A karst landform unit extraction method based on reverse topographic convergence features

Abstract

This invention discloses a method for extracting karst landform units based on inverse topographic confluence features: An inverse topographic DEM is obtained based on the projection of the study area; surface sinks are calculated; sinks are aggregated based on a distance threshold, with those less than the threshold assigned to the same karst landform unit; sinks belonging to the same karst landform unit are used as descent points to extract the inverse topographic watershed boundary, serving as the boundary of the extended karst landform unit in the actual topography; using the positive topographic DEM, a confluence network is extracted based on a confluence threshold; based on the confluence network and slope cost raster, the slope cost distance from the confluence network to other locations within the region is calculated; based on the slope cost distance results, a threshold is set; areas less than the threshold are considered non-mountainous areas; extended karst landform units are used to erase non-mountainous areas, resulting in karst landform units. This invention achieves automated extraction of karst landform units, with extraction results exhibiting strong continuity, good integrity, clear geomorphological boundaries, and high computational efficiency.

Description

Technical Field

[0001] This invention belongs to the field of special landform identification technology, specifically relating to a method for extracting karst landform units based on anti-topographic confluence features. Background Technology

[0002] Landforms are considered one of the most important natural geographical features and a key component of the Earth's surface physical processes. Peak forests and peak clusters are products of karst landform development at a certain stage. They are not only the most typical, widely distributed, and structurally complex type of karst landform, but also an important part of research in natural disaster prevention and control, soil erosion assessment, and quantitative geomorphological analysis. Karst landform units, through their different spatial combinations and distribution patterns, constitute unique peak forest and peak cluster landscapes, control the positive and negative topographic attributes of the Earth's surface, and constrain the scale and form of human activities. Therefore, accurately extracting the boundaries of karst landform units is fundamental to research on landform development and evolution, human production activities, and material and energy transformation.

[0003] Karst landform units are the main components of peak forest and peak cluster landscapes. Different regions exhibit significant differences in surface morphology due to factors such as the duration of karst activity, developmental stage, and material composition. This diversity results in the unique complexity of karst topography, posing considerable challenges to the effective extraction of karst landform units. While methods based on topographic features (such as slope, aspect, and undulation) are more efficient than traditional manual delineation methods, they generally suffer from incomplete, discontinuous, and blurred boundaries in the extracted results. Summary of the Invention

[0004] Technical problem solved: To address the above-mentioned technical problems, this invention provides a method for extracting karst landform units based on anti-topographic confluence characteristics. Under the premise of conforming to the real topographic features, it ensures the accuracy, completeness and continuity of the extraction results, which has important guiding value for karst research in practical work.

[0005] Technical solution: A method for extracting karst landform units based on anti-topographic confluence features, comprising the following steps:

[0006] S1. Obtain the inverse terrain DEM based on the DEM projection of the study area, calculate the surface sinks, aggregate the sinks based on the distance threshold, and assign the surface sinks smaller than the distance threshold to the same karst landform unit.

[0007] S2. Extract the anti-topographic watershed boundary using the surface sink point belonging to the same karst landform unit as the pouring point, and use it as the boundary of the extended karst landform unit in the actual topography.

[0008] S3. Using a positive terrain DEM, extract the confluence network based on the confluence threshold, and calculate the slope cost distance from the confluence network to other locations within the region based on the confluence network and the slope cost raster.

[0009] S4. Based on the slope cost distance results, set a slope cost distance threshold. Areas smaller than the slope cost distance threshold are non-mountainous areas. Use extended karst landform units to erase non-mountainous areas and obtain karst landform units.

[0010] Preferably, when projecting the DEM of the study area in S1, bilinear interpolation is used for resampling of the projection grid.

[0011] Preferably, the method for obtaining the reverse terrain DEM in S1 is as follows: subtract the DEM of the study area from a value greater than the maximum elevation value of the study area.

[0012] Preferably, the method for calculating the surface sink in S1 is as follows: the D8 algorithm is used to calculate the anti-topographic flow direction, and the anti-topographic surface sink in the study area is calculated based on the anti-topographic flow direction.

[0013] Preferably, when aggregating the sink in S1, the selected distance threshold is 60m.

[0014] Preferably, the method for extracting the confluence network in S3 is as follows: the positive topographic flow direction is calculated using the priority flooding algorithm, the cumulative amount of positive topographic confluence is calculated based on the positive topographic flow direction, and the area with a flow threshold greater than the confluence is extracted as the confluence network.

[0015] Preferably, the merging threshold in S3 is 100.

[0016] Preferably, the slope cost grid in S3 is calculated as follows: the slope of the positive terrain surface is calculated using the second-order difference method, and then a minimum value is added to obtain the slope cost grid.

[0017] Preferably, the extraction method for non-mountainous areas in S4 is as follows: Based on the slope cost distance threshold, the slope cost distance is classified into binary rasters, with those less than the slope cost distance threshold set to 1, and those greater than or equal to the slope cost distance threshold set to NODATA; the above binary rasters are converted into vector surfaces, and vector surface features with GRIDCODE values ​​not equal to 1 in the binary rasters are deleted to obtain the non-mountainous areas.

[0018] Preferably, the slope cost distance threshold in S4 is 500.

[0019] Beneficial effects: This invention uses digital terrain analysis methods to achieve automated extraction of karst landform units. The extraction results are characterized by strong continuity, good integrity, clear geomorphological boundaries, and high computational efficiency. It improves the extraction system of geomorphic feature elements in karst areas and has guiding and practical significance for karst research in actual work. Attached Figure Description

[0020] Figure 1 This is a flowchart of a karst landform unit extraction method based on anti-topographic confluence features;

[0021] Figure 2 This is a schematic diagram of the DEM data of the experimental sample area;

[0022] Figure 3 This is a schematic diagram of the inverse terrain DEM data of the experimental sample area;

[0023] Figure 4 This is a map showing the results of surface sink extraction from the reverse topography of the experimental sample area;

[0024] Figure 5 This is a map showing the results of extracting extended karst landform units from the experimental sample area;

[0025] Figure 6 This is a map showing the cumulative surface runoff in the experimental sample area;

[0026] Figure 7 This is a map showing the extraction results of the positive topographic confluence network in the experimental sample area;

[0027] Figure 8 This is a map showing the extracted topographic features of the experimental sample area, specifically the river valley plain.

[0028] Figure 9 This is a map showing the results of extracting karst landform units from the experimental sample area. Implementation

[0029] The present invention will be further explained below with reference to the accompanying drawings and specific embodiments. Example

[0030] This experiment selected a region in Anlong County, Guizhou Province as the experimental area (e.g., Figure 2 The experimental area features well-developed peak forests and peak clusters, representing a typical karst landform landscape, and can serve as a typical sample area for extracting karst landform units.

[0031] This invention provides a method for extracting karst landform units based on anti-topographic confluence features, such as... Figure 1 As shown, it includes the following steps:

[0032] S1. Obtain the inverse terrain DEM based on the DEM projection of the study area, calculate the surface sinks, and aggregate the sinks based on a certain distance threshold. Surface sinks smaller than the distance threshold are assigned to the same karst landform unit.

[0033] S1-1. The steps for projecting the DEM are as follows:

[0034] Using the "Projected Raster" tool in ArcGIS Pro, the output coordinate system was set to WGS_1984_UTM_Zone_48N in the projected coordinate system, and the "Resampling Technique" was selected as "Bilinear Interpolation". The DEM projection result is as follows. Figure 2 As shown.

[0035] The steps for creating an inverse terrain DEM (S1-2) are as follows:

[0036] Using the "Raster Calculator" tool, select an elevation value greater than the maximum value of the DEM in the study area (2000 in this experiment). Subtract the projected DEM from this value to obtain the projected inverse terrain DEM of the study area, as shown below. Figure 3 As shown.

[0037] S1-3. The steps for calculating the sink point are as follows:

[0038] S1-3-1. Using the "Flow Direction" tool, input the inverse terrain DEM obtained in S1-2, and use the D8 algorithm to calculate the inverse terrain flow direction;

[0039] S1-3-2. Using the "Sink" tool, input the anti-topographic flow direction obtained in S1-3-1 to calculate the anti-topographic sink point in the study area, such as... Figure 4 As shown.

[0040] S1-4. The aggregation steps for sink points are as follows:

[0041] S1-4-1. Using the "Buffer Wizard" tool, input the sink point obtained in S1-3-2, set the buffer distance to 30 meters (half of the threshold distance), set the buffer output type to "Merge barriers between buffers", and generate the buffer;

[0042] S1-4-2. Using the “Identify” tool, input the sink obtained in S1-3-2; input the buffer generated in S1-4-1 as the identifier element to identify the buffer vector surface information of the surface sink.

[0043] S2. Using the surface sink points belonging to the same karst landform unit as the descent points, extract the anti-topographic watershed boundary as the boundary of the extended karst landform unit in the actual topography, such as... Figure 5 As shown.

[0044] S2-1. The steps for extracting extended karst landform units are as follows:

[0045] Using the "Catchment Area" tool, input the aggregated sink point in S1-4-2 as the watershed discharge point to calculate the anti-topographic watershed boundary, which is the extended karst landform unit boundary.

[0046] S3. Use the positive terrain DEM to extract the confluence network based on a certain confluence threshold. Based on the confluence network and the slope cost raster, calculate the slope cost distance from the confluence network to other locations in the region.

[0047] S3-1. The steps for extracting a merging network are as follows:

[0048] S3-1-1. Calculate the flow direction of the positive topographic DEM using the priority flooding algorithm;

[0049] S3-1-2. Using the "Flow" tool, input the flow direction obtained in S3-1-1 to calculate the cumulative amount of positive topographic runoff, such as... Figure 6 As shown;

[0050] S3-1-3. Use the "Raster Calculator" tool to extract the positive terrain confluence network. Set the confluence threshold to 100. Areas with values ​​greater than the confluence threshold are designated as confluence networks. Figure 7 As shown.

[0051] S3-2. The steps for calculating the slope cost distance are as follows:

[0052] S3-2-1. Using the "Slope" tool, input the DEM of the normal topography of the study area and calculate the surface slope of the normal topography using the second-order difference method;

[0053] S3-2-2. Using the "Raster Calculator" tool, input the surface slope obtained in S3-2-1, and add 0.0001 as the slope cost raster.

[0054] S3-2-3. Using the "Cost Distance" tool, input the slope cost grid obtained in S3-2-2 as the cost grid; input the confluence network obtained in S3-1-3 as the source, and calculate the slope cost distance.

[0055] S4. Based on the slope cost distance results, set a slope cost distance threshold. Areas smaller than the slope cost distance threshold are non-mountainous areas such as river valleys and plains. Use extended karst landform units to erase non-mountainous areas and obtain karst landform units.

[0056] S4-1. The steps for extracting non-mountainous areas such as river valleys and plains are as follows:

[0057] S4-1-1. Use the "Raster Calculator" tool, input the slope cost distance obtained in S3-2-3 to extract non-mountainous areas such as river valleys and plains, and reclassify them into binary rasters based on the slope cost distance threshold. Set the value less than the threshold to 1, and set the value greater than or equal to the threshold to NODATA.

[0058] S4-1-2. Use the "Raster to Surface" tool to input the binary raster obtained in S4-1-1 and convert it into a vector surface;

[0059] S4-1-3. Delete vector surface features with GRIDCODE values ​​other than 1 in the binarized raster of S4-1-2 to obtain non-mountainous areas such as river valleys and plains, as shown below. Figure 8 As shown.

[0060] The steps for extracting S4-2 karst landform units are as follows:

[0061] S4-2-1. Using the "Erase" tool, input the extended karst landform unit obtained in S2-1; input the non-mountainous areas such as river valleys and plains obtained in S4-1-3 as erasure elements to obtain karst landform units;

[0062] S4-2-2. Using the "Multi-part to Single-part" tool, input the karst landform units obtained in S4-2-1 to obtain mutually independent karst landform unit vector surface features;

[0063] S4-2-3. Using the "Eliminate" tool, input the karst landform unit vector surface obtained in S4-2-2, and eliminate vector surfaces with an area less than 12000 square meters to obtain the final karst landform unit, such as... Figure 9 As shown.

[0064] This invention uses digital terrain analysis methods to achieve automated extraction of karst landform units. The extraction results are characterized by strong continuity, good integrity, clear geomorphological boundaries, and high computational efficiency. It improves the extraction system of geomorphic feature elements in karst areas and has guiding and practical significance for karst research in actual work.

Claims

1. A method for extracting karst landform units based on anti-topographic confluence features, characterized in that, Includes the following steps: S1. Obtain the inverse terrain DEM based on the DEM projection of the study area, calculate the surface sinks, aggregate the sinks based on the distance threshold, and assign the surface sinks smaller than the distance threshold to the same karst landform unit. S2. Extract the anti-topographic watershed boundary using the surface sink point belonging to the same karst landform unit as the pouring point, and use it as the boundary of the extended karst landform unit in the actual topography. S3. Using a positive terrain DEM, extract the confluence network based on the confluence threshold, and calculate the slope cost distance from the confluence network to other locations within the region based on the confluence network and the slope cost raster; S4. Based on the slope cost distance results, set a slope cost distance threshold. Areas smaller than the slope cost distance threshold are non-mountainous areas. Use extended karst landform units to erase non-mountainous areas and obtain karst landform units.

2. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, When projecting the DEM of the study area in S1, bilinear interpolation is used for resampling of the projection grid.

3. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The method for obtaining the inverse terrain DEM in S1 is as follows: subtract the DEM of the study area from a value greater than the maximum elevation value of the study area.

4. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The calculation method for the surface sink in S1 is as follows: the D8 algorithm is used to calculate the anti-topographic flow direction, and the anti-topographic surface sink in the study area is calculated based on the anti-topographic flow direction.

5. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, When aggregating the sink in S1, the selected distance threshold is 60m.

6. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The extraction method of the confluence network in S3 is as follows: the positive topographic flow direction is calculated using the priority flooding algorithm, the cumulative amount of positive topographic confluence is calculated based on the positive topographic flow direction, and the area with a flow threshold greater than the confluence is extracted as the confluence network.

7. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The confluence threshold in S3 is 100.

8. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The slope cost grid in S3 is calculated as follows: the slope of the positive terrain surface is calculated using the second-order difference method, and then a minimum value is added to obtain the slope cost grid.

9. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The extraction method for non-mountainous areas in S4 is as follows: Based on the slope cost distance threshold, the slope cost distance is classified into binary rasters. Those less than the slope cost distance threshold are set to 1, and those greater than or equal to the slope cost distance threshold are set to NODATA. The above binary rasters are converted into vector surfaces, and vector surface features with GRIDCODE values ​​not equal to 1 in the binary rasters are deleted to obtain the non-mountainous areas.

10. The method for extracting karst landform units based on anti-topographic confluence features according to claim 1, characterized in that, The slope cost distance threshold in S4 is 500.

Images