Due to the difficulties in outlining glacier using optical imagery methodology hampered by debris-cover, a new method using the combination of interferometric coherence and surface velocity to delineate debris-covered glacier limits is presented. Coherence images from difference interferometry of ALOS PALSAR data were classified using Maximum Likelihood classifiers (ML) based on Iterative Self-Organization cluster algorithm (ISO) to discern glacier limit from non-glacier area. The results were compared with historic limit from aerial photos and validated by GPS ground-truth data. The surface velocity derived from SAR feature-tracking was employed to validate the results and to discuss the glacier dynamic change. The glacier (coded 5Y663D0009) in Kongur Mountains was tested. We find that the long glacier tongue shows high coherence with low surface velocities while the middle moraine area shows low coherence with higher surface velocities. This implies that the interferometric coherence is a confident judgment for active or inactive debris-covered glacier area. This will make it possible to monitor glacier dynamic change under climate warming. The reliability and uncertainty were discussed.