黄土丘陵沟壑区重力侵蚀严重，深入理解重力侵蚀时空变化特征对区域水土保持和生态恢复具有重要意义。土壤侵蚀模型是理解侵蚀过程变化的重要工具，但黄土高原重力侵蚀模型研究严重不足。LImburg Soil Erosion Model（LISEM）模型为具有物理机制的小流域、次降雨土壤侵蚀模型，为黄土高原重力侵蚀模拟提供了潜在有力工具。然而，该模型在黄土高原应用较少，其适用性和精度仍需深入研究。鉴于此，本研究基于2021年8月19日（0819降雨事件）和10月5日（1005降雨事件）两场不同的降雨事件，利用激光雷达、摄影测量、野外采样、定位观测等多源手段，获取地形、植被、降水等环境要素，驱动LISEM模型开展黄土丘陵沟壑区桥沟流域重力侵蚀模拟，依据参数敏感性分析和野外实测数据进行模型校准和验证，构建桥沟流域LISEM模型，分析流域边坡稳定性和重力侵蚀分布特征，研究降雨和植被情景下重力侵蚀的响应。主要结论如下：

     （1）地形地貌确定下的模型参数敏感性分析结果表明，对于重力侵蚀面积，敏感性前三名的因子分别为土壤厚度、内摩擦角和粘聚力；对于土体沉积量，敏感性前三名的因子分别为土壤内摩擦角、粘聚力和土壤厚度，且降雨量对重力侵蚀模拟结果的影响大于植被覆盖度，此外，重力侵蚀模拟对饱和水力传导率也有一定的敏感性，故重点调试这些参数降低模型不确定性。

    （2）LISEM模型进行了校准和验证表明，重力侵蚀空间分布方面，模型表现出较好的适用性，校准与验证结果的全局准确率分别为99.95%和99.91%，召回率分别为76.50%和74.90%，表明模型可以较为准确的区分流域内是否发生重力侵蚀且正确预测实际重力侵蚀发生位置的能力较强，精确率分别为47.40%和37.60%，表明模型对重力侵蚀发生频数和侵蚀面积存在高估。沉积量的校准和验证结果中R²大于0.60，纳什系数大于0.40，均方根误差分别为0.18 m³和0.47 m³，平均绝对误差分别为0.26 m³和0.50 m³，模拟沉积量较实测值偏高。

    （3）重力侵蚀模拟与情景分析结果显示，桥沟流域大部分区域为稳定区域，不稳定区主要分布在40-80°坡度段。重力侵蚀基本发生在不稳定区，主要发生在50-70°坡度段和825 m-845 m高程区间内，分布在主沟道两侧的陡坎处、梯田和两条支沟部分区域。当植被覆盖度为60%时，桥沟流域土体沉积量最小。在降雨情景模拟中，随降雨量和降雨强度的增加，重力侵蚀发生频数、侵蚀面积和土体沉积量均有所增加，增加区域主要分布在梯田和西北部与北部的陡坎区。

Mass movement is one of the dominant soil erosion processes in the hilly and gully Loess Plateau. An in-depth understanding on the spatial pattern of mass movement provides a valuable reference for erosion control and ecological restoration. Soil erosion models are powerful means for understanding erosion processes, while mass movement models suitable for the Loess Plateau condition are lacking. The LImburg Soil Erosion Model is a catchment scale, event-based and physically-based model, providing a promising tool for the Loess Plateau mass movement modelling. However, few studies have applied the LISEM model on the Loess Plateau, which means the applicability of the model on the plateau has not been well evaluated. Therefore, in the study, based on two rainfall events on August 19 (0819 rainfall event) and October 5, 2021 (1005 rainfall event), LISEM model was used to simulate the mass movement in Qiaogou watershed of the hilly and gully Loess Plateau with utilizing the multi-source methods such as LiDAR, photogrammetry, field sampling, and positioning observations to acquire environmental elements including topography, vegetation, and rainfall. Based on the parameter sensitivity analysis and field investigation, model calibration and validation were conducted to construct the LISEM model for Qiaogou watershed. Furthermore, the study analyzed the slope stability of Qiaogou watershed and spatial pattern of mass movement, explored the response of mass movement in different rainfall and vegetation scenarios. Main findings are as follows:

(1) A sensitivity analysis of model parameters under the constant topography and landforms demonstrated that the three most sensitive factors for mass movement areas were soil depth, internal friction angle, and cohesion. Regarding the sediment volume, the top three factors in sensitivity were the internal friction angle, cohesion and soil depth. Moreover, the impact of rainfall on mass movement simulation was found to be greater than that of vegetation coverage. Additionally, the simulation of mass movement showed tempered sensitivity to the saturated hydraulic conductivity. Therefore, focusing on adjusting these parameters could help reduce model uncertainty.

(2) The calibration and validation of the LISEM model showed that, in terms of the spatial pattern of mass movement, the model showed good applicability and achieved high overall accuracy of 99.95% and 99.91% for calibration and validation, respectively, with recall of 76.50% and 74.90%. This suggested that LISEM can accurately distinguish whether mass movement occured within the watershed and predict actual locations of mass movement. However, the precision were relatively lower, at 47.40% and 37.60%, indicating that the model overestimated the frequency and extent. For sediment volume, the calibration and validation results yielded R² greater than 0.6, Nash-Sutcliffe efficiency coefficients greater than 0.4, and root mean square errors of 0.18 m³ and 0.47 m³, respectively. The mean absolute errors were 0.26 m³ and 0.50 m³, indicating that the model tended to overestimate sediment volume compared to actual volume.

(3) Mass movement modelling and secenario analysis demonstrated that most areas of Qiaogou watershed were stable, with unstable areas primarily concentrated in slope ranging from 40 to 80 degrees. Mass movement almost occurred in these unstable areas, particularly within slope of 50-70 degrees and elevations ranging from 825 to 845 meters. And mass movement were observed along steep slopes on both sides of the main gully, terrace, and part areas of two tributaries. When vegetation coverage reached 60%, the intensity of mass movement in Qiaogou watershed was lowest. In rainfall scenarios, the frequency, area, and sediment deposition of mass movement increased, with the expanded areas primarily located in terrace and steep slope in the northwest and north regions.

[1] Xia L, Bi R T, Song X Y, et al. Dynamic changes in soil erosion risk and its driving mechanism: A case study in the Loess Plateau of China[J]. European Journal of Soil Science, 2021, 72(3): 1312-1331.

[2] 李朋飞, 黄霖霖, 臧宇哲, 等. 黄土高原彬长矿区土壤侵蚀特征及其与环境变化的关系[J]. 农业工程学报, 2024, 40(5): 158-167.

[3] 张茜茜, 龚家国, 王浩, 等. 黄土区重力侵蚀研究进展与展望[J]. 水利水电技术(中英文), 2022, 53(12): 172-184.

[4] Guzzetti F, Mondini A C, Cardinali M,et al.Landslide inventory maps: New tools for an old problem[J]. Earth Science Reviews, 2012, 112(1): 42-66.

[5] 穆兴民, 李朋飞, 刘斌涛, 等. 1901-2016年黄土高原土壤侵蚀格局演变及其驱动机制[J]. 人民黄河, 2022, 44(9) : 36-45.

[6] Li P, Mu X, Holden J, et al. Comparison of soil erosion models used to study the Chinese Loess Plateau [J]. Earth Science Reviews, 2017, 170: 17-30.

[7] 高航, 徐向舟, 李依杭. 黄土高原植被对重力侵蚀的作用机理[J]. 中国水土保持科学(中英文), 2022, 20(6): 137-142.

[8] 杨吉山, 姚文艺, 郑明国, 等. 岔巴沟淤地坝小流域重力侵蚀产沙量分析, 水利学报, 2017, 48(2): 241-245.

[9] 高晨迪, 姚顽强, 李朋飞, 等. 黄土高原重力侵蚀研究进展[J]. 人民黄河, 2020, 42(6): 99-105.

[10] 张猛, 杨志全, 王振霖. 坡体重力侵蚀研究现状及展望[J]. 化工矿物与加工, 2023, 52(1): 48-56.

[11] Kariminejad N, Hosseinalizadeh M, Pourghasemi H R,et al. Change detection in piping, gully head forms, and mechanisms[J]. Catena, 2021, 206:105550.

[12] 赵兴阳, 徐向舟, 蒋云钟, 等. 暴雨条件下植被对黄土沟坡重力侵蚀的影响[J]. 水土保持学报, 2020, 34(1): 58-63.

[13] Gao C, Li P, Hu J,et al. Development of gully erosion processes: A 3D investigation based on field scouring experiments and laser scanning[J]. Remote Sensing of Environment, 2021, 265: 112683.

[14] 邢书昆, 张光辉, 王滋贯, 等. 黄土丘陵沟壑区浅层滑坡和崩塌形态特征与发育临界地形[J]. 水土保持学报, 2022, 36(2): 106-113.

[15] 赵新凯, 龚家国, 任政, 等. 黒垆土坡面细沟形态演变规律试验研究[J]. 水利水电技术, 2020, 51(2): 205-212.

[16] 金鑫, 刘喆, 王春振, 等. 人工黄土坡面重力侵蚀数值模拟[J]. 中国农村水利水电, 2020, 454(8), 40-45.

[17] Rui D, Ji M, Nakamur D, et al. Experimental study on gravitational erosion process of vegetation slope under freeze–thaw[J]. Cold Regions Science and Technology, 2018, 151: 168-178.

[18] De Roo A P J, Wesseling C G, Ritsema C J. LISEM: a single-event physically based hydrological and soil erosion model for drainage basins Ⅰ: theory, input and output[J]. Hydrological Processs, 1996b, 10(8): 1107-1117.

[19] 陈明, 唐川, 王飞龙, 等. 基于OpenLISEM模型的泥石流启动冲出预测[J]. 泥沙研究, 2020, 45(4): 59-65.

[20] 王晓迪. 汶川震区八一沟流域滑坡物源稳定性评价研究[J]. 人民珠江, 2020, 41(4): 134-139．

[21] Vieira D. C. S, Basso M, Nunes J. P, et al. Event-based quickflow simulation with OpenLISEM in a burned Mediterranean forest catchment[J]. International Journal of Wildland Fire, 2022, 31(7), 670-683.

[22] 李朋飞, 李豆, 胡晋飞, 等. 机载LiDAR监测黄土高原土壤侵蚀的能力评估[J]. 测绘学报, 2023, 52(08): 1342-1354.

[23] 曹斌挺, 焦菊英, 王志杰, 等. 2013年延河流域特大暴雨下的滑坡特征[J]. 水土保持研究, 2015, 22(6): 103-109.

[24] Wang S, Fu B, Lv Y, et al. Reduced sediment transport in the Yellow River due to anthropogenic changes[J]. Nature Geoscience, 2015, 9(1): 38-41.

[25] 杨吉山, 郑明国, 姚文艺, 等. 黄土沟道重力侵蚀地貌因素分析[J]. 中国水土保持, 2014(8): 42-45+69.

[26] 郭文召. 黄土沟坡重力侵蚀与产沙过程试验研究[D]. 大连: 大连理工大学, 2018.

[27] 杨吉山, 姚文艺, 王玲玲. 黄土沟道重力侵蚀规律及机理研究[J]. 人民黄河, 2014, 36(06): 93-96.

[28] 叶浩, 石建省, 程彦培, 等. ‘劈’砂岩重力侵蚀定量计算的GPS、GIS方法初探[J]. 地球学报, 2004(04): 479-482.

[29] 徐向舟, 王书芳, 赵超. 沟坡重力侵蚀的定量观测方法[J]. 人民黄河, 2012, 34(10): 20.

[30] Bremer M, Sass O. Combining airborne and terrestrial laser scanning for quantifying erosion and deposition by a debris flow event[J]. Geomorphology, 2012, 138(1): 49-60.

[31] 刘希林, 张大林. 基于三维激光扫描的崩岗侵蚀的时空分析[J]. 农业工程学报, 2015, 31(4): 204-211.

[32] 马修军, 谢昆青. GIS环境下流域降雨侵蚀动态模拟研究-以PCRaster系统和LISEM模型为例[J]. 环境科学进展, 1998, 7(5): 137-44.

[33] 张信宝, 柴宗新, 汪阳春. 黄土高原重力侵蚀的地形与岩性组合因子分析[J]. 水土保持通报, 1989 9, 9(5): 40-44+57.

[34] 金鑫, 郝振纯, 张金良, 等. 考虑重力侵蚀影响的分布式土壤侵蚀模型[J]. 水科学进展, 2008(02): 257-263.

[35] 朱同新, 陈永宗. 晋西黄土地区重力侵蚀产沙分区的模糊聚类分析[J]. 水土保持通报, 1989, 9(2): 27-34.

[36] 穆兴民, 李朋飞, 高鹏, 等. 土壤侵蚀模型在黄土高原的应用述评[J]. 人民黄河, 2016, 38 (10): 100-110+114.

[37] 高海东, 李占斌, 李鹏, 等. 淤地坝淤积高度对斜坡稳定性影响的定量评估[J]. 农业工程学报, 2012, 28(16): 127-132.

[38] 庄建琦, 彭建兵, 张利勇. 不同降雨条件下黄土高原浅层滑坡危险性预测评价[J].吉林大学学报（地球科学版）, 2013, 43(3): 867-876.

[39] 鲁克新, 李占斌, 于国强. 黄土高原小流域重力侵蚀稳定可靠度分析[J]. 干旱区地理, 2012, 35(4): 545-551.

[40] 张霞, 张振文, 李占斌, 等. 黄土高原小流域重力侵蚀稳定性评价[J]. 水土保持通报, 2012, 32(2): 236-239+301.

[41] 于国强, 张霞, 张茂省, 等. 植被对黄土高原坡沟系统重力侵蚀调控机理研究[J]. 自然资源学报, 2012, 27(6): 922-932.

[42] 王光谦, 薛海, 李铁键. 黄土高原沟坡重力侵蚀的理论模型[J]. 应用基础与工程科学学报, 2005, 13(4): 335-344.

[43] 邹兵华, 袁洁, 李占斌, 等. 淤地坝减轻坡沟系统滑坡侵蚀的数值模拟[J]. 水土保持通报, 2013, 33(1): 265-270.

[44] 张彤文. 极端降雨下黄土分布区群发性土水灾害的动力过程模拟[D]. 兰州: 兰州大学, 2023.

[45] 王飞龙. 基于OpenLISEM的震后泥石流启动-冲出过程的数值模拟研究[D]. 成都: 成都理工大学, 2018.

[46] 程子卿. 断裂构造影响区泥石流特征及危险性评价[D]. 成都: 西南交通大学, 2022.

[47] 马驰洋. Comparing and evaluating two physically-based models: OpenLISEM and Scoops3D, for landslide volume prediction[D]. 西安: 长安大学, 2018.

[48] Bout V B,Jetten V G. The validity of flow approximations when simulating catchment-integrated flash floods[J]. Journal of hydrology, 2018, 556: 674-688.

[49] De Roo A P J, Jetten V G. Calibrating and validating the LISEM model for two data sets from the Netherlands and South Africa[J]. Catena, 1999, 37(3-4): 477-493.

[50] Hessel R, Jetten V, Liu B, et al．Calibration of the LISEM model for a small Loess Plateau catchment[J]. Catena, 2003, 54: 235-254.

[51] Hessel R, Bosch, Den V, et al. Evaluation of the LISEM soil erosion model in two catchments in the East African Highlands[J]. Earth Surface Processes & Landforms, 2010, 31(4): 469-486.

[52] Starkloff T, Stolte J, Hessel R , et al. Integrated, spatial distributed modelling of surface runoff and soil erosion during winter and spring[J]. Catena, 2018, 166: 147-157

[53] 杨勤科, 李锐. LISEM: 一个基于GIS的流域土壤流失预报模型[J]. 水土保持通报, 1998(3): 83-90.

[54] 史培军, 刘宝元, 张科利, 等. 土壤侵蚀过程与模型研究[J]. 资源科学, 1999(05): 11-20.

[55] 傅伯杰, 邱扬, 王军, 等. 黄土丘陵小流域土地利用变化对水土流失的影响[J]. 地理学报, 2002(06): 717-722.

[56] 贾媛媛, 郑粉莉. LISEM模型及其应用[J]. 水土保持研究, 2004(04): 91-93,120.

[57] 张小文, 张世强, 蔡迪花, 等. 黄土高原西部不同土地利用与土壤侵蚀的相互作用[J]. 兰州大学学报(自然科学版), 2008(02): 9-14.

[58] 王玲玲. 黄土丘陵沟壑区不同空间尺度地貌单元水沙耦合机制[D]. 杨陵: 西北农林科技大学, 2017

[59] 柏佳, 孙睿, 张赫林, 等. 基于GF-1和Sentinel-2时序数据的茶园识别[J]. 农业工程学报, 2021, 37 (14): 179-185.

[60] 杨勤科, 贾大韦, 李锐, 等. 基于DEM的坡度研究——现状与展望[J]. 水土保持通报, 2007(01): 146-150.

[61] 余璐, 徐向舟, 张茂省, 等. 黄土高原重力侵蚀对地貌因素的敏感性分析[J]. 水土保持学报, 2019, 33(04): 119-125.

[62] 李宗梅, 魏锦旺, 满旺, 等. 基于D8算法和Dinf算法的水系提取研究[J]. 水资源与水工程学报, 2016, 27(05): 42-45.

[63] 杨艳芬, 王兵, 王国梁, 等. 黄土高原生态分区及概况[J]. 生态学报, 2019, 39(20): 7389-7397.

[64] 喻涵, 周子渊, 王一, 等. 极端暴雨下典型小流域重力侵蚀的分布及影响因素[J]. 水土保持学报, 2023, 37(04): 69-74.

[65] 马玉蕾. 重力侵蚀驱动的黄土塬边沟坡产沙机制[D]. 大连: 大连理工大学, 2022.

[66] 舒章康, 李文鑫, 张建云, 等. 中国极端降水和高温历史变化及未来趋势[J]. 中国工程科学, 2022, 24(05): 116-125.

[67] 杨波, 焦菊英, 马晓武, 等. 2022年黄土高原典型暴雨侵蚀及洪水灾害调查分析 [J]. 水土保持通报, 2022, 42(06): 1-13.

[68] 杜敏. 黄土高塬沟壑区河流水沙和侵蚀产沙格局变化[D]. 杨陵: 中国科学院大学(中国科学院教育部水土保持与生态环境研究中心), 2022.

[69] 凌成星, 刘华, 纪平, 等. 基于无人机影像VDVI指数的植被覆盖度估算——以陕西神木防护林工程研究区为例[J]. 森林工程, 2021, 37(02): 57-66.

[70] Zhao X, Su Y, Li W, et al. A comparison of LiDAR filtering algorithms in vegetated mountain areas[J]. Canadian Journal of Remote Sensing, 2018, 44(4): 287-98.

[71] 刘经南, 张小红. 激光扫描测高技术的发展与现状[J]. 武汉大学学报（信息科学版）, 2003, 28(2): 132-7.

[72] 陈剑南, 刘益麟, 李朋飞, 等. 1901-2016年黄土高原降雨侵蚀力时空变化[J]. 水土保持研究, 2022, 29(04): 39-46.

[73] Aston A R. Rainfall interception by eight small trees[J]. Journal of hydrology, 1979, 42(3-4): 383-396.

[74] Kamphorst E C, Jetten V, Guerif J, et al. Predicting depressional storage from soil surface roughness[J]. Soil Science Society of America Journal, 2000, 64(5): 1749-1758.

[75] Yan L, Li P, Hu J, et al. Monitoring small-scale mass movement using unmanned aerial vehicle remote sensing techniques[J]. Catena, 2024, 238: 107885.

[76] De Roo A P J, Wesseling C G, Cremers N, et al. LISEM: a new physically-based hydrological and soil erosion model in a GIS-environment, theory and implementation[J]. IAHS Publications-Series of Proceedings and Reports-Intern Assoc Hydrological Sciences, 1994, 224: 439-448.

[77] Hessel R, Messing I, Liding C, et al. Soil erosion simulations of land use scenarios for a small Loess Plateau catchment[J]. Catena, 2003, 54(1-2): 289-302.

[78] Takken I, Govers G, Jetten V, et al. The influence of both process descriptions and runoff patterns on predictions from a spatially distributed soil erosion model[J]. Earth Surface Processes and Landforms: The Journal of the British Geomorphological Research Group, 2005, 30(2): 213-229.

[79] Song Y, Deng H, Tang C, et al. Critical area identification and dynamic process simulation for landslide hazard chain formation in the upstream Jinsha River[J]. Frontiers in Earth Science, 2023, 11: 1051913.

[80] 杨几, 肖涛, 孙靓, 等. 植被根系对土体水力学特性影响的现状与定量分析[J]. 西北大学学报(自然科学版), 2024, 54(01): 53-62.