Method for quickly identifying landslide-prone terrain
By using DEM data and terrain position index algorithm and combined with GIS tools, we can identify landslide-prone terrain, solving the problems of simple division of slope morphology and data dependence in the existing technology, and achieving rapid and accurate identification of landslide-prone terrain.
Patent Information
- Application Number
- CN202410999924.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-12
AI Technical Summary
In the evaluation of landslide proneness, the slope morphology type is simple, and cannot truly reflect the real world terrain. It also requires historical disaster points to calculate and model training, resulting in a large number of data requirements and low recognition efficiency.
Using digital elevation model (DEM) data, the plane curvature, profile curvature and slope are extracted through the GIS surface analysis tool, combined with the topographic position index algorithm to calculate the slope index (TPI), divide the slope morphology types, and identify landslide-prone terrain, rely only on DEM data does not require historical disaster points.
It realizes rapid identification of landslide-prone terrain, can approximate and realistically divide slope patterns, improves identification efficiency, reduces dependence on historical disaster point data, and improves identification accuracy and efficiency.
Smart Images

Figure CN118941730A_ABST
Abstract
Description
Technical Field
[0001] The invention provides a method for quickly identifying landslide-prone terrain, belonging to the technical field of geological disaster prevention. Background Art
[0002] Landslide is a geological disaster that seriously affects the safety of life and property. In order to better prevent and reduce disasters, scholars have conducted long-term research on landslide susceptibility evaluation and hazard assessment. Existing studies usually use historical disaster points, and use multiple methods such as analytic hierarchy process and information method to optimize influencing factors and establish evaluation models. However, the terrain is constantly changing. In recent years, new landslides have occurred without warning, indicating that a method is needed to identify prone terrain using less data, to help people quickly identify target areas for detailed investigation, and to carry out further work.
[0003] The master's thesis of Liao Mingyong of Chongqing University, "Research and Application of Geological Hazard Area-Single Risk Assessment Method", discloses a surface-point combined geological hazard risk assessment method. This method is based on geological survey reports, combined with Landsat8 satellite images, GIS, Google remote sensing images, field surveys, the third land historical data and other multi-source data, to collect disaster factors of various factors such as topography, geological structure, hydrological meteorology, vegetation environment, and human engineering activities in the study area, and construct a risk assessment index system. It uses core technical methods such as machine learning and numerical simulation to conduct geological hazard risk assessment in a surface-point combined manner. Among them, the slope morphology type (slope shape) is divided by the curvature method, and the slope shape is divided into concave slope and convex slope according to the size of the curvature. However, this method has the following disadvantages: (1) The slope morphology type is simply divided into concave and convex, which cannot truly reflect the slope morphology in the real world; (2) It requires historical disaster points to form samples.
[0004] Zheng Yanan's master's thesis "Comparative Study on the Evaluation Effect of Collapse Proneness Based on Different Evaluation Units" from Taiyuan University of Technology discloses a method for evaluating collapse susceptibility. This method uses 10 influencing factors, including elevation, slope height, slope gradient, slope aspect, slope position, slope type, rock formation, landform, valley river, and road, to complete the evaluation of collapse susceptibility based on the CF sensitivity index model, CF binary logistic regression model, and AHP information volume model. The slope shape is distinguished by the value of the comprehensive curvature. The curvature value between -0.1 and 0.1 is defined as a straight slope, a concave slope when the curvature value is greater than 0.1, and a convex slope when the curvature value is less than 0.1. However, this method has the following disadvantages: (1) The slope morphology types are simply divided into three types: concave, convex, and flat, which cannot truly reflect the slope morphology in the real world; (2) Historical disaster points are required to complete the sensitivity index calculation and information volume calculation.
[0005] Huang Jianhao's master's thesis "Data Mining Method and Application Research on the Main Control Factors of Collapse and Landslide Disasters" in Chongqing University takes Wushan and Wuxi as the research areas, and uses the geographic detector method, CART decision tree method, and random forest method to construct an evaluation system for the main control factors of landslide disasters. It evaluates 17 terrain factors, 5 hydrological factors, 1 meteorological factor, 1 vegetation factor, 2 geological structural factors, and 2 human activity factors, and obtains that the evaluation system for the main control factors of landslide disasters consists of 19 disaster-causing factors. The CART decision tree prediction model, random forest prediction model, and SVM prediction model of landslide disasters are established. The 19 main control factors of landslide disasters include 2 geological structural factors, 10 terrain and geomorphological factors, 5 hydrological and meteorological factors, and 2 human activity factors. However, this method has the following disadvantages: (1) There are many factors, and data from different sources are required to complete the calculation; (2) Historical disaster points are required to complete the training of the prediction model. Summary of the invention
[0006] In view of the above technical problems, the present invention provides a method for quickly identifying landslide-prone terrain, which has achieved the following objectives:
[0007] (1) Use DEM to divide the slope morphology close to the actual terrain;
[0008] (2) No historical disaster sites are required to identify and extract prone terrain.
[0009] The specific technical solution of the present invention is:
[0010] A method for quickly identifying landslide-prone terrain comprises the following steps:
[0011] (1) Based on the digital elevation model (DEM) data, GIS surface analysis tools are used to extract the plane curvature, profile curvature, and slope;
[0012] (2) Based on the relationship between slope morphology type and plane curvature, profile curvature, and slope, complete the classification of slope morphology type;
[0013] (3) Based on the digital elevation model (DEM) data, the terrain position index algorithm is used to compare the elevation of each cell in the DEM with the average height of the specified neighborhood around the cell. Using ring analysis, the slope position index (TPI) is calculated as follows:
[0014] TPI i =M0-∑ n-1 M n / n
[0015] Among them, M0 is the grid elevation value of the TPI to be calculated, M n is the elevation value of the analysis area, and n is the total number of grids in the analysis area.
[0016] For slope location classification, TPI values are used to classify the terrain into different slope location types. This classification is based on the range of TPI values for each raster and the slope of this location. TPI values are higher near the top of the mountain, while TPI values are lower near the bottom of the valley. Where TPI values are close to zero, they are flat or intermediate slopes, and the slope value is used to distinguish between these two slope types.
[0017] (4) The obtained slope morphology type and slope position are put into the grid calculator to extract the grid cells that meet both the mid-slope position and the convex slope, thus completing the identification of landslide-prone terrain.
[0018] The present invention uses only DEM data, without historical disaster points, to quickly identify landslide-prone terrain and delineate target areas for further detailed investigation. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 The slope diagram of the present invention is
[0020] Figure 2 It is a schematic diagram of the slope position of the present invention;
[0021] Figure 3 It is the technical roadmap of the present invention;
[0022] Figure 4 is a flow chart of the slope classifier of the present invention;
[0023] Figure 5 It is a flow chart of the slope position classifier of the present invention;
[0024] Figure 6 It is the CV slope extraction result diagram of the embodiment;
[0025] Figure 7 It is a slope extraction distribution result diagram of the embodiment;
[0026] Figure 8 4 is a distribution map of main slope positions in Sichuan Province according to an embodiment. DETAILED DESCRIPTION
[0027] The technical solution of the present invention is explained in conjunction with embodiments.
[0028] The present invention provides a method for quickly identifying landslide-prone terrain, comprising the following steps:
[0029] (1) Based on digital elevation model (DEM) data, GIS surface analysis tools are used to extract plan curvature, profile curvature, and slope;
[0030] (2) Based on the relationship between slope morphology type and plane curvature, profile curvature, and slope in Table 1, the slope morphology type is divided into: Figure 1 .
[0031] Table 1 Slope morphology types and their geometric characteristics (ND means not defined)
[0032]
[0033] (3) Based on the digital elevation model (DEM) data, the terrain position index algorithm is used to compare the elevation of each cell in the DEM with the average height of the specified neighborhood around the cell. The ring analysis is used, and the slope position index TPI calculation formula is as follows:
[0034] TPI i =M0-∑ n-1 M n / n
[0035] Among them, M0 is the grid elevation value of the TPI to be calculated, M n is the elevation value of the analysis area, and n is the total number of grids in the analysis area.
[0036] For slope location classification, the TPI value is used to divide the terrain into different slope location types. This classification is based on the TPI value range of each grid and the slope of this location. The TPI value near the top of the mountain is higher, while the TPI value near the bottom of the valley is lower. The TPI value close to zero is the flat or intermediate slope, and the slope value is used to distinguish between these two slope types. The classification criteria are shown in Table 2. The slope diagram is shown in Figure 2 As shown (for details, refer to Weiss Andrew-D..Topographic position and landforms analysis[Z].San Diego,CA,USA:Esri User Conference,2001.).
[0037] Table 2 Slope classification standards
[0038]
[0039]
[0040] (4) The obtained slope morphology type and slope position are put into the grid calculator to extract the grid cells that meet both the mid-slope position and the convex slope, thus completing the identification of landslide-prone terrain.
[0041] The technical route of the present invention is as follows Figure 3 The process of slope shape classifier and slope position classifier is as follows. Figure 4 , Figure 5 shown.
[0042] This embodiment takes Sichuan Province as an example:
[0043] (1) In GIS, based on the 3*3 analysis window, the slope, plan curvature, and profile curvature of the Sichuan Province DEM were calculated respectively.
[0044] (2) Taking the CV slope as an example, when extracting the slope, it can be seen from Table 1 that the plan curvature feature of the CV slope is greater than zero, and the profile curvature feature is less than zero. Using the Raster Calculator, enter the formula in the pop-up window: CV = [plan]>0&[profile]<0&[slope_dem]>3 to obtain the CV slope extraction result, as shown in Figure 6 Other slope shapes are extracted using this method.
[0045] Then, according to the ten slope types, use the Raster Calculator and enter the formula in the pop-up window:
[0046] [CC]*1+[CL]*2+[CV]*3+[LC]*4+[LL]*5+[LV]*6+[VC]*7+[VL]*8+[VV]*9+[Flat]*10, select the comprehensive extraction result, realize terrain segmentation, and obtain the slope morphology extraction map, such as Figure 7 .
[0047] (3) Based on DEM, the terrain position index (TPI) method is used to complete the slope position division. The results are as follows Figure 8 shown.
[0048] (4) Extract the grids that meet the requirements of mid-slope position and convex slope shape. The grid calculation conditions are: (slope shape == LV or slope shape == VL or slope shape == VV or slope shape == VC or slope shape == CV) and (slope position == mid-slope position). Verify with historical disaster points. There are 1031 historical disaster points. Due to the non-standard and non-uniform collection of landslide points, there are a certain number of flat slope disaster points, 141 of which are located on the ridge, 34 on the ridge, and 104 in the valley. After removing these data, there are 752 effective disaster points. The number of disaster points that fall into the prone area extraction results is 512, accounting for 68.1%. It proves that the extraction results of this method have a high reference value.
Claims
1. A method for quickly identifying landslide-prone terrain, characterized in that: The following steps are involved: (1) Based on the digital elevation model (DEM) data, GIS surface analysis tools are used to extract the plane curvature, profile curvature, and slope; (2) Based on the relationship between slope morphology type and plane curvature, profile curvature, and slope, complete the classification of slope morphology type; (3) Based on the digital elevation model (DEM) data, the elevation of each cell in the DEM is compared with the average height of the specified neighborhood around the cell using the terrain position index algorithm, using annular analysis; For slope location classification, TPI values are used to classify terrain into different slope location types; this classification is based on the range of TPI values for each grid and the slope of this location; TPI values near the top of the mountain are higher, while TPI values near the bottom of the valley are lower; where TPI values are close to zero are flat or intermediate slopes, the slope value is used to distinguish between these two slope types; (4) The obtained slope morphology type and slope position are put into the grid calculator to extract the grid cells that meet both the mid-slope position and the convex slope, thus completing the identification of landslide-prone terrain.
2. A method for quickly identifying landslide-prone terrain according to claim 1, characterized in that: In step (3), the slope position index TPI is calculated as follows: TDI i =M0-∑ n-1 M n / n Among them, M0 is the grid elevation value of the TPI to be calculated, M n is the elevation value of the analysis area, and n is the total number of grids in the analysis area.
Citation Information
Patent Citations
Tunnel portal slope stability evaluation method based on weighted information amount method
CN118195364A
Landslide disaster risk evaluation method based on information amount model
CN119026913A
Empirical Formula for the Mobilization Criteria between Landslide and Debris Flow and Generation Method thereof
KR1020160137100A
Cited By
Land comprehensive renovation planning method based on three-dimensional modeling and auxiliary decision analysis
CN120235361A
Comprehensive land improvement planning method based on three-dimensional modeling and auxiliary decision analysis
CN120235361B
