Abstract
Numerical methods for geophysical fluid dynamics are a central piece in weather and climate modeling, furthermore, they are recently going through radical changes to become adequate to the new generation of supercomputers that demands massively parallel processing. This project concerns the investigation of numerical methods, considering time and space discretizations, aiming at adequately solving the equations of atmospheric or ocean dynamical cores. The main goal is to ensure adequate representation of relevant flow characteristics, mimicking the continuous equations, with particular attention to the development of methods that can be algorithmically competitive in modern supercomputers. Investigations will pursue different interconnected directions. The study of numerical schemes for quasi-uniform spherical grids will aim at avoiding scalability problems encountered in traditional models. Alongside, vertical discretizations will be investigated aiming at an adequate representation of flow at sharp topography (e.g. Andes Range) and the efficient solution of vertical non-hydrostatic dynamics. Additionally, the exploration of the time dimension will be investigated as a source of added parallelism, with methods that change the time integration way of thinking of traditional models. The project will be developed in partnership with national and international collaborators, including specialists in geophysical fluid modeling. We expect the project to directly contribute to the development of numerical methods of this new worldwide emerging generation of global dynamical cores. Also, we aim to contribute to solving local (Brazilian) climate and weather challenges, particularly those that can be tackled with better-suited numerical techniques. (AU)
| Articles published in Agência FAPESP Newsletter about the research grant: |
| More itemsLess items |
| TITULO |
| Articles published in other media outlets ( ): |
| More itemsLess items |
| VEICULO: TITULO (DATA) |
| VEICULO: TITULO (DATA) |