Abstract: On March 15,2024,a severe forest fire occurred in Baizi Village,Gala Town,Yajiang County,Ganzi Prefecture,Sichuan Province. The fire persisted for 10 days and burned an area of 278.8km². After the fire was extinguished,ash and sediment layers ranging from 1cm to 7cm in thickness were deposited on the hillslopes. These deposits are prone to triggering post-fire debris flows under heavy rainfall conditions,posing a serious threat to settlements such as Yajiang County and nearby villages,as well as infrastructures including the G318 National Highway and the Lianghekou Hydropower Station. Using the"3·15" Yajiang forest fire site as the study area,this research employed remote sensing images and field investigations to identify topographic characteristics,fire severity,and the distribution of mobilizable source materials across different catchments within the burned area. A total of 202 catchments(with 10 conditioning factors)and debris flow occurrences from six burned areas with similar geological environmental conditions were used as a test dataset for model training and transferability assessment. After collinearity and relative weight analysis of the conditioning factors,a logistic regression model was developed to predict the susceptibility of post-fire debris flows. Six key factors were identified: the proportion of areas with moderate to severe burn severity and slopes greater than 23°,the average thickness of the ash layer and burned soils,catchment morphology coefficient,gully density,maximum height difference,and catchment area. After validating the model's transferability,the susceptibility of post-fire debris flows in the Yajiang burned area was predicted. The results indicate that the prediction model exhibits strong predictive performance(AUC=0.963,SEN=0.975,ACC=0.936,TS=0.882)and good transferability(AUC=0.864,ACC=0.851,SEN=0.956,TS=0.813). Among the 276 potential post-fire debris flow catchments in the"3.15" Yajiang burned area,192 were classified as highly or extremely susceptible,accounting for 69.6% of the total; 31 were moderately susceptible,accounting for 11.2%; and 53 were of very low or low susceptibility,accounting for 19.2%. These results provide a scientific basis for emergency prevention and mitigation planning related to post-fire debris flows.