Nora Loose

Nora Loose is an ocean modeler. She obtained her PhD from the University of Bergen in 2019, where she worked on adjoint modeling in the North Atlantic.

Loose’s research sits at the intersection of physical oceanography, numerical modeling, and machine learning. She has worked with a range of regional and global ocean models, including MITgcm, MOM6, Oceananigans, and ROMS. Across these, her scientific focus has spanned ocean turbulence, eddy parameterizations, teleconnections, and carbon dioxide removal applications. A topic she is particularly excited about is integrating machine learning with traditional ocean model dynamics, and making Earth system models differentiable to enable more seamless coupling of physics and data-driven components. Across all her projects, she places an emphasis on open-source software development, and has led the development of community Python packages such as GCM-Filters and ROMS-Tools.

At the Nansen Center, Loose works with the physical ocean component of the TOPAZ forecasting system. Her focus is on improving the representation of ocean physics in the Arctic and North Atlantic to produce better forecasts and reanalyses.