|
Abstract
|
Change in natural and human systems are almost invariably measured using trends. Cause and effect are known to take other forms but these are rarely investigated and, if they are, the burden of proof tends to be greater. When dealing with complex systems, analysts (and their critics) need to be more discriminating.
Flood deaths in Australia 1900-2015 were investigated by Haynes et al. (2017)1 using the PerilAUS database constructed by Risk Frontiers 2 using a methodology also applied to bushfire fatalities (Haynes et al. 2010).3 This was combined with coronial information to determine causes of death and develop potential policy interventions. A total of 1,859 flood deaths occurred during 1900- 2015. About two-thirds of these deaths involved 2 or fewer people and most were due to drowning or suspected drowning. People of indigenous heritage are likely to be underrepresented in the data, especially earlier in the 20th Century (Haynes et al. 2010).
Between 1900 and 1960, the rate of deaths due to floods declined by 0.55 persons per 100,000 population per year, but a step change occurred around 1960 to lower mortality levels that have since remained stable. Numbers declined from 219 deaths per decade before 1960 to 99 per decade afterwards. Work by Haynes et al. (2017) determined a reason for this rapid decline, speculating that it may have been due to flood mitigation works, technology advancements, better warnings and the postwar expansion of emergency services response (i.e. the growth of the SES). However, records of state-based SES show staggered starts, with Queensland, the state with the most deaths, having no specific emergency response function until the early 1970s.
In 1957, the Australian Government directed the Bureau of Meteorology to set up a hydrology branch that subsequently operated in Queensland, New South Wales and Victoria (Pagano et al. 2016)4 , the states where most deaths had occurred during...
Language: en |