“The truth is much too complicated to allow anything but approximations.” — John von Neumann
Computer simulations of the ocean have become increasingly realistic. However, the ocean exhibits more processes and scales than can be resolved even with the world’s largest computers. Therefore, we must inevitably resort to approximations, which unfortunately result in errors that manifest as incorrect physics. My research seeks to understand and mitigate errors committed in numerical models, and ultimately to develop improved models.
Specifically, my research promotes nonhydrostatic ocean models, which solve the incompressible Navier-Stokes equations. Nonhydrostatic models do not resort to the commonly-used hydrostatic approximation, as traditional ocean models do. Consequently, the added nonhydrostatic physics allows for the representation of processes such as wave dispersion, overturning eddies, and small-scale turbulent mixing.
Additionally, I work on developing ocean models with layered, moving vertical coordinate systems for capturing stratified flows. Isopycnal (density-following) vertical coordinate systems provide natural representations of stratified flows, which in turn, reduce the number of vertical layers compared to coordinate systems used in traditional ocean models. This coordinate system can be applied to nonhydrostatic models to greatly improve computational efficiently.
Photo of Debordieu Sunrise by my friend and colleague Pat Limber