Effects of Riparian Land Use and Hydrological Conditions on Water Quality in the Middle and Down Streams of Han River

ABSTRACT

Understanding spatial heterogeneity in riparian land use impacts on water quality is crucial for watershed management. This study quantified spatially varying relationships between five riparian land use types (residential, grassland, forest, agricultural, and bare land) and stream water quality at 94 sites in China's Han River Watershed, using integrated geographically weighted regression (GWR) and redundancy analysis (RDA). Results showed significantly degraded water quality in tributaries versus the mainstream, exacerbated downstream by concentrated pollution and altered hydrology. GWR significantly outperformed ordinary least squares regression in predicting seasonal water quality variations, demonstrating high prediction accuracy by capturing critical spatial heterogeneity. GWR identified grassland, agricultural, and residential land as the dominant drivers of water quality. RDA established key hydrological controls (elevation, flow velocity, and slope) correlating with nutrient accumulation. We detected critical spatial non‐stationarity: Fragmented upper reach grassland buffers provided localized NO ₃ ‐N reduction, whereas middle/lower reaches experienced agricultural phosphorus flushing and intensified algal responses near urban areas. Model results suggest that spatially targeted interventions in Han River may reduce nonpoint source pollution: In agricultural mid‐reaches, converting croplands to grassland buffers could reduce nitrate loading; in urbanized zones, improved stormwater management and wastewater treatment could mitigate phosphorus; and in reaches affected by altered hydrology, optimizing flow regimes might reduce sediment resuspension effects. However, these interventions are specific to Han River conditions and would require localized feasibility assessment and cost–benefit analysis before implementation. While case‐specific, the integrated GWR–RDA framework demonstrates a novel approach correlating land‐use dynamics with water quality variations, offering a promising pathway for watershed management strategies.