Evaluation of Systematics in Euclid Spectroscopic Sample: from Simulations to Observations

Euclid is a mission of the European Space Agency (ESA) launched in July 2023 to explore the content and evolution of the Universe. Over at least six years, Euclid will survey one-third of the sky up to redshift z ∼ 2, aiming to constrain cosmological parameters, including the nature of dark energy and dark matter, as well as the sum of neutrino masses. Given the large survey volume, systematic uncertainties are expected to dominate over statistical errors. Mitigating these systematics is therefore critical to avoid biased cosmological inferences. In particular, reliable galaxy clustering measurements require a detailed understanding of the completeness and purity of the sample. To support this, a secondary survey, namely Euclid Deep Survey (EDS), will complement the main Euclid Wide Survey (EWS) by producing a purer and more complete galaxy catalog. However, the full depth of EDS data will only become available upon completion of the mission. This thesis investigates systematic effects in Euclid spectroscopic measurements and is divided into two main parts. The first part presents pre-launch end-to-end simulations used to estimate redshift errors, evaluate cross-contamination, and identify sources of bias before real data became available. However, these simulations are limited by their simplified modeling of instrumental effects, such as detector persistence and spurious photometric detections. To overcome this limitation, the second part of the thesis introduces a method that injects simulated spectra into real images, capturing both known and unknown observational effects. This dual approach enables an early evaluation of spectroscopic performance before the completion of the survey, and it provides a complementary tool to the EDS for characterizing and mitigating systematic uncertainties.