The use of empirical methods to predict landslide runout for use in rapid landslide risk asssessments following Cyclone Gabrielle, New Zealand

In February 2023, a severe weather event triggered widespread landslides across the coastal towns of Muriwai, Karekare and Piha, New Zealand, causing extensive damage and resulting in fatalities. In response, there was an urgent need to conduct area-wide landslide risk assessments on individual properties with respect to future debris flow hazards. This paper focuses on the development of rapid methods to predict landslide runout, using established empirical methods. With an inventory of 160 mapped landslides empirical-statistical relationships were established between runout distance, landslide volume, and downslope angle. Variation in data correlations ultimately necessitated the adoption of separate empirical models for Muriwai and Piha/Karekare.

These simple empirical-statistical models were combined with flow accumulation modelling to predict both landslide distance and direction. The method ultimately proved highly effective for the rapid delivery of risk assessments but highlighted the importance of using observational data, incorporating site-specific factors that influence landslide runout behaviour and applying a degree of judgement.