黄土高原是我国乃至全球土壤侵蚀最严重的地区之一，受限于监测技术，黄土高原土壤侵蚀的研究多集中于径流小区尺度，大规模的侵蚀研究和野外观测仍相对缺乏。机载激光雷达（LiDAR）技术的革新为高精度、大规模土壤侵蚀研究提供了新的可能。然而，该技术受复杂地形的影响会产生高程不确定性误差，导致其监测土壤侵蚀的能力缺乏深入研究。基于此，本文以我国黄土高原丘陵沟壑区典型小流域桥沟流域为研究对象，联合机载LiDAR测量和模糊推理系统（FIS），定量分析了两期DEM求差（DoD）不确定性的空间分布，研究小流域土壤侵蚀与沉积的空间分布特征。结果表明：（1） 地形形态、点云密度等对插值生成的DEM误差影响较大，地形平坦区域的DoD不确定性明显小于地形陡峭区域的DoD不确定性；（2） FIS算法通过将已知的误差源整合到一个稳定的误差模型中，减少了主观干预和人为误差，避免了在复杂的DEM表面对误差进行错误估计，提高了计算结果的准确性；（3） 梁峁坡地区侵蚀产沙体积为9.21 m³，占坡沟系统产沙体积的15.89%；沟谷坡地区侵蚀严重，侵蚀产沙体积为48.76 m³，占坡沟系统产沙体积的84.11%，在坡沟系统的产沙中起主导作用。沟床地区主要以沉积为主，梁峁坡和沟谷坡对侵蚀产沙贡献较大，而沟床则主要为产沙沉积区。研究成果不仅为黄土高原小流域的土壤侵蚀监测技术的拓展提供了新的视角，也为实施有效的水土保持措施提供了理论依据和参考。

The Loess Plateau is widely recognized as one of the most severely eroded regions, both within China and globally. Because of the limitations in monitoring technology, the study of soil erosion has primarily focused on areas of small-scale flow; indeed, large-scale erosion studies and field observations remain relatively scarce. The introduction of airborne LiDAR technology has opened new possibilities for high-precision, large-scale soil erosion research. However, LiDAR is impacted by complex terrain and introduces height uncertainty, which limits its capacity to effectively monitor soil erosion. This study analyzes the typical small watershed of Qiaogou in the hilly gully area of the Loess Plateau in China. To overcome the aforementioned challenges, LiDAR measurements were combined with a fuzzy inference system (FIS) to quantitatively analyze the spatial distribution of the DoD uncertainty and investigate the spatial distribution characteristics of soil erosion and sedimentation within the small watershed. The results show that: (1) Terrain shape and point cloud density greatly influence the DEM error generated by interpolation, with significantly smaller DoD uncertainty in flat terrain regions than in steep terrain regions; (2) By integrating known error sources into a stable error model, the FIS algorithm reduces subjective intervention and human error, avoiding error estimation on complex DEM surfaces and improving the accuracy of calculation results; (3) The sediment yield volume of the hillslope area is 9.21 m³, comprising 15.89% of the sediment yield volume of the slope gully system. Erosion in the gully slope area is severe, with a sediment yield volume of 48.76 m³, comprising 84.11% of the sediment yield volume of the slope gully system; hence, it is the main component of the sediment yield of the slope gully system. The trench bed area is primarily sedimentary. Ridge slopes and gully slopes have a larger contribution to sediment production, whereas the bottom of the gully is mainly a sediment-producing area. These research findings provide a new perspective on the development of soil erosion monitoring technology in the small watersheds of the Loess Plateau and offer a theoretical basis and reference for the implementation of effective soil and water conservation measures.