Keywords

flash flood, shallow water, multi-scale, adaptive grid, quad tree

Location

Session A1: Environmental Fluid Mechanics - Theoretical, Modelling and Experimental Approaches

Start Date

12-7-2016 3:50 PM

End Date

12-7-2016 4:10 PM

Abstract

We propose an open source solver to model flash floods, using the well-known shallowwater equations, in 2 dimensions. The full resolution of such equations have usually a high computational cost, so that majority of flood simulation softwares used for flood forecasting is using a simplification of this model, which is not relevant in the case of flash flood, where torrential flows are commons. To reduce drastically the cost of such 2D simulations, we propose a 2D shallow-water flow solver built with the open source code Basilisk (http://basilisk.fr/) which is using adaptive refinement on a quad-tree grid. We reproduce the flood of Cannes (France) which happened in October 2015 and which led to huge damage, human and material. We show that the simulation is faster than the event. We also show that our simulation predicts accurately the areas harshly touched by the flood.

 
Jul 12th, 3:50 PM Jul 12th, 4:10 PM

Modeling the flash flood of Cannes (Fr) with Basilisk

Session A1: Environmental Fluid Mechanics - Theoretical, Modelling and Experimental Approaches

We propose an open source solver to model flash floods, using the well-known shallowwater equations, in 2 dimensions. The full resolution of such equations have usually a high computational cost, so that majority of flood simulation softwares used for flood forecasting is using a simplification of this model, which is not relevant in the case of flash flood, where torrential flows are commons. To reduce drastically the cost of such 2D simulations, we propose a 2D shallow-water flow solver built with the open source code Basilisk (http://basilisk.fr/) which is using adaptive refinement on a quad-tree grid. We reproduce the flood of Cannes (France) which happened in October 2015 and which led to huge damage, human and material. We show that the simulation is faster than the event. We also show that our simulation predicts accurately the areas harshly touched by the flood.