Despite the development of the parallel computing, simulating a flood lasting several hours over a large urban area still requires a significant computational time. Porosity-based models rely on the analogy between a city and a porous medium: the buildings are the grains, and the streets constitute the pores. Using such an approach, it is possible to reproduce the effects of the buildings on the flow, e.g reduction of the storage capacity and of the flow propagation speed, using very coarse computational meshes.
In our group, we have participated to the pioneering developments of porosity-based models, together with Prof. Vincent Guinot from the University of Montpellier.