The map didn't make it personal enough.
The maps were supposed to scare people into paying attention. They didn't.
A study published in the peer-reviewed journal Nature Sustainability has found that flood maps — the kind governments and climate scientists have long used to show coastal residents how rising seas might swallow their neighborhoods — are not just failing to raise alarm. They appear to be making people less worried about climate change than they were before they saw them.
Matto Mildenberger, an associate professor of political science at the University of California, Santa Barbara, led the research. A Canadian originally from New Brunswick with a background in B.C. environmental politics, Mildenberger has spent recent years studying how climate and energy policy gets communicated to ordinary people. His team recruited more than 1,000 participants across four coastal U.S. communities: San Francisco, California; Norfolk, Virginia; Ocean County, New Jersey; and Palm Beach, Florida. The cities were chosen deliberately — the first two had populations broadly accepting of climate science, while the New Jersey and Florida communities were more skeptical, and both had lived through major flood events, Superstorm Sandy and repeated hurricane seasons respectively. All participants lived within a mile of a projected future flood boundary. Some were on the dry side of that line. Others had homes expected to be at least partially underwater by 2100.
Half the participants were shown personalized bird's-eye flood maps, each with a red dot marking their own home relative to the projected inundation zone. The other half were not. When researchers measured concern about sea level rise afterward, those who had seen the maps registered lower worry than those who hadn't. The effect held across all four cities, inside and outside the flood zone, and regardless of where participants stood on climate science to begin with. "We found evidence for the backlash effect," Mildenberger said.
The explanation, he believes, lies in how the human brain processes distant threats. Social psychologists call it construal theory: the mind struggles to make abstract, far-off risks feel real without concrete, personal detail to anchor them. A map showing a blue wash creeping across a neighborhood in 2100 is, in that sense, too tidy. It flattens a complex, lived danger into something that looks almost manageable — a technical problem for someone else to solve, decades from now.
To test what might work better, Mildenberger's team returned to San Francisco and tried a different approach. Using cellphone data to model commuting patterns, they layered future flood projections onto the city's road network and calculated how rising seas would lengthen commute times — in some neighborhoods, by up to 20 minutes. When participants were shown that information instead, their concern about climate change's personal impact rose significantly. The difference wasn't the data. It was the specificity. A flooded basement in 2100 is abstract. Twenty more minutes stuck in traffic next decade is not.
The stakes for places like Metro Vancouver are considerable. Nearly $30 billion in residential and commercial property sits within one meter of current sea level in the region. A 2021 report from the Canadian Institute for Climate Choices projected that annual flood damages in Metro Vancouver could reach $510 million by the end of the century under a low-emissions scenario — a 17-fold increase — and as high as $820 million under a high-emissions path. For individual homeowners, that translates to an average annual property damage cost of $4,400 within 50 years, more than seven times today's figure. Vancouver's city website currently hosts a flood map showing what a one-meter sea level rise would mean for its downtown core and western coastline. Mildenberger suspects the psychological backlash his team documented in the U.S. applies equally to Canadian coastal cities.
Not everyone is ready to retire the maps. Andrew Pershing, director of climate science at Climate Central — a U.S. research group that has made increasingly detailed sea level rise maps central to its public communications work — defended their value for what he called "power users": city planners, engineers, and policymakers who know how to read them. He doesn't anticipate pulling back from producing them. But Pershing acknowledged that scientists have often assumed the public sees these images the way they do, and that assumption has costs. Even something as small as the shade of blue used to represent floodwater can mislead — suggesting the water is shallow, nothing urgent. "Risk is just a really complicated thing for people to understand," he said.
Climate Central has already begun moving beyond static maps. Its FloodVision project uses LiDAR sensors and GPS units mounted on vehicles to capture street-level imagery, then combines that data with sea level rise projections and artificial intelligence to produce photo-realistic images of flooded neighborhoods — the kind of picture that puts a viewer on their own sidewalk, looking at their own street, under water. A colleague of Mildenberger's at the University of Montreal is pursuing similar AI-generated imagery and studying how effective it proves at shifting public perception. The underlying problem, Mildenberger said, is that for years, communicators have been sending messages without adequately testing whether those messages land. That, at least, is beginning to change.
Notable Quotes
These top-down maps may be focusing people's attention on one small sliver of the problem — a sliver that actually reduces concern for climate change.— Matto Mildenberger, associate professor of political science, UC Santa Barbara
Risk is just a really complicated thing for people to understand.— Andrew Pershing, director of climate science, Climate Central
The Hearth Conversation Another angle on the story
So the maps were meant to make the risk feel real — and they made it feel less real?
That's what the data showed. More than a thousand people, four cities, different political leanings — and in every case, seeing the map reduced concern rather than raising it.
Why would a visual of your own neighborhood flooding make you less worried?
The researchers think it's because the maps are too clean. A bird's-eye view of a flood zone in 2100 looks like a planning document, not a threat. It puts the danger at a distance — spatial, temporal, abstract.
But people whose homes were inside the flood zone still felt less worried after seeing the maps?
Yes, and that's the part that really surprised the team. Even people whose houses were projected to be underwater by 2100 showed the backlash effect. The map didn't make it personal enough.
What did make it personal enough?
Commute times. When researchers showed San Francisco participants how flooding would add up to 20 minutes to their daily drive, concern went up. That's a thing people feel in their bodies every day — not a scenario 75 years away.
So the problem isn't the science. It's the translation.
Exactly. The data behind the maps is solid. But the format asks people to do interpretive work they're not equipped for — to look at a blue wash on a screen and feel the weight of it.
Climate Central is still making maps, though.
They are, and they argue the maps serve planners and decision-makers well. But even their director admitted that scientists have been assuming the public reads these images the way experts do. That assumption has been costly.
What's the next experiment?
Street-level imagery — AI-generated photos of real neighborhoods under projected floodwater. The idea is to put someone on their own block, looking at their own house, and let the picture do what the map couldn't.
And nobody tested any of this before?
That's Mildenberger's sharpest point. For years, communicators sent messages without checking whether they worked. The maps felt elegant and scientific. Turns out feeling right and being effective are different things.