Research project A new epoch for applied stellar spectroscopy

 However, disentangling the planets' fingerprints in the observed spectra is a tremendous challenge that requires a detailed understanding of how starlight varies across the surface.

In recent years, stellar spectroscopy have seen revolutionary progress both on the modelling and theory side. In this project, we build upon the advances to address astrophysical problems with novel techniques.

The light we observe from stars is emitted from a very thin outer layer of gas known as the stellar atmosphere. The layer experiences constant large-scale motions and matter and radiation are not in equilibrium. Accounting for both these effects properly can be done with so-called 3D non-LTE modelling of the stellar spectrum, which is numerically extremely expensive compared to traditional models. However, only with such advanced modelling techniques can the spatial variation of light from the centre to the limb of the stellar disk be accurately predicted. The first goal of this project is to accurately separate the stellar and planetary imprints on the combined spectrum during a planet transit. The Sun, the only star for which we can study the spatially resolved light in exquisite detail, is our most important benchmark star.

Stellar surveys can nowadays routinely record data with high spectral resolving power for hundreds of stars simultaneously, a number that will rise further with the next generation of instruments. The assembly of million-star data sets is a major regime change that can only be mastered by introducing elements of machine learning in the analysis and scientific exploitation. The second goal of this project is to develop efficient and automated analysis tools for parameter and abundance analysis of stars that leverage the developments on the theory side.