Anticipating and monitoring water risks for agriculture

The landscape of water risks for agriculture

“Too much”, “too little” or “too polluted” water are often seen as the trifecta of water risks facing human systems, including agriculture. A more comprehensive view also considers systemic water risks in addition to these three quantifiable water-related pressures (Figure 1). The OECD Recommendation on Water (OECD, 2016[3]) recognises the “risk of undermining the resilience of water-related ecosystems” as a water-related risk, defined as exceeding the coping capacity of the surface and groundwater bodies and their interactions, possibly crossing tipping points, and causing irreversible damage to the system’s hydraulic and biological functions (OECD, 2013[4]). The disruption to freshwater ecosystems may be triggered by various water risks (e.g. water pollution) and, in turn, amplify them. Recent findings by the Global Commission on the Economics of Water call attention to an additional systemic risk: pressures from human actions are driving the global hydrological cycle out of balance, rendering supplies of freshwater less reliable (GCEW, 2024[1]).

This section defines water risks and summarises the latest evidence on water risks facing agricultural production, including the often-overlooked role of “green water” in water security. It also considers the issue of “imported water risk”, outlining how a country may be exposed to water risks beyond its borders via trade of agricultural commodities.

Figure 1. Water-related risks for agriculture

“Too much”: The risk of excess water for agricultural production

Water excess hazards include sudden-onset events such as floods or intense rainfall episodes, as well as less extreme phenomena that can nevertheless affect agricultural production, such as wetter average conditions and waterlogging. Longer-term shifts in rainfall patterns may lead to wetter average conditions that are sub-optimal for crops currently under cultivation. Excess rainfall can result in waterlogged soils that affect crop growth. Globally, flooding affects over 17 million km² of land surface annually, with around 10% of agricultural land affected by waterlogging or severe soil drainage constraints (Kaur et al., 2020[5]).

Considerable attention is paid to the impacts of water shortages for agriculture, yet too much water can pose as great a risk (OECD, 2016[6]). Zampeiri et al. (2017[7]) found that wheat production is more sensitive to water excess than to drought in several countries, while Li et al. (2019[8]) found that excessive rainfall in conjunction with poorly drained soils leads to maize yield loss of a comparable magnitude to extreme drought in the United States, especially in cooler areas. Even rice, a semi-aquatic crop that benefits from controlled inundation, can be vulnerable to flood events. Li et al. (2025[9]) assessed the causal impact of rice crop submergence on yield losses from 1980 to 2015 and found that globally, these floods reduced annual rice yield by 4.3% while the East Coast of the People’s Republic of China (hereafter “China”) experienced losses of 14%. In the livestock sector, flooding can result in injury or death of animals, nutritional restriction and disease spread, with specific impacts varying by type of livestock (Crist, Mori and Lee Smith, 2020[10]).

The impact of excess water hazards on agriculture depends on several factors including timing, place and crop type, as well as water management infrastructure and practices (e.g. tile drainage, flow management) that have been incorporated into the production system. Crops are more vulnerable to excess water during certain growth stages (Schmitt et al., 2022[11]; Li et al., 2019[8]). Climate, soil type and topography as well as water management infrastructure can mediate the impacts of excess water. For example, Li et al (2019[8]) found that excessive rainfall can either decrease or increase maize yields depending on whether the crop was grown in warmer or cooler climes and whether soils have poor drainage.

“Too little”: Water shortages

Water scarcity occurs when demand exceeds available supply, a relative and dynamic concept that varies based on local conditions and across time (OECD, 2025[12]), (FAO, 2012[13]). Definitions can encompass both physical factors (i.e. the physical availability of water in sufficient quantity and quality to meet demand), as well as broader socio-economic considerations (e.g. when human, institutional, infrastructural or financial capital limit access to water, even when it would be available to meet needs) (IPCC, 2022[14]). For example, high energy prices can make the cost of groundwater pumping prohibitive, even if water is physically available. Finally, water scarcity may be seasonal (e.g. in monsoon areas) or chronic (e.g. in arid regions) and can also be driven by pollution if water quality is too poor for use in agriculture.

Droughts can be an important underlying cause of water scarcity in agriculture, affecting the supply-side. They refer to periods of significant hydrological imbalance affecting various components of the water cycle, characterised by abnormal water deficit on different timescales. Dependent on their exogenous drivers, droughts can be classified as i) meteorological (a prolonged period of low precipitation), ii) agricultural or ecological (conditions when soil moisture levels fail to meet crop and vegetation needs) and iii) hydrological (prolonged periods with surface or groundwater water levels dropping below average) (OECD, 2025[12]). In addition, anthropogenic drought can be defined as drought events caused or intensified by human activities, considering the interactions of anthropogenic actions with natural systems (AghaKouchak et al., 2021[15]). Different types of droughts are often interlinked and can persist simultaneously (Liu et al., 2016[16]). Around 11% of rainfed croplands and 14% of pasturelands face frequent droughts and over 60% of irrigated cropland faces high water stress (FAO, 2020[17]).

The impacts of drought on agriculture are substantial. Agriculture bears the largest share of drought-caused economic costs across all sectors (Tikoudis, Gabriel and Oueslati, 2025[18]) and drought is the most impactful natural disaster for agriculture (OECD/FAO, 2021[19]). A metanalysis by Zhang et al (2018[20]) found that even mild drought stress intensity reduced average wheat and rice yields by 21% and 17% respectively. In 2021, drought conditions in California caused economic costs of USD 1.1 billion for the agricultural sector (Escriva-Bou et al., 2022[21]). FAO research shows that drought causes more harm to agricultural production than any other natural disaster in least developed and lower middle-income countries (FAO, 2021[22]).1

“Too polluted”: Inadequate water quality

Poor water quality limits agricultural productivity, compromises food security and undermines rural livelihoods. Poor-quality water, whether due to salinity, nitrate pollution, toxic elements, chemical imbalances or microbial contamination, can reduce crop yields, degrade soils, harm livestock and pose risks to human health. For example, Pica et al (2017[23]) found that switchgrass and rapeseed suffered biomass losses of 25‑90% when irrigated with water containing high levels of dissolved solids and organic carbon. Intensive horticulture faces additional stresses from contaminant mixtures. Closed irrigation loops in high-density nurseries accumulate nutrients, pesticides, microplastics and pathogens; extreme rainfall can flush these pollutants beyond facility boundaries, suppressing seedling growth and driving up treatment costs (Gomes et al., 2025[24]). Clean, reliable stock water is also critical to animal health and performance, but high nitrate concentrations pose significant risks to livestock health (FAO & IWMI, 2023[25]).

At broader scales, water quality constraints intensify water scarcity. Globally, adding water quality constraints to classic scarcity metrics reveals that river basins classified as water scarce could triple by 2050, threatening irrigation across 40 million km² and exposing an extra three billion people (Wang, 2024[26]).

Damaged resilience of freshwater ecosystems: Loss of ecosystem services

Freshwater ecosystems2 are vital natural assets for agriculture. They provide the water that sustains crops and livestock, regulate flows and recharge groundwater, filter pollutants to maintain water quality, and protect farmland from floods and droughts. Wetlands, rivers and lakes also deliver nutrient-rich sediments, support pollinators and pest predators, and maintain the biodiversity that underpins agricultural productivity.

Degraded freshwater ecosystems undermine the ecological foundations of agriculture. The risk of undermining the resilience of freshwater systems includes exceeding the coping capacity of the surface and groundwater bodies and their interactions, possibly crossing tipping points, and causing irreversible damage to the system’s hydraulic and biological functions (OECD, 2013[4]). Fluet-Chouindard et al (2023[27]) estimate that 3.4 million km² of inland wetlands have been lost since 1700, a net loss of 21%. They find that wetland loss has been concentrated in Europe, the United States and China and rapidly expanded during the mid-twentieth century. Freshwater biodiversity gains achieved since the 1970s have plateaued since 2010, with recovery weakest downstream of cropland, dams and urban areas (Haase et al., 2023[28]). Without decisive mitigation, the combined pressures of nutrient enrichment, higher temperatures and ongoing habitat conversion will further degrade freshwater ecosystems, threaten water security and undermine the resilience of agricultural production systems.

A destabilised hydrological cycle: Increasing hydro-climatic extremes

Rainfall is becoming more variable and extreme. Rising global temperatures combined with land-use changes are destabilising and accelerating the hydrological cycle, shifting precipitation patterns and undermining assumptions of a stable year-to-year supply of freshwater (GCEW, 2024[1]). The Global Commission on the Economics of Water call attention to this systemic risk, echoing similar warnings from the World Meteorological Organization that the hydrological cycle is becoming more erratic and unpredictable (WMO, 2023[29]; WMO, 2024[30]). The IPCC confirmed this trend of increasing hydro-climatic extremes (IPCC, 2022[14]).

A destabilised hydrological cycle may manifest locally as water excess or water shortages – two of the water risks identified above – yet the risk is qualitatively distinct. The risk concerns pushing year-to-year supplies of freshwater outside of the bounds of natural variability known for the last 12 000 years and under which existing agricultural and water management systems have developed. Part of the challenge for agriculture is coping with the increased variability of water availability and increased intensity of water-related hazards. Increased hydrologic variability can overwhelm human systems that were designed to manage a certain frequency or intensity of hazard event but may struggle as those events become more frequent or intense. The increased uncertainty means that farmers cannot necessarily depend on past patterns to plan planting or irrigation. It also raises questions about the reliability of risk assessment tools that draw solely on historical data.

Globally, average annual precipitation in the last two decades exceeded the average recorded between 1950‑2000. That said, several regions, such as the Mediterranean, most of Africa, the Middle East, eastern Australia and the western parts of North America and parts of South America experienced up to a 20% reduction in average annual precipitation over the same reference periods (OECD, 2025[12]). Interannual precipitation patterns have also changed, with regional seasonal increases and decreases affecting the seasonal availability of water critical for agriculture. Summer precipitation decreased by over 20% with each degree of observed global warming in parts of southeastern and western Australia, southeastern South America and parts of eastern Africa, while it increased by around 20% in northern Europe (IPCC, 2022[14]).

Water risks are dynamic and interlinked

Water risks to agriculture result from dynamic interactions between water-related hazards and the exposure and vulnerability of agricultural systems to those hazards. This framing is in line with the IPCC concept of risk (Reisinger et al., 2020[31]),3 defined as the “potential for adverse consequences to human or ecological systems” and where risks result from the dynamic interactions between hazards, exposure and vulnerability. Consequently, this paper considers water risks as the potential for adverse consequences for agriculture, also noting the vital importance of ecological systems for productive and sustainable agricultural systems, given the myriad ecosystem services they provide.

Agriculture faces water risks across different timescales and a range of tools to anticipate and monitor risks adapted to these different temporal scales are therefore needed. Rapid-onset hazards include flooding or intense rainfall episodes, where accurate weather forecasting and early warning systems play an important role in risk mitigation. Other hazards such as droughts can develop more slowly, affecting water availability over one or many growing seasons. Over a longer time horizon, agriculture may face water-related risks driven by a more variable climate or other megatrends such as population growth and urbanisation. Water risks to agriculture also vary across space. Regionally specific risks and variability across landscapes with different geographies, climates and production systems necessitate the need for locally adapted risk assessment and responses.

Drivers of water risk

Water risks in agriculture are driven by several factors. Agriculture is not only highly vulnerable to water-related risks but also has the potential to be a major contributor to them, as it places significant pressure on freshwater resources, i.e. part of the risk agriculture faces is endogenous. Furthermore, drivers range from the local level to regional and global scales. For risk managers, this implies being aware of sources of water risk that may be beyond the basin or catchment level – such as land cover change in distant places upwind – and having the right tools to monitor and assess that risk.

Climate-related drivers of water risks

At a global level, rising temperatures alter the hydrological cycle. Higher temperatures drive more evaporation and increase the atmosphere’s capacity to hold water: each 1°C rise in global temperatures increases water vapour in the atmosphere by 7% (NASA, 2022[32]). The increased energy in the system leads to more frequent and severe extreme weather events, including intense rainfall, floods and drought. Furthermore, increasing temperatures lead to glacier melt, affecting freshwater supplies and increasing flood risk in the short-term (The GlaMBIE Team, 2025[33]). More broadly, increasing rainfall variability year-to-year is altering the distribution and amount of water available and our ability to rebalance this disturbance via storage.

Looking forward, projections undertaken in the context of climate change analysis indicate increasing magnitude and frequency of water-related hazards, with regional variations (IPCC, 2022[14]). Under these projections, many regions are expected to experience increasingly volatile precipitation regimes characterised by “both longer dry spells and heavier events when precipitation does occur”. Extreme agricultural droughts are also projected to become 150‑200% more likely in western China, the Mediterranean and the northern parts of South America, North America and Eurasia under 2°C of warming. The magnitude and frequency of floods are projected to increase in Asia, central Africa, western Europe, Central and South America and eastern North America, and decrease in northern North America, southern South America, the Mediterranean and eastern Europe. The picture for annual mean precipitation is less clear. At 1.5°C warming, climate models consistently project increased precipitation in the central and eastern Sahel, south-central Asia, parts of Greenland and Antarctica, and the far northern regions of North America and Asia, but projections for other regions vary substantially.

Rising air and water temperatures, altered precipitation regimes and changing sea levels directly modify the dilution, transport and transformation of pollutants, affecting water quality.

Land cover, land management and landscape level changes

Land cover changes and land management practices can disrupt the hydrological cycle by altering how water is stored, evaporated and transpired, giving rise to diverse water risks for agriculture. A farm’s water security can be shaped by large-scale land cover changes occurring far beyond its immediate surroundings, as well as by on-farm land management practices that farmers directly control.

The feedback between land cover and rainfall has been under appreciated. The loss of natural ecosystems – particularly deforestation and loss of wetlands – disrupts the moisture that transfers from the land to the atmosphere4 (i.e. green water flows) that then falls as rain downwind, a process referred to as “terrestrial moisture recycling”. Given that 40‑70% of terrestrial rainfall originates from land, land cover change in one place can influence water security in another place, sometimes thousands of kilometres away (GCEW, 2024[1]). Precipitationsheds (the area supplying evaporation to a downwind region’s rainfall) and evaporationsheds (the downwind region where evaporation from upwind areas falls as rain) can be mapped to show the connections between upwind and downwind locations through these atmospheric rivers. Smith, Baker and Spracklen (2023[34]) found that even a 1% loss in tropical forest area can reduce precipitation by up to 0.25 mm per month in a 200 km radius of the deforested area, in addition to negatively affecting water balances locally. Several sectors are responsible for land cover change, but agriculture has altered the Earth’s land area more than any other anthropogenic activity (UNCDD, 2022[35]).

As well as the impact on rainfall, land cover change can exacerbate other water risks. Looking at the case of Northeast China, Ma et al (2024[36]) show that conversion of forests and peatlands to agricultural area reduced the water retention capacity of the landscape, leading to greater flood risk. The loss of freshwater ecosystems such as wetlands also deprives agriculture of the ecosystem services they provide.

Agricultural land management practices can impact water risks. Heavy machinery and deep ploughing can compact soils, inhibiting infiltration and water retention, and thereby reducing groundwater and soil moisture availability for crops (Timár, Jakab and Székely, 2024[37]). Monocropping can reduce the soil water-holding capacity, while bare or fallow fields or overgrazing can also lead to soil moisture loss. Soil compaction can also have a major effect on flood peaks after high intensity, short duration storms in small catchments (Alaoui et al., 2018[38]). Erosion and soil degradation can lead to sediment deposits that impact the ecological functioning of freshwater ecosystems (Ramsar, 2021[39]).

Certain agricultural and land management practices, on the other hand, may reduce some water-related risks. In Japan’s Suse region, paddy field dams (rice paddy fields combined with runoff control devices) were found to have reduced peak discharge by up to 48% during typhoon events (Kobayashi et al., 2021[40]).

Finally, landscape-level changes such as river regulation influence water risks. While river regulation may reduce flood risk and free up land for agriculture and other uses, it may also cause water shortages. As documented in Hungary, meander cut offs reduce water storage in the landscape and shorter rivers limit the time for water infiltration and groundwater recharge (Timár, Jakab and Székely, 2024[37]). Land drainage and diverting rivers for agriculture also impact wetland functioning.

On the other hand, certain landscape management, such as wetland restoration or gully woodland and riparian buffer zone planting can reduce flood risk (OECD, 2016[6]). Several countries are exploring the effects of rewilding of agricultural lands. As found in the Trollberget Experimental Area in Sweden, the rewetting of peatlands previously drained for agriculture and forestry can enhance hydrological balance by reducing peak flows and elevating low flows (Karimi et al., 2024[41]; Karimi et al., 2025[42]). The impact of peatland rewetting on reducing flood and drought risk, however, remains highly context-dependent (Elenius et al., 2025[43]).

Growing water demand and over-abstraction

The risk of water shortages for agricultural production may in part be driven by rising water demand and competition with other sectors. Population growth, economic development and changing consumption patterns have contributed to rising water abstraction, with water demand increasing by around 1‑1.8% annually over the past century (Boretti and Rosa, 2019[44]). Water demand worldwide is projected to grow by 20‑30% by 2050, compared to 2010. Domestic and industrial demand are projected to increase faster than agricultural demand, with water needs in the manufacturing sector projected to rise by 400% by mid-century. In addition to the water demands from conventional economic sectors, emerging sectors such as hydrogen production and data centres may create new challenges for water resource management.

Nevertheless, agriculture is projected to remain the major water user. Agriculture is the primary driver of water abstraction, responsible for around 70% of freshwater withdrawals worldwide. Globally, irrigated areas expanded by 11% between 2000‑15, with more than half of this expansion occurring in already water-stressed areas (Mehta et al., 2024[45]). In certain cases, irrigation expansion exceeded water recharge rates. For example, the High Plains region of the United States experienced severe decreases in groundwater levels, with groundwater irrigation exceeding aquifer recharge rates up to 10 times (Scanlon et al., 2023[46]). The water intensity of chosen crops shape agriculture’s demand on green water resources too. For example, in northern China, the expansion of agricultural area, fertiliser application and disproportionate increase in the cultivation of water intensive maize contributed to reduce soil moisture by about 29% per decade between 1983‑2012 (Liu et al., 2015[47]).

Excessive water consumption reduces the amount of water available for nature and negatively affects freshwater ecosystems. Around 30% of irrigated crop production is estimated to affect environmental flow requirements and biodiversity (Jägermeyr, 2017[48]).

Some countries have turned to non-conventional water sources for irrigation such as desalinisation and wastewater reuse as a strategy to reduce pressure on freshwater sources. This can be a way to improve the environmental sustainability of irrigation, increase water supply to agriculture and address drought. For example, Israel and Spain have invested significantly in modernising irrigation systems and increasing the use of non-conventional water sources for agriculture.

Salinisation, nitrate and other diverse pollution sources

Drivers of poor water quality and freshwater ecosystem degradation include salinisation and pollution from industry, urbanisation and agriculture.

Globally, secondary salinisation5 already affects over 1 billion ha across more than 100 countries, reducing yields of major food crops by up to 50 % (Singh, 2022[49]). At the basin scale, salinity export from irrigated lands is now the dominant water‑quality stressor. A global assessment of 458 river basins showed that 57 % exhibit significant upward salinity trends, with irrigation‑related salt fluxes explaining over 80 % of the observed increases; basin aridity index and irrigation intensity emerged as the best predictors of trend magnitude (Thorslund et al., 2021[50]). Likewise, projections for coastal aquifers indicate 10‑25% increases in salinity by 2050 as sea‑level rise and groundwater over‑extraction promote seawater intrusion (Khondoker et al., 2023[51]).

Nitrate pollution significantly degrades water quality in both groundwater and surface water systems. Groundwater is particularly vulnerable due to the high mobility of nitrate in soils, resulting in contamination of aquifers and drinking-water supplies. Prolonged exposure can adversely affect human health (Ward et al., 2018[52]). In some localities, elevated nitrate inputs from agriculture, municipal and industrial discharge promote eutrophication in rivers, lakes and marine waters, driving excessive algal growth, oxygen depletion, the proliferation of invasive species and declines in aquatic biodiversity (Bijay-Singh and Craswell, 2021[53]; Akinnawo, 2023[54]). Eutrophication is now widely recognised as a persistent, system level problem rather than an occasional episode, especially in shallow, low volume lakes that receive heavy external nitrogen and phosphorus loads from cropland and settlements (Zhou et al., 2022[55]).

Intensive agricultural use of synthetic pesticides poses some of the most serious diffuse threats to fresh‑ and coastal‑water quality. At regional and national scales, the chemical footprint of pesticide use is growing faster than that of nutrients (Yi, Gerbens-Leenes and Aldaya, 2024[56]). Certain synthetic pesticides are persistent, bio‑accumulative and toxic to humans and wildlife (Kaur et al., 2024[57]). Past agricultural pesticide use included several substances now classified as Persistent Organic Pollutants (POPs), such as aldrin, chlordane, dieldrin, and heptachlor. Ecological consequences are also increasingly well quantified. Even in remote, non‑agricultural headwater streams, atmospheric deposition of insecticides is sufficient to alter ecosystem community structure (Schweiger et al., 2025[58]).

Additional threats to water quality include other legacy POPs, such as PCBs and dioxins, heavy metals as well as ultra-persistent emerging contaminants that are spreading rapidly. While the effect on agricultural production is not yet clear, they bring with them public health risks. Agricultural soils also hold 23 times more micro- and nanoplastics than the oceans. These particles can be taken up by crops like lettuce, wheat and carrots, and have been shown to reduce the water-holding capacity of sandy loam soils by up to 12%, worsening drought stress (Boctor et al., 2025[59]).

Imported water risk: Exposure via international trade

As well as water risks affecting domestic agricultural production, countries may be exposed to agriculture-related water risks originating beyond their borders via their imports of agricultural commodities. The interconnected nature of global markets means that water risks in one locality can have impacts through trade channels, with potential consequences for food security. This can be through a short-term shock, such as a flood or drought, or longer-term phenomena. Different tools are needed to anticipate and assess water risks operating on these different timescales.

Awareness and assessment of this “imported water risk” can have implications for countries’ trade strategies (i.e. ensuring sufficiently diverse sources of supply) and decisions around domestic production capacities (i.e. ensuring domestic production of goods no longer accessible on international markets), although this needs to take account of the water risks posed by domestic production. Indeed, trade can be a mechanism to alleviate water risks, as the ability to access agricultural goods on international markets reduces vulnerability to food insecurity and can relieve pressure on local water demand (OECD, 2017[60]).

Acute shocks

Production shocks can lead to food shortages, price increases and increasing food insecurity in producer and importing countries alike. Following the 2023 drought in Brazil, global coffee prices were 55% higher in August 2024 compared to the previous year. Similarly, a prolonged drought combined with a heatwave affecting cocoa production in Ghana and Ivory Coast tripled year-on-year global cocoa prices in April 2024 (Kotz et al., 2025[61]). Market responses to agriculture production shocks depend on multiple market variables and prices do not always respond to changes in yields (OECD, 2017[60]). Nevertheless, in the case of large production shocks in a larger market for a product with relative inelastic demand and no immediate substitute, relative prices will rise (OECD, 2017[60]).

Chronic water risk exposure

As well as short-term shocks with immediate disruptions to supply chains, slower developing phenomena may over time increase the exposure of importing countries to water risks. For example, countries may become increasingly exposed to water risks if they rely on agricultural commodities from localities with unsustainable water use, such as the over-abstraction of groundwater for irrigated agriculture, or from agricultural regions projected to become more drought prone. Ercin, Chico and Chapagain (2019[62]) estimate that more than 44% of the EU agricultural imports from the global market under current trade patterns will become highly vulnerable to drought in the future. To understand their vulnerability and exposure to disruptions in agricultural production, including water related risks, the EU has assessed the impacts of climate change on the supply of agricultural commodities to Europe (EEA, 2021[63]).

The concepts of virtual water trade and water footprint (Box 1) have been applied to assess imported water risk. While many early studies using this approach calculated simply the volume of water embedded in a good, more recent research refines the concept to consider the sustainability of water resources. Hoekstra and Mekonnen (2016[64]) quantify and map the United Kingdom’s direct and indirect water needs to assess the “imported water risk” by evaluating the sustainability of the water consumption in the source regions. They find that 49% of the United Kingdom’s global blue water footprint – the direct and indirect consumption of ground and surface water resources behind all commodities consumed in the United Kingdom – is in places where the blue water footprint exceeds the maximum sustainable level.

That said, countries’ own domestic management of water risks will impact the extent to which they could pose water-related risks to others as exporters. This is particularly the case given that most countries are not simply exporters or importers in global agriculture and food markets, but play both roles (being exporters of some products and importers of others).

Trade can be a mechanism to alleviate water risks

At the same time, international trade can be a key mechanism to alleviate water risks. As noted by Muller and Bellmann (2016[65]), empirical studies have shown that “trade has helped maintain food security in countries where water availability limits domestic production and plays an important role in buffering the impact […] of floods or droughts”. Similarly, Allan (2001[66]) argues that imports of water-intensive goods contributed to solving water shortages in the Middle East.

International trade is a mechanism by which countries can ensure the supply of goods that cannot be produced domestically (as, for example, can be the case for products that only grow in certain climatic conditions, such as tropical products) or where the quality, price or environmental trade-offs of domestic production are too high. Trade also allows countries to adjust to domestic supply shocks and maintain food security. Although there may be an initial shortage after a water-related event such as a flood, importing countries (and the affected producer country) can source goods from other producer countries. The degree of trade openness and purchasing power are two important factors that determine the extent to which countries can turn to trade to buffer the impacts of domestic supply shocks. Well-functioning food markets play a key role in diluting price effects and ensuring that when a problem arises it does not have a ripple effect (OECD, 2017[60]). Furthermore, to address risks to food security, including water-related risks, several countries such as Singapore and the United Kingdom set out to strategically diversify their import sources, alongside ramping up domestic production (Teng, 2020[67]); (Defra, 2025[68]). Open trade and diversified sourcing can be elements of country strategies to enhance resilience, lower prices and support food security.

Box 1. Virtual water trade

The relationship between water and trade has been explored through the concepts of virtual water trade (VWT) and the water footprint. Allan (1997[69]) introduced the concept of virtual water as the volume of water needed to produce a certain good (or “water footprint”), and VWT as the flows of virtual water resulting from the trade in these goods. The term virtual is used to indicate that the water is not physically present in the traded commodity but is embedded in its production (D’Odorico et al., 2019[70]). Agricultural, food and textile products account for about 66% of total global virtual water trade. Between 1986 and 2022, the global volume of VWT related to agricultural products almost tripled (Mekonnen  et al., 2024[71]).

The usefulness of these concepts for policy purposes has been questioned, however. In particular, the lack of information on the economic implications of water use – notably the opportunity cost or the scarcity value of the input – limits the use as a guide for water resource strategy (Wichelns, 2015[72]). Furthermore, two products can have the same water footprint but very different environmental impacts depending on the local context, i.e. a litre of water use in a water abundant country does not have the same environmental implications as a litre of water use in an arid country (Berger et al., 2021[73]).

Nevertheless, the concepts have been used to observe the interplay between trade and water resources. Empirical studies have shown that the distribution of the water content of consumption across countries was not improved via trade (Afkhami et al., 2020[74]). At the same time, Carleton, Crews and Nath (2024[75]) note that differences in relative agricultural productivity and relative arable land endowments “can cause virtual water to flow from [water] scarce regions to abundant ones.”

Notes

← 1.  The seven types of natural disasters that adversely impacted agricultural production systems were drought, followed by floods, storms, earthquakes/landslides/mass movements, plant and animal pests/diseases, extreme temperatures and wildfires.

← 2.  Freshwater ecosystems encompass wetlands, river, lake and groundwater systems as well as forests. Water-related ecosystem services from trees and forests to agriculture include freshwater supply, regulation of water flows, local climates and temperatures, supporting an intact and reliable hydrological cycle, water damage mitigation and water purification (FAO, SEI, CI & TNC, 2025[192]).

← 3.  Developed to guide climate risk assessment, the IPCC framing is well adapted to considering water-related risks given that many climate hazards are water hazards, such as floods and drought.

← 4.  Water is transferred from the land to the atmosphere via evaporation from the soil and other surfaces and transpiration by plants, together known as evapotranspiration.

← 5.  Secondary salinisation is salinity that results from human activities (e.g. farming, irrigation) as opposed to naturally occurring salinity.