Global NEST Journal, Vol 19, No 1, pp 37-48 Copyright© 2017 Global NEST Printed in Greece. All rights reserved Demirarslan K.O., Çetin Doğruparmak Ş. and Karademir A. (2017), Evaluation of three pollutant dispersion models for the environmental assessment of a district in Kocaeli, Turkey, Global NEST Journal, 19(1), 37-48. Evaluation of three pollutant dispersion models for the environmental assessment of a district in Kocaeli, Turkey Demirarslan K.O. 1,* , Çetin Doğruparmak Ş. 2 and Karademir A. 3 1 Department of Environmental Engineering, Artvin Çoruh University 08000 Artvin, Turkey 2 Department of Environmental Engineering, Kocaeli University, 41380 Kocaeli, Turkey 3 Department of Environmental Engineering, Kocaeli University, 41380 Kocaeli, Turkey Received: 21/01/2016, Accepted: 18/12/2016, Available online: 15/02/2017 *to whom all correspondence should be addressed: e-mail: onurdemirarslan@artvin.edu.tr Abstract Air Quality Modeling is a method used to manage urban air quality. Various pollutant dispersion models are available, and each of these models is characterized by its own advantages and disadvantages. Thus, we aimed to evaluate the advantages and disadvantages of the models and to determine their performance by applying them to a specific district. This study also enabled the determination of the contribution of pollution sources to the total pollution and the current air quality of the study area according to the selected pollutants. In this study, both steady-state models (the American Meteorological Society/Environmental Protection Agency Regulatory Model-AERMOD and the Industrial Source Complex Short Term Model-ISCST-3) and the Lagrangian model (the California Puff Model-CALPUFF) were used as the dispersion models. The Körfez district of Kocaeli was selected as the study area. SO2 and PM10 emissions were observed as pollutants. The statistical methods of mean squared error (MSE) and fractional bias (FB) were employed to evaluate the performance of these models. The results of the study revealed that the highest concentration varied according to the models and time options. However, when the modeling results for all of the sources were examined, the highest concentration was calculated by ISCST-3. The effect of the line source was less than the other sources (point and area). The contributions of the pollution sources differed according to each modeling program. The results of the statistical methods, which were used for evaluating the performance of the models, varied according to both the pollutant type and the time option. An overall ranking regarding modeling performance is as follows: CALPUFF > AERMOD > ISCST-3 for PM10 and ISCST-3 > CALPUFF > AERMOD for SO2. The MSE/FB results demonstrated that the predicted values were lower than the measured outcomes. Similarly, a comparison of the predicted and measured values with national and international limits revealed that various measures are necessary to reduce SO2 and PM10. Keywords: AERMOD, CALPUFF, ISCST-3, Körfez district, Pollutant dispersion modeling 1. Introduction The aim of dispersion modeling is to evaluate the concentrations (at the receptor points) of pollutants released into the atmosphere from any source. A dispersion model represents mathematical or physical relationships, which are established on scientific principles for the concentrations of the released pollutants. Information regarding air quality in a study area can be derived from the modeling results. Modeling results can aid in selecting the location of a new factory and designing the chimney of the factory and can be used to determine the areas that are affected the most by the discharge source and to execute different scenarios in these areas (AQMG, 2010; Silva et al., 2013; Clappier et al., 2015). Different methods are used to study pollutant dispersion. One method of modeling is the physical method in which the dispersion of pollutants is analyzed by recreating the environmental conditions in the laboratory. Another method involves mathematically modeling the pollutant dispersion (Markiewicz, 2008). Various scientific dispersion models are available, each of which is characterized by its own advantages and disadvantages. Three different dispersion models, the American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD VIEW 6.5.0), the Industrial Source Complex Short Term model (ISCST-3 6.5.0), and the California Puff Model (CALPUFF VIEW 5.8), were used to mathematically model the dispersion of pollutants. The Körfez district in Kocaeli was selected as the study area. SO2 and PM10 emissions were observed as the pollutants. The mean squared error (MSE) and fractional bias (FB) statistical methods were used to evaluate the performance of the selected models. This study was important for two reasons. Firstly, by evaluating three different dispersion models rather than one, their advantages and disadvantages could be