遥感估算绿洲-沙漠下垫面地表温度及感热通量

Land Surface Temperature and Sensible Heat Flux Estimated from Remote Sensing over Oasis and Desert

利用古浪非均匀近地层野外观测试验资料,验证了卫星反演地表温度的局地分裂窗算法在非均匀地区的适用性,进一步给出了提高反演精度的修正方案,并利用修正的反演地表温度算法和改进的空气动力学阻抗算法反演了试验区的感热通量。结果表明,Becker算法最适用于非均匀地区。与近地层野外观测试验观测结果相比,修正后的反演算法反演的地表温度和感热通量更加合理,均方根误差分别为2.38K和23.49W.m-2。地表温度和感热通量的区域分布特征相似,都表现为在绿洲和沙漠区的变化梯度较小,而在过渡区的梯度较大。

Using the Gulang Heterogeneous Underlying Surface Layer Experiment （GHUSLE） data, the applicability of local split-window algorithms in the heterogeneous underlying surface are verified, and further the modified algorithms are provided to improve the retrieval accuracy. The sensible heat flux in the experiment area is retrieved by using the modified local split-window algorithm and the improved aero- dynamics resistance algorithm. The results show that the Becker algorithm is the best one for retrieving the land surface temperature （LST） in the heterogeneous underlying surface. Comparing with the observed data, the LST and sensible heat flux obtained by the modified algorithms are more reasonable, and the root mean square errors are 2.38 K and 23.49 W.m-2. The regional distribution characteristics of the LST and sensible heat flux are similar which are expressed as： The change gradients of LST and sensible heat flux in the oasis and desert regions are smaller than those in the transition region.