@article {bnh-7824,
title = {Flood risk communication - final project report},
number = {645},
year = {2021},
month = {01/2021},
institution = {Bushfire and Natural Hazards CRC},
address = {Melbourne},
abstract = {
The context for the research
The\ focus of this research project was informed by the findings of earlier BNHCRC research that investigated human fatalities from natural disasters (Haynes et al., 2017). This project focused on the two behaviours most frequently associated with flood fatalities:
- Driving into floodwater in a motor vehicle, and
- Recreating in floodwater.
Although there were pockets of information and more detailed knowledge about these behaviours held in some jurisdictions, there was no national picture of either the details surrounding vehicle-related flood fatalities or the incidence of these two general behaviours and the detailed contexts in which the public enters floodwater. Therefore, by studying the general public more comprehensively, there was an opportunity to quantify behaviours and identify the potential challenges and additional levers for communicating flood risk, as well as enabling greater insight into differences across Australia (both within and across jurisdictions).
In addition, the information that was already known about behaviour in floodwater related solely to the general public. Emergency services, in particular State Emergency Services (SES) personnel, were previously identified as an at-risk group for entering floodwater and flood fatalities.
In the context of established approaches to the investigation of risk perception, comprehension, and risk-taking behaviour, SES personnel represent an {\textquoteleft}expert{\textquoteright} group. This meant they could also provide insights about risk perception and risk communication when studied and compared to the {\textquoteleft}lay public{\textquoteright} (expert-novice paradigm), for example enabling identification of differences in use of words/ language, conceptualisation of {\textquoteleft}flood{\textquoteright} and {\textquoteleft}flooded roads{\textquoteright}, and identification and use of environmental cues when judging flood risk.\
Finally, as a risk group there is an organisational imperative to study the behaviour of SES personnel entering floodwater when on/off duty. This research supports the evaluation, understanding, and provides opportunities to assist in the mitigation of a range of organisational risks, including
- workplace health and safety/duty of care, i.e., potential injuries to personnel,
- financial risk, i.e., through damage to vehicles, equipment and other assets, and
- reputational risk, i.e., through being seen to disregard the public risk messaging to never enter floodwater and/or being seen to take unnecessary risks and damage vehicles.
This project commenced in July 2017 and comprised broad two phases:
- Understanding behaviour in and around floodwater, and
- Collating flood risk communications and co-developing Community Service Announcements (CSAs) for flood for use in National broadcasting by the ABC.
Project strategy for research and collaboration
The research strategy involved a suite of research studies, and employed mixed qualitative and quantitative methods, including experimental research, questionnaires, and in-depth interviews employing a mental models approach to ensure the research problem is well understood, the needs of those at risk could be met, and knowledge is co-produced with end-users and community.
Consultation and collaboration with key stakeholders and various end-users were key aspects of the overall research approach and helped to determine the research foci through the lifetime of the project. Following this collaborative approach, the research was responsive to stakeholder contexts. This enabled the team to take advantage of opportunities to explore some new topics in greater depth (e.g., SES behaviours around floodwater), but has also meant there were some delays and changes to the planned research activities (e.g., being unable to progress some approaches and outputs due to COVID-19). As a consequence, some of the resulting outputs and projects differ to the outputs outlined at the start of the project.
Project overview - extending what is known and addressing the gaps
Previous research has established that floods in Australia are a significant, and often preventable, cause of death, and this research has provided some useful insights to the risk perceptions and planned behaviour among the public in relation to driving and floodwater (Fitzgerald, Du, Jamal, Clark, \& Hou, 2010; Hamilton, Peden, Pearson, \& Hagger, 2016; Haynes et al., 2017).
This project builds on existing knowledge in several important ways:
- Detailed analysis of recent vehicle-related flood fatality records (2001-2017) held in the National Coronial Information System (NCIS) provides a current synthesis of the specific demographic, situational and environmental conditions in which Australians have lost their lives entering floodwater in vehicles.
- Survey data provides insights to how often, and under what circumstances, the Australian public enters floodwater in vehicles and for other purposes.
- Surveys with SES personnel provide similar insights to how often, and under what circumstances, SES personnel enter floodwater in vehicles when both on and off duty.
- Mental models interviews with SES personnel and the general public provide nuanced insights into how each group understands flood risks, and supports the identification of shared understandings and differences.
- The development of a driving in floodwater version of an online tool (EXPERTise 2.0) allows for the assessment of the use of environmental cues when evaluating flood risk and deciding whether it is safe to enter floodwater. This research provides early support for an objective approach to understanding flood risk assessment, identifying and understanding individual differences in the use of environmental cues, and provides an additional metric to strengthen research into decision-making under conditions of risk and uncertainty.
Overall, these various studies highlight some of the complexity and challenges in flood risk communication and flood risk assessment and factors that need to be considered when developing communication materials. The detailed nature of the data collected also provides insights to how interventions and flood risk communication and engagement work can be targeted.
A snapshot of key findings
As the project\ comprised a number of studies it also generated a substantial number of findings and insights. These have been condensed into a series of practitioner-focused Research into Practice Briefs. In addition, the BNHCRC Communications Team has developed a short series of videos to showcase some key research findings and augment these Briefs.
A snapshot of key findings across the various studies is provided below:
- During the period 2001-2017 there were 96 vehicle-related flood fatalities in 74 separate incidents in Australia, with an average of 5.65 fatalities per year over the study period. The majority of deaths (87\%) occurred when people were attempting to cross low bridges, creeks or causeways and in just under two-thirds of incidents (64\%) the driver was travelling alone. At least 29\% of incidents were in four-wheel drive vehicles, however this is probably an underestimate because the vehicle details were not always reported in NCIS records.
- The nationally representative survey of the general public found that 26\% had engaged in activities in floodwater on land and 19\% had engaged in activities in flooded rivers. For the latter, swimming and wading were the most frequently reported activities, but for the former there was a much greater variety of practical and discretionary reasons given for entering floodwater. Public enjoyment of water activities, their familiarity with it as a source or risk, and the multitude of reasons given for entering floodwater pose challenges for those promoting the blunt- but-safe official message of {\textquoteleft}never enter{\textquoteright} floodwater.
- Regarding driving through floodwater, 56\% of respondents in the public survey had driven through floodwater and a similar proportion (53\%) could recall the details of an event of driving through floodwater within the last five-year period. Although the depth of water that was driven through was generally shallow and still or slow moving (70\% 30cm or less, and 89\% still/slow flow) there were a number of respondents who had driven through deeper water, and in all cases (based on the definition provided in the survey) respondents were reporting driving through water in situations where they couldn{\textquoteright}t see the road/road markings underneath the water and therefore could not judge the integrity of the road surface.
- In 90\% of incidents reported in the public survey, there were no negative consequences of driving through floodwater, i.e., no vehicle damage or issues for personal safety.
- Findings of the public survey research highlight the commonplace nature and potential risks of driving through floodwater on roads, but the lack of consequences for most people highlight the challenges for risk communicators in engaging the public with the potential risks and influencing behaviour change.
- Overall, only 40\% of the public sample could recall seeing any official campaigns aimed at preventing people driving or playing in floodwater. However, these proportions were higher for respondents in Queensland (68\%) and the Northern Territory (64\%). Only 13\% of the sample overall could recall any elements of campaign risk messaging and only 6\% of the sample overall could recall the most commonly used {\textquoteleft}if it{\textquoteright}s flooded, forget it{\textquoteright} message (either in its complete form, or the key phrase {\textquoteleft}forget it{\textquoteright} or the general sentiment {\textquoteleft}don{\textquoteright}t do it{\textquoteright}).
- Surveys with SES personnel found that 37\% had driven through floodwater in an SES vehicle (as the driver) in the last two years, and 52\% had driven through floodwater in their private vehicle in the last two years. Similar to the public survey findings, water depths were generally shallow, and flow was still or slow (57\% 30cm or less and 86\% still/slow flow) and generally they were driving large/heavy vehicles (68\% in light truck/dual cabs or medium/heavy trucks). However, also like the public, in all incidents reported they were unable to see the road surface/markings so there was a degree of uncertainty about the integrity of the road surface and heavier vehicles could be more of a hazard and cause more damage to an unstable road.
- In 75\% of incidents of driving through floodwater reported by SES personnel there were passengers in the vehicle, and these passengers were reported to have influenced the decision to drive through in around a quarter of incidents. Further investigation indicated that passengers were more likely to have been part of cooperative discussions and risk assessment and, as such, a resource for deliberative decision-making, rather than advocates for risk taking. This finding supports the existing value placed on team training and dynamic risk assessment and general safety advocacy.
- Mental models interviews with (expert) emergency personnel and (lay) general public uncovered complexities in defining, assessing, and communicating flood risk. Although some public interviewees could articulate sophisticated approaches to judging risk when encountering flooded roads, emergency services personnel perceived the public generally to be underestimating these risks.
- Complexities in communicating about risk were based on the multiple variables that influence risk, differences in perceptions about risk from flood/water, e.g., compared to fire as a risk to self, and the realities of needing to live with floodwater as {\textquoteleft}a part of life{\textquoteright} in some rural and remote areas.
- Experimental research on cue utilisation and floodwater risk assessment has been promising. Individual differences in the use of environmental cues have been successfully recorded, however at this point it has not been possible to map higher cue utilisation to higher risk assessment performance. It is not clear if this is a short-coming of the performance tasks used in the research to date, or if this is due to other factors. Further development and evaluation in this area is required.
Areas for research utilisation
This\ project provides an in-depth understanding of how the public and emergency service professionals (SES) behave around, and understand the risks of, floodwater. The translation of these findings into utilisation and impact is an ongoing process that will continue beyond the formal end of the research.\
There are four main areas of research utilisation that are being pursued:
- Co-development of National Community Service Announcements (CSAs) for flood with the Australian Institute for Disaster Resilience (AIDR), the ABC, and the AFAC Community Engagement Technical Group (AFAC CETG) this is an existing and final deliverable of the project that was delayed and is being completed in 2021.
- Training and engagement: Through Phase 1 of the research survey tools and findings about how people behave, perceive risks, and make decisions around floodwater were developed. The research findings can be used to inform more targeted communications and safety-related training, and the survey tools themselves can be used as an engagement tool both with communities and SES personnel. Surveys with SES personnel about entering floodwater included evaluation of organisational safety climate as well as risk taking behaviour, these data provide a baseline against which a number of SES jurisdictions could evaluate WH\&S improvements and the effectiveness of a range of safety interventions.
- EXPERTise 2.0: in this project we have developed, trialled and we{\textquoteright}re in the process of validating an online module of tasks to assess cue utilisation when evaluating floodwater risk. Effective cue utilisation is an important component of expertise that has been assessed and validated in a number of occupational domains. This tool could provide a low-cost approach to support and evaluate the effectiveness of targeted training interventions among SES staff and volunteer members.
- Research\ \ into Practice Briefs: Throughout the project, a series of Research into Practice Briefs has been developed. These are concise summaries of academic research and provide an accessible evidence base for end-users and practitioners to guide their work, whether that is community engagement, development of flood risk communication materials, or formulation of organisational WH\&S policy.
In addition\ to the above areas of potential and emerging utilisation, the project team has worked actively with end-users and a range of additional stakeholders to co-produce project outputs, including conference presentations (Taylor, Wiebusch, Tofa, Haynes; AFAC 2019) and co-authored peer-reviewed articles (Taylor, Tofa, Haynes, McLaren, Readman, Ferguson, Rundle, Rose, 2019).
Research Capacity Building
The\ research project team has also supported research capacity building through the mentoring and inclusion of students in research activities, including a successful PhD completion (Ahmed, 2019), five Masters of Organisational Psychology students, two psychology honours students, and a number of short-term research and engagement placements. The project funding has supported early career researchers (ECRs) and has enabled these ECRs and students to progress through co-authored peer-reviewed publications and utilisation outputs, i.e., Research into Practice Briefs.
},
keywords = {communication, communities, Flood, risk},
issn = {645},
author = {Mel Taylor and Matalena Tofa and Katharine Haynes}
}