A triple increase in global river basins with water scarcity due to future pollution

Water is an essential resource for life and ecosystems, yet only a very small fraction is readily available to people, plants, and animals. Global river discharge (~45,500 km³/year) masks strong spatial and temporal variability that creates regional mismatches between water availability and demand, leading to water scarcity. Water scarcity is generally defined as the condition where available water cannot meet environmental and societal demands. Future scarcity is expected to intensify due to both climate change, which alters the hydrological cycle and river discharge patterns, and socio-economic changes, including land-use change, irrigation expansion, dam construction, and rising withdrawals driven by population and economic growth. Prior studies suggest that in some major economic regions (e.g., India, China), direct anthropogenic alterations may exceed climate change effects in driving runoff declines. Socio-economic development also increases water demand across sectors. Despite growing awareness that degraded water quality can limit usable water, most global scarcity assessments focus on quantity alone. This study addresses that gap by assessing clean-water scarcity, jointly considering water quantity and nitrogen-driven water quality constraints at the sub-basin level worldwide for 2010 and future scenarios for 2050.

Most global water scarcity assessments have emphasized water quantity, with only a few integrating water quality explicitly. Two notable indicators are the Water Quality Dilution (WQD) indicator, applied globally for several constituents, and the Quantity-Quality-Environmental flow requirement (QQE) indicator, applied in China. Existing quantity-focused studies broadly identify hotspots in regions such as Central Europe, India, China, and parts of Africa; however, neglecting water pollution can underestimate scarcity. Prior work estimated 25–65% of the global population lived in severe water-scarce areas (1995–2005) when considering quantity only. In 2010, this study’s quantity-only estimate is about 45% of the population under severe scarcity, but including quality issues suggests much higher exposure, aligning with other studies indicating up to ~80% of the world’s population faces scarcity from combined quantity and quality pressures. Threshold choices influence quality-based scarcity classification: this study adopts a 1 mg N/L TDN threshold to avoid eutrophication (following De Vries et al.), recognizing that sector-specific thresholds (e.g., agriculture, drinking water) or stricter environmental standards could alter classifications. Overall, integrating water quality is essential to more accurately identify hotspots and inform management.

The study develops a global clean-water scarcity assessment that integrates water quantity and water quality (nitrogen) constraints at >10,000 sub-basins. An integrated modeling framework links: (1) VIC (Variable Infiltration Capacity) for hydrology and water availability/withdrawals; (2) MAgPIE (Model of Agricultural Production and its Impact on the Environment) for land-system dynamics and nitrogen budgets; and (3) MARINA-Nutrients-Global-1.0 (Model to Assess River Inputs of pollutants to seA) for river nitrogen pollution from diffuse (agriculture, natural) and point (sewage) sources. Quantity-based scarcity (Squantity) is computed using natural river discharge at sub-basin outlets (before withdrawals), sectoral surface water withdrawals (domestic, industrial, livestock, irrigation), and environmental flow requirements (EFRs) derived using a Variable Monthly Model approach. Quality-based scarcity (Squality) is based on total dissolved nitrogen (TDN = DIN + DON) loads and resulting concentrations at sub-basin outlets simulated by linking MAgPIE-derived nitrogen budgets (downscaled to 0.5°) with MARINA-Nutrients for source apportionment and in-river processing/transport; VIC provides hydrological context. Clean-water scarcity classification combines Squantity and Squality to identify hotspots where either or both indicate high scarcity (threshold matrix provided), using nitrogen concentration-based thresholds linked to eutrophication risk (e.g., >0.45 mg N/L for transition toward meso/eutrophic conditions; ~1 mg N/L for eutrophication risk). The analysis is conducted for baseline (2010) and future (2050) under three SSP-RCP scenarios: SSP1-RCP2.6 (sustainability; ambitious N policies; diet shifts; improved sewage connection/treatment; sustainable withdrawals; strong climate mitigation), SSP2-RCP2.6 (middle-of-the-road trends; moderate N policies; medium meat/dairy diet; limited sewage and withdrawal improvements; strong climate mitigation), and SSP5-RCP8.5 (fossil-fueled development; low-ambition N policies; meat-rich diet; some connection but limited treatment; higher withdrawals; weak climate mitigation). Sub-basin-level outputs include water availability changes, withdrawals, TDN inputs and dominant sources, and scarcity indicators for 2010 and 2050. Hotspots are defined where scarcity is high for quantity, quality, or both.

- Incorporating nitrogen pollution substantially increases estimated water scarcity. In 2010, classical quantity-only scarcity identifies 984 hotspot sub-basins, whereas the clean-water scarcity (including quality) identifies 2,517 hotspot sub-basins (out of ~10,226). These CWS hotspots cover ~32% of global drainage area and about half of the global population, versus ~16% area and ~42% population for quantity-only hotspots. - By 2050, clean-water scarcity hotspots remain high or increase depending on scenario: 2,587 (SSP1-RCP2.6), ~2,755 (SSP2-RCP2.6), and 3,061 (SSP5-RCP8.5). Classical quantity-only hotspots remain roughly similar (~980–998). In the worst case (SSP5-RCP8.5), CWS hotspots expand to ~48% of global drainage area with ~91% of global population living in these basins (vs. ~32% area and ~50% population in 2010). - The number of sub-basins with scarcity approximately triples due to future nitrogen pollution (from ~984 quantity-only to up to ~3,061 CWS hotspots in 2050). This implies up to an additional ~40 million km² of basin area and ~3 billion more people potentially facing water scarcity by 2050. - Quality-induced scarcity dominates in many regions already in 2010 and increasingly by 2050, particularly in South America, Central and North America, Europe, Africa, the Middle East, South and East Asia, where intensive agriculture, urbanization, and high nitrogen inputs (fertilizer, manure, sewage) elevate TDN loads. - Hotspot sub-basins disproportionately concentrate agricultural activity and nitrogen inputs. In 2010, hotspots contain ~44% of agricultural land and receive large shares of global N inputs (e.g., ~84% fertilizer, ~53% manure); by 2050 under SSP5-RCP8.5, hotspots are projected to receive ~89% of global N inputs from fertilizer and manure. - Water quantity remains important: water availability increases in ~69–72% of sub-basins but decreases in the rest; continent-level totals rise modestly in Africa, Asia, and North America (+4–6%) and decline in Central America, Europe, Oceania, and South America (down to −4%) between 2010 and 2050. Global withdrawals increase by ~10–12% by 2050, driven by urbanization and food production. - Sectoral drivers of quantity-induced scarcity vary: irrigation dominates withdrawals globally and in China, India, South America; industry is a major driver in Europe and North America (e.g., energy and manufacturing). For quality-induced scarcity, primary TDN sources vary: low nitrogen use efficiency (e.g., China, South America), high agricultural production (Europe), and nitrogen deposition/fixation on natural lands (South America, Central Africa). By 2050, agriculture remains the dominant TDN source in most hotspots across scenarios.

This study addresses the research gap by quantifying how water quality constraints (nitrogen pollution) compound or even dominate water scarcity relative to traditional quantity-only assessments. The findings show that many basins previously not classified as water scarce by quantity alone are reclassified as scarce once nitrogen pollution is considered, dramatically increasing the number of hotspots and affected area and population. The results emphasize that future scarcity is not only a matter of hydrological supply-demand balance but also of maintaining water quality suitable for use and ecosystem health. Regionally, drivers differ: irrigation and industrial withdrawals shape quantity-induced scarcity, whereas agricultural nitrogen inputs, low nitrogen use efficiencies, sewage loads, and atmospheric deposition/fixation drive quality-induced scarcity. Although climate change increases water availability in many sub-basins, it decreases it in others and, together with rising withdrawals, maintains or exacerbates quantity pressures. The dominance of quality-induced scarcity in 2050 underscores that pollution control is critical to achieving water security. The study’s hotspot mapping provides actionable insights on where targeted interventions—nutrient management in agriculture, improved sewage connection and treatment, and sustainable withdrawals—can most effectively mitigate scarcity. The strong linkages to SDGs (e.g., SDG 6, 2, 11, 12, 13) reveal potential trade-offs (e.g., fertilizer reduction vs. food production) that can be turned into synergies through efficiency improvements, circular nutrient management, and cleaner technologies.

The paper introduces and operationalizes a global clean-water scarcity concept by integrating quantity and nitrogen-based quality constraints using an interlinked modeling framework (VIC, MAgPIE, MARINA-Nutrients) at sub-basin scale. It demonstrates that accounting for nitrogen pollution substantially increases the extent and severity of water scarcity—potentially tripling hotspot basins by 2050 compared to classical quantity-only assessments—with profound implications for area and population exposure. The study identifies global hotspots, their sectoral and pollution drivers, and highlights that water quality management is as critical as water quantity management for meeting SDG targets. Future research should extend the framework beyond nitrogen to multiple pollutants (e.g., salinity, temperature, dissolved oxygen, heavy metals, pathogens, antibiotics, plastics, pesticides) and develop integrated, possibly WQI-based, multi-pollutant indicators across temporal and spatial scales. Advancing sector-specific thresholds, improving representation of intra-annual variability, and reducing model uncertainties will further refine scarcity estimates and support effective, region-specific policy interventions.

- Temporal resolution: Annual-scale assessment aggregates intra- and inter-annual variability, not capturing seasonal dynamics of flow, withdrawals, or nitrogen concentrations (e.g., higher winter concentrations, summer uptake/denitrification). - Hydrological representation: Annual metrics do not distinguish baseflow vs. stormflow contributions; EFRs applied as annual fractions (derived from monthly methods) may vary seasonally (30–60%). - Threshold selection: Using a stringent environmental threshold (~1 mg N/L TDN for eutrophication risk; moderate thresholds ~0.45 mg N/L) may bias toward environmental protection; alternative sector-specific thresholds could yield different scarcity classifications. - Model and input uncertainties: Uncertainties stem from MAgPIE nitrogen budgets, VIC hydrology and withdrawals, and MARINA-Nutrients pollution estimates, as well as scenario assumptions for land use, diet, sewage connection/treatment, and climate. - Pollutant scope: Focus on nitrogen excludes other important pollutants (e.g., salinity, dissolved oxygen, heavy metals, emerging contaminants) that could exacerbate or alter scarcity patterns. - Data and typographical inconsistencies: Some reported counts and percentages vary slightly across sections, reflecting scenario nuance and potential transcription/rounding issues.