNG and Magneticum, and the distribution of the offset parameter from dark matter only simulations. However, the tails of the distributions are different. Baryonic effects cause a decrement (increment) in the low (high) offset regime compared to the Xoff distribution from dark matter-only simulations. Finally, we find a correlation between the offset predicted by hydro simulations and Xoff measured in their parent dark matter-only run and calibrate a relation between them, which allows us to recover the full Xoff distribution with excellent precision. The work developed in this thesis is essential to characterize the real eRASS1 sample, hat will soon be available to the public. The development of the innovative mass function framework and its link to data with the offset between different definitions of the cluster center will allow the minimization of uncertainties in cluster count experiments due to selection effects related to the cluster dynamical state.