Teleconnections detection of land use carbon emissions beyond geographical proximity promotes the construction of cross-city networks

Abstract

Considering the spatial association network of land use carbon emissions (LUCE) between cities is essential for supporting the sustainable development of urban agglomerations and achieving carbon neutrality. The Yangtze River Delta urban agglomeration (YRDUA), as one of China's most developed regions, has seen rapid growth in all aspects of the cities due to rapid urbanization. However, urban interactions in some areas can no longer meet the goal of regional integration for the YRDUA. Therefore, this study first identifies LUCE at the pixel scale from both direct and indirect perspectives. Then a complex network model is used to construct the spatial correlation strength, network structure, and network clustering among cities to identify spatial associations between cities within the YRDUA. Finally, the CatBoost machine learning regression model is used to analyze the factors influencing LUCE. The results show that: (1) The levels of LUCE in the YRDUA have significantly changed, showing a spatial distribution characterized by local concentration and overall fragmentation. The total LUCE from land use in the YRDUA increased by 3.96 times, while the growth rate decreased from 100.62% to 10.38%. (2) The intensity of the associations between cities is continuously rising, with large cities exerting an increasing influence on smaller cities. Shanghai, Suzhou, Wuxi, and Nanjing are situated at the core of the network structure, exerting the most significant influence on the cities within the region. (3) The interactions between different sectors have become more frequent, and the spillover effect is more evident. However, cities tend to collaborate with others within the same block. (4) GDP is the primary factor affecting LUCE, with its contribution amounting to 27.6% in 2000 and 32.5% in 2020. Moreover, the impact of climate change on LUCE is becoming increasingly complex.

Keywords

• Carbon emissions
• Land use
• Complex network
• Machine learning
• Urban agglomeration