Abstract Urban river water quality is critically influenced by external environmental factors across spatiotemporal scales, particularly rainfall and land use characteristics. This study investigates the Xunsi River in Wuhan, China, using an integrated framework that couples multi-scale redundancy analysis (RDA) with stable hydrogen and oxygen isotopes (δ²H and δ¹⁸O) and the MixSIAR model. The aim is to quantify the effects of rainfall and land use on total nitrogen (TN) and total phosphorus (TP), and to trace water sources for hydrological validation.Key findings include: (1) Wastewater treatment plants (WWTPs) of combined sewer systems effectively stabilize receiving water quality, though stricter effluent standards and combined sewer overflow control are needed for sustained improvement. (2) TN is predominantly controlled by point sources (WWTPs effluent) and exhibits scale-invariant correlations with urbanized land uses, whereas TP is governed by non-point sources (stormwater runoff from forest/park) and shows strong scale-dependent behavior, with forest/park vegetation (FV) emerging as the dominant contributor only at larger spatial scales (≥ 500 m). (3) Expanding building-attached vegetation alone is insufficient for pollution mitigation; low-impact development (LID) interventions (e.g., permeable pavements, bioswales) are required to enhance hydraulic connectivity and vegetation-based purification. (4) Strengthening river-lake hydrological connectivity can further mitigate urban waterway degradation. These findings provide actionable insights for spatiotemporal pollution control strategies and urban spatial planning in rapidly developing watersheds.
Wang et al. (Wed,) studied this question.