Once-in-a-Century Coastal Floods Now 12× More Likely

TL;DR

• Once-in-a-century coastal floods are now 12 times more frequent — what was a 1% annual risk in 1900 is now roughly an 8-year event, on average, across more than 100 tide-gauge sites worldwide.
• The cause is unambiguous. Since the 1960s, human-driven sea-level rise — not natural variability, not land subsidence — has been the dominant force. Anthropogenic forcing alone has quadrupled the frequency of these extremes.
• The findings almost certainly understate today's risk. The study covers 1900–2005. Emissions and sea-level rise have only accelerated since.
• In some places, the shift is staggering. Wellington, New Zealand, now experiences what was once a 100-year coastal flood roughly twice a year.

What Happened

A study published June 10 in Nature Climate Change — and widely covered in the week since — has produced the most comprehensive attribution yet of how human-driven sea-level rise has reshaped coastal flood risk. The paper, led by Sönke Dangendorf of Tulane University with co-authors from the University of Central Florida, Harvard, the University of Bremen, Utrecht University, NIOZ Royal Netherlands Institute for Sea Research, and IMEDEA in Spain, combined tide-gauge records from more than 100 sites with climate model simulations to disentangle natural variability from anthropogenic forcing.

The headline finding: floods that historically carried a 1% annual probability — the so-called "100-year flood" — are now, on average, 12 times more likely. The global mean sea level has risen roughly 20 centimetres (nearly 8 inches) since 1900. That sounds modest. It is not.

A companion study published in Science Advances, co-authored by Ben Strauss of Climate Central, found that climate change was responsible for roughly 58% of the days with major floods between 2000 and 2018, and that the number of days where the sea rises above extreme flood levels has nearly tripled since the 1970s.

"Essentially every coastal flood today has human fingerprints on it through climate change," Strauss said. "Without the extra bit of sea level rise caused by global heating, most of these events would not have reached the status of flood."

What It Actually Means

The study's real contribution is not the 12× figure — arresting as it is — but the attribution. The authors isolated the human signal from the noise of natural climate variability, local land movement, and other confounding factors. The conclusion: anthropogenic forcing alone accounts for a fourfold increase in the likelihood of 100-year events. Since the 1960s, it has been the dominant driver.

This matters because it transforms the conversation from "is climate change making floods worse?" to "by how much, and what do we do about it?" The attribution is now precise enough to inform infrastructure planning, insurance pricing, and coastal zoning.

The mechanism is straightforward but often misunderstood. A 20-centimetre rise in mean sea level does not simply add 20 centimetres to a storm surge. It shifts the entire baseline from which extremes are measured. A surge that once crested just below a seawall now overtops it — not because the storm is worse, but because the sea started higher. As Ben Marzeion of the University of Bremen put it: "These 20 centimetres aren't just added on top — they shift the entire baseline."

Thomas Wahl of UCF offered a metaphor that has travelled widely in coverage of the study: "If you live within FEMA's 100-year flood zone, you have a 100-sided die that you roll every year. Now, because of sea level rise, that die is losing sides."

What This Is Not

This is not a projection. It is a measurement of what has already happened — from 1900 to 2005. The researchers explicitly note that their findings "probably understate today's risk, because human contributions to changes in coastal extremes have only increased since then."

It is also not uniform. The 12× figure is a global average. In some regions, the shift is far more dramatic — Wellington now sees historical 100-year floods roughly twice a year. In others, local factors such as land uplift partially offset the effect. The study is a global assessment, not a local risk map, and should not be read as a replacement for site-specific flood modelling.

Stakeholder Landscape

Who is directly affected: The hundreds of millions of people living in low-lying coastal zones — from Bangladesh to the Gulf Coast of the United States, from the Netherlands to the Mekong Delta. Also: coastal infrastructure operators, port authorities, and anyone holding a 30-year mortgage on coastal property.

Who needs to act now: Municipal planners, civil engineers, and flood-management agencies whose design standards are based on historical flood-frequency estimates that no longer represent reality. Jeff Williams, a retired USGS oceanographer not involved in the study, noted that the current protections for New Orleans "will likely not be adequate beyond the next couple decades."

Who benefits from the noise: This is not a story with obvious commercial beneficiaries, but the insurance industry — which has been quietly repricing coastal risk for years — will find in this study a data point that supports decisions already underway.

Cross-Layer Implications

Insurance and finance. The 100-year flood is the actuarial bedrock of coastal insurance. If it is now an 8-year event, the mathematics of insurability shifts. Expect accelerating withdrawal of private insurers from high-exposure coastal markets, and growing pressure on government-backed flood insurance programmes.

Infrastructure planning. Most coastal defences were designed against historical return periods. This study provides the clearest evidence yet that those return periods are obsolete. The question is not whether to rebuild but whether to retreat — and the study makes clear that the window for managed retreat is narrowing.

The emissions connection. Dangendorf noted a "silver lining": the world is no longer on track for the worst-case warming scenario, and emissions choices still matter. But the study also makes clear that even the sea-level rise already locked in will continue to drive up flood frequency for decades.

What This Means for You

If you live near the coast: Check whether your local flood maps reflect post-2005 sea-level data. Many do not. The FEMA flood zone you are in may already be outdated. Consider that flood insurance priced for a 100-year event is now priced for something closer to an 8-year event — and premiums have not yet caught up.

If you are a planner or engineer: The study's methodology — separating anthropogenic forcing from natural variability using tide-gauge records and climate model simulations — provides a template for updating local flood-frequency estimates. The historical record is no longer a reliable guide.

If you are a policy-maker: The study strengthens the case for both mitigation (emissions reduction slows the rate of sea-level rise) and adaptation (coastal defences, zoning, managed retreat). The two are not alternatives; both are necessary.

If you are none of the above: The honest answer is that there is little an individual can do beyond staying informed and supporting policies that address both emissions and adaptation. The study's value to the general reader is clarity: the risk is not theoretical, not distant, and not natural.

Uncertainty Ledger

• The study ends at 2005. The acceleration since then is not captured. The true present-day risk is almost certainly higher than the 12× figure.
• Regional variation is enormous. The global average masks places where the shift is far worse (Wellington) and places where it is less pronounced. Local modelling is essential.
• Storm behaviour is not factored in. The study examines sea-level extremes, not changes in storm frequency or intensity. If climate change also strengthens storms — as separate research suggests — the combined effect on coastal flooding would be greater than either factor alone.
• The Science Advances companion study found that 58% of major flood days from 2000–2018 were attributable to climate change. The remaining 42% reflects natural variability — a reminder that attribution is not 100%, even when the signal is dominant.

Bottom Line

The 100-year coastal flood is dead as a planning concept. What was once a statistical abstraction — a 1% annual chance — is now, on average, a once-per-decade reality, and in some places an annual event. The cause is not natural variability, not bad luck, and not local subsidence. It is human-driven sea-level rise, and it has been the dominant factor since the 1960s. The study's most uncomfortable finding is not the 12× figure but the implication that it is already out of date. The data stops at 2005. The seas have not.

Sources:

• Dangendorf, S., Sun, Q., Maduwantha, P., Wahl, T., Marcos, M., Marzeion, B., Slangen, A. B. A., & Mitrovica, J. X. (2026). Human-driven sea-level rise has quadrupled the frequency of coastal sea-level extremes since 1900. Nature Climate Change. DOI: 10.1038/s41558-026-02659-0. [Tier 1]
• St. John, A. (2026, June 17). Climate change is making once-rare coastal floods more likely, study says. Los Angeles Times / Associated Press. [Tier 1]
• University of Central Florida. (2026, June 15). UCF coastal expert: Extreme coastal water level events are now 12 times more likely to occur. EurekAlert. [Tier 2]
• University of Bremen. (2026, June 10). What was previously a once-in-a-century storm surge now occurs every eight years. EurekAlert. [Tier 2]
• Strauss, B. et al. (2026). Science Advances (companion study on attribution of extreme flood days). Referenced via LA Times/AP coverage. [Tier 1]