Bressan M.,ARPAV Environmental Protection Agency of Veneto Region |
Elvini E.,ARPAV Environmental Protection Agency of Veneto Region |
Liguori F.,ARPAV Environmental Protection Agency of Veneto Region |
Pillon S.,ARPAV Environmental Protection Agency of Veneto Region
HARMO 2010 - Proceedings of the 13th International Conference on Harmonisation within Atmospheric Dispersion Modelling for Regulatory Purposes | Year: 2010
This work is addressing the decision making process involved in the selection of the most suited air quality model to assess the impact of a cement plant located within a complex terrain domain in Northern East of Italy. The overall objective is the description and analysis of operational issues arising from the case study which are among others: complexity of problem setting, assumptions and range of model applicability (pollutant types, spatial and temporal scales, orography description, wind calm treatment) and the input data availability. The use of a specific model and its setting up is always a trade off between consistency and accuracy of results. Two air quality models (CALPUFF vs ADMS-Urban) were used to simulate emissions from the kiln of a cement plant over a complex orography domain (6 Km × 6 Km). Both models were run under different configurations and sensitivity analysis of results was performed. On strict terms this 'field testing' was not a comparison of model performances but rather a cross-check of air quality impacts calculated by different modelling approaches. Hence a key issue was to evaluate at which point the differences between the model outputs make one model inadequate for evaluating impacts. Modelling exercise showed a discrete agreement when comparing long term averages but a strong contrast in short term outputs (high hourly percentiles). Analysis showed that a crucial choice in determining the differences of CALPUFF outputs was the alternative use of continuous parameterisation of the turbulence properties within the atmospheric surface layer as computed by the similarity approach or by the discrete characterization through the Pasquill-Gifford dispersion coefficients; the comparison of CALPUFF vs. ADMS showed a significant contrast in modelling outputs due to the different computational treatment of the similarity approach. Short term outputs are crucial for a regulatory point of view because they imply different compliance to air quality standards. The use of multiple models of varying complexities applied to the same case study allows useful insights into how sensitive results are to the different computational choices and how much trust should be put in the results from any one model. This also has important implications for interpretation of model results by final users (i.e. stakeholders, policy makers).