Influence of urban green space landscape pattern on river water quality in a highly urbanized river network of Hangzhou city

Optimizing urban green space landscape patterns is a key goal for improving river water quality. However, little is known about the spatial heterogeneity of the impact of urban green spatial patterns on river-water environments. This study investigated the influence of urban green space patterns on the spatiotemporal heterogeneity of river water quality in the Hangzhou section of the Beijing-Hangzhou Canal through exploratory regression analysis and a combination of geographically weighted regression analysis (GWR) and spatial interpolation. The results show that (1) total nitrogen (TN) and nitrate nitrogen (NO3–-N) are the leading indicators of river pollution in the study area, (2) green space configuration is more crucial for improving water quality than composition, and (3) GWR can effectively explain the impact of urban green space on river water quality. For example, landscape shape and edge indices have a great impact on ammonia nitrogen (NH4+-N), TN and total phosphorus (TP); the more complex the shape and edge of the green space, the more beneficial it is for water purification. The interpretation of NO3–-N is complex and mainly influenced by the largest patch index (LPI) and landscape composition. Given the limited land availability in urban area, the spatial configuration of urban green space should be optimized without additional land use to minimize the non-point source (NPS) pollution with the smallest possible green space area. The proposed approach provides a new understanding of the interaction between spatial patterns of green space and the urban water environment, and valuable information for developing green space planning policies for local sites.