Numerical treatment of large-scale air pollution models (Q1104817)

Studying air pollution phenomena over a large space domain by a fairly general mathematical model is discussed. The space discretization of such a model leads to huge systems of ODEs. The integration algorithms used in the solution of these systems must be efficient. If a 3-D model is considered, then a vector processor is to be used. Moreover, the most time-consuming parts of the code are to be vectorized. The calculated results, concentrations of different air pollutants, must be reliable, because they have to be used by other specialists. Therefore different checks concerning the reliability of the results are carried out. In this paper it is shown that an efficient and reliable code for studying both sulphur and nitrogen pollution has been developed at the Danish Agency of Environmental Protection. Some simulation processes are also described briefly.