On the Design of Dynamical Cores for GCMs: Physical and Computational Challenges

Christiane Jablonowski
University of Michigan
Department of Atmospheric, Oceanic & Space Sciences

The design of a dynamical core not only incorporates numerical and scientific choices but also faces many physical and computational challenges.
For example, there is a high demand for built-in physical conservation laws, like the conservation of mass, or the monotonic transport of tracer substances. Secondly, the anticipated increases in grid resolution put strong demands on the computational efficiency of the numerical algorithms, especially on modern parallel computing architectures.
The lecture focuses on the physical and computational dynamical-core requirements, and reveals the sometimes hidden features in the design process. In particular, the subgrid-scale diffusive and filtering processes are addressed. In addition, the lecture gives an overview of evaluation techniques and selected dynamical core test cases. The lecture utilizes many examples from existing dynamical cores and invites to an open discussion of the pros and cons of the (co-)design decisions.

Presentation (PDF File)

Back to Model and Data Hierarchies for Simulating and Understanding Climate: Tutorials