FRESENIUS ENVIRONMENTAL BULLETIN, cilt.12, sa.10, ss.1173-1179, 2003 (SCI-Expanded)
Air pollution continues to be a major problem in many countries. Mathematical models are useful in relating emissions to air quality under a variety of meteorological conditions and source emission concentrations over an urban area. Meanwhile, the forecasting capability of sophisticated models is limited to very large and complex terrains. In this study, an adaptive neuro-fuzzy logic method has been developed to estimate the impact of meteorological factors on SO2 pollution levels. The model satisfactorily forecasts the trends of SO2 concentration levels with a performance between 78-90%.