A systematic review published in Nature finds that roughly 90% of coastal hazard and sea level rise exposure studies contain vertical-datum errors that make baseline water heights appear lower than they actually are. The average underestimation is about 0.3 meters, or roughly 1 foot, and the problem is most pronounced in the Global South and the Pacific. The findings suggest that millions of people living along vulnerable coastlines face greater flood risk than current models indicate.
What the Study Actually Found
The research, a meta-analysis in Nature, reviewed hundreds of individual studies and hazard assessments that estimate how many people and how much infrastructure sit in flood-prone coastal zones. Its central finding is straightforward but alarming: the vast majority of these assessments start from a baseline sea level that is too low. That error does not stem from bad tide gauges or satellite altimeters. It comes from a mismatch between the elevation datasets researchers use to map the land and the reference surfaces they use to represent the sea.
In practical terms, if a study says a neighborhood sits 2 meters above sea level, the actual margin may be closer to 1.7 meters once the vertical-datum error is corrected. That 0.3-meter gap matters enormously for low-lying deltas and island nations, where a few centimeters can determine whether a storm surge overtops a seawall or stays just below it. When that misalignment is baked into national risk assessments, it can quietly distort everything from evacuation plans to long-term coastal zoning.
The authors emphasize that this is a structural problem, not a handful of sloppy case studies. By systematically comparing how land elevations were referenced to geodetic surfaces and how sea levels were tied to tidal datums, they found that about nine in ten exposure estimates were skewed in the same direction: they made coasts look safer than they really are. In some regions, the discrepancy between modeled and actual water heights exceeded the global 0.3-meter average.
How Elevation Data Creates a Blind Spot
Most global coastal hazard work relies on digital elevation models, or DEMs, to determine how high the land sits relative to the ocean. One of the most widely used products is the Copernicus DEM, maintained by the European Space Agency and the European Union. These datasets measure land height relative to a mathematical surface such as an ellipsoid or a geoid, which approximates the shape and gravity field of the Earth. Sea level, by contrast, is typically referenced to a tidal datum, a surface defined by years of local tide measurements that capture how water actually behaves along a particular stretch of coast.
The problem arises when researchers overlay land elevation data referenced to one surface onto sea level data referenced to a different one without properly converting between the two. That conversion requires detailed knowledge of local gravity, ocean dynamics, and tide-gauge records; skipping it is tempting when working at global scale or in data-poor regions. The result is a systematic bias: the ocean appears lower relative to the land than it truly is, and coastal communities appear to sit higher and drier than they actually are.
The Nature study identifies this class of error in approximately 90% of the assessments it reviewed, making it not a rare oversight but a near-universal feature of the field. Because the bias is directional (almost always making risk look smaller), it cannot be dismissed as random noise that cancels out over large areas. Instead, it acts like a hidden discount applied to flood danger, one that policymakers and residents never explicitly agreed to.
Why the Global South Bears the Brunt
The vertical-datum mismatch appears more frequently in the Global South and the Pacific, where tide gauge networks are sparser and local geoid models are less precise. Countries in West Africa, Southeast Asia, and small island states often lack the dense ground-truth data needed to calibrate satellite-derived elevation products against actual measured coastal water levels. Where those reference surfaces are poorly constrained, the errors in converting between ellipsoidal height and tidal datum can easily exceed the global average.
Consider a place like Suriname, where stretches of coast are already experiencing erosion and saltwater intrusion driven by rising seas. If hazard assessments there rely on global DEMs without correcting for local tidal datums, the resulting flood maps will systematically undercount the land and population at risk. In low-lying deltas, a nominal 0.3-meter underestimate can translate into tens or hundreds of square kilometers of additional floodplain, much of it occupied by informal settlements with limited protective infrastructure.
These blind spots have equity implications. Wealthier countries typically have long-running tide-gauge records, detailed geodetic surveys, and technical agencies capable of performing complex datum transformations. Many poorer countries do not. Yet it is precisely those nations, with rapidly growing coastal cities and limited resources, that depend most heavily on international risk assessments to guide adaptation funding and disaster preparedness. When the underlying data systematically understates their exposure, they risk being underprioritized in global climate finance and insurance markets.
A New Dataset Aims to Close the Gap
Alongside the critique, the researchers have released what has been described as a corrected elevation product designed to address these baseline errors. The goal is to give hazard modelers a dataset that already accounts for the difference between ellipsoidal heights and local tidal datums, removing the need for each research team to perform its own conversion, a step that many have evidently been skipping or performing incorrectly.
In parallel, a related effort known as the DeltaDTM initiative is building a global coastal digital terrain model tailored to low-lying coasts and river deltas. DeltaDTM focuses on exactly the kinds of landscapes where small elevation errors have the largest consequences, blending satellite data with local measurements to refine ground heights at high resolution. Together, these products represent a practical attempt to move the field past a known flaw rather than simply documenting it.
The authors argue that adopting such corrected datasets should become standard practice for any national or regional flood-risk assessment. Where bespoke local models already exist, those can be reconciled with the new global products; where they do not, the corrected datasets offer an immediate upgrade over unadjusted DEMs. The key is to make the choice of vertical reference explicit and consistent from the outset, rather than an invisible afterthought.
What This Means for Flood Risk Planning
The implications extend well beyond academic publishing. National adaptation plans, multilateral climate finance, and local building codes all depend on estimates of how much land will flood under various sea level rise scenarios. If those estimates systematically start from a baseline that is roughly 1 foot too low, the cascade of consequences is significant. Seawalls may be built to the wrong height. Evacuation zones may be drawn too narrowly. Insurance models may underprice risk in the very places where losses will be greatest.
For international development agencies and climate funds that allocate resources based on vulnerability indices, the study raises a pointed question: are the numbers guiding those allocations reliable? If correcting the baseline shifts the boundary of flood-prone land outward by even a modest amount in densely populated deltas like the Ganges-Brahmaputra or the Mekong, the count of people living in high-risk zones could jump substantially. That, in turn, would change the math on where adaptation money is most urgently needed and how it should be sequenced over time.
The findings also intersect with how media organizations and their audiences engage with climate risk. Readers who regularly follow in-depth coverage, whether through a weekly print subscription or other formats, increasingly expect that headline numbers about people at risk rest on solid technical foundations. As new evidence emerges about systematic errors, it forces a re-examination not only of models but also of how their results are communicated to the public.
A Critique of Standard Practice, Not of Climate Science
One distinction worth drawing clearly is that this study does not challenge the science of sea level rise itself. It does not dispute that oceans are warming and expanding, or that ice sheets are losing mass. What it challenges is the engineering step that translates those physical changes into on-the-ground risk estimates. The error lies in the measurement framework and data handling, not in the underlying climate signal.
That distinction matters because it changes the nature of the fix. Addressing a vertical-datum mismatch is, in principle, a solvable technical problem: standardize reference surfaces, improve geoid models, expand tide-gauge networks, and require that exposure studies document their vertical references as rigorously as their emissions scenarios. It also calls for better collaboration between geodesists, oceanographers, and risk modelers, so that choices about datums and transformations are treated as core design decisions rather than obscure back-end settings.
For institutions that commission or rely on coastal risk assessments, the immediate step is due diligence. Agencies can require that new studies explicitly state which vertical references they use and how they convert between them, and they can prioritize corrected elevation datasets as defaults. Newsrooms and civil society groups, supported by engaged readers who may log in through digital accounts or contribute via reader support, can then scrutinize and explain these technical shifts rather than simply repeating outdated figures.
Ultimately, the Nature review is a reminder that in climate risk, seemingly arcane details, like which invisible surface you measure “height above sea level” from, can have very concrete human consequences. Correcting a 0.3-meter error on paper may mean the difference between a community being counted in the danger zone or left outside the lines, between a funded seawall and an unprotected shoreline. As sea levels continue to rise, getting those baselines right becomes not just a technical nicety but a matter of climate justice.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
