Blue-green infrastructure for urban resilience enhancement pathways: A hybrid empirical study integrating machine learning and Panel-QCA.

Under the dual pressures of climate change and rapid urbanization, blue-green infrastructure (BGI) has emerged as a core strategy for enhancing urban resilience (UR). Existing studies have verified the functions of individual BGI elements or measured aggregate UR levels, yet how multiple BGI elements jointly shape UR through nonlinear, substitutive, and configurational mechanisms remains underexplored, limiting refined decision-making under resource constraints. Using panel data from 22 Chinese mega- and super-large cities spanning 2014 to 2023, this study develops a hybrid framework integrating spatiotemporal analysis, Random Forest, XGBoost-SHAP, and Panel Qualitative Comparative Analysis (Panel-QCA) to examine how BGI drives UR through nonlinear mechanisms and multiple configurational pathways. Results show that UR generally increased but remained spatially uneven, and urban scale expansion did not guarantee resilience improvement. BGI structural balance mattered more than scale alone. High resilience is achievable through equivalent pathways of facility network dominance and greenway-centered configuration, while low resilience is jointly constrained by BGI misallocation, institutional absence, and structural imbalance. Greenway stocks mitigate the negative impact of insufficient fiscal investment on UR. Prioritizing well-connected greenways is cost-effective under fiscal pressure, whereas large-scale water body expansion yields diminishing marginal returns and requires sustained funding. UR governance thus requires differentiated strategies: high-resilience cities should transition to digital operation of existing facilities, stable-growth cities should strengthen cross-regional connectivity through greenways, and low-resilience cities should address facility gaps and institutional deficiencies simultaneously. This study integrates machine learning with Panel-QCA and offers empirical guidance for resilience strategies in large cities under resource constraints.