Abstract: Looking into the future "air to ground" high-throughput interconnection scenarios composed of 5G and satellite network, and in order to realize the early warning of aircraft landing risk, a set of landing warning system based on multi-source real-time data, historical statistics and expert knowledge is established based on statistics and models. Then, in view of the existing research results lag problem, through the clustering analysis of ARJ21 aircraft landing QAR data, the pilot landing operation modes are divided into four categories. After that, a pilot landing mode prediction model based on decision field theory is constructed to calculate and discuss the landing mode selection of pilots with different personalities in different scenarios. Aiming at the complexity and uncertainty of landing process, a hierarchical confidence rule base reasoning method is proposed, which integrates qualitative and quantitative information to realize landing dynamic risk assessment and early warning. Finally, through the reasoning calculation of the landing process of ' 2020.10.16 Panzhihua runway grounding event ' and ' 2010.8.2 Yichun air disaster ', the results verify the effectiveness of the method, and the early warning time of Panzhihua event can reach 13 seconds.