Effects of Topographic Variables on Traffic-related Pollutant Concentrations: Comparison of AERMOD and CAL3QHCR Models

Kazim Onur Demirarslan^*

• Artvin Çoruh University, Seyitler, Artvin, Türkiye

This study investigates the spatial dispersion of CO, NOₓ, and PM₁₀ emissions resulting from highway traffic along a major transport corridor connecting the Eastern Black Sea Region to other parts of northern Turkey. The primary objective is to compare pollutant concentration estimates produced by two atmospheric dispersion models-AERMOD and CAL3QHCRanalyze their behavior, and evaluate the influence of independent topographic variables on model outputs. Both models predicted higher pollutant concentrations near roadways; however, AERMOD consistently estimated higher maximum values (CO: 0.78 ppm; NOₓ: 1.48 ppm; PM₁₀: 26.59 µg/m³). In contrast, CAL3QHCR produced significantly lower estimates (CO: 0.20 ppm; NOₓ: 0.09 ppm; PM₁₀: 2.70 µg/m³), yet showed better agreement with observed concentrations for certain pollutants. Correlation analyses revealed strong negative relationships between pollutant levels and physical environmental factors-most notably between CO and elevation (r = -0.87) as well as road distance. These findings underscore the critical role of topography in shaping pollutant dispersion patterns. Both models also captured the decline in concentrations with increasing distance from the highway, with AERMOD showing steeper gradients within the first kilometer. To evaluate model performance, Spearman's rank correlation was employed to compare AERMOD and CAL3QHCR outputs, followed by validation against observations from ten monitoring stations. Statistical performance indicators-Fractional Bias (FB) and Normalized Mean Square Error (NMSE)were used for quantitative assessment. Results indicated that AERMOD tends to overpredict pollutant levels, while CAL3QHCR demonstrates better alignment with measured CO and NOₓ values. Nevertheless, both models showed poor performance for PM₁₀, characterized by high NMSE values and generally underestimated concentrations.