Quantifying effects of explicit representation of buildings in tsunami inundation simulations

Abstract

One of the critical steps in quantifying tsunami hazard is modelling of inundation. There are many factors influencing the complex inundation process, including the role played by features such as buildings affecting flow patterns which, if not simulated well, can affect the accuracy of tsunami intensity measure (IM) estimates (e.g. flow depth) and consequently, the reliability of outputs produced (e.g. economic loss estimates) by other models that utilise the estimated IM. In this paper, the effects of two alternative approaches to represent buildings in inundation modelling are investigated: (a) Equivalent Surface Roughness (ESR) approach, in which buildings and other ground surface features are removed and their effects on tsunami flows are approximated by spatially averaged surface roughness values; and (b) Explicitly Represented Building (ERB) approach, in which buildings are explicitly presented in the simulations to directly influence tsunami inundation. Using comparative simulations, this study quantifies the effects of explicit representation of buildings in inundation modelling on tsunami flow depth estimates and investigate how such effects consequently influence loss modelling results. We assess four plausible major earthquake-triggered tsunami scenarios for Napier, New Zealand. The modelling results suggest that the ERB approach better captures tsunami flow patterns and hence provides more realistic hazard and loss estimates. However, the difference in the estimated loss arising from potential direct damage caused by tsunami inundation (i.e. ERB vs. ESR) is shown to reduce in scenarios causing large-scale inundation, both in terms of severity and spatial extent, over the built-up area in the study region.