Home Contact us Gallery Site Map Resource Centre Search About us People News Publications Education & Training Events Publications | Education and Training | Current & Recent Projects | Events | Research Opportunities

Download issue (989KB PDF) Cover Page Executive Summary Introduction Data Sources Deaths Due to Natural Hazards A Building Damage Index 20th Century Building Damage Alternative Perspectives on Damage Spatial Variation in Damage A More Refined View Discussion Conclusion Further Reading Acknowledgements

Issues in Risk Science Natural Hazards Risk Assessment: An Australian Perspective - Russell Blong Discussion It is evident that the answers to the questions with which we began are not simple. We can say that the record of human deaths and building damage is dominated by tropical cyclones and floods. We can be confident that we have captured enough of the sparse record that a few more years digging through historical accounts will not change that conclusion in any significant way. It is also clear that geological hazards have been supremely unimportant during the couple of hundred years of European supremacy. However, just 20 of the 1,200 events (i.e., 1.7% of the events) in the building damage database contributed about 50% of the total damage. Similarly, just a handful of events have contributed a substantial proportion of the death total. We have also noted the diversity of perils that produced the large death tolls and the most damaging events. Thus, while we have a fairly good idea of the past it is less easy to say which peril will be the most deadly or the most damaging in the future. We could argue that the overwhelming dominance of meteorological events will continue, but a single earthquake or a tsunami could make a substantial dent in the patterns of the past. The range of perils that could produce a substantial single-event death toll (say, >>100 deaths) include tropical cyclone, flood, earthquake, tsunami, and bushfire. Notwithstanding this possibility, most years produce a few deaths in floods, bushfires, and thunderstorms. Most commonly, those killed are young males – sometimes because it is firefighters who are killed, but more often because it is young males who try to drive across flooded rivers or who play golf while there is lightning around. While the dramatic decline in the death rate (Figure 2) is encouraging, the details of the death statistics identify the audience for any targeted campaign aimed at further reduction. Before we bask too long in the warm inner glow of confidence that our assessment of natural hazards deaths is spot on we need to reflect on the perils that we might have left out of the list in Table 1. In fact, while we have made only a reasonable effort to count them, as opposed to the determined effort we have made to count the deaths from perils listed in Table 1, heatwaves have killed at least 4,287 people since European settlement began; i.e. more than any other single hazard and about 70% as many as all other hazards combined. Moreover, past heatwaves have preferentially killed the elderly, in contrast to the bushfires, floods, and thunderstorms mentioned above. Maybe we need to rethink that targeted campaign as well. While we are still basking in that inner glow we should note that drought kills no one and damages few buildings in Australia yet it is easily the most costly natural hazard in economic terms – one estimate of the cost of the El Niño-induced drought in 1983 suggests it produced losses greater than the total of all the insured losses listed in Table 5. Drought resides in a different part of the Australian psyche – the focus is rural rather than urban, inland rather than coastal, agricultural rather than service-based in its direct impacts. While each state has a State Emergency Service and the feds have Emergency Management Australia (EMA – in the Attorney-General’s department) and the Department of Transport and Emergency Services (busily capturing the hazard mitigation agenda from EMA) to deal with the Prevention, Preparedness, Response and Recovery of the rapid onset hazards, slow-onset drought is dealt with elsewhere in the bureaucracy. As the current drought – endemic for much of the last 15 years – tightens its grip on the dwindling water supplies of the urban majority this dichotomy might change. Moreover, any nexus between drought, water supply, El Niño, and climate change might encourage serious reconsideration of Australia’s negative position on ratification of the Kyoto Protocol. Whether the Risk Frontiers’ database shows evidence of the impact of climate change on natural hazards magnitude, frequency or consequence is not clear. No hypothesis has been tested – likely, the data are not of sufficient quality. If we can characterize the poles of the debate with Munich Re’s assertions that the role of climate change is clear in the climbing global insurance losses of the last 35 years on the one hand and van der Vink et al.’s (1998) explanation of trends in U.S. losses through population movements and property values on the other, my personal view lies closer to the latter. In any case, for Australia, explication is complicated by links between El Niño and climate change. Drought and bushfire increase in El Niño periods. Flood and tropical cyclone frequency decrease. A link with hail fall frequency has been identified but is not simple. With the four hazards we identified earlier as the most important in terms of deaths and building damage all linked to the El Niño cycle, we are a long way from adding climate change to a convincing natural hazards risk assessment. The near-total destruction of dwellings in Darwin by Cyclone Tracy on Christmas Eve, 1974 produced at least one positive – the development of one of the world’s best wind loading codes for residential (and other) buildings. Damage to buildings in the 1989 Newcastle earthquake led to the further strengthening of the earthquake loading code and its application to dwellings in some circumstances. Most urban areas have quite stringent controls on the types of buildings that can be constructed and their layout in bushfire-prone areas. Similarly, local governments control suburban development on floodplains – some allow almost no buildings below the Average Recurrence Interval (ARI) 100-year flood line; others have been less cautious. These building codes and land use planning controls suggest that building damage, or at least building damage per 100,000 population, should continue to decrease. The risk assessment process has not proceeded far enough to demonstrate that this is so; the varying dates at which controls have been implemented in local areas, and the relatively low frequency of hazard impacts at a site suggest the task might be too difficult. Nonetheless, two issues stand out. Firstly, there are no controls, or even “recommendations” concerning suitable roofing materials for hail-prone areas such as Sydney and Brisbane where the vast majority of roofs are clay tile or, increasingly, concrete tile. Wind-driven hailstones >4 cm in diameter can penetrate these materials (and slate and fiber-cement roofing), leading to water damage to ceilings and building interiors. Surprisingly, research into the hail resistance of the commonly-used Australian roofing materials has hardly begun. This is even more surprising as it now seems clear that hail is the most important natural hazard in terms of building damage in these cities in the 1-100 year (possibly 1-200 year) time frame. Secondly, a recent re-survey of major Australian flood damage studies indicates all too clearly that about one-quarter of all residential building damage (excluding building contents) produced by flood is damage to built-in furniture – kitchen cupboards, built-in wardrobes, and bathroom vanity units. Usually, these units are constructed from non-waterproof chip-board that swells when wet. Should built-in furniture in dwellings on floodplains be constructed from something more resilient? Given the cycle of replacement of kitchen cupboards in an affluent society would more resilient materials be more cost-effective in houses located below, say, the ARI-75 year flood? With the recognition that we have done a lot to mitigate damage to houses in cyclone, bushfire and, even, earthquake-prone areas, why have we done so little in relation to floods and thunderstorms? Isn’t the risk assessment evidence in Figure 4, for example, clear enough? «back to top«