| With the speeding up of urbanization, industrialization, air pollution has become one of the most serious environmental problem in our country, which has caused many negative impacts to the public health and ecological environment. The urban area and population continued expand constantly, industrial emissions and energy consumption increases at the same time, which led to an obvious heterogeneity in the city underlying surface, causing that the spatial differentiation of city problem is outstanding and urban air pollution is no exception. Based on specific platform and data, the distribution of regional air quality is simulated effectively and quickly, which can lay the foundation for atmospheric environment on impact assessment, planning, management and decision-making.This article selected air-quality data of Wuhan for about three years, during the period from May 2014 to December 2016, which was came from 10 ground monitoring points in Wuhan, and aimed to analyze the space-time characteristics of atmospheric pollution in Wuhan. Then obtained the significant correlation variable factor based on the bivariate correlation analysis between the air-quality data and its influence factors, such as the remote sensing data, MODIS data and meteorological data, population,socioeconomic data and so on, then builded the multiple regression analysis model based the above data. The fitting model, which was verified and adjusted, could established the analysis model for land use with the help of ArcGIS platform. Simulated and predicted air pollution in Wuhan based on the regression model of land use and obtained the high-precision spatial distribution of air quality of Wuhan at last. By consulting the related research literature at home and abroad, on the basis of predecessors' research achievements for reference, the main conclusion of this paper were as follows:1. Based on the air-quality data of Wuhan from 2015 to 2016, through the air quality index analysis in different times and seasons, the law of the timescale of air quality in Wuhan were shown as the following two aspects. First of all,in a daily distribution, nearby six clock, morning and evening three time periods had poor air quality, the air quality was the best when it was afternoon. Secondly, in different seasons,pollution was most serious in winter, the best in summer,compare with the spring and autumn, better performance in the spring.2. On the basis of the concentration of air quality data from different site monitoring in Wuhan city, then through the analysis of its year and different season average index, it is concluded that the change rule of spatial scale of air quality in Wuhan city were as follows: the Qingshan Ganghua has a high performance of air quality index especially in spring and winter; Wujiashan was population concentrated area, the data of annual air quality index was only less than that of Qingshan Industrial Park; Hanyang Yuehu, Wuchang Ziyang, Hankou Huaqiao, the seasonal variation between This three station were very similar; The years means of Donghu Gaoxin, Dunkou Xinqu, Donghu Liyuan were isometric decreased, and the air quality index in spring and summer were very close, the Dunkou Xinqu is the worst and Donghu Xinqu has a high performance in winter, but adverse in autumn; The value of air quality in the Hankou Jiangtan was lowest exceptly Chenhu Qihao located on the outskirts of control points, the region around Hankou had the lowest value throughout the year.3. According to the above part, with the collection of impact factors of the air quality for pretreatment, then do the bivariate correlation analysis with the air quality,and finally concluded that the larger correlation coefficient factors related to the index as the independent variable factor to build the fitting model. From the spatial scale, the factors concluded the water area and road density in the range of 1km, NDVI and LST_DAY in the range of 3km; From the time scale, the correlation between The monthly Temperature/The monthly Precipitation/NDVI/LST_DAY.4. With the significantly related factors concluded above, as the independent variable factor for LUR building, based on the analysis of the multivariate linear regression to build the land use regression model, the spatial and temporal prediction model of air quality index both has good fitting effect. R2 of the time simulation model has a good performance for 0.855, adjusted R2 was 0.773. The seasonal prediction model to fitting air quality index of Wuhan have a good effect,Their R2 were 0.692,0.871,0.886, 0.648 respectively. And get a fitting figure of AQI of Wuhan through sample selection and kriging interpolation,and a simulation prediction results in a short time.The fitted values are in good agreement with the actual value . |
PDF Full Text Request