Future global urban water scarcity and potential solutions

The share of the global population living in cities has grown rapidly and is projected to reach 6.7 billion (68.4%) by 2050. Water scarcity—where human demand exceeds available supply—already affects many urban populations and has profound consequences for health, environmental quality, and socio-economic development. Industrial and domestic water demand are expected to rise by 50–80% over the next three decades due to population growth, urbanization, and socio-economic development, while climate change is altering the magnitude, timing, and spatial distribution of water availability. These dynamics threaten progress toward SDG6 (Clean Water and Sanitation) and SDG11 (Sustainable Cities and Communities). Traditional engineering responses—reservoirs, desalination, and inter-basin transfers—can help but are costly, resource-intensive, geographically constrained, and pose environmental risks. Significant uncertainty persists in the extent of current and future urban water scarcity because prior assessments often focused on subsets of cities, considered availability but not withdrawals, or lacked forward-looking analyses. This study aims to provide a comprehensive, catchment-based global assessment of urban water scarcity for 2016 and 2050 under multiple socio-economic and climate scenarios, and to evaluate the feasibility of major solution pathways for water-scarce cities worldwide.

Previous global assessments of urban water scarcity have important limitations: many analyzed only large cities or specific regions, some considered only water availability (omitting withdrawals), and many lacked future scenario analysis. Resulting estimates vary widely—from roughly 0.2 to 1.0 billion urban residents affected around 2000 and 0.5 to 4.0 billion by 2050—reflecting methodological and scope differences. Studies have discussed solutions such as groundwater exploitation, seawater desalination, reservoir expansion, inter-basin transfers, improving water-use efficiency, and managing urban landscapes. However, few have evaluated these solutions’ global feasibility city-by-city, considering severity of scarcity, geography and hydrogeology, socio-economic conditions, infrastructure constraints, and environmental carrying capacity. Addressing these gaps requires a comprehensive assessment of both exposure and solution feasibility to guide effective and sustainable investments in urban water security.

Overall design: The study quantified global urban water scarcity in 2016 and projected 2050 under four SSP–RCP combinations, then evaluated the feasibility of major solution options for water-scarce cities. The analysis proceeded in four steps: (1) reconstruct spatially explicit global urban population, (2) identify water-scarce catchments using a catchment-based water stress index (WSI), (3) overlay urban population with water-scarce areas to estimate exposed populations at global and national scales, and (4) assess feasibility of seven solution categories for large cities (>1 million residents). Spatial urban population: Global urban population was mapped at 1 km resolution by integrating HYDE population data with higher-resolution urban extent and density, using downscaling and the LUSD-urban model to simulate urban expansion for 2016 and 2050. Urban polygons were generated from raster layers to delineate individual cities and aggregate population within city boundaries. Water scarcity identification: Annual and monthly WSI were computed at the catchment scale following a water-withdrawal relative to availability framework. WSI_{t,i} = TWW_{t,i} / CWW_{t,i}, where TWW sums withdrawals for irrigation, livestock, industrial, and domestic sectors, and water availability accounts for local renewable resources (runoff), net inter-basin transfers, and upstream contributions. Upstream contributions were calculated as the excess of upstream availability over upstream consumption, and inter-basin transfers were incorporated using City Water Map data. Sectoral water withdrawals drew on AQUEDUCT 3.0 and related datasets; livestock withdrawals were held constant where data were lacking. Multi-year average runoff for 2005–2014 (baseline) and 2011–2050 (future) was derived from an ensemble of 10 CMIP6 GCMs. Water-scarce areas were identified using a WSI threshold of 1.0 to represent extreme water stress; annual and monthly metrics distinguished perennial from seasonal scarcity. Current (circa-2016) scarcity used 2014 WSI overlain with 2016 urban population; future scarcity used 2050 WSI and projected urban population under four scenarios (SSP1–RCP2.6, SSP2–RCP4.5, SSP3–RCP7.0, SSP5–RCP8.5). Exposure estimation: Urban populations within perennial and seasonal water-scarce catchments were summed globally and by country. Large cities (>1 million in 2016, per UN World Urbanization Prospects) were classified by scarcity type. Uncertainty and contribution analyses: Uncertainty across GCMs was quantified by repeating the WSI and exposure calculations with each model’s runoff and summarizing the range. A contribution analysis separated the effects of socio-economic factors (population growth, urbanization, and sectoral water demand changes) from climate-driven availability changes on the change in exposed urban population between 2016 and 2050. Solution feasibility assessment: Seven solution categories were evaluated for water-scarce large cities: seawater desalination, groundwater exploitation, reservoir construction, inter-basin water transfer, international virtual water trade, domestic virtual water trade, and systemic measures represented by the SSP1–RCP2.6 pathway (water-use efficiency improvements, slowed population growth, and climate mitigation). Feasibility was determined by combining geographic and hydrogeologic conditions, economic development levels, coastal access, proximity to donor basins, and environmental constraints, summarizing the proportion of cities for which each measure could alleviate scarcity.

- In 2016, 933 million urban residents (32.5% of the global urban population) lived in water-scarce regions: 359 million (12.5%) faced perennial scarcity and 573 million (19.9%) faced seasonal scarcity. India (212 million) and China (159 million) had the largest water-scarce urban populations. Among large cities (>1 million), numerous cities already experienced scarcity; examples with perennial scarcity include Los Angeles, Mosul, Lahore, Delhi, Bangalore, and Beijing; seasonal examples include Mexico City, Istanbul, and Karachi. - By 2050, the urban population facing water scarcity is projected to reach 1.693–2.373 billion people (about one third to nearly half of the global urban population), with an ensemble estimate around 2.0 billion. Of these, roughly 476–595 million would face perennial scarcity and 0.902–1.647 billion seasonal scarcity. India’s water-scarce urban population is projected to grow from about 222 million (2016) to 570–644 million (2050), constituting about 19–33% of the global water-scarce urban total. - Nearly half of the world’s large cities are projected to be water-scarce by 2050 under at least one scenario. The number of large cities exposed increases to roughly 292, including 19 megacities, with 10 additional megacities becoming water-scarce (Cairo, Dhaka, Jakarta, Lima, Manila, Mumbai, New York, Sao Paulo, Shanghai, and Tianjin). - Drivers: Growth in urban population and water demand dominates the increase in scarcity exposure, adding 0.909–1.135 billion people (about 80–91% of the total increase). Climate change impacts on availability add about 52 million people on average (range −72 to +229 million), about 5% of the increase; under SSP2–RCP4.5 climate adds ~51 million by 2050, while under SSP1–RCP2.6 this addition drops to ~16 million. - Solution feasibility: 276 large cities (94.5%), including 17 megacities (89.5%), could alleviate scarcity via one or more measures. Sixteen large cities, including megacities Delhi and Lahore, have severely limited feasible options due to geographic and economic constraints. - Effectiveness by option (share of large cities potentially alleviated): domestic virtual water trade 71.2% (208 cities, including 14 megacities); inter-basin transfer 68.5% (200; 14 megacities); groundwater exploitation 65.8% (192; 11 megacities); international transfer/trade 65.1% (190; 10 megacities); reservoir construction 51.7% (151; 10 megacities); seawater desalination 50.0% (146; 12 megacities). Additionally, 68 large cities (23.3%), including New York, Sao Paulo, Mumbai, Dhaka, and Jakarta, could mitigate scarcity via the SSP1–RCP2.6 pathway (efficiency, slowed population growth, and climate mitigation).

The study addresses the key question of how many urban residents and cities face water scarcity now and under future socio-economic and climate scenarios, and which solution pathways are feasible. Findings show that urban water scarcity is already widespread and will intensify markedly by 2050, driven primarily by urban population growth and rising water demand, with climate change compounding risks to a lesser but uncertain extent. The expansion of scarcity to many large and megacities underscores threats to SDG6 and SDG11 and highlights systemic dependencies across sectors and regions, including the implications of virtual water trade. While most water-scarce cities have at least one feasible technical or demand-side option, each carries significant environmental and socio-economic trade-offs: reservoirs and transfers can harm river ecosystems and alter regional hydrology; desalination affects coastal and marine environments; and expanded virtual water trade influences regional economies, transport emissions, social equity, and environmental burdens at production locations. Feasibility is uneven—particularly constrained in low- and middle-income contexts—implying that targeted support, improved water-use efficiency, demand management, controlled urbanization in water-scarce regions, and climate mitigation are critical. Integrated local sustainability assessments are essential to weigh benefits, costs, and risks, supported by robust regulatory oversight and environmental monitoring to ensure equitable and sustainable urban water security.

Global urban water scarcity is projected to intensify substantially from 2016 to 2050, with 1.693–2.373 billion urban residents and 19 megacities exposed by mid-century. Demographic and socio-economic growth in cities is the dominant driver, with climate change adding uncertain but meaningful pressure. Most affected cities have feasible pathways—including demand management and efficiency, desalination, groundwater development, reservoirs, inter-basin transfers, and virtual water trade—but these solutions entail significant financial, environmental, and social trade-offs, and are not accessible or sufficient for all cities. To advance the SDGs and urban livability, the study recommends: improving water-use efficiency and conservation; managing urbanization and population growth in water-scarce regions; mitigating climate change to reduce hydrological impacts; and conducting integrated, place-based sustainability assessments to guide investments, safeguard ecosystems, and ensure equitable water access. Future research should refine scarcity metrics (including environmental flow needs), better represent groundwater dynamics and energy–water nexus demands, evaluate adaptive changes in irrigation and industry, and develop decision-support tools that integrate effectiveness, costs, and multi-dimensional impacts of solution portfolios at city scales.

Uncertainties arise from data and methodological constraints. Current (circa-2016) scarcity uses 2014 water availability/withdrawal data and does not fully represent all groundwater sources or flow regulation; only documented inter-basin transfers from available datasets were included. Future withdrawal projections incorporate some irrigation changes (area, intensity, efficiency) but do not comprehensively capture adaptation to climate or water demands from energy systems (e.g., bioenergy, mining, fossil fuel extraction). Environmental flow requirements were not explicitly enforced. The choice of WSI thresholds (using 1.0 for extreme stress) affects scarcity delineation; applying a more conservative threshold (e.g., 0.4 for high stress) would yield higher exposure estimates. Runoff and availability are based on CMIP6 GCMs with inherent model spread, and downscaling of urban population, while improved, still introduces spatial uncertainty. Feasibility assessments of solutions are indicative and rely on generalized geographic and socio-economic criteria rather than detailed local studies.