Long-term calculations with large air pollution models

The air pollution levels in a given region depend not only on the emission sources located in it, but also on emission sources located outside the region under consideration, and even on sources that are far away from the studied region. This is due to the transboundary transport of air pollutants. The atmosphere is the major medium where pollutants can be transported over long distances. Harmful effects on plants, animals and humans can also occur in areas which are long away from big emission sources. Chemical reactions take place during the transport. This leads to the creation of secondary pollutants, which can also be harmful. The air pollution levels in densely populated regions of the world, such as Europe, must be studied carefully to find out how the air pollution can be reduced to safe levels and, moreover, to develop reliable control strategies by which the air pollution can be kept under certain prescribed critical levels. This can be done only when large scale mathematical models, in which all physical and chemical processes are adequately described, are used. The numerical treatment of such models leads to huge computational tasks, which can successfully be solved only on high speed computers. An improved version of the Danish Eulerian Model, which can efficiently be run on several vector and parallel computers, have been prepared and tested. It was demonstrated that the new version runs efficiently on several different high speed computers. The possibility of applying successfully the improved version in a comprehensive study of a wide range of air pollution phenomena in a prescribed European region was demonstrated by carrying out long series of scenarios in order to investigate the influence of different variations of the emission sources to the corresponding concentrations and depositions. Some results from an extensive study, in which the improved version of the Danish Eulerian Model was used, will be presented in this paper.