The semi-prognostic method
An overview is given of the semi-prognostic method, a new and novel technique that can be used for adjusting models to correct for systematic error. Applications of the method to a regional model of the northwest Atlantic Ocean, and to an eddy-permitting model of the entire North Atlantic, show impr...
| Published in: | Continental Shelf Research |
|---|---|
| Main Authors: | , , , , , |
| Format: | Article in Journal/Newspaper |
| Language: | English |
| Published: |
2016
|
| Subjects: | |
| Online Access: | https://doi.org/10.1016/j.csr.2004.07.009 https://ora.ox.ac.uk/objects/uuid:c1f256f1-4718-4d25-971d-8e30ccd10367 |
| Summary: | An overview is given of the semi-prognostic method, a new and novel technique that can be used for adjusting models to correct for systematic error. Applications of the method to a regional model of the northwest Atlantic Ocean, and to an eddy-permitting model of the entire North Atlantic, show improvement in the handling of the Gulf Stream/North Atlantic Current systems, especially in the "northwest corner" region southeast of Newfoundland where prognostic models show systematic errors of as much as 10°C in the temperature field. Use of the semi-prognostic method also leads to improvement in the modelled flow over the eastern Canadian shelf. An advantage of the semi-prognostic method is that it is adiabatic; in particular, in spite of the improvement seen in the modelled hydrography, the potential temperature and salinity equations carried by the model are unchanged by the method. Rather, the method introduces a correction term to the horizontal momentum equations carried by the model. Adiabaticity ensures that the method does not compromise the requirement for the flow in the ocean interior to be primarily in the neutral tangent plane, and also ensures that the method is well-suited for tracer studies. The method is also easy to implement, requiring only an adjustment in the hydrostatic equation carried by the model. We also describe the use of the method as a diagnostic tool, for probing the important dynamic processes governing a phenomenon, and finally as a technique for transferring information between the different subcomponents of a nested modelling system. © 2004 Elsevier Ltd. All rights reserved. |
|---|