Identifying urban haze islands and extracting their spatial features

Lei Zhu,Qingxu Huang,Qiang Ren,Huanbi Yue,Chentai Jiao,Chunyang He
DOI: https://doi.org/10.1016/j.ecolind.2020.106385
IF: 6.9
2020-08-01
Ecological Indicators
Abstract:<p>Air pollution has great impacts on both human health and economic development. The difference in air pollution concentrations between urban and rural areas (urban haze island) is a widespread phenomenon. However, a method that can effectively depict the features of this phenomenon remains lacking. This study developed a method based on the concentric ring approach and modified sigmoid function to identify urban haze islands and extract three features: the background value of urban haze islands (BUHI), the intensity of urban haze islands (IUHI) and the extent of urban haze islands (EUHI). Using this method, we analyzed the PM<sub>2.5</sub> concentration data from 346 Chinese cities in 2016 and examined the spatial characteristics and driving factors of the three features of urban haze islands. The results showed that about two-thirds (218/346) of Chinese cities had a significant urban haze island phenomenon, with an average BUHI of 36.8 μg/m<sup>3</sup>, an IUHI of 11.3 μg/m<sup>3</sup>, and an EUHI of 7.7 km. The developed method can successfully capture the variations of PM<sub>2.5</sub> concentration, and the RMSE of the fitting curve was 0.46 μg/m<sup>3</sup>. On the national scale, natural factors had a greater impact (with the standardized regression coefficients varying from −0.74 ~ 0.80) on the BUHI and IUHI than socioeconomic factors (-0.32 ~ 0.32), while socioeconomic factors had a greater impact (0.29 ~ 0.38) on the EUHI than natural factors (−0.23 ~ 0.17). This study proposes a new and simple method to identify urban haze islands with high efficiency and wide applicability. Our results can be helpful for understanding urban-rural environmental differences and providing guidance for policymakers.</p>
environmental sciences
What problem does this paper attempt to address?