Abstract:

The microwave radiation brightness temperature is closely related to the physical, chemical and geographical distribution characteristics of the materials on the Lunar surface. The distribution characteristics of the Lunar surface microwave radiation brightness temperature are investigated by using the method of Support Vector Machine (SVM). A support vector regression model is set up to analyze the 2C level brightness temperature data acquired by the microwave radiometer for the CE\|1 Lunar satellite. Then, this model is further optimized by using the particle swarm optimization algorithm. Thus, the relationship between the Lunar time and the microwave radiation brightness temperature of four frequency channels (3GHz, 7.8GHz, 19.35GHz, 37GHz) at different locations is established. The distribution of microwave radiation brightness temperature on full Lunar surface from four frequency channels in narrow period of
a lunar day is obtained. More specific features on the Lunar surface distribution of the microwave radiation brightness temperature are unveiled and described in detail. Finally, physical factors influencing the brightness temperature on the Lunar surface are systematically discussed.〖JP〗

Key words: Brightness temperature, SVM, Particle swarm optimization, Microwave radiometer, CE-1