Abstract
Like in so many other areas, machine learning (ML) has become an important topic also in the provision of meteorological information. This includes modeling of the atmosphere but also post processing and impact based forecasting. The German Weather Service (DWD) is (co) developing various applications along the product chain. After an introduction to some ML basics the main focus will be on AICON, a data driven atmospheric model currently under development at DWD. This is followed by an overview over AI data assimilation approaches and post processing products.