为了更好地研究泸定地震同震滑坡的空间分布，获取了Sentinel-2在2022年泸定地震震前7月8日和震后10月1日的卫星影像，经过裁剪得到145.6 km²的研究区域。对两幅影像分别使用最小距离法、最大似然法、神经网络法进行监督分类，通过计算总体精度、Kappa系数对分类结果进行验证。最终选择两幅影像的神经网络法监督分类结果进行对比得到变化检测结果，研究区域内共识别到同震滑坡2 247处，占地面积22.61 km²，占总研究区域的15.53%。在对坡向的统计中发现，同震滑坡坡向以发震断层的垂直方向向两侧辐射的形势，说明同震滑坡受到发震断层的影响。高程分布分析指出，滑坡主要集中在1 000 ~ 2 000 m的高程范围内，占比超过90%。与其他因素的分析表明，滑坡事件主要集中在河流、道路附近及山坡上。主成分分析结果表明，对同震滑坡发生率最主要的影响因素是地形沉积物的分布。

In order to better study the spatial distribution of co-seismic landslides of the Luding earthquake, satellite images of Sentinel-2 on July 8 before and October 1 after the 2022 Luding earthquake were acquired, and a study area of 145.6 km² was obtained after cropping. Supervised classification was performed on the two images using the minimum distance method, maximum likelihood method, and neural network method, respectively, and the classification results were verified by calculating the overall accuracy and Kappa coefficient. Finally, the supervised classification results of the neural network method of the two images were selected for comparison to obtain the change detection results, and a total of 2 247 co-seismic landslides were identified in the study area, covering an area of 22.61 km², which accounted for 15.53% of the total study area. In the statistics of the slope direction, it was found that the slope direction of the co-seismic landslides radiated in the vertical direction of the originating fault to both sides, indicating that the co-seismic landslides were influenced by the originating fault. The analysis of elevation distribution pointed out that the landslides were mainly concentrated in the elevation range of 1 000 m to 2 000 m, accounting for more than 90% of the total. The analysis with other factors indicated that the landslide events were mainly concentrated near rivers, roads and on mountain slopes. The results of the principal component analysis indicates that the most important factor influencing the incidence of landslides in the same earthquake is the distribution of topographic deposits.