NATO Advanced Research Workshop
"ADVANCES IN AIR POLLUTION MODELLING FOR ENVIRONMENTAL SECURITY"

8 – 12 May 2004, Borovetz, Bulgaria

The NATO ARW on Advances in Air Pollution Modelling for Environmental Security was organized by Istvan Farago (Eotvos Lorand University, Budapest, Hungary) and Krassimir Georgiev (Institute for Parallel Processing of Bulgarian Academy of Sciences, Sofia, Bulgaria). This was an interdisciplinary workshop, where physicists, meteorologists, chemists, computer scientists and specialists in numerical analysis discussed different ways for improving the performance and the quality of the results of different air pollution models. The main objectives of this meeting were discussions on:

1. improving the abilities of the air pollution models to calculate in real time reliable predictions of the pollution levels in a given domain by using adequate description of the physical and chemical processes;
2. implementation of advanced numerical methods and algorithms in the models;
3. efficient utilization of up-to-date computer architectures;
4. development of mechanisms for studying particles (including here fine and ultra-fine particles), biogenic emissions, etc.;
5. optimization techniques in the study of the pollution levels, etc.

The total number of the participants was 46 including 20 Key speakers from 18 NATO and eligible partner countries. During the ARW they discussed the following main topics:

• preventive actions which are urgently needed, because some damaging effects may soon become irreversible including robust and reliable control strategies which must be developed in order to find out where and by how much the emissions of harmful pollutants should be reduced;
• economical problems which may appear when the task of optimal reduction of the emissions is not correctly solved;
• optimal (or nearly optimal) solution which can be successfully found only by carrying out long simulation experiments consisting of many hundreds runs of comprehensive mathematical models.
• non﷓realistic simplifying assumptions which are always made in all existing large﷓scale air pollution models in order to be able to treat them numerically on the computers available and if they steel absolutely necessary (because both the computers and the numerical methods are much faster now and will become even faster in the near future);
• need of parallel computations and even the need of using GRID computations;
• data assimilation and its implementation in large-scale models;
• need of reliable emission inventories and the improvement of the resolution of the emission inventories (both the spatial resolution and temporal resolution);
• splitting techniques which are commonly used in large-scale air pollution modelling, etc.

The workshop conclusion was that much closer connections and more extensive exchange of information among the specialists working in environmental modelling, numerical analysis and computer science are required in the efforts to improve the performance of the existing large scale environmental models as well as in the efforts to develop new models which are both more efficient and more powerful.