Abstract

Contributed Talk - Splinter ExoPlanets

Tuesday, 16 September 2025, 15:35

Rapid Interior Characterization of Exoplanets for the PLATO Era

Anna Julia Poser (1), Philipp Baumeister (1), Nicola Tosi (2), Heike Rauer (1,3)
(1) Freie Universität Berlin, Fachbereich Geowissenschaften, Fachrichtung Planetologie und Fernerkundung (2) DLR, Deutsches Zentrum für Luft- und Raumfahrt, Institut für Weltraumforschung, Berlin (3) DLR, Deutsches Zentrum für Luft- und Raumfahrt

PLATO, the upcoming ESA space mission set to launch at the end of 2026, will detect a large number of small to intermediate-sized exoplanets. These planets will come with well-constrained fundamental parameters, such as planetary mass, radius, and age of the planetary system [1, 2]. Combined with stellar properties, this PLATO dataset will allow fundamental questions to be addressed: What are these planets made of? What can they reveal about the formation and evolution of planets and their systems? To analyse the large number of expected detections efficiently, rapid interior characterization methods are required. ExoMDN is a machine-learning-based tool for rapid interior inference that is available to the community [3]. However, in its current implementation, ExoMDN is limited to treat planets up to ~25 Earth masses, and does not yet include volatile-rich or gas giant planets. Our work builds on this foundation by adapting the underlying mixture density network framework and extending it to cover the expected diversity in the PLATO dataset, In this talk, we present the conceptual design and implementation progress of an extended version of the ExoMDN framework. Our tool allows for fast interior characterization, supports the interpretation of PLATO discoveries, informs theoretical models of planet formation and evolution, and helps identify promising targets for atmospheric and ground-based follow-up studies, such as JWST, ARIEL, or the ELT.