The PM2.5 Estimation And Tracing To The Source In Urban Area Based On Multi Source Data

With the continuous development of China’s social and economic construction, the problem of air fine particulate matter pollution has become more and more intense. By satellite remote sensing data to study the pollution characteristics of fine particulate matter in the atmosphere, and establish the model of estimating the quality concentration of fine particles and trace the source are hot research topics, and they also have a very important practical significance.In this paper, this paper studies Chengdu which is an important central city in western region of China. This paper studies the pollution characteristics in Chengdu by data of aerosol optical thickness and mass concentration, the correlation between fine particle mass concentration and aerosol optical thickness through vertical and humidity correction, and to improve glowworm swarm optimization(GSO), and verify the reliability of the algorithm by experiments. The research contents are as follows:(1) From the point of view of daily mean and monthly mean, this paper studies the time variation of the fine particle mass concentration data in 2014 and 2015 published by the ground environment monitoring sites in Chengdu; this paper analysis and discusses the correlation of fine particulate matter and particulate matter, carbon monoxide and other pollutants from the perspective of the source of pollutants and the existence form; this paper analysis and discusses the effects of meteorological factors on the mass concentration of fine particles, discusses the relationship between rainfall, temperature, air pressure and the mass concentration of fine particles.(2) Based on the MOIDS 1B data, Aerosol optical thickness of Chengdu area in 2009 of the same year with the pollution source data is retrieved by the mature dark pixel algorithm. Based on the pollution source data and the administrative division of Chengdu City, the paper calculates the average value of overall regional, enterprise group, mountain region and road in the 15 regions of Chengdu and analysis their differences and reasons.(3) The correlation between aerosol optical thickness and fine particle mass concentration was increased from 0.344 to 0.593 through the correction of vertical and humidity in this paper. According to the season, the paper makes a variety of model fittings and multivariate linear regression model with meteorological elements for them. And compare these various models by correlation coefficient and significance.(4) This paper studies and improves GSO to apply to trace the source of pollution. The target function is the space weight of the firefly to each pollution source in improved GSO, and improved GSO add the influence of wind elements and aerosol optical thickness values in location update process. The parameters of the algorithm are determined by analysis and experiment. Finally, an experiment verifies the reliability of the algorithm using a small regional enterprise data.