A Comparative Diagnostic Framework for Urban Heat-Island and Cool-Island Networks: Structural Heterogeneity and Targeted Mitigation Strategies in Wuhan Central Urban Area

Rapid urbanization intensifies the urban heat island (UHI) effect while degrading urban cooling island (UCI) functions. Although macro-scale thermal connectivity is well-studied, internal topological differences lack precise differentiation, hindering targeted interventions. Focusing on Wuhan’s urban area, this study applied Morphological Spatial Pattern Analysis (MSPA) to map core patches, circuit theory to simulate thermal flow paths, and NetworkX-based complex network metrics to quantify topological structures. A LightGBM-SHAP framework optimized against nine comparison algorithms constructed an interpretable resistance surface, overcoming subjective weighting limitations. The results reveal fundamental structural divergences: the UHI network exhibits a core-radiation topology with a degree centrality of 0.188, constrained by 33 pinch points totaling 7.0 square kilometers. Conversely, the UCI network displays a polycentric synergy structure with a closeness centrality of 0.146, restricted by 55 barrier points totaling 30.8 square kilometers. Louvain community detection deconstructed these networks into functional subunits, identifying heat exchange hubs and composite cooling units anchored by vegetation and water. Addressing this dual vulnerability, we propose a three-tier governance strategy involving zonal regulation, corridor activation, and node remediation. By shifting the analytical paradigm from macro-connectivity assessment to precise structural diagnosis, this framework identifies specific intervention targets, providing a transferable methodology for urban thermal environment regulation.