Coastal flooding is usually discussed on one of two timescales: the next storm, or the sea-level rise expected by the middle or end of the century. A new study adds a less familiar interval between them. It finds that changes in Atlantic Ocean circulation may help predict how often high-tide flooding occurs along the U.S. Southeast Coast as much as three years in advance.
That is a meaningful scientific development. It is not evidence that the Atlantic Meridional Overturning Circulation, or AMOC, is about to collapse. Nor does it tell a particular city which street will flood, how deep the water will be or when a hurricane will arrive.
The result is narrower and more useful: some of the ocean variability that raises or lowers regional sea levels appears predictable far enough ahead to inform planning.
Three causes of coastal flooding are being separated
The study, published in Communications Earth & Environment in June 2026, examines the Southeast Coast rather than the entire Atlantic seaboard. The authors report that flooding frequency has increased since approximately 2010 as a result of rising background sea levels, storms and internal ocean variability. Background water levels largely determine how often minor flooding occurs; storms remain responsible for the most severe individual events.
The distinction matters. A higher baseline means that an ordinary high tide can become disruptive without a hurricane. A storm then operates on top of that elevated water level, potentially producing more serious flooding than it would have produced in the past.
The researchers argue that AMOC-related changes in the distribution and movement of heat and water affect regional sea level. They combined a high-resolution regional ocean model with AMOC predictions from the National Oceanic and Atmospheric Administration’s Geophysical Fluid Dynamics Laboratory decadal prediction system. The resulting forecasts showed useful skill in predicting flood frequency up to three years ahead, according to the NOAA Institutional Repository record for the study.
Predictable does not mean certain
“Three years ahead” should not be confused with a deterministic forecast. The study does not predict the exact number of flood days for every community, nor does it forecast individual storms. It identifies a regional probability signal that may help indicate whether a period is likely to be more or less flood-prone than the longer-term trend would suggest.
That is the same distinction that applies to seasonal climate outlooks, although the physical processes are different. A forecast can provide useful information for decisions without specifying the outcome of every event. The practical question is whether the uncertainty is small enough, and the potential consequences large enough, to justify action.
NOAA’s existing high-tide flooding outlooks already combine long-term sea-level change with large-scale climate influences. Its 2026–27 national outlook projects a median of seven to 12 high-tide flood days at water-level stations from May 2026 through April 2027, with local results varying by region and station. NOAA also emphasizes that flooding thresholds are site-specific and that the consequences depend on local geography and infrastructure. The agency’s annual high-tide flooding outlook is therefore a useful reminder that a flood day at one gauge is not automatically equivalent to the same level of disruption somewhere else.
The AMOC is not the Gulf Stream
Public discussion often treats the AMOC and the Gulf Stream as interchangeable. They are related, but they are not the same system. The AMOC describes a basin-scale circulation involving the movement of heat and mass through the Atlantic. The Gulf Stream is one component of the broader circulation.
That distinction is important because a study of AMOC-related sea-level variability should not be presented as evidence that the Gulf Stream is simply “stopping.” The paper makes a more limited claim about regional sea-level effects and their predictability.
The observational record also has limits. Direct monitoring of the AMOC is comparatively recent, so scientists use a combination of instruments, satellite measurements, models and reconstructed records. That approach is standard in climate research, but it means that some conclusions remain dependent on model structure and on how different observations are combined.
What the study establishes—and what it does not
The evidence supports several conclusions:
- Sea-level rise is increasing the background conditions for recurrent coastal flooding.
- Regional ocean circulation can temporarily amplify or moderate those conditions.
- AMOC-related variability appears to provide useful predictive information for Southeast Coast flood frequency.
- That information may extend several years beyond ordinary weather and storm forecasts.
It does not establish that the AMOC is collapsing, that a collapse is imminent or that every coastal flood is caused by ocean circulation. It does not show that local emissions controls or flood-protection projects are ineffective. And it cannot replace tide gauges, storm forecasts, local engineering studies or observations of land subsidence, tides, winds, rainfall and infrastructure conditions.
The broader debate over future AMOC weakening remains legitimate. So do questions about how much of recent regional sea-level variability should be attributed specifically to the AMOC rather than to winds, atmospheric pressure, other climate modes or land movement. But those questions should not obscure the study’s narrower contribution.
The institutional test comes next
A three-year flood-frequency signal could, in principle, inform drainage and pump maintenance, road and bridge schedules, utility assessments, port operations, emergency staffing and insurance reviews. It might help a public agency decide whether to advance a maintenance project, inspect a vulnerable system or avoid treating a temporary reduction in flooding as evidence that the underlying risk has disappeared.
But a forecast does not answer who should act or who should pay. A city may know that several unusually flood-prone years are more likely without having the money to raise a road, replace a drainage system or relocate a vulnerable facility. Agencies may also hesitate to alter budgets in response to a probabilistic signal that has not been independently tested across many regions and forecast periods.
That is why the next stage is not simply better prediction. It is evaluation of use. Researchers will need to show how often the system produces false alarms, misses elevated flood periods or performs differently across coastal locations. Public agencies will need to determine whether the forecasts improve decisions enough to justify the costs of incorporating them into planning.
The most important development here is not that the ocean has become predictable in a general sense. It has not. It is that scientists may be identifying a specific, measurable source of intermediate-term information about coastal risk.
We know that rising seas are making high-tide flooding more common. We know that regional ocean dynamics can make some periods worse or better than the long-term trend alone would suggest. We now have evidence that part of that variability may be predictable several years ahead.
The unresolved question is whether institutions can use that warning before the water arrives—or whether the forecast remains another scientifically credible product that sits outside the decisions governing roads, pumps, utilities, insurance and public safety.
