Kinematic Hierarchical Filling-&-Spilling vs. fully 2D schemes for pluvial inundation

Urban areas are especially vulnerable to the effects of pluvial flooding due to the high population density and concentration of valuable resources as well as the fact that heavy precipitation events are becoming more common and more intense because of climate change. The use of high-resolution hydrologic-hydraulic numerical models for pluvial flood risk assessments in large metropolitan areas is still very resource-intensive. Several studies have pointed out the potential of fast-processing DEM-based methods, like Hierarchical Filling-&-Spilling Algorithms (HFSAs), considering the increasing availability of LiDAR (Light Detection and Ranging) high-resolution DEMs (Digital Elevation Models). We have developed a fast-processing HFSA, as part of a web-based digital twin solution for flood risk intelligence (see https://saferplaces.co/), that enables building-by-building pluvial flooding hazard and risk modelling in large urban areas by accounting for spatially distributed rainfall input and infiltration processes through a pixel-based Green-Ampt model. In this study, we upgrade SaferPlaces’ HFSA based on kinematic wave approximation to depression points and analysing the impact of flow contributions from one cell to another on the formulation of travel time, as well as the backwater effect in depression-related watershed areas. We present the first applications of the kinematic HFSA, comparing it with two state-of-the-art fully 2D rain-on-grid inundation models (i.e. HEC-RAS and UnTRIM). We discuss the potential and limitations of Saferplaces' upgraded HFSA and identify future research possibilities.