Abstract: Rapidly and accurately mapping areas with debris flow traces (DFTAs) following debris flow disasters holds great significance for emergency rescue and post-disaster assessment. Conventional approaches are time-consuming since they rely heavily on professional-grade unmanned aerial vehicle (UAV) photogrammetry and manual delineation. This study investigated the digital orthophoto maps (DOMs) of a disaster-stricken area that were derived from a consumer-grade UAV-DJI Mavic 3 Pro. Using the zero-shot image segmentation capability and interactive point prompts of the SAM2 model, this study investigated the potential of the Segment Anything Model 2 (SAM2) for rapid, accurate information extraction of post-disaster DFTAs. The SAM2-based method was verified through field application in the study area located in the Qiongshan Gully, Badi Town, Danba County, Sichuan Province. The results indicate that this method achieved DFTA mapping on the day when the debris flow disaster occurred in the study area, yielding a Kappa coefficient of 0.889, an F1 score of 0.896, and an intersection over union (IoU) of 0.812. Comparative experiments across different regions and datasets further verified the stability and generalizability of the method. The results of this study confirm that the SAM2 model can achieve efficient information extraction of DFTAs from UAV images while eliminating the need for sample training. The developed method provides a time-efficient, cost-effective technical approach with great promotion potential for post-disaster emergency responses.