Exploratory Analysis of the Anemoi Framework for Machine Learning in Weather Forecasting 

Docente:  Giovanna Venuti

Tutor Aziendale: Nicolò Taggio Planetek Italia (taggio@planetek.it

Area di ricerca: 

Machine Learning in weather forecasting 

Keywords Machine Learning, Weather Forecasting, Anemoi 

Description

The proposed activity aims to explore Anemoi, the open-source framework developed by ECMWF to support the full machine learning lifecycle for weather forecasting applications, from dataset preparation to training, inference, and operational deployment. Weather monitoring and forecasting are increasingly strategic for several sectors, including civil protection, energy systems, agriculture, and critical infrastructure management, requiring more efficient and accurate predictive tools. In recent years, machine learning has gained significant relevance in meteorology by complementing traditional numerical weather prediction systems, offering reduced computational costs and faster forecast generation. Within this context, the activity will consist of an exploratory analysis of Anemoi aimed at understanding its architecture, core functionalities, usability, and potential applicability, with particular focus on identifying its main strengths, advantages, and possible limitations in experimental or operational scenarios. 

Scroll to Top