Review of existing shelf flux models of the NOWESP partners
Yu, J.C.S.; Monbaliu, J.; Berlamont, J. (1996). Review of existing shelf flux models of the NOWESP partners, in: NOWESP: 2. Compilation of scientific reports. pp. 12 [1-45] In: (1996). NOWESP: 2. Compilation of scientific reports. North-West European Shelf Programme (NOWESP): Hamburg. 324 pp., more |
| Authors | | Top | - Yu, J.C.S.
- Monbaliu, J., more
- Berlamont, J., more
| | |
| Abstract | Modelling shelf sea dynamics is faced with gradients and processes on many space and time scales.Hydrodynamic, transport and ecological processes are interrelated and have to be treated in a balanced and integrated way. Present developments, however, either include advanced 3D hydrodynamic and transport models but use only a simplified ecological parameterisation, or they feature a sophisticated ecological model ( including many different state variables and interactive processes) but a simple hydrodynamic parameterisation, i.e. using so called 'box model' which describes processes in an averaged environment over a large space. Therefore, one of the tasks for the modelling group (Task Group-B) is to review the existing shelf flux models which are used by the NOWESP partners. A set of inquiry forms for different types of models, i.e. hydrodynamic models, transport models and ecological models, has been designed by KUL and distributed to the NOWESP partners on different occasions. Information are summarized in the text, according to those replied inquiry forms. In a general view, different models have been used or been developed following the conventional trend: ºHydrodynamic (HN) models are 2D or 3D using numerical techniques for solving the partial differential equations. Besides the hydrodynamic parameters (u, v, w, z), 3D models include often the baroclinic state variables (i.e. salinity and temperature). Since tide is the predominate feature in the NOWESP area, it is calculated by all HN-models. Storm conditions have been included in every model as surface forcing. These models have different resolution of the horizontal mesh and vertical layer. Only a few fine-grid 3D models have a detailed turbulence closure scheme. ºTransport models are mostly developed for suspended sediment transport and other physical state variables. Both Eulerian and Lagrangian types of models are in use. º IfM has reported their 3D ecological model while IFR has reported their fine-boxes Channel model. Since most of the institutes involved with the development of HN-models are either having or developing 3D models, more interactions (e.g. atmospheric-ocean, surface-bottom) or processes (e.g. turbulence, density, ...) will be considered. The question of how fine the resolution of the mesh size has to be has long been discussed. A resolution of 1 km² seems an appropriate grid for the NOWESP area. But, the mu-CSM (joint development of KUL and MUMM) still has the finest resolution (i.e. 2.5' x 5' ) among the HN-models. Local refinement of meshes can be an alternative for the suspended sediment transport models, since most of the dominant features are occurring within coastal and estuarine areas. The numerical techniques for developing models are matured, it is therefore possible and desirable to develop fine mesh ecological model coupled with advective-diffusion models. Every refinement of the numerical model is, however, asking for better computational tools and performance. Because advanced computing tools, i.e. High Performance Computer (HPC), are available to most of the NOWESP partners, most of the models involved certain degree of optimization for implementation on HPCs. This mostly still remained on a job-level optimization, which means that the improvement of the model performance is marginal. The MAST -II concerted action MMARIE is an valuable project to this extent. Since most of the NOWESP partners who are involved with numerical models are participating in MMARIE, code level improvements as well as new algorithms developments for advance shelf flux models. |
|
