Air Pollution Monitoring And Spatiotemporal Feature Analysis Based On Remote Sensing Of Atmospheric Components

With the rapid development of urbanization and the improvement of industrialization,the air pollution has become a common concern of people all over the world which is closely related to our health.The generation of gaseous pollutants is connected with the combustion of fossil fuels,automobile exhaust and other factors,which seriously threaten the life and health of our people.Due to the characteristics of satellites with all-day and high coverage,the measurement of trace gas column density in the troposphere based on remote sensing data is an important method to realize the monitoring of atmospheric pollutants in gaseous states.The thesis uses the L1B spectral data of Sentinel-5 Precursor to carry out the research of atmospheric trace gas retrieval,and mainly improves the retrieval of the total slant column density and the subsequent stratospheric-tropospheric NO₂ column density separation process.The main contents of the thesis are as follows:1.The basic principle and realization process of DOAS algorithm are studied.Based on the L1B level data of Sentinel-5P,the spectral fitting window of 405-450 nm was selected based on the characteristics of the features of the absorption cross-section curve,and the optical thickness fitting method was used to fit the gas molecular absorption cross-section during the measurement process.At the same time,the pseudo absorption cross-section processing method is adopted for the Ring effect.The retrieval method was used on the strip images in eastern China,and the correlation coefficient R between the retrieval results of the slant column density and the official product was 0.905,which was highly correlated.2.The Stream algorithm is applied to the stratosphere-troposphere stratification process of the total slant column density,and the Sentinel-5P data from previous years are used to improve the calculation of the pollution weight factor,decreasing the instrumental measurement error between different sensors.The Stream algorithm can better correct the misestimation of stratospheric NO₂ column density in some regions using the official separation algorithm.Finally,the correlation coefficient R of the tropospheric vertical column density data after deducting the stratospheric NO₂ column density and the Sentinel-5P official data reached 0.902,which was extremely significant.At the same time,since the results of the secondary products are underestimated by 30%-50%compared to the ground data according to other research,and there is an upward offset of 35.12%between the vertical column density retrieval results in the thesis and the official products.It can be considered that the retrieval results in the thesis are closer to near-ground truth data.3.Based on the retrieval algorithm of this thesis,the retrieval of tropospheric NO₂vertical column density was carried out in mainland China.The results showed that the geographical distribution of NO₂ pollution is closely associated with the distribution and density of urban agglomerations such as Beijing-Tianjin-Hebei,Yangtze River Delta,and Pearl River Delta.Based on the time series analysis in the Yangtze River Delta region,it is found that both natural factors and human activities are important factors affecting tropospheric NO₂ emissions.During the initial outbreak of the COVID-19 in 2020,the tropospheric NO₂ column density in the Yangtze River Delta region decreased in January and February 2020 compared with last year with 56.05%and 53.12%,the downward trend is obvious.Finally,using the backward trajectory model,a cluster analysis was implemented on the backward trajectory of high air mass transport in the Yangtze River Delta region,and it was found that the main sources of local pollution were not only local emissions but also from the Northeast area.