月球是理想的在轨辐射定标源，为提高卫星等在轨航空器的长波红外辐射测量精度，本文开展了月球长波红外辐射特性研究。通过分析月球辐射来源类型、辐射探测机理和建立辐射探测模型，探寻点目标/圆盘目标模式下月球的长波红外辐射规律，并依据天基实测数据计算其定标应用精度。首先，从月球自身辐射和反射太阳辐射两方面研究月球辐射，精确拟合月表温度，建立月球红外辐射模型，仿真两部分天基卫星辐射探测结果与月相角的数学关系；其次，对卫星红外载荷进行地面辐射定标，获得图像灰度-辐照度之间转换的关键参数，进而建立了月球长波红外辐射探测模型；最后，通过“吉林一号”对月拍摄实测数据，反演计算了探测模型中月表的反射率和发射率。实验结果表明：本文所得的月表长波红外辐射特性结果准确、鲁棒性高，与Apollo 12070实验室测量结果、Diviner实测数据拟合结果相比，月球发射率和反射率的误差分别约为7.18%和5.71%。

The moon is an ideal on-orbit radiation calibration source. In order to improve the measurement accuracy of long-wave infrared radiation for satellites and other on-orbit aircrafts, the characteristics of lunar long-wave infrared radiation are studied in this paper. The law of lunar long-wave infrared radiation is explored taking the moon as a point target or disk target, and the calibration accuracy is calculated based on the measured space-based data by analyzing the types of lunar radiation sources and radiation detection mechanism, and by establishing a radiation detection model. First, the lunar radiation is studied from two aspects of lunar self radiation and reflected solar radiation. The lunar surface temperature is fitted accurately, the lunar infrared radiation model is established, and the mathematical relationship is simulated between the radiation measurement results of space-based satellites and the lunar phase angles. Second, the key parameters of conversion between gray-scale and irradiation images are obtained through calibrating the radiation of satellite infrared loads in the ground. The lunar long-wave infrared radiation measurement model is then established. Finally, the reflectance and emissivity of the lunar surface are calculated based on the measured data taken by the “Jilin-1” satellite on the moon. The experimental results show that the long-wave infrared radiation characteristics of the lunar surface obtained by the proposed menthod are accurate and robust. Moreover, there are nearly 7.18% and 5.71% of the fitting errors in the obtained data compared with Apollo 12070 laboratory measurement results and diviner measured data, respectively.