Skip to Content

Saving Holland

With much of its land already below sea level, the Netherlands is charting a course around the ominous trends of climate change.

The lowest point in western Europe is 6.74 meters below sea level and falling. It lies in a boggy area of decomposing peat outside the cheese mecca of Gouda, the Netherlands, and is identified by a seven-meter marker plunked into a brackish pool at the entrance to the Van Vliet truck dealership. (The dealership’s owner erected the marker, taking a little license with the facts; the actual low spot is a few hundred meters away.) The Fodor’s travel guide does not mention this corner of ­Holland, but it’s a focal point for the question of how to plan for the risks and realities of climate change.

Never again: A deadly 1953 flood prompted a massive Dutch effort to build seawalls and surge barriers. Climate change is prompting a new shift, toward intensive flood-risk analysis and resilient construction.

That’s because the town of Gouda is considering whether to erect 4,000 houses–some of which might float–just two kilometers from this continental nadir. Subdivisions may rise on portions of the sparsely developed farmland near the truck dealership, a 50-square-kilometer area surrounded by dikes and a canal. Such reclaimed lowlands are called polders; they’re kept dry by pump houses that suck away rainwater and ground­water seepage. The Dutch have always built on polders, but doing so now, as flood risks rise across the country, will require new approaches that could get an early test in this particularly low region, called the southwest polder or Zuidplaspolder. “It sounds, sometimes, somewhat illogical,” concedes Marco van Steekelenburg, an urban planner with the regional province of South Holland, who took me to the site. “But that is what we have to investigate: how illogical it is. We have been given a challenge: can we find solutions which are climate-proof?”

The country faces ominous trends as global temperatures rise. Already, 55 percent of the Netherlands’ land area is below sea level, protected by a vast system of seawalls, storm-surge barriers, and thousands of dikes that crisscross the countryside. Dutch scientists say sea levels in the region will rise between 25 and 85 centimeters (10 and 33 inches) this century. In addition, weather worldwide is expected to become more extreme, on average. This means a higher likelihood of flooding along the Rhine and other rivers, and a greater risk of droughts. All the while, Dutch land will continue to sink–at a rate of 0.2 centimeters annually in some areas–as the peat soil underlying much of it decomposes, exposed to air by Dutch drainage efforts.

Now, in an effort being watched around the world, the Dutch government and several prominent research institutions are trying to figure out how to adapt a whole country to the realities of climate change. The Zuidplaspolder is one of several regions under consideration for developments that float–or can float, or at least are flood resistant. Apart from one well-hyped residential demonstration project elsewhere in the country, though, no actual construction has yet begun. Behind the scenes, the Dutch are taking a hard look at their growing vulnera­bilities, conducting new analyses and running computer simulations. The researchers hope to submit a plan for adapting to climate change to the Dutch parliament this fall and to take action on it by next year.


  • View images of the Dutch's approach to climate change.

  • Watch video of the floating houses, of flood simulations, and of animations of historical floods.

0707END_C_x600.jpg0707END_A_x600.jpgWater world: Prototype “amphibious” houses shown here flank a marina in rural Maasbommel. One such house (above) can rise four meters, guided by pilings. Similar concepts await implementation until a national climate adaptation plan is complete.
Credit: Siebe Swart

It’s all about understanding risk. This means understanding exactly how breaches in dikes and seawalls could lead to flooding in farms and cities–how deep the waters could get and how fast they might rise. It means understanding how those risks will change under various conditions, such as higher sea levels and different weather patterns. It means calculating how many people and how much property lie in the path of such floods, and how future changes in land use could worsen or mitigate any damage. Once they understand all that, the Dutch can plan, build, and fortify in ways that meet specific local needs. This could mean designating certain areas as no-build zones, developing new building techniques, adding floatable roadways as escape routes, or replacing farms with floating greenhouses. All this represents a major shift in focus, from preventing floods to minimizing damage when they occur. “You are looking at a different type of planning,” says Pieter ­Bloemen, a program manager with the Netherlands’ National Spatial Planning Agency.

Historically, the Netherlands–like most other countries– has thought about disaster protection in simple but reassuring terms. The Netherlands’ North Sea flank is sheltered by great seawalls and tidal-surge barriers erected after a 1953 flood that killed nearly 2,000 people. The strongest of these were engineered to withstand all but the rarest disasters: floods that have only a 1-in-4,000 chance–even a 1-in-10,000 chance–of happening in a given year. But these probabilities were calculated around 1960 and are understood to be obsolete. What’s more, the Dutch population surged from 11.5 million in 1960 to 16.6 million this year, greatly raising the stakes. “Most people in the Netherlands have the idea they are safe from flooding, because of all the investment,” says Hans Balfoort, a senior policy advisor in the Netherlands’ Ministry of Transport, Public Works, and Water Management in The Hague. “It gives people a false idea of absolute protection.”

The Dutch program for intensive risk analysis and planning is gaining attention in geographically similar parts of the world, including the Mississippi River Delta in Louisiana and the Sacramento River Delta in California. “There is a lot of pressure that places like New Orleans should adopt this kind of approach for planning and protection,” says Rafael Bras, a hydroclimatologist at MIT. He notes that in the U.S., as in the Netherlands, planners and engineers have historically focused on the strength of seawalls and levees, not the extent of the destruction that could occur if they failed. “In the past, our approach has been, ‘We will protect to a certain-level hurricane,’ without trying to translate that into, ‘What does that mean in terms of risk to the population?’”

Dams of Rotterdam
The risk analysis the Dutch are performing requires basic information about the country’s network of seawalls and dikes, but collecting it is no easy task. Much of the nation was reclaimed from the sea piecemeal over the past 800 years: farmers drained land, dug canals, and built dikes, giving rise to the saying “God made the world, but the Dutch made the Netherlands.” Local democratic bodies called water boards–hundreds of them–emerged to manage and maintain flood barriers and pumping stations. But the accretion of locally built fortifications did not leave a legacy of centralized, accurate records. “We are not Swiss,” laments Jaap Kwadijk, a geologist with the engineering firm Delft Hydraulics.

To understand how un-Swiss the Dutch are, consider the political history of the southwest polder, the area near Gouda where the 4,000 houses may rise. Historically, a chunk of the polder was managed by a water board called the Schieland, organized in 1273. An adjacent water board, the ­Krimpenerwaard, was founded in the 1430s, cobbled together from smaller boards. Dramatic change came in 2005. After 535 years of peaceful coexistence, the ­Krimpenerwaard joined with the Schieland. The merger is marked by a promi­nent sign inside the region’s biggest pump house, designating the board the “Hoogheemraadschap van Schieland en de Krimpenerwaard.” Pump-house engineer Harry Berkouwer, whose family has lived for more than 600 years in the hamlet of Berkenwoude (its name means “birch forest”; his means “birch cutter”), beams when he speaks of the event.

Despite such mergers, there are still 27 separate water boards in a country that’s only about twice the size of New Jersey. And the hodgepodge of record-keeping makes painstaking work for engineers like Arthur Mynett, director of research and development at Delft Hydraulics and an engineering professor at the Delft University of Technology. His group is probing for potential failure points in existing Dutch dikes; that requires knowing the precise height and engineering characteristics of every one of them. “We are trying to integrate everything,” Mynett says. “If one of these dikes goes, collapses, that has an effect on the probability that others will go. Some might have a higher, others might have a lower probability. It is not that trivial to find out. From history, the Netherlands is a country which is also separated in rather small administrative units–not only the water boards, but municipalities and railroads. All these organizations have their own databases. It’s already quite an effort, let’s say, that you can even use it all.”

But his group is making progress; it is now running ­simulations to show how floodwater could cascade from polder to polder, farm to farm, and street to street under various failure scenarios. (Video of flood simulations and animations.) Nathalie Asselman, a staff hydrolo­gist at Delft Hydraulics, gave me a demonstration with a few clicks of a mouse. On her computer flashed a map of the city of Rotterdam. She ran a recently created model of two ­possible disasters. In the first, a major levee failed, and blue, representing water, rapidly filled an empty area impounded by a second levee. From there, various shades of blue–indicating different depths–trickled out slowly to parts of the city, rising to a height of about a meter over several days. That would be serious, but not life-threatening or city-wrecking. Then Asselman showed what would happen if the second levee weren’t there and a major storm-surge barrier several kilometers away were left open. What unfolded would, if it happened in real life, dwarf the New Orleans catastrophe. In a matter of hours, much of Rotterdam was awash in blues, with flooding as high as three meters in some areas.


Reassessing risk: The Dutch dike system was tailored to flood probabilities calculated around 1960. Faced with climate change and population growth, the Dutch are now seeking systemic forms of protection, from upstream water impoundments to floatable buildings and roadways.
Credit: Carol Zuber-Mallison

That the Dutch haven’t previously tried to understand the consequences of calamities in such detail points to the irony of having strong defenses. No catastrophic flood has befallen the nation since 1953. That freedom from disaster has bred complacency. “Sometimes a plane falls down and you can investigate why it falls down,” says Kwadijk. “The trouble is, we never get any flooding, so you can’t test anything, and you can’t convince the public [of the danger].” But all that changed in 2005, when the Dutch were transfixed by the destruction in New Orleans following Hurricane Katrina. “Katrina raised awareness in the Netherlands,” says Mynett. “To the general public, it wasn’t, ‘Silly Americans can’t take care of water management.’ It was, ‘Oops–this can happen.’ It is more a feeling of solidarity.”

Modeling the Rhine
Pinpointing weaknesses in existing water barriers is just a first step toward understanding the Netherlands’ flood risk. Rising sea level is, of course, the elephant in the room. But for the moment, the elephant is moving slowly enough to rank lower on the list of Dutch concerns than certain near-term threats. One is that the Rhine could burst its banks in areas such as Rotterdam. And as peat decomposes, land is sinking faster than the sea is rising. (To make matters worse, peat decomposition, triggered by centuries of Dutch land drainage, throws off greenhouse gases.) Finally, new roads and developments could increase runoff, and population growth could put more people in the path of disaster.

At Wageningen University’s Alterra research institute, 20 earth scientists and climate scientists are trying, among other things, to develop an accurate way of forecasting the water level of the Rhine. The goal is to understand the entire river as a system, from its headwaters in the Swiss Alps, through Germany, and finally through Rotterdam to the North Sea–to figure out how much precipitation it receives, what hydrological processes shape it and its watershed, and how development will change these factors. The Alterra group is trying to integrate meteorological and hydrologi­cal models and use them to evaluate various scenarios of climate and land-use change. “All the different components of the model are available somewhere,” says Eddy Moors, a hydrometeorologist at Alterra. “There’s a Dutch model, a German model, a hydrological model, a meteorological model. It’s a matter of finding a way to combine those components, more than inventing something new.”

One of the factors the researchers are considering is that radical changes are expected in European land use. Dutch planners say that by 2050, European agricultural land totaling an area larger than Germany will give way to development or, in some cases, revert to forest. That, in turn, will affect the way floods propagate–by altering the ability of the land to absorb water, for example. “If we take these changes into account and look into land use, we can perhaps promote some land-use changes which will assist us,” says Moors. “We want to see the feedback of those changes.” As part of his study, Moors has even been investigating how trends in land use might change the local weather. By running meteorological models, he found that turning farmland into forest is likely to stimulate more local precipitation. Whether this is good or bad depends on whether it happens during a summer drought or a winter flood. Either way, analyzing such effects is important to understanding the larger system.

In addition to forecasting the effects of changing land use, the Alterra group is trying to predict the effects of intensified weather extremes. The Intergovernmental Panel on Climate Change, the U.N.-sponsored body whose work represents the global scientific consensus on the subject, recently predicted that warmer temperatures worldwide could make droughts harsher and precipitation more intense. And in winter, precipitation will tend to fall more as rain than snow in some areas, including the Swiss Alps, at the Rhine’s headwaters. As a result, wintertime river flooding–already a problem in the Netherlands–could get far worse. Buffer systems of some kind will probably be required. These might be impoundment areas upriver, perhaps even in Germany, or underground tanks beneath Dutch developments. But planners need to know where to put these buffer systems, and how to manage them so they’re empty when deluges are expected but full when droughts are near. This calls for better long-range forecasts. “To adapt, it is important to have forecasting systems,” says Moors. “To be able to do that, you have to couple meteorology and hydrology models.”

Developing these sharper predictive tools is a pursuit common to planners in The Hague, New York City, and California (see “Planning for a Climate-Changed World,” May/June). But in the Netherlands, the need is acute. “Our whole country is at stake,” says Piet Dircke, director of the water program at Arcadis, an engineering and consulting firm based in Maastricht that is participating in national planning efforts. “So we are moving from an engineering kind of approach to a systems approach. You never know which part is going to change, and which one will be relevant, until you look at the complete system.”

Floating Houses
Once the risk analysis is complete, it will help guide decisions on where and how to build. Development is already restricted in some stretches of river floodplains and may soon be in others. But with population continuing to rise, there’s great pressure to develop tracts of low-lying areas like those in the southwest polder. And besides, defying the sea is a point of patriotic pride. “It is our culture to cope with water,” says Chris Zevenbergen, director of business development at the construction company Dura Vermeer and a professor at the Unesco-IHE Institute for Water Education in Delft. “Retreat would give a very bad signal to the world. Suppose we are not building; this will have an enormous impact on the climate for foreign investment. And from a technological point of view, [construction] is feasible. We need to adapt to that kind of development.”

To demonstrate what is possible, Dura Vermeer has built a floating housing development in a hamlet called ­Maasbommel, in the rural province of Gelderland, near the center of the country. There, 46 amphibious houses are perched on the outer edge of a dike that holds back the River Maas, adjacent to a marina. Sixteen are floating at the river’s edge, in conjoined pairs, with sealed hollow basements providing buoyancy. Between each pair of houses are two vertical concrete piles (one pile is shown below); if water levels rise, the houses rise around the piles. Flexible water, sewer, and electrical connections are unaffected. Thirty similar houses sit on slightly higher ground, on concrete slabs a meter or so above the waterline. They, too, have piles and hollow basements that will let them float if necessary. All 46 houses can tolerate a four-meter rise in water levels. (Video of the floating houses.)Though the houses have been finished since 2006, the higher ones have not yet faced a flood high enough to test them. “Everybody wants to see it happen. Including the builder and the architect,” jokes Cees Westdijk, who owns one of the houses. His two-bedroom house offers beautiful water views; on the downside, an algae bloom last year made for a nasty, if temporary, smell.


The Netherlands has designated 15 areas near riverbeds as possible sites for amphibious developments, including variations on the Maasbommel prototype. For the southwest polder, researchers at Delft Hydraulics and Wageningen ­University have already produced the first risk maps, showing which areas within its 50 square kilometers are most vulnerable. National planners “look inside the dike ring and make zones for how the water comes–fast and deep, fast and undeep, slow,” says Bloemen of the National Spatial Planning Agency. Buildings could be customized accordingly; some might always float, others might rise and fall if needed, and still others might simply be built to survive inundation without sustaining major damage. “You [could] have building restrictions appropriate to the relative dangers and flooding probabilities within each subzone,” Bloemen says.

For the southwest polder, university and government researchers are considering what kinds of development might be suitable. One option is to raise water or ground levels in parts of the polder and build amphibious or floating structures as appropriate; as a side benefit, raising water levels would halt the decomposition of peat. Other areas of the polder would be left at lower grades and could absorb floods. A decision on what to do in the southwest polder is expected in the next two years. Visions for other parts of the country include floating towns, floatable roadways that could be used for evacuation, and tanks beneath buildings that could hold floodwater. All of this would be a big departure from the traditional Dutch development method: throw a couple of meters of sand on top of the ubiquitous peat, install pilings, and pour the concrete. Still, impressive though it sounds, erecting floating or floatable structures is the easy part. The hard part of adapting to climate change is the planning, which requires intensive forecasting, sophisticated modeling, and risk mitigation strategies.


High and dry concept: A developer’s vision of a floating town, complete with greenhouses, is still just a vision. The Dutch are doing risk analyses to decide where to build, where to forbid development, and where to change construction techniques. To experience historical and possible future flooding in the Netherlands, visit
Courtesy of Dura Vermeer

The Dutch approach is gaining adherents around the world. “They are taking a systems approach that includes smart development,” says Lewis E. Link, a former director of research and development at the U.S. Army Corps of Engineers and now a professor at the University of Maryland, who led a postmortem federal investigation of Louisiana’s levee, pumping, and drainage systems after Hurricane ­Katrina. That means not just restricting building in certain areas but being smart about how to build in others. “I think in the U.S., we have been far too prone to let people build in vulnerable areas that then have to be protected,” he says. “It is just a mad cycle. We have trapped ourselves into this, over and over and over again.” Part of the problem in the United States, Link notes, is that the federal government has little control over land use, and local governments are often unwilling to challenge developers in areas that may face higher threat levels. In the Netherlands, the federal government can take more control, says Balfoort. “Sometimes you must make a top-down decision for the benefit of the nation as a whole,” he observes. “You do not discuss Christmas with the turkey.”

While it’s not clear whether the United States’ federal government will try to start making top-down decisions about land use in threatened areas, at least Dutch-­American research cross-pollination is well under way. Link has just completed an assessment of the New Orleans area–similar to the one Mynett is performing at the Delft University of Technology–to gauge the current risks posed to the Gulf Coast by various storm scenarios. Mynett sits on a U.S. panel making a similar assessment of the Sacramento and San Joaquin River valleys in California.

All parties are watching to see how the Dutch fare with climate-resilient housing. Given the dangers faced by coastal areas and river deltas around the globe, the rest of the world may soon beat a path to the Netherlands, clamoring for technical expertise. But before anyone will come, the Dutch must build it.

David Talbot is Technology Review’s chief correspondent.

Keep Reading

Most Popular

This new data poisoning tool lets artists fight back against generative AI

The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models. 

The Biggest Questions: What is death?

New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.

Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist

An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.

How to fix the internet

If we want online discourse to improve, we need to move beyond the big platforms.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.