顾及尺度效应的多源遥感数据“源”“汇”景观的大气霾效应

Effect of Atmospheric Haze Based on Multi-source Remote Sensing Data Considering the Size Effect of Landscape Sources and Sinks

以武汉市为例,利用多源遥感数据研究城市"源""汇"景观格局与大气霾污染的相关关系.首先,基于武汉市Landsat8数据的地表覆盖分类结果,计算不同尺度下地表覆盖的整体异质性景观指数,选择异质性最大的尺度作为"源""汇"景观分析的最优尺度;在此基础上,用MODIS数据的气溶胶光学厚度(AOD)产品作为大气霾污染程度的度量,基于自相关性较小的"类别层"景观指数,使用地理加权回归分析模型对"源""汇"景观与AOD进行局部回归分析,并在工业区、商业区和居民区3种功能区内分析建筑物对大气霾污染影响的差异.结果表明:6 km是本文分析的最优尺度;大气霾污染的"源"景观为建筑物,"汇"景观为灌木和林地;减小"源"景观面积所占比例、增大其破碎化程度、"源""汇"穿插均匀分布,可以有效减小气溶胶光学厚度,降低大气霾污染;对于武汉市来说,其大气霾污染的主要来源为中心城区的商业区和居民区,即来自于市民生活.针对中心城区要发挥其经济、交通等功能,已有的建设用地不宜大面积改动这一现状,可以采取小幅度优化措施,而对于非建成区可采取建前合理规划"源""汇"景观空间分布的措施.

Based on a geo-weighted regression model, this study analyzes the relationship between the landscape patterns of sources and sinks in urban areas and atmospheric haze pollution. First, the classification result of the study area is divided into a number of square grids with side lengths of 2-9 km. Heterogeneity indices at different scales are calculated and compared to select an optimal scale for the following analysis. Then, the category level landscape indices, such as PLAND, PD, COHEDION, LPI, and FRAC MN, are calculated in each grid. The aerosol optical thickness （AOD） of Wuhan is used to represent the degree of atmospheric haze pollution. Furthermore, the mean value of the pixels in each grid is regarded as the value of the grid＇s center. Next, the landscape sources and sinks of atmospheric haze pollution are selected based on the analysis of the correlation between landscape indices and AOD. To make the following analysis more efficient, the indices selected previously are determined using their correlation coefficients. Finally, the geo-weighted regression analysis model is used to analyze the relationship between the landscape indices of the category level and AOD. In addition, the influences of industrial, commercial, and residential areas on haze pollution are analyzed based on the result of the classification of urban functional areas. The results show that the heterogeneity of the whole landscape is most obvious at a 6 km scale, so 6 km is the optimal scale for the analysis. The landscape sources of atmospheric haze pollution are the buildings, and the landscape sinks are shrubs and woodland. Reducing the proportion of landscape source area and increasing the degree of fragmentation can cut down aerosol optical thickness. Distributing the landscape sources and sinks evenly and interspersedly could effectively reduce aerosol optical thickness, which represents atmospheric haze pollution. For Wuhan City, the main sources of haze pollution are commercial and residential areas in the city center, representing public sources. Since it is not easy to adjust existing facilities and infrastructure, adjusting built-up areas slightly and planning reasonably for those areas that are not yet built up can reduce atmospheric haze pollution.