Computational Methods for Geophysical Fluid Dynamics
Develop advanced numerical methods for solving fluid dynamics PDEs on supercomputers, focusing on accuracy, parallelization, and adaptive grids.
This grant is no longer accepting proposals
NRC Research Associateship Programs has archived this opportunity.
Funder: NRC Research Associateship Programs
Due Dates: May 1, 2025
Funding Amounts: Base stipend approximately $67,000/year plus $3,000 travel allowance; supplements based on experience; typical tenure 2-3 years.
Summary: Postdoctoral and senior researchers develop advanced numerical methods for geophysical fluid dynamics PDEs on supercomputers, focusing on accuracy, parallelization, and adaptive grids.
Key Information: Open to U.S. citizens, permanent residents, and non-U.S. citizens; strong programming and mathematical background required; relocation and health insurance benefits included.
Description
This fellowship opportunity at the Naval Postgraduate School (NPS) supports research in developing cutting-edge computational methods for geophysical fluid dynamics, including aerodynamics, atmospheric, and ocean modeling. The project emphasizes element-based Galerkin methods such as spectral element, discontinuous Galerkin, entropy-stable, and kinetic-energy-preserving methods to approximate spatial derivatives. Advanced time-integration techniques like Jacobian-free Newton-Krylov, implicit-explicit, parallel-in-time, and multirate methods are employed to achieve highly accurate solutions for large-scale problems on parallel supercomputers, including multi-core and many-core architectures.
Key research areas include iterative solvers, preconditioners, Message-Passing Interface (MPI), GPU computing, and the use of unstructured and adaptive grids. The project also involves improving non-reflecting boundary conditions for the Navier-Stokes equations.
Applicants should have expertise in fluid dynamics and/or geophysical fluid dynamics, with a strong foundation in numerical analysis, scientific computing, and partial differential equations. Additional knowledge in super-parameterization, reduced-order modeling, and scientific machine learning is advantageous. Proficiency in object-oriented programming languages such as Fortran, C/C++, and Julia is essential.
