SCHEDULE

1200-1430 (Hybrid session)

Andrew Stuart - Learning Solution Operators For PDEs: Algorithms, Analysis and Applications

Wei Pan - Stochastic bathymetry in TQG

Long Li - Stochastic Ekman models

Speaker: Alberto Carrassi, Dept of Physics and Astronomy “Augusto Righi”, University of Bologna, IT

Title: Using machine learning, data assimilation and their combination to improve a new generation of Arctic sea-ice models 

With: L. Bertino (NERSC, NO), M. Bocquet (ENPC, FR), J. Brajard (NERSC, NO), Y. Chen (U Reading, UK), S. Driscoll (U Reading, UK), C. Durand (ENPC, FR), A. Farchi (ENPC, FR), T. Finn (ENPC, FR), C. Jones (U Reading, UK), I. Pasmans (U Reading, UK) and F. Porro (U Bologna, IT)

Abstract: We present an overview of the research efforts and results obtained in the context of the international project SASIP aimed at understanding and prediction the Arctic changes. We have been working on developing novel data assimilation, machine learning and their combination adapted to a new generation of sea-ice models that treats the ice as a brittle solid instead of as a fluid. These models present unique physical challenges such as sharp gradients, anisotropy and multifractality. In this talk we will first present the application of an ensemble variational method to estimate the state and parameters of the sea-ice model based on synthetic, satellite-like, data, illustrating the power and limitation of the available measurements. Second, we will show how to adapt the data assimilation procedure to the use of discontinuous Galerkin model, a modification that makes possible to assimilate very dense data (such as satellite) as well as to develop a scale-aware localisation procedure. 

To incorporate multifractal, anisotropic, and stochastic-like processes in sea ice, we envision the combination of geophysical sea-ice models together with neural networks in a hybrid modelling setup. On the one hand, deep learning can surrogate computationally expensive sea-ice models, on the other hand, deep learning can parametrize subgrid-scale processes in sea-ice models and correct persisting model errors, improving the forecasts by up to 70 % across all model variables on an hourly timescale. Finally, we will show how to use neural networks to emulate and replace a physical parametrization of the sea-ice melt ponds that create on the ice surface, and that have a major role on the albedo and thus on the general energy balance. Overall, our results show the potential of data assimilation and machine learning to extract much information from available data to correct model prediction and the models themselves.

SCHEDULE

12:00-12:45 Alberto Carrassi

12:45-13:15 Discussion

13:15-13:30 Break

13:30-13:45 Mini-talk 1 – Alexander Lobbe

13:45-14:00 Mini-talk 2 – Mael Jaouen

14:00-14:30 Discussion

Speaker: Peter Korn

Title: Remarks on Virtual Euler Flows and Numerical Ocean Models  

Abstract: In this talk I present a discretisation of the incompressible Euler equations on polygonal prismatic meshes and describe associated discrete equivalents of continuous conservation laws. Capitalizing on the fact that the Euler equations form the heart of the dynamical core of numerical ocean circulation models I discuss implications of the numerical approach for the computational design and structural understanding of ocean models. The talk closes with outlining future research directions.  

SCHEDULE

12:00-12:45 Peter Korn

12:45-13:15 Discussion

13:15-13:30 Break

13:30-13:45 Mini-talk 1 - James Woodfield

13:45-14:00 Mini-talk 2 - Wei Pan

14:00-14:30 Discussion & Close

SCHEDULE

12:00-12:20 Solange Coadou Chaventon

12:20-12:40 Yan Barabinot

12:40-13:10 Etienne Memin

13:10-13:20 Break

13:20-14:00 SWOT data TBD (likely Fabrice Collard)