Milan is a Schmidt AI in Science Fellow working on combining climate modelling and machine learning.
He received his PhD from Oxford in climate computing with Tim Palmer and Peter Düben at the European Centre for Medium-Range Weather Forecasts. Afterwards he went for a PostDoc to the Massachusetts Institute of Technology where he developed SpeedyWeather.jl, a modern intermediate-complexity atmospheric model written in Julia. He enjoyed a very comprehensive education in climate physics from University of Kiel/GEOMAR and the Alfred Wegener Institute AWI in Germany; IUEM Brest in France and UNIS Svalbard on the Norwegian archipelago in the Arctic.
His research interests include:
- Climate modelling: Atmosphere and ocean, grid-point and spectral, dynamical core development, stochastic parameterizations, turbulence closures.
- Machine Learning: Online learning, automatic differentiation, generalization and interpretability.
- Computing: High-performance, low-precision, parallel, CPU and GPU, number formats, posit arithmetic, stochastic rounding, efficiency.
- Data compression: Lossy and lossless, information theory, data formats.
- Predictability of weather and climate: Chaos, uncertainty, ensemble prediction, error growth, weather forecasting.
- Software engineering: Open source, multiple dispatch and code composability, automatic differentiation, and the Julia programming language.
- Decarbonisation: Aviation's contribution on global warming, carbon footprints, decarbonising science.
