Caroline Sullivan has both a personal and professional connection to the recent devastating floods in eastern Australia.
She recently headed up a flood research centre based at Southern Cross University in Lismore, and has a house up the hill from the city's centre.
The floods didn't reach Dr Sullivan's place, but she still sees the pain they caused all around her.
"An awful lot of my friends lost every single thing they have in the world," she says.
"It's just heartbreaking."
There is now much soul-searching about the best long-term responses to disasters like this, and many face the agonising question of whether to stay or move.
The community as a whole needs to be involved in "redesigning" the town so it has long-term resilience against such disasters, says Dr Sullivan, whose main affiliation is now as an adjunct professor at the UNSW Water Research Centre.
Part of that redesign could involve relocating those most at risk.
To some extent, she says, it's about letting nature run its course, an approach pioneered by countries like The Netherlands.
As part of a program that literally made "room for the river", the low-lying country bought back many farms for the flood plain.
"Instead of trying to control the river into channels and make it do what we say, the Dutch said, 'We know rivers are dynamic and therefore we must leave room for them to move'."
But there's so much more that could be done to work with nature to reduce the risk of flooding, Dr Sullivan adds.
Restoring river and wetland ecosystems, covering more of our towns in vegetation, and keeping our catchments forested could all help reduce the impact of floods.
They're interventions inspired by the processes and functioning of nature.
The tropical island state of Singapore — which gets a lot of rain — has already embraced such nature-based solutions.
"Singapore used to have very large concrete drains to reduce flooding, but this is ugly and doesn't have any ecological or social benefits," says urban sustainability expert, Perrine Hamel of Nanyang Technological University in Singapore.
But over a decade ago it launched a water management program that included "naturalising" the city's river by greening its banks and surrounding it with beautiful, biodiverse parklands.
Urban areas are particularly prone to flash flooding because impermeable surfaces like concrete and asphalt cause water to run off and accumulate in low-lying areas.
The park around the river is not only a great place for people to gather when it's dry, but provides an area where excess water can soak into the ground and cause less havoc.
Other nature-based flood mitigation initiatives in Singapore include using swales — vegetated drains on slopes, which slow down water run-off, reducing its destructive potential; and wetlands on the lowest-lying areas, which help protect the city from coastal flooding.
Dr Hamel says traditional engineering solutions (sometimes called "grey infrastructure") like dams and levees are still important, but have their limitations.
"They have a huge carbon footprint, and they can fail."
And, she adds, they also displace the problem downstream.
"If you are getting rid of the water as fast as possible, this water still goes somewhere, and it can lead to flooding downstream."
And dams can be controversial.
Jason Byrne from the University of Tasmania studies climate change adaptation and says Australia has a long way to catch up to places like Singapore.
"They are deploying nature-based solutions right across the city state … they're light years ahead of us."
Professor Byrne says some "green infrastructure" techniques to reduce run-off, like green roofs, have been trialed in cities like Melbourne.
Other options include permeable pavements that soak up water, and street trees that do the same, and can have the added benefit of cooling cities.
But evidence suggests the specific benefits of urban greening will depend on local conditions.
A recent modelling study, for example, found that in most cities of the world, greening was unlikely to both cool the city and reduce flooding at the same time.
"Urban greening is not a panacea or silver bullet solution," Professor Byrne says.
Experts say it will also be important to manage other issues that can occur with more nature in urban areas — things like bushfire risk or mosquitoes — and to avoid varieties of trees that suck up too much water, leaving little for the rest of the environment.
"It really will be about making a complex set of trade-offs," Professor Byrne says.
"Planting the right tree in the right place will be critically important."
If greenspace helps soak up water into the landscape, this could not only help guard against flooding, but save more groundwater to feed rivers in the dry times — something that's important for water security.
The deep roots of trees are particularly good at helping water infiltrate into the soil, which has led to studies of how deforestation can make flooding worse in downstream towns.
"If we deforest those catchments for agricultural purposes, it means there's less vegetation around to catch that rainfall and evaporate it back out into the atmosphere," Professor Byrne says.
But nature will always have the capacity to surprise us with large floods, especially in a changing climate, and there are limits to how much forests can absorb.
If there is too much rain, a catchment gets soaked and can't absorb any more, which is why we need to have land downstream set aside that can be flooded without causing too many problems.
Despite these limits, Dr Hamel and colleagues argued in a recent paper that models often underestimated the impacts of nature-based solutions like forests.
Even small reductions in flood peak and flow could make a big difference to how badly communities are affected, they say.
And since smaller floods are more frequent, their impacts add up.
Putting exact figures on the benefits of nature-based solutions can be tricky and models have to be specific for individual catchments, Dr Hamel says.
Still, momentum is growing in support of nature-based solutions, likely in part because they can have a multitude of benefits.
The World Bank's Global Program on Nature-Based Solutions for Climate Resilience is funding projects all over the world.
And there's even talk about getting the insurance industry to invest in nature-based solutions to save money in the long run.
Back in Australia, Dr Sullivan, who knows the Lismore community well, is among a group of researchers who wants to look more at the potential of this approach.
Unfortunately, the interdisciplinary national flood research centre she headed closed not long after it was opened, due to lack of funding
But this hasn't stopped her applying for federal grants to study how nature-based solutions might contribute to mitigating floods in the Richmond catchment, which includes Lismore.
This is something the local community itself has said it wants to explore.
Dr Sullivan is aware of the challenges.
"Nature-based solutions feel a bit intangible because it's lots of little things, not one single big thing like a dam wall that everyone can focus on."
And research will need to look at the best way of getting the community to work together to benefit the whole catchment.
This might require removing economic and regulatory disincentives, she says, citing the example of a farmer who had to pay extra rates after establishing a wetland on their property — because it was classed as a farm dam.
"Why would a farmer want to do that if they have to pay more money," Dr Sullivan says.
"It's much broader than just hydrological modeling. It's a whole system issue, so we have to take a wholistic approach."
Dr Sullivan says she's aware of skepticism about nature-based solutions, but thinks it's imperative we add them to our toolkit of flood responses.
"Rather than trying to hold nature back, we need to work with nature, and let nature work for us."