A methodology for development of flood-depth-velocity damage functions for improved estimation of pluvial flood risk in cities

Abstract

Globally, flooding is a persistent challenge in many rapidly urbanising cities. Effective flood risk management requires reliable and accurate approaches for quantifying potential flood damages, yet city specific flood damage functions are often unavailable in many cities. Due to this, estimation of flood damages in most data scarce cities is undertaken using global averaged flood − depth − damage functions and does not consider complex interactions between flood depth, velocity and urban form that influence the resulting flood damages. In this research, a new methodology that combines field questionnaire surveys, coupled 1D-2D modelling and polynomial regression was applied to derive three absolute flood damage functions that is; flood depth-damage, flood velocity-damage and flood depth-velocity-damage functions for a highly urbanised catchment in Kampala City, Uganda. The performance of the developed functions in describing the expected flood damage at given flood depths and velocities was evaluated using the coefficient of determination (R²), Akaike Information Criterion (AIC) and the modified Akaike Information Criterion (AIC_c). The study results suggest that the developed flood-depth-velocity damage function provides a more accurate means of estimating direct tangible flood damages when compared to individual flood depth-damage and flood velocity functions. Furthermore, the developed methodology incorporates flood flow velocity, a key factor in structural damage and loss of life during flooding conditions and thus provided empirical evidence that consideration of both flood depth and velocity parameters can lead to more accurate estimation of annual expected flood damage in cities.