Langridge, R.M.; Wolter, A.; Stahl, T.; Underwood, A.; Dellow, G.D.; Potangaroa, J.; Potangaroa, P.; Seebeck, H.; Andrews, J.R.; Rosser, B.J.; Lukovic, B.; Bloom, C. 2026 Large landslides as ground-motion calibrators in the Hikurangi margin. Lower Hutt, NZ: Earth Sciences New Zealand. GNS Science report 2025/39. 28 p.; doi: 10.21420/N35K-AT10
Abstract
This report summarises the work and outcomes of the Ministry of Business, Innovation & Employment (MBIE) Smart Ideas project entitled ‘Large landslides as ground motion calibrators in the Hikurangi margin’ undertaken between 2022 and 2025. The 2022 National Seismic Hazard Model of Aotearoa New Zealand (hereafter referred to as ‘NSHM 2022’) forecasts elevated ground motions across much of Aotearoa New Zealand, and the Wairarapa area is in one of the highest regions. An important aim of this project was to try to calibrate the NSHM 2022 ground motions against landslides in the uplifted forearc of the Hikurangi subduction margin to better understand future earthquake-induced geohazard scenarios. Investigation of historical earthquake-induced landsliding is addressed through a combination of mapping, field studies and modelling of earthquake sources in the Wairarapa region. Regional airborne lidar data was used to map the distribution and density of moderate to large landslides across the eastern Wairarapa, focusing on the eastern coastal hills and coastline. The study area is dominated by uplifted Upper Cretaceous to Pliocene marine rocks of the uplifted forearc and accretionary prism, which typically are weak and prone to landsliding. Detailed mapping covered priority areas and highlighted concentrated areas of large landslides along active fault corridors. Field studies focused on using paleoseismic techniques to investigate a historical (post-1840 CE) co seismic landslide complex, another pre-historic (pre-1840 CE) landslide along a major active fault corridor (both near Tauweru) and landslides along the eastern coastline. Trenching and dating of the largest slope movements from the 24 June 24 1942 Wairarapa I (Masterton) earthquake at Te Matai farm indicated that the headscarp had previously failed during the Mid-Holocene, while other parts of the south-facing hillslope had failed as recently as 346–486 calendar years before present. Excavation of two trenches at Hapua farm provided evidence for landslide initiation at the Pleistocene–Holocene boundary (c. 11,700 years ago) and for multiple Holocene re-activations of the Hapua landslide. These were identified using proxy indicators such as headscarp (‘fault’) movements, post-event sediment burial (of paleosols) and liquefaction. Drilling was undertaken to intercept the slide plane of the Hapua landslides. Reconnaissance studies of coastal landslides are ongoing and show potential for dating landslides, along with correlation to coastal uplift from the offshore Palliser–Kaiwhata Fault and Hikurangi subduction earthquake proxies. Earthquake fault sources were updated from the New Zealand Community Fault Model (Seebeck et al. 2024) from existing and recently published fault mapping to test the shaking potential ofvarious earthquake scenarios through peak ground acceleration (PGA) maps in combination with an earthquake-induced landslide (EIL) tool. Earthquake sources were developed for the Carterton and Palliser-Kaiwhata faults, for the southern Hikurangi subduction source and for bespoke sources for the 1942 Wairarapa I and II earthquakes. For the bespoke Wairarapa I source, we found that the earthquake is best explained as a MW 6.9–7.0 earthquake occurring along a NNE-striking reverse dextral fault, extending north from the Huangarua Fault towards Tauweru. Rupture of the ENE-striking Carterton source, which was much closer to Te Matai, could not generate sufficient PGAs and thus landslide potential through the EIL tool, nor match the known landslide distribution from the 1942 earthquake sequence. Rupture of the Palliser–Kaiwhata fault source would uplift and shake the Wairarapa coastline and could therefore be a primary driver of the landslide distribution along the coast. Three Hikurangi subduction interface sources were also modelled. Two smaller interface patch sources generated relatively low PGAs (~0.3 g), while a full rupture of the southern Hikurangi interface generated high PGAs across the Wairarapa east of the main divide. In terms of origin, a problem remains thatwe cannot distinguish landslides triggered by something other than seismic shaking where there is no earthquake source, for example, rainfall-induced landsliding. The project has benefited greatly through partnering with Potangaroa Education Ltd to engage with school, kura and university classes, as well as with public groups and through the Ahunuku summer research scholar programme. In particular, many schools and groups were taken to the Hidden Lakes landslide in northern Wairarapa to consider the impact of the historical 1855 Wairarapa earthquake on iwi living at that location. (auths)
