NSW-wide probabilistic tsunami inundation hazards from subduction earthquakes

Tsunamis are a low-frequency yet potentially high-consequence hazard, with impacts growing rapidly with the flow depth and inundation extent. For this reason, tsunami risk management often considers return periods from hundreds to thousands of years for which modelling uncertainties are substantial. Probabilistic Tsunami Hazard Assessment (PTHA) is a valuable framework for quantifying these hazards because modelling uncertainties can be integrated into the analysis by weighting alterative models. The last decade has seen major progress in “offshore” PTHA for earthquake generated tsunamis, with studies such as the 2018 Australian PTHA (PTHA18) providing deep-water models of hundreds-of-thousands of earthquake-tsunami scenarios and their uncertain occurrence rates. However, due to the computational expense of inundation models, it remains challenging to accurately translate offshore PTHAs into onshore tsunami hazards. This study presents a novel solution to this challenge. Tsunami inundation hazards are computed throughout NSW by combining PTHA18 with a NSW-wide inundation model, assuming a conservative background sea-level. The inundation model simulates tsunamis from the earthquake source through to inundation, resolving the entire NSW coast at 1/54 arcminute resolution (about 30 m). It is tested against observations of nine historical tsunamis at NSW tide gauges, and the numerical convergence is tested against a 1/162 arcminute model for a large inundation scenario. A novel Monte Carlo technique is developed to rigorously approximate the onshore tsunami hazards implied by PTHA18. The approach is computationally tractable because inundation simulations are only required for a random sample of PTHA18 scenarios. Multiple importance sampling allows focusing on scenarios that are hazardous in NSW, greatly increasing the accuracy without introducing bias. Consistent inundation hazard results are obtained from three separate Monte Carlo samples, using 370 random scenarios each, demonstrating the precision of the approach. The final results combine all three solutions, and demonstrate the feasibility of large-scale onshore PTHA. Presented at the 2025 Australasian Coasts & Ports Conference