Optimization of atomic data for improved kilonova modellingHYBRID

The recent observations of several neutron-star merger events and associated electromagnetic transients, particularly AT2017gfo, have provided strong evidence that heavy r-process elements like, lanthanides and actinides, are synthesized in these explosive events. However, pinpointing specific spectral features continues to be a challenge due to the lack of accurate atomic data, necessary as input for the spectral modeling of the ejecta. The atomic structure and radiative properties of the heaviest neutron capture elements are still largely uncharted, particularly in the infrared spectral region, which is prominent in kilonova spectra and critical for late-stage, non-local thermodynamic equilibrium (non-LTE) conditions.
In this presentation, the main challenges in computing accurate atomic data for kilonova spectra will be discussed. Different methods to overcome these challenges will be presented, including improvement and refining configuration interaction techniques and the application of machine learning ideas to improve data accuracy. Emphasis is place on balancing accuracy and completeness of the data while striving for efficient computational times to ensure a comprehensive dataset. These efforts aim to refine spectral predictions and enhance our understanding of r-process nucleosynthesis in neutron-star mergers.