In October 2025, the University of New South Wales hosted ‘One Big Idea’, a showcase of select UNSW academics, researchers and PhD students to share how they could change the world as we know it. Selected participants had 10 minutes to share how one concept in their field of expertise could revolutionise life on Earth.
Imagine: a drone pans across the sky, you begin to see where the floodwaters end and the dark clouds begin. What were green sugarcane fields are covered in grey water that matches the sky. Out of the corner of your eye, a dark block floats into view. As you begin to descend towards it, there are people standing on what looks like the roof of a building. They’re trapped. They’re surrounded by water. And they’re waiting for you to rescue them. They wave frantically as they realise help is on the way.
But it didn’t have to be this way. We can – and should – flip the script to think about evacuation as forced movement (McAdam 2025). That subtle shift opens us up to ways of preventing and responding to disasters that are attentive to people’s experience and knowledge of their communities. It also opens us up to not only recognising but prioritising the traditional knowledge and sovereignty of First Nations peoples. Our laws and emergency management plans must recognise this reality and they must support and enable it.
The people trapped on the roof are not a one-off. They are waiting, just like hundreds of other people who could not be immediately reached by overwhelmed emergency services teams, for example in Texas and Malibu in the United States and Taree and Mallacoota in Australia (Tillman and MacKenzie 2025; Casas and Yousif 2025; Lee et al. 2025; McGuire and Butt 2020). And it’s not just floods. During 2025, many people were unable to leave their homes before wildfires tore through Los Angeles. In 2020, the Royal Australian Navy evacuated over 1,000 people and 150 pets that had fled to a beach in Mallacoota on Victoria’s southeast coast as bushfires moved into the town.
In each case, evacuations undoubtedly saved lives and minimised harm. They became necessary as hazardous conditions overwhelmed communities. But those conditions, and our susceptibility to them, are not inevitable. With climate change, hazards, like floods and bushfires are increasing in intensity and frequency (DHA and NEMA 2023). Events will unfortunately continue to unfold. And, just as in these cases, emergency services personnel won’t be able to assist every single person (Australian Climate Service 2025). Not because they’re not trying; but because the numbers are simply too great.
Evacuation as forced movement
This situation appears bleak, but our ability to prepare for and prevent the need for evacuation doesn’t have to follow the same script. We can chart a path forward that places people and communities at the centre of how disaster response is done. If we take a step back, evacuations are not equivalent to rescue. Instead, people are forced to move either as a result of the hazard itself or by some other decision-making authority (McAdam 2025). This is true regardless of whether a Navy ship scoops people off a beach or people leave before the disaster hits.
When we look at evacuations only as a way that people are rescued from dire situations, it makes it hard to think about them in any other way. However, at their core, evacuations are a process that forces people to move from one place to another, even when it appears voluntary. Evacuations are deeply disrupting and involve different risks of harm for different people and communities (Oliver-Smith et al. 2017; Howard et al. 2017; McAdam 2025). They are a process that may not end in people returning home quickly – or at all.
Many people in Taree, for example, have still been unable to return home (Burgess and Fisher 2025) and they won’t be able to until their homes are rebuilt or they move again. For them, and countless others, evacuation isn’t temporary. During the summer bushfires of 2019–20, some Aboriginal and Torres Strait Islander people were turned away when they tried to access evacuation centres and supplies (Williamson 2022). People with medical conditions, or who don’t have documents matching their gender identity, may regard staying put as their only option (Yamashita 2024). Farmers, and people with animals, may see limited, or no choice about whether to leave or stay (AIDR 2024).
Putting people at the centre of planning, response and funding
Traditional knowledge and local experience are critical to disaster response (DCCEEW 2025). Law and policy must change from placing ‘combat agencies’, like police, at the centre of planning, response and funding. Instead, people – and especially First Nations peoples – should be at the centre of planning, response and funding.
Rather than rescuing people trapped on a roof, reducing the risk of disaster should empower people and resource them directly (Sharpe 2021; Cretney 2015). This may mean individual payments to assist residents to evacuate ahead of a cyclone or flood. It may mean providing building and other materials to communities to make structures resilient to bushfires. Or it may mean providing vehicles to assist people who do not have transport. Local communities, whether a town or a group of friends, and First Nations peoples as traditional custodians of the land, are best placed to know their needs and context (Copes-Gerbitz 2024). People are key to preventing the need for evacuation and they can be key to disaster response, if properly empowered and supported.
During the 2019–20 summer bushfires, local Indigenous community organisations worked on the ground to provide evacuation assistance (Williamson 2022). This included providing food and shelter to evacuees. In many parts of the country, First Nations peoples were first responders who filled a gap in response without being funded to take on that role. They repurposed schools as evacuation centres because someone had a key to the building not because it was set up for evacuations.
During Hurricane Katrina in 2005, neighbours called one another to provide information or to urge evacuation (Li et al. 2008). Before evacuating, they knocked on each other’s doors to check if anyone was there, looking out for elders and people without vehicles. Neighbours packed and evacuated together, sharing cars and other resources.
In response to the 2011 tsunami in Japan and the Fukushima disaster, people throughout the country measured radiation in the air, food and soil (Kenens et al. 2020). They couldn’t access the type of local data that they needed to make informed decisions about whether their food was safe nor whether they should evacuate because of dangerous levels of radiation. These data were made publicly available in a variety of ways via websites, word of mouth, organised meetings, whatever worked in a particular community.
In Australia, there is a decentralised and Indigenous-led movement of cultural burning; a fire management practice with a long history (Williamson et al. 2020). The movement has collaborated with land and fire agencies across Australia to re-establish First Nations peoples participation in fire management. Yet, to date, that work has mostly been incorporated into existing processes of land and fire risk management.
In 2017, bushfires tore through more than one million hectares of British Columbia. The Tsilhqot’in Nation was heavily affected (Verhaeghe et al. 2019). In the face of a rapidly spreading fire, one community decided to remain and defend their territory. The decision to not evacuate was based on traditional knowledge such as the topography of the land and forest as well as past fire and weather behaviour and traditional decision-making arrangements. The Tsilhqot’in successfully defended their territory despite lacking the equipment and funding of the Royal Canadian Mounted Police. That experience reinforced the importance of First Nations knowledge and relationships to the land. This highlights the need to provide material support and funding directly – as well as decision-making authority – to those on the front lines.
The knowledge and the capability of First Nations to lead is often sidelined by current legal frameworks. This must change. Rather than filtering funding, equipment and materials through layers of bureaucracy and people, communities and First Nations should be at the front of the line because they are on the frontlines.
Building a new system with people as partners
Shifting our thinking of evacuation as forced movement should lead us to think about people on both sides of the equation. For example, law enforcement agencies should not lead evacuation planning in over-policed or marginalised communities (National Justice Project 2025). State or regional organisations may not be best placed to make decisions for a community they do not know (Copes-Gerbitz 2024).
So, where does all of this leave the people standing on the roof, surrounded by floodwater? If we saw the people as partners and not just potential evacuees, perhaps they wouldn’t have had to climb onto the roof, waiting for rescue. Maybe they would have had somewhere safe to go, trusted a warning that their house would be inundated or have been supported to relocate to higher ground before this flood or future floods.
People routinely point to family and friends, community and trustworthy information as vital during evacuations. Neighbours knock on doors to assist people to leave. Parents use open-source technology to check the real-time radiation levels in their homes. First Nations peoples rely on generational knowledge about fire behaviour and deep relationships to lands to prepare for and decide whether to evacuate. This isn’t about leaving people to fend for themselves. And it’s never as simple as just sending in a helicopter. Evacuating people isn’t always the right move, but empowering people always is. Let’s stop designing systems that rescue people after the fact. Let’s start building systems that trust, resource, fund and place decision-making authority with people before evacuation becomes necessary.