In regions with steep slopes, the classification of permanent gully (PG) into hillslope gullies (HG) and valley floor gullies (VG) was obvious before the study began. However, in the rolling hill region (slope < 5°), the difference between HG and VG was often overlooked. Moreover, there is no standardized approach for classifying gullies, which results in ambiguity regarding the impact of topography on various developmental stages of gullies and significantly impairs the effective management of gully erosion in the rolling hill region. We propose a remote sensing-based classification method to categorize 1081 PG into VG and HG in the cropland of the rolling hill region. The high-resolution satellite images (0.7 m) from 2010 and 2021 of all PGs were used to obtain the gully development rate and distribution patterns. Among these, 79 typical PGs were investigated using UAV to acquire high-resolution DEM (5 cm), which was used to analyze the relationship between the single or composite topographic factor and gully development rates. The results show that the proposed classification method can effectively recognize the VG and HG in the rolling hill region. The average length, area, and volume of the VG were found to be 2.31, 3.15, and 6.59 times that of the HG, respectively. The rate of gully head retreat, expansion area, and volume of the VG were also 1.59, 2.48, and 5.81 times faster than that of the HG, respectively. In HG and VG, the retreat rate of gully head (Δl) both showed a positive linear correlation with the distance from the gully head to the catchment divide (LA). The extension rate of gully area (Δa) was positively linearly and exponentially correlated with composite topographic factor SA (product of local slope (S) and contributing area above the gully head (A)). Additionally, the Δa of VG was strongly related to the shape and size of A, while the Δa of HG was strongly related to S. The gully volume expansion rate (ΔV) of both HG and VG was influenced by factors such as the existing size of the gully, the contributing area of the outlet (Ao), and the elevation difference from the outlet to the gully head (h). Hence, the changes in Δl and Δa for PG are primarily attributed to hydraulic erosion, while ΔV is influenced by both hydraulic and gravitational erosion. The study has shown the non-negligible influence of HG and VG in the rolling hill region. And the composite topographic factors can also better predict the PG development rate. This study contributes to the formulation of effective soil erosion prevention strategies and sustainable land management practices.
Read full abstract