Effect of uncertainty in source-receptor relationships on transboundary air pollution control strategies

Optimized emission reduction costs produced by the IIASA Regional Acidification Information and Simulation (RAINS) model were examined for robustness with respect to year-to-year meteorological variability and model uncertainty. It was found that optimized emission reductions to meet target depositions in a small target receptor area (Austria) were not especially robust with respect to meteorological variability because emission reductions are required mainly in the neighbouring countries to the extent where marginal costs, and the sensitivity of costs to changes in emission reductions brought about by meteorological variability, were great. When all of Europe was used as a target receptor, the robustness increased because the wide geographical separation of binding receptor points required moderate (and less sensitive) emission reductions and costs in a larger number of countries in which changing meteorology would have compensating effects. A chance-constraint method indicated that optimized costs in RAINS are not robust with respect to model uncertainty but it is felt that this method is unrealistically severe; an alternative test is proposed