Thèse en cours

Traitement d’images avancé pour la détection directe d’exoplanètes avec les imageurs haut-contraste au sol

FR  |  
EN

Accès à la thèse

AttentionLa soutenance a eu lieu en 2023. Le document qui a justifié du diplôme est en cours de traitement par l'établissement de soutenance.
Auteur / Autrice : Chen Xie
Direction : Arthur ViganÉlodie Choquet
Type : Projet de thèse
Discipline(s) : PHYSIQUE & SCIENCES DE LA MATIERE - Spécialité : ASTROPHYSIQUE ET COSMOLOGIE
Date : Soutenance en 2023
Etablissement(s) : Aix-Marseille
Ecole(s) doctorale(s) : École Doctorale Physique et Sciences de la Matière (Marseille)
Partenaire(s) de recherche : Laboratoire : LAM - Laboratoire d'Astrophysique de Marseille
Jury : Président / Présidente : Jean-Luc Beuzit
Examinateurs / Examinatrices : Arthur Vigan, Remi Soummer, David Mouillet, Anne-Marie Lagrange, Sasha Hinkley, Hervé Le coroller, Elodie Choquet
Rapporteur / Rapporteuse : Remi Soummer, David Mouillet

Mots clés

FR  |  
EN

Résumé

FR  |  
EN

Current high-contrast imaging techniques have limitations (i.e., self-subtraction and oversubtraction) at short angular separations (i.e., <0.3''), precisely where most exoplanets are situated. It is thus necessary to develop advanced post-processing techniques to overcome these limitations, further pushing the limit of current high-contrast imagers. The goal of this thesis is to explore the potential of reference-star differential imaging (RDI) by using all the archival data for the detection and characterization of exoplanets and disks. The RDI technique I developed for SPHERE/IRDIS lays out the very fundamental framework in this thesis, and it enabled a wide range of applications and further developments. I show that RDI is a promising imaging technique for SPHERE, which can outperform angular differential imaging (ADI) at short angular separations. To expand the application to the SPHERE/IFS data, I further developed the RDI technique to include spectral diversities in the reference library without compromising sensitivity while reducing computational demands in post-processing. The concept of using archival data as references inspire the development of a new background subtraction for IRDIS. Combining RDI with data imputation using sequential nonnegative matrix factorization (DIsNMF), I demonstrate that RDI-DIsNMF is an optimized and model-free disk imaging technique that minimizes self-subtraction and oversubtraction. RDI-DIsNMF can accurately recover the disk morphology and preserve the disk flux with a throughput approaching 100%. At last, I apply the multiple techniques I developed in this thesis on IRDIS observations of HD 100453 as an astrophysical application. By recovering the spiral features via RDI-DIsNMF, I measure the spiral motion across 4 yr to perform dynamical motion analyses. The spiral pattern motion is consistent with the orbital motion of the eccentric companion. With this first observational evidence of a companion driving a spiral arm among protoplanetary disks, we directly and dynamically confirm the long-standing theory on the origin of spiral features in protoplanetary disks.