Executive Summary


New Zealand is comprised of two main islands, the North Island and South Island, with over 15,000 km of coastal area (Bell & Gibb, 1996). Located in the Pacific region, about a third of all New Zealand’s shallow earthquakes occur offshore. Subduction zones occur at a convergent boundary where one tectonic plate moves under another and sinks into the earth’s mantle. Submarine faults and subduction zones create earthquakes that can cause tsunami and pose serious hazards for coastal cities and towns. The east coast of New Zealand lies in close proximity to the Hikurangi subduction zone.

The last major tsunami to hit New Zealand, however, was the 1868 Peru-Chile tsunami, which caused substantial damage to the country’s infrastructure. Due to the historic infrequency of tsunami in New Zealand, natural hazard mitigation organizations such as the Institute of Geological and Nuclear Sciences (GNS), fear that public concern may be low. The recent disasters in the Indian Ocean regions, Samoa, and Japan have illustrated the importance of disaster planning and awareness to mitigate damage.

Wellington, the capital of New Zealand, lies on the south coast of the North Island and currently has developed an evacuation map (Figure A) suggesting various escape routes in the event of a tsunami. This map features the locations of the tsunami blue-line, which represents the maximum distance a tsunami will reach inland.

Figure A 2013 Project 6

Figure A – Island Bay tsunami danger zones. Grey land indicates safe elevation (Wellington Region Emergency Management Office; Wellington City Council, 2012)

The Wellington Region Emergency Management Office (WREMO) provides extensive information regarding earthquake and tsunami preparation on their website and through their “Get Ready Get Thru” public readiness program. The city also utilizes civil defense sirens to alert the public of potential tsunami. However, proper warning systems and evacuation plans are only as effective as residents are aware, willing, and able to adhere to them. Long or strong earthquakes occurring at the subduction zone and local fault lines can produce tsunami that can arrive in Wellington within minutes; which does not give local authorities enough time to utilize the warning system. Residents can learn that long or strong earthquakes can produce tsunami, and respond to the threat immediately, saving valuable time for evacuation.


The purpose of our research was to design and test a pilot study to assess the awareness and preparedness of the Greater Wellington Region on tsunami threat. The results of our research will contribute to the framework for a future, nation-wide study sponsored by GNS Science. Our research will also assist the Wellington Region Emergency Management Office (WREMO) to develop enhanced programs to improve emergency response.

Findings & Analysis


To accomplish our goal, we developed an optimal interview location guide (Figure B) that targeted communities with high tsunami risk. This included tsunami blue-line communities such as Owhiro Bay and Island Bay. Porirua and nearby high-risk Wellington Harbour communities, such as the Wellington Central Business District, Petone, and Lyall Bay, were also targeted.

Figure B 2013 Project 6

Figure B – Interview location guide

We implemented a sample of convenience to identify interview participants. We used this strategy to rapidly identify subjects from locations with high human foot traffic. In certain localities with low foot traffic, face-to-face interviews were conducted with residents at their homes. We conducted short (10-15 minute) face-to-face interviews with pedestrians in well- populated public areas in the targeted communities (Figure C). We designed the interviews to quickly cover demographic information as well as to generate in-depth responses. We explained to participants that they would remain anonymous and their responses would not be used to identify them. Once the interview was completed, we provided information sheets that outlined the interview goal and contact information of GNS representatives. Additionally, to improve our interview design, we requested feedback from participants who provided inappropriate or irrelevant responses.

Figure C 2013 Project 6

Figure C – Gaining a community perspective through public face-to-face interviews

Our interview included questions designed for open-ended responses. This inherently yielded varied responses from participants. In order to analyze our data quantitatively and qualitatively, we developed representative categories into which we could sort our responses. Based on our review of our data, we established a coding guide to ensure accurate categorization. The coding guide was developed by identifying key tsunami awareness concepts that we want to capture, and then reviewing raw responses and generating or modifying representative categories based on relevant themes. Once the guide was developed, the categories were reviewed a second time against all interview responses to ensure their accuracy.

We initially analyzed our data qualitatively by reviewing collected responses and forming an overall impression of the participant‟s awareness, preparedness, attitudes, and general knowledge. We formed these impressions for our entire sample as well as for specific communities that we were interested in understanding independently. We were able to form impressions based on the content and tone of participant‟s responses.

To support our qualitative analysis, our team developed a collection of algorithms to analyze our data quantitatively, using the Haskell Programming LanguageTM. We used these algorithms to analyze theme frequencies of coded data as well as identify correlations on all individual variables. After analyzing the frequencies, we used the information to produce graphs in Microsoft Excel to highlight main themes of our data set. The qualitative analysis was simultaneously supported by quantitative analysis to deduce implications, which were essential in generating recommendations. Throughout the analysis, we employed a grounded theory approach to develop a hypothesis about the underlying phenomena driving the observed trends. These hypotheses affected our interview content and strategy, which improved the quality and accuracy of our responses.


We then studied responses and evaluated the awareness of participants and compared the findings of separate communities. Specifically, it was important to understand if the presence of tsunami blue-lines affected risk perception and preparedness. Additionally, we compared social status and geography across communities in order to explain differences in awareness and preparedness.

Back to Top

Findings and analysis

Interview design and testing

The design and testing of the interview itself underwent several iterations that served to strengthen participant response. The interpretation of terminology in our interview varied depending on location and cultural background, requiring that we rephrase certain questions to some participants. Overall we encountered a general reluctance to be interviewed from the public. Cited reasons include interview length and suspected affiliations with religious and political organizations.

Categorization (coding) of responses

Originally attempting to use IBM’s SPSS packages to analyze data, we found manual categorization of responses to be more accurate and concise than SPSS Text Analytics. An iterative method of categorization through group input proved critical in developing complete and consistent categories by eliminating individual bias.

Analysis of responses from the GWR

Earthquakes and tsunami are the most-cited natural disasters from our research. However, most participants either failed to recognize the threat of tsunami due to local earthquakes or could not effectively recognize local warning signs of a tsunami. Since the recent occurrence of the devastating 2011 Christchurch earthquake, most individuals seem more concerned about this natural disaster rather than a possible tsunami. This was evident when our studies revealed that only half of the respondents expected a tsunami in their lifetime. Furthermore, less than 7% of respondents could recognize the natural warnings of a local tsunami: an earthquake too strong to stand during or an earthquake lasting more than a minute. Even for those interviewees who recognized earthquake warnings, most expected confirmation from an official source in the form of an alarm or radio before considering evacuation. Though an official alarm is unlikely to be sounded during tsunami generated close enough to arrive within an hour, over two thirds of those who expected an alarm warning also expected no more than 30 minutes of notice prior to impact.

While many would evacuate immediately if prompted, the most common source of delay was a desire to remain and assist others in evacuation. Cars were a common form of transportation in evacuation scenarios, with more respondents claiming to evacuate by car when given more warning time, which can cause dangerous traffic jams and delay evacuation. It appears that most participants are concerned about the risk of earthquake damage to roads, but do not recognize the danger of mass evacuation traffic.

There appears to be a general lack of tsunami awareness from individuals visiting coastal areas of New Zealand from more inland locations of New Zealand. Though individuals from more inland regions of New Zealand recognize tsunami as a potential hazard, they are much more focused on the earthquake threat and base their evacuation behavior around more immediate earthquake dangers such as falling objects and broken roads. Visitors from other countries generally do not identify tsunami as a threat and are unaware of tsunami warning signs.

Compare responses of communities within the GWR

Residents of blue-line communities such as Owhiro Bay and Island Bay were more acutely aware of tsunami threat than the residents of non-blue-line communities, though it was not evident that they were also better prepared. We found no significant difference in preparedness between blue-line and non-blue-line communities, with the exception that almost all blue-line residents knew of the blue-line program. Even in blue-line communities, there seemed to be confusion about the placement of the blue-line and general distrust of the program. The distrust of the program seems to have resulted from a misunderstanding of how the placement of the blue-line was calculated.

Additional Observations

While interviewing, we observed participants displaying varied levels of confusion regarding question phrasing and terminology between locations. Class and education may have played a role in understanding but overall, the issue of class distinctions is a complicated one, and we could identify no strong associations between wealth and tsunami preparedness regardless of confusion with terminology. It seems that the perception of tsunami risk and initial reactions to disaster derive from a deep human response, which is common between people.

We also feel that the inability of the inhabitants to recognize a tsunami threat when presented with the earthquake scenarios is likely because of two reasons. Disaster survivors doubt reoccurrences and those who experience frequent earthquakes are desensitized to their danger. The frequency of earthquakes and tendency to draw on past experiences are likely causing individuals to perform only standard earthquake evacuation behaviors, despite an earthquake‟s potential to induce a tsunami.

Through our observations, we have realized that individuals have been conditioned to expect confirmation of tsunami threat from authority. A possible explanation is that New Zealand residents do not trust their own instincts to judge natural warnings since they have not experienced a devastating tsunami induced by a local earthquake. By seeking confirmation, people feel more in control of the situation.
Back to Top


Based on our analysis, we have generated recommendations for GNS Science and WREMO. The recommendations are categorized under the following two tiers:

    • Recommendations for improving future interview/survey response rate and data quality
    • Recommendations for improving awareness and preparedness of tsunami in the GWR

These recommendations are derived from the analysis of data collected in this pilot study as well as available supplemental information and programs available to residents and visitors through WREMO’s website.

To improve survey/interview design and implementation

GNS, WREMO/CD (Civil Defence) logos could be present on surveys or interviewer clothing to help establish the credibility of the study. Clearly presenting logos also distinguishes the theme of the study.
Carefully worded survey questions are essential to the success of the study. A simpler format will avoid confusion or random guessing from participants. A sample of our final interview questions is listed in Appendix A: Version 2. Future surveys and interviews could omit the distinction between questions related to “earthquakes that last more than one minute” and “earthquakes that are difficult to stand in”. Participants did not provide distinct responses  between these two scenarios. This will decrease survey time and reduce the burden on participants.

Mail-in surveys can be used to conduct a larger study where face-to-face interviews are infeasible. We have developed suggestions for a pre-coded mail-in survey based on the responses to our original interviews.

To improve awareness and preparedness in the GWR

WREMO‟s effort to develop public awareness with the blue-line program is a strong commitment to tsunami awareness. Our work supports the idea that the agency should continue to implement blue-lines in communities. Our team discovered that blue-lines raise awareness and therefore more blue-lines would increase awareness. Additionally, WREMO could provide information to homes in blue-line communities or static information (such as signs) near the blue-lines explaining their purpose to resolve misconceptions.

Educating communities on alarm sounds and when they would be sounded could reduce confusion surrounding alarm usage. This can be performed though alarms clips provided on emergency websites as well as audio messages with an alarm preview sent to homes.

To help protect less informed waterfront visitors, emergency officials could work with Wellington city waterfront stores and restaurants to train employees on proper earthquake and tsunami response. Another idea is to develop official stickers on buildings in tsunami danger zones indicating if the building is tsunami safe (Figure D).Figure D 2013 Project 6

Figure D – A draft illustration of the “tsunami safe” building sticker

Given the success of the “duck, cover, hold” slogan, it could be beneficial for WREMO to develop a slogan in a similar vein that emphasizes the dangers of earthquakes that last more than one minute and earthquakes that are difficult to stand in. The slogan could possibly be “Long? Strong? Gone!” This recommendation has arisen from the overwhelming failure to recognize long and strong earthquakes as tsunami threats.

We recommend that WREMO continue to stress evacuation by foot or bicycle. Specifically, WREMO could include images indicating proper evacuation methods on tsunami warning signs or could designate specific cars as tsunami transportation vehicles. This recommendation has arisen from the large number of participants stating that they would flee by vehicle, especially in lengthier tsunami warnings. Figure E below shows an illustration of a possible sign we created.

Figure E 2013 Project 6

Figure E – A draft illustration signage conveying specific warning signs and transportation methods Image of bicycle and car inspired by (Map symbols bike clip art, 2013) and (Bednell holiday homes, 2013)


Lastly, WREMO could distinctly outline personal evacuation responsibility for residents so that they can feel confident in the actions they should take before evacuating. This suggestion has resulted from the numerous participants who stated that they were unsure of how much time they should spend assisting others before evacuating. GWR inhabitants could adopt a self- preservation belief similar to Japan‟s “tsunami tendenko” or another belief that appeals more to Kiwi values.

Back to Top


This study has confirmed low tsunami preparedness in the GWR. We have also piloted, tested, and revised questions for a greater survey to be distributed by GNS Science related to tsunami awareness and preparedness. Lastly, we have assessed the impact of education efforts such as the tsunami blue-line and its influence on communities‟ awareness and preparedness.

New Zealand‟s position on seismically active ground establishes the importance of conducting research to ensure the preparedness and safety of communities. Using what has been learned from this pilot study, a larger study could evaluate the awareness and preparedness of a greater portion of the GWR. Similar pilot studies can be conducted in areas where little is known about preparedness and awareness in order to prepare for a larger study. Once levels of awareness and preparedness have been measured, WREMO can develop or improve education programs to address the gaps in tsunami awareness and preparedness of GWR residents and visitors. Developing effective disaster education programs is vital in saving lives during earthquakes and resultant tsunami in countries with large coastal areas such as New Zealand.

Without proper education in threatened areas, individuals will not be able to react quickly enough to a locally induced tsunami, potentially leading to severe damage and loss of life as reflected in the 2004 Indian Ocean Tsunami. Improving education efforts has been proven to help mitigate the loss of life, as seen in the Great East Japan Tsunami in 2011. As research is conducted on tsunami awareness and preparedness, education programs can properly evolve to minimize the loss of life during natural disasters.

Back to Top