The Study Of Source Tracking And Quantitative Remote Sensing Of The Air Pollution In Urban Areas

In the past 20 years,despite remarkable achievements in air pollution control in China,air pollution in urban areas is still serious,affecting people's daily life,physical and mental health.It is a major environmental problem that urgently needs to be solved in China.In order to grasp the current situation of air pollution and the sources of pollutants,China has established more than 1,400 ground air quality monitoring stations for real-time observation of China's atmospheric conditions.However,these ground stations are still relatively sparse in space,with large distance between stations,high maintenance cost,and it is difficult to obtain continuously distributed pollutant concentration.Therefore,the ability of these ground stations to indicate local pollution sources is limited.The most concerned target of air pollution is particulate matter with an aerodynamic diameter of less than 2.5 microns,known as PM_2.5,which is the main pollutant causing haze pollution.In addition to anthropogenic pollution sources,natural sources of pollution such as wildfires and volcanic eruptions are aslo the main sources of PM_2.5.Compared with traditional ground station monitoring,satellite remote sensing technology can quickly monitor the continuous distribution of PM_2.5 and other air pollutants across large space,and can continuously observe for a long time,which has its unique advantages in tracking the occurrence and development trend of air pollutants.This dissertation aims at the spatial distribution of PM_2.5,and information acquisition of occurrence and development of air pollution,and studies of the inversion methods of PM_2.5 based on satellite remote sensing data.By analyzing the current characteristic of atmospheric monitoring satellite data and its relationship to PM_2.5,the AOT inversion methods were established by studying the dual-star synergistic AOT?Aerosol Optical Thickness?and NDVI?Normalized Difference Vegetation Index?with low-resolution.Meanwhile,this dissertation carried out the in-depth studies of improving the accuracy of PM_2.5 inversion,the identification of wildfire smoke based on multi-source data fusion and its incidence.The main innovations in this dissertation are as follows:?1?An algorithm of dual-star synergy AOT inversion in urban area with complex background was proposed.In view of the difficulty of AOT inversion in urban area consisting of surface with high reflectance,by using dual-star remote sensing data and atmospheric radiation transmission model,the relationship between reflectance at the top of atmospheric and surface reflectance of multi-band satellite was established,then the surface reflectance and AOT were reversed,realizing the AOT acquisition of urban area using satellite remote sensing data.?2?An algorithm model of AOT inversion of high spatial resolution using NDVI of low spatial resolution was proposed.In view of the difficulty of AOT inversion of high spatial resolution in urban area with complex surface,the relationship between NDVI and the surface reflectance of visible light band was constructed by using spectral library data,the surface reflectance of high spatial resolution was calculated by satellite remote sensing NDVI product,and the surface reflectance of remote sensing image of high spatial resolution was reversed by method of mutual information maximization.The AOT inversion algorithm was verified by using MODIS NDVI and HJ-1 A/B data,and the AOT inversion with high spatial resolution was realized,and the difficult problem of AOT inversion with high spatial resolution in urban areas was solved.?3?An improved PM_2.5 estimation model was proposed,which greatly improved the inversion accuracy of PM_2.5.In order to avoid the estimation error caused by the spatio-temporal inconsistency between AOT and meteorological factors in the PM_2.5 estimation model,an improved PM_2.5 estimation model was established by using the AOT and meteorological factors obtained from the same sensor and adding the population distribution parameter.The model was applied to estimate PM_2.5 in Urban Agglomeration of Chengdu Plain,and the results showed that this improved model had high precision.?4?The AOT inversion algorithm of using geostationary satellite to monitor the occurrence and development trend of air pollution was studied.For the study of air pollution source tracking in urban areas,basd on the traditional dark pixels method and the spectral range of geostationary satellite sensors,this dissertation put forward an algorithm of using NDVI values to identify dark pixels in the visible light band,and combining the look up table created by the 6S?the Second Simulation of the Satellite Signal in the Solar Spectrum?atmospheric radiation transmission model to achieve AOT inversion.The algorithm was verified by Gao Fen 4 satellite data,and the inversion results comfirmed the validity of the algorithm.?5?The method of identifying wildfire smoke based on multi-source data fusion was studied,and the influence pattern of large area wildfire on urban air quality was analyzed.Aiming at the problem of air pollution caused by frequent wildfire disasters,this dissertation studied the impact of wildfire events in western North America on air quality in New York City from 2015 to 2017 by using the proposed wildfire smoke identification method.The statistical results showed that the air quality in New York City was mostly impacted in August,with the PM_2.5 increased by 113.37%due to smoke from wildfires,and PM_2.5 components also increased significantly.It was also found that the smoke from wildfires could travel thousands of kilometers and remain in the air for days or even a week.