Nat Hazards (2008) 45:99–122 DOI 10.1007/s11069-007-9172-8 ORIGINAL PAPER
Testing the use of a ‘questionnaire survey instrument’ to investigate public perceptions of tsunami hazard and risk in Sydney, Australia Deanne Bird Æ Dale Dominey-Howes
Received: 21 December 2006 / Accepted: 24 July 2007 / Published online: 6 November 2007 Springer Science+Business Media B.V. 2007
Abstract The Indian Ocean tsunami (IOT) of December 2004 has demonstrated that the coasts of Australia are vulnerable to tsunami flooding. As a consequence of the IOT, the Australian Federal Treasurer announced in 2005 that the Bureau of Meteorology and Geoscience Australia will jointly develop and implement the Australian Tsunami Warning System. Effective response to tsunami warnings is highly dependent on public awareness and perception of tsunami hazard and risk. At present, no efforts have been made to investigate and publish public awareness of tsunami hazard and risk and as such, emergency managers have little idea of the likely challenges to effecting appropriate tsunami risk management. We develop a short questionnaire survey instrument and trial that instrument in order to investigate its suitability for generating information about the perceptions of tsunami hazard and risk in the Sydney region. We found that the design, layout and format of the questionnaire were suitable for our purpose and should be useful for generating information appropriate to emergency management agencies tasked with the responsibility of developing tsunami education campaigns and risk mitigation strategies in Australia. However, certain limitations, such as individual question design and format, should be considered before a much larger survey of various stakeholders is conducted. Keywords Tsunami risk Sydney Public perception Risk mitigation Questionnaire survey instrument Questionnaire structure
D. Bird (&) Department of Physical Geography, Macquarie University, North Ryde, Sydney, NSW 2109, Australia e-mail:
[email protected] D. Dominey-Howes Risk Management Research Group, School of Safety Science, University of New South Wales, Sydney 2052, Australia e-mail:
[email protected]
123
100
Nat Hazards (2008) 45:99–122
1 Introduction and aims The Indian Ocean tsunami (IOT) disaster of December 2004 was significant not just because of the number of lives lost but because it demonstrated that large, regionally destructive tsunami are not confined to the Pacific Ocean. For countries with coastal landscape areas bordering the Indian Ocean (including Australia), the effects of the IOT provided a stark ‘wake-up call’ concerning the actual threat posed by tsunami—a hazard not fully realised until now. As a consequence of the IOT, the Australian Federal Treasurer announced in 2005 that the Bureau of Meteorology (BoM) and Geoscience Australia (GA) would jointly develop and implement the Australian Tsunami Warning System (ATWS) (Geoscience Australia 2005). GA and the BoM will co-ordinate the detection of tsunamigenic events, monitor tsunami as they approach Australia and undertake forecast assessments of probable tsunami impacts (Greenslade et al. 2007). These agencies will evaluate what type (if any) of information and/or warning messages should be issued. When required, the BoM will release initial warning messages to State or Territory emergency service organisations and the media through the Bureau’s regional offices. Messages will also be passed to EMA’s National Emergency Management Coordination Centre (NEMCC) (Sullivan 2006). Where required, the BoM Regional Offices will issue detailed warning messages to the SES and the public. Meaningful public response to any evacuation order is partly dependent on the clarity and accuracy of those orders; the time available prior to tsunami arrival; the efficiency of the coordinating emergency services and significantly, upon the public’s understanding and perception of (tsunami) hazard and risk (Hurnen and McClure 1997; Johnston and Benton 1998; Anderson-Berry 2003; Dominey-Howes and Minos-Minopoulos 2004; Johnston et al. 2005). Therefore, having a clear understanding of the public’s awareness and perception of tsunami is a vital element in developing hazard education programmes and risk management procedures (Hurnen and McClure 1997; Johnston and Benton 1998; Gough and Hooper 2003). In order to obtain information about what the public understands in relation to a particular hazard type, it is necessary to design and execute appropriate information gathering techniques such as questionnaire surveys (McGuirk and O’Neill 2005; Parfitt 2005). In light of the introduction, the aims of this article are to: (1) (2)
(3)
(4) (5)
Provide a short overview of the tsunami hazard and risk to, and exposure within, the Sydney region, Australia; Briefly review the importance of public awareness and perception of hazard and risk and consider the role of education in the mitigation of risk and vulnerability; Develop a short questionnaire survey instrument and trial that instrument in order to investigate its suitability for generating information about the perceptions of tsunami hazard and risk in the Sydney region; Report the findings of this trial; and to, Consider the significance of our results to future studies of tsunami risk in Australia and the emergency management community and make recommendations to improve our understanding of public awareness of tsunami hazard and risk.
123
Nat Hazards (2008) 45:99–122
101
2 Identifying tsunami hazard, risk and exposure in the Sydney coastal region Tsunami affecting Australia since European occupation have all been rather small, causing little in the way of damage (Risk Frontiers 1995; Rynn and Davidson 1999; DomineyHowes 2007). In contrast, geological research since the late 1980’s suggests that the coast of SE Australia has been repeatedly impacted by palaeo-(prehistoric) tsunami—some of them very large (Bryant 2001; Bryant and Nott 2001; Bryant and Young 1996; Bryant et al. 1992a, b; Kelletat and Scheffers 2003; Nott 1997, 2003a, b, 2004; Young and Bryant 1992; Young et al. 1995, 1996). In fact, Bryant and Nott (2001) report tsunami flood heights south of Sydney to heights in excess of 100 m above sea level (m asl) with waves sweeping as much as 10 km inland. Whilst much more research needs to be conducted on the evidence for the reported palaeo-tsunami in order to increase our confidence in their occurrence and likely effects (Dominey-Howes 2007; Dominey-Howes et al. 2006), it is at least clear that historic tsunami have inundated the eastern seaboard (Rynn and Davidson 1999). Work by Gusiakov (2005) is useful in defining regions within the Pacific that can generate tsunami potentially damaging to eastern Australia. Gusiakov examined the total number of earthquakes with a magnitude of Ms C 7.0 and a depth h \ 100 km that occurred within the Pacific basin between 1901 and 2000. He then compared these events with those that generated a tsunami. This allowed him to calculate a Tsunami Efficiency ratio (TE%) for each tsunamigenic region. Gusiakov’s tsunamigenic regions of the Pacific are shown in Fig. 1. The TE% ratios are shown in Table 1. South America is the region most efficient at generating tsunami (followed by Indonesia and the Philippines). New Zealand–Tonga and the Kuril–Kamchatka regions are least efficient at generating tsunami. There are however, two significant limitations to using these data. First, the analysis of events between 1901 and 2000 do not adequately reflect the total cumulative tsunami risk for that period because the tsunamigenic events only relate to earthquakes and do not include other mechanisms such as volcanic eruptions and submarine sediment slides (Dominey-Howes and Keating 2005). Second, the approach used by Gusiakov may not properly reflect the long-term earthquake-tsunami risk (i.e. longer than 100 year period) within different regions of the Pacific because his analysis is only based on a short time series of data from 1901 to 2000. It may be that over longer time periods, other geographic regions dominate in terms of their tsunamigenic recurrence. However, at the present time, the work of Gusiakov (2005) is the most comprehensive (and publicly available) analysis of tsunamigenicity in the Pacific so we use his ‘tsunamigenic regions’ in our study. Table 2 details return periods of tsunami for selected areas around the Pacific. From Fig. 1 and Tables 1 and 2, it is apparent that Sydney is at risk from tsunami generated in many of these regions—particularly those across the Pacific. Indeed, the 1960 Chile tsunami and the 1964 Alaska tsunami both affected the NSW coast including Sydney generating wave heights at shore of more than 1 m (Opper and Gissing 2005; Dominey-Howes 2007). Approximately 330,000 people in New South Wales live within 1 km of the ocean or a coastal river and at an elevation of no more than 10 m asl (Molino Stewart 2005; Chen and McAneney 2006). More than 20% of these people are over the age of 65 years (Opper and Gissing 2005). Within the Greater Sydney region specifically, there are a total of *2.5 million addresses (residential, commercial, industrial etc) and of these, 46,000 are located \1 km from the coastline and at an elevation of \3 m asl (Fig. 2) (Risk Frontiers 2005; Chen and McAneney 2006). Whilst these 46,000 addresses are not uniformly
123
102
Nat Hazards (2008) 45:99–122
Fig. 1 The epicentres of tsunamigenic earthquakes that occurred in the Pacific between 1981 and 1997 are represented by the circles. The size of the circle indicates the magnitude of the earthquake and colour variation denotes intensity. Based on the clustering of these events, Gusiakov (2005) identified what he referred to as ‘tsunamigenic regions’ of the Pacific. The boundaries of these regions are shown in black. Whilst this figure only shows earthquake events between 1981 and 1997, the regions were based on a much longer record of events from 1900 to 2000. The regions are: Alaska–Aleutians (A–A), Central America (CAM), South America (SAM), New Zealand–Tonga (NZT), New Guinea–Solomon Islands (NGS), Indonesia (IND), Philippines (PHI), Japan (JAP), Kuril–Kamchatka (K–K) and Hawai’i (HAW) (after Gusiakov 2005)
vulnerable to tsunami damage (Papathoma and Dominey-Howes 2003; Papathoma et al. 2003; Dominey-Howes and Papathoma 2007), they do at least represent what is ‘exposed’ to tsunami inundation and damage.
3 Public awareness of hazard in risk communication and education The physical occurrence of a tsunami cannot be prevented, but a tsunami disaster can be. Disaster management comprises a variety of elements that are undertaken before, during and after a (tsunami) disaster event. These elements are illustrated in Fig. 3. Embedded within and central to the disaster management cycle, is communicating with and educating the public and others about tsunami hazard, risk, vulnerability and disaster preparedness (Kurita et al. 2006). Risk communication is undertaken to persuade people, formally or informally, to adopt self-protective behaviours and practices. Residents exposed to a tsunami threat need to be well informed about the various types of risk management and mitigation procedures available to them (Rohrmann 2003).
123
Nat Hazards (2008) 45:99–122
103
Table 1 Total number of earthquakes (EQ) (Ms C 7.0 and depth, h \ 100 km) and tsunami (TS) in each region (between 1901 and 2000), and the tsunami generating efficiency (TE%) of these earthquakes (after Gusiakov 2005) Region
Earthquake (EQ)
Tsunami (TS)
South America
Tsunami efficiency (TE) %
122
102
84
Indonesia
86
68
79
Philippines
73
55
75
New Guinea–Solomon Islands
130
86
64
Central America
112
62
55
Japan
255
123
48
Alaska–Aleutians
108
49
46
Kuril–Kamchatka
150
68
45
New Zealand–Tonga
162
62
38
3
13
100
Hawaiia a
Hawai’i has experienced many locally generated tsunami but only three far-field tsunami of earthquake origin, which skews the data. The TE is 100% which says little about the efficiency of earthquake generating tsunami in the Hawai’i region
Table 2 Tsunami frequency/recurrence intervals for selected geographic regions based upon the published literature (after Dominey-Howes and Keating 2005) Region
Tsunami frequency
Reference
Kamchatka
1 large event per 1,000 years or 1 event every 30 years
Pinegina et al. (2003)
Chile
1 large event every 200 years
Salgado et al. (2003)
Cairns, Australia
1 event every 600 years
Nott (1997)
Japan
1 major event every 500 years
Nanayama et al. (2003)
Hawai’i
1 event every 30 years
Dudley and Lee (1998)
Cascadia, NW USA
6 major events in 3,000 years
Hutchinson and McMillan (1997)
Note: For Kamchatka, Chile, Australia and Japan, the frequency is calculated from the records of near-field (local) tsunami only. For Hawai’i, the frequency is calculated from the records of both near-field and farfield tsunami
Risk communication and education may take several forms but education through the school curriculum and effective home-based preparedness are closely linked (Johnston et al. 2005). For example, a risk perception and preparedness survey conducted by Ronan et al. (2001) on school children revealed that those who had been involved in hazard education programs displayed more stable risk perceptions, reduced hazard-related fears and demonstrated an increased awareness of important hazard-related protective behaviours compared to non hazard-educated children. Following hazard education, children are likely to interact and educate their parents on hazard issues, in turn, increasing home-based preparedness (Ronan et al. 2001; Johnston et al. 2005). The importance of hazard education in schools was highlighted in Thailand during the 2004 IOT when a 10-year-old British girl saved hundreds of lives by recognising the on-coming tsunami because of the recession of coastal water from Phuket beach—a clear physical sign of a tsunami which she had learnt about at school (King and Gurtner 2005).
123
104
Nat Hazards (2008) 45:99–122 Greater SydneyRegion Number of G-NAF addresses
(Sydney Basin,Central Coast and South Coast) 250,000 200,000 150,000
