Pollution exacerbates China’s water scarcity and its regional inequality

Water supports human survival and development, and global demand has risen nearly eightfold over the last century. China, with per capita water resources roughly one quarter of the global average, faces significant water scarcity challenges due to geographic and temporal mismatches between water availability and demand. Regional inequality is pronounced, with North China experiencing extreme pressures not captured by national averages, leading to conflicts among sectoral uses and growing policy attention. While many prior assessments quantified scarcity based on water quantity, they often neglected inadequate water quality as a critical constraint on water usability. Rapid industrialization and urbanization have degraded aquatic environments through untreated wastewater and diffuse pollution. Few studies have explicitly quantified the implications of pollution on water scarcity by comparing sectoral quality requirements to available water quality. High-resolution nationwide assessments that include water quality have been limited by data coverage, leaving the impact of inadequate quality and its contribution to regional inequality unclear. This study addresses these gaps by quantifying water scarcity across China incorporating both quantity and quality, at multiple spatial (grid and multi-order basins) and temporal (annual, seasonal, monthly) scales, and accounting for environmental flow requirements.

Earlier assessments acknowledged water quality as an influencing factor but did not quantify it, often relying on the gray water footprint to represent the dilution needed to meet environmental standards. National-level analyses using gray water footprint provided initial insights but did not explicitly incorporate sector-specific water quality requirements versus available quality. Many large-scale studies used country-level statistics (e.g., FAO AQUASTAT) or model simulations, with grid or watershed-scale assessments that may not capture actual water supply networks and inter-basin transfers. Prior work also noted that annual assessments can mask intra-annual variability. This study builds on these findings by integrating sectoral quality thresholds, using observed water quality data, and evaluating scarcity at multiple geographic and temporal scales to reveal inequality and the influence of pollution on scarcity.

The study assesses water scarcity considering both quantity and quality across China for 2012–2016 at multiple spatial scales (0.25° grid; first-, second-, and third-order basins) and temporal scales (annual, seasonal, monthly). Environmental flow requirement (EFR) is set to 80% of water availability. - Quantity-based water scarcity (WSqua): ratio of regional water withdrawal to water availability after reserving EFR. Conceptually, WSqua = (water withdrawal) / (water availability − EFR). - Quality-based water scarcity (WSpol): uses a dilution approach to translate inadequate water quality into an equivalent volume of water needed to dilute pollutants to sector-specific acceptable thresholds. For each sector i and water quality parameter j (COD, NH4-N, EC), the required dilution water d_ij is computed to meet the sectoral maximum threshold Cmax for parameter j given observed concentration Cj. The maximum across parameters and sectors (dg = max(d_ij)) represents the binding dilution requirement. WSpol = dg / (water availability − EFR). - Combined water scarcity (WScom): WScom = WSqua + WSpol. Data: - Water availability: Provincial annual natural water availability (2012–2016) disaggregated to monthly, 0.25° grids using the VIC hydrological model’s validated monthly runoff; aggregated to basin levels. - Water withdrawals: Sectoral (agriculture/irrigation, industry, domestic, eco-environmental compensation) provincial statistics downscaled spatially: agriculture via cropland distribution and net irrigation requirements; industry via gridded industrial GDP and land use; domestic via urban/rural population distributions with temperature-dependent seasonality; eco-environmental compensation (~2% of total) allocated to urbanized areas and distributed uniformly across months. - Water quality: Monthly COD, NH4-N, EC (and other reported indicators) from ~620 monitoring sites (2012–2016) interpolated to 0.25° grids using inverse distance weighting (power 2); aggregated to basin scales and temporal means. Regional inequality: Theil’s index (mean log deviation) weighted by water availability quantifies spatial inequality of scarcity across regions and time scales.

- Spatial extent: At the grid level, including quality increases both the extent and intensity of scarcity compared to quantity-only assessments. 28.8% of China’s area has WSqua > 1, while 32.0% has WScom > 1. - Regional patterns: Water-scarce areas concentrate in North China. Over half of areas in the Huai, Hai, Yellow, and Liao River basins, and 45.4% of the Songhua basin are under WSpol. In parts of the middle/lower Yangtze and southeast coasts, quality drives scarcity in places not scarce by quantity alone, indicating seasonal/quality-driven scarcity in South China despite abundant water. - Basin-level: First-order basins in South China mostly show low WScom (except Southeastern Rivers slightly > 1), while six northern basins face both WSqua and WScom. In highly stressed basins (e.g., Hai), WScom more than doubles WSqua, highlighting degraded water quality’s large effect. Areas with WSqua > 5 rely on unsustainable fossil groundwater and costly inter-basin transfers (e.g., South-to-North Water Transfer). - Seasonality: Scarcity is most severe in spring (dry season, high agricultural demand). In spring, basins under WScom comprise 70.0% (first-order), 69.7% (second-order), and 71.3% (third-order); 39.3% of grid cells have WScom. Summer/autumn generally have fewer units under WScom depending on scale. - Sectoral: Agriculture (~67% of total withdrawals) dominates scarcity contributions annually, peaking in spring/summer. In a minority of second- (6.6%) and third-order (8.6%) basins, industry or domestic uses exceed agriculture in driving WScom. Agriculture is most vulnerable when competition arises, but also offers the largest conservation potential. - Population exposure: On an annual basis, 31.1%, 75.2%, 56.2%, and 86.1% of China’s 1.36 billion people live under WScom depending on geographic scale. Considering seasonality, 61.5%, 97.7%, 96.1%, and 92.9% are under WScom for at least one season; at monthly scale, 100%, 99.2%, 98.8%, and 94.7% experience WScom for at least one month. About 31.3%, 82.6%, 94.4%, and 91.6% face severe WScom (WScom > 2) for at least one month. Between 0.62 and 1.2 billion people face severe WSqua (WSqua > 2) for at least one month; ~0.34 billion are exposed to severe monthly WScom year-round. - Inequality: Lorenz-type curves show strong inequality between withdrawals and availability; adding quality moves curves further from equality. Theil’s index is higher for WScom than WSqua, increases with finer spatial resolution, and peaks in spring (drier months amplify pollution and scarcity). South China basins cluster at low scarcity; North China at high scarcity. - Carrying capacity: Quantity-included carrying capacities < 1 in all North China basins and the Southeast River basins imply insufficient resources for current populations/economies; in Southeast Rivers, quality limits carrying capacity despite adequate quantity.

Incorporating water quality into scarcity assessments reveals that pollution substantially aggravates scarcity and intensifies regional inequality. North China endures chronic, year-round scarcity from both limited quantity and degraded quality, whereas South China’s scarcity is often seasonal and quality-driven despite abundant flows. These findings indicate that supply augmentation alone is insufficient; water quality improvement is essential to secure usable supplies. The pronounced seasonality suggests targeted, time-specific management, especially in spring when agricultural demand peaks and self-purification capacity declines. The sectoral analysis underscores agriculture’s central role: it is both highly vulnerable and a major lever for conservation through efficiency and management improvements. The inequality analysis (higher Theil’s index for WScom) highlights that pollution control can reduce disparities and human exposure, particularly in heavily impacted northern basins such as the Hai, Yellow, Huai, and Liao. The results support multi-scale planning—recognizing that grid- or annual-only views can misrepresent actual scarcity due to inter-basin transfers, reservoir operations, and spatial disconnects between abstraction and use. Incorporating infrastructure operations and targeted pollution abatement into allocation and planning could mitigate seasonal and regional scarcity hotspots.

This study provides a comprehensive, multi-scale assessment of China’s water scarcity that explicitly integrates water quality. Including quality elevates the extent and severity of scarcity, increases the number of affected people, and widens regional inequality, with North China under persistent pressure and South China experiencing seasonal, quality-driven scarcity. Key contributions include a nationwide synthesis of observed water quality data with sector-specific thresholds, VIC-informed water availability, and a multi-scale inequality analysis using Theil’s index. Policy implications point to prioritizing water quality improvements (e.g., wastewater treatment, pollution source control), particularly in northern basins, alongside sustainable agricultural water management and demand-side measures. Future research should integrate dynamic infrastructure operations, refine sector-specific quality thresholds, expand monitoring networks to reduce interpolation uncertainty, and evaluate scenario-based interventions (e.g., pollution abatement, allocation rules, irrigation efficiency) on scarcity and inequality across scales.

- Data coverage and interpolation: Water quality observations (~620 sites) required spatial interpolation to grids, introducing uncertainty, especially in sparsely monitored areas. - Metric scope: The dilution-based WSpol focuses on selected indicators (COD, NH4-N, EC) and sector-specific thresholds; it does not encompass all contaminants or ecosystem health dimensions. - Infrastructure representation: Reservoir operations, inter-basin transfers, and water supply networks are not fully modeled; thus, fine temporal/spatial scarcity may be over- or under-estimated relative to managed conditions. - Assumptions: Environmental flow requirement assumed as 80% of availability; sectoral downscaling relies on proxies (e.g., GDP, land use, population, temperature), which carry uncertainty. - Scale effects: Results are sensitive to geographic/temporal scale; annual assessments can mask intra-annual variability, while grid-based assessments may not reflect actual abstraction-use distances.