Lab Home | Phone | Search
Center for Nonlinear Studies  Center for Nonlinear Studies
 Colloquia Archive 
 Postdoc Seminars Archive 
 Quantum Lunch 
 Quantum Lunch Archive 
 CMS Colloquia 
 Q-Mat Seminars 
 Q-Mat Seminars Archive 
 P/T Colloquia 
 Kac Lectures 
 Kac Fellows 
 Dist. Quant. Lecture 
 Ulam Scholar 
 CNLS Fellowship Application 
 Student Program 
 Past Visitors 
 History of CNLS 
 Maps, Directions 
 CNLS Office 
Monday, May 18, 2009
10:00 AM - 11:00 AM
CNLS Conference Room (TA-3, Bldg 1690)


An adaptation of Multiscale Finite Element Method (MsFEM) to efficient simulation of Urban Transport of Pollution

Loubere Raphael
Université de Toulouse

In the context of an industrial accident or terrorism attack, which could take place in highly populated urban area, the need for efficient numerical simulation is of national importance. Efficiency has a twofold meaning: a real time simulation capability, and an accurate simulation in order to provide fast, precise and accurate materials for decision-makers.

Consider a physical model (convection/reaction/diffusion of passive pollutant) and a city of characteristic length L. The smallest detail to be taken into account has length l< In this talk we present the basics of the method and its adaptation to urban transport of pollution. Indeed the key point is the choice of these basis functions and, more precisely, the internal boundary conditions on the coarse cells that one has to "guess". We will show that an oversampling technique is mandatory to avoid numerical artifacts. Moreover we will show through a penalisation method that the mesh does not have to exactly fit the complex topography of the city: it may be a structured Cartesian mesh which further improves the efficiency of the method. Through 2D satelite image of urban area we will show how to:
- reconstruct an advection field (wind) given few meteorological data sites within the city, and
- simulate the transport of pollution after an emergency situation (explosion, accident).
Comparisons will be performed against the fine scale computation in terms of accuracy and efficiency.

Host: Mikhail Shashkov