A new study suggests many coastal risk assessments have been starting from the wrong baseline. Researchers found measured sea level along shorelines is on average about 10.6 inches (0.27 meters) higher than what many studies treated as “zero,” and in some regions the gap tops 3.3 feet (1 meter).
This is not a claim that the ocean suddenly rose an extra foot overnight. It is a warning that a common shortcut in research can quietly shrink the safety margin in flood maps and planning tools. After reviewing 385 peer-reviewed studies from 2009 to 2025, the authors say fixing the mismatch changes who counts as “at risk” when seas rise.
A baseline error that adds up
Coastal flood studies often combine sea level and land elevation from different measurement systems. Many papers treated a global “geoid” model as if it were the same thing as local mean sea level at the shoreline. If those references are not aligned, you are stacking projections on top of a starting line that is already off.
Seeger and Minderhoud report that 90% of the hazard assessments they evaluated assumed coastal sea level from geoid models instead of using actual sea level measurements.
They also found that more than 99% handled sea level and elevation data inadequately, which matters because the coast is where a few inches can decide whether water reaches a doorstep or stops at the curb.
A big part of the problem is basic transparency. The Nature paper says 73% of the studies had incomplete or missing documentation about the vertical datums being used, and only a tiny share correctly described and aligned the datasets. That makes it harder for reviewers, planners, and the public to notice the error before it spreads.
Why the ocean does not match the geoid
A geoid is an idealized “level” surface based on Earth’s gravity and rotation. It is useful, but it is not the same as the living, moving ocean you see at a beach. What happens when wind patterns, currents, and water density push the sea surface up or down for years at a time.
Climate change adds heat, and warm seawater expands. NASA has highlighted thermal expansion as a key reason global sea level rose faster than expected during 2024, which is a reminder that the ocean stores much of the extra heat trapped by greenhouse gases.
The Nature authors also point out that global geoids depend on gravity observations, and uncertainty can reach several feet where gravity data are sparse. Those gaps are more common in parts of the Global South, so the same shortcut that looks harmless on a well-studied coastline can be risky elsewhere.
Hotspots where “zero” is most wrong
The largest underestimates show up in regions already juggling big populations and high exposure. The study flags Southeast Asia and Oceania as hotspots, and says measured coastal sea level can be more than 3.3 feet (1 meter) above global geoids in parts of the Pacific, with the biggest differences in the Indo-Pacific.
Other areas with large discrepancies include parts of Latin America, East Africa, the Caribbean, the Middle East, and sections of the west coast of North America.
Meanwhile, the lowest mismatches tend to be in data-rich regions such as eastern North America and northern and western Europe, which may help explain why the “geoid equals sea level” habit became so common.
It is also not a one-direction story. The paper notes that in some places the discrepancy can flip, including along the northern Mediterranean coast and in Antarctica, meaning geoid-based workflows can sometimes overestimate sea surface height. That is why local measurements and careful conversions matter more than ever.
A faster clock for coastal life
A 10.6-inch baseline error lands differently when the trend line is rising. Copernicus data show the global mean sea level trend rose from about 2.9 millimeters per year in 1999 to 2009 to about 4.2 millimeters per year in 2014 to 2024, while the World Meteorological Organization reports around 4.7 millimeters per year over 2015 to 2024 (about 0.17 to 0.19 inches per year).
IPCC projections suggest global mean sea level could rise about 11 to 40 inches by 2100 depending on emissions. Global averages are only part of the story, though, because local relative sea level can climb faster where land is sinking.
Now think about everyday infrastructure. Storm drains that already burp seawater during the highest tides, roads that flood during rush hour, and wells that turn salty are all sensitive to inches. When the baseline starts too low, the “later” in long-term plans can quietly become “sooner.”
What better risk maps look like
When the authors redo the math using measured coastal sea levels, the exposure picture changes sharply.
With a hypothetical 3.3-foot (1 meter) rise in relative sea level, they estimate 31 to 37% more land and 48 to 68% more people would fall below sea level than geoid-based assessments suggested. That pushes the population estimate to about 77 million to 132 million people under that scenario.
In plain area terms, their analysis suggests the land below sea level could rise from about 114,000 to 166,000 square miles to about 178,000 to 259,000 square miles, depending on which datasets are used. That extra land at risk, roughly 64,000 to 92,000 square miles, includes ports, homes, farmland, and power infrastructure.
So what do we do with a problem that is part physics and part bookkeeping. The authors call for re-evaluating existing impact studies and tightening standards so elevation and sea level data are aligned and fully documented, and they suggest journals and data providers can make correct workflows easier to follow.
The original study was published in Nature.
