汾渭平原PM_2.5浓度的影响因素及空间溢出效应

Based on data collected by real-time monitoring and remote sensing retrieval from 2015 to 2017, the paper probed into the spatial and temporal change of PM_2.5 concentration and its influence factors on Fen-wei Plain via spatial autocorrelation analysis and spatial regression analysis. The results showed that: 1) The growing trend of the concentration during these three years was a result of a rapid increase during the heating period (from November to next March), while there was no significant inter-annual variation during the non-heating period (from April to October). 2) The average monthly change of PM_2.5 concentration was in a U shape, with a much higher concentration during the heating period. And days with PM_2.5 non-attainment during the heating period to the total yearly PM_2.5 polluted days increased from the 75.0% in 2015 to the 83.4% in 2017. 3)Cities on the Plain were all in an increasing trend except Tongchuan and Sanmenxia, among which plains from Xianyang to Yuncheng and Luoyang Basin were experiencing the worst PM_2.5 pollution with a subtle rural-urban difference, and, thus, formed a highly polluted area. Then it followed by Linfen and plains along side the upper reach of Fen River, which were also in a bad condition but with an evident urban-rural difference. 4) Based on spatial regression analysis, there was a significant spatial spillover effect for the PM_2.5concentration on the Plain. Driving factors including annual average temperature, urbanization rate, and energy consumption positively effected the PM_2.5 concentration, and additionally, they drove the PM_2.5 pollution of neighboring areas into a worse situation. On the contrary, annual precipitation and relief amplitude were not only negatively correlated with the concentration of PM_2.5, they also helped for a lower PM_2.5 concentration in neighboring areas. Moreover, the transmission effect by wind facilitated the PM_2.5 pollution, while vegetation coverage discourage PM_2.5 concentration, but neither of their indirect effect was significant.