中文摘要：

运用MODTRAN大气模型和通道光谱响应函数对2003年10月7日过境广西北海市的ASTER数据进行大气校正，从Planck公式和热辐射传输方程出发实现了地表温度的反演，采用ASTERVNIR和SWIR数据计算归一化植被指数（NDVI）和归一化湿度指数（NDWI），并分析了5个TIR通道地表温度、地表温度与NDVI、地表温度与NDWI之间的相关关系。结果表明，该方法反演5个TIR通道的地表温度具有显著的一致性，且与实测气温基本相符；地表温度与NDVI、NDWI之间具有明显的负相关关系；利用ASTER数据进行城市热环境效应定量分析，取得了较为理想的结果。

英文摘要：

Taking an ASTER data of the study region in Beihai city, as the data source,the moderate resolution transmittance （MODTRAN） radiation model and the spectral response function of ASTER TIR bands have been applied in atmospheric correction, then land surface temperature（LST） were retrieved from five TIR bands by using an integrated algorithm based on the PLANCK function and radioactive transfer equation. The NDVI and the NDWI were calculated from the ASTER VNIR and SWIR bands after atmospheric correction. Compared the LST derived from TIR band with air temperature from observatory, there was an obvious corresponding between the average value of five channels LST and the air temperature. The results demonstrate that LST possessed evidently stronger negative correlation with the NDVI and NDWI. According to the quantitative analysis, the conclusion is that the LST algorithm is effective.