| Atmospheric inhaled particulate matter PM10which has become the primary air pollutants in large and medium-sized cities in China affects urban ecological environment and health of residents seriously, so it has an important theoretical and practical significance to explore the spatial distribution characteristic of the regional PM10concentrations and its influencing factors. In this paper, we intend to use aerosol optical depth (AOD) data by the MODIS satellite remote sensing and PM110quality concentration data by the ground sites during2008to2010year in Wuhan City, to establish the regression models between AOD and PM10quality concentration. Then selecting the representative MODIS-AOD space distribution data in September the last three years, and applying remote sensing and geographic information system technology to calculate the spatial concentration distribution of atmospheric inhaled particulate matter PM10for each period in Wuhan City, by using the above models. On the basis of those, through collecting the built-up land area proportions; water area proportions; relief amplitude; normalized difference vegetation index NDVI and land surface temperature,5PM10concentrations space distribution influencing factors data, we took a quantitative analysis of various factors impact on the spatial distribution of PM10concentrations by Arcgis and SPSS. The main results are shown as follows:(1) Spatial distribution of aerosol optical depth (AOD) in Wuhan City the three periods are all showed significant spatial differences, comprehensively, the high AOD value area has been including Hongshan District and Wuchang District of Wuhan main city zone, as well as Dongxihu District, the northeast part of Caidian District; the northern part of Jiangxia District of Wuhan suburbs, there are generally intensive population and building distribution, frequent human interference and production activities in those areas. And the low AOD value area has been including the northern part of Huangpi District, the eastern part of Xinzhou District, and the southern part of Jiangxia District of Wuhan suburbs, there are more farmland and forest, less industrial activities in those areas. It indicates that human activities influence the spatial distribution of AOD greatly.(2) By comparison of the correlation analysis between annual and each season of aerosol optical depth (AOD) and PM10quality concentration during2008to2010year, it was showed that the correlation coefficient between revised AOD with elevation and revised of the PM10concentration with humidity all have more increase than the original coefficient. And then contrasting five kinds of regression models between AOD and PM10concentration, linear, logarithmic functions, exponential function, power function and a quadratic equation, five kinds of regression models’determination coefficients R2between AOD and PM10concentration in spring are all slightly higher than0.3, generally lower than the annual and the other three quarters; the annual and the summer’s determination coefficients R2of AOD and PM10quality are all higher than0.4, and the autumn’s and winter’s determination coefficients R2of AOD and PM10quality all can be higher than0.5. Through sorting R2by the size to find out the optimal mathematical model of the whole year and four seasons tested by regression equation in Wuhan City, the results are shown as follows:the linear model is to the whole year and the autumn, the exponential function model is to the spring and the winter, a quadratic equation model is to the summer. The model accuracy results of the respective tested data can be more than70%, indicates that these models all have practical significance for using.(3) The spatial distribution of PM10concentration showed a significant distribution characteristic in Wuhan City during the three periods, and this characteristic is basically consistent with spatial distribution characteristic of AOD. On the whole, in the main city, Hongshan District and Wuchang District have always been being the high value area of the PM10concentrations, frequent human interference and production activities caused serious pollution in these two areas. In the suburbs, the northeastern area of Caidian District; the Hannan District and the northern area of Jiangxia District have always been being the high value area of PM10concentration. Caidian District has a large area of the mining industry, forming a large number of mined lands, Hannan District has numbers of downtowns and dry farmlands, lakes and woodlands are relatively less, northern area of Jiangxia District is nearby the main city, with the accelerating of urbanization process, it has formed a landscape pattern with dense population and constructions. While in the suburbs, northern area of Huangpi District, eastern area of Xinzhou District and southern area of Jiangxia District have always been being the low-value area of PM10concentration, because of arable land and forest land taking the main land use types in these areas, the water area is larger and the industrial activity is less relatively, indicates that human activities have a great contribution to PM10concentrations.(4) The partial correlation between the concentration of PM10concentration and five influencing factors during three periods showed that, in addition to the PM10concentrations and the proportion of water area on September7,2009, all other factors of the three periods and PM10concentrations had shown a highly significant correlation. Relief amplitude, vegetation index NDVI, surface temperature and atmospheric PM10concentration showed a highly significant negative correlation, which indicated that PM10concentration reduced with increasing of relief amplitude, vegetation index NDVI and surface temperature. Proportion of the built-up land area, the proportion of water area and PM10concentrations showed a highly significant positive correlation, which indicated that the concentration of PM10increased with the increasing of proportion of construction land area and water body area. But it is not consistent that the absolute value sorting results of partial correlation coefficients between the various factors in three period and PM10concentration,while it is consistent that the proportion of water area have been in the terminalplace, indicates it less influence to the spatial distribution of PM10concentration than other factors’, and the vegetation index NDVI have always been in the top three place, indicates it affects the spatial distribution of PM10concentration more greatly.(5) The principal component analysis of five influencing factors during three periods showed that, there were two components’ vaule more than1of eigenvalues during the three periods, but their cumulative contribution rates are in the range of60%-70%, the values are relatively low. That indicates there are other factors affecting the spatial distribution of PM10concentration, these five factors can not explain the intensity of spatial distribution of PM10concentration completely. The results of the component matrix during three periods are not entirely consistent, but vegetation index NDVI has been having a higher load in the first principal component, the three periods’ vaules are as follows,0.822,0.820,0.877. It indicates that the urban green space makes a great contribution to eliminating atmospheric inhaled particulate matter PM10pollution. |
PDF Full Text Request