Effects of green infrastructure composition and configuration on runoff regulation in watersheds: A global meta-analysis

Green infrastructure is a resilient, nature-based approach to regulating watershed runoff. However, at the global watershed scale, quantitative understanding of how GI composition and configuration influence runoff characteristics—such as generation and concentration processes—remains inconsistent, constraining evidence-based planning for efficient runoff regulation. This study conducted a meta-analysis of 121 peer-reviewed studies to systematically quantify the global effects of green infrastructure composition and configuration on runoff characteristics in watersheds and to identify the sources of heterogeneity in the results. The findings indicated that in terms of green infrastructure composition, shrubland exhibited slightly stronger regulatory effects on runoff characteristics compared with forest land and grassland, and forest land significantly influenced multiple runoff indicators. Regarding green infrastructure configuration, higher patch density significantly increased runoff. Moreover, moderators—including climate, topography, soil and geological conditions, and watershed area—significantly influenced the relationship between the composition and configuration of green infrastructure and runoff characteristics. The study also identified Forest% (proportion of forest land area) thresholds for different climate zones and watershed scales, recommending a minimum of 35 % to ensure watershed ecological security. These findings provide a scientific foundation for optimizing land-use planning and runoff regulation globally.