After epic floods in India, South Africa, Germany, New York and Canada killed hundreds in the past year, droughts are now parching landscapes and wilting crops across the western US, the Horn of Africa and Iraq. The responses have included calls for higher levees, bigger drains and longer aqueducts. But these concrete interventions aimed at controlling water are failing. Climate extremes are revealing a hard truth: our development choices – urban sprawl, industrial agriculture and even the concrete infrastructure designed to control water – are exacerbating our problems. Because sooner or later, water always wins. Water might seem malleable and cooperative, willing to flow where we direct it. But as human development expands and the climate changes, water is increasingly swamping cities or dropping to unreachable depths below farms, often making life precarious. Signs of water’s persistence abound. Supposedly vanquished waterways pop up in inconvenient places. Seasonal creeks emerging in basements are evidence that those houses encroach on buried streams, while homes built on wetlands are the first to flood. When our attempts at control fail, we are reminded that water has its own agenda. Water finds its chosen path through a landscape, moulding it and being directed by it in turn. To reduce the impact of today’s more frequent and severe droughts and floods, a new global cohort of “water detectives” – restoration ecologists, hydrogeologists, biologists, anthropologists, urban planners, landscape architects and engineers – are asking a critical question: what does water want? Figuring out what water wants – and accommodating those desires within our human landscapes – is now a crucial survival strategy. The water detectives start by uncovering what water did before generations of humans so radically transformed our landscapes and waterways. How did water interact with local rocks, soils, ecosystems and climates before we scrambled them? As those detectives make more discoveries, we can begin to understand why certain areas flood repeatedly, or how our tendency to speed water off the land deprives us of urgently needed local rainfall. Then we can start to think creatively about how to solve these problems, by making space for water within our existing habitats. The answers that the detectives are finding – in cities, fields, swamps, marshes, floodplains, mountains and forests – are that we should be conserving or repairing natural systems, or mimicking nature to restore some natural functions – and not building more concrete infrastructure. These reparative approaches go by various names: nature-based systems, green infrastructure, low-impact development, water-sensitive urban design. In China, the “sponge cities” initiative aims to make urban regions better able to absorb rainfall and release it when needed. Because these kinds of solutions are based on working with or simulating natural systems, they offer benefits beyond just reducing floods and droughts. They can help us address the dramatic declines of animal species. They can help us adapt to climate change, or even slow its progression. Protecting biodiversity and storing carbon dioxide are not just peripheral effects of solving water problems – they are integral to healthy natural systems. So what does water want? In its liquid state, with sufficient quantity or gravity, water can rush across the land in torrential rivers or tumble in awe-inspiring waterfalls. But it is also inclined to linger to a degree that might surprise many of us, because the infrastructure of the modern world has erased so many of water’s slow phases, instead confining it, containing it, or speeding it away. These slow stages are particularly vulnerable to human interference, because they tend to occur in flatter places – once floodplains and wetlands – that we blocked or drained so that we could settle. But when water slows and stalls on land, that’s when the magic happens, providing habitat and food for many forms of life above and below. The key to greater resilience, say the water detectives, is to find ways to let water be water, to reclaim space for it to interact with the land. Innovative water management projects aim to slow water on land in some approximation of natural patterns. For that reason, I’ve come to think of this movement as “Slow Water”. Like the Slow Food movement – founded in Italy in the late 20th century, in opposition to fast food and all its ills – Slow Water approaches are bespoke: they work with local landscapes, climates and cultures rather than try to control or change them. Slow Food aims to preserve local food cultures and to draw people’s attention to where their food comes from, and how its production affects people and the environment. Similarly, Slow Water draws attention to the ways in which speeding water off the land causes problems. Its goal is to restore natural slow phases in order to support local availability, flood control, carbon storage and diverse forms of life. Just as Slow Food is local – supporting local farmers and protecting a region’s rural land from industrial development while reducing food’s shipping miles and carbon footprint – ideally, Slow Water is, too. The engineered response to water scarcity has usually been to bring in more water from elsewhere. But desalinating and transporting water consumes a lot of energy: in California, for example, the giant pumps that push water south from the Sacramento Delta are the state’s largest user of electricity. Withdrawing water from one basin and moving it to another can also deplete the donor ecosystem, or introduce invasive species to the receiver. Water engineering is also an environmental justice issue. Between 1971 and 2010, 20% of the global population gained water from human interventions on rivers, including dams, but 24% were left with less water, according to a 2017 study. Bringing in water from elsewhere can also harm the people receiving it. A big new reservoir imparts a false sense of security – when we live long distances from the source of our water, we don’t understand the limits of its supply, so we’re less likely to conserve it. We also don’t understand how the water we use supports its local ecosystem. By over-expanding human populations in places where there isn’t enough local water – such as in the US south-west, southern California and the Middle East – we make those places and those people more vulnerable to drops in supply. Water transfers also create a well documented cycle of scarcity, akin to the way that adding more lanes to a motorway just attracts more cars. Slow Water is in the spirit of many Indigenous traditions. Kelsey Leonard is a citizen of the Shinnecock nation, a Native American tribe whose traditional territory lies in modern-day New York state. She is also assistant professor in the school of environment, resources and sustainability at the University of Waterloo in Ontario. As she explained in an online talk in 2020, Indigenous traditions don’t consider water to be a “what” – a commodity or threat – but a “who”. Many Indigenous peoples around the world believe not only that water is alive, but that it’s kin. “That type of orientation transforms the way we make decisions about how we might protect water,” she said. The water detectives are a diverse bunch, with varying beliefs. But they share an openness to moving from a mindset of control to one of respect. As our long-held illusion that we can control water is crumbling in the face of escalating disasters, we are beginning to understand that it’s better to learn how to accommodate water, and to enjoy the benefits that cooperation can bring. Nowhere has urbanisation happened more rapidly than in China, where a mass exodus from the countryside over the last 40 years has seen the number of urban dwellers boom, from around 20% of the population in 1980 to almost 64% in 2020. To house and employ all these people, cities sprawled and new ones were built from scratch. Builders paved floodplains and farmland, felled forests and channelised rivers, leaving stormwater that once filtered into the ground with nowhere to go but up and over levees. Then, one notable flood struck the national government where it lives. On 21 July 2012, Beijing was hit by its largest storm in 60 years. As much as 46cm of rain fell on some parts of the city, filling underpasses and flooding roads a metre deep. Landscape architect Yu Kongjian barely made it home from work. “I was lucky,” he says. “I saw many people abandon their cars.” As the deluge descended, the city was plunged into turmoil. Seventy-nine people died, many of them drowned in their vehicles, electrocuted or crushed by collapsed buildings. The damage stretched across 14,000 sq km, costing nearly $2bn. Yu, co-founder of the acclaimed landscape architecture firm Turenscape and a leading figure in the Slow Water movement, was frustrated. He had warned the government years earlier that disaster was coming. He had led a research team in mapping what he called the city’s “ecological security pattern”, showing the government which parcels of land were at high risk of flooding, and urging it to block development and instead to use them to absorb stormwater. They ignored his recommendations. “The 2012 flood taught us the lesson that the ecological security pattern is a life-and-death issue,” Yu told me when I met him in Beijing in 2018. Urban sprawl is also exacerbating water scarcity in China, especially in the north and west. In some of China’s densest cities, because of rain running off buildings, streets and car parks, only around 20% of precipitation soaks into the soil. Instead, as in so many other cities around the world, drains and pipes funnel it away – lunacy, Yu thinks, in a place with water shortages. In common with other cities in China’s north, Beijing is pretty dry outside the summer monsoon season. For decades, the city has pumped groundwater to supply its growing population and rising rates of consumption. This is lowering the water table by about a metre each year, causing the ground to sink as well. This phenomenon is also happening elsewhere, such as in Mexico City and California’s San Joaquin Valley. But now Yu is leading the way as China re-engineers old cities and designs new ones to accept rather than fight natural water flows. His landscape architecture projects incorporate Slow Water principles in order to lessen floods, save water for dry spells and reduce water pollution. The 2012 Beijing disaster was a turning point. A month later, a Turenscape stormwater project in Harbin, a city about 800 miles north-east of Beijing, won a top US design prize. Chinese state television broadcast a high-profile interview with Yu. He said a government minister told him afterwards that President Xi Jinping had seen it. Less than a year later, Xi stood in front of China’s national urbanisation conference and announced his sponge city initiative, boosting the idea from fringe concept to national mission. It is part of Xi’s Ecological Civilisation agenda, which aims to clean up the pollution, flooding hazards and associated costs caused by his predecessors’ industrial civilisation. With its centralised government, China built its industry and its economy at a blistering pace. Similarly, now it is pursuing sponge cities on a scale difficult for most countries to even consider. Globally, urban flooding has become particularly acute as the land area covered by cities worldwide has doubled since 1992. Researchers from Johns Hopkins University calculated how impervious surfaces increase flooding: every time a city increases coverage of absorbent soil with roads, pavements or car parks by 1%, runoff boosts the annual flood magnitude in nearby waterways by 3.3%. To counteract this trend, sponge cities seek places throughout urban areas for water to sink into the ground. The system works best when these features are linked together so the water can travel along some approximation of its natural path. Cities can convert old industrial areas beside rivers into parks, and cut through paving to make way for run-off channels lined with water-loving plants, infiltration ponds and seepage wells. The idea is to mimic nature as much as possible. Where human space is non-negotiable, designers sometimes use surrogates, such as permeable paving and green roofs that can absorb water. In 2015, the government began demonstration projects in 16 cities, adding 14 more in 2016. Each project covered at least 13 sq km, although some were much larger. Objectives included reducing urban flooding, retaining water for future use, cleaning up pollution and improving natural ecosystems. The goal was, by 2020, for each project to retain 70% of the average annual rainfall on site, both to help prevent flooding and to store water underground for the dry season. The Chinese government said it has met these targets. Still, even though its scale is more ambitious than related projects elsewhere, it is likely insufficient. During heavy rains in 2021, one pilot city, Zhengzhou, still suffered significant flooding and deaths. Absorbing rainfall across 13 sq km of a city that spans thousands wasn’t enough to avert disaster. Yu and other urban water detectives are looking to manage water to a grander extent, seeking out connected routes for water to slow and flow across entire watersheds, which often extend beyond jurisdictional boundaries. Solving a city’s flooding problems requires coordination with communities and landowners upstream. Ideally urban designers could absorb water where it falls, reducing stormwater runoff at every rooftop and at every farm field upstream. Yu is dreaming beyond sponge cities to sponge land. “This is a philosophy for taking care of the continental landscape,” he tells me. “It’s time to expand the scale.” When planning a project, Yu and other urban designers start by trying to figure out what water did before a city spread, and what it does now within its current confines. Like many of the water detectives I met, the staff at Turenscape use spatial mapping software from Environmental Systems Research Institute (ESRI), which can map watersheds from mountains to ocean, modelling floods, plant succession, infrastructure and much more. The tool allows designers to comprehend complex systems and interrelated challenges, such as how to reduce flooding while also preserving other species, building smarter cities and reducing resource waste. The first thing planners plot is topography, or the highs and lows of the landscape – a primary factor in how water flows. Models also include soil type, which can dramatically affect how water drains; and vegetation, because that affects how much water soaks in, runs off or evaporates from plants into the air. Plus, soil acidity can affect which plants will thrive or die in a restored area. Turenscape also models historic and ecological data, as well as information on the local population, economy and transportation. The data comes from various sources. Hydrology records can help to predict rainfall and flooding more accurately. Topography data can be gathered by aeroplanes with lidar sensors, which use lasers to survey under buildings. City maps can show transportation corridors, parks, domestic yards and industrial buildings with giant roofs. Getting good soil data in urban areas can be tricky because builders often move soil from one place to another. To know for sure what’s down there, engineering firms typically drill a hole and take a core sample. With this information, Slow Water practitioners can better understand how a particular variable affects the way water behaves. When their landscape maps are complete, they send test floods through the digital model they’ve created. These experiments allow them to identify pinch points where water is constrained and will flood first. Then they experiment with a topography adjustment or the addition of a wetland or pond to see how each affects stormwater behaviour. Yu told me he traces his passion to repair humans’ relationship with water back to the agricultural commune where he grew up, in Zhejiang province, south-west of Shanghai. There he observed the Chinese “peasant wisdom” for managing water, practised for thousands of years. Farmers maintained little ponds and berms to help rainfall soak into the ground, storing it for a dry day. The seasonal creek next to his village swelled and retreated with the seasons. “For me, flood is a time of excitement because the fish come to the field, the fish come to the pond.” He saw that flooding need not be the enemy. “If you have wise ways to deal with flood, water can also be friendly.” A week after meeting Yu, I visited one of Turenscape’s projects in progress, Yongxing River Park, located in Daxing, a far-flung exurb of Beijing. Satellite images from three years earlier showed open land surrounding the river, which was straightened and confined by steep concrete walls. Today those images are chock-a-block with buildings around a more generous, meandering path for water. The project was nearly complete when I saw it in April 2018. About 4km long and perhaps two city blocks wide, the park follows the river. Workers removed concrete along the river channel and excavated soil to widen the riverbed. That dirt was then moulded into a large berm running down the centre, creating two channels. The river flows on one side, while the other channel has large holes of varying depths that act as filtration pools and direct the water flow. During the dry season, the filtration side is filled with partially cleaned effluent from a sewage treatment plant. Wetland plants in the pools slow the water, further cleaning it and allowing some of it to filter into aquifers. During monsoon season, that channel is reserved for floodwaters, and the effluent is treated industrially. The broader riverbanks, newly freed from concrete, are dotted with thousands of small plants in closely set rows to hold the earth. As we walk the path between the two channels, we pass young willow trees scaffolded together with sticks for strength while they grow. Willows, a native streamside plant beloved by beavers, have roots that reach for the air, like cypress and mangroves, allowing them to survive extensive periods of flooding. Elsewhere, reeds, small bushy willows, dwarf lilyturf and other native plants stabilise the soil. Existing large trees, including elms and poplars, were retained. During big rains in 2020, Yu sent me photos of Yongxing River Park. The trees and grasses had grown considerably since I’d seen it two years earlier, turning it into a lush, green oasis. The channel contained a good amount of water, but was nowhere close to over-topping. All Slow Water projects must factor in local climate, soil and hydrogeology. Consider two Chinese cities with diametrically opposed water needs. Kunshan, in Jiangsu province near Shanghai, is built on polders – land reclaimed from water with levees. The water table is so high that surface water does not soak away, but filtration – cleaning the water – is necessary. Hotan, a desert city in far western Xinjiang province, gets less than 4cm of rainfall a year on average, so it needs to protect its groundwater supply. If China ignores this specificity, its broad ambition for sponge cities may falter, says Chris Zevenbergen, an expert in urban flood-risk management at the IHE Delft Institute for Water Education in the Netherlands and a visiting professor at China’s Southeast University. The rush to develop cities in the past 20 years did not allow builders time to understand imperfections in design and make changes. That is what led to cities across China experiencing the same problem at the same time: widespread urban flooding. Rushed implementation of sponge cities initiatives could also lead to missteps. Xi’s programme has strict deadlines, which may not allow time to monitor performance, adjust if necessary and transfer knowledge. It “takes time to learn and to reflect”, Zevenbergen warns. A paper written by Chinese government research institutes in 2017 expressed similar concerns about a cookie-cutter approach. Although Zevenbergen expects the Chinese will make a lot of mistakes along the way, he thinks that “in the end, they will become the leaders in sponge cities. The same happened in the realm of renewable energy.” The country has a culture of getting things done. “Every year I’m in China, I make designs with students, and the next year the projects have been implemented. And that is really astonishing.” The flipside of getting things done quickly, however, is a culture with less devotion to maintenance and aftercare. And green infrastructure requires maintenance, such as pruning or replacing plants. A Chinese–European peer-learning exchange programme with which Zevenbergen is involved is helping to accelerate the pace of learning. China needs to learn fast. Grey infrastructure – so called because it is usually built with concrete – struggled during recent heavy summer monsoons, which pushed several giant dams to the brink of failure and killed more than 200 people. Meanwhile, the Yangtze area alone is choked with so many dams that 333 rivers have dried up to varying degrees. Zevenbergen calls the massive dams examples of “stupid infrastructure”. Giant grey infrastructure projects like these are unlikely to have long lives in the age of climate change, because they can take a decade to build and are engineered to accommodate certain maximum flows. To do that effectively “means that we need to know how much climate change we can expect”, Zevenbergen told me. “The problem is, we do not know.” Yet China is still dam-oriented. Despite the national promotion of sponge cities, Yu noted that China’s schools continue to train engineers using 20th-century principles. And in the offices where decisions are made, Yu still confronts a penchant for stronger dams, bigger pipes and larger stormwater storage tanks – a refrain I heard repeatedly from water detectives around the world. Shifting the dominant culture will mean adopting a new ethos of water and land management. Yu said: “We are fighting so hard to try to get people to think in an ecological way.” This is an edited extract from Water Always Wins: Thriving in an Age of Drought and Deluge by Erica Gies, published by Head of Zeus and available at guardianbookshop.com Follow the Long Read on Twitter at @gdnlongread, listen to our podcasts here and sign up to the long read weekly email here.