Spatiotemporal Distribution Of PM_2.5 In Southern Fujian Delta Region Based On Land Use Regression Model

In the southern Fujian delta region(Xiamen,Zhangzhou and Quanzhou),the urban integration and traffic integration are in rapidly developing stage,and human activity is also intensifying,these factors will have a certain impact on the air quality in this region.In spring and winter,severe Gray haze often occurs in the southern Fujian delta region.PM2.5(Particulate matter with aerodynamic diameter?2.5?m)is the causes of Gray haze,and it is a kind of air pollutants that can affect people's production activities and daily lives and seriously endanger human health.It is of great significance to study the spatiotemporal distribution of PM2 5 in the southern Fujian delta region in order to know the present situation of local air pollution,ensure the local air safety and promote the healthy development of the local economy.Land use regression model is a kind of efficient approach for the study of spatial and temporal variation of atmospheric pollutants.The model is applicable to small area,and it has a lower requirement on data classification and accuracy,and it can be used to draw the high-resolution image of pollutant distribution.The application of this model to study the temporal and spatial variation of fine particulate matter in the city area can help to capture the pollution condition of PM2.5,to explore the sources of PM2.5 pollutants,and to explore factors that affect the distribution of atmospheric pollutants in some specific areas.This study intends to explore the applicability of land use regression model in the southern Fujian delta region.Firstly,this study analyzes PM2.5 data from 11 air quality monitoring stations in the southern Fujian delta region.The result shows that PM2.5 pollution is more serious in spring and winter in the southern Fujian delta region,and in the other seasons PM2.5 pollution is relatively slight.So this study choose the mean PM2.5 concentration from winter of 2015(December 1,2015 to February 29,2016)to spring of 2016(March 1,2016 to May 31,2016)as the dependent variable of the model.And then,this study chooses land use,road traffic,population density and elevation data as independent variable to establish land use regression model to simulate the spatial distribution of PM2.5 concentration in the southern Fujian delta region.The application of geographic information systems(GIS)can create a series of buffers(0.1km?5km).And the buffer analysis function of geographic information systems can help to generate 60 independent variables.The extracted independent variables are used to characterize the influence of each characteristic variable on PM2.5 at different scales.Correlation analysis can be used to pick out 14 independent variables which meet the requirements of the algorithm,and these independent variables can be used to construct model.After that,this study constructed a land use regression model by multiple stepwise linear regression.There are 2 independent variables in the final model,which are the water area within 5000 meters radius and elevation.The coefficient of determination(R2)of the final model is 0.82 and Adj.R2 is 0.77,root mean square error(RMSE)of cross validation is 5.68 micrograms per cubic meter,the results show that performance of the model is good.The model can be used to predict the PM2.5 concentration in the study area,which can help to draw the high-resolution image of spatial distribution of PM2.5 concentration.This study simulated spatial distribution of PM2.5 concentration in the southern Fujian delta region by the data from 11 air quality monitoring stations,so it is feasible to build a land use regression model based on fewer monitoring sites.In the period from May 13,2014 to December 31,2016,the PM2.5 pollution in the southern Fujian delta region was relatively slight,and the seasonal variation of PM2.5 shows that there is a higher PM2.5 concentration in spring and winter,and lower PM2.sconcentrations in summer and autumn.In the southern Fujian delta region,the actual situation of PM2.5 pollution is Zhangzhou>Xiamen>Quanzhou.In the winter of 2015 to the spring of 2016,there are several serious PM2.5 events in the southern Fujian delta region.During the Spring Festival,setting off fireworks has led to a strong PM2.5 pollution.And control of southwest warm airflow and stable weather conditions also causes several serious gray haze in this study area.As can be seen from the image,in the southern Fujian delta region,there is relatively higher PM2.5 concentration in the southeast region,and relatively lower PM2.5 concentration in the Northwest region.The distribution of PM2.5 is closely related with the elevation.The higher PM2.5 concentration distributes in low altitude area,lower PM2.5 concentration distributes in the mountain area.And PM2.5 concentration is also lower in the areas on the verge of waters.The simulation results are consistent with the actual situation.In this study,we identify the geographic factors which is related to PM2.5 in the study area,and obtain the spatial distribution image of PM2.5 concentration,which can provide support for the further study of pollution exposure assessment and health impact analysis.