关键词: 雾霾污染, 空间关联, 网络分析, 模体, 指数随机图模型

Abstract: Based on the nonparametric approach of Granger causality, the paper identifies the spatial dependence relationship of haze pollution among 279 cities in China. It uses the diversified network analysis method and rolling window technology to reveal the global characteristic and microstructure of haze pollution spatial correlation. Finally, it utilizes exponential random graph mode to investigate the determining factors in the formation of the spatial correlation. It finds that there exists spatial dependence relationship of haze pollution among sample cities. In recent years, the serious haze pollution in China may be due to the increasingly strengthened regional characteristics of haze pollution. The result shows that human-related factors also play important roles in the formation of spatial correlation of haze pollution. The paper refutes the misperceptions about haze pollution and its regional transport. The conclusions of the paper can help to establish people's subjective initiative in haze controlling.

Key words: Haze Pollution, Spatial Correlation, Network Analysis, Motif, ERGM