Machine Learning Scientist for Weather and Climate (Varda project)

• Further develop and train machine learning model in Anemoi with a focus on regional weather predictions
• Improve the system by integrating observation data or working on ensemble methods
• Work on integrating the machine learning pipeline into production
• Fine tune and validate model against existing numerical model and observations
• Curate and validate ML training datasets

We are looking for a motivated Machine Learning Scientist to join the development team of the Varda machine learning weather prediction system. The model is being trained using archive data from MeteoSwiss operational forecasts and observations, with the objective to provide accurate and fast forecasts for the short- to medium term. A central component of this effort is Anemoi, a framework developed by the European Centre for Medium-Range Weather Forecasts (ECMWF) to support AI applications in weather and climate science., * University degree (MSc or PhD) in data science, computer science, physics or a related field

• Experience in training and validating large-scale deep-learning models on distributed systems
• Strong programming skills in Python and familiarity with a modern ML stack (e.g., PyTorch, hydra, zarr, dask) and best practices MLOps
• Experience in handling and processing large datasets or experience in high-performance computing (HPC) is an advantage
• Experience with weather and climate applications and weather ensemble forecasting is an advantage
• You are creative, solution-oriented and have excellent communication skills and the ability to work with interdisciplinary teams
• Good knowledge of spoken and written English

80%-100%, Zurich, fixed-term

The Center for Climate Systems Modeling (C2SM) at ETH Zurich, in partnership with the Federal Office of Meteorology and Climatology (MeteoSwiss), is pioneering innovative methods to leverage machine learning for numerical weather forecasting and climate modeling., * Unique opportunities to develop state-of-the-art Machine Learning system and shape the future of weather forecasting

• You will join a dynamic team operating at the intersection of cutting-edge research and real-world applications
• We are committed to fostering a diverse and inclusive workplace and offer flexible working arrangements to support work-life balance for all team members