Poster details
Sampling of the number density in the context of Halo Occupation Distribution models
Abstract
Lots of resources and human efforts in Stage IV cosmological surveys (such as DESI or Euclid) are dedicated to constrain the nature of dark matter and dark energy. In particular, galaxies will be used as a biased tracers of the total matter present in the Universe.
In this work we use Halo Occupation Distribution (HOD) models to populate dark matter-only simulations with galaxies following eBOSS Emission-Line Galaxies (ELG) data. Then, we establish a connection between numerical simulations and observations of our selected dark matter biased tracers: galaxies. The HOD model makes different assumptions for the distribution of galaxies in haloes, in particular for satellite galaxies. One of these assumptions is the probability to find a given number of satellite galaxies in a halo. In general, a Poisson distribution is assumed but it has not been proven that galaxies follow this particular distribution. In this work we cover the entire possible range of variance adding the binomial distribution to our set of functions. We find that varying the variance of the PDF have a very relevant effect on galaxy clustering, and the addition of the binomial PDF allows to explore this variable in a continuous way. Furthermore, we also want to measure if the HOD inferred from data, could depend on the cosmology assumed/simulation used. For this purpose we implement our improved HOD model to populate the UNIT (1000Mpc/h side + mp) and OuterRim (3000Mpc/h side +mp) DM-only simulations with galaxies. Finally, we use eBOSS clustering data and other statistics in order to constrain the values of the parameters of our HOD models.
In this work we use Halo Occupation Distribution (HOD) models to populate dark matter-only simulations with galaxies following eBOSS Emission-Line Galaxies (ELG) data. Then, we establish a connection between numerical simulations and observations of our selected dark matter biased tracers: galaxies. The HOD model makes different assumptions for the distribution of galaxies in haloes, in particular for satellite galaxies. One of these assumptions is the probability to find a given number of satellite galaxies in a halo. In general, a Poisson distribution is assumed but it has not been proven that galaxies follow this particular distribution. In this work we cover the entire possible range of variance adding the binomial distribution to our set of functions. We find that varying the variance of the PDF have a very relevant effect on galaxy clustering, and the addition of the binomial PDF allows to explore this variable in a continuous way. Furthermore, we also want to measure if the HOD inferred from data, could depend on the cosmology assumed/simulation used. For this purpose we implement our improved HOD model to populate the UNIT (1000Mpc/h side + mp) and OuterRim (3000Mpc/h side +mp) DM-only simulations with galaxies. Finally, we use eBOSS clustering data and other statistics in order to constrain the values of the parameters of our HOD models.
poster-030-.pdf