MODIS Combine Ground Monitoring Station Data To Monitor The Quality Concentration And The Diffusion Of PM2.5-shuang Liu District In Cheng Du

With the rapid development of city and industry,the increase of source energy consumption and the high intensity emission of atmospheric pollutant are far beyond the capability of our environment.The problem of air polluting is becoming more and more significant,which seriously threat People's daily life and physical and mental health.Haze is not only the problem of a city,but the problem of the whole world.And PM2.5 is the main cause of urban atmospheric environmental pollution,which contains a lot of harmful substances such as Polycyclic Aromatic Hydrocarbons,Metallica and so on.Almost all of them have strong ability of absorbing and scattering the visible light,which can reduce the visibility of the air,and they can stay for a long time in the atmosphere,so that cause the“haze”weather.Therefore,through take research on the distribution of PM2.5 and monitor the quality of PM2.5to reduce the harmless to human is the top priority of our country even of the world task.At present the traditional monitoring method of atmospheric environment is through establishing ground monitoring stations,but this way cannot take large scale of monitor the PM2.5 because of the existing number of station is too small to monitor the large area.The rapid progress of satellite remote sensing technology makes up for the lack of ground monitoring stations,and through satellite remote sensing image retrieval we can obtain large scale,high resolution and continuous PM2.5 concentration data.Satellite remote sensing has the unique advantages of extensive coverage,and is able to obtain atmospheric conditions quickly.Using remote sensing images to detect pollutants in the air has gradually become an important monitoring method.In this thesis,ShuangLiu district in ChengDu is selected as a study case.Some Terra-Modis satellite remote sensing samples from July in 2017 to January in 2018which covered by fewer cloud are selected in this paper.Using the mature dark pixel method,basing on ENVI5.3 platform,and showing the everyday aerosol optical thickness of ShuangLiu.At the same time,collecting PM2.5 instantaneous concentration of the 12 monitoring stations in ShuangLiu district which is obtained at the same time with the remote sensing images,and using the spatial statistical analysis method to analysis PM2.5 concentration with some influences factors.Due to the method of dark pixels algorithm to retrieve aerosol optical thickness should conform to some special application conditions,such as the inversion effect in the regions with high vegetation density is better than the lower vegetation coverage.The image inversion effect of more clouds is low,and the less the cloud,the better the inversion effect.Therefore,it is necessary to choose the better part of the aerosol optical thickness inversion results.It relates to that extraction of the optical thickness value of the aerosol optical thickness of each monitoring station and the concentration value of PM2.5 obtain by the monitoring station,and then the correlation analysis is made on the two data,and a regression model based on linear function,exponential function,logarithm function and multiplication power function is respectively established,and the optimal fit model is selected by comparing the regression model fitting with the comparison of the accuracy of the model inspection.At the same time,the influence of meteorological factors and land use types on PM2.5 concentration are also investigated.The correlation analysis is carried out,and the regression fitting model between each influencing factor and PM2.5 concentration was established to predict the PM2.5concentration of the ground.In order to verify the accuracy of the construction model.The main research results and cognition of this paper are shown as follows:(1)The optical thickness of aerosol in the ShuangLiu region is inversed by using the dark-pixel algorithm to get an image of AOD inversion at different times.(2)The change trend of PM2.5 in ShuangLiu district is analyzed in the time dimension.The PM2.5 value in winter is generally higher,and the PM2.5 value in summer and autumn is lower.(3)For monitoring site centered around a range of land use types,and analyze land utilization factor of farming,buildings,traffic,water land using on the influence of the concentrations of PM2.5 and mutual relations.(4)The correlation between aerosol optical thickness and PM2.5 was analyzed by the Pearson correlation coefficient and bilateral significance of aerosol optical thickness and PM2.5.In this paper,we establish four regression models of the exponential function of the exponential function of an exponential function of aerosol optical thickness and PM2.5.We selected the one-dimensional model with the highest degree of goodness of fit,and determined that the one-dimensional function was the best fitting model in autumn and winter by comparing their models with the goodness of fit~#.(5)Based on the regression model of PM2.5 concentration and AOD thickness value,meteorological factors and land use factors were added to optimize the one-dimensional function,and a multivariate regression prediction model was established.It is proved that the accuracy of the method is higher than that of one unitary function,and it is proved that the optimization method is feasible in the prediction model.