複雑な汚染サイトにおける適応型地下水管理：段階的環境健康リスク評価とゾーニング

"One toxicant at a time" regulations overlook the joint risks posed by realistic coexposure to complex toxicant mixtures in groundwater. Here, 1,016,590 quarterly concentration records for 277 toxicants from 367 monitoring points were integrated into an adaptive zonation and classification framework for mixture-oriented risk management. Ecological mixture risk was quantified with species-sensitivity-distribution models combined in a tiered concentration addition → response addition (CA → RA) scheme, while Monte Carlo simulations captured probabilistic carcinogenic and noncarcinogenic human health risks. These dimensions were merged into an environmental health risk index (EHRI; 0.42 ecology: 0.58 health). Quadrant analysis identified four coexposure patterns and showed that ≤10% of detected toxicants─chiefly inorganics, metals, BTEXs, haloalkanes, and phenols─drive >95% of cumulative EHRI. Spatial interpolation converted point-level EHRIs to grid-scale and, when embedded in a "source-plume-buffer" treatment train, reduced the great concern zone (EHRI > 0.60) from 19 km² to <0.1 km² after removal of 45 highest-leverage toxicants. The framework unifies deterministic and probabilistic perspectives, pinpoints priority toxicants and zones, and supplies tier-graded actionable triggers for adaptive intervention. This study provides the first field-scale evidence that mixture-aware risk assessment can be directly coupled to risk-zoned management, offering a transferable blueprint for the sustainable regulation of complex, multipollutant groundwater systems worldwide and a framework potentially relevant to other management zones such as surface water.