Gaia : de la validation des données aux paramètres du Red Clump

The Gaia mission of the European Space Agency (ESA) aims to map our galaxy with an unprecedented astrometric precision. It is therefore very important that the data that will be published be rigorously validated to ensure an optimal quality in the Catalogue. These validations are done by one of the teams of the coordination unit CU9 of the Gaia DPAC Consortium (Data Processing and Analysis Consortium) commissioned by ESA of the Gaia catalogue production. As part of this thesis, we implemented all the necessary infrastructure to validate the Gaia catalogue by comparison with external catalogues. This last manages all the interactions with the global environment of validations and with the Gaia database. Then we developed a set of statistical tests to validate the data from the first Gaia catalogue (DR1). These tests relate in particular to the homogeneity of data on the sky, the quality of the positions and of photometry of all the stars of DR1 (more than a billion stars, $V <20$) as well as that of the parallaxes and proper motions for textit{Tycho-Gaia} Astrometric Solution (TGAS) stars, around two million stars in common in Gaia and Tycho-2 catalogues ($V <12$).These DR1 statistical tests are operational and were already applied very recently on preliminary data. This has improved the data (thus the quality of the catalog) as well as allowed to characterize the statistical properties. This characterisation is essential for a correct scientific exploitation of the data. The first Gaia catalogue will be released in late summer 2016.Among the objects that Gaia observes, there is a population of stars particularly interesting, the Red Clump (RC) stars, widely used for distance indicators. We developed and tested two methods to model the colour-colour (CC) and effective temperature - colour relations in all photometric bands, from the ultraviolet to the near infrared. They will allow us to characterize the RC in the Gaia G band upon publication of the catalogue: 1. using theoretical models, and 2. empirically, based on a Monte Carlo Markov Chain (MCMC) method. For this we have very carefully selected samples of stars with a good photometric quality, good metallicity determined by spectroscopy, an homogeneous effective temperature and a low interstellar extinction.From these CC and temperature-colour calibrations, we then developed a Maximum Likelihood method that allows to derive absolute magnitudes, temperatures and extinctions of the RC stars. Estimates of colours and extinctions are tested on stars with spectroscopically measured effective temperatures and an extinction determined by the measurement of Diffuse Interstellar Bands (DIB). These intrinsic properties of RC stars will allow to characterize the Gaia RC and calibrate, within the Gaia G band, the absolute magnitude of this standard candle, first essential step of determining distances in the Univers.La mission Gaia de l'Agence Spatiale Européenne (ESA) a pour objectif de cartographier notre galaxie avec une précision astrométrique jamais atteinte auparavant. Il est donc particulièrement important que les données qui seront publiées soient rigoureusement validées afin d'assurer une qualité optimum au Catalogue. Ces validations sont faites par l'une des équipes de l'unité de coordination CU9 du Consortium Gaia DPAC (Data Processing and Analys Consortium) chargé par l'ESA de la production du Catalogue Gaia. Dans le cadre de cette thèse, nous avons mis en place toute l’infrastructure nécessaire à la validation du catalogue Gaia par comparaison avec des catalogues externes. Celle-ci gère toutes les interactions avec l'environnement global des validations et avec la base de données Gaia. Ensuite nous avons développé un ensemble de tests statistiques pour valider les données du premier catalogue Gaia (DR1). Ces tests concernent notamment l’homogénéité des données sur le ciel, la qualité des positions et de la photométrie de l'ensemble des étoiles de DR1 (plus d'un milliard d'étoiles, $V<20$) ainsi que celle des parallaxes et mouvements propres des étoiles de textit{Tycho-Gaia} Astrometric Solution (TGAS), environ deux millions d'étoiles communes aux catalogues Gaia et Tycho-2 ($V