Recent tsunami disasters around the world have demonstrated the necessity for effective tsunami risk management. Evacuation is considered the most effective tsunami risk management measure for preventing injury and death. However, empirical research to date has focused on specific aspects of the tsunami evacuation response process, such as whether people evacuated or not, instead of seeking to understand behaviour across the full process. Existing studies have also focused mostly on urban areas with prior experience of tsunami events as this provides reasonable grounds for prioritised science research funding. This means there is a lack of understanding concerning tsunami hazard risk awareness and evacuation intentions in highly exposed remote coastal communities that are yet to experience a tsunami. Further research is required to better understand 1) evacuation behaviour and intentions across the full tsunami evacuation response process, from warning response and pre-evacuation actions to the evacuation dynamics and the process of returning home, and 2) tsunami hazard awareness, preparedness and evacuation intentions in highly exposed remote rural coastal communities that are yet to experience a tsunami.

This thesis contributes to filling these two global research gaps using a New Zealand case study. The overall aim of this thesis is to increase understanding of tsunami hazard risk awareness and evacuation intentions in highly exposed remote rural coastal communities to support increased rates of successful evacuation in future tsunami. This involves addressing three research objectives. The first is to synthesize the existing knowledge of tsunami hazard and risk management and evacuation behaviour and intentions in remote rural coastal communities, with a focus on the North Island, New Zealand. A review of relevant literature was undertaken to collate this knowledge. The second is to develop and refine a survey tool for assessing tsunami hazard risk awareness and evacuation intentions in remote rural coastal communities. Survey questions asked about prior tsunami hazard awareness, preparedness actions undertaken and intended evacuation response to three scenarios (local, regional and distant source tsunami), including expected warning sources, time available to move to safety, pre-evacuation actions, evacuation route and intended transport mode. Finally, the third research objective is to use a Wairarapa case study to test the refined survey tool to better understand tsunami hazard risk awareness and evacuation intentions in these remote rural coastal communities. The survey responses were analysed to identify trends in the data and potential factors influencing awareness, preparedness and evacuation intentions. Key findings from this research were compared to those of existing studies to identify areas of improvement, decline and consistency.

A total of 415 surveys were distributed to Wairarapa rural communities: Riversdale Beach, Castlepoint, Lake Ferry and Cape Palliser areas, with 87 in total returned (21% response rate). Several key findings were identified from the results of the survey. The positive impact that the recent implementation of tsunami blue lines had on awareness and evacuation intentions was evident throughout the survey responses. However, preparedness appeared to be relatively low, and three key findings of concern were identified from the survey results, including: • Almost half of the respondents (49%, n = 38) thought they would have ten or more minutes available to move to safety from a tsunami, meaning this portion of respondents did not intend to evacuate within the recommended ten-minute natural warning evacuation threshold for local source tsunami. This suggests that there is a heightened risk to life for this group in a local source tsunami event. • Awareness of natural warnings for local tsunami events was higher than it had been in existing studies (73%, n = 58). However, the level of awareness is still inadequate given the life safety risk that is posed by not recognising natural warning signs. This was coupled with an ongoing reliance on official warnings, particularly tsunami sirens (61%, n = 49), for local source tsunami events. • There is an ongoing reliance on vehicles for evacuation in local source tsunami (72%, n = 56), suggesting that many respondents had not considered the risk of traffic gridlocks and congestion in these small rural communities when there are population influxes during the summer months.

These findings are consistent with existing literature and remain prevalent despite the national public education campaign rolled out over recent years to address these issues. They highlight a serious lifesafety risk to remote rural coastal communities that are highly exposed to local source tsunami, which can arrive within minutes, such as those in Wairarapa.

A set of recommendations and recommendations for future work are presented at the close of this thesis for future tsunami risk management and scientific research. Key recommendations include: • consider additional measures alongside standard public education measures to focus on key issues that are not being addressed by current public education initiatives. Additional measures should reach enable highly exposed remote rural communities and encourage their engagement and participation; • foster more community-led, bottom-up preparedness initiatives that focus on rural coastal communities, particularly during the summer months when many semi-permanent residents are most likely to be present in the communities and available to participate; • implement tsunami blue lines in all highly exposed rural coastal communities, particularly those that are still without blue lines in Wairarapa. Continue to monitor their effectiveness past the early implementation phase. • evaluate criteria used to allocate funding for science programmes and emergency management initiatives to ensure funding is going towards those who are most at risk. • undertake further scientific research using alternative data collection methods in remote rural coastal communities, to better capture the diverse views and behaviour of whole communities, and; • prioritise partnership with local iwi for future research in Wairarapa and other areas of New Zealand.

From an emergency management perspective, the findings from this thesis indicate gaps in tsunami hazard risk awareness and preparedness that are not being addressed by current public education measures, to inform future education initiatives. This insight is particularly useful for underresearched, high-risk remote rural coastal communities. From a science perspective, this thesis contributes to the gap in research undertaken in highly exposed remote rural coastal communities and is one of few studies that seeks to understand individuals’ full tsunami evacuation response. It builds on the methodology used for undertaking survey research in rural communities and contributes to the nationally consistent dataset built over many years of New Zealand tsunami survey research.