Markus Andrews

To develop flood risk management strategies, governments need to consider what really matters, namely how and over what period floods affect societal welfare. To do so, we advocate the adoption of a resilience lens in flood risk management. Here, resilience is understood as the ability of a society to cope with flood hazards by resisting, absorbing, accommodating, adapting to, transforming and recovering from the effects of floods on people’s welfare3,4. To analyze and enhance resilience, we need to consider how and over what period floods affect societies and how measures could affect flood impacts and society5. Questions to consider include whether floods will hamper economic activities; whether people can earn sufficient income or their livelihoods are destroyed and whether their health will be affected.Adopting a resilience lens means taking societal welfare as our starting point. From there, the interaction with flood hazards and flood risks can be considered6. For frequent events resistance may be required to allow societies to continue functioning without facing frequent damage. Damage as a result of rare and extreme events may not be avoidable, but such events must be included in our considerations in order to make sure that those events, although damaging, do not turn into disasters. This requires a deep understanding of what makes people vulnerable to floods and how resilience can be improved. We offer four elements linked to this resilience lens to understand what makes a flood disastrous. We aim to enable an informed discussion on how to arrive at appropriate flood risk management strategies (see Fig. 1).Fig. 1: Adopting a resilience lens by operationalizing the four elements into an integrated flood risk management approach.A welfare and recovery capacity (element 1 and 2): Different effects of floods on different areas or societal groups: some have a larger deterioration of welfare or a slower recovery than others. Both the maximum impact and the recovery together determine the impact of a flood disaster. B include beyond-design events (element 3). The grey curve shows the impacts as a function of event extremity. The standard assessment integrates over this curve and uses the resulting expected annual damage as risk measure; this aggregation undermines the role of high-impact but low-probability events. The extreme events must be given attention as well; (C) distributional impacts (element 4). Distributional impacts can be considered spatially or for different social groups. Welfare economics principles can be applied to capture the utility of different communities and vulnerable groups. By aggregating the effects, we may not see how some groups benefit from measures while others pay for them, or still face large risks. Therefore, next to total cost and benefits, also distributed impacts must be used and weighted to enhance equity.Full size imageImpacts on welfare, instead of on asset lossesFloods hit socially vulnerable people harder, because poorer communities often lack the capacity to recover quickly. Vulnerable people or communities have a lower capacity to anticipate, cope with, resist or recover from the impact of hazards7. They may be forced to live in hazardous places, have less access to flood warnings, a less effective network to enhance recovery, and fewer resources to protect their homes or livelihoods. Especially people that already live in poverty may need to shift to destructive strategies such as selling land or cattle or consume seeds to meet other short-term needs. Such strategies can lead to a vicious circle.Using absolute asset-based damages as yardsticks, as is often done in flood risk management, largely underestimates the disproportionally large welfare impact relatively small absolute losses can have on poor people and may lead to biased planning8. As one dollar does not count equally for all people, flood risk planning should move beyond asset-based valuations and put the welfare of people at the core of the assessment9. This can be done, for example, by considering social impacts such as loss of houses (irrespective of their value), deprivation cost, loss of percentage of income, or considering the effect on income generating ability.There are further merits to placing welfare upfront. First, it opens the possibility of better aligning flood risk management with the larger development agenda3, for instance by linking flood risk management to spatial and economic planning. Second, it allows for a better inclusion of non-structural measures in flood risk management strategies, such as adaptive social protection systems that can quickly disburse financial assistance to households when a disaster hits10. Such measures may not reduce asset-based damages but can have significant benefits of increasing recovery rate and dampening welfare losses.Recovery capacityWhen recovery from floods takes longer, the impact of the floods is more disastrous because of the many indirect and cascading effects, which often exceed the direct damage11. Differences in flood impacts across societal groups often link to differences in their ability to recover from flood impacts. To recover, physical damage must be repaired and income generating options must be restored. Accounting for disruption of services of critical infrastructure, cascading impacts12 or addressing people’s recovery capacity are thus crucial to understand the impact of floods on societal welfare. If we consider recovery as part of flood risk management, the effect of recovery enhancing measures can be included to reduce longer-term welfare loss. Measures such as citizen training, micro-credits, affordable insurance to compensate for flood losses and improving critical infrastructure (enhancing its robustness, redundancy, or flexibility) then become relevant.Beyond-design eventsThe July 2021 floods in Europe have shown the devastating impact of beyond-design events, events that exceed the known risks. The flood peak discharge in July 2021 in the Ahr valley was roughly five times higher than the extreme event scenario of the official flood map13 and its return period was estimated to be around 500 years. Such an event was beyond the imagination of people and authorities, which led to high numbers of fatalities and massive destruction.The complexity of flood risk systems, limitations of scientific knowledge but also motivational and cognitive biases in perception and decision making contribute to such surprises14,15. In many regions, climate change and other drivers of change, such as population growth or increasing vulnerability, lead to more frequent situations where current protection systems are overwhelmed. Our third element targets this blind spot of flood risk management: extreme events beyond current design standards to prevent disastrous surprises.This can be done for example by using a storyline approach, narrative scenarios or training exercises and simulation games that stimulate decision-makers to think through the full disaster cycle. Such exercises are known to inspire discussion of potentially long-term unexpected or unintended cascading effects across different systems16. Outliers in ensemble forecasts may be used as a starting point for such scenarios. These explorations guide dialogues towards achieving the desired level of protection and preparedness for extreme events, to reduce the impact to the most crucial objects, locations, or groups of a society, and provide the basis for training of decision-makers.Distributional impacts and equityA resilience lens requires asking the distributional questions of “the five Ws“17: for whom, when, what, where, and why? Most flood risk analyses aggregate risks and flood protection benefits and disregard their distribution across people, space and time. The resilience lens requires unpacking this aggregation by assessing the distributional impacts of alternative measures. Making explicit who wins and who loses can support distributive justice and prevent unintended distributional consequences. Additional measures for compensating worse-off groups can also be prepared. It is one option, for example, to target flood risk protection measures18 at the most socially vulnerable instead of selecting measures based on utilitarian principles. To do so, a risk analysis that shows distributed impacts on a range of social groups and regions must be carried out. These distributional questions also play out between current and future generations (intergenerational justice).The distributional performance of alternative plans can be assessed through a normative analysis. Various ethical principles drawn from theories of distributive justice can be operationalized to evaluate the fairness of alternative measures19. Multiple principles can also be combined. In the Netherlands, the flood protection standard is designed such that every person has at least a minimum level of safety (sufficientarian principle), while additional safety margin is allowed if it is economically sensible (utilitarian principle)20. More

Peru’s water utility companies are protecting peat bogs because of their ability to hold water.Credit: Erica Gies

In just a few months this year, abnormally low water levels in rivers led China to shut down factories and to floods in one-third of Pakistan, killing around 1,500 people and grinding the country to a halt. A dried-up Rhine River threatened to tip Germany’s economy into recession, because cargo ships could not carry standard loads. And the Las Vegas strip turned into a river and flooded casinos, chasing customers away. It seems that such water disasters pepper the news daily now.Many businesses have long lobbied against changing their practices to safeguard the environment, by refusing to implement pollution controls, take climate action or reduce resource use. The costs are too high and would harm economic growth, they argue. Now we are seeing the price of that inaction.With mounting climate-fuelled weather disasters, social inequality, species extinctions and resource scarcity, some corporations have adopted sustainability programmes. One term in this realm is ‘circular economy’, in which practitioners aim to increase the efficiency and reuse of resources, including water — ideally making more goods (and more money) in the process.
Part of Nature Outlook: The circular economy
But the term has its roots in decades of alternative economic theories — known variously as environmental economics, ecological economics, doughnut economics and steady-state economics. These frameworks recognize that the mainstream economics’ goal of eternal growth is impossible on a planet with finite resources.These ideas are beginning to filter into the mainstream, a mark of both the persuasiveness of advocates’ arguments and the declining state of the natural world. But the economists and scientists behind these principles say that some businesses and governments are engaging in greenwashing — claiming their actions to protect the environment are more significant than they really are — rather than making the kinds of fundamental change required to move the global economy onto a truly sustainable path.Because the dominant culture prioritizes human demands, water is generally viewed as either a commodity or a threat. That perspective inspires single-focus problem solving that ignores the complexity and interconnectedness of water’s relationships with rocks and soil, microbes, plants and animals, including humans, inevitably resulting in unintended consequences.Pumping out groundwater when rivers run low further depletes surface water because the two are linked. Erecting dams to provide water to one group of people deprives other people and ecosystems. Leveeing up rivers and building on wetlands removes space for water to slow, pushing flooding onto neighbouring areas. Paving cities and whisking water away creates localized scarcity.Some corporations are making ‘water neutrality’ or ‘water positive’ pledges, which are a big step forward but not enough, says Michael Kiparsky, director of the Wheeler Water Institute at the University of California, Berkeley’s Center for Law, Energy and the Environment. “If corporations are really serious about water stewardship, they would throw their political and financial heft behind reform of the governance systems that set up this extractive economy around water,” Kiparsky says.More than 11,000 scientists from 153 countries agree that tweaks around the margins are insufficient. In a 2019 letter in the journal BioScience they called for “bold and drastic transformations”, including a “shift from GDP growth and the pursuit of affluence toward sustaining ecosystems and improving human well-being”1. In February, the Intergovernmental Panel on Climate Change, agreed, calling for integrating “natural, social and economic sciences more strongly,” in part by conserving 30–50% of Earth’s ecosystems (see go.nature.com/3sccm6h).A growing group of ecologists, hydrologists, landscape architects, urban planners and environmental engineers — essentially water detectives — are pursuing transformational change, starting from a place of respect for water’s agency and systems. Instead of asking only, ‘What do we want?’ They are also asking, ‘What does water want?’. When filled-in wetlands flood during events such as the torrential 2017 rains in Houston, Texas, researchers realized that, sooner or later, water always wins. Rather than trying to control every molecule, they are instead making space for water along its path, to reduce damage to people’s lives.Broadly speaking, the detectives are discovering that water wants the return of its slow phases — wetlands, floodplains, grasslands, forests and meadows — that human development has eradicated. People have destroyed 87% of the world’s wetlands since 17002, dammed almost two-thirds of the world’s largest rivers3, and doubled the area covered by cities since 19924. All these have drastically altered the water cycle. The water detectives’ projects — part of a global ‘slow water’ movement — all restore space for water to slow on land so it can move underground and repair the crucial surface–groundwater connection.Although the uses of slow-water approaches are unique to each place, they all reflect a willingness to work with local landscapes, climates and cultures rather than try to control or change them. Slow water is distributed throughout the landscape, not centralized. For instance, wetlands and floodplains are scattered across a watershed — an area of land drained by a river and its tributaries — in contrast to a dam and giant reservoir. Around the globe, water detectives are beginning to scale up these projects.Slow waterFor most of California’s state history, groundwater and surface water have been treated as separate resources from both a legal and regulatory perspective. But physically they are linked — by gravity and hydraulic pressure. When river levels run high and spill over into wetlands and floodplains, the flow slows down and seeps underground, raising the water table. Later, that groundwater feeds wetlands, springs and streams from below. “It is hydrologically ridiculous to treat groundwater and surface water differently,” says Kiparsky. “That is as non-circular as you can get.”That legal separation has resulted in overtaxing California’s water supply. The state’s massive water infrastructure — huge dams, levees and long-distance aqueducts — prevents the great rivers of the Central Valley region from occupying their floodplains and naturally recharging groundwater. Plus, when surface water is scarce, people aggressively pump groundwater. But because the two are connected, that further decreases surface water. This depletion means that people have to drill deeper, more expensive wells to reach water. It can also collapse the land, destroying infrastructure. And pumping groundwater near the ocean can allow seawater to push salt inland.Since passage of the 2014 Sustainable Groundwater Management Act (SGMA), California has prioritized recharging groundwater by spreading excess winter water and floodwater on land so it filters underground, or injecting it underground through wells. Various state programmes include incentives for farmers to percolate water on fallow fields, flood management that sets back levees, allowing floodplains to once again serve their purpose, and a search for palaeo valleys — special geological features that could rapidly move heavy water flows underground.But key hurdles remain to seize the bounty of winter floods, says Kiparsky. The main problem is that, despite the SGMA, legal legacies of the artificial divide between surface water and groundwater linger. Colorado is managing this better, he says, because it has integrated the rights systems for groundwater and surface water. Connecting them legally facilitates multipurpose projects such as routing winter water to recharge ponds, which provides habitats for birds and human recreation. The water infiltrates the ground and rejoins the river, effectively making that same water available to farmers later in the year.Peru is also focused on the connection between surface water and groundwater. Almost two-thirds of its population live on a desert coastal plain that receives less than 2.5 centimetres of rain per year and relies on water from the Andes, including from melting glaciers. In 2019, the World Bank predicted that drought-management systems in Lima — dams, reservoirs and under-city storage — would be inadequate by 20305. Over the past decade, Peru has passed a series of laws that recognize nature as part of water infrastructure and require water utilities to invest a percentage of user fees in wetlands, grasslands and groundwater systems.One type of investment is the protection of rare high-altitude wetlands called bofedales, or cushion bogs, which slow water runoff that might otherwise cause flooding or landslides, and hold onto wet-season water, releasing it in the dry season. Bofedales are peatlands, which cover just 3% of global land area but store 10% of freshwater and 30% of land-based carbon6. Unfortunately, these bogs have been subject to peat thievery for the nursery trade. Utility investments are introducing surveillance to protect bofedales and restoring damaged wetlands. Scientists have also studied a local practice of carving out more space for water in the landscape to expand the bofedales, and found that these expansions can store similar quantities of water as the original bogs7.Peru’s water utilities are also investing in a practice innovated by the Wari people 1,400 years ago. In a few Andean villages, Wari descendants still build hand-cobbled canals called amunas. The amunas route wet-season flows from mountain creeks to natural infiltration basins, where the water sinks underground and moves downslope much more slowly than it would on the surface. It emerges weeks to months later from lower-altitude springs, where farmers tap it to irrigate crops.“If we plant the water, we can harvest the water,” says Lucila Castillo Flores, a communal farmer in the Andes village of Huamantanga above the Chillón River valley in Peru. Their culture of reciprocity, with the landscape and with each other, governs how communal farmers care for the water and share the bounty. Because much of the water they use for irrigation seeps back underground, it eventually returns to rivers that supply Lima. Hydrological engineer Boris Ochoa-Tocachi, chief executive of the Ecuador-based environmental consultancy firm ATUK, and his co-researchers used dye tracers, weirs and surveys of traditional knowledge to calculate the impact of restoring amunas throughout the highlands. Lima already has 5% less water than its consumers need. The researchers showed that restoring amunas throughout the largest watershed that supplies Lima could make up that water deficit and give the capital an extra 5%, extending availability into the dry season by an average of 45 days8.Working with wildlifeTaking a holistic approach is also paying off in Washington state and in the United Kingdom, where people are allowing beavers space for their water needs. The rodents in turn protect people from droughts, wildfires and floods. Before people killed the majority of beavers, North America and Europe were much boggier, thanks to beaver dams that slowed water on the land, which gave the animals a wider area to travel, safe from land predators. Before the arrival of the Europeans, 10% of North America was covered in beaver-created, ecologically diverse wetlands.Environmental scientist Benjamin Dittbrenner, at Northeastern University in Boston, Massachusetts, studied the work of beavers that were relocated from human-settled areas into wilder locations in Washington state. In the first year after relocation, beaver ponds created an average of 75 times more surface and groundwater storage per 100 metres of stream than did the control site9. As snowfall decreases with climate change, such beaver-enabled water storage will become more important. Dittbrenner found that the beaver’s work would increase summer water availability by 5% in historically snowy basins. That’s about 15 million cubic metres in just one basin, he estimates — almost one-quarter of the capacity of the Tolt Reservoir that serves Seattle, Washington.

Beavers help to protect people from floods.Credit: Troy Harrison/Getty Images

Beavers have fire-fighting skills too, says Emily Fairfax, an ecohydrologist at California State University Channel Islands in Camarillo. When beavers are allowed to repopulate stretches of stream, the widened wet zone can create an important fire break. Their ponds raise the water table beyond the stream itself, making plants less flammable because they have increased access to water.And beavers can actually help to prevent flooding. Their dams slow water, so it trickles out over an extended period of time, reducing peak flows that have been increasingly inundating streamside towns in England. Researchers from the University of Exeter, UK, found that during storms, peak flows were on average 30% lower in water leaving beaver dams than in sites without beaver dams10. These benefits held even in saturated, midwinter conditions.Beaver ponds also help to scrub pollutants from the water and create habitats for other animals. The value for these services is around US$69,000 per square kilometre annually, says Fairfax. “If you let them just go bananas”, a beaver couple and their kits can engineer a mile of stream in a year, she says. Because beavers typically live 10 to 12 years, the value of a lifetime of work for two beavers would be $1.7 million, she says. And if we returned to having 100 million to 400 million beavers in North America, she adds, “then the numbers really start blowing up”.System changeFor the most part, mainstream economics doesn’t take into account the many crucial services provided by healthy, intact ecosystems: water generation, pollution mitigation, food production, crop pollination, flood protection and more.Value calculations such as Fairfax’s are increasingly tabulated by scientists but usually ignored by the market. One early effort to put a monetary value on those services was a landmark report11 in Nature in 1997, co-authored by Robert Costanza, an ecological economist at the Institute for Global Prosperity at University College London. At the time, global ecosystem services were worth tens of trillions of dollars, more than global gross domestic product (GDP). In an updated paper published in 2014, the global economy had grown but ecosystem services were still worth considerably more12.Another problem: the degradation of those services is typically not counted against profits; instead, those costs are paid by the environment and people. Hannah Druckenmiller, an environmental economist and data scientist at the non-profit organization Resources for the Future in Washington DC, has calculated that permitting development on one hectare of wetlands incurs property damages of more than $12,000 per year13. That’s because water that has been displaced from an area that used to absorb it floods surrounding communities. Druckenmiller estimates the value of wetlands nationwide, just for flood absorption, to be $1.2 trillion to 2.9 trillion. And that is a conservative estimate, based on flood damage data covering just around 30% of households in floodplains.The overarching problem is that the main measure of economic health, GDP, has a narrow focus on market-based production and consumption and does not accurately measure human well-being, Costanza asserts. “A circular economy that similarly limits itself to production will also fall short,” he says. If the goal is well-being, “the question becomes: should you be producing and consuming all those things in the first place?”. Protecting and restoring natural resources and rebuilding social capital, he says, are more likely to achieve well-being.
More from Nature Outlooks
One way to do that is to put more natural ecosystems into a common asset trust, or ‘the commons’. Creating state or local parks, hunting reserves, or wildlife refuges can restrict development and provide significant benefits to the community, says Druckenmiller. Communities that invest in protecting a wetland to prevent flood damages will see the benefit of avoided costs quickly, she says, often with a payback period of less than five years.Another strategy to protect the commons, says Costanza, is the ‘rights of nature movement’, which began in the early 1970s and has gained ground over the past 15 years. It includes enshrinements in the constitutions of Bolivia and Ecuador, local government changes across the United States, and personhood for the Whanganui River in New Zealand, the Ganges River in India and the Magpie River in Canada. That might sound unusual to some people, but in the United States, some corporations have personhood. Granting personhood to a river enables people to argue in court on behalf of its rights. A river’s rights can include freedom from pollution, protection of its cycles and evolution, and space to fulfil its ecosystem functions. The rights of nature movement recognizes that healthy ecosystems make everything work, and “people are part of that system and not separate from it”, says Costanza.States reforming century-old water rights, utilities investing in wetlands and Indigenous techniques and scientists deploying beavers for their engineering prowess are definitive shifts from business as usual. “We’ve made a lot of progress integrating [natural capital] into the system, where it doesn’t get pushed aside because other things are higher priority,” says Druckenmiller.But Costanza thinks much deeper change is needed. “A lot of the things that we’re talking about with the circular economy — regenerating wetlands, planting forests, dealing with climate change — are difficult to implement because the underlying goal is still GDP growth, and these things get in the way of that,” he says.People applying slow-water approaches are doing what they can in the dominant economy. But Costanza says that people can better protect social capital and environmental systems by switching from GDP to metrics such as the Genuine Progress Indicator or one of “literally hundreds” of alternatives, he says.Society’s fundamental goals might seem like a high bar to set, but some of these metrics have already been adopted by governments in Maryland, Vermont, Bhutan and New Zealand. Such shifts move beyond greenwashed versions of a circular economy and help to facilitate water detectives’ work in caring for water systems so that they can sustain human and other life. More

Körner, C., Paulsen, J. & Spehn, E. M. A definition of mountains and their bioclimatic belts for global comparisons of biodiversity data. Alp. Bot. 121, 73–78 (2011).Article 

Google Scholar 
Humboldt, A. v. & Bonpland, A. Ideen zu einer Geographie der Pflanzen nebst einem Naturgemälde der Tropenländer (Cotta, 1807).Barry, R. G. Mountain Weather and Climate (Cambridge Univ. Press, 1992).Paulsen, J. & Körner, C. A climate-based model to predict potential treeline position around the globe. Alp. Bot. 124, 1–12 (2014).Article 

Google Scholar 
Huss, M. et al. Toward mountains without permanent snow and ice. Earth’s Future 5, 418–435 (2017).Article 

Google Scholar 
Smith, R. B. 100 years of progress on mountain meteorology research. Meteorol. Monogr. 59, 20.1–20.73 (2019).Article 

Google Scholar 
Immerzeel, W. W. et al. Importance and vulnerability of the world’s water towers. Nature 577, 364–369 (2020).Article 
CAS 

Google Scholar 
Bradley, R. S., Keimig, F. T. & Diaz, H. F. Projected temperature changes along the American Cordillera and the planned GCOS network. Geophys. Res. Lett. https://doi.org/10.1029/2004GL020229 (2004).Zappa, G., Ceppi, P. & Shepherd, T. G. Time-evolving sea-surface warming patterns modulate the climate change response of subtropical precipitation over land. Proc. Natl Acad. Sci. USA 117, 4539–4545 (2020).Article 
CAS 

Google Scholar 
Payne, A. E. et al. Responses and impacts of atmospheric rivers to climate change. Nat. Rev. Earth Environ. 1, 143–157 (2020).Article 

Google Scholar 
Mooney, H., Dunn, E., Shropshire, F. & Song, L. Vegetation comparisons between the Mediterranean climatic areas of California and Chile. Flora 159, 480–496 (1970).Article 

Google Scholar 
di Castri, F. in Mediterranean Type Ecosystems (eds di Castri, F. & Mooney, H. A.) 21–36 (Springer, 1973).Cody, M. L. & Mooney, H. A. Convergence versus nonconvergence in Mediterranean-climate ecosystems. Annu. Rev. Ecol. Syst. 9, 265–321 (1978).Article 

Google Scholar 
Morales, M. S. et al. Six hundred years of South American tree rings reveal an increase in severe hydroclimatic events since mid-20th century. Proc. Natl Acad. Sci. USA 117, 16816–16823 (2020).Article 
CAS 

Google Scholar 
Viviroli, D., Kummu, M., Meybeck, M., Kallio, M. & Wada, Y. Increasing dependence of lowland populations on mountain water resources. Nat. Sustain. https://doi.org/10.1038/s41893-020-0559-9 (2020).Siirila-Woodburn, E. et al. A low-to-no snow future and its impacts on water resources in the western United States. Nat. Rev. Earth Environ. 2, 800–819 (2021).Article 

Google Scholar 
Pepin, N. et al. Elevation-dependent warming in mountain regions of the world. Nat. Clim. Change 5, 424–430 (2015).Article 

Google Scholar 
Sturm, M., Goldstein, M. A. & Parr, C. Water and life from snow: a trillion dollar science question. Water Resour. Res. 53, 3534–3544 (2017).Article 

Google Scholar 
Saavedra, F. A., Kampf, S. K., Fassnacht, S. R. & Sibold, J. S. Changes in Andes snow cover from MODIS data, 2000–2016. Cryosphere 12, 1027–1046 (2018).Article 

Google Scholar 
Garreaud, R. D. et al. The Central Chile Mega Drought (2010–2018): a climate dynamics perspective. Int. J. Climatol. 40, 421–439 (2020).Article 

Google Scholar 
Milly, P. C. D. & Dunne, K. A. Colorado River flow dwindles as warming-driven loss of reflective snow energizes evaporation. Science 367, 1252–1255 (2020).Article 
CAS 

Google Scholar 
Muñoz, A. A. et al. Water crisis in Petorca Basin, Chile: the combined effects of a mega-drought and water management. Water https://doi.org/10.3390/w12030648 (2020).Overpeck, J. T. & Udall, B. Climate change and the aridification of North America. Proc. Natl Acad. Sci. USA 117, 11856–11858 (2020).Article 
CAS 

Google Scholar 
Serrano-Notivoli, R. et al. Hydroclimatic variability in Santiago (Chile) since the 16th century. Int. J. Climatol. 41, E2015–E2030 (2021).Article 

Google Scholar 
Hock, R. et al. in Special Report on the Ocean and Cryosphere in a Changing Climate (eds Pörtner, H.-O. et al.) Ch. 2 (IPCC, Cambridge Univ. Press, 2019).Held, I. M. & Soden, B. J. Robust responses of the hydrological cycle to global warming. J. Clim. 19, 5686–5699 (2006).Article 

Google Scholar 
Xu, Y. & Ramanathan, V. Latitudinally asymmetric response of global surface temperature: implications for regional climate change. Geophys. Res. Lett. https://doi.org/10.1029/2012GL052116 (2012).Friedman, A. R., Hwang, Y.-T., Chiang, J. C. & Frierson, D. M. Interhemispheric temperature asymmetry over the twentieth century and in future projections. J. Clim. 26, 5419–5433 (2013).Article 

Google Scholar 
Putnam, A. E. & Broecker, W. S. Human-induced changes in the distribution of rainfall. Sci. Adv. https://doi.org/10.1126/sciadv.1600871 (2017).Allan, R. P. et al. Advances in understanding large-scale responses of the water cycle to climate change. Ann. N. Y. Acad. Sci. 1472, 49–75 (2020).Article 

Google Scholar 
Amatulli, G. et al. A suite of global, cross-scale topographic variables for environmental and biodiversity modeling. Sci. Data 5, 180040 (2018).Article 

Google Scholar 
Shea, J. M., Whitfield, P. H., Fang, X. & Pomeroy, J. W. The role of basin geometry in mountain snowpack responses to climate change. Front. Water 3, 4 (2021).Article 

Google Scholar 
Patricola, C. M. et al. Maximizing ENSO as a source of western US hydroclimate predictability. Clim. Dyn. https://doi.org/10.1007/s00382-019-05004-8 (2019).Eidhammer, T., Grubišić, V., Rasmussen, R. & Ikdea, K. Winter precipitation efficiency of mountain ranges in the Colorado Rockies under climate change. J. Geophys. Res. Atmos. 123, 2573–2590 (2018).Article 

Google Scholar 
Lynn, E. et al. Technical note: precipitation-phase partitioning at landscape scales to regional scales. Hydrol. Earth Syst. Sci. 24, 5317–5328 (2020).Article 
CAS 

Google Scholar 
Bales, R. C. et al. Mountain hydrology of the western United States. Water Resour. Res. https://doi.org/10.1029/2005WR004387 (2006).Jennings, K., Winchell, T. S., Livneh, B. & Molotch, N. P. Spatial variation of the rain–snow temperature threshold across the northern hemisphere. Nat. Commun. https://doi.org/10.1038/s41467-018-03629-7 (2018).Colombo, R. et al. Introducing thermal inertia for monitoring snowmelt processes with remote sensing. Geophys. Res. Lett. 46, 4308–4319 (2019).Article 

Google Scholar 
Demory, M. et al. The role of horizontal resolution in simulating drivers of the global hydrological cycle. Clim. Dyn. 42, 2201–2225 (2013).Article 

Google Scholar 
Rhoades, A. M., Ullrich, P. A. & Zarzycki, C. M. Projecting 21st century snowpack trends in western USA mountains using variable-resolution CESM. Clim. Dyn. 50, 261–288 (2017).Article 

Google Scholar 
Kapnick, S. B. et al. Potential for western US seasonal snowpack prediction. Proc. Natl Acad. Sci. USA 115, 1180–1185 (2018).Article 
CAS 

Google Scholar 
Palazzi, E., Mortarini, L., Terzago, S. & Von Hardenberg, J. Elevation-dependent warming in global climate model simulations at high spatial resolution. Clim. Dyn. 52, 2685–2702 (2019).Article 

Google Scholar 
Haarsma, R. J. et al. High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6. Geosci. Model Dev. 9, 4185–4208 (2016).Article 

Google Scholar 
O’Neill, B. C. et al. The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6. Geosci. Model Dev. 9, 3461–3482 (2016).Article 

Google Scholar 
Körner, C. et al. A global inventory of mountains for bio-geographical applications. Alp. Bot. 127, 1–15 (2017).Article 

Google Scholar 
Conover, W. J. Practical Nonparametric Statistics Vol. 350 (John Wiley & Sons, 1999).Woodhouse, C. A. & Pederson, G. T. Investigating runoff efficiency in Upper Colorado River streamflow over past centuries. Water Resour. Res. 54, 286–300 (2018).Article 

Google Scholar 
Lehner, F., Wahl, E. R., Wood, A. W., Blatchford, D. B. & Llewellyn, D. Assessing recent declines in Upper Rio Grande runoff efficiency from a paleoclimate perspective. Geophys. Res. Lett. 44, 4124–4133 (2017).Article 

Google Scholar 
Berghuijs, W., Woods, R. & Hrachowitz, M. A precipitation shift from snow towards rain leads to a decrease in streamflow. Nat. Clim. Change 4, 583–586 (2014).Article 

Google Scholar 
Li, D., Wrzesien, M. L., Durand, M., Adam, J. & Lettenmaier, D. P. How much runoff originates as snow in the western United States, and how will that change in the future? Geophys. Res. Lett. 44, 6163–6172 (2017).Article 

Google Scholar 
Livneh, B. & Badger, A. M. Drought less predictable under declining future snowpack. Nat. Clim. Change 10, 452–458 (2020).Article 

Google Scholar 
Trujillo, E. & Molotch, N. P. Snowpack regimes of the western United States. Water Resour. Res. 50, 5611–5623 (2014).Article 

Google Scholar 
Musselman, K. N., Clark, M. P., Liu, C., Ikeda, K. & Rasmussen, R. Slower snowmelt in a warmer world. Nat. Clim. Change 7, 214–219 (2017).Article 

Google Scholar 
Barnhart, T. B., Tague, C. L. & Molotch, N. P. The counteracting effects of snowmelt rate and timing on runoff. Water Resour. Res. 56, e2019WR026634 (2020).Article 

Google Scholar 
Bambach, N. E. et al. Projecting climate change in South America using variable-resolution Community Earth System Model: an application to Chile. Int. J. Climatol. https://doi.org/10.1002/joc.7379 (2021).Rhoades, A. M. et al. The shifting scales of western U.S. landfalling atmospheric rivers under climate change. Geophys. Res. Lett. 47, e2020GL089096 (2020).Article 

Google Scholar 
Rhoades, A. M., Risser, M. D., Stone, D. A., Wehner, M. F. & Jones, A. D. Implications of warming on western United States landfalling atmospheric rivers and their flood damages. Weather Clim. Extrem. 32, 100326 (2021).Article 

Google Scholar 
Milly, P. C. D. et al. Stationarity is dead: whither water management? Science 319, 573–574 (2008).Article 
CAS 

Google Scholar 
Cosgrove, W. J. & Loucks, D. P. Water management: current and future challenges and research directions. Water Resour. Res. 51, 4823–4839 (2015).Article 

Google Scholar 
Fernández, A. et al. Dendrohydrology and water resources management in south-central Chile: lessons from the Río Imperial streamflow reconstruction. Hydrol. Earth Syst. Sci. 22, 2921–2935 (2018).Article 

Google Scholar 
Castilla-Rho, J., Rojas, R., Andersen, M., Holley, C. & Mariethoz, G. Sustainable groundwater management: how long and what will it take? Glob. Environ. Change 58, 101972 (2019).Article 

Google Scholar 
Scanlon, B. R., Reedy, R. C., Faunt, C. C., Pool, D. & Uhlman, K. Enhancing drought resilience with conjunctive use and managed aquifer recharge in California and Arizona. Environ. Res. Lett. 11, 035013 (2016).Article 

Google Scholar 
Sterle, K., Hatchett, B. J., Singletary, L. & Pohll, G. Hydroclimate variability in snow-fed river systems: local water managers’ perspectives on adapting to the new normal. Bull. Am. Meteorol. Soc. 100, 1031–1048 (2019).Article 

Google Scholar 
Dillon, P. et al. Sixty years of global progress in managed aquifer recharge. Hydrogeol. J. 27, 1–30 (2019).Article 

Google Scholar 
Delaney, C. J. et al. Forecast informed reservoir operations using ensemble streamflow predictions for a multipurpose reservoir in Northern California. Water Resour. Res. 56, e2019WR026604 (2020).Article 

Google Scholar 
Szinai, J. K., Deshmukh, R., Kammen, D. M. & Jones, A. D. Evaluating cross-sectoral impacts of climate change and adaptations on the energy–water nexus: a framework and California case study. Environ. Res. Lett. 15, 124065 (2020).Article 

Google Scholar 
Vicuña, S. et al. in Water Resources of Chile (eds Fernández, B. & Gironás, J.) 347–363 (Springer International, 2021).Williams, J. H. et al. Carbon-neutral pathways for the United States. AGU Adv. 2, e2020AV000284 (2021).Article 

Google Scholar 
Fasullo, J. T. Evaluating simulated climate patterns from the CMIP archives using satellite and reanalysis datasets using the Climate Model Assessment Tool (CMATv1). Geosci. Model Dev. 13, 3627–3642 (2020).Article 

Google Scholar 
Hirai, M. et al. Development and validation of a new land surface model for JMA’s operational global model using the CEOP observation dataset. J. Meteorol. Soc. Japan II 85A, 1–24 (2007).Article 

Google Scholar 
Baldwin, J. W., Atwood, A. R., Vecchi, G. A. & Battisti, D. S. Outsize influence of Central American orography on global climate. AGU Adv. 2, e2020AV000343 (2021).Article 

Google Scholar 
Rhoades, A. M. et al. Sensitivity of mountain hydroclimate simulations in variable-resolution CESM to microphysics and horizontal resolution. J. Adv. Model. Earth Syst. 10, 1357–1380 (2018).Article 

Google Scholar 
Hawkins, E. & Sutton, R. The potential to narrow uncertainty in regional climate predictions. Bull. Am. Meteorol. Soc. 90, 1095–1108 (2009).Article 

Google Scholar 
Hawkins, E., Smith, R. S., Gregory, J. M. & Stainforth, D. A. Irreducible uncertainty in near-term climate projections. Clim. Dyn. 46, 3807–3819 (2016).Article 

Google Scholar 
Lehner, F. et al. Partitioning climate projection uncertainty with multiple large ensembles and CMIP5/6. Earth Syst. Dyn. 11, 491–508 (2020).Article 

Google Scholar 
Marshall, A. M., Abatzoglou, J. T., Link, T. E. & Tennant, C. J. Projected changes in interannual variability of peak snowpack amount and timing in the western United States. Geophys. Res. Lett. 46, 8882–8892 (2019).Article 

Google Scholar 
Huning, L. S. & AghaKouchak, A. Global snow drought hot spots and characteristics. Proc. Natl Acad. Sci. USA 117, 19753–19759 (2020).Article 
CAS 

Google Scholar 
Hatchett, B. J., Rhoades, A. M. & McEvoy, D. J. Monitoring the daily evolution and extent of snow drought. Nat. Hazards Earth Syst. Sci. 22, 869–890 (2022).Article 

Google Scholar 
Svoboda, M. et al. The Drought Monitor. Bull. Am. Meteorol. Soc. 83, 1181–1190 (2002).Article 

Google Scholar 
Sexstone, G. A., Driscoll, J. M., Hay, L. E., Hammond, J. C. & Barnhart, T. B. Runoff sensitivity to snow depletion curve representation within a continental scale hydrologic model. Hydrol. Process. 34, 2365–2380 (2020).
Google Scholar 
Mote, P. W., Li, S., Lettenmaier, D. P., Xiao, M. & Engel, R. Dramatic declines in snowpack in the western US. NPJ Clim. Atmos. Sci. 1, 2 (2018).Article 

Google Scholar 
Huning, L. S. & AghaKouchak, A. Approaching 80 years of snow water equivalent information by merging different data streams. Sci. Data 7, 333 (2020).Article 

Google Scholar 
Mote, P. W. et al. Perspectives on the causes of exceptionally low 2015 snowpack in the western United States. Geophys. Res. Lett. 43, 10980–10988 (2016).Article 

Google Scholar  More

Seneviratne, S. I. et al. Investigating soil moisture–climate interactions in a changing climate: a review. Earth Sci Rev. 99, 125–161, https://doi.org/10.1016/j.earscirev.2010.02.004 (2010).Article 
ADS 
CAS 

Google Scholar 
Bolten, J. D., Crow, W. T., Zhan, X., Jackson, T. J. & Reynolds, C. A. Evaluating the utility of remotely sensed soil moisture retrievals for operational agricultural drought monitoring. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 3, 57–66, https://doi.org/10.1109/JSTARS.2009.2037163 (2010).Article 
ADS 

Google Scholar 
Orth, R. & Seneviratne, S. I. Using soil moisture forecasts for sub-seasonal summer temperature predictions in Europe. Clim Dyn. 43, 3403–3418, https://doi.org/10.1007/s00382-014-2112-x (2014).Article 

Google Scholar 
Wanders, N., Karssenberg, D., de Roo, A., de Jong, S. M. & Bierkens, M. F. P. The suitability of remotely sensed soil moisture for improving operational flood forecasting. Hydrol. Earth Syst. Sci. 18, 2343–2357, https://doi.org/10.5194/hess-18-2343-2014 (2014).Article 
ADS 

Google Scholar 
Martínez-Fernández, J., González-Zamora, A., Sánchez, N., Gumuzzio, A. & Herrero-Jiménez, C. Satellite soil moisture for agricultural drought monitoring: assessment of the SMOS derived Soil Water Deficit Index. Remote Sens. Environ. 177, 277–286, https://doi.org/10.1016/j.rse.2016.02.064 (2016).Article 
ADS 

Google Scholar 
O, S., Hou, X. & Orth, R. Observational evidence of wildfire-promoting soil moisture anomalies. Sci. Rep. 10, 11008, https://doi.org/10.1038/s41598-020-67530-4 (2020).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Kroll, J. et al. Spatially varying relevance of hydrometeorological hazards for vegetation productivity extremes. Biogeosciences 19, 477–489, https://doi.org/10.5194/bg-19-477-2022 (2022).Article 
ADS 
CAS 

Google Scholar 
Famiglietti, J. S., Ryu, D., Berg, A. A., Rodell, M. & Jackson, T. J. Field observations of soil moisture variability across scales: Soil moisture variability across scales. Water Resour. Res. 44, https://doi.org/10.1029/2006WR005804 (2008).Brocca, L., Ciabatta, L., Massari, C., Camici, S. & Tarpanelli, A. Soil moisture for hydrological applications: Open questions and new opportunities. Water 9, 140, https://doi.org/10.3390/w9020140 (2017).Article 

Google Scholar 
Rodell, M. et al. The global land data assimilation system. Bull. Amer. Meteor. 85, 381–394, https://doi.org/10.1175/BAMS-85-3-381 (2004).Article 
ADS 

Google Scholar 
Naz, B. S., Kollet, S., Franssen, H.-J. H., Montzka, C. & Kurtz, W. A 3 km spatially and temporally consistent European daily soil moisture reanalysis from 2000 to 2015. Sci. Data 7, 111, https://doi.org/10.1038/s41597-020-0450-6 (2020).Article 
PubMed 
PubMed Central 

Google Scholar 
Muñoz-Sabater, J. et al. ERA5-land: a state-of-the-art global reanalysis dataset for land applications. Earth Syst. Sci. Data 13, 4349–4383, https://doi.org/10.5194/essd-13-4349-2021 (2021).Article 
ADS 

Google Scholar 
Koster, R. D. et al. On the nature of soil moisture in land surface models. J. Clim. 22, 4322–4335, https://doi.org/10.1175/2009JCLI2832.1 (2009).Article 
ADS 

Google Scholar 
Petropoulos, G. P., Ireland, G. & Barrett, B. Surface soil moisture retrievals from remote sensing: Current status, products & future trends. Phys. Chem. Earth. 83-84, 36–56, https://doi.org/10.1016/j.pce.2015.02.009 (2015).Article 
ADS 

Google Scholar 
Dorigo, W. et al. ESA CCI Soil Moisture for improved Earth system understanding: state-of-the art and future directions. Remote Sens. Environ. 203, 185–215, https://doi.org/10.1016/j.rse.2017.07.001 (2017).Article 
ADS 

Google Scholar 
Chan, S. et al. Development and assessment of the SMAP enhanced passive soil moisture product. Remote Sens. Environ. 204, 931–941, https://doi.org/10.1016/j.rse.2017.08.025 (2018).Article 
ADS 
PubMed 

Google Scholar 
Yao, P. et al. A long term global daily soil moisture dataset derived from AMSR-E and AMSR2 (2002–2019). Sci. Data 8, 143, https://doi.org/10.1038/s41597-021-00925-8 (2021).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Park, S. et al. Downscaling GLDAS soil moisture data in East Asia through fusion of multi-sensors by optimizing modified regression trees. Water 9, 332, https://doi.org/10.3390/w9050332 (2017).Article 

Google Scholar 
Mao, H., Kathuria, D., Duffield, N. & Mohanty, B. P. Gap filling of high–resolution soil moisture for SMAP/sentinel–1: A two–layer machine learning–based framework. Water Resour. Res. 55, 6986–7009, https://doi.org/10.1029/2019WR024902 (2019).Article 
ADS 

Google Scholar 
Guevara, M., Taufer, M. & Vargas, R. Gap-free global annual soil moisture: 15 km grids for 1991–2018. Earth Syst. Sci. Data 13, 1711–1735, https://doi.org/10.5194/essd-13-1711-2021 (2021).Article 
ADS 

Google Scholar 
O, S. & Orth, R. Global soil moisture data derived through machine learning trained with in-situ measurements. Sci. Data 8, 170, https://doi.org/10.1038/s41597-021-00964-1 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
Guo, Z., Dirmeyer, P. A., Gao, X. & Zhao, M. Improving the quality of simulated soil moisture with a multi-model ensemble approach. Q.J.R. Meteorol. Soc. 133, 731–747, https://doi.org/10.1002/qj.48 (2007).Article 
ADS 

Google Scholar 
Bai, W. et al. The performance of multiple model-simulated soil moisture datasets relative to ECV satellite data in China. Water 10, 1384, https://doi.org/10.3390/w10101384 (2018).Article 

Google Scholar 
Reichstein, M. et al. Deep learning and process understanding for data-driven Earth system science. Nature 566, 195–204, https://doi.org/10.1038/s41586-019-0912-1 (2019).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Geer, A. J. Learning earth system models from observations: machine learning or data assimilation. Phil. Trans. R. Soc. A. 379, 20200089, https://doi.org/10.1098/rsta.2020.0089 (2021).Article 
ADS 
CAS 
PubMed 

Google Scholar 
Zhang, Y., Keenan, T. F. & Zhou, S. Exacerbated drought impacts on global ecosystems due to structural overshoot. Nat. Ecol. Evol. 5, 1490–1498, https://doi.org/10.1038/s41559-021-01551-8 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
Bastos, A. et al. Vulnerability of European ecosystems to two compound dry and hot summers in 2018 and 2019. Earth Syst. Dynam. 12, 1015–1035, https://doi.org/10.5194/esd-12-1015-2021 (2021).Article 
ADS 

Google Scholar 
O, S. et al. The role of climate and vegetation in regulating drought-heat extremes. J. Clim. 35, 5677–5685, https://doi.org/10.1175/JCLI-D-21-0675.1 (2022).Meng, X., Wang, H., Chen, J., Yang, M. & Pan, Z. High-resolution simulation and validation of soil moisture in the arid region of Northwest China. Sci. Rep. 9, 17227, https://doi.org/10.1038/s41598-019-52923-x (2019).Article 
ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Peng, J. et al. A roadmap for high-resolution satellite soil moisture applications – confronting product characteristics with user requirements. Remote Sens. Environ. 252, 112162, https://doi.org/10.1016/j.rse.2020.112162 (2021).Article 
ADS 

Google Scholar 
Sabaghy, S., Walker, J. P., Renzullo, L. J. & Jackson, T. J. Spatially enhanced passive microwave derived soil moisture: Capabilities and opportunities. Remote Sens. Environ. 209, 551–580, https://doi.org/10.1016/j.rse.2018.02.065 (2018).Article 
ADS 

Google Scholar 
Nayak, H. P. et al. High-resolution gridded soil moisture and soil temperature datasets for the indian monsoon region. Sci. Data 5, 180264, https://doi.org/10.1038/sdata.2018.264 (2018).Article 
CAS 
PubMed 
PubMed Central 

Google Scholar 
Vergopolan, N. et al. Field-scale soil moisture bridges the spatial-scale gap between drought monitoring and agricultural yields. Hydrol. Earth Syst. Sci. 25, 1827–1847, https://doi.org/10.5194/hess-25-1827-2021 (2021).Article 
ADS 

Google Scholar 
Abbaszadeh, P. et al. High-resolution SMAP satellite soil moisture product: Exploring the opportunities. Bull. Amer. Meteor. 102, 309–315, https://doi.org/10.1175/BAMS-D-21-0016.1 (2021).Article 
ADS 

Google Scholar 
Hochreiter, S. & Schmidhuber, J. Long Short-term memory. Neural Comput. 9, 1735–1780, https://doi.org/10.1162/neco.1997.9.8.1735 (1997).Article 
CAS 
PubMed 

Google Scholar 
Gao, P. et al. Modeling for the prediction of soil moisture in litchi orchard with deep long short-term memory. Agriculture 12, 25, https://doi.org/10.3390/agriculture12010025 (2021).Article 

Google Scholar 
Li, Q. et al. An attention-aware LSTM model for soil moisture and soil temperature prediction. Geoderma 409, 115651, https://doi.org/10.1016/j.geoderma.2021.115651 (2022).Article 
ADS 

Google Scholar 
Dorigo, W. A. et al. The International Soil Moisture Network: a data hosting facility for global in situ soil moisture measurements. Hydrol. Earth Syst. Sci. 15, 1675–1698, https://doi.org/10.5194/hess-15-1675-2011 (2011).Article 
ADS 

Google Scholar 
Dorigo, W. et al. The International Soil Moisture Network: serving Earth system science for over a decade. Hydrol. Earth Syst. Sci. 25, 5749–5804, https://doi.org/10.5194/hess-25-5749-2021 (2021).Article 
ADS 

Google Scholar 
O, S., Dutra, E. & Orth, R. Robustness of process-based versus data-driven modeling in changing climatic conditions. J. Hydrometeorol. 21, 1929–1944, https://doi.org/10.1175/JHM-D-20-0072.1 (2020).Article 
ADS 

Google Scholar 
Beck, H. E. et al. Global-scale regionalization of hydrologic model parameters. Water Resour. Res. 52, 3599–3622, https://doi.org/10.1002/2015WR018247 (2016).Article 
ADS 

Google Scholar 
Mittelbach, H. & Seneviratne, S. I. A new perspective on the spatio-temporal variability of soil moisture: temporal dynamics versus time-invariant contributions. Hydrol. Earth Syst. Sci. 16, 2169–2179, https://doi.org/10.5194/hess-16-2169-2012 (2012).Article 
ADS 

Google Scholar 
Amante, C. ETOPO1 1 arc-minute global relief model: Procedures, data sources and analysis. National Geophysical Data Center, NOAA https://doi.org/10.7289/V5C8276M (2009).Wieder, W. Regridded harmonized world soil database v1.2. ORNL Distributed Active Archive Center https://doi.org/10.3334/ORNLDAAC/1247 (2014).Kratzert, F. et al. Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets. Hydrol. Earth Syst. Sci. 23, 5089–5110, https://doi.org/10.5194/hess-23-5089-2019 (2019).Article 
ADS 

Google Scholar 
LeCun, Y. A., Bottou, L., Orr, G. B. & Müller, K.-R. Efficient BackProp. In Neural networks: tricks of the trade, Second Edition, 9–48, https://doi.org/10.1007/978-3-642-35289-8_3 (Springer Berlin Heidelberg, Berlin, Heidelberg, 2012).Gauch, M. et al. Rainfall–runoff prediction at multiple timescales with a single long short-term memory network. Hydrol. Earth Syst. Sci. 25, 2045–2062, https://doi.org/10.5194/hess-25-2045-2021 (2021).Article 
ADS 

Google Scholar 
O, S., Orth, R., Weber, U. & Park, S. K. High-resolution european daily soil moisture derived with machine learning (2003–2020). Figshare https://doi.org/10.6084/m9.figshare.c.5957127 (2022).Martens, B. et al. GLEAM v3: satellite-based land evaporation and root-zone soil moisture. Geosci. Model Dev. 10, 1903–1925, https://doi.org/10.5194/gmd-10-1903-2017 (2017).Article 
ADS 

Google Scholar 
Bogena, H. R. et al. COSMOS-europe: A european network of cosmic-ray neutron soil moisture sensors. Earth Syst. Sci. Data. 14, https://doi.org/10.5194/essd-14-1125-2022 (2022).Koster, R. D. & Suarez, M. J. Soil moisture memory in climate models. J. Hydrometeorol. 2, 558–570, https://doi.org/10.1175/1525-7541(2001)0022.0.CO;2 (2001).Orth, R., Koster, R. D. & Seneviratne, S. I. Inferring soil moisture memory from streamflow observations using a simple water balance model. J. Hydrometeorol. 14, 1773–1790, https://doi.org/10.1175/JHM-D-12-099.1 (2013).Article 
ADS 

Google Scholar 
Wu, W. & Dickinson, R. E. Time scales of layered soil moisture memory in the context of land–atmosphere interaction. J. Clim. 17, 2752–2764, 10.1175/1520-0442(2004)0172.0.CO;2 (2004).McColl, K. A. et al. The global distribution and dynamics of surface soil moisture. Nature Geosci. 10, 100–104, https://doi.org/10.1038/ngeo2868 (2017).Article 
ADS 
CAS 

Google Scholar 
Laaha, G. et al. The european 2015 drought from a hydrological perspective. Hydrol. Earth Syst. Sci. 21, 3001–3024, https://doi.org/10.5194/hess-21-3001-2017 (2017).Article 
ADS 

Google Scholar 
Ionita, M. et al. The European 2015 drought from a climatological perspective. Hydrol. Earth Syst. Sci. 21, 1397–1419, https://doi.org/10.5194/hess-21-1397-2017 (2017).Article 
ADS 

Google Scholar 
Meyer, H. & Pebesma, E. Predicting into unknown space? estimating the area of applicability of spatial prediction models. Methods Ecol. Evol. 12, 1620–1633, https://doi.org/10.1111/2041-210X.13650 (2021).Article 

Google Scholar 
Ghiggi, G., Humphrey, V., Seneviratne, S. I. & Gudmundsson, L. GRUN: an observation-based global gridded runoff dataset from 1902 to 2014. Earth Syst. Sci. Data 11, 1655–1674, https://doi.org/10.5194/essd-11-1655-2019 (2019).Article 
ADS 

Google Scholar 
Jung, M. et al. The FLUXCOM ensemble of global land-atmosphere energy fluxes. Sci. Data 6, 74, https://doi.org/10.1038/s41597-019-0076-8 (2019).Article 
PubMed 
PubMed Central 

Google Scholar 
Kraft, B., Jung, M., Körner, M., Koirala, S. & Reichstein, M. Towards hybrid modeling of the global hydrological cycle. Hydrol. Earth Syst. Sci. 26, 1579–1614, https://doi.org/10.5194/hess-26-1579-2022 (2022).Article 
ADS 

Google Scholar 
Dabrowska-Zielinska, K. et al. Soil moisture in the Biebrza wetlands retrieved from Sentinel-1 imagery. Remote Sens. 10, https://doi.org/10.3390/rs10121979 (2018).Brocca, L. et al. Soil moisture estimation through ASCAT and AMSR-E sensors: An intercomparison and validation study across Europe. Remote Sens. Environ. 115, 3390–3408, https://doi.org/10.1016/j.rse.2011.08.003 (2011).Article 
ADS 

Google Scholar 
Zreda, M. et al. COSMOS: the COsmic-ray Soil Moisture Observing System. Hydrol Earth Syst Sci. 16, 4079–4099, https://doi.org/10.5194/hess-16-4079-2012 (2012).Article 
ADS 

Google Scholar 
Ikonen, J. et al. The Sodankylä in situ soil moisture observation network: an example application of ESA CCI soil moisture product evaluation. Geosci. Instrum. Methods Data Syst. 5, 95–108, https://doi.org/10.5194/gi-5-95-2016 (2016).Article 
ADS 

Google Scholar 
Al-Yaari, A. et al. The AQUI soil moisture network for satellite microwave remote sensing validation in south-western France. Remote Sens. 10, https://doi.org/10.3390/rs10111839 (2018).Cobley, A., Hemment, D., Rowan, J., Taylor, N. & Woods, M. GROW soil moisture data. University of Dundee https://doi.org/10.15132/10000156 (2020).Bircher, S., Skou, N., Jensen, K. H., Walker, J. P. & Rasmussen, L. A soil moisture and temperature network for SMOS validation in Western Denmark. Hydrol. Earth Syst. Sci. 16, 1445–1463, https://doi.org/10.5194/hess-16-1445-2012 (2012).Article 
ADS 

Google Scholar 
Morbidelli, R., Saltalippi, C., Flammini, A., Rossi, E. & Corradini, C. Soil water content vertical profiles under natural conditions: matching of experiments and simulations by a conceptual model. Hydrol. Process. 28, 4732–4742, https://doi.org/10.1002/hyp.9973 (2014).Article 
ADS 

Google Scholar 
Biddoccu, M., Ferraris, S., Opsi, F. & Cavallo, E. Long-term monitoring of soil management effects on runoff and soil erosion in sloping vineyards in Alto Monferrato (North–West Italy. Soil Tillage Res. 155, 176–189, https://doi.org/10.1016/j.still.2015.07.005 (2016).Article 

Google Scholar 
Beyrich, F. & Adam, W. Site and Data Report for the Lindenberg Reference Site in CEOP – Phase 1. Berichte des Deutschen Wetterdienstes (2007).Sanchez, N., Martinez-Fernandez, J., Scaini, A. & Perez-Gutierrez, C. Validation of the SMOS L2 soil moisture data in the REMEDHUS network (Spain. IEEE Trans. Geosci. Remote Sens. 50, 1602–1611, https://doi.org/10.1109/TGRS.2012.2186971 (2012).Article 
ADS 

Google Scholar 
Schaefer, G. L., Cosh, M. H. & Jackson, T. J. The USDA natural resources conservation service soil climate analysis network (SCAN. J. Atmos. Ocean. Technol. 24, 2073–2077, https://doi.org/10.1175/2007JTECHA930.1 (2007).Article 
ADS 

Google Scholar 
Calvet, J.-C. et al. In situ soil moisture observations for the CAL/VAL of SMOS: the SMOSMANIA network. In 2007 IEEE International Geoscience and Remote Sensing Symposium, 1196–1199, https://doi.org/10.1109/IGARSS.2007.4423019 (IEEE, 2007).Marczewski, W. et al. Strategies for validating and directions for employing SMOS data, in the Cal-Val project SWEX (3275) for wetlands. Hydrol. Earth Syst. Sci. Discuss. 7, 7007–7057, https://doi.org/10.5194/hessd-7-7007-2010 (2010).Article 
ADS 

Google Scholar 
Zacharias, S. et al. A network of terrestrial environmental observatories in Germany. Vadose Zone J. 10, 955–973, https://doi.org/10.2136/vzj2010.0139 (2011).Article 

Google Scholar 
Schlenz, F., dall’Amico, J. T., Loew, A. & Mauser, W. Uncertainty assessment of the SMOS validation in the upper Danube catchment. IEEE Trans. Geosci. Remote Sens. 50, 1517–1529, https://doi.org/10.1109/TGRS.2011.2171694 (2012).Article 
ADS 

Google Scholar 
Bell, J. E. et al. U.S. Climate Reference Network soil moisture and temperature observations. J. Hydrometeor. 14, 977–988, https://doi.org/10.1175/JHM-D-12-0146.1 (2013).Article 
ADS 

Google Scholar 
Kirchengast, G., Kabas, T., Leuprecht, A., Bichler, C. & Truhetz, H. WegenerNet: a pioneering high-resolution network for monitoring weather and climate. Bull. Amer. Meteor. 95, 227–242, https://doi.org/10.1175/BAMS-D-11-00161.1 (2014).Article 
ADS 

Google Scholar 
Petropoulos, G. P. & McCalmont, J. P. An operational in situ soil moisture & soil temperature monitoring network for West Wales, UK: The WSMN network. Sensors 95, 227–242, https://doi.org/10.3390/s17071481 (2014).Article 

Google Scholar 
Beck, H. E. et al. MSWX: Global 3-hourly 0.1° bias-corrected meteorological data including near-real-time updates and forecast ensembles. Bull. Amer. Meteor. 103, E710–E732, https://doi.org/10.1175/BAMS-D-21-0145.1 (2022).Article 

Google Scholar 
Beck, H. E. et al. MSWEP V2 global 3-hourly 0.1° precipitation: Methodology and quantitative assessment. Bull. Amer. Meteor. 100, 473–500, https://doi.org/10.1175/BAMS-D-17-0138.1 (2019).Article 
ADS 

Google Scholar 
Fischer, G. et al. Global agro-ecological zones assessment for agriculture (gaez 2008). IIASA, Laxenburg, Austria and FAO, Rome, Italy (2008). More

Fishing vessels have legitimate reasons to turn off their position-tracking systems — but there are some suspicious reasons, too.Credit: Anthony Wallace/AFP/Getty

When fishing vessels hide their locations, they sometimes reveal a wealth of information. Gaps in tracking data can hint at illegal activity, finds a modelling study1.Some ships carry automatic identification systems (AIS), which pinpoint their locations and help to prevent collisions, but can be turned off manually. Researchers studied gaps in the tracking data to identify hotspots where fishing vessels frequently disabled their devices on purpose — and to explore the possible reasons. The findings suggest that vessels hid up to 6% of their activity — more than 4.9 million hours between 2017 and 2019. Some of these gaps could mask illegal fishing, finds the study, which was published in Science Advances this month..The study uses holes in tracking data “to tell us more about what we’re not seeing, what we’re missing”, says Juan Mayorga, a marine data scientist based in Santa Barbara, California, who is part of the National Geographic Society’s Pristine Seas project. “That is a really valuable contribution.”Expensive problemIllegal, unreported and unregulated fishing costs the global economy up to US$25 billion each year. It is also detrimental to marine life, and some evidence suggests that it is linked to human-rights violations such as people trafficking. Heather Welch, a spatial ecologist at the University of California, Santa Cruz, and her colleagues analysed more than 3.7 billion signals from vessels, sent over three years and recorded in the Global Fishing Watch AIS data set. The team used a model to distinguish between gaps caused by vessels intentionally turning off their AIS and those that were due to technical issues. Gaps of 12 hours or more when ships were at least 50 nautical miles from shore in areas with adequate signal reception were suspected to be intentional disabling.

Source: Ref 1.

The team found that 82% of time lost to AIS disabling happened on ships flagged from Spain, the United States, Taiwan and the Chinese mainland (see ‘Flag of origin’). Although most vessels that use AIS come from middle- and upper-income countries, so the data are biased towards those countries, the study says. “AIS is not feasible for a lot of countries globally at the moment,” says Claire Collins, a marine social scientist at the Zoological Society of London.There are many reasons vessels intentionally turn off their AIS, says Welch, and not all of them are nefarious. For instance, crews might hide their location in areas where pirates are a threat, or might obscure their position from competitors when fishing in a bountiful area. More iniquitous reasons to hide a ship’s location include trying to mask illegal fishing or unauthorized transshipment — transfers of cargo between ships at sea — she says.The team used another model to investigate what was behind the intentional AIS signal gaps, looking at factors such as how productive an area is for fishing, the risk of piracy and the level of transshipment activity. The results indicate locations in which the signal gaps are potentially nefarious, but they cannot definitively say whether these gaps hide illegal activity, says Welch.HotspotsThe model revealed 4 hotspots for intentional AIS disabling: 16% of gaps occurred next to Argentina’s exclusive economic zone, 13% in the Northwest Pacific Ocean, 8% adjacent to the exclusive economic zones of West African nations and 3% near Alaska. Apart from Alaska, these hotspots are already regions of concern for illegal, unreported and unregulated fishing. They produce a lot of fish and have limited management, partially because of their locations in the high seas. Signal gaps near exclusive economic zones indicate that vessels could be hiding that they are crossing boundaries without authorization to fish in restricted areas, says Welch. “If they were allowed to go in that zone, why would they disable their AIS?” she says.Drifting longlines were the fishing vessels found to disable their AIS most often, followed by tuna purse seines (see ‘Out of sight’). Intentional AIS disabling events were also common near transshipment hotspots. Offloading catch at sea helps to reduce costs, but past research has linked it to human trafficking and slipping illegal catch on to the market.

Source: Ref 1.

The research is a good way to start exploring what AIS-disabling data can expose, and could help researchers to conduct finer-scale studies in the future, says Collins. “It’s a really important study.”Mayorga agrees that the data will aid fishery managers in understanding the magnitude and patterns of illegal fishing, helping them to zero in on specific problematic regions and improve enforcement of laws at sea. More

Figure 4 shows a sample of the seasonal cycle and spatial distribution of SWE over HUC2 basins and the entire WUS domain in WY 2019. No SWE or snow depth measurements are assimilated in deriving the WUS–SR dataset. Thus, in situ SWE and snow depth measurements, and ASO SWE and snow depth estimates are used as independent verification datasets. Landsat fSCA measurements are assimilated into the snow reanalysis framework assuming a measurement error (standard deviation) of 10%34. Though Landsat fSCA cannot be used for independent verification, the WUS–SR posterior fSCA estimates, which are fitted to these measurements using a likelihood function, are expected to have comparable bulk error. The snow reanalysis framework has been successfully applied previously to generate datasets over the Sierra Nevada, Andes, and High Mountain Asia33,50,52.Fig. 4Illustrative results from the WUS–SR SWE estimates in WY 2019. (a) Seasonal cycle of SWE volume (km3) integrated over HUC2 basins. (b) Spatial distribution of SWE (meters) over part of the Sierra Nevada on March 1st, WY 2019. (c) Spatial distribution of WUS SWE (meters) on March 1st, 2019. The boxed area in (c) represents that shown in (b).Full size imageVerification with in situ dataIn this section, grid-averaged reanalysis SWE and snow depth are compared with point-scale in situ measurements. It should be acknowledged a priori that there are inevitable representativeness issues in the comparison between point-scale in situ data and grid-averaged snow reanalysis data. The WUS–SR estimates are modeled with assumed sub-grid heterogeneity within each ~500 m grid cell (which is modeled via a lognormal distribution) meant to account for the complex sub-grid variations in terrain (elevation, slope, aspect), forest cover, and meteorological forcings. Given that in situ stations are often sited in non-representative regions of a grid cell (i.e., in sheltered flat forest clearings), it is unlikely that the grid-averaged SWE/snow depth (spanning ~ 250,000 m2) should match the point-scale in situ SWE/snow depth (spanning ~10 m2). Nevertheless, in situ measurements, from the SNOTEL and CA Department of Water Resources (CADWR) networks, represent the best available data that covers much of the WUS and extends back several decades. While not expected to match each other, the verification herein is meant to illustrate consistency between the in situ measurements and WUS–SR estimates.Peak SWE comparison with in situ dataIn situ SWE measurements from WY 1985 to 2021 are taken from 1) the SNOTEL network (https://www.wcc.nrcs.usda.gov/snow/) managed by the U.S. Natural Resources Conservation Service (NRCS), and 2) CADWR (https://cdec.water.ca.gov/dynamicapp/staSearch from sensor type: “SNO ADJ (82)”), collections of automated snow pillows in the WUS. For in situ verification, we pair each in situ site with the closest snow reanalysis grid based on the geolocation of these two datasets. The precision of in situ coordinate values varies from 0.000001° (1 km). Considering the potential for geolocation mismatch, the nine nearest pixels32,33,55 are additionally used to compare in situ and WUS–SR peak SWE. In this latter approach, the differences between in situ peak SWE and the neighboring WUS–SR grid cell peak SWE with the smallest difference among the nine nearest snow reanalysis grids are used. To compare the SWE on the same day, peak SWE day determined by in situ SWE is used to extract peak SWE from both datasets throughout the paper.Figure 5 presents the density scatter plots comparing in situ peak SWE values against collocated grid-cell posterior peak SWE values. Peak SWE values less than 1 cm are screened out from the comparison. In total, 928 in situ sites are used in the comparison with the WUS–SR SWE estimates. To understand the performance of the WUS–SR dataset across different regimes in the WUS, verification is conducted for each HUC2 basin. The comparison is quantified using correlation coefficient (R), mean difference (MD), and root mean square difference (RMSD). Table 5 summarizes the number of total site-years, and statistics for both prior and posterior reanalysis SWE against in situ SWE within each HUC2 basin and over the WUS.Fig. 5Density scatter plot of in situ (snow pillow) peak SWE and collocated posterior (grid-average) peak SWE grouped by HUC2 basins over WYs 1985 to 2021. The solid black line is the 1:1 line. The correlation coefficient (R), mean difference (MD), and root mean square difference (RMSD) are shown for each HUC2 basin. In situ data with peak SWE values greater than 1 cm are included in the comparison.Full size imageTable 5 Number of in situ sites and comparison metrics between in situ (snow pillow) peak SWE and collocated grid-averaged snow reanalysis prior and posterior (post.) peak SWE grouped by HUC2 basins.Full size tableCompared with the performance of the prior peak SWE estimates (i.e., not constrained by Landsat fSCA), posterior SWE estimates show a better correlation (higher R) with less bias and random error (lower MD and RMSD) than the prior SWE over most of the HUC2 basins. Posterior SWE in CA has the highest correlation against in situ SWE (R = 0.82). The correlations with in situ SWE over the entire WUS are improved from 0.74 (prior) to 0.77 (posterior). Posterior peak SWE in UCRB has lower bias and uncertainty compared against in situ data with a relatively small MD of 0.06 m in absolute value (reduced by 62% from prior MD) and RMSD of 0.19 m (reduced by 27%). Over the WUS, in situ peak SWE is (on average) larger than the WUS–SR peak SWE (negative MD). Sub-grid topographic variability, snow-forest interactions, and wind-driven snow redistribution may all cause differences seen between grid-averaged peak SWE and point-scale in situ peak SWE. The statistics for PN indicate comparable correlation of in situ and both prior and posterior snow reanalysis, however the MD and RMSD do not get improved from posterior to prior. Fewer cloud-free fSCA measurements are available in PN, which limits the improvement of snow reanalysis SWE via data assimilation.To acknowledge the potential geolocation mismatch, Fig. 6 provides verification of in situ peak SWE and posterior reanalysis peak SWE using an approach comparing to the best match among the nine nearest pixels. The WUS-wide correlation coefficient (R), MD and RMSD of posterior peak SWE and in situ peak SWE is 0.91, −0.08 m, 0.18 m, respectively. Compared to the approach used in Fig. 5, the posterior reanalysis peak SWE in Fig. 6 (as expected) is more correlated with in situ peak SWE (R values above 0.9), and has lower MD ( More

Evaporation structure design and fabricationFor conventional salt-rejection solar evaporation systems, water evaporation is confined to the solar absorber surface, and the salt backflow is accompanied by an undesired heat dissipation from the solar absorber to bulk water, thus resulting in a low evaporation rate. This limitation can be solved to a considerable extent by our 3D evaporator. As illustrated in Fig. 1a, the top surface of our evaporator is a solar absorber layer used for light-to-heat conversion to generate vapour. Beneath the solar absorber are a number of vertically aligned MTBs connecting the saline water to the solar absorber. MTBs have hydrophilic microchannels that can pump saline water to the solar absorber via a capillary force. Furthermore, excessive salt can flow back into the bulk water through these brine-filled microchannels via diffusion and convection (Fig. 1b-1). The adequate mass transfer via a high density of MTBs ensures a continuous water supply and an efficient salt backflow, thus enabling a unique salt rejection capability. Unlike conventional salt-rejection systems, where the heat conducted from the solar absorber to the bulk water is simply dissipated and considered “wasted,” the MTBs can efficiently recover this conductive heat to generate additional vapour from the brine flowing through their microchannels (Fig. 1b-2). The microchannels within the MTBs and macrochannels between the spaced MTBs together form a highly open structure that allows the generated vapour to be easily released from the MTB surfaces in all directions. We envision that by optimizing the MTB height, conductive heat can be largely confined in them for vapour generation, thereby significantly improving the water evaporation efficiency.Fig. 1: Design and fabrication of the 3D salt-rejection evaporation structure.a Schematic of the 3D salt-rejection solar evaporator. b Working principle includes salt rejection and evaporation enhancement. c UV–Vis–NIR spectra of the GFM, CNT-coated GFM, and standard solar irradiation spectrum of AM 1.5 G. d SEM image of the CNT-coated GFM surface. e SEM image of the GFM. f Image of the water drop hanging above the GFM and the moment it touches the GFM surface. g Anti-gravity transport of water along a GFM. h 3D salt-rejection evaporator prototype. i Schematic illustration of the fabricating process of the evaporator.Full size imageWe achieved the designed structure by fabricating the top solar absorber layer by loading carbon nanotubes (CNTs) with a diameter of about one hundred nanometres on a glass fibre membrane (GFM). The solar absorption of wet CNT-coated GFM can reach ~96% (Fig. 1c) because of the porous fibrous light-trapping structure (Fig. 1d) and the inherent black property of the CNT27. Considering their abundant hydrophilic microchannels formed by intertwined glass fibres (Fig. 1e), the GFMs were also selected for use as MTBs. A GFM can immediately absorb a water droplet upon touching it because of its high affinity to water (Fig. 1f). Moreover, vertically aligned GFMs (i.e., MTBs) can pump water to 25 cm height in 60 min, demonstrating its strong capillary force for water transfer (Fig. 1g). A complete evaporation system was fabricated by assembling a number of MTBs and the solar absorber in a plastic frame (Fig. 1h, 1i and Fig. S1).Salt rejection capabilityTo avoid salt crystallization, excess salt must be efficiently transported back to maintain the top surface salinity below the saturation point. In this system, salt can be rejected via diffusion and convection through brine-filled microchannels under the driving force of the concentration gradient (osmosis) and gravity25. Its mass flow rate ((J)) can be described by the diffusion–convection equation as follows28,29:$$J={J}_{{diff}}+{J}_{{conv}}={nA}varepsilon ({k}_{d}({C}_{{evp}}-{C}_{0})/l+{k}_{c}({rho }_{{evp}}-{rho }_{0}))$$
(1)
where ({J}_{{diff}}) and ({J}_{{conv}}) are the mass flow rate caused by diffusion and convection, respectively; (n) is the number of MTBs; (A), (varepsilon), and (l) are the cross-section area, porosity, and height of the MTBs, respectively; ({k}_{d}) and ({k}_{c}) are the diffusion and average convective coefficients of salt, respectively; ({C}_{{evp}}) and ({C}_{0}) are the salt concentrations on the evaporation surface and in the bulk saline water, respectively; and ({rho }_{{evp}}) and ({rho }_{0}) are the salt solution densities on the evaporation surface and in the bulk saline water, respectively.In Eq. (1), the mass transport rate is proportional to the bridge number (n). We validated this relation by fabricating MTB structures with different bridge numbers ranging from 2 to 32 [Fig. 2a, cross-section area ((A)): ~0.135 cm2; height ((l)): 3 cm; porosity ((varepsilon)): ~65%] and evaluating their evaporation performance using high-salinity water (10 wt.% NaCl). The evaluation was performed under 1 sun illumination for 12 h. Figure 2b shows that salt crystals massively accumulated on the 2-bridge evaporator surface because of insufficient mass transfer. This salt accumulation was mitigated with increase in the bridge number. For the evaporator containing 32 MTBs, no salt crystals were observed on the surface after the 12 h operation (Fig. 2b). At an insufficient number of MTBs (≤16), the evaporation rate gradually decreased as the vapour generation progressed because of the increased evaporation surface salinity (Fig. 2c, see the corresponding mass change curves in Fig. S2). In contrast, with sufficient MTBs (e.g., 32 bridges), the excess salt can be efficiently rejected to maintain the evaporation surface at a relatively low salinity. Remarkably, the evaporation rate of the 32-bridge evaporator was ~1.44 kg/m2/h without degradation during the 12 h operation.Fig. 2: Salt rejection performance.a Photograph of evaporators with various bridge numbers (bridg height: 3 cm). b Photographic recordings of the salt accumulation on the 3D evaporators with different MTB numbers. c Evaporation rate variations of evaporators during long-term operations. d Photos of salt redissolving from the surface of a 32-bridge evaporator.Full size imageSubsequently, we performed a complementary experiment to more intuitively demonstrate the salt backflow introduced by the 32-bridge evaporator. In this experiment, the evaporator was placed in a high-concentration saline water (10 wt.% NaCl solution) and exposed to 1 sun illumination, and 1 g of NaCl salt was added on its surface (upper panel, Fig. 2d). It was seen that during vapour generation, the added salt was gradually dissolved and completely removed in 11 h (lower panel, Fig. 2d; more details in Fig. S3). This experiment demonstrated that the salt backflow rate of the 32-bridge evaporator in the 10 wt.% NaCl solution was higher than the salt generation rate, thus confirming the salt rejection feature of the proposed MTB architecture. We further increased the brine salinity to test the maximum applicable salt concentration of this evaporator. Because the effects of diffusion and convection backflow decreased as the salinity (i.e., ({C}_{0}) and ({rho }_{0})) increased, salt started to crystallize at the edges of the solar absorber after 12 h operation when 14 wt.% NaCl solution was used for the test (Fig. S4). Based on the corresponding evaporation rate, the salt backflow along the MTBs was calculated as ~1.1 g/cm2/h. Interestingly, this unique mass transport feature is intertwined with its heat transport feature, as demonstrated in the subsequent section.Heat managementWe fabricated 32-bridge evaporators with different bridge heights (Fig. 3a) and evaluated their evaporation performance. Under dark conditions, the evaporator without MTBs (i.e., bridge height: 0 cm) exhibited a natural evaporation rate of 0.15 kg/m2/h, which became more pronounced with the incorporation of MTBs due to the increased surface area (Fig. S5). Specifically, it linearly increased by ~0.04 kg/m2/h for every 1 cm increase in the MTB height. Under 1 sun illumination, the evaporation rate of the evaporator without MTBs was only 0.99 kg/m2/h because of the massive conductive heat dissipation to the bulk water (Fig. 3b, see the mass change curves in Fig. S6). The MTB usage considerably promoted solar evaporation. The evaporation rate increased to 1.58–1.73 kg/m2/h when the bridge height reached 2–5 cm (Fig. 3b). These values are even higher than the theoretical upper limit for solar evaporation (~1.44 kg/m2/h, Supplementary Note 1 and Fig. S7), which can be attributed to the natural evaporation contribution (Fig. S8). When the MTB height exceeded 3 cm, the evaporation rate increased by ~0.04 kg/m2/h for every 1 cm increase in MTB height (Fig. 3b), which was consistent with the result obtained under dark conditions. This consistency suggests that the 3 cm height is sufficient for the MTB structure to maximize solar evaporation (note that additional increase in MTB height only increases natural evaporation). To reveal the mechanism of this observation, we analyzed the heat transport in this unique architecture.Fig. 3: Evaporation performance, heat management, and stability evaluation.a Photograph of evaporators with various bridge heights (bridge number: 32). b Evaporation rate of evaporators with different MTB heights under 1 sun illumination (error bar type: standard deviation). c Internal temperature variation at different distances from the solar absorber. d Demonstration of the bulk water temperature after 3 h operation with different evaporators. e Photograph of the enclosed evaporator after 3 h evaporation. f Mass change curves and evaporation rate during the cycling experiment.Full size imageThe energy loss channels for this evaporation system primarily include conductive heat loss into the bulk water (({P}_{{cond}.})), radiative heat loss (({P}_{r{ad}.})), and convective heat loss to the environment (({P}_{{convec}.})). Therefore, the power flux available for evaporation (({P}_{{evp}})) can be described as follows16:$${P}_{{evp}}={P}_{{solar}}-{P}_{{cond}.}-{P}_{{rad}.}-{P}_{{convec}.}$$
(2)
where the solar energy input ({P}_{{solar}}={{{{{rm{alpha }}}}}}{C}_{{opt}}{q}_{i}); ({{{{{rm{alpha }}}}}}) is the light absorption coefficient; ({C}_{{opt}}) is the optical concentration; and ({q}_{i}) is the direct solar illumination. The conductive heat flux ({P}_{{cond}.}=k({T}_{{sa}}-{T}_{{bw}})/l), where (k) is the thermal conductivity; ({T}_{{sa}}) and ({T}_{{bw}}) are the temperatures of the solar absorber and the bulk water, respectively; and (l) is the heat conduction path referring to the MTB height in our model. The radiative heat flux ({P}_{{rad}.}=varepsilon {{{{{rm{sigma }}}}}}({{T}_{1}}^{4}-{{T}_{2}}^{4})), while the convective heat flux ({P}_{{convec}.}=hleft({T}_{1}-{T}_{2}right),) (varepsilon) is the optical emission, ({{{{{rm{sigma }}}}}}) is the Stefan–Boltzmann constant, (h) is the convection heat transfer coefficient, and ({T}_{1}) and ({T}_{2}) are the temperatures of the evaporator and environment, respectively.The energy loss caused by the heat transfer from the top surface to the bulk water (i.e., ({P}_{{cond}.})) can be minimized by increasing the MTB height (i.e., (l)) to confine the conductive heat within the MTB structure. This effect was visualized using infrared imaging to display the temperature gradients along MTBs with different heights. The results showed that the temperature at the bottom of the evaporator was similar to the ambient temperature when the MTB height reached or exceeded 3 cm (Fig. S9). This temperature distribution agreed well with the simulation modelled by COMSOL (Fig. S10). To obtain more insights into the heat transport in the architecture, we carefully recorded the internal temperature variation at different distances to the solar absorber under solar illumination. The results showed that the temperature stabilized after 60 min, when the internal temperature at 3 cm to the solar absorber was similar to that of the surrounding environment (Fig. 3c), indicating that the conductive heat was completely confined in the top 3 cm of the MTB structure. This confinement effect was also demonstrated by the temperature change of the bulk water (Fig. 3d): for the evaporator without MTBs, the bulk water temperature increased from ~21 to ~26.2 °C after a 3 h operation due to the continuous heat input (left panel); for the evaporator with 3-cm MTBs, however, the bulk water temperature was maintained at room temperature (~21.3 °C) (right panel), thus confirming the suppression of heat dissipation into the bulk water.Importantly, the confined heat energy can be exploited to generate additional vapour from the MTB surfaces, which can be efficiently released via the highly open interbridge spaces. To reveal this additional vapour generation from the vertical surfaces of MTBs, we used an evaporator having 32 MTBs (3 cm high) to perform a control experiment. In this experiment, the evaporator body was enclosed with an airtight polypropylene film, thus leaving only the upper surface exposed to the open space for vapour release (Fig. S11). After a 3 h operation, many water droplets condensed on the inner film surface, thus confirming that the MTBs released vapour (Fig. 3e). Compared to the completely open evaporator, the evaporation rate of the partially enclosed system decreased by ~31% (Fig. S12), demonstrating the importance of the open-channel design for enhanced interfacial evaporation.Furthermore, we performed a cycling experiment to evaluate the evaporator stability. In each cycle, the evaporator ran for 12 h under 1 sun illumination and in a dark environment for another 12 h to simulate day and night alternation. Figure 3f shows that during this long-term test (with 10 wt.% NaCl solution), the mass change of the NaCl solution in each cycle linearly evolved and the evaporation rate stabilized at ~1.44 kg/m2/h. No performance degradation was observed after a seven-day cycling experiment.Compared with the previously reported salt-rejection evaporators (evaporation rate: from 1.24 to 1.28 kg/m2/h for 10 wt.% NaCl solution)9,26,30, our evaporator demonstrated a higher evaporation rate under similar conditions due to the heat confinement effect and the natural evaporation contribution. However, high evaporation efficiency alone is not sufficient for water production applications. If the evaporated moisture is not collected, it can only be considered as a pollutant to the environment considering that it has the greatest greenhouse effect among various components in the atmosphere31. Water collection that is equally important as vapour generation has been largely ignored in many previous studies on salt-rejection evaporators.Therefore, we enclosed the evaporator with a transparent cover made of polymethyl methacrylate (PMMA) plates, creating a system that can produce water by condensing the evaporated moisture, and investigated the effects of bridge number and bridge height on the water production capacity of this system (Fig. S13a). When the bridge height was fixed at 3 cm, the amount of collected water increased with the number of bridges (Fig. S13b), which is consistent with the observation in the open system, confirming that the enhanced salt backflow facilitates water evaporation. When the bridge number was fixed at 32, the amount of collected water increased with the bridge height and reached the maximum at 3 cm, while further increasing the bridge height did not produce more water (Fig. S13c). This result is consistent with the conclusion above that 3 cm is sufficient to confine the conductive heat while further increasing bridge height only increases natural evaporation that does not contribute to water production. According to the three-hour test results, the water production rate of the enclosed evaporator in the optimal configuration (32 bridges; 3 cm high) is calculated to ∼0.68 kg/m2/h (Fig. S13).We also investigated the water generation performance of the enclosed system under different salinity conditions using NaCl solutions (3.5−20 wt.%). The results showed that the water production efficiency monotonically decreased from ~0.73 kg/m2/h for 3.5 wt.% NaCl solution to ~0.63 kg/m2/h for 20 wt.% NaCl solution (Fig. S14a). The relatively low water production efficiency associated with the high-salinity brines is mainly due to their low saturated vapour pressure, partly due to the decreased photothermic conversion efficiency caused by salt precipitation. For instance, when using brine containing 20 wt.% NaCl, salt precipitation emerged at the periphery of the evaporator after three hours of testing (Fig. S14b).Field testsAs per the recently announced “best practice for solar water production”32, the daily water yield is an important evaluation criterion that deserves additional consideration in practical implementations. Therefore, we prepared closed system based on the MTB structure and measured their water generation capacity under practical outdoor conditions.Rooftop experimentThe fabricated solar-driven water generation system has a 15 × 26 cm2 evaporator area (see Fig. S15). We first tested the system on the rooftop in KAUST, Thuwal, Saudi Arabia (Fig. S16). In this experiment, we employed the discharged water from an RO system of the KAUST Seawater Desalination Plant as the source water (salinity: ~8.7%). Our daily evaluation started at 8:00 and ended at 17:00. As shown in Fig. 4a, the evaporator surface was heated by solar light to a temperature 4–15 °C higher than the environment. However, the temperature at the bridge bottom was almost the same as the environment temperature, indicating that the conductive heat was confined, with only a small amount transferred to the bulk water. Consequently, saline water can be efficiently evaporated and condensed at the cover surface for the water collection. Figure 4b and Supplementary Movie 1 illustrate the relevant details. The total collected water was ~175 ml, of which ~110 ml flowed in the graduated cylinder, and ~65 ml was retained in the PMMA cover. Based on the evaporator area (390 cm2), the daily water productivity was calculated as ~5.0 L/m2. We measured the ion contentions of our water samples to evaluate the water quality. Compared with the discharged water from the RO plant, the ion concentration of condensed water was reduced by at least four orders of magnitude, thus fully meeting the WHO drinking water requirements (Fig. 4c).Fig. 4: Field tests.a Real-time temperature variation of the solar absorber, environment, bottom of bridges and bulk water, and solar flux from 8:00 to 17:00 on Apr. 11, 2022. b Timelapse photos of the collected water in the graduated cylinder from 8:00 to 17:00. c Ion concentration in the effluent water collected from the RO facility and collected freshwater from our system. d Daily water generation, solar insolation, and solar–water collection efficiency from Apr. 7 to 11, 2022. e Photograph of the evaporator after five-day operation. f Photograph of the floating system for the ocean test. g Schematic illustration of the structure of the floating system. h Daily water collection, solar insolation, and solar–water efficiency during the ocean test from Apr. 17 to 21, 2022.Full size imageWe calculated its practical solar–water collection efficiency of the system, ({eta }_{{prac}}), using Eq. (3):$${eta }_{{prac}}={m}_{{cond}}{h}_{{lv}}/left({A}_{{evp}}int {q}_{{solar}}left(tright){dt}right)$$
(3)
where ({m}_{{cond}}) is the daily water collection amount; ({h}_{{lv}}) is the total enthalpy of the liquid–vapour phase transition; ({A}_{{evp}}) is the evaporator area; and ({q}_{{solar}}) is the time-dependent solar flux. Benefiting from the highly efficient vapour generation, the overall solar–water collection efficiency of our system reached ~41.6%, representing a considerable improvement compared to the previously reported salt-rejection solar evaporation systems (e.g., maximum efficiency of a rooftop system: ~24%25). We performed a continuous test from Apr. 7 to Apr. 11, 2022 to evaluate the performance stability (Fig. 4d). The daily water collection rate fluctuated in the range of 4.7–5.2 L/m2 depending on the specific solar insolation of the day. The corresponding solar–water collection efficiency was 39%–42%. Remarkably, no salt accumulation was observed during this five-day outdoor operation (Fig. 4e). These results demonstrate the potential of the fabricated evaporator to extract freshwater from the wastewater discharged by RO plants.Floating testAfter the 5-day rooftop experiment, the same MTB-based evaporation system was tested in a floating configuration in the Red Sea (salt content: ~4.3%) to demonstrate its potential for practical seawater desalination (Fig. 4f, g). The test started and ended at 8:00 and 17:00, respectively, each day and lasted for five days from Apr. 17 to Apr. 21, 2022. As shown in Fig. 4h, the daily freshwater productivity ranged from 5.0 to 5.8 L/m2 with a stable solar–water collection efficiency of 42%–45%, which was consistent with the rooftop test. This freshwater productivity was approximately two times higher than the previous record of the salt-rejection solar evaporator (~2.5 L/m2 per day)25. The field test demonstrated a high-performance solar evaporator that will help in disaster relief or strengthen the resilience of individuals living on boats and coastal areas. More

Joachim Ayiiwe Abungba from the University of Science and Technology in Ghana and colleagues from India used a water evaluation and planning model to estimate runoff, streamflow and future water demand under different climate scenarios. They reveal that between 1990 and 2019, human settlements, open savannah woodland, croplands and waterbodies increased, while closed savannah woodlands, wetlands and grasslands decreased. The model shows increased water availability from river discharge compared with the current scenario. However, the uncertainties in future changes exceeded the predicted increases. There is an urgency to further improve model certainty and to develop integrated water management in order to ensure sustainable livelihoods for local people. More