摘要
由于城市人口急剧增长,城市化进程成为全球变暖的主要贡献来源:城市裸露、半裸露土地和不透水面温度较高。针对城市地表物质丰度与温度的关系,以武汉市2013年夏季Landsat-8遥感数据为数据源,使用多端元光谱混合分析法动态提取每个像元中的植被-不透水面-土壤(V-I-S)端元组分,并采用大气顶层辐射值反演地表亮温,通过聚类分析典型地表物质类型的地表亮温区位表现,结合三角图分析每个像元内的端元丰度与相应地表亮温之间的关系,拟合多元回归方程探究端元丰度对地表温度的影响能力,以及地表亮温对端元丰度数值变化的表现。实验结果表明,V-I-S端元丰度情况可解释夏季地表亮温98.563%的变化,其中不透水面丰度对夏季亮温的影响最大,而植被丰度的大小与植被温度调节能力成正比。
The relationship between fraction of urban land surface material and temperature has always been a hot issue in the research of urban environment.In this paper,a multiple endmember spectral mixture analysis method was applied to extract vegetation-impervious surface-soil fraction in every pixel,and surface brightness temperature was derived by using the radiation on the upper atmosphere based on Landsat-8image.Then a clustering analysis and a ternary triangular Chart,as well as a multivariate statistical analysis,were applied to ascertain the relationship between the fraction in each pixel and the surficial brightness temperature.The results show that,the endmember fraction of V-I-S can explain 98.563% of the urban surface temperature changes,and the fraction of impervious surface has a positive impact on land surface temperature while the fraction of vegetation has a negative impact on it.
出处
《遥感信息》
CSCD
北大核心
2017年第2期113-120,共8页
Remote Sensing Information
基金
国家自然科学基金(61601418)
湖北省自然科学基金面上项目(2012FFB06501)
国家863计划重点资助项目(2012AA121303)
关键词
三角图
混合像元分解
多端元光谱混合分析法
Landsat-8
城市地表亮温
ternary triangular chart
sub-pixel unmixing
multiple endmember spectral mixture analysis
Landsat-8
urban land surface brightness temperature