Author(s): Xuefang Li; Sebastien Erpicum; Valentin Scheen; Emmanuel Mignot; Pierre Archambeau; Michel Pirotton; Benjamin Dewals

Linked Author(s): Xuefang LI, Sébastien Erpicum, Emmanuel Mignot, Pierre Archambeau

Keywords: Urban flooding; Laboratory measurement; Computational modelling; Urban layouts; Flow processes

Abstract: Urban planning is known to influence flood vulnerability but also flood hazard. Indeed, the layout of buildings in floodplains alters the flow paths, so that flood hazard is changed both in magnitude and in terms of spatial distribution (upstream and downstream effects). Based on computational modelling, Bruwier et al. (2018) analysed flooding in 2,000 realistic urban districts. They unveiled significant correlations between the magnitude of flood danger and geometric characteristics of the urban district (e.g. typical street length, width and curvature, building layout etc.). However, the studied configurations involve specific flow patterns for which neither field nor laboratory validation data exist. In the present research, we consider a more limited number of urban forms (i.e. layouts and geometric characteristics of buildings); but we aim to combine computational modelling and laboratory experiments. We consider a setup with two main streets along one direction and two main streets in the normal direction, forming four crossroads (with three- and four-branches). In the central area in-between these crossroads, a modular space is available in which buildings can be positioned to create various urban forms. We vary the number of additional smaller streets in this modular space as well as their relative width compared to the main streets, thus also the conveyance porosity of the central area. We additionally investigate the influence of the radius of curvature of the corners of the buildings. This is a parameter hardly analysed in previous studies of urban flooding which usually assume that all buildings have a rectangular footprint. We conduct numerical simulations, in which we vary systematically the controlling parameters (number and relative width of the smaller streets, radius of curvature of the building corners) and we assess their influence on flood danger throughout the urban district. Based on the outcomes of these simulations, a limited number of configurations will be investigated experimentally in a laboratory setup at the University of Liege. Water depths and discharge partition in-between the streets will be measured, as well as surface velocity using a LSPIV technique. The findings of this study will advance our understanding of the influence of urban forms on urban flooding and pave the way for guidelines towards more flood-resilient urban planning and emergency response. Additionally, the experimental dataset will prove valuable for the validation of various types of urban flood models (e.g., porosity-based models).

DOI: https://doi.org/10.3850/IAHR-39WC252171192022560

Year: 2022