Characterizing erosion processes on a convex slope based on 3D reconstruction method

• The erosion processes on a convex slope were analyzed by a 3D reconstruction method. • Rill and interrill erosion contributed 60 and 40% to the slope erosion based on DEM. • The differences of slope erosion between the convex slope and uniform slope were discussed. • Slope erosion rate was positively correlated with rill density and rill dissection degree. Rill erosion is one of the main forms of soil erosion on hillslopes in hilly and gully regions of the Loess Plateau. For simplicity, many studies assume hillslopes with a uniform sloping profile. However, the most loess hillslopes are convex shape with variable degrees in this area. Estimating soil losses and planning soil conservation practices on such slopes have required much local judgment. The objective of this study was to clarify soil erosion evolution qualitatively and quantitatively on a convex slope based on a three-dimensional (3D) reconstruction technique and geographic information system (GIS) in simulated rainfall experiments. The results showed that: (1) The erosion processes of the convex slope could be reasonably described with the following stages: splash and sheet erosion – drop pits – head-cut erosion – intermittent rill – continuous rill – rill network development – transition from rill to ephemeral gully. (2) Soil loss rate from the convex slope varied from 0.127 Kg m −2 min −1 in the sheet erosion stage to 0.342 Kg m −2 min −1 in rill network development stage, which was much greater compared with the results of uniform slope erosion reported by others in this area. Once rill erosion evolved into the dominant erosion pattern (after 30 mm of rainfall), rill length, width, and depth enlarged noticeably with headward and lateral erosion. (3) The contribution of sediment from rill and inter-rill erosion to total erosion was accurately depicted, and rill erosion contributed up to 60% of the total sediment amount. (4) Soil erosion rate was significantly correlated with rill morphological parameters—fractal dimension, rill density, and degree of rill dissection (correlation coefficients of 0.978, 0.989, and 0.980, respectively). These results are helpful for the qualitative and quantitative understanding of the erosion processes of hillslopes in Chinese Loess Plateau, and have an important reference for the rational arrangement of erosion control measures.