中心城区地表温度空间分布及地物降温效应——以南京市为例

Spatial distribution of land surface temperature in central city proper and the cooling of objects effect: A case study of Nanjing

基于TM图像,采用灰度图像分割算法,自动提取南京市中心城区的不透水面、植被及水体;并用单窗算法反演了地表温度;依据离市中心的距离将中心城区分为45级缓冲区,依次统计分析每一级缓冲区内的平均温度、不透水面温度、植被温度、水体温度及各类地物的面积比例等;建立了地表温度与不透水面比例、植被比例、水体比例之间的关系模型。研究结果表明:随离市中心距离的增大,地表温度降低;植被、水体具有降温效应,水体的降温效应是植被的2.43倍;地表温度与不透水面比例、植被比例、水体比例的综合关系模型模拟效果较好。

By using TM imagery as the raw data in this paper, the authors first classified the objects of the studyarea into three classes, L ,e. , impervious surface （ IS）, vegetation, and water based on the algorithm of OTSU, and retrieved land surface temperature （LST） of the study area by using the mono -window algorithm. Then, the studyarea was separated into 45 pieces by setting a muhi - level buffer in a discriminating scheme, which was based on the distance between each pixel and the center of Nanjing, and the work of statistics and analysis was carried out for the mean temperature of all pixels, the mean temperature of IS, the mean temperature of vegetation, and the mean temperature of water in each buffer. Finally, the authors set up the relational models between LST and the area ratio of IS, LST and the area ratio of vegetation, LST and the area ratio of water. The results showed that the land surface temperature decreased obviously with the increasing distance from the city＇ s center. Vegetation and water could reduce the temperature in central city proper, and the cooling effect of water is 2.43 times that of vegetation. The integrated relational model between LST and the area ratio of IS, the area ratio of vegetation and the area ratio of water was performed well