Abstract: Frequent adjustments of flights in actual operations have a negative impact on airline operations. Using visual analysis methods to analyze historical flight operation characteristics can find defects in flight plans and provide support for flight plan adjustments. For this reason, the FlightOM-Vis system for multi-dimensional spatio-temporal flight data is designed to support the progressive analysis of flight operation characteristics from top to bottom. The system used multiple views to display information, integrated clustering analysis method for multi-dimensional information of spatio-temporal flights data, retained the contextual information of the analysis results, and provided a wealth of interactive tools. The experiment used the Eastern Airlines flight data set for verification. The results show that the system can assist analysts in quickly mining the flight operation characteristics hidden in the data subset, and make guiding suggestions based on the visualization results.