Richard Perkins 教授 特別講演会

There are many situations in which it is necessary to model environmental flows in the presence of complex boundaries; typical examples of this include the dispersion of pollutants in the urban canopy, the dispersion of pollen in rural areas and the propagation of flood waves over complex flood plains and through urban zones. One of the current paradoxes in the modelling of these flows is that Geographical Information Systems now routinely provide a great deal of data about the boundary conditions, often in much greater detail than can be incorporated in computational models. The challenge is therefore to develop ways of exploiting this data which remain coherent and compatible with the modelling techniques that will be used to compute flow and dispersion. This requires a much greater understanding of the interaction between complex terrain and the flow over it.

Recent research at the LMFA has lead to the development of a series of nested operational models for predicting the transport and dispersion of pollutants in the urban canopy. These models are based on an analysis of the dominant physical processes at different scales. In this seminar I will present the different models and show how they have been developed using a combination of wind tunnel experiments, numerical simulations and theoretical analysis. Their use will be illustrated through their application to a particular real problem ? modelling urban air quality in one of the arrondissements in the Lyon agglomeration. The modelling results agree well with data measured by a pollution monitor in the district, and a detailed comparison illustrates the necessity of combining models at different scales in order to provide a reliable and useful estimate of pollutant concentrations.

The first part of the seminar will be devoted to a general presentation of the laboratory, and to current research activities in the environmental fluid mechanics group.