Massey, C.I.; Jones, K.E.; Kundu, J.; Hamling, I.J.; Lukovic, B.; Kellett, R.L.; Thingbaijam, K.K.S.; Sadashiva, V.K.; Farr, J.A.; Lin, S-L.; Townsend, D.B. 2025 Identifying and forecasting earthquake- and rainfall-induced movement patterns of large, populated landslides: a Nelson case study. Lower Hutt, NZ: Earth Sciences New Zealand. GNS Science report 2025/12. 141 p.; doi: 10.21420/ET1G-0474
The Tāhunanui ‘slump’ is one of many large, active and populated landslides in New Zealand. Numerous large, populated landslides exist around New Zealand, and many are only now being identified given the increasing availability of high temporal and spatial resolution InSAR and other remotely sensed datasets. Both long periods of wet weather, which lead to increased pore water pressures and reduced effective stresses, and earthquake shaking are known to re-activate large landslides. As our climate changes, extreme variations in weather are more likely to occur, including more frequent periods of wet weather, droughts and high-intensity rainfall events. The re-activation of the Tāhunanui landslide in response to the August 2022 rainfall event, as well as the availability of historical and recent monitoring datasets relating to it, provide an opportunity to use this landslide to investigate how these dormant and/or episodically active slow-moving landslides may respond to a changing climate; how strong earthquake shaking might impact their movement patterns given these changing climate conditions; and how such movements will affect people, buildings and infrastructure exposed to them. To answer these questions, regional remotely sensed datasets have been blended with InSAR displacements and landslide-specific monitoring datasets. Regional statistical- and site-specific physics-based simulations have been carried out to link landslide occurrence and displacement to rainfall-induced changes in pore water pressure and earthquake-induced ground shaking. The Tāhunanui landslide is an outlier in both location and size compared to other landslides mapped in similar materials (Port Hills Gravel) within the Nelson region. Most of these landslides cluster towards the Flaxmore and Waimea faults, indicating that their initiation may possibly have been related to earthquakes associated with these faults. Most of the other known large landslides appear to be at least partially dormant, based on the InSAR displacement analyses (2018 to present) and their apparent lack of response to the August 2022 rainfall event. In contrast, movement of the Tāhunanui landslide is episodic: it is known to have re-activated at least six times between 1929 and 2022, as well as more recently in 2024. Such periods of re-activation have been linked to 60- to 120-day periods of rain comprising 5–13 days of consecutively wet days. The annual exceedance probabilities for the 10- and 20-year average recurrence intervals of the five-day rain amounts that have triggered past displacement of the Tāhunanui landslide are likely to increase because of climate change. This means that the landslide might re-activate more frequently than in the past. Given that the slope has not historically been subjected to strong earthquake ground-shaking (typically greater than a modified Mercalli intensity [MMI] of 8), the effects of earthquakes on the landslide and region are largely unknown. Earthquake-induced displacements of the Tāhunanui landslide may not occur frequently, due to the low annual exceedance probability of earthquakes able to generate large ground motions, but when these do, the associated permanent ground displacements of the landslide are likely to be larger than those historically associated with rain. Conversely, rainfall-induced displacements occur frequently but at lower rates to those caused by earthquakes. If such displacements were to occur, then it might be possible for larger areas of the coherent landslide mass to break down – and the material properties to reduce in strength – allowing the movement mechanism to transition from coherent sliding to flowing/avalanching. Such changes could generate larger and more rapid landslides, which could impact exposed people and infrastructure (auths)
