Remote Sensing Estimation Of PM 2.5 In Hangzhou Bay And Its Relationship With Land Surface Characteristics

In recent years,with the improvement of economic level,China's air quality has gradually deteriorated,and haze and fog weather phenomena have frequently occurred,which seriously affect people's traffic and physical health.PM_2.5,the main pollutant in haze weather,has an important impact on air quality and public health.The traditional PM_2.5 site has the advantages of high time resolution and high accuracy of observation,but the site construction cost is high and the number is small,so it is impossible to monitor a wide range of PM_2.5 concentrations.In addition,China's PM_2.5 ground monitoring network was built relatively late.The urban population is dense and the type of land cover is complex.It is urgent to obtain high-temporal resolution PM_2.5 data.The rapid development of satellite remote sensing technology compensates for the lack of ground monitoring.It is possible and convenient for studying the spatial and temporal distribution pattern and driving force factors of atmospheric pollution to obtain large-area,long-term serial PM_2.5 concentration data based on the remote sensing image.Therefore,this study uses the MODIS L1B remote sensing imagery to invert the PM_2.5concentration data from March 2016 to November 2016 in the Hangzhou Bay,and analyzes the relationship between PM_2.5 and the land cover type,vegetation coverage,landscape pattern indices.The main conclusions are as follows:?1?In this study,using MODIS L1B remote sensing data as a data source,a dark pixel algorithm was used to invert the AOD distribution image of Hangzhou Bay through IDL language programming,and the inversion results were verified with AERONET ground observation data.The inversion accuracy was good.Both R² reach0.9064,demonstrating the feasibility and reliability of inversion of AOD based on remote sensing images.?2?The correlation analysis of AOD and PM_2.5 concentration in Hangzhou Bay was conducted in this study.The overall correlation coefficient reached 0.571,the spring reached 0.577,the summer reached 0.376,and the autumn reached 0.716.There was a significant positive correlation between the two.Based on this,six kinds of regression models including exponentiation,linearity,quadratic,cubic,exponential and logarithm were es Tab.lished.After the model tests,it was found that the power model fit between the two was the best,and the spring,summer,and autumn power model predictions The coefficient of determination is 0.6763,0.6785,and 0.6013,respectively.?3?Based on the relationship between AOD and PM_2.5,this study obtains the distribution map of PM_2.5 concentration in spring,summer and autumn based on remote sensing images,and analyzes the distribution map of PM_2.5 concentration obtained by interpolating data from PM_2.5 stations.The spatial and temporal distribution characteristics of air pollution in the Hangzhou Bay area.The agglomeration characteristics of PM_2.5 were found in Hangzhou Bay.The high value areas were mainly concentrated in Shanghai,while the low value areas were concentrated in Hangzhou and Shaoxing.Seasons showed the characteristics of spring>summer>autumn.?4?This study analyzed the relationship between PM_2.5 concentration and land cover types,vegetation coverage,and landscape pattern indices in the Hangzhou Bay.The concentration of PM_2.5 in various types of land cover is represented by forest land,grassland,wetland,and cultivated land.There was a significant negative correlation between PM_2.5 concentration and vegetation coverage,and the correlation coefficient was 0.7828.There was a positive correlation between LPI,PLAND,AI,ED and PM_2.5concentrations on artificial surface type.LPI,PLAND,AI,ED and PM_2.5 values were significantly negatively correlated on Woodland type.