地表参数反演及城市热岛时空演变分析  

Estimation of Land Surface Parameters and Spatio-temporal Characteristics of Urban Heat Island

作　　者：王天星[1] 陈松林[2] 阎广建[1] 

机构地区：[1]北京师范大学地理学与遥感科学学院遥感科学国家重点实验室,北京100875 [2]福建师范大学地理科学学院,福建福州350007

出　　处：《地理科学》2009年第5期697-702,共6页Scientia Geographica Sinica

基　　金：国家973项目(2007CB714402);国家自然科学基金(40871164)项目;福建省自然科学基金项目(2007J0345)资助

摘　　要：采用TES算法实现了ASTER地表温度的反演，在对发射率估算方法改进基础上，利用单窗算法反演1989年TM地表温度。结合C地形校正，利用线性光谱模型提取植被覆盖度与城市不透水面密度，利用反射率提取NDVI。从多方面对城市热岛时空演变进行综合研究，研究表明，与等间距法相比，均值-标准差法可以较合理地刻画地表热场的分布，一定程度上可以避开不同时相的差异。最后时空对比及空间统计学分析显示，1989—2004年间福州市城市热岛面积、热岛强度都有所增加，城市热岛总体趋势为西北一东南走向，并逐渐向北一南方向偏移，而且城市热岛重心向东南方向偏移。Fuzhou was chosen as the study site, land surface temperatures were estimated form both ASTER images acquired on April 6, 2004 and Landsat TM image collected on June 15, 1989. Then the linear spectral unmixing-LSU model was used to estimate the vegetation abundances and percent of impervious surface from ASTER images. Before employing the LSU, the water body within the study area was masked and the C-terrain correction was applied to the images. Subsequently the NDVI image was also calculated using reflectance data of TM. With the inversion of land surface parameters mentioned above, the spatio-temporal changes of urban heat island （UHI） in Fuzhou from 1989 to 2004 were analyzed in various aspects. Comparisons between different methods used to determine the scopes of urban heat island showed that the mean-standard deviation method was more suitable for studying UHI than equal-interval method. Based on the researches discussed above, the dynamic changes of urban heat island were investigated and the spatial statistics technologies was then introduced, results indicated that after 15 years, the area extent of UHI was significantly enlarged and the urban heat intensity was also augmented, moreover, the standard deviation ellipse indicated that the center of UHI was moved towards southeast, which is in line with the planning policy of Fuzhou.

关 键 词：ASTER TM 植被覆盖度 不透水面密度 地形校正 线性光谱模型 城市热岛 

分 类 号：TP722.5[自动化与计算机技术—检测技术与自动化装置] P423.3[自动化与计算机技术—控制科学与工程]