Wolter, A.; Uma, S.R.; Mahmood, Y.; Liu, A.; Sadashiva, V.K.; Lin, S-L. 2026 Unravelling challenges to dwellings on slopes under climate change: Getting smarter with adaptive mitigations – Final Report. Lower Hutt, NZ. GNS Science report 2026/26. 47 p.; doi: 10.21420/S6MV-FK05
Abstract
Landslide hazards, and associated impacts and risks, are increasing as climate change triggers more frequent and intense rainfall events. Recent storms have shown that homes built on or below hillsides are especially vulnerable to these hazards. This vulnerability highlights several connected problems: (i) gaps in knowledge about how land and buildings respond to landslides; (ii) regulations that are unclear or insufficient and (iii) operational difficulties that lead to inconsistent practice. Climate effects and impacts therefore need to be considered within the whole site–building system. Research identifying which building elements are most vulnerable, and in what order, allows practical, sustainable and adaptable solutions to be developed for both existing and new homes. It is also important to understand the challenges faced by the sectors responsible for improving resilience so that barriers to implementation can be removed. Our exploratory research project used a dwelling-focused approach to produce evidence-based outputs that can support long-term change to landslide hazard and impact legislation, guidance and practice. First, end-user partners from several sectors were engaged from the beginning to co-design project scope and deliverables. Second, detailed field data were compiled from four recent weather events in Marlborough, Napier and Wellington. Although small in sample size, the uniquely detailed data sample combined Earth Sciences New Zealand field investigations during GeoNet emergency event responses and Natural Hazards Commission Toka Tū Ake (NHC) insurance claims and covered structural and land damage caused by debris inundation. Third, the data were analysed rigorously to rank the vulnerable building elements of typical timber-frame dwellings. Fourth, the study proposed conceptual solutions to improve the resilience and sustainability of hillside homes. Finally, the research outputs were linked to opportunities for long-term, cross-sector impact. Detailed analysis was completed on 14 landslides and 21 properties affected by debris inundation in Marlborough during 2021 and 2022 storm events, as these had the most varied and complete hazard and building damage data and are representative of damage due to rainfall-induced landslides. Statistical analysis determined that debris height (i.e. the maximum height of debris measured at a given building) is currently the most reliable measure of landslide-hazard intensity, given the small data sample analysed. Using debris height and other hazard-intensity metrics, such as debris volume and pressure, the study developed classification schemes with hazard levels ranging from low to extreme. A simple scoring method was then used to assess the observed damage level to building elements. The resulting damage scores showed a clear order of vulnerability. Retaining walls, where present, were the first to be affected by landslide debris, and hence were the most vulnerable. In our sample, most of these were not designed with landslide hazard in mind but were landscaping features.These were followed by non-structural building elements, such as windows and doors; then walls, roofs and foundations. Comparison case-study data from Napier and Wellington showed similar overall patterns of building-element vulnerability. Based on these results, the study proposed a conceptual framework for mitigation measures that considers improvements to both the site and dwelling itself. It also examined how sections 71–74 of the Building Act 2004, together with insurance requirements, may affect the practical implementation of these measures. The findings were shared through two workshops with impact partners from the research, engineering and geotechnical practice, regulation, local government and insurance sectors. Participants provided their sector-specific views on future research that could build on the current study and deliver practical impact. A cross-sector engagement exercise was also used to identify potential operational barriers, gaps and opportunities for improved coordination among organisations. The study recommends further technical research, stronger alignment between guidance and legislation and stronger collaboration and communication across all relevant sectors to support effective implementation of resilience measures for hillside homes vulnerable to landslides. The work may be used as a framework and basis for future similar projects, especially as more data become available. Ultimately, reducing the consequences of rainfall-induced landslides and related hazards, especially in the era of climate-change impacts, will require coordinated effort across disciplines. Our work represents a small piece of this larger ambition (auths)