Header image

Breakout Session 3C: Environmental data, exposure, and risk

Tracks
Day 1 - July 23 2026
Thursday, July 23, 2026
1:30 PM - 3:00 PM

Speaker

Dr Ryan Essex
Research Fellow
The George Institute

Pulling data from thin air: The potential of satellite data to strengthen global drowning prevention

Abstract

Background and Objectives: Drowning prevention efforts continue to be constrained by substantial data deficits, particularly in settings where surveillance systems are limited or absent. As a result, prevention planning is often reactive rather than proactive. This presentation explores how satellite data could be used to identify drowning risk, strengthen prevention planning and support decision-making in areas where local drowning data is limited.

Methods: The presentation synthesises emerging analyses that link satellite-derived environmental data with drowning outcomes at both population and individual levels. It draws on three related strands of work, using satellite data to 1) profile environmental characteristics associated with drowning risk at a population level; 2) improve understanding of behaviour and circumstances preceding drowning events, and 3) estimate exposure and drowning hazard across different settings and the potential to leverage this for prioritisation and planning.

Results: Preliminary findings will be presented across these three strands. At the population level, preliminary findings show how satellite data has the potential to expand existing understandings of drowning risk. Looking to geospatial data of the built environment for example, while urbanisation is generally protective rapid expansion with limited infrastructure can negate these protective effects at population level. Using the example of West Bengal, India we will also show how such data could be utilised to understand behaviours related to drowning, providing critical insights into weather preceding and during drowning events. Finally, we will show how data could be utilised to better predict drowning risk.

Conclusions: This presentation will consider future applications of satellite data for drowning prevention, including their relevance to the growing challenges posed by climate change. It will discuss how this data can complement existing surveillance and prevention systems, particularly in data-poor settings and shift drowning prevention from reactive planning toward more proactive, equitable and targeted action.
Mr Robert Andronaco
Senior Visitor Safety Planning Officer
Parks Victoria

Developing and validating a geospatial drowning risk exposure model for inland and coastal waterways in Victoria

Abstract

Background: Open waterways represent a major contributor to drowning, with recent increased incidents in Victoria, including events linked to extreme weather. Despite this, systematic exposure-based modelling to support drowning prevention and emergency management decision-making has been limited. This study aimed to develop and evaluate a geospatial drowning risk exposure model to identify and prioritise inland and coastal waterways at elevated risk, aligned with the Australian Water Safety Strategy 2030 emphasis on evidence informed, place-based prevention.

Methods: A mixed-methods, iterative modelling approach was used. Quantitative datasets included historical drowning records, waterway spatial data, land management parcels, Australian Bureau of Statistics geography, and recreational exposure proxies derived from activity datasets. Spatial analysis processes were applied to generate relative risk estimates across inland and coastal environments. Stakeholder evaluation was embedded throughout model development. An online workshop convened 35 participants from emergency management agencies, government departments, local government, land managers, and aquatic safety organisations. This was followed by a workshop with over 80 stakeholders. A targeted hybrid evaluation session involving operational and policy stakeholders further assessed model usability and relevance.

Results: The model produced geographically exposure-specific relative risk profiles, enabling identification of key high-risk waterbodies not apparent from incident counts alone. Stakeholder feedback supported the model’s conceptual validity, operational relevance, and value for prioritising prevention, preparedness, and mitigation activities across sectors. Participants emphasised the importance of combining incident data with exposure measures to support resource allocation and cross agency planning. Feedback also assisted with the development of a second model incorporating further relevant geospatial datasets. Full results, and detailed geospatial outputs (including risk maps) will be presented at the conference.

Conclusions: This study demonstrates the feasibility and utility of an exposure-based, geospatial approach to drowning risk assessment in open water environments. The model provides a scalable framework for evidence-informed decision making and targeted intervention.
Agenda Item Image
Dom Fitzgerald
Research Analyst
Surf Life Saving Australia

Building a holistic picture of visitation and risk of the Australian coastline

Abstract

Background: Surf Life Saving has long relied on indirect or localised measures to understand how people use the coastline - patrol statistics, incident reports, surveys, and anecdotal knowledge from clubs and lifeguards. While valuable, these sources do not provide a consistent, national, time-resolved view of how many people visit coastal locations, when they do so, or how this varies across patrolled and unpatrolled areas. This gap limits the ability to: Quantify exposure to coastal risk; Compare visitation patterns across regions and seasons; Evaluate patrol coverage relative to demand; Support evidence-based decisions around prevention, resourcing, and policy.

Description: The National Visitation Project was established to address this gap by creating a repeatable, scalable, and nationally consistent visitation dataset for Australia’s coastline. The project uses aggregated, de-identified mobile device data to estimate visitation patterns. This data is transformed into visits using defined rules and then normalised to improve comparability over time and geography. The dataset is unique in its inclusion of times and places not usually covered by more traditional data-collection methods.

Lessons Learned: Phase 1 has demonstrated the value of this dataset, with SLS team members seeing its potential across diverse use cases from lifesaving service agreements at a club level to LGA collaborations. Key challenges faced are centred around effective normalisation of the data and distribution of the dataset to users inside and outside of the Surf ecosystem.

Conclusions: Given Places / Beach is one of the AWSS 2030 Midpoint Focus Areas requiring urgent attention and that coordinated and collaborative lifesaving systems; and managing risk at unpatrolled beaches have been highlighted as things that matter most within that area, this project offers an opportunity to address the shortcomings that might be blocking efforts to reduce drowning in line with the AWSS 2030 goals
Agenda Item Image
Dr Juliana Albertoni De Miranda
Coastal Safety Officer
Surf Life Saving New Zealand

From hazard science to operational safety: Implementing a national rip current forecast system in Aotearoa New Zealand

Abstract

Rip currents are one of the leading causes of surf rescues in Aotearoa New Zealand and contribute substantially to beach drowning risk. To support proactive coastal safety management, Surf Life Saving New Zealand (SLSNZ), in partnership with the University of Plymouth (UK), has been developing a national-scale rip current forecast system based on environmental conditions associated with elevated rip hazard.

The forecast system was developed using more than 20 years of lifeguard incident data linked with wave and tide conditions across New Zealand beaches. Simple thresholds in wave and tide forcing associated with increased rip current activity were used to generate a five-level Rip Index designed to support lifeguard operations and future public-facing warning systems.

During the 2024–25 patrol season, lifeguards from beaches across New Zealand recorded in-situ assessments of perceived rip hazard, allowing comparison between forecast conditions and operational observations. Preliminary findings indicate the forecast provides conservative warnings, more commonly overpredicting than underpredicting hazardous conditions, and over the highest 4 Rip Index levels, agreed exactly with lifeguard perceptions 64-86% of the time. Operational implementation also highlighted important challenges, including differences between forecasted and visually perceived hazard, variability between beaches, and the potential for warning fatigue at persistently hazardous locations where elevated hazard levels occur frequently.

The project demonstrates the value of integrating long-term incident datasets into practical drowning prevention tools while highlighting the importance of balancing precautionary warning systems with user confidence and local operational context. The findings provide insight into how simplified hazard forecasting systems may support lifeguard preparedness, public risk communication, and future proactive coastal safety strategies in New Zealand.
Agenda Item Image
Ms Mary-Jane McLeod
National Community Engagement Lead
Bureau Of Meteorology

The Ripple Effect: Turning Bureau Forecasts into Action to Reduce Drowning Risk

Abstract

Background:
While real-time information about weather, water, and hazard conditions such as forecasts, marine outlooks, and flood warnings—is widely available through the Bureau of Meteorology (the Bureau), its role in drowning prevention is often under-recognised. A critical gap exists between the availability of forecasts and warnings, and the actions people take based on that information. This presentation explores how Bureau tools can support safer decision-making in and around water, aligning with the Australian Water Safety Strategy 2030 priorities of localised action, disaster resilience, and behaviour change.
Description:
This presentation showcases three key Bureau tools that influence water safety decisions: app-based notifications, real-time warnings, and publicly accessible knowledge resources. Using practical scenarios, it demonstrates how these tools inform decisions such as whether to enter the water, where to recreate, and when to avoid high-risk conditions. The session highlights how available information can be translated into simple, actionable insights for individuals, families, and communities across diverse aquatic environments including beaches, rivers, and inland waterways.
Lessons Learned:
Effective drowning prevention requires more than access to information—it depends on understanding, trust, and timely action. Key insights include the importance of delivering clear, accessible messaging and recognising the role of social sharing in amplifying impact. When one person acts on a forecast or warning, that decision often influences others, extending the reach of safety messages beyond the individual.
Conclusion:
The Bureau plays a critical role in drowning prevention by providing trusted, real-time environmental intelligence. Strengthening public engagement with these tools—and encouraging the sharing of knowledge within communities—can create a “ripple effect” that supports safer behaviours and contributes to reducing drowning risk across Australia.

Agenda Item Image
Miss Katherine Millar
Research Assistant
Surf Life Saving NSW

AI‑Enabled Quantification of Beach Attendance to Inform Lifesaving Operations

Abstract

Background
Understanding beach attendance patterns is crucial for delivering effective and relevant lifesaving services. Traditional methods of beach attendance data collection rely on manual counts, which can be inconsistent and time-consuming for lifesaving personnel. To address these challenges, Surf Life Saving NSW (SLSNSW) partnered with Surfline to implement artificial intelligence (AI) cameras to gather precise, continuous data on beach and water users, fundamentally improving resource allocation and safety outcomes.

Methods
AI‑enabled cameras were deployed at selected NSW beaches to automatically count beach and in‑water users throughout the day, comprehensively capturing attendance across both patrolled and non‑patrolled periods. The data can be used to identify peak visitation times, monitor exposure trends, and inform patrol deployment and service planning. By automating this process, lifesavers can focus on supervising swimmers rather than conducting regular manual beach counts.

Results
Preliminary analyses confirm the cameras provide detailed attendance data, enabling the identification of peak usage times and patterns beyond the capability of manual data collection. These insights facilitate targeted resource allocation strategies, including the optimisation of patrol scheduling and the provision of adequate coverage during identified high-risk periods.

Conclusion
The project underscores the importance of implementing standardised and objective data collection protocols to minimise reporting burden and inconsistencies to enhance evidence-based decision-making. AI‑enabled beach attendance monitoring ultimately provides a practical and scalable approach to strengthening lifesaving services by improving understanding of beach exposure and risk. Expansion to additional and unpatrolled beaches may further support targeted local responses and contribute to reductions in drowning risk.
Agenda Item Image
Mr Sean Kelly
Phd Candidate
UNSW

Environmental Drivers of Beach Visitation and Water Use and the Associated Impacts of Climate Change

Abstract

Background
Beach visitation is strongly shaped by environmental conditions, with attendance highest on warm, low wind, low rainfall days. These favourable conditions not only bring more people to the beach but increase exposure to coastal hazards. Climate change is expected to alter environmental conditions in ways that will make coastal environments more favourable for recreation year-round. This study examines the effects of environmental conditions on beach visitation and water entry, and project how climate change may alter future patterns of coastal recreation.
Methods
Daily visitation counts from multiple beach sites were linked to historical weather observations and modelled as a function of environmental and temporal predictors using generalised additive models (GAMs). Future patterns were projected by applying fitted GAM smooth functions to climate data from the NARCliM2.0 regional climate model ensemble across low, medium, and high emissions scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0) for mid-century (2040-2059) and end-of-century (2080-2099) epochs.
Results
Beach recreation is projected to increase across all study sites, scenarios, and epochs. Mid-century daily visitation is projected to rise by 6.1% (ensemble range: 2.9-9.7%) under low emissions and 9.1% (5.4-15.4%) under high emissions. End-of-century projections range from 6.5% (2.5-11.6%) to 16.4% (10.8-23.2%). Water entry showed comparable but consistently larger increases, with mid-century projections of 8.9-13.0% and end-of-century projections of 9.5-23.6%. Considerable spatial and seasonal variation exists, with projected increases reaching up to 28% for visitation and 41% for water entry at some sites under high emissions.
Discussion
Beach recreation in NSW is projected to increase substantially under climate change, driven by rising temperatures and drying conditions. These changes will be experienced as both higher peak-season volumes and an extended beachgoing season. These findings underscore the need for proactive adaptation to ensure that future increases in beach recreation do not translate into associated increases in the aquatic injury and drowning burden.

Session Chair

Agenda Item Image
Bernadette Matthews
Head of Research, Evaluation and Insights
Life Saving Victoria

loading