Speaker
Description
The wealth and complexity of X-ray data provided by modern observatories has seen a dramatic improvement in recent years, in part due to new facilities like NICER and IXPE. This trend will accelerate further with new techniques like polarimetric timing, as well as future missions like NewAthena. On the other hand, the software modelling tools utilized by the community to model X-ray data are mostly legacy stand-alone packages which are tailored to one-dimensional spectral analysis, and have relatively limited tools for advanced statistical analysis. In my talk, I will discuss how the limits of the existing software infrastructure can bias the retrieval of physical parameters from spectral models, using the measurement of black hole spin as an example. I will also discuss potential short and long term solutions to this problem, from more advanced Bayesian sampling techniques to surrogate models based on machine learning methods. Finally, I will discuss the current implementation of some of these methods in the newly released nDspec analysis package, which I hope to contribute to the NewAthena software suite.
