Unlocking a Bayesian view on protoplanetary disks
PhD ceremony: | T.F. Kaeufer |
When: | October 29, 2024 |
Start: | 12:45 |
Supervisors: | I.E.E. (Inga) Kamp, Prof, prof. dr. P.T. Woitke |
Co-supervisor: | dr. M. Min |
Where: | Academy building RUG |
Faculty: | Science and Engineering |

Protoplanetary disks are a natural by-product of the star-formation process. These disks of dust and gas orbit the emerging star and are the birthplace of planets. To understand the planet formation process better, it is crucial to understand the conditions in these disks. This is typically done by comparing observations to models. Bayesian analysis offers a statistically robust way of doing so.
Till Kaeufer's thesis unlocks a Bayesian view on protoplanetary disks. He developed different methods and simple but fast models to allow for a statistical comparison between models and observations.
Kaeufer developed Neural Networks that emulate existing disk modelling software with a fraction of the computational cost to analyse the energy emitted by disks over a large wavelength range (from ultraviolet to radio wavelengths). He examined a small sample of 30 well-known disks in detail to extract typical disk properties and uncertainties of the fitting process. He also studied a larger sample of 672 disks, looking for trends between different groups of disks.
Additionally, Kaeufer introduces the Dust Continuum Kit with Line Emission from Gas (DuCKLinG), which is a fast model to analyse the recent mid-infrared spectra captured by the James Webb Space Telescope. This thesis uses these models to analyse the spectra of two protoplanetary disks, extracting the molecular and dust properties, and highlights the benefit of the developed model over simpler commonly used models.