Cyclone Gabrielle landslides, New Zealand

Between 12-16 February 2023, an extreme rainfall event, referred to as Cyclone Gabrielle affected much of the northern and eastern North Island of New Zealand, causing widespread damage. It was a severe event that required a national level response. States of emergency were declared for seven regions of New Zealand. Along with surface, coastal, and river flooding, the event triggered more than 100,000 landslides, destroying infrastructure and lives across the North Island. From discussions with stakeholders it became apparent early on that they wanted to know – with a good level of spatial and positional accuracy – where landslides triggered by Cyclone Gabrielle had occurred, and where future landslides could occur in other rain events, which may pose a risk to life and/or lifeline infrastructure. With guidance from central and local government agencies, scientists, and engineers at GNS Science, Manaaki Whenua, University of Canterbury, and the University of Auckland developed a systematic workflow and methodology to map the landslides using available pre- and post-event aerial photography and satellite imagery covering the affected regions. The approach aimed to achieve rapid systematic collection of landslide data along with regular dissemination to end-users, prioritizing the mapping in areas where people and lifeline infrastructure are at risk from landslides. This presentation provides a description and explanation of the workflow and methodology, the data used and the outputs. The main output from this work is a spatially accurate landslide inventory that represents the landslides – and their selected attributes – triggered by Cyclone Gabrielle. Landslides were mapped using orientated poly lines, where the first point of the line represents the centroid of the source area and the polyline represents the centerline of the debris trail. A series of landslide attributes were collected that provided additional information on landslide movement types, size, materials, mapping certainty, and quality control metadata. This rich dataset is now being used to develop landslide susceptibility, runout and risk models, which underpin the post-event recovery decision making. USGS video: