HY-1B/COCTS热红外波段的交叉辐射定标方法研究

HY-1 B/COCTS Thermal Infrared Radiation Calibration

遥感传感器辐射定标是定量遥感应用与分析的基础。该文以具有高辐射精度的TERRA/MODIS传感器热红外第31、32通道为参考,采用了目标跟踪交叉定标法完成了对HY-1B/COCTS传感器热红外第9、10通道的辐射定标。选取2009年4月17、21、22、25、26日的5次双星同步观测太湖的白天数据,通过大气辐射传输模型MODTRAN对两传感器相应热红外通道进行大气透过率和大气热辐射的模拟,从MODIS的入瞳辐亮度等效模拟出COCTS的入瞳辐亮度,再结合COCTS影像上提取对应的DN值,进行多点线性回归,最终获得COCTS第9和第10通道的定标系数分别为:增益0.008 Wm^(-2)sr^(-1)μm^(-1)/DN,偏置2.873 Wm^(-2)sr^(-1)μm^(-1);增益0.006 Wm^(-2)sr^(-1)μm^(-1)/DN,偏置3.913 Wm^(-2)sr^(-1)μm^(-1)。将定标结果用于OCTS热红外通道的辐射校正,与MODIS等效计算的辐亮度相比,第9和第10通道的平均偏差分别为-0.070 Wm^(-2)sr^(-1)μm^(-1)和-0.008 Wm^(-2)sr^(-1)μm^(-1),均方根误差分别为0.532 Wm^(-2)sr^(-1)μm^(-1)和0.083 Wm^(-2)sr^(-1)μm^(-1);将经辐射校正后的COCTS图像用于太湖水温反演,所得的平均偏差为1.47K。

Remote sensing sensor calibration is the basis of remote sensing image quality assurance. TERRA/MODIS which has high accuracy of radiation band 31 and band 32 was selected as the criterion for doing the radiometric cross-calibration of HY-1B/COCTS thermal band 9 and band 10, respectively. April 17,21,22,25 and 26,2004, five day synchronous images of two sensors passing through the Lake Taihu were selected to obtain the cross-calibration data. TERRA/MODIS band data were used to conduct out the pupil radiance of HY-1B/COCTS band based on the atmospheric radiation transmission mode MODTRAN to simulate the two sensors corresponding thermal infrared channel atmospheric transmittance and the atmospheric thermal radiation, combining the DN values picked-up from the COCTS data in the same area. A new model to calculate the radio-metric calibration coefficients was carried out in this study. The model is multi-points linear regression method with 5 times synchronous images. The radiometric cross-calibration coefficients are 0. 008 Wm^-2 sr^-1μm^-1/DN （gain） and 2. 873 Wm^-2 sr^-1μm^-1 （offset） in band 9 and 0. 006 Wm^-2 sr^-1μm^-1/DN （gain） and 3. 913Wm^-2 sr^-1μm^-1 （offset） in band 10, respectively. The result will be used for COCTS thermal infrared channel radiation correction. Compared with the MODIS equivalent calculation of radiance, the average deviation of 9 and 10 channels are -0. 070 Wm^-2 sr^-1μm^-1 and -0. 008 Wm^-2 sr^-1μm^-1 respectively and the root mean square error are 0. 532 Wm^-2 sr^-1μm^-1 and 0. 083 Wm^-2 sr^-1μm^-1 respectively. Radiation corrected COCTS images are used to Lake Taihu water temperature inversion, the average deviation is 1.47 K. In this paper, the calibration result is reliable and effective.