Comparison of Gaussian and Lagrangian models for predicting pollutant concentrations

Abstract

Accurate modeling of pollutant dispersion is essential for effective air quality assessment. This study evaluates and compares the Gaussian Plume Model (GPM) and the Lagrangian Model (LM) in predicting ground-level sulfur dioxide (SO2) concentrations from an industrial stack. Both models were applied under identical emission and meteorological conditions over a 1 km × 1 km receptor grid. The GPM, based on a steady-state formulation, tended to overestimate concentrations near the source due to simplified turbulence representation. In contrast, the LM, formulated in a time-dependent framework, accounts for evolving wind fields and spatial variability in pollutant transport, resulting in a more accurate representation of dispersion behavior. At 500 meters downwind, the LM showed better agreement with observed data, demonstrating higher predictive accuracy. While the GPM remains advantageous for rapid regulatory screening, the LM offers improved performance under dynamic atmospheric conditions. These findings underscore the importance of selecting dispersion models based on application needs, balancing computational efficiency with predictive reliability