Computational Astrochemistry
by
In Astrochemsitry - the field that studies molecules in space - building accurate chemical models is central to our understanding of how molecules can form and be destroyed in space. The accuracy of such models however is heavily dependent on the availability of complete chemical gas-phase and solid-state reaction networks.
Due to the very non linear nature of chemistry in space, as well as the complex interactions between the chemistry and the physics of the gas and dust, determining the right sets of species and possible reactions between them is not trivial. In this talk I shall give an overview of the challenges we face when designing and optimizing networks as well as possible avenues for solutions, including sensitivity analysis, data-driven methods and machine learning.
