Spatiotemporal dynamics of land-use carbon emission efficiency in the Yangtze River Delta Urban Agglomeration : insights from a directed weighted network

Abstract

Clarifying the spatial interactions by which socio-economic and spatial factors govern land-use carbon emission efficiency (LUCEE) is critical for collaborative urban climate governance and sustainable development in the face of increasing carbon emissions. However, most existing studies rely on oversimplified binary network models, overlooking the nuanced roles of regional connectivity and hierarchical city relationships in shaping emission patterns, which can introduce systematic policy bias. This study evaluates spatiotemporal patterns of LUCEE in the Yangtze River Delta Urban Agglomeration (YRDUA) from 2018 to 2023 using a Super-SBM model. A directed and weighted network is constructed via a modified gravity model integrating LUCEE values, a directional migration index, and refined distances. PageRank, Infomap, and social network indicators are compared to binary results, while the Quadratic Assignment Procedure (QAP) identifies network drivers. Results show core cities declined due to metabolic saturation, while secondary cities demonstrated a “catch-up effect”. High-efficiency areas expanded toward northern Zhejiang, whereas low-efficiency clusters persisted in underdeveloped regions. The network formed a “core-subcore-periphery” hierarchy transcending geography. Shannon Entropy and K-S tests confirm a “binarization trap,” where binary models systematically underestimate secondary cities. QAP reveals economic development as the dominant driver, with government intervention and industrial structure gaining significance by 2023.These insights offer a new, network-based perspective on polycentric governance, providing actionable guidance to tailor land-use policies and reallocate resources according to the true functional roles of cities within the agglomeration.