Abstract:
Background The gully shoulder line is the topographic structure line that can best reflect the morphological characteristics of loess landform. The shape, level, and spatial distribution of the gully shoulder line show the regional differences of loess geomorphology and is also an important basis for measuring gully erosion and landform evolution. The research of the gully shoulder line has gradually matured and has widely used. This paper systematically summarizes the current research results and scientific theoretical progress of the gully shoulder line, deeply ponders over the current research status, aiming to explore future research emphases and promote the subsequent research.
Methods In this study, using "gully shoulder line" and "positive and negative terrain" as the keywords, relevant references were selected from CNKI, Web of Science core database and other domestic and foreign databases. Starting from the definition, this paper reviewed and summarized the conceptual model and extraction method of the gully shoulder line, and used VOSviewer software to analyze it, so as to clearly grasp the research hotspots and application directions of gully shoulder line.
Results 1) The gully shoulder line can reflect the internal evolution process of loess landforms. At present, the cognition of the gully shoulder line is relatively perfect, but its classification system is shortage. 2) There are currently 4 commonly used extraction methods: Image hand-painted method, geomorphological morphology method, image segmentation, and point cloud data method, but the scope of application is not wide and the extraction accuracy is different, the traditional image hand drawing method has the highest accuracy, but it needs a lot of manpower and material resources. Compared with the traditional extraction method, the method based on geomorphic morphology is of higher efficiency, but the effect on extraction accuracy and continuity is poor. The method based on image segmentation is better in accuracy and extraction efficiency, but the algorithm of model method is more complex. The accuracy of point cloud data extraction method is second only to the traditional method. It can reflect the rich detailed information of the gully shoulder line, but it needs strong computer carrying capacity. 3) The application research of the gully shoulder line is relatively simple, and the wide range application is less. The most widely used research is based on the loess landform, followed by the research on soil erosion of the gully, while the research on land use and vegetation cover is less.
Conclusions The study proposes that the future research on the gully shoulder line should focus on the following aspects: High-precision automatic extraction, global efficiency and local fidelity of the extraction algorithm of the gully shoulder line, the coupling relationship between the gully shoulder line and the gully of the Loess Plateau, providing theoretical reference and new thoughts for researchers.