Hi all,
I have a question related to the implications of the Boussinesq approximation on simulated sea level in ROMS.
The missing non-Boussinesq steric effect and the consequent missing global steric effect in Boussinesq models have been explained by several studies for global models (e.g., Greatbatch 1994, Griffies & Greatbatch, 2012). Normally, for global models an a-posteriori correction can be made using global mean density. Presumably, this also works for regional models if there are no net mass fluxes across the boundaries. However, often open lateral boundaries are used. If Flather radiation conditions are applied, sea level in the interior will be further adapted by the correction to the barotropic currents.
I’m unclear what the consequence of the Boussinesq approximation is for sea level in long-term integrations with open boundaries. Have ROMS users or developers given this any thought before? Have comparisons been made between sea level in Boussinesq and non-Boussinesq regional models with open boundaries? I would like to learn more.
Best,
Tim