Retrieval And Analysis Of Aerosol And Haze Based On MODIS

In recent years,the haze in northern China is frequent,and the occurrence of haze in most areas of China is related to PM_2.5.PM_2.5is a part of aerosol,and there is a strong correlation between them in urban areas.The aerosol inversion based on satellite data can quickly get the distribution of air pollution,which provides the basis for air treatment.Therefore,the aerosol in the study area is retrieved mainly based on MODIS data.Firstly,based on IDL language,the deep blue algorithm is written to retrieve the aerosol in the study area.The inversion results do not reflect the distribution of geossol in the study area.By analyzing the relationship between apparent reflectance,emissivity data and aerosol optical thickness,the multi regression and ridge regression model are established based on Python language to retrieve the distribution of geossol in the study area,respectively,with the study area MOD04＿The correlation of 3km aerosol was above 0.9.Although the accuracy of multivariate regression model is high,it is easy to appear the situation of more invalid values,and the ridge regression model is relatively stable.The correlation between the results of multiple regression aerosol and PM_2.5is only 0.22.The correlation coefficient between the aerosol results and PM_2.5 is 0.72.The ridge regression model is less affected by the outside world,which better reflects the distribution of air pollution.The correlation coefficient between PM_2.5and humidity is 0.71,and the relative humidity is relatively high when PM_2.5 is high,and certain humidity can promote the aggregation of fine particles.Through comprehensive analysis,we know that the mathematical model based on ridge regression better reflects the relationship between apparent reflectance data and emissivity data and aerosol,and provides better pollution distribution for air treatment.Figure 13;Table 9;Reference 47...