TERRA/MODIS热红外通道陆地晴空大气可降水分裂窗反演

Retrieval Atmospheric Water over Land Using TERRA/MODIS Thermal Split-window Algorithm in Clear Sky

大气水汽在遥感反演地表温度中起到关键作用,精确的大气水含量对于提升反演精度有着重要意义,而热红外方法是夜晚获得区域大气水汽含量的唯一方法。本研究采用热红外方法中的改良分裂窗算法,首先,使用TIGR大气廓线与MODIS波谱响应函数通过MODTRAN辐射传输模式进行模拟,将大气廓线的水汽含量与大气透过率进行回归;然后,根据MODIS数据产品特点,使用2015年夏季MODIS数据观测角小于20°的范围进行反演;最后,将白天反演水汽分别与全球GPS地基大气水汽观测网络和MODIS近红外水汽产品进行比较,均方根误差分别为5.5 g/cm2和6.4 g/cm2,显示了高度的区域一致性。由于MODIS观测角度较大,本研究未对其他角度进行反演,若对其余角度的MODIS数据进行反演,可根据余弦角度变化,针对不同角度范围进行模拟与回归。

Atmospheric water vapor plays a key role in remote sensing inversion of surface temperature, and it would improve a lot with accurate atmospheric water content. And thermal algorithm is the only way to retrieval atmospheric water vapor. This study uses a new split window algorithm. First, using the TIGR atmospheric profiles and MODIS spectral response function to simulate the atmospheric water vapor and transmittance through the MODTRAN radioactive transfer model. Next, Retrieval with the summer MODIS data of 2015 that the scope of observation angle less than 20 degrees. There trivial results were compared with GPS observation station and MODIS near Infrared water product, which show highly consistent with the RMSE of 5.5g/cm2 and 6.4 g/cm2, respectively. Because of the large observation angle of MODIS, this study does not carry on the inversion to residue angles. If the MODIS data of there side angles are inversed, the simulation and regression can be carried out according to the cosine.