Flash flooding case studies to improve predictions and the communication of uncertainty

Flash flooding leads to adverse social, economic and environmental impacts, including loss of life, infrastructure and property damage, ongoing health effects and community disruption. Flash flood events are generally difficult to predict with much lead time, making it challenging to prepare, warn and take protective actions for flash flood in an effective and timely manner.

The primary responsibility for flash flood warnings in Australia sits with state and territory emergency services and local councils. The Bureau of Meteorology provides weather forecasts, nowcasts and observations and issues severe weather warnings for heavy rainfall that may lead to flash flooding. In contrast, jurisdictions issue targeted warnings for locations at risk, which provide the community with information about the impact of the flash flood and guidance on how to respond effectively. The effective communication of forecasts and warning information is critical to emergency services and public safety.

This mixed methods research project focused on gaining an overview of the public understanding of flash flood and the influence of forecast uncertainty on decision-making by emergency management and the public. The project components comprised:

• Case study analysis
  • Three case studies of flash flood events from New South Wales (NSW), Tasmania and South Australia (SA).
  • Examination of the full warning value chain and flows of information.
  • Testing of the warning value chain tools developed by the World Meteorological Organization (WMO) HiWeather project.
• Public survey
  • Baseline assessment of the understanding of flash flooding, including comprehension of uncertainty terms and rainfall language used in severe weather warnings.
  • Total of 1239 responses (31% emergency management sector, 69% general public; 55% with flood experience, 45% without).
• Collaboration and engagement
  • Collaboration with emergency management stakeholders through one-on-one interviews, steering group meetings, case study workshops, Natural Hazards Research Forum workshops and the AFAC Flood and Severe Weather Intelligence Services Technical Group (FSWISTG).
  • Ten presentations on the project at conferences and events, including the February 2025  Centre Hazardous webinar.

The warning value chain is core concept used in the case studies to describe the building blocks of information (e.g. weather forecasts), the actors (e.g. the Bureau and emergency management) and the flows of information (e.g. data) that operate to produce value (e.g. reduced harm from the hazard).

The project outputs comprise this final summary report, a detailed report on the three case studies and a manuscript submitted to the Australian Journal of Emergency Management discussing the survey results and implications.

Findings from this research will inform how future strategies, procedures, services and products can be developed to improve risk and uncertainty communication to emergency management and the public during flash flood events.

The research findings showed that the flash flood warning value chain is complex and includes many actors, information nodes, linkages and information flows. Uncertainty is endemic across the warning value chain, affecting all components individually to varying extents, and there are flow-on effects where uncertainty cascades from one component to another. The value of information can be enhanced or reduced by the way it is communicated along the chain.

The baseline survey highlighted that flash flooding is not well understood. People overestimated how much they know about flash flooding and there was little difference in knowledge between the public and emergency management. The survey also highlighted that the terminology used in weather warnings is confusing. Uncertainty terms such as 'likely' or 'possible' are understood differently, and there is limited understanding of the difference between 'heavy' and 'intense' rain.

Understanding the complexities involved and improving the communication of uncertainty will assist decision-making for these types of events. A multi-faceted approach of public awareness and education, use of best-practice communication strategies (such as consistent terminology, contextual information, use of words, numbers and visual prompts) and co-designed decision support (such as understanding of warning thresholds and response triggers for users) will yield the optimum benefits.