This is the third newsletter from the Fatalities and Building Losses project, and the first for 2015, with updates for the project end-users and associates. A big welcome to Corey Shackleton from the NSW RFS, our new cluster lead end-user. Thank you to those who have helped Lucinda in her mission to find more data about… everything. In this newsletter we focus on just what this project is providing, what it means, and where we are up to. We welcome feedback at any stage – on the newsletter content but also on any aspect of the project. Please contact Lucinda in the first instance, via lucinda.coates@mq.edu.au.
What is this thing called PerilAUS?
A large part of this project will be based upon PerilAUS, Risk Frontiers’ database of natural hazard event impacts, and most of the work that has occurred since the start of this project has been about improving it, with a current focus on floods. But just what is PerilAUS?
It’s a database of the impacts of natural hazards in Australia.
The various hazard databases that comprise it were begun as long ago as 1985.
About 35 person-years have been spent on researching, inputting and checking.
It has 12 types of natural hazards: floods, bushfires, tropical cyclones, earthquakes, heatwaves, severe storm – windstorms, hail, lightning, rain, tornado [& flash flood] – tsunami and landslides.
It covers the period from European settlement (1788) to now, with a more complete record from 1900.
As of 31 Jan 2015, there are 14,743 event reports from 1900 onwards (& quite a few before then).
It’s based on data from news media, government reports, published literature, Coroners’ inquests – so far, some 19,473 references have been used.
It contains data on:
economic, social and environmental impacts
number of people killed, injured, homeless or evacuated
physical attributes of the event: duration, intensity/ magnitude e.g. area covered; presence of other natural perils or contributing factors… depending on peril type
fatality details: age, sex, cause of death, occupation, circumstances at time of death – including activities undertaken prior to death and why (e.g. inside house; retrieving possessions; received no/ little warning etc).
It differs from other databases we know because of:
the wealth of descriptive detail around the circumstances of death.
the measure of building damage caused – the “housing equivalent” – which allows comparison between different events, locations, years and peril types.
We used to scroll through microfilm in the basement of the uni library; now most of our research is done online (many local newspapers have now been scanned).
The heat is on: what can you do about it?
The first deliverable of this project was a published paper on the history, consequences and future threat of heatwave fatalities in Australia, based on the current data within PerilAUS (1844-2010). To assist our end-users to understand just what its conclusions might mean for them, we present here a summary of our findings and recommendations. The paper in its entirety may be accessed at: http://www.sciencedirect.com/science/article/pii/S1462901114000999
Some interesting facts – did you know that:
There is no universal definition of a heatwave: the term is relative to the usual weather of a location.
From 1844-2010 extreme heat events have killed at least 5,332 people in Australia.
From 1900, that figure is at least 4,555 – this is more than the combined totals of deaths from all other kinds of natural hazards.
Over 30% of deaths since 1900 occurred in just nine large scale events.
January 27 stands out historically as the date with the most heat-associated deaths.
The highest ranked heatwaves in terms of fatalities are:
1896 (450 deaths), 2009 (432) then 1939 (420) in terms of raw numbers; and
1896 (13), 1939 (6), 1908 (6) and 2009 (2) in terms of death rates (number of deaths per 100,000 population).
Current and future threats
Deaths and death rates fluctuate widely but both show a decrease over time.
More males than females have perished but the numbers are more even in recent times.
The elderly are significantly more vulnerable and this vulnerability increases with age.
Infants, especially those aged <1 year, are also more at risk – more so in the earlier years of the record.
Those involved in physical work, especially if exposed to sun or other heat sources or in poorly ventilated areas (for example, labourers, tradespersons, farmers, cooks, miners).
Who is most at risk during extreme heat events?
In a warmer climate, extreme heat events will probably be more frequent, longer and hotter.
Australia’s population is increasing – in numbers, and in age.
The heat island effect of urbanisation and high density housing is increasing.
Many people are increasingly living in homes not designed to reduce heat stress and more reliant on air conditioning – which cannot be guaranteed to work during a heatwave.
What you can do about it
Reducing future impacts of extreme heat will be especially challenging – many of the most vulnerable groups represent those sectors of society most marginalized, lacking resources and difficult to reach.
Most planning currently relies on reducing risks through information and education to influence and change public behaviour, and emergency response when an event unfolds.
Long term risk reduction must also consider urban planning, building design, community development and social equity.
A new use for coronial inquests
A coronial inquest may be carried out if, amongst other reasons, a death is sudden or untimely. Such inquest reports have been found, during this project, to be a crucial means of “ground truthing” or verification and adding even more detail to the circumstances surrounding natural hazard fatalities. This provides more detailed information for analysis so that we can better inform emergency management with regard to, for example, policy and procedures, resourcing during hazard event management, and public messaging. If we know what groups of the population are most at risk, and what governed the decisions/actions that led to death, we can minimise future risk.
To make the most of State Archive Office visits, the PerilAUS team has been busy expanding and improving the Flood Database, centred mainly on trying to find names for all the flood fatalities (see also next section). Challenges faced so far:
Access to coronial inquests varies from state to state and can be difficult:
No public access to Northern Territory records is allowed and our approach to have the office staff find the relevant files was unsuccessful (our dataset was a bit large).
To access Western Australian records, approval needs to be granted by an ethics committee (we are still working on that one: any assistance greatly appreciated).
States have varying holding periods – there are generally restrictions around how recently inquests can be made publicly available: this period can be up to 75 years. However, we intend to apply for access to these more recent records
Coronial inquests will not necessarily be carried out for every flood fatality that has occurred.
Some records sometimes no longer exist: for example, NSW has no records from 1900-1915 and decimated records 1916-1939; WA has no records prior to the early 1970s.
Names can be misreported in the media, or a person may have aliases; often first names can be interchangeable.
Handwritten registers and records are sometimes difficult to interpret.
Progress update: what has the research team been doing all this time?
Finding out more names for flood fatalities:
965 more flood events entered, 528 of which had one or more deaths associated with it
592 new flood fatalities have been entered and, importantly,
a total of 953 more names have been added to existing and new records
Armed with lots of names, we visited the state archives offices SA, Queensland (twice), NSW (a few times) and Victoria (twice) to find coronial inquests of flood fatalities. So far:
About 29 person-days have been spent in the records offices.
333 inquest files have been located and photographed.
The data – concerning deaths and physical aspects of the event – has been entered in PerilAUS for SA, most of Vic, half of NSW and about a third of the Qld records.
The “hit” rate of finding inquests from PerilAUS’s listed named fatalities has varied across states, from 14% to 53%.
Finding better data on environmental conditions about floods:
The 6,072 BoM references used to source data on the meteorology of the hazard have been supplemented by a few more and, more notably, by 150+ local media articles (via online information source Trove, which supplies scans of media reports)
Over 200 hours have been spent on this thus far
Gaps have been filled from 1900-1916, ~1994-2013 (more available data; much faster process!) and we have restarted from 1917.
Adding injury, rescue and near miss data, focussing on floods:
We tried to access a comprehensive dataset on injuries from the Australian Institute of Health and Welfare. Unfortunately, a fee of $15,000 was requested to access the past 10 years of hospitals admissions data. An additional dataset is available for the decade before then that would also add greatly to the outcomes of this project. Such a sum was not envisaged at the planning stage, but this is very valuable data and it would be great if we could somehow access it… any ideas?
A start has been made on adding injury, rescue and near miss data from our usual online sources. This is a very time-consuming process and has been put on hold while we complete the floods fatalities work.
Other means of obtaining at least a sample of this kind of data are being investigated.
Requesting help from end-users in data from their organisations:
SES: NSW recommended their website for annual reports and floodsafe/com.au for flood event reports and Local Flood Plans (which often have a history of events). Activity reports are created after large events and given to the Ministry of Police. Initial contact was made from a few staff in Qld.
WA Dept of Fire & Emergency Service: a couple of very useful teleconferences were held and there is a plan to visit and interview some staff from the office and further afield after the fire season.
SA Metropolitan Fire Service accessed their AIRS database – unfortunately, very few relevant records were located.
Tasmanian Fire Service have been helpful in passing on other contacts; of those the most valuable will be those in the Coroner’s office.
The year 2013 flood events were added to the Flood Database of PerilAUS.
Research plans for 2015
Our fourth deliverable for this project is a report and submission of a journal article on flood fatalities alongside a presentation of the results to end-users. This is scheduled for 29 May 2015 and is the focus of the majority of work carried out to date. To reach that milestone, it is further planned to:
Finalise the addition of names to the PerilAUS flood fatality records.
Add the year 2014 flood events to the Flood Database within PerilAUS.
Research additional data on the environmental circumstances of fatal flood events from 1925 to ~1993.
Access the remaining available inquest data from the Victorian, New South Wales and South Australian state archive offices.
access all of the available inquest data from the Australian Capital Territory and Tasmanian state archive offices
make reasonable attempts to access the inquest data from the Western Australian and Northern Territory state archive offices
make direct application to the State/ Territory Coroners for access to the more recent inquests:
Qld: 1980 onwards
NSW: July 1963 onwards
Vic: 1986 onwards
Tas: ~1941 onwards
SA: 1954 onwards
The remainder of the year will be spent focussing on all the other natural hazards of interest, and finding out the same sorts of data around those as was done for floods, starting with bushfire. Recent collaborations with CSIRO and the former Bushfire CRC have meant that our record of fatalities is pretty good although, as we have found with floods, it can always be improved.
We hope to catch up with all end-users at either the upcoming Research Advisoy Forum in Sydney on 8-9 April (we will be giving a 20 minute talk on our work to date [flood fatalities] and ensuring what we’re doing is what you want us to do) or the annual Bushfire and Natual Hazards CRC and AFAC conference in Adelaide on 1-4 September.
How you can help
Does your organisation have any records - reports, databases, other documentation - on:
Events that caused death, damage or injury.
Injuries (to your staff or to the public).
Rescues and other near misses and fatalities resulting from flood and other natural hazard impacts/call-outs.
Physical and environmental details of flood events that have led to a fatality?
Please let Lucinda know: the more data we gather, the better our final results are for you.
Teleconferences: how regularly would you like us to hold these to keep us all connected and up to date? Once every couple of months? More often? Less often? Please let us know.
Items for the next newsletter: any meetings, conferences, news, etc you would like to share, please send them to Lucinda. Editor’s deadline: Friday 27 March 2015!
Researchers’ contact details
Finally, please contact us if you have any questions, suggestions or feedback on the project:
Please note: Katharine is on maternity leave until March 2015. In her absence Lucinda Coates will be managing the fatality, injury and rescue component of the project.