Thèse Révéler les Populations de Pulsars Millisecondes au Coeur de notre Galaxie H/F - Doctorat.Gouv.Fr

Établissement : Université Grenoble Alpes École doctorale : PHYS - Physique Laboratoire de recherche : Institut de Planetologie et d'Astrophysique de Grenoble Direction de la thèse : Gilles HENRI ORCID 0000000216567525 Début de la thèse : 2026-10-01 Date limite de candidature : 2026-05-12T23:59:59 Une importante population de pulsars millisecondes (MSP) a été prédite dans les régions internes de notre Galaxie, mais très peu ont été découverts jusqu'à présent. Ceci qui remet en question soit notre capacité à les détecter, soit l'existence d'un si grand nombre de MSP dans cette région. L'interprétation des détections obtenues nécessite une modélisation précise des populations sous-jacentes, couplée aux seuils de détection des observations effectuées. Le but de ce projet de thèse est de développer un cadre statistique d'inférence basé sur des simulations pour contraindre les populations Galactiques de MSP et tester l'existence d'une population dans le bulbe. Les résultats obtenus et les propriétés multi-longueur d'onde des MSP seront comparés aux prédictions de l'évolution stellaire et utilisés pour comprendre la contribution des MSP à l'excès de Fermi au GeV, avec des implications importantes pour l'interprétation en termes de matière noire.

Millisecond pulsars (MSPs) are magnetised, rapidly rotating neutron stars. Beamed emission enables their identification at multiple frequencies, from radio to gamma-ray, with hundreds Galactic field MSPs known today. However, detecting MSPs at distances much larger than a few kiloparsecs from Earth in the direction of the Galactic center is severely limited by the scattering of radio waves o the dense interstellar medium (in radio) and to the bright GeV background from cosmic-ray interactions (in gamma-ray). This probably explains why, despite predictions of large populations of MSPs and intensive searches, only very few objects are currently detected from these inner regions. In recent years, our team has developed innovative strategies to identify new sources at bulge distances, with significant success, and precise modelling of the MSPs 3D distribution is now needed to understand whether a bulge MSP population does exist or if all detections still belong to the Galactic disk.
Deep radio observations, including with new facilities such as MeerKAT and FAST, are discovering MSPs at a higher rate than ever before. The main objective of the PhD project is to provide a comprehensive view of the underlying Galactic MSP populations by developing a proper statistical framework, built upon simulation based inference, for population studies of radio MSPs to : (i) assess the properties of the disk population, (ii) test the evidence for a bulge population, (iii) provide the most complete description of the populations detected, including multi-wavelength properties. If a bulge MSP population is found, the results will be used to constrain its contribution to the high-energy sky, and in particular to the excess of emission detected by Fermi at GeV energies, with important implications on Dark Matter models that have also been proposed to account for this signal. If no evidence for a bulge population is found, the model-to-data comparison will be also be used to derive a stringent upper limit on the total number of MSPs that could be hiding in these inner regions and investigate the fate of the MSP population predicted by stellar evolution. The PhD student will take part in the MSP searches lead by our group thanks to the proprietary on-going MeerKAT survey and up-coming Parkes survey of the bulge region, as well as targeted follow-up of promising bulge MSP candidate, with the goal of unveiling new MSPs towards the inner region of our Galaxy. In parallel, they will develop an simulation-based inference statistical framework for population studies of radio MSPs. They will rely on a specific implementation of truncated marginal neural ratio estimation (i.e. Swyft), which has already been applied to several physical problems (including by our team). The PhD student will construct a multi-parameter model of the Galactic MSP population, incorporating bulge and disk components, characterised by different period, luminosity and spatial distributions informed by stellar tracers. Observational biases and systematics will be carefully taken into consideration. This framework will be first tested on simulated data and then applied to existing Galactic MSP samples from past and ongoing uniform surveys (e.g., HTRU, FAST, and the MeerKAT one lead by our group), to derive the MSP populations properties. It will then be complemented by multi-wavelength information available, especially at high energy, to investigate the contribution of these populations to the Fermi GeV excess.

Au cours de son doctorat, l'étudiant ou l'étudiante devra utiliser, modifier et développer des codes visant à simuler des population de sources, notamment avec des techniques d'apprentissage automatique et de MCMC. Une expertise en programmation est donc requise. Le ou la doctorante devra également avoir un bon niveau d'anglais pour communiquer au sein de l'équipe, présenter ses travaux lors de conférences et publier ses résultats. Enfin, une formation en astronomie et astrophysique et un goût pour la modélisation des données d'observations seront également appréciés.