Hello CROCO users,
I am slowly working my way up to build a realistic simulation of seafloor hydrothermal discharge. I have met some problems with implementing open boundary conditions in the NBQ mode. The diagrams below illustrate the problem. In this experimental simulation, a buoyant plume rises from the middle of the bottom of a 50 m by 50 m by 50 m box that is open on all four sides (see the second diagram). There is no forcing other than the heat coming from the bottom. The background is linearly stratified. The problem occurs at the southern boundary. The surface elevation (zeta) shows artifacts that appear to generate ripples that propagate through the domain. The artifacts grow with time and will eventually lead to a blow-up. Here, I used sponge to dampen the amplitudes of the artificial surface anomalies to keep the simulation stable.
I am wondering if the fact that the artifacts only show up at the southern boundary in this symmetric setup suggests there could be some issues in the southern obc in the NBQ mode. I did the same simulation in the hydrostatic mode and the artifacts did not show up.
I would appreciate if anyone can provide suggestions on diagnosing this issue.