Abstract: The construction of debris flow disaster virtual scenes based on virtual reality(VR) technology provides a new method for interaction and analysis of disasters. However, there are many deficiencies in the visualization of disaster scenes, such as low efficiency of VR scene modeling, single interactive mode, weak query and analysis. Therefore, this paper proposes a fusion modeling and interaction method with spatial semantic constraint for debris flow VR scene, the fusion modeling method is addressed in detail. Multimode interaction scenes are designed by symbol augmented expression and the acquisition of ray focus technologies. A prototype system was implemented and a case study region is selected for experiment analysis. The experimental results show that the proposed method can efficiently construct debris flow VR scene, and meet the user requirements for immersive perception of disaster scenes and disaster information query and analysis.