Remote Sensing (Jul 2022)

Gully Morphological Characteristics and Topographic Threshold Determined by UAV in a Small Watershed on the Loess Plateau

  • Ziguan Wang,
  • Guanghui Zhang,
  • Chengshu Wang,
  • Shukun Xing

DOI
https://doi.org/10.3390/rs14153529
Journal volume & issue
Vol. 14, no. 15
p. 3529

Abstract

Read online

Gully erosion is an important sediment source in small watershed, and causes severe land degradation, particularly in semi-arid regions. Accurately measuring gully morphological characteristics, and determining its topographic threshold, are vital for gully erosion simulation and control. In this study, 910 gullies were visually interpreted by unmanned aerial vehicle (UAV) technology combined with field measurement. Ten gully morphological characteristics were extracted from the digital orthophoto map (DOM) and digital elevation model (DEM) generated by UAV images, including gully length (L), circumference (C), plane area (PA), surface area (SA), volume (V), depth (D), top width (TW), mean width (MW), cross-sectional area (CSA), and ratio of top width to depth (TW/D). The morphological characteristics of 30 reachable gullies were measured by a real time kinematic (RTK) to validate the parameters extracted from the UAV images. The topographic thresholds were determined based on the local slope gradient (S) and upland drainage area (A), using a dataset of 365 gully heads and their corresponding land-use types. The results show that the mean absolute percentage errors (MAPE) of the 2D and 3D gully characteristics are less than 10% and 20%, respectively, demonstrating a high accuracy of gully characteristic extraction from UAV images. Gully V is significantly related to the other nine parameters. Significant power functions were fitted between V, and L, C, PA, and SA. The gully volume could be well-estimated by SA (V = 0.212 SA0.982), with a R2 of 0.99. For all land-use types, the topographic threshold could be described as S = 0.61 A0.48, implying that water erosion is the dominant process controlling gully erosion in this region. The topographic threshold is land-use-dependent, and shrubland is hardest for gully incision, followed by grassland and cropland. The results are helpful to rapidly estimate gully erosion, and identify the areas for gully erosion mitigation in small watershed.

Keywords