Daniele Nerini

Dr. Nerini research has contributed to bridging stochastic modeling and machine learning in the context of weather and climate prediction. Early on, I worked on stochastic simulation of precipitation fields, improving how variability and uncertainty are represented in nowcasting applications. Building on this expertise, I later extended my work to machine learning–based approaches to forecasting, with a particular focus on robustness for operational use. These efforts resulted in new methods and practical implementations that demonstrate how AI can complement traditional numerical weather prediction, especially in nowcasting and statistical postprocessing. This line of work has informed collaborations between research institutions and operational weather services, where my contributions helped shape the integration of AI into forecasting pipelines.

Beyond methodological advances, he has played a central role in building collaborative infrastructures that enable data-driven research in meteorology and climate. At MeteoSwiss, in close collaboration with international partners, I have coordinated and contributed to projects that provide open-source software, shared datasets, and reproducible workflows for AI in weather prediction. I have emphasized practices that ensure scientific rigor, transparency, and long-term usability of data and code. These efforts have allowed the wider community — from researchers to operational forecasters — to access, test, and improve state-of-the-art methods. This work has supported the growth of an international research network at the interface of atmospheric sciences and data science.