基于地形复杂度的山区水土流失风险评价

Terrain complexity-based assessment of soil and water loss risk in mountainous regions

水土流失研究中通常以大范围区域数据为主导,探究小范围山地地区精细化表达,这对评价小范围山地地区水土流失情况具有重大现实意义。利用无人机数据生成高精度的数字高程模型(DEM)与三维模型,基于DEM数据计算多个地形因子,结合相关性分析、聚类分析实现地形因子的分类,再使用变异系数法和主成分分析法实现地形因子的筛选与确权并建立地形复杂度模型;随后,在地形复杂度模型中引入常数偏移量,并进行归一化处理,得到水土流失风险评价因子模型;最后,将其与三维模型和实地踏勘数据进行验证分析。结果表明:1)水土流失风险评价因子模型中坡度、地表切割深度、剖面曲率、平面曲率的系数分别为1.933、0.338、0.206和2.633;2)对于整个试验区,中风险区面积比例为28.50%,分布较为分散。极高风险区比例为4.42%,集中分布在南侧以及西北侧。总体而言,该区域以中低风险为主;3)F1和F2区域水土流失风险较高,因为其存在陡峭地形、低植被覆盖率和土壤沙化等问题。F3区域由于平坦地形和种植大量农作物,表现出相对轻微的水土流失情况。该模型能够准确提取水土流失区域,为小范围山地地区的水土流失和土壤侵蚀等地学研究提供有益的参考。

[Background]In studies related to soil and water loss,the primary emphasis is frequently placed on large-scale regional data,presenting a challenge in effectively articulating the specific conditions found in small-scale mountainous areas.Investigating fine-scale conditions in small mountainous regions is of significant practical importance for assessing soil and water loss.[Methods]This study utilized unmanned aerial vehicle(UAV)data to generate high-precision digital elevation model(DEM)and 3D model.Based on DEM data,multiple terrain factors were calculated,and the classification of terrain factors was achieved through correlation analysis and cluster analysis.Subsequently,the selection and weighting of terrain factors were accomplished using the variation coefficient method and principal component analysis,leading to the establishment of a terrain complexity model.Later,a constant offset was introduced and normalized within the terrain complexity model to derive a soil and water loss risk assessment factor model.Finally,validation analysis was conducted by comparing it with 3D model and on-site reconnaissance data.[Results]1)Coefficients for slope,surface incision depth,profile curvature,and plan curvature in the soil and water loss risk assessment factor model were 1.933,0.338,0.206,and 2.633,respectively.2)In the entire study area,the medium risk area accounted for 28.50%,with a dispersed distribution.The very high risk area accounted for 4.42%,concentrated in the south and northwest.Overall,the area was predominantly at a medium to low risk.3)The risk of soil and water loss was higher in regions F1 and F2 due to factors such as steep terrain,low vegetation coverage,and soil desertification.The F3 region exhibited relatively mild soil and water loss due to its flat terrain and extensive cultivation of crops.[Conclusions]The model is capable of accurately delineating areas prone to soil and water loss,providing valuable insights for geoscientific research on soil and water loss and sedimentation in small-scale mountainous regions.