基于FTIR和MODIS数据估算新疆沙漠宽波段地表比辐射率

基于FTIR和MODIS数据，建立了新疆沙漠宽波段（8～13.5 μm）地表比辐射率的最优估算模型。首先，利用傅立叶变换热红外光谱仪观测的塔克拉玛干沙漠地表比辐射率光谱数据，结合同期MODIS温度/比辐射率产品MOD11A1的29、31和32波段比辐射率值和MOD09A1的第7波段反射率值，建立宽波段地表比辐射率估算模型，并分别采用观测数据和光谱库数据验证了模型的精度，估算结果的均方根误差分别为0.0041和0.0081。其次，选择最优估算模型，利用MODIS数据，估算了新疆4个沙漠的宽波段地表比辐射率，得到了沙漠地表比辐射率的空间分布特征。结果表明：塔克拉玛干沙漠和库鲁克库姆沙漠气候干燥稳定，地表比辐射率分布较为均匀，范围为0.850～0.915；古尔班通古特沙漠受到植被和地表水分的影响，比辐射率空间分布不均匀，范围为0.890～0.915；库木塔格沙漠的地表比辐射率分布与其羽状地表类似，范围为0.860～0.910。

Estimating the Surface Broadband Emissivity of Deserts in Xinjiangbase on MODIS and FTIR Data

We proposed an optimal model to estimate the surface broadband emissivity of deserts in Xinjiang for the spectral domains 8-13.5 μm by the FTIR and MODIS data. Firstly, using the broadband emissivity spectral data measured by Fourier Transform Infrared spectrometer (FTIR) over Taklimakan Desert, and the MODIS data during same period in three thermal infrared channels 29 (8.4-8.7 μm), 31 (10.78-11.28 μm) and 32 (11.77-12.27 μm) and one near-infrared channel 7 (2.105-2.155 μm), an empirical regression equation to convert these spectral emissivity and reflectance to broadband emissivity was developed. The accuracy of the equation was verified by the observed FTIR data and the MODIS spectral library data, and the root mean square errors (RMSE) were 0.0041 and 0.0081, respectively. Secondly, the optimal equation was applied to build a distribution map of emissivity of four deserts in Xinjiang based on MODIS data. The emissivity in the Taklimakan Desert and Culukekum Desert is in the range of 0.850-0.915. Because there is dry and stable climate, the broadband emissivity is fairly well-distributed in the deserts. The emissivity in the Gurbantonggut Desert is in the range of 0.890-0.915. Because the land surface is influenced by vegetation and surface water, so its emissivity spatial distribution is not uniform. The emissivity in the Kumtagh Desert is in the range of 0.860-0.910, and its distribution is similar to its plume surface.