U-TAE算法提取城市热岛信息的空间尺度分析  

Spatial Scale Analysis of the Methods Using U-TAE Algorithm to Extract the Information of Urban Heat Island

作　　者：刘宇鹏[1] 杨波[1] 陈崇[2] 

机构地区：[1]湖南师范大学GIS研究中心资源与环境科学学院,长沙410081 [2]南京大学国际地球系统科学研究所,南京210093

出　　处：《地理科学进展》2011年第9期1118-1124,共7页Progress in Geography

基　　金：科技部中德国际科技合作重点项目(2007DFA90500);湖南省自然科学基金项目(10JJ3022)

摘　　要：U-TAE算法是一种使用不同范围的滑动窗口来提取城市热岛信息并根据稳健估计法设定窗口阈值的算法,该算法根据像元被记为热岛像元的累积总次数是否为0来区别热岛区与非热岛区,根据累积总次数的多少来判断热岛强度的强弱,具有动态阈值的特点和无偏特性。文章以长沙地区为例,对U-TAE算法在6个不同空间尺度上(5×5,11×11,25×25,51×51,101×101,201×201)对城市热岛信息提取的效果从提取范围、提取强度、空间构成以及大面积热岛提取稳定程度4个方面进行了评估。结果认为大小为11×11的窗口在4个方面均取得较好的效果,提取的城市热岛信息更加完整和有效。U-TAE is an algorithm which uses moving windows of different areas to extract the information of UHI and sets each moving window＇ s threshold according to robust estimation method. This algorithm distinguishes non-UHI area and UHI area according to the cumulative times in which pixels have thermal anomalies and determines the intensity of UHI according to the cumulative times with high or low values. It has dynamic threshold and unbiased characteristics. The U-TAE algorithm is used to evaluate its performance in six different scales： 5×5, 11×11, 25×25, 51×51, 101×101 and 201×201. The performance evaluation includes four aspects： area, intensity, space structure of UHI and stability of large-scale UHI. The results show that the best size of the moving window is 11 × 11. At this size the information of UHI is more complete and effective than at other sizes.

关 键 词：遥感 U-TAE 城市热岛 空间尺度 长沙 

分 类 号：X16[环境科学与工程—环境科学]