Journal of Geophysical Research Atmospheres · 2008 · 52 citations · 28 references
Natural SciencesEarth SciencesEnvironmental SciencesAtmospheric ChemistryAtmospheric ScienceUncertainty QuantificationAir PollutantsAir Pollution ModelingAir QualityEmission InventoriesOzone ChemistryUrban Atmospheric ChemistryBayesian Model UncertaintyEmission ModelsMeasurement Uncertainty EvaluationModel UncertaintyMonte Carlo AnalysisObservational ConstraintsOzoneOzone Formation Modeling
The approach of Bayesian Monte Carlo analysis has been used for evaluating model uncertainty in ozone production and its sensitivity to emission changes. This approach has been applied to the ozone fields calculated by the CHIMERE regional model in the Ile‐de‐France area during the 1998 and 1999 summer seasons. The AIRPARIF network measurements of urban NO and O 3 concentrations and rural O 3 over the Ile‐de‐France region have been used for constraining the Monte Carlo simulations. Our results yield the following major conclusions: (1) The simulated formation of O 3 plumes are mainly located in southwestern to southeastern directions downwind the Paris area. (2) Uncertainties on simulated ozone concentrations and several derived quantities are evaluated and reduced using the BMC approach; simulated urban and plume ozone concentrations are enhanced through the observational constraint as compared to those obtained from the unconstrained model. (3) The chemical regime over the urban area of Paris and within plumes is clearly VOC sensitive on the average over two summers. This statement is robust with respect to the BMC uncertainty analysis.
28