M/F PhD in Astrophysics and Data analysis

Job Description

Thesis Subject

Machine learning and statistical tests for detecting rings around exoplanets. Application to data from the Kepler, TESS, and PLATO space missions:

Although the four giant planets in the Solar System have rings, their origin and long-term stability are still debated. Recent discoveries of rings around smaller objects, such as Chariclo or Haumea, have reignited the debate, showing that a significant fraction of icy bodies likely possess rings. In this context, the clear detection of rings around exoplanets, even in small numbers, would provide unprecedented insight into the formation mechanisms, structure, and composition of rings in planetary systems—including clues for the Solar System. To detect exorings, the preferred method is transit photometry, which allows for observing anomalies in the light curve of a star when it is partially obscured by a planet and its rings. Although several detections have already been suggested, none have yet been confirmed in an evolved system like the Solar System.
The objectives of this funded doctoral position are to develop advanced exo-ring detection techniques, evaluate their performance, and apply them to large datasets to obtain new detections. Two types of approaches will be compared: a statistical testing approach based on likelihood ratios, and a Machine Learning (ML) approach based on standard methods (convolutional neural networks, SVM, etc.) followed by less generic approaches more suited to the statistical nature of the problem. An important aspect of this work concerns the modeling of the statistical properties of noise in light curves. Our preliminary results on the analysis of Kepler data show, for certain light curves, a significant deviation from the usual Gaussian assumption.
This requires "tailor- made" noise models that will lead, for both the classical and ML approaches, to specific methods with more reliable control of the false alarm rate and increased detection power compared to conventional methods. To this end, the internship will consider three important questions: first, what are the theoretical performance limits for exo-ring detection algorithms? Second, what statistical modeling of noise affecting light curves should be implemented, and how does this impact detection methods? Third, what astrophysical discoveries have been made by detection methods developed on large-scale data: Kepler,TESS, and then PLATO? The search for theoretical or algorithmic evidence to guarantee the reliability and robustness of the modeling and detection methods will constitute a methodological thread that will guide this work.

Your Work Environment

This PhD project is part of an ANR project (called WRAPS: «Where do Rings Appear in Planetary Systems?») funded in 2025 for 5 years to address the issue of exo-rings (PI:
S. Sulis). The PhD will be carried out over three years within the Planetary Systems Group (GSP) at the Laboratoire d'Astrophysique de Marseille (UMR7326).

Constraints and risks

The student will be expected to participate in international conferences
in France, across Europe, and worldwide.

Compensation and benefits

Compensation

2300 € gross monthly

Annual leave and RTT

44 jours

Remote Working practice and compensation

Pratique et indemnisation du TT

Transport

Prise en charge à 75% du coût et forfait mobilité durable jusqu’à 300€

About the offer

About the CNRS

The CNRS is a major player in fundamental research on a global scale. The CNRS is the only French organization active in all scientific fields. Its unique position as a multi-specialist allows it to bring together different disciplines to address the most important challenges of the contemporary world, in connection with the actors of change.

CNRS

The research professions