Spatio-temporal And Driving Analysis Of PM_2.5 Pollution In Typical Urban Agglomerations Of China

Urban agglomerations are strategic core areas for national economic development and the main areas for new urbanization,playing an important role in the shifting process of the global economic gravity center to China.However,urban agglomerations often face more serious ecological and environmental problems compared with other regions,where air pollution is a particular problem,especially fine particulate matter(PM_2.5)pollution.It has become one of the important environmental problems that urban agglomerations need to overcome in order to achieve sustainable development goals.Therefore,it is necessary to gain an in-depth understanding of the current status and influence mechanism of PM_2.5 at the scale of urban agglomerations.Additionally,heterogeneous natural geographical conditions and urbanization development levels may form different regional air pollution characteristics,and further complicate their influence mechanisms.In view of this,this study took 11 nation-level urban agglomerations in China as typical case study area and selected a variety of data including PM_2.5 concentration data,natural and socioeconomic data from 2015-2019 to conduct the following study based on a multi-urban agglomeration perspective:First,time series analysis,standard deviation ellipse and spatial autocorrelation analysis were used to explore the temporal variation characteristics and spatial distribution characteristics of PM_2.5 pollution.Then,ANOVA was further used to compare the spatio-temporal evolution characteristics among 11 urban agglomerations.Second,various regression analysis methods such as geographic detector model and optimal curve fitting were employed to explore the key natural and socioeconomic factors affecting PM_2.5 and the evolutionary relationship between each factor and PM_2.5.Then,comparing and analyzing the differences in the influencing factors of PM_2.5 in each urban agglomeration.The following are the main findings of this study.（1）During 2015-2019,PM_2.5 concentrations in the 11 urban agglomerations showed significant downward trend of up to 24.5%（from 52.53μg/m³to 39.66μg/m³）.For each urban agglomeration,the pollution situation also showed significant improvement trend.The BTH experienced the most significant improvement,decreasing from 76.89μg/m³ in 2015 to49.82μg/m³ in 2019,a decrease of 35.21%.Followed by HC and LX,with decreases of 32.40%（from 52.28μg/m³to 35.34μg/m³）and 30.62%（from 47.84μg/m³to 33.19μg/m³）,respectively.The GP showed the smallest decrease with 9.81%（from 52.51μg/m³ to 47.36μg/m³）.The remaining urban agglomerations were,in descending order,CC（28.09%）,YRD（26.46%）,MYR（23.08%）,CP（22.65%）,PRD（19.07%）,BG（17.23%）and HBEY（14.34%）.From this,the improvement of PM_2.5 pollution varied among the 11 urban agglomerations.（2）The standard deviation ellipse results showed that PM_2.5 pollution in all urban agglomerations exhibited spatial distribution pattern in the"northeast-southwest"direction.For each urban agglomeration,the PM_2.5 pollution in BTH,CC,GP and HBEY was distributed in"northeast-southwest"direction,while in YRD,MYR,HC,BG and LX was distributed in"southeast-northwest"direction.The pollution in PRD was distributed in"east-west"direction,and the distribution direction in CP was not obvious.During the study period,there was significant positive spatial correlation for all urban agglomerations,and the spatial clustering characteristics were obvious.The PM_2.5 pollution in most individual urban agglomerations also showed positive spatial correlation.The spatial clustering types are mainly of"high-high"or"low-low"types,except for HBEY and LX.In general,PM_2.5 pollution showed heterogeneous distribution pattern and clustering characteristics in space among urban agglomerations.（3）As for all urban agglomerations,the natural and socioeconomic factors that had the greatest influence on PM_2.5 were TEM and HUM,as well as POP and SI,respectively.For the individual urban agglomeration,ELE and HUM,as well as SI and POP were the natural and socioeconomic factors that contributed the most to PM_2.5in the BTH,respectively.In the YRD,it was TEM&ELE and SI&POP,respectively.PRD was WS and NDVI,as well as GDP and BA.For the MYR,TEM,HUM and VQ,NLI had greater impact on PM_2.5.In the CC,it was WS&HUM and SI&BA.HC was WS and HUM as well as POP and VQ.CP was ELE and TEM,as well as NLI and POP.BG was ELE,TEM and BA,VQ.For the GP,ELE and TEM,as well as GDP and POP were the important factors affecting PM_2.5.HBEY was ELE,WS and POP,BA.LX was WS&HUM and SI&POP.In summary,the key influencing factors on PM_2.5 pollution differed significantly across 11 urban agglomerations.（4）For all urban agglomerations,PM_2.5 concentration mainly showed linear and quadratic relationships with their influencing factors.At the individual level,curves feature between natural factors and PM_2.5 were dominated by linear relationship.Specifically,the change trend of PM_2.5with TEM,HUM and WS showed regular regional characteristics in 11 urban agglomerations;with ELE and NDVI did not show obvious regional regularity;and with PRE presented less difference.The curves feature between socioeconomic factors and PM_2.5 were diverse.Specifically,POP,BA,GDP mainly showed inverted U-shaped curve feature in most urban agglomerations.SI and VQ mainly showed curve feature with upward trend,and NLI mainly showed linear and inverted U-shaped curve feature.To sum up,there are regional differences in the curve feature fitted by PM_2.5 concentration and its influencing factors among 11 urban agglomeration.