STUDY OF THE DIRECT AEROSOL EFFECT IN THE PARIS METROPOLITAN AREA USING A COUPLED METEOROLOGICAL - CHEMICAL TRANSPORT MODEL

摘要

The radiative effects of atmospheric aerosols have been recognised as one of the main modes of interaction between meteorological phenomena and atmospheric pollution. Aerosol radiative forcings on the surface radiation budget include the combined effect of scattering and absorption of solar radiation and have a profound impact on the total amount of absorbed radiation, as well as low-level atmospheric stability. Nowadays, two-way coupling of numerical weather prediction and chemical transport mesoscale and urban-scale models has been an area of active research. In an online coupled mesoscale model system, the interaction between pollutant fields and meteorology is usually implemented through the use of specialised submodules that aim to account for the contribution of aerosol feedbacks on the main radiative processes. In the present work, a coupled model system consisting of the mesoscale Eulerian meteorological model MEMO and the chemical transport model MARS-aero is applied to study the impact of the aerosol direct effect on the urban meteorology and the variation of urban air pollution. During the coupled model's operation, the aerosol concentration fields calculated by the dispersion model are introduced back as input in the enhanced radiation module of a three-dimensional non-hydrostatic mesoscale meteorological model by means of the OPAC (Optical Properties of Aerosols and Clouds) software library. For the assessment of the system's performance, simulations of meteorological fields and dispersion of atmospheric pollutants were performed for the Paris metropolitan area during the MEGAPOLI study period of summer 2005. The performance of the new formulation was assessed by evaluating the response of the primary meteorological variables governing the transport and dispersion of pollutants, as well as trends in particulate matter concentrations. The introduction of the direct effect leads to a consistent reduction of calculated mean wind speeds over most of the domain, while a notable decrease of the turbulent kinetic energy production at the first layer of the model is also evident. At the same time, PM_(10) concentration fields calculated by the coupled model system show a significant increase over almost the entire computational domain, compared to the standalone calculations, as well as higher loads in the southern part of the domain. Further evaluation of the coupled model's predictive skill was performed by comparing calculated PM_(10) concentrations with timeseries of measurements at several AIRPARIF stations, which indicated an improved ability of the coupled system in reproducing the variation of particulate concentrations throughout the area of study.