Marcin Chrust

Senior Scientist
Research, Earth System Assimilation, Data Assimilation Methodology

Summary:

Marcin works on algorithmic aspects of the ocean ensemble variational data assimilation system NEMOVAR and leads the development and maintenance of the Object-Oriented Prediction System (OOPS), the key framework underpinning ECMWF's operational data assimilation system, including 4D-Var for the Integrated Forecasting System (IFS). He is actively engaged in variational ocean data assimilation research, including the development of the Ensemble of Data Assimilations (EDA) for improved background error covariances in ORAS6, and leads the implementation of the Copernicus Climate Change Service (C3S) contract enhancing ocean reanalysis capabilities. He set up and hosts a collaboration platform for NEMOVAR developers at Cerfacs, INRIA and UK Met Office. Marcin actively engages in researching machine learning applications in data assimilation, particularly neural network-based online model error correction within the weak-constraint 4D-Var framework. 

Revised Expertise List

  • Data assimilation methodology

  • Object-Oriented Prediction System (OOPS) development and leadership

  • Variational ocean data assimilation research and NEMOVAR development leadership

  • Leading C3S contract implementation for ocean reanalysis

  • Machine learning for hybrid model development

  • Numerical methods and high-performance computing

  • Ocean and atmosphere ensemble data assimilation

Professional interests:
  • Data assimilation methodology
  • Machine learning
  • Numerical methods
  • High performance computing
Career background:

Marcin obtained his PhD in computational fluid mechanics from the University of Strasbourg in 2012. After defending his dissertation he took a post-doctoral position at the University of Ottawa and subsequently at the Ecole Polytechnique de Montreal. Before joining ECMWF in 2015 he worked as a research associate at Bombardier Aerospace in the Advanced Aerodynamics department in Montreal. Marcin's background is in numerical methods, applied mathematics and high performance computing.

External recognitions
  • Member of OceanPredict DA-TT (2017 - )
  • Co-chair of OceanPredict DA-TT (2026 - )