The air target has the following characteristics: high-speed maneuver, strategic deep penetration and long-range precision strike. Building wide-area, high-efficiency, accurate detection and continuous monitoring for air targets has become a new challenge for air target detection. In view of the optical detection requirements of air targets, a radiation characteristic model of aircraft plume was established under the combined effect of complex environmental factors. An accurate prediction model of the full chain including aircraft plume-sea/cloud background-environmental atmosphere-optical system-imaging detector was formed. The infrared multi-spectral self-radiation of the aircraft plume was derived based on the measured data. At the same time, the FY-2G remote sensing data was used to invert the sea surface temperature and cloud top temperature distribution in a certain sea area of the South China Sea, and the background radiation model of sea/cloud background was established. On this basis, the effects of background radiation, atmospheric path radiation and atmospheric attenuation on the spectral radiation of the aircraft plume were considered. A model for the upstream spectral radiation of the target at the entrance of the optical system was established. Combined with the diffraction effect of optical imaging system, the detectability of the infrared imaging system in the geostationary orbit under different spectral segments and cloud types was discussed. Research shows that detection efficiency is better in the atmospheric shielding band than in the wide band, and reasonable detection band selection in different backgrounds is conducive to efficient detection of targets.