The Importance of Remote Forcing for Regional Modeling of Internal Waves

Matthew R. Mazloff,Bruce Cornuelle,Sarah T. Gille,Jinbo Wang
DOI: https://doi.org/10.1029/2019jc015623
2020-01-01
Abstract:Regional ocean general circulation models are generally forced at the boundaries by mesoscale ocean dynamics and barotropic tides. In this work we provide evidence that remotely forced internal waves can be a significant source of energy for the dynamics. We compare global and regional model solutions within the California Current System. Both models have similar inputs, forcings, and identical grids and numerics. The global model has a steric height power spectrum consistent with mooring observations at superinertial frequencies, while the regional model spectrum is weaker. The regional model also has less sea surface height variance at high wavenumber than the global model. The vertical velocity variance is significantly larger in the global model, except in the sheltered Southern California Bight. While the regional model has roughly equal high-pass baroclinic and barotropic kinetic energy levels, the global model high-pass baroclinic kinetic energy is 28% (0.39 PJ) greater than the barotropic energy. An internal wave energy flux analysis reveals that the regional model domain boundaries act as a sink of 183MW, while in the global model the analysis domain boundaries act as a source of 539MW. This 722MW difference can account for the relative increase of 0.39 PJ high-pass baroclinic energy in the global model, assuming a baroclinic kinetic energy dissipation time in the domain of approximately 6.3 days. The results here imply that most regional ocean models will need to account for internal wave boundary fluxes in order to reproduce the observed internal wave continuum spectrum. Plain Language Summary Global ocean simulations can be too computationally expensive, so many researchers prefer to use lower cost regional ocean models. The open ocean boundary conditions of these regional models must be prescribed from other products. It is common practice for these prescribed conditions to lack high-frequency oceanic variability. Here we show that this missing component of the dynamics may account for a large amount of the overall high-frequency energy in the region. Without this, remotely forced energy input regional models may underestimate processes important to the overall state (e.g., upwelling processes fundamental to the ocean ecosystems). Regional models must account for these remotely forced dynamical signals in order to produce simulations with realistic short time- and space-scale variability.
What problem does this paper attempt to address?