中文摘要：

云南石林喀斯特地区属于典型的生态脆弱区,地表蒸散是喀斯特地区的重要地表环境参数.采用 Land- sat5 TM遥感影像数据和地面气象测站辅助数据,利用改进的 Priestlry-Tayley方程算法反演实验区地表蒸散,与 传统方法相比,该方法不需要过多的气象参数和繁琐的计算,并且具有较好的反演精度.采用 M0D16蒸散发数 据和实际测站蒸散发数据与TM 反演蒸散发数据进行精度检验,研究区 TM 反演蒸散发数据与实际测站数据进行 对比分析,绝对误差为 - 0 .32 mm,测站点处 M0D16蒸散发数据与实际测站数据进行对比分析,绝对误差为- 1.05 mm.最后随机选取地表土壤和植被类型的测试点,实验结果表明TM反演蒸散发数据和M0D16蒸散发数据 在土壤测试点的平均绝对误差为- 0. 78 mm,植被测试点的平均绝对误差为3.795 mm.

英文摘要：

As a typical ecological vulnerable karst area, the surface evapotranspiration in Shilin of Yun-nan is the vital environmental parameter. Using Landsat TM remote sensing data and the surface meteoro-logical data in observing station, this paper retrieves the area’s surface evapotranspiration with improved Priestlry-Tayley,by means of which, the retrieval is highly accurate without too many data and too much calculating, compared with the traditional one. Checking the accuracy between two groups of M0D16 evapotranspiration data vs data in observing station, TM evapotranspiration data vs data in ob-serving station, the absolute error of TM data is -0 . 32 mm, while the M0D16 is - 1. 05 mm. Randomly picking two testing areas, one featuring with the soil and another vegetation, it carries out an experiment, whose result shows that the average absolute error between TM data and MODI 6 data in soil testing area is -0. 78 mm, the average absolute error in vegetation testing area is 3. 795 mm.