Cosmologia amb redshifts fotometrics
- Autores: Andrea Pocino Yuste
- Directores de la Tesis: Francisco Javier Castander Serentill (dir. tes.)
- Lectura: En la Universitat Autònoma de Barcelona ( España ) en 2020
- Idioma: español
- Tribunal Calificador de la Tesis: Héctor Martín Crocce (presid.), Roser Pelló Descayre (secret.), William Hartley (voc.)
- Programa de doctorado: Programa de Doctorado en Física por la Universidad Autónoma de Barcelona
- Materias:
- Enlaces
- Tesis en acceso abierto en: TDX
- Resumen
Current and future photometric surveys will observe a large volume of the universe that will allow us to accurately constrain the cosmological model. However, the constraining power from cosmological probes of photometric surveys highly relies on the accuracy and precision with which we can determine the galaxies redshifts. Therefore, the determination of photometric redshifts (photo-zs) and their effect in cosmological analysis should be treated and studied carefully.
In the first part of this thesis, we transform the photometry of existing simulations to mimic the photometric measurements of the Dark Energy Survey (DES). With this exercise, we expect to recover the real photo-z distribution in simulations, thus creating a more realistic environment to crosscheck the performance of DES in cosmological analyses that use photo-z. We transform the simulations using several method to transfer the statistical properties from the real observations photometry to the simulations.
In the second part of the thesis, we use the Self-Organizing Map technique to select spectroscopic targets for the C3R2 program aimed at establishing the mapping between color and redshift space. We also explore the color space defined by the photometry of galaxies from the Physics of the Accelerating Universe Survey (PAUS) in order to study the spectroscopic redshift coverage of its color space. We want to quantify the regions of color space without spectroscopic redshifts because the lack of spectroscopic representation can be a source of bias when the accuracy of photo-zs is evaluated by comparing it to spectroscopic redshifts and when the spectroscopic redshifts are used to determine the photo- z with training-based algorithms.
Lastly, we explore how the variation of the depth of ground-based observations combined with Euclid observations affects the accuracy and precision of the photo-z and thus the cosmological constraining power of Euclid focusing on photometric galaxy clustering and galaxy-galaxy lensing analyses. We also study how the number density of photometric galaxy samples affects the constraining power and which tomographic redshift binning configuration returns the maximum information to constrain the cosmological parameters. To perform such analyses, we create several realistic photo-z distributions based on the Euclid Flagship simulation and we use the Fisher forecast and the cosmological inference code, CosmoSIS, over the different configurations of the galaxy samples to determine the cosmological constraining power.
