分裂窗算法在热红外遥感陆面温度反演中的应用

Application of split window algorithm in land surface temperature retrieval from thermal infrared remote sensing data

在分析分裂窗算法原理的基础上，介绍常用的几种算法，分析并选取Becker-Li分裂窗算法作为温度反演的最优方法，对ASTER数据进行辐射定标、大气纠正和几何校正，同时结合对植被反映比较敏感的Valor方法计算地表比辐射率，反演出山东省临沂地区的陆面温度．结果表明：城区温度最高，范围为315．09-319．12K，而水体温度最低，范围为299．10-303．12K，反演结果较为理想；此外，Becker-Li分裂窗算法和Valor算法能够较好地提高反演精度，为复杂区域的温度反演提供良好的应用示范．

On the basis of analyzing the split window algorithm, some commonly employed algorithms were introduced. The Becker-Li split window algorithm was analyzed and selected as the best algorithm for temperature retrieval. Radiometric calibration, atmospheric and geometric correction for ASTER data were performed. At the same time, Valor algorithm sensitive to the reflection of the vegetation was used to evaluate the specific radiativity of land surface and the land surface temperature of Linyi region was hence retrieved. The result showed that the temperature of urban area was the highest, being from 315. 09 K to 319.12 K and the temperature of water body the lowest, being from 299.10 K to 303.12 K. These retrieval results were more ideal. Besides, Becker-Li split window algorithm and Valor algorithm could improve the accuracy of temperature retrieval, offering a good example of application of temperature retrieval in complicated areas.