Firstly, improvement in the model run-time was sought. By developing a parallel implementation using High Performance Fortran, a speed-up by a factor of 69 is achieved on a Cray <i>T3E</i> (128 processors). In addition, an alternative diffusion discretisation, the implicit Finite Volume method, is implemented, producing a speed-up of 34. Combined, these two approaches yield a speed-up of 2346. The results of this new version compare well with those of the previously validated sequential code. Secondly, attention is focused on the wet deposition process for reduced nitrogen, a common and recognised weakness of existing UK models. Here, in contrast with previous approaches, the directional orographic enhancement of precipitation and wet deposition from the seeder-feeder effect is investigated and then explicitly taken into account. Furthermore, a more realistic description of the wet deposition parameterisation is developed. Comparison with measurement data shows that these two considerations improve significantly the reduced nitrogen wet deposition results. These and other developments of the models result in a robust and significantly improved tool for atmospheric reduced nitrogen predictions. Moreover, they provide a foundation for further use of the model in a variety of applications. The model is applied here to consider the transport and deposition of oxidised nitrogen and sulphur, to evaluate regional budgets of reduced nitrogen for the British Isles and to test scenarios for 2010.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:651061 |
| Date | January 2002 |
| Creators | Fournier, Nicolas |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/10908 |
Page generated in 0.0018 seconds