A systematic review of nearly 400 coastal hazard studies has found that the vast majority relied on flawed assumptions about where sea level actually sits, leading to significant underestimates of flood exposure in low-lying regions worldwide. The research, led by Christopher Seeger and Philip Minderhoud and published in Nature on March 4, 2026, concludes that about 90% of assessments underestimated baseline coastal water heights by roughly 30 centimeters on average, with offsets reaching up to 150 centimeters in parts of Southeast Asia and the Indo-Pacific. The finding carries direct consequences for adaptation planning, climate finance, and the hundreds of millions of people living in coastal zones.
A 30-Centimeter Blind Spot in Flood Models
The core problem is deceptively simple. Most coastal hazard assessments determine how high the sea sits by referencing a mathematical model of Earth’s gravity field, known as the geoid, rather than using actual tide gauge or satellite altimetry measurements of local mean sea level. The geoid provides a useful global approximation, but it does not account for persistent ocean currents, temperature gradients, and salinity patterns that push real water levels above or below the modeled surface. Seeger and Minderhoud’s review of global coastal assessments found that more than 99% handled sea-level and land-elevation data inadequately, and roughly 90% assumed coastal sea levels directly from global geoid models.
The resulting gap is not trivial. Supplementary exposure spreadsheets from the study put the global mean offset at 0.24 to 0.27 meters. That quarter-meter difference may sound modest, but in flat deltaic terrain where a few centimeters of elevation determine whether land floods or stays dry, it changes the calculus entirely. In Southeast Asia and the Indo-Pacific, the offset reaches up to 150 centimeters, meaning some of the world’s most densely populated coastlines face far greater present-day exposure than models have indicated.
Why the Error Persisted for Over a Decade
The technical roots of this problem have been documented for years, which makes the scale of its persistence striking. Dean Gesch published a vertical datum guide in 2018 laying out the requirements for using elevation data in sea-level rise and coastal flood analyses, including proper vertical alignment, uncertainty quantification, and appropriate use of digital elevation models. Yet the new review shows that the field largely continued to skip those steps, defaulting instead to off-the-shelf global products.
Part of the explanation is convenience. Global geoid-referenced elevation datasets are freely available and easy to plug into geographic information systems. Converting them to local mean sea level requires integrating mean dynamic topography data, a step that adds complexity and computing time. Seeger and Minderhoud addressed this barrier directly by producing processed elevation layers already converted to local mean sea level, hosted on Zenodo alongside the code used for their datum conversions and exposure calculations. The intent is to remove the practical excuse for skipping the correction and to standardize better practice for future work.
Another factor is institutional momentum. When early studies used geoid-based assumptions and were later cited by subsequent research and policy documents, the error propagated through the literature. The supplementary screening of cited papers in the new review catalogued which of the 385 evaluated analyses appeared in major assessment reports, including the IPCC’s AR6 and Special Report on the Ocean and Cryosphere. That linkage raises uncomfortable questions about whether the projections informing international climate negotiations rest on systematically biased baselines and whether adaptation priorities have been skewed as a result.
Delta Case Studies Foreshadowed the Problem
The new meta-analysis did not emerge from a vacuum. Minderhoud co-authored a 2019 study showing that datum errors in the Mekong Delta materially changed flood risk estimates. That research demonstrated how elevation referencing could cause large errors in sea-level rise impact assessments for one of the world’s most vulnerable lowland regions, with implications for agriculture, infrastructure, and rural livelihoods. When the land surface was correctly referenced to local mean sea level, much larger areas of the delta were already at or below high-tide levels than previous maps had suggested.
Separately, Seeger led a 2023 assessment of the Ayeyarwady Delta in Myanmar, which reinforced that coastal elevation relative to local sea level must be handled with care and that vertical datum alignment and digital elevation model limitations strongly affect flooding and sea-level rise vulnerability conclusions. That work found that widely used global elevation products smoothed over embankments, canals, and subsiding polders, leading to underestimates of the population at risk from storm surges and monsoon-driven floods.
Those delta-scale studies served as proof of concept. The 2026 paper scales the same logic globally, and the results suggest the Mekong and Ayeyarwady were not outliers but symptoms of a systemic methodological gap across the entire field of coastal hazard assessment. By systematically comparing geoid-based assumptions with observed coastal water levels, the authors show that the same kinds of misalignments recur in deltas, estuaries, and low-lying coastal plains on every continent.
What Higher Baselines Mean for Flood Exposure
The practical stakes become clear when the corrected baselines are applied to future scenarios. Under a hypothetical 1 meter of relative sea-level rise, the authors report that using local mean sea level instead of the geoid increases the estimated global population exposed to annual flooding by tens of millions of people. Their comparison with geoid-based assumptions shows that exposure estimates in some low-lying regions more than double once the higher starting water level is taken into account. In parts of Southeast Asia, West Africa, and small island states, land previously classified as safe under mid-range climate scenarios is already within reach of extreme tides and storm surges.
The shift in baseline also alters the perceived benefits of protective infrastructure. Sea walls, levees, and surge barriers designed using geoid-referenced elevations may in reality offer less freeboard than intended, especially where land subsidence is ongoing. When the water surface is 30 centimeters higher than assumed, the margin before overtopping shrinks accordingly. That miscalculation can affect cost–benefit analyses, insurance pricing, and the prioritization of nature-based solutions such as mangrove restoration.
For planners and financiers, the message is that existing exposure maps and risk models should not be taken at face value without checking how sea level was defined. Projects that appeared marginal under older assessments may now clear benefit–cost thresholds once the higher baseline is included, particularly in densely populated deltas where a small vertical shift translates into millions more people and billions of dollars in assets at risk.
Implications for Policy and Practice
The authors argue that correcting this blind spot is both technically feasible and urgently necessary. Because the error stems from how data are referenced rather than from deep uncertainty about future climate trajectories, it is amenable to rapid improvement. The availability of pre-processed, mean sea level–referenced elevation datasets, along with openly shared code for datum conversion, lowers the barrier for agencies and consultancies that lack in-house geodesy expertise.
At the same time, the review underscores that technical fixes alone are not enough. Journals, funding agencies, and international assessment bodies may need to tighten methodological standards, requiring explicit documentation of vertical datums, uncertainty ranges, and validation against local observations. Without such guardrails, the path of least resistance will continue to favor quick geoid-based analyses that understate present-day exposure and delay difficult adaptation decisions.
For communities on the front lines, the consequences of inaction are concrete. Underestimated flood risk can lead to underdesigned defenses, misplaced development, and inadequate disaster preparedness. Conversely, more accurate baselines can support earlier and better-targeted investments in relocation, elevation of critical infrastructure, and ecosystem restoration that reduces wave energy and erosion. As coastal cities and deltas grapple with accelerating sea-level rise, getting the starting line right may prove just as important as narrowing the range of future climate outcomes.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
