Improved representation of upper ocean dynamics and mixed layer depths in a model of the North Atlantic on switching from eddy-permitting to eddy-resolving grid resolution
Two configurations of a primitive-equation model of the North Atlantic are analyzed with respect to the simulated cycling of energy, mass, and heat in the upper ocean. One model is eddy-permitting (1/3° horizontal resolution), the other one is eddy-resolving (1/9° resolution), with both models using...
| Main Author: | |
|---|---|
| Format: | Article in Journal/Newspaper |
| Language: | unknown |
| Published: |
2002
|
| Subjects: | |
| Online Access: | https://eprints.soton.ac.uk/12723/ http://ams.allenpress.com/amsonline/?request=get-abstract&issn=1520-0485&volume=32&page=2277 |
| Summary: | Two configurations of a primitive-equation model of the North Atlantic are analyzed with respect to the simulated cycling of energy, mass, and heat in the upper ocean. One model is eddy-permitting (1/3° horizontal resolution), the other one is eddy-resolving (1/9° resolution), with both models using identical topographies and identical forcing fields at the surface and lateral boundaries. Besides showing some improvement in the simulated mean circulation and heat budgets, the eddy-resolving model reaches good agreement with satellite altimeter measurements of sea surface height variability. An unexpected finding of the model intercomparison is that simulated winter mixed layer depths in mid and high latitudes turn out to be systematically shallower by some 50 to 500 m in the higher resolution run, thereby agreeing better with observations than the 1/3° model results. This model improvement is related to enhanced levels of baroclinic instability leading to a decrease in potential energy and an associated increase in stratification. In the high-resolution model, shear-induced tilting of lateral density gradients generates stratification within the mixed layer itself, at a rate sufficient to set off an average surface heat loss of 5 W m–2 in mid and high latitudes. Although this is small compared to present uncertainties in surface heat fluxes, the resulting reduction in mixed layer depths may be important for an accurate simulation of water mass formation, air–sea gas exchange, and marine biological production. With traditional formulations of mixed layer physics assuming that properties are set by purely vertical mixing, and parameterizations of lateral subgrid-scale mixing often being tapered to zero in the mixed layer, present mixing schemes would have to be modified in order to account for eddy-induced generation of stratification in the surface mixed layer in noneddy-resolving ocean models. |
|---|