Sea levels are much higher than often assumed. How is that possible?

Minderhoud looks around, across the vast delta. The landscape is as flat as a pancake. And almost everywhere he looks, he sees the same thing: the water level is far higher than the maps suggest. In many parts of the delta, the surface water level — directly connected to the sea — is not one and a half metres but only a few decimetres below the land surface. The maps simply do not match reality. And that is not because of the weather or the tides.

The maps Minderhoud consults in Vietnam come from reports by international organisations. How can it be, he wonders, that these reports present such a distorted picture? And how widespread are these inaccuracies? In the years that follow, he delves deeper into the methods and calculations underpinning the relative height of sea level and land elevation.

In the Netherlands, sea level elevation is no mystery. Since the end of the 19th century, the Netherlands has used Normal Amsterdam Level (NAP) as a reference frame for measuring the country's elevation. Nieuwerkerk aan de IJssel is 6.78 metres below NAP, while the Vaalserberg is 322 metres above it. Many people intuitively think that 0 metres NAP equals sea level, but sea level has now risen to about 10 centimetres above NAP. A house 1.50 metres above NAP is therefore actually about 1.40 metres above the current average sea level for the Dutch coast. 

In many other countries, especially in the global south, the situation is much more complicated. There is often less public data available. When coastal researchers want to know how high the water is (relative to the land) at the coast, they often use a different method: they assume local sea level to equal geoid models. These are mathematical models that indicate sea level on the globe based on gravity and the Earth's rotation. They show, as it were, the height of the “calm ocean”. 

These models do not account for other ocean dynamics that also influence local sea levels. Wind, tides and currents, for example, and the temperature of the seawater. It is therefore to be expected – and well known in the field of oceanography – that the actual average sea level at the coast will differ from what a geoid model indicates. 

The differences between global geoid models and direct measurements are relatively small for regions like Northern Europe and the east coast of the United States, says Minderhoud: ‘This could explain why Western coastal researchers, when comparing geoid models with direct sea level measurements in their own “backyard”, may have concluded these models were accurate enough for their analyses of other regions. It seems they were not fully aware that the differences between the models and the actual measurements in other parts of the world could be much greater.'

For many southern parts of the world, in particular, the difference between the geoid and the actual measured sea level is often much larger than for the west. There are two reasons for this. First, there is generally less data available in the global south to create geoid models, which means that the calculations are generally less accurate. Secondly, ocean dynamics tend to be stronger in these areas, so local sea levels are more influenced by currents, tides and prevailing winds. It is therefore very important to also use direct sea level measurements in these regions when assessing coastal impacts. 

‘There is another factor to consider, says Minderhoud: ‘In order to be able to say anything about the impact of sea-level rise on the coast, you need to know the level of the sea and the level of the land. Both are measured globally using different satellites and their elevation measurements are usually made available relative to different vertical reference frames.’ In order to correctly link the measurements of land level and sea level, researchers must therefore apply a conversion. 

Such a conversion is common practice in the field of geodesy, the science of measuring the Earth’s shape and gravity field, but it is quite a technical step. ‘If you want to know the elevation of your land relative to sea level, you have to convert the different datasets to a common reference frame first. Then you can correctly determine the relative height between the two.’ 

To calculate the impact of sea-level rise on the coast, scientists therefore require accurate measurements, as well as careful alignment of land and sea data. It takes Minderhoud several years to gather all the pieces of the puzzle for his calculations of the Mekong Delta in Vietnam.

In 2019, he published his findings in Nature Communications: the delta is elevated approximately one and a half metres lower above local sea level than was concluded in previous international studies. Presciently, Minderhoud wrote that “Our results imply major uncertainties in sea level rise impact assessments for the Mekong delta and deltas worldwide, with errors potentially larger than a century of sea level rise.” 

In the years that followed, Minderhoud's suspicion grew that the sea level rise analyses of many scientific studies might be inaccurate. German geographer Katharina Seeger, supervised by Minderhoud, discovered during her PhD research that the same inaccuracies also affected the Ayeyarwady Delta in Myanmar.

Minderhoud and Seeger jointly decided to systematically examine scientific studies on the vulnerability of coastal areas to sea level rise. What starts as a side project for Seeger alongside her PhD research soon grew into a huge, complex undertaking. The research will eventually take more than two years. 

Seeger and Minderhoud search for scientific literature published between 2009 and 2025 in the field of coastal research, addressing vulnerable coastal areas and sea level rise, among other topics. Of the tens of thousands of publications that meet their initial search criteria, 385 relevant publications remain. In addition, there are many hundreds of underlying studies and supplementary studies that are referenced for methodology. ‘There are really huge spreadsheets behind our study,’ says Minderhoud. ‘Katharina really dug into that, to figure everything out and record it down to the last detail.’ 

The systematic review of all these publications confirms their original suspicion: 99% of studies into coastal vulnerability to sea level rise do not use sea level measurements, combine them incorrectly or fail to adequately describe their methodologies. Most studies (more than 90%) only looked at land elevation and assumed the current sea level to be equal to the geoid. As a result, the coastal sea level in these studies was often underestimated. Of the studies that did use sea level measurements, some did so incorrectly or did not provide a complete and reproducible methodological description. 

Now, in March 2026, Minderhoud and Seeger are publishing their findings in the renowned scientific journal Nature. According to their new assessment, in which the conversion has been applied correctly, sea levels relative to the coast are higher than assumed in many studies almost everywhere in the world. In some regions, such as South-East Asia and the Indo-Pacific, the water is sometimes one to two metres higher than previously assumed. The reverse can also be observed: in some regions, especially around Antarctica, sea levels are slightly lower. 

How is it possible that so many scientific studies continued to use assumed sea level based on geoid models, when more accurate, global sea level measurements had been available for at least twenty years? This may have been due to the fact that most prominent studies were conducted by Western scientists, who were accustomed to the reliable results of geoid models in their own regions. As geoid assumptions worked relatively well, they may not have realized the need to use direct sea level measurements in their analyses. This assumption has led to an underestimation of coastal exposure in these studies, especially at the coasts of South-East Asia and the Indo-Pacific. 

Another possible cause: the blind spot lies precisely in between a number of traditionally disconnected scientific disciplines. Minderhoud observes that scientists often focus on their own research field: ‘Researchers who study land elevation or sea levels try to make their elevation models as accurate as possible. For them, the steps required to perform correct data coupling are common knowledge and practice. However, it seems that their data and knowledge have only recently and only sparsely reached coastal research studies. Most researchers there seem to be unaware that it is necessary to use and correctly align measurements of both land and the sea when performing coastal impact assessments.’ 

If it is up to Minderhoud and Seeger, assuming sea levels based on geoid models will soon be a thing of the past in coastal research. Using supercomputers, the Wageningen researchers have correctly linked four global elevation models to the most recent sea-level measurements. The resulting datasets have been made openly available, enabling countries and organisations to carry out more accurate coastal analyses straight away. Minderhoud hopes that this integration of land and sea-level measurements will become the new standard in coastal research.

The findings of Seeger and Minderhoud are likely to have an impact not only within the scientific community, but beyond it as well. Globally, average coastal sea levels are around 30 centimetres higher than the levels assumed in the majority of the studies examined. This means that, in reality, many more people live in low-lying areas close to sea level. If sea levels continue to rise as a result of climate change, more people could be directly affected than was previously assessed. 

Should countries such as Vietnam or tropical island groups such as the Maldives now completely overhaul their policies? Should they urgently start building dykes, because the local sea level is already much higher than the analyses indicated? That depends, says Minderhoud: ‘In the Mekong Delta, the Vietnamese government was already relying on its own, more accurate data and projections, rather than international studies. So the impact of the new findings will be less significant there. But at this point, it is unknown whether that applies to all countries.’ 

‘Our publication demonstrates how to improve our coastal impact studies. That is how science works’, Minderhoud concludes. ‘Now that we have discovered this blind spot, the scientific community can make more accurate assessments for coastal areas and cities around the world. This will help, for example, to determine which areas are most vulnerable to future sea level rise and where coastal adaptation strategies are most urgently needed. 

‘Our findings also suggest that governments and policymakers, especially in the most affected regions, may need to evaluate whether the methodological issues that we discovered have affected the information they’ve been using. That could be relevant for their coastal adaptation and sea-level rise protection strategies.’ 

Because in the meantime, sea levels continue to rise.